Donate
Independent, objective, nonpartisan research

R 107TGR

Authors

R 107TGR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3742 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(12) "R_107TGR.pdf" ["wpmf_size"]=> string(7) "2284892" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(472904) "Public Policy Institute of California Fiscal Realities: Budget Tradeoffs in California Government ••• Tracy M. Gordon Jaime Calleja Alderete Patrick J. Murphy Jon Sonstelie Ping Zhang Fiscal Realities: Budget Tradeoffs in California Government ••• Tracy M. Gordon Jaime Calleja Alderete Patrick J. Murphy Jon Sonstelie Ping Zhang 2007 Library of Congress Cataloging-in-Publication Data Fiscal realities: budget tradeoffs in California government / Tracy M. Gordon ... [et al.]. p. cm. Includes bibliographical references. ISBN 978-58213-118-4 1. Budget—California. 2. Fiscal policy—California. 3. Intergovernmental fiscal relations—California. I. Gordon, Tracy M., 1970– HJ325.F57 2007 336.794—dc22 2006102045 Copyright © 2007 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword Issues of state and local governance and public finance took center stage in California’s most recent election. Voters approved five bond issues worth $43 billion. Still, a recent report on California’s fiscal outlook by the Legislative Analyst’s Office warns of “operating shortfalls in excess of $5 billion in both FY 2007–08 and 2008–09, which will require significant actions to eliminate.” California’s growing population is putting considerable strain on the state’s health, education, and social services programs as well as in many other areas. With the persistent shortfalls in operating budgets and the uncertainty about who and what will receive benefits from the bond issues, planning for and financing future programs and projects is increasingly difficult. “What kind of California do you want?” is a difficult question to answer when residents are not sure that there is enough money for even the most modest visions of the future. Ten years ago, PPIC published its first study on California public finance. That study evaluated the accuracy and timeliness of the revenue data reported by more than 7,000 local government entities. Since then, the institute has published nearly two dozen reports examining a wide range of governance and fiscal policy issues, including the effects of Proposition 13, the relationship between local sales taxes and land-use decisions, the changing balance of fiscal power between state and local governments, and the challenges that future growth and changing demography present for future planning. In this volume, five research fellows analyze and compare California’s revenue and spending patterns to those of several other large states and to those of the nation as a whole. Their objective is to give Californians a better sense of how they are doing compared to other states facing similar public spending challenges. California voters have a strong say in how much they are willing to pay for services and the kind and degree of services they expect to receive in return. However, the authors find a considerable gap between the iii services state residents would like their government to provide and the taxes they are willing to pay to receive those services. The authors find that California collects more taxes, fees, and miscellaneous revenues per capita than the average state in the rest of the nation. They also find that California differs markedly in its mix of tax revenues: Whereas the average state in the rest of the nation relies on property taxes for 32 percent of its tax revenues and on income taxes for 25 percent, these proportions are reversed in California as a result of its property tax restrictions and highly progressive income tax. The authors also find that some of Californians’ strongest preferences for policy change are very expensive. And although many observers believe that Proposition 13 has severely limited California’s revenues and expenditures, the authors find that the state’s total spending on public services per capita is higher than the average in the rest of the nation. Like all employers in California, state and local governments must pay higher wages to attract employees who can afford to live in the state. Yet, comparative analyses show that labor market conditions are responsible for only a portion of the differences in government payroll expenditures between California and other states. Nonpayroll expenditures also figure into the equation. So higher spending levels in California do not necessarily translate into more services per capita or higher staffing ratios in K–12 education or in police and fire protection. For example, California spends 50 percent more per capita on public safety than the average in all other states, but it employs proportionately fewer police and fire protection personnel. This report provides a unique explanation of these and other factors involved in California finance—benchmarking California against other states. The authors’ goal was to provide a clearer picture of where California spends its public revenues and how much it takes in. They have met this goal most effectively and have set the stage for a more informed discussion of the state’s budget policies and priorities. David W. Lyon President and CEO Public Policy Institute of California iv Summary California’s recent fiscal tumult has prompted calls for a frank public discussion of state and local budget priorities. In short, what do Californians want from their governments and what are they willing to pay for? Despite exhortations from many quarters, this conversation has yet to occur. This omission may have contributed to delays in the passage of state and local budgets in recent years and to an increasing detachment of citizens from their governments. We argue that deciding what California should raise in revenues and spend on services requires first understanding what it currently does raise and spend and why. The usual framing of state and local budget choices can be an obstacle to this understanding. For example, focusing on the state budget alone, and the General Fund within this budget, obscures the activities of other layers of government and their contributions to overall revenues, expenditures, and service delivery. More important, limiting considerations of public spending to one category at a time (such as K–12 education, health and human services, or public safety) prevents a recognition of the explicit or implicit tradeoffs California has made across spending categories or between revenues and public spending. This report takes a different approach. It considers state and local government finances together using a common framework. This framework is comparative, evaluating California’s fiscal choices against those of other states and of California in earlier periods. The purpose of these comparisons is not to make the case that California should be more like one state or another. Rather, it is to provide a benchmark of comparison. These comparisons also provide insights into the drivers of revenue and public spending. Although this framework ignores certain institutions of California budgeting—for example, the two-thirds voting requirement for enacting a budget and constitutional spending mandates v in K–12 education, transportation, and other areas—these rules are also ultimately policy choices. Throughout the report, we compare California’s fiscal choices to those of the United States excluding California and to those of other large states including New York, Illinois, Florida, and Texas. Although no set of states provides a perfect comparison group, this group shares certain features with California, such as a large and diverse population, high median incomes, and a substantial proportion of the population living in poverty. We rely on the most timely, complete, and accurate source of information on state and local budgets in all states over time. Although the most recent data available for all states at the time of our study were for fiscal year 2001–2002, we feel that this base year provides a reliable indicator of where California stands in relative terms today. Our principal findings are presented below. Revenues On a per capita basis, California’s total revenues were $7,166 in 2002, 16 percent higher than the average in the rest of the nation. California also collected more taxes, fees, and miscellaneous revenues than the average state. In 2002, California state and local governments raised $5,099 per state resident, compared to an average of $4,530 in the United States excluding California, $6,232 in New York, $4,480 in Illinois, $4,299 in Florida, and $3,969 in Texas.1 These differences narrow once we adjust for higher personal incomes in this state. In 2002, California’s own-source revenues represented $155 per $1,000 of personal income, 5 percent higher than the average in the rest of the nation, 7 percent higher than in Florida, 12 percent higher than in Texas, 13 percent higher than in Illinois, and 12 percent lower than in New York. California also relied on roughly the same composition of taxes, charges, miscellaneous revenues, and federal income as other states. Where California differs more dramatically from other states is in its tax mix. The average state in the rest of the nation relies on the property _____________ 1In this report, we refer to fiscal years by their terminal years. For example, fiscal year 2001–2002 is referred to as 2002. vi tax for 32 percent of tax revenues and on income taxes for 25 percent. In California, these proportions are reversed as a result of both limits on the property tax and the state’s highly progressive personal income tax. Spending California’s total per capita spending is higher than that of other states ($8,386 versus $6,939 for the rest of the nation in 2002). Focusing on the core activities of government (excluding public sector utilities and insurance trusts), California’s general expenditures are closer to the average for the rest of the nation but still 15 percent higher.2 Spending differences between California and the rest of the nation are particularly large in public safety ($266 per capita in 2002), social services ($222 per capita), environment and housing ($142), and government administration ($142 per capita) (Figure S.1). What are the sources of these differences? Do California’s higher per capita expenditures reflect the particular challenges California faces, such as a large and diverse population, high labor costs, and stewardship over a rich endowment of natural resources? Or do they reflect policy choices—such as decisions about what level of services to provide and who should receive these services? K–12 Education Contrary to perceptions that California is near the bottom in K–12 education spending, it spends more per capita than the average for all other states, at $1,592 compared to $1,489 in 2002. However, California is also home to more students. As a result, spending per pupil in California is slightly below the average for the rest of the nation, at $7,391 compared to $7,523 per student in 2002. Another difference between California and other states is that California pays higher salaries to school district and state education employees. In 2002, California teachers earned an average of $54,348, _____________ 2Throughout this report, we include mass transit in the transportation category, which is part of the U.S. Census Bureau’s general government sector. In contrast, the bureau considers transit part of the utility sector along with water, gas, and electricity supply. Our concept of general expenditures therefore differs slightly from theirs. vii Expenditures (2002 $) 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 California U.S.–California aHidKHagST–enrhmoi1PadaecElGin2rtiunnoahslhsbveetvlioirprddsseucaroouueetsriirrnncstcvvonaaaaiimmtfttngiiicceeeoooetennnsytstnn Figure S.1—General Expenditures per Capita, 2002 the highest salary for any state in the country according to the National Center for Education Statistics (2003). These salaries partly reflect the premium that all California employers must pay to attract collegeeducated workers from other states. However, about half of the compensation difference between California and other states is not explained by labor market conditions. Higher compensation for public school employees and more K–12 students per capita imply lower staffing ratios per student in California than in other states, all else being equal, and this is the case in California. In 2002, there were 49 teachers per 1,000 students in California, compared to 65 per 1,000 in the rest of the nation. Lower staffing ratios were also evident for other types of school employees, including instructional aides, librarians, principals, and administrative assistants. It is also true that California’s student achievement scores lag those of other states. These differences persist even after taking into account differences in English language proficiency, race or ethnicity, and family characteristics, such as income and parental education (Carroll et al., viii 2005). Although the link between resources and outcomes is not straightforward (see, for example, Hanushek, 1997, and Krueger, 2003), California’s recent focus on accountability standards raises the question of whether schools have adequate resources to meet these goals. Higher Education California spends about 8 percent more per capita than the average for all other states on its public colleges and universities, including the University of California, California State University, and California Community College systems ($634 compared to $586). Moreover, California has a tradition of low tuition and fees and a tendency to collect less in revenue from auxiliary services, such as dormitories, cafeterias, and bookstores. As a result, its expenditures net of revenues exceeded the average for other states by an even greater margin (15%) in 2002. As in the case of elementary and secondary education, California has more college and university students per capita than other states. As a result of higher enrollments, net support per full-time-equivalent student (FTES) is lower in California than in other states. In 2002, California spent 7 percent less per student than the average for all other states ($11,329 versus $12,177). Compared to other large states, California was similar to Florida in net support per FTES but lagged Texas by 10 percent, Illinois by 13 percent, and New York by 21 percent. California’s broad participation in higher education stems from its lower tuition and fees as well as its reliance on the California Community College system, which enrolls approximately two million full-time-equivalent students. In fact, California is one of only a handful of states where a majority of higher education students (60%) are enrolled in two-year institutions. Lower costs per FTES in community colleges could bring down average spending for all students. However, California also lags other states in a key measure of productivity in higher education—the number of degrees conferred per student. In 2002, California’s colleges and universities as a group conferred fewer degrees per student than other states did (an average of 0.15 compared to 0.19 degree per FTES). The gap relative to the rest of ix the nation was particularly large (38%) for associate degrees granted relative to enrollment in two-year institutions. Health Services California spends virtually the same amount per capita as the rest of the nation on health services, including medical care for low-income residents, publicly administered hospitals, and public health ($1,173 versus $1,175 per capita). However, California spends less per capita on the largest individual program in this category, Medicaid (known as Medi-Cal in California), and even less per Medicaid recipient relative to the rest of the nation ($3,113 compared to $4,878). California’s lower Medicaid expenditures stem in part from its younger population. California enrolls more nonelderly adults and children in its Medicaid program than other states do and health care costs for these enrollees are generally about one-fifth of those for the elderly and disabled. California also enrolls a higher proportion of MediCal beneficiaries in managed care plans; and it reimburses these and other health care providers at relatively low rates, despite the higher costs of practicing medicine in this state. As in any public program, high enrollment rates can reflect demographics, eligibility rules, or the rate at which individuals apply for and receive benefits (also referred to as the “take-up” rate). A review of these factors suggests that both generous eligibility rules and high take-up rates are at work. For all groups—nonelderly adults, elderly and disabled individuals, and children—California’s income criteria for receiving Medi-Cal benefits were higher than those of other states in 2002. California also extended Medicaid eligibility to groups for whom the federal government did not mandate coverage, such as working parents. Social Services Beyond medical care, state and local governments provide direct cash payments and other subsidized services to low-income individuals. California spent $533 per capita on these services in 2002, roughly 70 percent more than the average for all other states although only a modest share (6%) of total state and local government spending. x The largest of California’s social services programs is Temporary Assistance for Needy Families (TANF), more commonly known as welfare, or CalWORKs in California. CalWORKs is more generous than TANF programs in other states along several dimensions, including higher maximum allowable benefits per family, milder sanctions for noncompliance, and longer time limits before termination of benefits (MaCurdy, Mancuso, and O’Brien-Strain, 2002). Nevertheless, actual expenditures per individual recipient were lower in California than in other states in 2002 ($4,718 compared to $5,105). This discrepancy may be due to shorter average welfare spells, larger family sizes, or a different allocation of cash assistance versus in-kind benefits in California than in other states. Apart from CalWORKs, state and local governments in California support a variety of community-based services, such as child care, child welfare (including foster care and adoption assistance), and programs geared toward low-income elderly and disabled adults. These services are often characterized by low caseloads and high costs per recipient. For example, California spent roughly $15,000 and $29,000 per case for foster care and adoptions, respectively, in 2002. Although exceeding the average for the rest of the nation, these expenditures were not out of line with those of New York or Illinois. In contrast, California’s expenditures on other community services apart from child care and child welfare were $6,690 per case in 2002, much higher than the average for the rest of the nation or for any of our comparison states. Finally, California spends much more than the average state on supplements to the federal Supplemental Security Income (SSI) program, although the per capita amount ($79) is low in absolute terms. California also has higher-than-average caseloads in this program. These caseloads do not stem from a greater concentration of low-income elderly and disabled individuals in the state but rather from a higher rate of participation within this target group. Transportation In 2002, California’s per capita spending on transportation was roughly the same as the average for all other states. However, in every other year since 1977, California’s spending lagged that of the rest of the xi nation. California is also distinctive in the way it allocates transportation expenditures, with less money going toward highway construction ($154 versus $240 in 2002) and more toward mass transit ($181 compared to $117 in 2002). Mass transit use is higher in California than in many other states, comparable to levels in Illinois. However, automobile use is also high, rivaling that of Texas or Florida. This intensive highway use, coupled with lower capacity, contributes to road congestion and delays. Public Safety California’s third-largest area of government spending is public safety. Although California spends 50 percent more per capita on this function than the average for all other states, it employs fewer police and fire protection personnel. California lags even further behind other large states in police protection (with 72 employees per 1,000 crimes in 2002, compared to 89 and 165 per 1,000 in Illinois and New York, respectively) and fire protection (with 0.9 staff per 1,000 residents in 2002, compared to 1.4 per 1,000 in Florida). As in K–12 education, low staffing ratios in public safety go hand in hand with high operating costs, including payroll. Payroll cost differences are only partly explained by the salary premium that all employers in California must pay. Differences in nonpayroll expenses were even wider, perhaps due to more supplemental services provided in California or to different practices of contracting out for services. Higher payroll costs may reflect differences in quality or other personnel attributes. For example, crime rates have fallen more rapidly in California than in the rest of the nation since the mid-1990s. This trend may be due to the effectiveness of law enforcement in this state as well as to the improving economy, changing demographics, and stricter sentencing policies. On the other hand, falling crime rates and relatively low numbers of inmates per capita may also signal an opportunity to reexamine resource allocation in this area. Housing and Community Development In 2002, California spent roughly 50 percent more on housing and community development than the average for all other states. The majority of these expenditures were pass-throughs of federal funds for xii programs administered by local public housing authorities on behalf of the U.S. Department of Housing and Urban Development. Despite these higher expenditures per capita, fewer households report receiving direct housing aid in California. This result may be due to higher costs per recipient or a greater emphasis on redevelopment in California than in other states. Waiting lists for housing assistance are also longer in California than in other states. Environment California’s per capita expenditures on natural resources were 80 percent higher than the average for all other states in 2002. However, they were less than those of other western states or Florida, which face similar water management issues. Water supply is the largest component of natural resources spending in California. Government Administration In 2002, nearly 5 percent of California’s state and local government expenditures went toward administration. At $447 per capita, California’s spending on administration exceeded the average for all other states in that year and even surpassed that of New York, which spent $390 per capita. Part of the explanation for California’s high administrative costs is that these functions are labor-intensive and labor costs are high in this state. However, as in other areas we have considered, only part of the difference between California and other states is attributable to labor market conditions. Nonpayroll expenditures are also part of the explanation. Unfortunately, it is unclear what these expenditures include and whether they are excessive in the absence of appropriate performance measures. Alternative Tradeoffs The final chapter of this report draws these findings together and identifies the implicit or explicit choices California has made over time, either between revenues and public spending or across spending categories. It also compares California’s budget tradeoffs to the expressed xiii preferences of its residents and, where actual choices deviate from public opinion, it poses some alternative scenarios. These alternative tradeoffs are intended not as policy prescriptions but as illustrations of the kinds of changes necessary to achieve desired policy goals. Perhaps the most important theme of this chapter is that some of Californians’ most desired policy goals are costly. For example, increasing K–12 staffing levels to national levels (129 employees per 1,000 students versus 92 per 1,000) would cost five times as much as reducing corrections costs per inmate to the average for all other states and twice as much as raising property taxes per $1,000 of personal income to the average for all other states. These and other tradeoffs provide an analytical foundation for a far-reaching public conversation on what kind of state California should be. xiv Contents Foreword......................................... iii Summary......................................... v Figures .......................................... xix Tables ........................................... xxiii Acknowledgments................................... xxix Acronyms ........................................ xxxi 1. INTRODUCTION .............................. 1 2. ANALYTICAL TOOLS ........................... Overview ...................................... Comparative Approach ............................ Data Sources ................................... Decompositions and Wages ......................... Treatment of Capital.............................. 3. CALIFORNIA’S REVENUE SYSTEM ................ Total Revenues.................................. General Revenues ................................ Own-Source Revenues ............................ Taxes......................................... Conclusions .................................... 4. ELEMENTARY AND SECONDARY EDUCATION ..... Total Expenditures ............................... Current Expenditures ............................. Capital Expenditures.............................. Outcomes ..................................... Conclusions .................................... 5 5 7 8 10 13 19 19 22 24 25 33 35 35 39 50 53 56 5. HIGHER EDUCATION .......................... Total Expenditures ............................... Higher Education Revenue ......................... Net Support for Higher Education .................... Net Support per Student ........................... 57 57 59 61 62 xv Participation and Student Fees ....................... Staffing and Salaries .............................. Capital Expenditures.............................. Outcomes ..................................... Conclusions .................................... 65 69 73 74 77 6. HEALTH SERVICES ............................ 79 Total Expenditures ............................... 80 Medicaid ...................................... 83 Own Public Hospitals ............................. 95 Other Health Services ............................. 97 Outcomes ..................................... 98 Conclusions .................................... 101 7. SOCIAL SERVICES ............................. 103 Total Expenditures ............................... 103 Welfare (TANF/CalWORKs) ....................... 108 Community Services .............................. 114 State Supplementary Program for Supplemental Security Income.................................... 120 Conclusions .................................... 123 8. TRANSPORTATION ............................ 125 Total Expenditures ............................... 125 Mass Transit ................................... 130 Highways...................................... 137 Outcomes ..................................... 146 Conclusions .................................... 148 9. PUBLIC SAFETY ............................... 151 Total Expenditures ............................... 151 Corrections .................................... 153 Police Protection ................................ 160 Fire Protection .................................. 163 Protective Inspection and Regulation .................. 167 Conclusions .................................... 169 10. ENVIRONMENT AND HOUSING ................. 171 Total Expenditures ............................... 172 Housing and Community Development ................ 174 Natural Resources................................ 181 xvi Parks and Recreation.............................. 184 Sewerage, Solid Waste, and Utilities ................... 188 Conclusions .................................... 189 11. GOVERNMENTAL ADMINISTRATION............. 191 Total Expenditures ............................... 191 Finance and Central Staff .......................... 194 Judicial and Legal ................................ 198 Conclusions .................................... 201 12. CALIFORNIA’S FISCAL TRADEOFFS ............... 203 Californians’ Fiscal Preferences....................... 203 Alternative Tradeoffs.............................. 206 Raising Staffing Levels in K–12 Schools............... 208 Expanding Coverage in Health and Human Services ...... 212 Investing in Roads and Infrastructure ................ 217 Conclusions .................................... 219 Appendix: Data Sources .............................. 221 References ........................................ 235 About the Authors .................................. 241 Related PPIC Publications............................. 243 xvii Figures S.1. General Expenditures per Capita, 2002.............. viii 3.1. General Revenues per Capita, 1977–2002 ............ 23 3.2. General Own-Source Revenues per $1,000 of Personal Income, 1977–2002 ........................... 25 3.3. California State and Local Major Taxes per $1,000 of Personal Income, 1977–2002..................... 27 4.1. Total Expenditures on Elementary and Secondary Education per Capita, 1977–2002 ................. 37 4.2. Number of Public School Students per Capita, 1977–2002 ................................. 43 4.3. Current Expenditures on Elementary and Secondary Education per Student, 1977–2002 ................ 44 4.4. Capital Expenditures on Elementary and Secondary Education per Student, 1977–2002 ................ 51 5.1. Higher Education Total Expenditures per Capita, 1977–2002 ................................. 59 5.2. Higher Education Total Revenues per Capita, 1977–2002 ................................. 61 5.3. Higher Education Net Support per Full-Time- Equivalent Student, 1991–2002................... 64 5.4. Percentage of Population Ages 18 to 24 Enrolled in Public Higher Education, 2002 ................... 66 5.5. Higher Education Revenues per Full-Time-Equivalent Student, 1991–2002 ........................... 68 5.6. Higher Education Revenues and Expenditures per Full- Time-Equivalent Student, 1991–2002 .............. 69 5.7. Percentage of Higher Education Enrollment at Two-Year Institutions, 2002 ............................. 72 5.8. Higher Education Capital Expenditures per Capita, 1977–2002 ................................. 74 6.1. Total Health Expenditures per Capita, 1977–2002 ..... 82 xix 6.2. Own Public Hospital Expenditures per Capita, 1977–2002 ................................. 96 6.3. Other Health Services Expenditures per Capita, 1977–2002 ................................. 97 7.1. Social Services Expenditures per Capita, 1977–2002..... 106 7.2. TANF Caseload per Capita, 1998–2002 ............. 114 8.1. Net Expenditures on Transportation per Capita, 1977–2002 ................................. 129 8.2. Capital Expenditures on Mass Transit per Capita, 1977–2002 ................................. 132 8.3. Current Expenditures on Mass Transit per Capita, 1977–2002 ................................. 132 8.4. Capital Expenditures on Highways per Capita, 1977–2002 ................................. 139 8.5. Freeway and Other Principal Arterial Lane Miles per Million Residents, 1980–2002 .................... 139 8.6. Annual Hours of Delay per Capita in the Largest Urban Areas, 2002 ................................. 149 9.1. Public Safety Expenditures per Capita, 1977–2002 ..... 153 9.2. Corrections Expenditures per Capita, 1977–2002 ...... 155 9.3. Inmates Relative to Population, 1977–2002 .......... 156 9.4. Police Protection Expenditures per Capita, 1977–2002... 161 9.5. Reported Offenses per Capita, 1990–2002 ........... 164 9.6. Fire Protection Expenditures per Capita, 1977–2002 .... 165 9.7. Fire Protection Employees per 10,000 Residents, 1993–2002 ................................. 167 9.8. Protective Inspection and Regulation per Capita, 1977–2002 ................................. 169 10.1. Housing and Community Development Expenditures per Capita, 1977–2002 ......................... 175 10.2. Percentage of All Households Receiving Rent Subsidies or Public Housing, 1977–2002 ................... 178 10.3. Natural Resources Net Expenditures per Capita, 1977–2002 ................................. 184 10.4. Parks and Recreation Net Expenditures per Capita, 1977–2002 ................................. 186 11.1. Governmental Administration Expenditures per Capita, 1977–2002 ................................. 194 xx 11.2. Finance and Central Staff Expenditures per Capita, 1977–2002 ................................. 195 11.3. Total Judicial and Legal Expenditures per Capita, 1977–2002 ................................. 198 xxi Tables 2.1. 2.2. 2.3. 2.4. 2.5. 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 3.7. 4.1. 4.2. 4.3. 4.4. 4.5. 4.6. 4.7. 4.8. 4.9. 4.10. 4.11. State and Local Government Expenditures per Capita, 2002 ...................................... Geographic and Economic Features of California and Comparison States ............................ Job Zones from the O*NET Database .............. Average Wages, by State and Job Zone, 2002.......... Wage Indexes Relative to the United States, by Job Zone, 2002 ................................. State and Local Total Revenues per Capita, 2002 ....... State and Local General Revenues per Capita, 2002 ..... State and Local Own-Source Revenues, 2002.......... State and Local Taxes per $1,000 of Personal Income, 2002 ...................................... State and Local Retail Sales Tax Rates, 2004 .......... Selective Sales Tax Rates, 2005.................... Local Revenues per $1,000 of Personal Income, by Source, 2002 ................................ Education Expenditures per Capita, 2002 ............ Total Expenditures on Elementary and Secondary Education per Capita, 2002 ...................... Components of Current Expenditures per Capita, 2002 ...................................... Components of Public School Enrollment, 2000 ....... Components of Current Expenditures per Student, 2002 ...................................... Components of Compensation per Student, 2002 ...... Public School Staff per 1,000 Students, 2002.......... Components of Average Compensation, 2002 ......... Average Public School Salary Adjusted for Location, 2002 ...................................... Capital Outlay per Additional Student, 1992–2002 ..... NAEP Scores for All Students, 2003 ................ 6 8 14 15 15 21 23 24 26 29 30 32 36 38 40 42 46 46 47 48 49 52 54 xxiii 4.12. NAEP Scores for Students Proficient in English, 2003 ... 5.1. Expenditures on Higher Education per Capita, 2002 .... 5.2. Higher Education Revenues per Capita, 2002 ......... 5.3. Higher Education Net Support per Capita, 2002 ....... 5.4. Net Support per Full-Time-Equivalent Student, 2002 ... 5.5. Residence and Migration of First-Year Undergraduate Students, 2002 ............................... 5.6. Total Payroll, Costs per Employee, and Number of Employees per Full-Time-Equivalent Student, 2002..... 5.7. Wages for Full- and Part-Time Instructional and Noninstructional Employees, 2002................. 5.8. Full-Time Instructional Employee Salary Comparisons, 2002 ...................................... 5.9. Increases in Enrollment and Real Capital Outlays, 1992–2002 ................................. 5.10. Degrees Conferred at Public Colleges and Universities per 1,000 State Residents, Academic Year 2002–2003 ... 5.11. Degrees Conferred per Full-Time-Equivalent Student and Net Support per Degree, 2002................. 5.12. Estimated Degrees Conferred per Full-Time-Equivalent Student at Two- and Four-Year Institutions, 2002 ...... 6.1. Health Expenditures per Capita, 2002 .............. 6.2. Components of Medicaid Expenditures per Capita, 2002 ...................................... 6.3. Composition of State Population, by Eligibility Groups, 2002 ..................................... 6.4. Components of Medicaid Recipients per Nonelderly Adult, 2002 ................................. 6.5. Components of Medicaid Recipients per Elderly or Disabled Adult, 2002 .......................... 6.6. Components of Medicaid Recipients per Child, 2002.... 6.7. Components of Medicaid and SCHIP Enrollment, 2001 ...................................... 6.8. Medicaid Expenditures per Recipient, by Eligibility Group, 2002 ................................ 6.9. Components of Medicaid Expenditures per Recipient, 2002 ...................................... 6.10. Medicaid Fee Index, 2003 ....................... 55 58 60 62 63 66 70 71 71 74 75 76 77 81 85 86 87 89 90 91 92 93 94 xxiv 6.11. Medicaid-to-Medicare Fee Indexes, 2003 ............ 94 6.12. Components of Own Public Hospital Expenditures per Capita, 2002 ................................ 96 6.13. Average Preventable Hospitalizations per 1,000 Persons, 1999 ...................................... 99 6.14. Birth Outcome Indicators per 100 Births, 1999 ........ 100 6.15. Access to Care Measures: Percentage of Population Below 200 Percent FPL, 1999–2000................ 100 7.1. Social Services Expenditures per Capita, 2002 ......... 106 7.2. Welfare Expenditures per Capita, by Source of Funding, 2002 ...................................... 109 7.3. Components of Welfare Expenditures per Capita, 2002 .. 110 7.4. Components of Welfare Caseload, 2002 ............. 111 7.5. Welfare Child Care Expenditures per Capita, by Source of Funding, 2002 ............................. 113 7.6. Total Community Services Expenditures per Capita, by Source of Funding, 2002 ........................ 115 7.7. Components of Foster Care Expenditures per Capita, 2002 ...................................... 116 7.8. Components of Adoptions Expenditures per Capita, 2002 ...................................... 117 7.9. Expenditures on Other Community Services per Capita, by Source of Funding, 2002...................... 118 7.10. Components of Expenditures on Other Community Services per Capita, 2002........................ 119 7.11. Other Community Services Expenditures per Beneficiary from State-Only Sources, 1998–2002 ............... 120 7.12. SSI Expenditures per Capita: Use and State Supplementation, 2002 ......................... 121 7.13. Components of SSI Expenditures per Capita, 2002 ..... 122 7.14. Components of SSI Caseload, 2002 ................ 122 8.1. Expenditure on Transportation per Capita, 2002 ....... 127 8.2. Net Expenditures on Transportation per Capita, 2002 ... 128 8.3. Net Expenditures on Mass Transit per Capita, 2002..... 130 8.4. Expenditures on Mass Transit and Urban Density, 2002 ...................................... 131 8.5. Expenditures on Public Mass Transit per Capita, 2002 ...................................... 131 xxv 8.6. Passenger Miles per Trip, by Transit Mode, 2002 ...... 134 8.7. Operating Cost per Passenger Mile, by Mode, 2002 ..... 134 8.8. Components of Operating Cost per Passenger Mile, by Mode, 2002 ................................. 135 8.9. Components of Employee Compensation per Passenger Mile, 2002.................................. 136 8.10. Mass Transit Salary Adjusted for Location, 2002 ....... 137 8.11. Expenditures on Highways per Capita, 2002 .......... 138 8.12. Current and Capital Expenditures on Highways per Capita, 2002 ................................ 138 8.13. Capital Expenditures on Highways per Capita, 1957–1997 ................................. 140 8.14. Freeways and Other Principal Arterial Growth in Lane Miles and Population, 1980–2002 ................. 141 8.15. Traffic on Freeways and Other Principal Arterials, 2002 .. 142 8.16. Traffic on Freeways and Other Principal Arterials in Urban and Rural Areas, 2002..................... 143 8.17. Current Highway Expenditures per Capita, 2002....... 143 8.18. Components of Current Highway Expenditures per Capita on Local Roads and Highways, 2002 .......... 144 8.19. Components of Current Highway Expenditures per Capita on State Roads and Highways, 2002........... 145 8.20. Annual Highway Payroll Adjusted for Location, 2002 ... 145 8.21. Road Congestion and Delay in 85 Urban Areas, 2002 ... 148 9.1. Public Safety Expenditures per Capita, by Type of Service, 2002 ................................ 152 9.2. Corrections Expenditures per Capita, by Type, 2002 .... 154 9.3. Components of Corrections Expenditures per Capita, 2002 ...................................... 156 9.4. Components of Corrections Expenditures per Inmate, 2002 ...................................... 158 9.5. Average Yearly Payroll in the Correctional System, 2002.. 158 9.6. Average Occupancy of Correctional Facilities, 2002 ..... 160 9.7. Real Capital Outlay per Additional Inmate in State Prisons, 1992–2002 ........................... 160 9.8. Police Protection Expenditures per Capita, by Type of Government, 2002 ............................ 161 xxvi 9.9. Components of Police Protection Expenditures per Capita, 2002 ................................ 163 9.10. Average Yearly Payroll for Police Protection, 2002 ...... 163 9.11. Fire Protection Expenditures per Capita, by Type of Government, 2002 ............................ 164 9.12. Components of Fire Protection Expenditures per Capita, 2002 ...................................... 166 9.13. Average Yearly Payroll for Fire Protection, 2002........ 166 9.14. Protective Inspection and Regulation Expenditures, by Type of Government, 2002 ...................... 168 10.1. Environment and Housing Expenditures per Capita, 2002 ...................................... 174 10.2. Housing and Community Development Expenditures per Capita, 2002.............................. 175 10.3. Housing and Community Development Federal Intergovernmental Revenue per Capita, 2002 ......... 176 10.4. Housing Assistance Caseloads and Components, 2002 ... 177 10.5. Rent Burdens as a Percentage of Household Income, 2002 ...................................... 179 10.6. Average Months on Waiting List for Housing Program Participants, 2002............................. 179 10.7. Statewide Average Fair Market Rents Weighted by Population in Each Market Area, 2002 .............. 180 10.8. Housing and Community Development Real Capital Outlay, 1992–2002 ........................... 181 10.9. Natural Resources Expenditures per Capita, 2002 ...... 182 10.10. Natural Resources Net Expenditures per Capita, 2002 ... 183 10.11. Parks and Recreation Expenditures per Capita, 2002 .... 185 10.12. Parks and Recreation Net Expenditures per Capita, 2002 ...................................... 185 10.13. Federal, State, and Local Parks and Natural Resource Lands, 2002................................. 187 10.14. Parks and Recreation Expenditures per Acre, 2002 ...... 187 10.15. State and National Park Visits per Capita, 2002........ 188 10.16. Utility Expenditures per Capita, 2002............... 188 10.17. Net Enterprise and Utility Expenditures per Capita, 2002 ...................................... 189 xxvii 11.1. Governmental Administration Expenditures per Capita, 2002 ...................................... 193 11.2. Components of Total Expenditures per Capita on Finance and Central Staff, 2002 ................... 195 11.3. Components of Finance and Central Staff Payroll Expenditures, 2002............................ 196 11.4. Average Payroll for Finance and Central Staff Employees, 2002 ...................................... 197 11.5. Components of Total Judicial and Legal Expenditures per Capita, 2002.............................. 199 11.6. Components of Judicial and Legal Payroll Expenditures per Capita, 2002.............................. 200 11.7. Average Payroll for Judicial and Legal Employees, 2002 .. 201 12.1. California Preferences for Governmental Spending, January 2006 ................................ 204 12.2. Budget Priorities in Other States, 2002 and 2003....... 207 12.3. Alternative Scenario #1: Raising Staffing Levels in K–12 Schools .................................... 209 12.4. Alternative Scenario #2: Expanding Medi-Cal and CalWORKs ................................. 215 12.5. Alternative Scenario #3: Investing in Roads and Other Infrastructure ................................ 219 A.1. Urban Areas in Texas Transportation Institute Study, 2002 ...................................... 234 xxviii Acknowledgments This report reflects the contributions of several individuals and institutions. We are grateful to The James Irvine Foundation for providing a generous planning grant that enabled us to think broadly about the goals of the report and to consult with experts in government, academia, journalism, business, and the nonprofit sector. We are also indebted to these experts for providing us with their counsel. This report would not have been possible without the comprehensive, high-quality data produced by the U.S. Census Bureau. We spent several days on site at the Governments Division and engaged in frequent consultations with Census Bureau staff by telephone and email. We are especially indebted to Henry Wulf, Chief of the Governments Division, as well as David Kellerman, Christopher Pece, Steven Poyta, Jeffrey Little, Jacques C. Kusseling, and Joseph Bruzzese. We are also indebted to those who reviewed this report in its entirety: Donald Boyd of the Rockefeller Institute of Government at SUNY, Albany; Heather Rose of PPIC; Kim Rueben of the Tax Policy Center and the Urban Institute; Peter Schaafsma of the California State Assembly Republican Caucus; and Fred Silva formerly of PPIC and now consultant to the New California Network. In addition, several individuals provided comments on specific subject areas: California Legislative Analyst, Elizabeth Hill, as well as several members of her staff including Brad Williams, David Vasché, Robert Manwaring, Steve Boilard, Daniel Carson, Todd Bland, Dana Curry, Gregory Jolivette, Michael Cohen, Mark Newton, and Brendan McCarthy; Margaret O’Brien-Strain of the SPHERE Institute; Ellen Hanak and Marianne Bitler of PPIC; and Phyllis Myers of State Resource Strategies. Mark Baldassare and Max Neiman of PPIC and Paul Lewis formerly of PPIC and now at Arizona State University offered guidance on this project from the initial stages to its completion. PPIC’s communications group improved the report’s clarity and exposition. Finally, this report xxix would not have been possible without the contributions of several excellent research associates at PPIC: Shelley de Alth, David Haskel, Supreet Kaur, and Kate Jordan. Although we are pleased to thank each of these persons and organizations, all findings, conclusions, errors, and omissions in this report are our own. xxx Acronyms AC Transit ACF AFDC AHRQ AMT BART BEA CalPERS CalSTRS CalWORKs CCC CCDF CDBG CMS CMSA CPEC CPI CPS CSR CSU CVP DMV DSH DWR FFS FMR FPL FTA FTES GAO Alameda–Contra Costa Administration for Children and Families Aid to Families with Dependent Children Agency for Healthcare Research and Quality Alternative Minimum Tax Bay Area Rapid Transit Bureau of Economic Analysis California Public Employees’ Retirement System California State Teachers’ Retirement System California Work Opportunity and Responsibility to Kids California Community Colleges Child Care and Development Fund Community Development Block Grant Centers for Medicare and Medicaid Services Consolidated MSA California Postsecondary Education Commission Consumer Price Index Current Population Survey Class Size Reduction California State University Central Valley Project Department of Motor Vehicles Disproportionate Share Hospital Department of Water Resources Fee for Service Fair Market Rent Federal Poverty Line Federation of Tax Administrators Full-Time-Equivalent Student Government Accountability Office xxxi HCBS HUD IPEDS LACMTA LAMTA LAO LEA LIHEAP LIHTC MIC MOE MSA MSIS MUNI NAEP NCES NEA NEC NOL NTD O*NET OES PAB PHA PMSA RDA RVU SAT9 SCHIP SCORP SSBG SSD SSI SSP SSTP SWP Home- and Community-Based Services Housing and Urban Development Integrated Postsecondary Education Data System Los Angeles County Metropolitan Transportation Agency Los Angeles Metropolitan Transit Authority Legislative Analyst’s Office Local Education Agency Low-Income Home Energy Assistance Program Low-Income Housing Tax Credit Manufacturer’s Investment Credit Maintenance of Effort Metropolitan Statistical Area Medicaid Statistical Information System Municipal Railway National Assessment of Educational Progress National Center for Education Statistics National Education Association Not Elsewhere Classified Net Operating Loss National Transit Database Occupational Information Network Occupation Employment Statistics Private Activity Bond Public Housing Authority Primary Metropolitan Statistical Area Redevelopment Agency Relative Value Unit Stanford Achievement Test, Ninth Edition State Children’s Health Insurance Program State Comprehensive Outdoor Recreation Plan Social Services Block Grant State Summary Datamart Supplemental Security Income State Supplemental Program Streamlined Sales Tax Project State Water Project xxxii TANF TTI UC Temporary Assistance for Families with Dependent Children Texas Transportation Institute University of California xxxiii 1. Introduction California has been on a fiscal rollercoaster in recent decades. The 1990s started with a recession that was mild by national standards but strongly felt in this state. At the time, many observers expressed alarm about a “structural deficit,” or a permanent mismatch between state revenues and expenditures, apart from effects of the business cycle. They also pointed out that local governments were still struggling with uncertain revenues after the passage of Proposition 13. These concerns diminished as government revenues soared in the late 1990s, fueled by rising income from stock options and capital gains. Revenue increases funded popular program expansions in education, health, social services, and tax relief. In hindsight, of course, these gains were short-lived. The collapse of the stock market and economic downturn of 2001 generated massive losses for the state ($12 billion out of a roughly $100 billion budget). Budget shortfalls widened as economic conditions worsened and corrective actions yielded disappointing results. By May 2003, the Department of Finance projected a $38 billion gap between revenues and expenditures. More recently, state coffers have rebounded thanks to stronger than expected personal and corporate income tax receipts. Local governments are also in a more secure fiscal position, having won passage of Proposition 1A, which prohibits future reductions in allocations of state funds. Nevertheless, concerns persist about an underlying imbalance between state revenues and expenditures. For example, the Legislative Analyst’s Office (LAO) has warned that higher energy costs and a steeper-than-expected slowdown in real estate markets could expose the state to more budget risk. The LAO has also noted that the state will face substantial borrowing costs and large unfunded liabilities for state government retiree health plans (Legislative Analyst’s Office, 2006b). Throughout these budgetary swings, there have been numerous calls for reform. A report by the California Research Bureau tabulates 1 recommendations from nearly a dozen state commissions, business groups, civic task forces, and academic studies since 1991 (Wear Simmons, 2002). Although there appears to be a consensus that California’s budget process is “broken,” there is less agreement about what should be done to fix it. An oft-cited reason for inaction on budget reform is that Californians have not confronted a fundamental disconnect in preferences for both lower taxes and higher government spending. By this argument, a “something for nothing” mentality dating to the 1970s tax revolt pervades budget discussions and precludes consideration of revenue increases needed to support publicly favored programs. On the other hand, an equally vociferous group argues that there is no disconnect but instead an inefficient public sector containing enough waste to allow spending reductions without harming the quality of services. Interestingly, a common refrain in both arguments is that California needs to have a frank public conversation about its budget priorities. In short, what kind of a state do Californians want and what are they willing to pay for? Yet, despite these exhortations on both sides of the political spectrum, such a conversation has not occurred. One obstacle to this kind of far-reaching budget conversation is insufficient information. Given the current framework for budget decisions in California, it is not clear how to assess what state and local governments raise, spend, and produce. For example, budget decisions are typically framed within a category of spending (e.g., categorical versus unrestricted aid for schools), by level of government (e.g., cities and counties versus school districts), and by source of funds (e.g., the state General Fund versus special funds). These distinctions are not helpful for understanding broader tradeoffs between spending categories (e.g., K–12 education versus health services) or between revenues and public spending. Moreover, they do not permit a discussion of whether to continue these tradeoffs. This report takes a different approach. It analyzes revenues and spending in every major category (e.g., education, health, social services, and so on) using a common framework that ignores distinctions between levels of government and sources of funds. The framework is 2 comparative, evaluating California’s fiscal tradeoffs against those of other states. The purpose of these comparisons is not to argue that California should be more like one state or another. Rather, it is to provide a basis of comparison. These comparisons further allow us to identify common drivers of spending—such as program caseloads or workload, input costs, and the level or quality of services provided—and to highlight which tradeoffs reflect the particular challenges this state faces and which stem from policy decisions. According to this framework, institutions such as voter-approved limits on revenues and spending are also policy choices. Finally, the report assesses current tradeoffs against public opinion. Where actual tradeoffs deviate from these expressed preferences, it presents alternative scenarios. These scenarios are not policy prescriptions. Indeed, some of them may be impractical or infeasible. Instead, these scenarios are intended to reinforce our findings and to illustrate the magnitudes of past tradeoffs in California finances as well as the difficulty in some cases of undoing these tradeoffs. Taken together, the findings in this report provide an analytical foundation for a public conversation about California’s budget priorities. As in other previous studies, we identify some potentially conflicting goals and tough choices. However, there is no reason to expect that these vexing questions about the goals of government should not be amenable to the same creativity and innovation for which California is known in other sectors. The remainder of this report is organized as follows. The next chapter explains the overarching approach, concepts, and analytical tools used throughout the report. Chapter 3 analyzes California’s revenue system, including taxes, fees, charges, other state and locally generated revenues, and federal contributions. Subsequent chapters examine state and local government spending in each major category: elementary and secondary education, higher education, health, social services, transportation, public safety, environment and housing, and government administration. The report concludes with a review of findings from each chapter and an exploration of alternative budget tradeoffs. 3 2. Analytical Tools This chapter reviews the analytical approach, data, and methods used in this report. It begins with a snapshot of California’s government finances in fiscal year 2001–2002, the most recent year for which our primary data were available.1 The discussion in this chapter is not a summary but a point of departure for the chapters that follow. The remainder of this chapter explains how we will analyze California’s revenues and expenditures in greater detail using the experiences of other states. Overview In 2002, California’s state and local governments collected nearly $251 billion in revenues and administered over $290 billion in spending.2 The gap between these two numbers arises from several features of that particular year. First, as discussed in the previous chapter, a collapse of the stock market and economic downturn in 2001 led to a plunge in state revenues of about $12 billion in 2002. It also generated investment losses of approximately $23 billion for California’s insurance trust funds, such as the public employee retirement plans California Public Employees’ Retirement System (CalPERS) and California State Teachers’ Retirement System (CalSTRS). An added complication in California was the state energy crisis of 2001, during which the state purchased electricity on the wholesale market on behalf of investor-owned utilities. Although later repaid from a bond secured by consumer rate payments, this loan appeared in 2002 as a state expenditure of approximately $7 billion. _____________ 1These data, through 2004, are available on PPIC’s website (www.ppic.org) in a user-friendly format. They are also available from the Census Bureau (ftp://ftp2.census. gov/pub/outgoing/govs/special60/). 2In the remainder of this report, we refer to fiscal years by their terminal year. For example, fiscal year 2001–2002 is referred to as 2002. 5 On a per capita basis, California’s total expenditures were $8,386 and its revenues were $7,166. Other analysts have used personal income rather than population as a benchmark for expenditure comparisons. We use per capita figures here because controlling for income incorporates several factors that we will investigate separately below (e.g., higher labor costs). We use the average for the rest of the nation rather than the United States overall because California is such a large state that it can skew national averages. Rather than total expenditures, we can also focus on the core activities of government by excluding public sector utilities and insurance trusts. By this measure, California’s general expenditures were $6,735 per capita, roughly 15 percent higher than the average for all other states. Per capita spending differences between California and the rest of the nation were particularly large in public safety ($266), social services ($222), environment and housing ($142), and government administration ($142) (Table 2.1). Table 2.1 State and Local Government Total Expenditures per Capita, 2002 K–12 education Health Public safety Higher education Environment and housing Social services Government administration Transportation Utilities Insurance trusts Interest on general debt Other Liquor Total California ($) 1,592 1,173 778 634 590 533 447 430 824 746 248 391 — 8,386 U.S.–California ($) 1,489 1,175 512 586 448 311 305 482 438 568 263 344 17 6,939 Difference $% 102 6 –2 0 266 34 47 7 142 24 222 42 142 32 –52 –12 385 47 179 24 –15 –6 47 12 –17 — NOTES: Throughout this report, numbers in the tables may not sum to totals because of rounding. Also, we often refer to percentage differences in the text although we do not report these percentages in the tables for the sake of brevity. 6 Comparative Approach These figures raise a series of questions. Is California a high-tax and high-spending state? Has it always been so? What is high or low? California’s higher per capita expenditures may reflect policy choices— such as decisions to collect more in revenues and provide more in services—or background conditions—such as higher costs of living and thus higher wages for both public and private sector workers in the state. California’s public finances are also undoubtedly shaped by its population. More school-age children place greater demands on the K–12 education system, and more low-income families and elderly or disabled individuals can strain state and federally funded social programs. Answering these questions requires a basis for comparison. One benchmark is historical experience. Is California raising and spending more or less now than in 1977? However, intervening events— fluctuations in the economy, demographic shifts, and technological improvements—may lead to observed differences between now and the past. Disentangling which events matter for revenues and public spending is complicated in practice. We can better understand which factors affect government finances by broadening our analysis to include the experiences of other states, which are presumably affected by the same general trends. These comparisons are even more meaningful if these states are similar to California. California is a large and complex state, which many claim defies comparison with other states, if not with other countries. Yet, California shares certain features with other states, particularly other large states including Florida, Texas, Illinois, and New York (Table 2.2). Like California, all of these states except Texas are urbanized and densely populated. Similar to New York and Illinois, California is also a relatively high-income state. At the same time, a substantial share of Californians lives in poverty, as in Florida, New York, and Texas. Apart from its size, California’s population is notable for its diversity. Like Texas, California is a “majority minority” state, meaning that more than half of its residents are nonwhite. Both states are also home to large immigrant populations, which tend to be predominantly younger. In the remainder of this report, we will compare California to the rest of the 7 Table 2.2 Geographic and Economic Features of California and Comparison States California U.S. total Florida Illinois New York Texas Population 36,132,147 296,410,404 17,789,864 12,763,371 19,254,630 22,859,968 Median Below NonTotal Area Household Poverty Hispanic Foreign(sq. miles) Income ($) Level (%) White (%) Born (%) 163,696 50,220 13.4 44.5 26.5 3,794,083 43,564 12.7 67.4 11.9 65,755 39,871 13.1 62.8 17.6 57,914 47,977 11.3 66.2 13.3 54,556 46,195 13.5 61.1 20.8 268,581 40,674 16.3 49.8 15.6 nation (the United States excluding California) and to these other large states, often referring to them as our “comparison group.” Data Sources This report relies primarily on government finance and payroll data collected through the U.S. Census Bureau’s Census of Governments and Annual Government Finance series. These data are the most complete, accurate, and timely source of information available on state and local government finances. The Census Bureau canvasses all state and local governments in the nation in years ending in 2 and 7 and all states and a sample of local governments annually.3 Data are available for all states in the same basic format going back to the 1950s. To maximize the consistency of the data over time, we rely mainly on a series starting in 1977. Although every state has its own reporting system, the Census Bureau collects and recasts this information using its own standard format. This format ignores distinctions between funds (including borrowed funds and prior year balances) although it does track intergovernmental revenues and capital expenditures separately.4 The format classifies revenues and expenditures according to functions rather _____________ 3Local government totals in the annual survey are generally subject to a sampling variability of 1 percent or less. 4These expenditures do not include debt service payments, which have their own category. 8 than programs. In what the Census Bureau refers to as the general government sector, these functions are elementary and secondary education, higher education, health, social services, transportation, public safety, environment and housing, and government administration. In addition, the Census Bureau tracks revenues and expenditures of public water, gas, electric, and transit utilities separately. In our discussion we include the first three utilities with the environment and housing function and transit with transportation. The Census Bureau also gathers data on the finances of insurance trusts such as public employee retirement, unemployment compensation, workers’ compensation, and disability insurance systems. We do not analyze these systems here. These data have their limitations. The most recent year for which Census Bureau data were available at the time of this writing for both states and localities was 2002. In some sense, this year represents a highwater mark for government finances because it reflects higher spending during the economic boom of the 1990s. However, as the most recent year available, it was also the best predictor of government finances today. The Census Bureau’s standard reporting format also means that these data may differ from totals in state and local government reports. For example, some observers have expressed concern that Census Bureau data on state and local spending separately count both state assistance to local governments and local expenditures of these funds. Moreover, relationships between functions and programs are not always clear. For example, Medicaid expenditures are split between several categories, including vendor payments, health, hospitals, and cash assistance. To address these concerns, we performed an extensive crosswalk between the Census Bureau data and California state and local sources, including the Governor’s Budget and Annual Financial Reports from the State Controller’s Office. The results of this crosswalk (outlined in de Alth and Haskel, 2007) explain the relationship between government programs in California and Census Bureau functional categories. Through this exercise, we were also able to reproduce Census Bureau figures for expenditures and revenues within 2 and 3 percent, respectively. Together with the reputation of the Census of 9 Governments for producing high-quality data for forecasting and research, this result makes us reasonably confident in the data for all states and previous years. Using the Census Bureau data, we consider state and local government finances together in each major functional category. We do so mainly because we are interested in the overall activities of the public sector in California and not in the state-local fiscal relationship. There is also a more practical reason. States have different traditions of allocating revenues and program responsibility, and ignoring these differences by focusing on the state alone can lead to mistaken inferences. For example, California’s state expenditures on K–12 education are higher than most other states’ expenditures, although state and local expenditures are near the national average. This pattern is an artifact of the centralization of education finance in this state after passage of Proposition 13 in 1978. Finally, the Census of Governments collects data on the government payroll and employment using the same classification scheme used for tracking government finances. Payroll data do not include information on benefits such as paid leave, overtime, and employer contributions to health and disability insurance, savings and retirement plans, workers’ compensation, and unemployment funds. Where possible, we discuss differences in nonwage compensation in each functional area. Decompositions and Wages We analyze public spending in this report using a series of basic accounting relationships. First, for any area of public spending, expenditures per capita depend on caseloads—the number of recipients or users of public goods and services—and the costs per recipient: Expenditures = Caseload  Expenditures Population Population Case (2.1) Second, as discussed above, the workload for any public program depends on a state’s demographic characteristics, such as the size of its population and its age or income distribution, as well as on the number of service users within this target group. We can separate the influence of these two factors using the following relationship: 10 Caseload = Participants  Potentially Eligible Population Potentially Eligible Population (2.2) Here, the second term reflects the demographic features of a state whereas the first term represents the combined effect of policymaker decisions and individual behaviors. For example, policymakers may decide who is eligible for a program, for what services they are eligible, and for how long they may receive benefits. These rules may also be determined at the federal level as a condition for states to receive matching funds. Eligible individuals then decide whether to use or “take up” services. When demand for services outstrips capacity, an administrative process may determine who receives benefits, or rationing may occur based on wait times or congestion. In some cases, we are able to further decompose the first term of Eq. (2.2) into take-up rates and program rules, but often there are insufficient data for this task. Finally, as noted above, expenditures per capita also depend on the costs of labor and other inputs such as energy or land prices. Expenditures per case can be represented as the product of these costs and the average quantity of inputs used: Expenditures = Units  Price Case Case Unit (2.3) In the same way that the above relationships shed light on the determinants of expenditures per capita within a state, they can also help explain spending differences between states. For example, where  represents the difference between California and another state or the average for all other states:  Expenditures   Caseload +  Expenditures Population Population Case (2.4) 11 That is, the difference in expenditures per capita between two states is equal to the difference in cases per capita plus the difference in expenditures per case.5 Labor costs deserve special attention in any analysis of government expenditures because public services tend to be labor-intensive. Indeed, perceptions of quality in the public sector often depend on how much labor is involved (e.g., smaller class sizes in public schools). In California, public sector wages may be particularly high because of the higher costs of living in the state. At the same time, California also offers certain amenities, such as its climate, geography, and culture. Market-basket cost adjustments, such as the U.S. Bureau of Labor Statistics Consumer Price Index (CPI), fail to account for how these amenities might affect prices. Although this point is well understood in real estate, where similar houses with additional rooms, swimming pools, or larger lots command a price premium, it is frequently overlooked in discussions of regional cost differences. To capture both the costs and benefits of living in California, we rely on the concept of a “location premium.” This premium is the difference in wages that employers in California must pay to compete in a national labor market. In principle, it may be either positive (i.e., if employers have to pay more to attract and retain workers) or negative (i.e., if workers have to give up some of their earnings to live in a desirable place). For any occupation, the location premium is based on the average salary difference for a range of occupations requiring comparable education, skills, and experience. For example, the location premium for teachers is based on other occupations that typically require a four-year college degree. We use a range rather than simply comparing average public school teacher salaries across states to avoid conflating labor costs with policy decisions about teacher pay.6 _____________ 5In fact, this is an approximation because of an arbitrary weight that is introduced when differences are taken in discrete terms rather than continuously (e.g., a Taylor expansion). 6This is equivalent to the Comparable Wage Index approach (e.g., Taylor, 2004). 12 In this report, we use labor market zones as defined by the Occupational Information Network database, or O*NET (see the appendix). This database uses surveys of employers and employees to define ranges of skill, knowledge, and experience known as labor market zones. Zone 1 corresponds to occupations requiring no previous workrelated skills, knowledge, or experience, whereas Zone 5 refers to jobs that require extensive preparation, often including an advanced degree. Table 2.3 describes these and intermediate zones more fully. In all labor market zones, California employers must offer higher wages than the U.S. average (Table 2.4). The gap is largest in Zones 3 through 5, where more education and experience are required. In New York, the gaps are even larger, whereas in Texas and Florida, the gaps are negative, suggesting that comparable workers earn less than the national average in these states. Table 2.5 summarizes the percentage difference in average salaries between each state and the U.S. average for each labor market zone. To illustrate what this means for understanding California’s expenditures per capita, consider a government agency hiring a worker from Zone 3. It would have to pay an annual salary of nearly $43,000 for the same worker who would earn $37,000 per year in the rest of the country, a 14 percent difference as shown in the third column of Table 2.5. Suppose we actually observe the California agency paying salaries at 25 percent above the average for a similar occupation in all other states. The remaining 11 percent is unexplained by labor market conditions. Treatment of Capital State and local budgeting practices are usually based on cash accounting, or measuring current revenues against current expenditures. This method captures debt service and rental payments but otherwise ignores the value of long-lived investments such as buildings, roads, or land and mineral rights. A capital budget would take into account these assets as well as their depreciation. Capital budgeting is not widespread, however. In this report, we focus on current revenues and expenditures, adjusting historical figures for inflation using 2002 dollars (CPI-U). We 13 Table 2.3 Job Zones from the O*NET Database Zone 1 Description No previous work-related skill, knowledge, or experience. May require high school diploma or GED or formal training course to obtain license. Training on the job from experienced co-worker for a few days to months. Skills involve following instructions and helping others. Examples include bus drivers, forest and conservation workers, general office clerks, home health aides, and waiters/waitresses. 2 Some previous work-related skill, knowledge, or experience. Usually requires a high school diploma and may require vocational training or job-related coursework. Training from a few months to one year with experienced employees. Skills involve using knowledge and skills to help others. Examples include drywall installers, fire inspectors, flight attendants, pharmacy technicians, salespersons (retail), and tellers. 3 Previous work-related skill, knowledge, or experience. Usually requires training in vocational schools, related on-the-job experience, or an associate’s degree. One or two years of training involving on-the-job experience and informal training with experienced workers. Skills involve using communication and organizational skills to coordinate, supervise, manage, or train others. Examples include dental assistants, electricians, fish and game wardens, legal secretaries, personnel recruiters, and recreation workers. 4 Minimum of two to four years of work-related skill, knowledge, or experience. Most require four-year bachelor’s degree. Usually need several years of workrelated experience, on-the-job training, or vocational training. Skills involve coordinating, supervising, managing, or training others. Examples include accountants, chefs and head cooks, computer programmers, historians, pharmacists, and police detectives. 5 Extensive skill, knowledge, and experience. Bachelor’s degree is minimum formal education required, but many require graduate school (master’s degree, Ph.D., M.D., or J.D.). Usually, a person will already have required skills, knowledge, work-related experience, or training. Very advanced communication and organizational skills. Examples include athletic trainers, lawyers, managing editors, physicists, social psychologists, and surgeons. do not discuss the value of assets, including trust funds set aside to cover future liabilities (e.g., public employee pension systems such as CalPERS and CalSTRS) or outstanding debt. As mentioned above, we ignore 14 Table 2.4 Average Wages, by State and Job Zone, 2002 (dollars) California U.S.–California Florida Illinois New York Texas United States 1 22,067 20,681 19,139 21,162 22,446 19,336 20,831 2 29,771 26,896 25,000 28,283 30,055 24,894 27,223 Zone 3 42,660 36,759 34,261 38,629 42,875 34,570 37,432 4 62,711 54,294 51,958 55,026 67,591 52,048 55,285 5 83,331 73,242 73,370 72,214 85,652 73,535 74,371 Table 2.5 Wage Indexes Relative to the United States, by Job Zone, 2002 California U.S.–California Florida Illinois New York Texas Zone 1 2 3 45 106 109 114 113 112 99 99 98 98 98 92 92 92 94 99 102 104 103 100 97 108 110 115 122 115 93 91 92 94 99 distinctions between funds, including so-called “nongovernmental cost funds” such as bond funds. We do this because we are trying to capture total spending in each major area and, although these funds may be predetermined or “off-budget” in a given year, they are subject to change over time. Nevertheless, the uneven nature of capital spending and the durable nature of these investments require that we treat these expenditures differently. Thus, in addition to examining capital outlays in a given year, we evaluate trends in capital spending per additional user over a longer period.7 This method is not perfect: It ignores differences in the _____________ 7These comparisons are usually from 1992 to 2002. Because the Census Bureau does not report expenditures for state and local governments separately by state for 2001, they include cumulative spending over nine years and population growth over ten years, except in Chapters 4 and 8, which interpolate the data for 2001. We could also consider 15 age of a capital stock at the start of the period, which may affect the need for additional spending (e.g., retrofitting). It also does not consider the efficiency with which resources are used and how the intensity of use might affect the need for additional spending. Wherever possible, we note these factors and how they might affect relative spending differences. Finally, we refer throughout this report to “fiscal tradeoffs,” or choices across budget categories. The concluding chapter goes even further, asking what alternative tradeoffs might look like. Of course, these choices may be implicit rather than explicit. That is, individual government revenue or spending decisions may be made in isolation, or without regard for “spillover” effects on other areas. However, at the state, local, or even household level, these choices are necessarily interdependent. If state and local governments choose to collect less in revenue or spend more on a particular program, or are compelled to do so because of demographic and economic pressures, there will be fewer resources left over for other purposes. Having budget information cast in these terms is therefore essential to evaluate the consequences of current budget decisions and to decide whether the state ought to do things differently. ______________________________________________________________ debt service payments per year as a way to smooth capital expenditures over time. However, these payments reflect interest rates and terms of borrowing as well as capital outlays. 16 A Note on Budget Concepts Discussions of budget concepts can be confusing. To clarify, we define a few terms. For federal, state, and local governments, the debt is the total amount of resources owed, just as it is for households. At the federal level, the deficit is the difference between revenues and spending in any given year. It is also the incremental addition to the debt. When state revenues are insufficient to cover expenditures, this is referred to as a budget shortfall or gap. Some claim that states cannot run deficits because, unlike the federal government, they face constitutional balanced budget rules or limits on borrowing. In fact, the picture is more complicated. In about onethird of all states, including California, budgets must be balanced at the start but not at the end of a fiscal year. Also, although state constitutions often require voter and legislative approval for the issuance of “full faith and credit” debt, or general obligation bonds, other types of “nonguaranteed” borrowing are available as long as capital markets are willing to supply the funds. Nevertheless, there are limits to how much spending can outpace revenues. These limits are captured by what is known in economics as the intertemporal budget constraint. In brief, this concept states that the value in today’s dollars of all current and future revenue streams must be sufficient to cover all current and future spending obligations and indebtedness net of assets such as public buildings or land. In a two period example, the difference between taxes net of transfers, T1, and government purchases net of debt service, G1, is the deficit, d1. The next year, d1 becomes the accumulated debt D, which must be paid back with interest, r. Therefore, in period 2, T2 = (1 + r) D + G2. Recall that D came about because of the gap between G1 and T1. Therefore, T2 = (1 + r) (G1 – T1) + G2/(1 + r). In other words, a tax cut in period 1 will raise taxes in period 2, all else being equal. A gap between the net present value of all future outlays on one hand and revenues plus debt on the other is a fiscal imbalance. This imbalance may have negative consequences for social equity if future generations are compelled to pay higher taxes or receive less in public services and current investments do not yield benefits to them. This scenario further assumes that current generations will not offset this imbalance by leaving higher bequests, the so-called Ricardian equivalence result. 17 3. California’s Revenue System State and local governments depend on a variety of revenues to fund public programs. Major sources include taxes, fees or charges for services rendered, and contributions from the federal government. This chapter compares California’s overall level of revenues and its reliance on each of these sources to other states. As shown in this chapter, whether California is considered a “highrevenue” state depends on measurement choices. On a per capita basis, California collected more total and general revenues than the average for all other states in 2002. However, its state and locally generated revenues (i.e., taxes, fees or charges, and miscellaneous sources of income) were close to the average for the rest of the nation as a share of personal income. California relied on taxes, charges, miscellaneous revenues, and federal intergovernmental revenues in roughly the same proportions as the rest of the nation. California’s tax mix differs from that of other states because of its greater reliance on income taxes and lower reliance on property taxes. Total Revenues The U.S. Census Bureau classifies state and local government revenue sources using four major categories: taxes, charges, miscellaneous revenues, and federal intergovernmental revenue. Taxes are the predominant source of government income in all states except Alaska, where oil and mineral sales and investments provide more funds. Most states rely on a combination of income, sales, and property taxes. Income taxes are typically assessed at the state level. Personal income taxes generally follow the federal tax code, using federal adjusted gross income or taxable income as the basis for calculating tax liability and allowing federal itemized deductions. Beyond these provisions, states may offer additional tax credits and deductions as well as exempting certain types of income such as Social Security benefits, capital gains, and 19 dividends or interest payments. As discussed below, states define their own marginal tax rates and brackets, including the level at which income is untaxed. Corporate income taxes operate analogously to the individual income tax, applying to corporate profits less credits, exemptions, and deductions. Certain types of businesses (sole proprietorships, partnerships, and “S” corporations) may pay individual rather than corporate income taxes. Provisions for taxing corporate income vary widely by state. The most important sources of variation are the connection or “nexus” that must exist between a corporation and a state for the corporation to pay taxes in that state, which sources of corporate income are included in the tax base, and how much of this income will be apportioned to the state. The Census Bureau also includes in this category levies on specific industries—such as banks and other financial institutions—as well as franchise or license taxes that apply to corporate net income. Sales and use taxes are levied at both the state and local level. General sales taxes apply to the value of goods or services sold within a jurisdiction, with exemptions for specific items such as food or prescription medicines. Selective sales taxes, also known as excise taxes, apply only to specific commodities, such as motor fuels, alcoholic beverages, and tobacco products. The amount owed may depend on price or volume of a good that is sold. Use taxes are analogous to sales taxes except that they are imposed on items purchased out of state for use in the state of residence. Property taxes are levies on real property, including land and structures, as well as on personal property, including vehicles, equipment, and “intangibles” such as bank accounts. They are assessed and collected mainly at the local level, although states may be responsible for distributing these revenues as is the case in California. Property taxes do not include levies based on characteristics other than a property’s value, such as its area or frontage. These are included in special assessments or one-time levies to fund public capital improvements. In addition to taxes, state and local governments rely on income from user fees or charges. Charges are distinct from taxes because they must be reasonably related to the costs of providing a service or product. 20 They differ from special assessments because they vary with use of a service or good. Examples include fees for higher education and charges for patient care in public hospitals. Miscellaneous revenues include special assessments, interest earnings, lottery revenues, private donations, fines and forfeitures, rents, royalties, and other unallocated sums. This category does not include license taxes, permit fees, or other payments to offset the costs of regulation, which have their own category under taxes. Finally, state and local governments receive intergovernmental revenue from the federal government in the form of grants for specific purposes, such as building or maintaining highways and mass transit systems, or reimbursements for administering federal or joint federalstate programs, such as Medicaid. State governments also provide assistance to local governments. To avoid duplication, the Census Bureau subtracts the value of the latter transactions from combined state and local expenditure totals. In 2002, California state and local governments collected $251 billion in revenues, equivalent to $7,166 per capita (Table 3.1). By comparison, the average state not including California took in $6,155 per capita, or about 14 percent less than California. Other states in our comparison group had lower revenues per capita than California, ranging from 18 percent lower in Illinois to 27 percent lower in Texas. On the other hand, New York collected 25 percent more in total revenues than California. Table 3.1 State and Local Total Revenues per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 7,166 6,155 5,553 5,847 8,928 5,249 General 6,478 5,764 5,195 5,491 8,122 5,024 21 General Revenues Total revenues include payments to public water, gas, electric, and transit utilities as well as contributions to insurance trusts such as public employee retirement plans and investment earnings from these funds. Excluding these sources of funds leaves what the Census Bureau refers to as the general government sector.1 Compared to total revenues, California’s general revenues per capita are closer to those of other states. In 2002, general revenues were 12 percent higher than the average for all other states, or $6,478 per capita, compared to $5,764. Similarly, the gap between California and certain other large states narrows to 15 percent for Illinois and 22 percent for Texas. New York, however, still collected 25 percent more than California in 2002 by this measure. Figure 3.1 shows that the gap between California and the rest of the nation in general revenues per capita has been fairly consistent over time with a few notable exceptions. Prior to the passage of Proposition 13 in 1978, California collected nearly one-third more in general revenues per capita than the average state. The following year, this gap narrowed to 13 percent as Proposition 13 cut property taxes in half by rolling back assessed values to 1976 levels, among other provisions. (Ongoing revenue effects of Proposition 13 are discussed below.) For the most part, California maintained this distance from the average for all other states until the recession of the early 1990s. Because California was more deeply affected by this recession than most other states, its general revenues per capita fell from 15 percent above average in 1990 to 8 percent above average in 1993. As economic growth accelerated during the late 1990s, California’s general revenues per capita rose from a low of 6 percent above the average for all other states in 1997 to 14 percent above average in 2000. Although its collections exceed those of other states, California is similar to the rest of the nation in the composition of these revenues. In 2002, more than half of California’s general revenues (53%) came from _____________ 1The general government sector also excludes publicly owned liquor stores, which do not exist in California and which provide only a small share of revenues in all other states. 22 7,000 6,000 5,000 Revenues ($) 4,000 3,000 2,000 1,000 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.1—General Revenues per Capita, 1977–2002 taxes, one-quarter were derived from charges and miscellaneous revenues, and the remaining 20 percent were revenues from the federal government (Table 3.2). The contributions from each of these sources were nearly identical in the rest of the nation on average. Within our comparison group, Illinois and New York collected a greater share from taxes (60% and 57%) and less from charges (11% and 12%). Florida gained more income from charges and miscellaneous Table 3.2 State and Local General Revenues per Capita, 2002 (dollars) California U.S.–Californi a Florida Illinois New York Texas Total 6,478 5,764 5,195 5,491 8,122 5,024 Taxes 3,442 Charges 1,029 Federal Miscellaneous Intergovernmental Revenue Revenue 628 1,379 3,102 2,688 3,303 4,641 2,715 859 901 627 949 744 570 710 550 642 510 1,235 897 1,010 1,889 1,055 23 revenues (31%), and New York received more federal aid as a share of general revenues (23%). Own-Source Revenues Focusing on state and locally generated revenues, California state and local governments raised $5,099 per capita from their own sources (i.e., excluding federal contributions) in 2002, 13 percent more than the average for all other states. One way to understand the burden of these collections is to adjust for personal income. States with higher incomes also include more resources and economic activity to tax.2 Performing this adjustment narrows the gap between California and other states considerably. In 2002, California’s own-source revenues represented $155 per $1,000 of personal income, 5 percent higher than the $148 average for the rest of the nation (Table 3.3). Within our comparison group, California’s revenues were 7 percent higher than in Florida, 12 percent higher than in Texas, 13 percent higher than in Illinois, and 12 percent lower than in New York. Nationally, California ranked 19th in its revenue collections per $1,000, immediately below Montana and the District of Columbia and above Ohio and Alabama. Table 3.3 State and Local Own-Source Revenues, 2002 Per Capita ($) U.S. Rank Per $1,000 of Personal Income ($) U.S. Rank California 5,099 10 155 19 U.S.–California 4,530 — 148 — Florida 4,299 30 145 37 Illinois 4,480 26 138 42 New York 6,232 3 176 3 Texas 3,969 38 138 41 _____________ 2The widely used personal income series from the Bureau of Economic Analysis (BEA) provides an imperfect adjustment because it excludes certain types of taxable income and includes other nontaxable sources. It is not clear how adjusting for these differences would affect California’s relative position. 24 Revenues ($) Adjusting for income also changes California’s relative position over time (Figure 3.2). In 1978, California’s general own-source revenues per $1,000 were 15 percent higher than the average for all other states; by 1979, they were 5 percent below this average. Revenues per $1,000 remained below the U.S. average excluding California throughout most of the 1980s. California kept pace with the rest of the nation during the late 1980s and 1990s, surpassing this average in 2000 and remaining at this level in 2002. 180 160 140 120 100 80 60 40 California U.S.–California 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.2—General Own-Source Revenues per $1,000 of Personal Income, 1977–2002 Taxes Tax systems vary from state to state. Four states (Delaware, Montana, New Hampshire, and Oregon) do not levy a sales tax and one state (Alaska) does not have a state sales tax but allows cities and counties to have one. Seven states (including two in our comparison group, Florida and Texas) do not have a tax on personal income. Another two states have a personal income tax but apply it only to unearned income such as interest and dividends. Although the property tax exists in all U.S. states, several states have limits on tax rates, assessment increases, or total revenues from this source. 25 Most states rely on a “three-legged stool” of income, sales, and property taxes. Table 3.4 shows these collections per $1,000 of personal income in 2002. In percentage terms, the average U.S. state excluding California derived 25 percent of total tax revenue from income taxes, 36 percent from sales and use taxes, and 32 percent from the property tax (Table 3.4). California derived a similar share of tax revenues from sales taxes overall, although it collected less from selective sales taxes or taxes on particular goods (9% of total taxes). Compared to the average for all other states, income and property tax shares were reversed in California, with property taxes contributing 25 percent of total taxes and income taxes providing 32 percent. Among the states in our comparison group, Florida and Texas stand out because they do not have a personal income tax. As a result, these states depend more on collections from sales taxes (51% and 49% of total taxes, respectively) and property taxes (35% and 42% of total taxes, respectively). California collects a greater share of tax revenues from income taxes than any other state in our comparison group except New York, which takes in 40 percent of tax revenue from this source. At the same time, California receives less in property tax revenues than any of the other large states considered here. This has not always been the case. As noted above, property tax collections in California fell by half immediately after the passage of Proposition 13 in 1978. Since then, personal income and sales taxes have provided a greater share of state and local tax revenues (Figure 3.3). In particular, the share of state and local tax revenues from the personal Table 3.4 State and Local Taxes per $1,000 of Personal Income, 2002 (dollars) California U.S.–California Florida Illinois New York Texas License Individual Corporate General Selective and Total Income Income Sales Sales Property Other 105 29 5 27 9 26 9 102 22 3 25 12 32 8 91 — 2 31 16 32 10 102 18 3 18 16 39 7 131 45 8 25 9 40 6 95 — — 29 17 39 9 26 Revenues per $1,000 of personal income ($) 60 Property General sales 50 Selective sales Individual income Corporate income 40 License and other 30 20 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.3—California State and Local Major Taxes per $1,000 of Personal Income, 1977–2002 income tax grew rapidly during the late 1990s, from 26 percent in 1997 to 33 percent in 2000, only to fall again to 27 percent in 2002.3 The rising importance of the personal income tax stems in part from its progressivity. As incomes grew in the state during the late 1990s, the state raised more funds from this source. California has one of the highest top marginal tax rates in the nation at 9.3 percent, compared to 7.7 percent in New York as of 2004 (6.85% in 2002) and 3 percent in Illinois.4 California’s personal income tax also has more rate brackets (six) than other large states in our comparison group, with the exception _____________ 3Because they are of state and local taxes, these shares differ from widely reported figures for the personal income tax as a share of state General Fund revenues: roughly 50 percent, compared to less than 20 percent in 1964 (Legislative Analyst’s Office, 2004). 4In addition, high-income taxpayers in California may be subject to an Alternative Minimum Tax analogous to the AMT at the federal level. Unlike its federal counterpart, however, California’s AMT exemptions are indexed to inflation, preventing “bracket creep.” Also, in November 2004, voters approved Proposition 63, which imposed an additional 1 percent tax on the portion of a taxpayer’s income in excess of $1 million to fund mental health programs. 27 of New York. For example, Illinois has only one tax bracket (i.e., a flat tax) on income. Beyond its high marginal tax rates and graduated structure, California’s personal income tax is also more progressive than in the average state because of credits, exemptions, and deductions that preclude many households from paying state income taxes. For example, in 2000, two-parent families of four earning less than $37,000 were exempt from state income taxes in California. In New York and Illinois, comparable income thresholds were $23,800 and $14,000, respectively. As a result of this progressivity, in 2001 the top 1 percent of income taxpayers in California contributed 34 percent of all income tax receipts, whereas the bottom 40 percent provided 1 percent of these revenues. By comparison, in New York the top 1 percent of tax filers paid 30 percent of all individual income taxes and the bottom 40 percent paid 7 percent. The difference between California and Illinois was even more dramatic because of that state’s flat rate structure. There, the top 1 percent paid only 17 percent of all personal income taxes. In contrast to the personal income tax, sales and use taxes are generally viewed as regressive because consumption represents a larger share of income for less affluent households. The collections and incidence of sales and use taxes depend on the tax rate and tax base. Like all states in our comparison group, California has a relatively high combined state and local sales tax rate compared to the rest of the nation. In 2004, the maximum combined state and local sales tax rate was 8.75 percent in California, lower than that of Illinois but higher than that of Florida or Texas (Table 3.5).5 _____________ 5California’s combined rate includes one state tax at 5 percent with revenues dedicated to the state General Fund. There are also two state taxes at one-half of one percent with proceeds going to special funds for local government. In addition, there is a statewide local tax of 1.25 percent (the so-called Bradley-Burns rate). Finally, localities may levy their own sales taxes with local voter approval. Proposition 57, approved by California voters in March 2004, diverts a portion of the uniform Bradley-Burns rate to a special fund to repay the state’s deficit financing bonds. While this repayment continues, the state will replace local sales tax revenues with property taxes from schools and provide schools with a greater infusion of resources from the state General Fund. This is the socalled “triple flip.” 28 Table 3.5 State and Local Retail Sales Tax Rates, 2004 (percent) California U.S.–California Florida Illinois New York Texas Maximum Combined 8.75 7.37 7.5 9.25 8.75 8.25 State 6 5.31 6 6.25 4.25 6.25 Maximum Local 2.75 2.83 1.5 3 4.5 2 In California, as in all of our comparison states, general sales taxes exempt certain purchases. For example, California and all comparison states except Illinois exempt food for home consumption. (Illinois taxes these purchases at a lower rate.) California also exempts utility payments, prescription drugs, custom computer programs, snacks, animal feed, bottled water, and free newspapers or periodicals (California State Board of Equalization, 2002). Some states in our comparison group also exempt prescribed and nonprescribed medications or tax these purchases at a lower rate (Illinois). More significantly, California is one of several states that do not tax most services, even though these purchases constitute the bulk of consumer expenditures. According to a 2004 survey by the Federation of Tax Administrators (FTA), California taxes 23 out of 168 potential services, compared to 17 in Illinois, 56 in New York, 62 in Florida, and 81 in Texas. Similarly, the Center for Budget and Policy Priorities notes that California currently taxes only two out of 40 “readily taxable services” other than health care, housing, education, legal, banking, public transit, insurance, and funeral services. By comparison, Illinois taxes three out of 40 such services, whereas Florida taxes 22, New York taxes 23, and Texas taxes 24 (Mazerov, 2002).6 _____________ 6A related issue is enforcement of the “use” tax on remote sales by catalogue or through the Internet. The Government Accountability Office (GAO) estimates that California loses more revenues to untaxed remote sales than any of our comparison states (Government Accountability Office, 2000, pp. 59–61). California has joined the national Streamlined Sales Tax Project (SSTP), in part to raise congressional attention to 29 The more items that are exempt from taxation, the narrower the base of the tax. At the extreme are selective sales taxes, including “sin” taxes designed to reduce consumption of particular goods such as alcoholic beverages, tobacco products, or motor fuels and perhaps also to fund programs counteracting harmful consequences of these commodities (e.g., smoking cessation programs). Although California derives less income from selective sales taxes as a group than other states (Table 3.4), “sin tax” rates are often close to the national median (Table 3.6).7 Like some of the other taxes we have considered, California’s corporate income tax rate is relatively high at 8.84 percent, compared to 5.5 percent in Florida, 7.3 percent in Illinois, and 7.5 percent in New York. In California, banks and other financial institutions also pay a surcharge of 2 percent, whereas subchapter S firms pay a lower rate (1.5%), and limited liability companies pay a flat fee.8 How much corporate income is subject to taxation depends on rules for determining the tax base and apportioning income to the state. California determines the amount of income subject to taxation using “combined reporting” for all holdings of a unitary business group rather than using separate reporting for each incorporated entity. Within Table 3.6 Selective Sales Tax Rates, 2005 California U.S. median Florida Illinois New York Texas $ per Pack of Cigarettes 0.87 0.70 0.34 0.98 1.50 0.41 Liquor 3.30 3.75 6.50 4.50 6.44 2.40 $ per Gallon of Wine Beer 0.20 0.20 0.64 0.19 2.25 0.48 0.73 0.19 0.19 0.11 0.20 0.19 Gas 0.18 0.20 0.40 0.19 0.80 0.20 ______________________________________________________________ this issue. It has not, however, conformed its sales and use tax laws to those of the Streamlined Sales and Use Tax Agreement (Jones, 2005). 7These rates are exclusive of federal taxes and general sales taxes on these goods. California’s Proposition 42 earmarks the latter revenues from gasoline sales to transportation purposes. 8Like individuals, corporations are also subject to an AMT at 6.64 percent of income. 30 our comparison group, Illinois also requires combined reporting, but New York and Florida do not. California apportions income from multistate and multinational corporations to the state based on a formula that weights property, payroll, and sales in the state. Like Florida and New York, California assigns double weight to sales to encourage businesses to invest capital and labor in the state. In addition, firms may elect to report all business activity or that which stops at the “water’s edge” (i.e., within the United States only). California’s corporate income tax also includes several exclusions, exemptions, deductions, and credits. Notably, firms can claim research and development credits and can carry forward net operating losses (NOLs) as deductions against future earnings.9 Understanding the net effects of these credits on revenue collections and economic competitiveness is complicated. However, it is interesting to note from Figure 3.3 that corporate tax earnings declined as a share of personal income in the 1990s, consistent with the promulgation of credits over this period, such as the Manufacturer’s Investment Credit (MIC) on intermediate purchases, as well as a trend toward corporate reorganizations. As noted throughout this chapter, the passage of Proposition 13 in 1978 dramatically altered the public finance landscape in California. This measure rolled back assessed property values to 1976 levels, capped property tax rates at 1 percent, and limited the growth in assessed values to 2 percent per year except for properties that are sold in that year. It also instituted a two-thirds popular vote requirement for all new state taxes and local special taxes. As a result, California has collected less in property tax revenues than the average for all other states since 1977. Local governments in particular have become more dependent on intergovernmental revenue, contrary to the trend in the rest of the nation and in all of our comparison states (Table 3.7). Although charges and miscellaneous _____________ 9After two years of suspension, the carry forward provision was reinstated in 2004, and in January 2005, the portion of losses eligible for deduction increased to 100 percent. 31 Table 3.7 Local Revenues per $1,000 of Personal Income, by Source, 2002 (dollars) California U.S.-California Florida Illinois New York Texas California U.S.-California Florida Illinois New York Texas Taxes 59 44 33 45 81 37 37 42 40 47 67 49 Charges and Miscellaneous Revenues 1977 17 17 25 12 21 17 2002 33 25 37 20 27 24 Intergovernmental Revenues 55 46 45 35 85 31 69 41 35 37 60 31 sources have become a more important source of revenue in all states since 1977, this growth is more pronounced in California. As in other states, California’s property tax includes many provisions designed to reduce the burdens on specific taxpayers. Like many states, California offers a homestead exemption that reduces the assessed value subject to taxation for an owner-occupied home. It also provides limited relief to elderly homeowners through “circuit-breakers,” or credits against property taxes when liabilities exceed a certain amount, and property tax deferrals. Whereas California’s property tax may fare as well as any state’s in terms of its “vertical equity,” it has been criticized on “horizontal equity” grounds for treating similarly situated homeowners differently based on the purchase date of a property. On the other hand, these effects have been diminishing over time with the turnover of properties and periods of lower price inflation (e.g., Sheffrin and Sexton, 1998). 32 Conclusions California collects more revenues than the average state in the rest of the nation. In 2002, it collected 16 percent more in total revenues, 12 percent more in general revenues, and 13 percent more in own-source (i.e., nonfederal) revenues than the average for all other states. Among other large states in our comparison group, Florida and Texas collected less per capita than California whereas New York had higher revenues per capita than California. California is also a high-income state. When we consider revenues as a proportion of personal income, California looks more like the rest of the nation. In 2002, own-source revenues per $1,000 in California were 5 percent higher than the average for all other states. Differences with respect to other large states are also narrower. California’s tax mix is relatively balanced, relying on roughly equal proportions of the income, sales, and property tax, although tilting more than other states toward income taxes and away from property taxes. California’s heightened reliance on the income tax stems from the progressivity of this tax, from rising incomes in the state in recent years, and from an increasing proportion of income from nonwage sources, such as stock options and capital gains. For these reasons, several analysts have pointed out that California’s tax system has also been more volatile than that of other states (Legislative Analyst’s Office, 2005). 33 4. Elementary and Secondary Education Elementary and secondary education, more commonly known as K–12 education, is the largest category of state and local government expenditures. In 2002, this category constituted 19 percent of all expenditures in California and 21 percent in the rest of the country. This chapter analyzes expenditure differences between California and other states in this important area and also briefly summarizes expenditures for public libraries and a variety of other activities related to K–12 education. We focus first on expenditures per capita because this is a better measure of the cost of public schools to taxpayers than is the commonly used measure of expenditures per pupil. We then consider expenditures per pupil and its components, including staffing levels and average employee compensation. As shown below, expenditures per capita and staff per pupil together often paint a more revealing picture than do expenditures per pupil alone. Per capita expenditures on K–12 education in California are similar to those in other states. However, public schools in California have far fewer staff resources per student than do schools in other states. This outcome stems primarily from two factors: California has more students per capita than other states, and school districts in California compensate their employees more highly than districts in other states. Although the link between resources and outcomes in K–12 education is not always clear, it is interesting to note that student achievement in California also lags that of other states. Total Expenditures The activities considered in this chapter are relatively easy to describe. Elementary and secondary education as defined by the U.S. Census Bureau includes the instructional activities of schools overseen by 35 public school districts and county offices of education. It also includes the administrative support for these activities and other support and auxiliary services such as school lunch programs, counseling services, and pupil transportation. As a practical matter, K–12 expenditures in California are the expenditures of its school districts, their joint powers agencies, and the county offices of education. Any expenditure of those three types of agencies is counted as an expenditure on elementary and secondary education. Public libraries are the libraries run by cities, counties, and states. This category excludes the libraries of public schools, colleges, and universities, which are included in the larger categories of elementary and secondary education or higher education. The Census Bureau defines a third area, “other education,” which includes services related to education that are not provided directly by libraries, public schools, and public colleges and universities. In California, this category includes the Office of the Secretary for Education and the operations of the State Department of Education. It also includes the vocational rehabilitation services provided by the Department of Rehabilitation Services. In per capita terms, library and other education expenditures are relatively small (Table 4.1). Although expenditures per capita in those two areas differ across states, the differences are not large enough to be an important factor in comparing educational expenditures in one state to Table 4.1 Education Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 1,592 1,490 1,175 1,531 2,059 1,518 Elementary and Secondary Education Libraries 1,521 1,415 1,116 1,426 1,999 1,476 29 29 23 45 39 16 Other Education 42 46 36 60 21 26 36 those in another. As a consequence, the remainder of this chapter focuses on the Census Bureau category of elementary and secondary education. In contrast, there are large differences across states in expenditures on elementary and secondary education. In 2002, California was roughly representative of other states. However, New York spent $478 more per capita than California (31% more), and Florida spent $405 less (27% less). In Texas and Illinois, public school spending per capita was approximately equal to that of California. California’s position relative to other states has changed significantly over time (Figure 4.1). In 1977, California was spending 13 percent more per capita than the rest of the nation; by the end of the decade, this margin had essentially disappeared. The 1970s was also the period during which California transformed its school finance system from one in which school districts levied their own property tax rates to the current system in which the property tax rate is capped statewide and the legislature decides how much property tax revenue each district receives (Sonstelie, Brunner, and Ardon, 2000). 1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 California 800 U.S.–California 700 600 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.1—Total Expenditures on Elementary and Secondary Education per Capita, 1977–2002 37 Expenditures (2002 $) After falling relative to other states in the 1970s, K–12 education expenditures per capita were roughly the same in California as in other states during the 1980s. This parity ended during the recession of the early 1990s, which was felt strongly in California. California’s real spending per capita fell by 10 percent between 1991 and 1994, while real spending per capita rose by 3 percent in the rest of the nation. From this low point, however, public school spending per capita has risen much faster in California than in other states. Between 1994 and 2002, real spending per capita rose by 45 percent in California, compared to 21 percent in the rest of the nation. Total expenditures on K–12 education can be partitioned into current and capital expenditures. Current expenditures include the salary and benefits of teachers, instructional aides, counselors, principals, and other staff. They also include the purchase of goods and services such as textbooks and utilities. Capital expenditures include outlays for constructing new schools and renovating existing schools. They also include purchases of equipment such as desks and computers. For California and its comparison states in 2002, capital outlays constituted 12 to 16 percent of total expenditures (Table 4.2). Because capital expenditures have long-lasting effects, however, outlays in one year are not particularly informative. Later in the chapter, we examine trends in these expenditures over many years. Table 4.2 Total Expenditures on Elementary and Secondary Education per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita 1,521 1,415 1,116 1,426 1,999 1,476 Current Expenditures per Capita 1,320 1,232 936 1,197 1,748 1,229 Capital Expenditures per Capita 201 182 180 229 251 247 38 Current Expenditures California’s current expenditures on K–12 education followed the same general trend as total K–12 expenditures: Real spending per capita declined during the 1970s and rose in each year after with the exception of the early 1990s. Compared to other states, California began the 1970s spending significantly higher amounts per capita, fell to parity in the 1980s, declined below spending in the rest of the nation in the 1990s, and returned to parity at the end of the decade. Underlying the trends in current expenditures per capita are fundamental trends in school enrollment. During the 1970s, for example, enrollments were falling throughout the country, so the decline in expenditures per capita did not necessarily translate into fewer resources for the typical school. Similarly, if one state has fewer students per capita than another, lower expenditures per capita in that state do not necessarily imply that its schools have fewer resources. In concept, expenditures per capita measure the cost of public schools to the average taxpayer, and expenditures per student measure the resources provided to the typical school. The two measures are tied together by Eq. (2.1) in Chapter 2. In the particular case of K–12 education, public school students play the role of caseload, yielding the following version of that equation: Expenditures = Public School Students  Expenditures (4.1) Population Population Public School Student Conceptually, the ratio of public school students to total population is the cost to the average taxpayer of increasing public school spending per student by one dollar. Raising expenditures per student is less expensive when there are fewer students. For example, doubling the number of public school students would double the cost in taxes to the average taxpayer of increasing expenditures per pupil by one dollar. Compared to other states in 2002, California had high student enrollments, making K–12 education per student relatively expensive (Table 4.3). Using Eq. (2.4) in Chapter 2, the effect of this higher cost is reflected in the following way: 39 Table 4.3 Components of Current Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 1,320 1,232 936 1,197 1,748 1,229 Students Expenditures per Capita per Student ($) 0.179 7,391 0.164 7,523 0.150 6,245 0.165 7,271 0.150 11,653 0.192 6,412  Expenditures   Public School Students +  Expenditures Population Population Public School Student (4.2) In this equation,  represents the percentage difference in a variable between two states or between California and all other states. Taking the first term in Eq. (4.2), in 2002 expenditures per capita in California were 7 percent higher than in other states. However, public school students per capita, the second term, were 9 percent higher in California than in other states. As a result, expenditures per student, the third term, were 2 percent less in California than in other states. California’s higher spending per capita was not quite enough to offset its higher public school enrollments per capita. Comparisons between California and Texas are also informative. In per capita terms, Texas spent 7 percent less than California. It also had 7 percent more students per capita, resulting in expenditures per student that were 14 percent less than in California. In this comparison with California, New York was the mirror image of Texas. It spent more per capita (32%) and had fewer students per capita (16%), resulting in expenditures per student that were 58 percent higher than in California. Each year, the National Education Association (NEA) ranks states according to their current expenditures per pupil on K–12 education (National Education Association, 2003). This well-known ranking is based on a slightly different definition of current expenditures than that used by the Census Bureau. The Census definition includes all current expenditures of school districts and county offices of education. The 40 NEA definition excludes some of those expenditures, particularly expenditures for summer schools, adult education, and community services. For 2002, the NEA measure of current expenditures per pupil in California was 4.5 percent less than the Census measure. There were also significant differences between the NEA and the Census measures for Illinois and Texas. Despite these differences, however, the NEA measure paints a similar picture to that of Table 4.3: Illinois spent about the same per pupil as California, Texas and Florida spent less, and New York spent much more. Equation (2.2) in Chapter 2 describes how the caseload for a public service can be decomposed into a take-up rate and an eligibility rate. In the case of K–12 education, this equation is: Public School Students = Public School Students Population Children Eligible for Public School  Children Eligible for Public School Population (4.3) For both the take-up rate and the eligibility rate, the difficulty lies in determining the number of children eligible for public school. Children are generally eligible for 13 years of schooling, kindergarten through high school, although they can also be retained for one or more grades, implying more years in school. In addition, at certain ages, compulsory attendance laws require that children attend either a public or a private school. In California, children must start school at age six and attend school until they either receive a high school diploma or reach the age of 18. Among other states, the starting age differs from age five to age seven, and the dropout age differs from age 16 to age 18. Accordingly, we define children eligible for public schools to be all children in public and private schools plus those who have dropped out of school before graduating as well as immigrant children who were never enrolled. The Census Bureau provides a working definition of the latter two groups in the number of children between ages 16 and 19 who do not have a high school diploma and are not enrolled in school. The cutoff at 41 age 19 is somewhat arbitrary; a twenty-year-old who has not finished high school is still eligible to enroll in school. However, a person without a high school diploma is less likely to enroll in school as his or her age increases, making some cutoff age a reasonable restriction on the definition of those eligible to enroll in school. Table 4.4 uses the Census restriction to define children eligible for school. The number of public school students per capita is relatively high in California, partly because of the state’s relatively high take-up rate. The take-up rate is the percentage of children eligible for a public school who actually attend such a school. This rate is the product of two factors: the percentage of eligible children who attend any school (public or private) and the percentage of those students who attend a public school. The first factor is presented in the third column of Table 4.4. Across all states, 96 to 98 percent of children eligible for public school are enrolled in a public or private school. This factor is not a significant source of difference among states in the take-up rate. However, the percentage of these students who attend a public school (the second column) varies from 86 percent to 93 percent. California has a relatively high public school enrollment rate of 90 percent, which is the primary source of its relatively high take-up rate. A more important source of California’s high number of public school students per capita is its high eligibility rate, the last column in Table 4.4. In California, 21.3 of every 100 residents were eligible for public school, as opposed to 19.6 in the rest of the country, a difference Table 4.4 Components of Public School Enrollment, 2000 California U.S.–California Florida Illinois New York Texas Public School Students per Capita 0.188 0.170 0.154 0.167 0.162 0.194 Students in Public School (%) 90 89 88 87 86 93 Eligible Children in School (%) 97 97 97 97 98 96 Children Eligible for Public School per Capita 0.213 0.196 0.180 0.198 0.194 0.217 42 of 9 percent. Among our comparison states, only Texas had a higher eligibility rate, with 21.7 of every 100 residents eligible. In fact, in both Texas and California, a relatively high eligibility rate combined with a relatively high take-up rate produce a high rate of public school students per capita. The relative size of a state’s public school enrollment may change significantly over time, with significant consequences for the cost of K–12 education (Figure 4.2). From 1977 through 1990, public school students per capita fell steadily outside California. Since 1990, that ratio has been relatively constant. In California, the ratio also fell in the 1970s and 1980s, but the decline was steeper and shorter. By 1985, the number of public school students per capita had reached its lowest level in California and had begun to rise. In the 1990s, while other states were benefiting from a relatively stable ratio of students to population, California was coping with an increase in that ratio and thus in the cost per capita of public elementary and secondary education. During the 1990s, three of our four comparison states also saw increases in students per capita. However, none experienced an increase as rapid as California’s. During the 1990s, the number of students per capita increased by 7 percent in California. In comparison, the ratio rose 0.21 0.20 California U.S.–California 0.19 0.18 0.17 0.16 0.15 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.2—Number of Public School Students per Capita, 1977–2002 43 Number of students by 5 percent in New York, 4 percent in Florida, and 2 percent in Illinois. Texas actually experienced a 4 percent decline in its ratio during this period. These demographic trends are evident when the focus shifts from expenditures per capita to expenditures per pupil (Figure 4.3). Throughout the country, real expenditures per capita fell from 1977 to 1985. However, expenditures per pupil actually rose during this period because of falling enrollments. In other words, the resources of public schools in the rest of the nation were improving during this period, even though relatively less tax revenue (on a per capita basis) was being allocated to them. On the other hand, California’s expenditures per pupil were flat from 1977 to 1985, and the steady increase in California’s studentpopulation ratio from 1985 onward magnified the decline in spending per capita during the early 1990s, resulting in a decrease in real expenditures per pupil of 13 percent from 1991 to 1994. This decline reversed in the late 1990s, however. Although pupils per capita continued to rise in California throughout the 1990s, real expenditures rose faster than enrollment from 1995 through 2002. In 8,000 7,500 7,000 Expenditures (2002 $) 6,500 6,000 5,500 5,000 4,500 4,000 3,500 California U.S.–California 3,000 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.3—Current Expenditures on Elementary and Secondary Education per Student, 1977–2002 44 fact, California’s real expenditures per pupil rose by 30 percent from 1995 to 2002. In contrast, in the rest of the country, real expenditures per pupil rose by only 15 percent during this period. Thus, by 2002, California had largely erased the gap in expenditures per pupil that had developed between itself and other states during the recession of the early 1990s. From 1977 to the present, expenditures per pupil in our comparison states rose significantly relative to California. In 1977, expenditures per pupil in Texas were 66 percent of expenditures per pupil in California. The ratios for Florida, Illinois, and New York were 82 percent, 90 percent, and 122 percent, respectively. By 2002, spending per pupil in Texas was 87 percent of that in California. For Florida, Illinois, and New York, the ratios were 85 percent, 98 percent, and 158 percent, respectively. Each of the four states had gained in expenditures per pupil relative to California. The most dramatic gains were in the mid-1990s. In 1997, for example, New York, Florida, and Illinois were spending more per student than California, and Texas was at 95 percent of California’s level. In addition to pupils per capita, another important element in the cost of education is the salaries and benefits paid to school district employees. For the nation as a whole in 2002, employee salaries and benefits constituted 81 percent of current expenditures on K–12 education. Employee benefits represented 20 percent of total compensation. The salaries and benefits of teachers and instructional aides accounted for only 69 percent of all employee compensation. Table 4.5 partitions current expenditures per student into staff compensation per student and other expenditures per student, including purchases of good and services such as textbooks, utilities, office supplies, and so on. In California, staff compensation per student was slightly less than in the rest of the nation and other expenditures were slightly higher (Table 4.5). Following Eq. (2.3) in Chapter 2, staff compensation per student can be decomposed in the following way: Compensation = Staff  Compensation Student Student Staff (4.4) 45 Table 4.5 Components of Current Expenditures per Student, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Current Expenditures per Student 7,391 7,523 6,245 7,271 11,653 6,412 Staff Compensation per Student 6,041 6,300 4,896 6,339 9,350 5,355 Other Expenditures per Student 1,350 1,223 1,349 932 2,303 1,058 Staff per student is a measure of the personnel resources schools have, and compensation per staff is a measure of the cost of each unit of resources. These two components are listed in Table 4.6. Despite nearly average staff compensation per pupil, employees in California’s K–12 were paid more on average, and thus California schools had relatively low staff-student ratios. California schools spent about 4 percent less per student on staff compensation than schools in the rest of the country. Where California differed more dramatically was in average staff compensation, which was 34 percent higher than in other states. With staff compensation per student approximately the same as in other states, these higher employee costs necessarily implied a lower ratio of staff to students. As the third column of Table 4.6 shows, the Table 4.6 Components of Compensation per Student, 2002 California U.S.–California Florida Illinois New York Texas Compensation per Student ($) 6,041 6,300 4,896 6,339 9,350 5,355 Compensation per Staff ($) 65,710 48,982 43,307 51,387 63,456 38,269 Staff per Student Ratio 0.092 0.129 0.113 0.123 0.147 0.140 46 ratio of staff to students in California was only 71 percent of the ratio in other states. Because of average staff compensation per pupil and above average employee compensation, California schools were thinly staffed relative to schools in other states. New York also had high compensation costs for public school employees, with an average nearly equal to the average in California. However, its very high expenditures per student (58% higher than California) meant that it could achieve a higher staff-student ratio than California (60% higher), despite its high compensation levels. Texas also had a much higher staff-student ratio than California (52% higher), but that ratio was not due to higher spending per pupil. Texas actually spent less per pupil than California. However, its average compensation was 58 percent of the average in California. In these comparisons, Illinois and Florida were less dramatic versions of New York and Texas, respectively. Both states had higher staff-student ratios than California’s ratio. California’s relatively low staffing ratio holds for all categories of employees (Table 4.7). On a per student basis, other states had 33 percent more teachers than California. In terms of support staff such as instructional aides, counselors, and librarians, other states had 43 percent more staff per student than California. Similar comparisons hold for administrators, such as superintendents, principals, and administrative assistants, and for other staff, such as janitors, bus drivers, and security guards. In the administrative category, other states had 19 percent more Table 4.7 Public School Staff per 1,000 Students, 2002 California U.S.–California Florida Illinois New York Texas Teachers 49 65 54 63 73 68 Support Staff 17 24 21 24 22 20 Administrative Staff 11 14 15 13 15 17 Other Staff 15 26 24 24 37 34 47 staff per student than California; in the “other” category, this margin was 73 percent. Employee compensation is the sum of salaries and benefits, and California’s relatively high compensation for public school employees extends to both components (Table 4.8). The ratio of benefits to total compensation is 20 percent in California, slightly lower than the average for all other states. Three of California’s four comparison states have similar benefit ratios. The only exception is Texas, in which the benefit ratio is only 13 percent. School districts must compete for most of their employees in the market for college-educated workers. In California, half of public school employees are teachers, who must have a college degree. Thirty percent are administrators, counselors, and other support staff, who generally have college degrees. As a consequence, the best comparison group for public school employees is Zone 4 workers (as described in Chapter 2), who are also employed in occupations requiring a college degree. In 2002, the average salary for these workers was 13 percent higher in California than the average in all other states. We used this difference to adjust the salaries of public school employees for differences in labor market conditions among states. The result is shown in the second column of Table 4.9. After adjustments for labor market conditions, the salaries of public school employees were about 17 percent higher in California than in other states. Before that adjustment, the differential was 35 percent. Thus, about half of the difference between the salaries of public Table 4.8 Components of Average Compensation, 2002 California U.S.–California Florida Illinois New York Texas Average Compensation ($) 65,710 48,982 43,307 51,387 63,456 38,269 Average Salary ($) 52,449 38,777 34,234 40,797 50,078 33,133 Average Benefit ($) 13,261 10,205 9,072 10,590 13,377 5,137 Benefit Ratio (%) 20 21 21 21 21 13 48 Table 4.9 Average Public School Salary Adjusted for Location, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Salary of Public School Employee 52,449 38,777 34,234 40,797 50,078 33,133 Average Salary Adjusted for Location 46,238 39,485 36,426 40,989 40,961 35,193 school employees in California and other states can be explained by differences in general labor market conditions in California and other states. The general compensation pattern portrayed in Table 4.9 holds in particular for public school teachers. According to the annual survey of the National Education Association, the average salary of California teachers in 2002 was $54,348, the highest average of any state in the country (National Center for Education Statistics, 2003). Teachers in Florida had an average salary that was 70 percent of the California average, a little higher than the 65 percent that we calculate for all school employees as reported in Table 4.9. For Texas, the comparisons are similar—71 percent for teachers versus 63 percent for all school employees. Illinois was somewhat different, however. Its teachers had an average salary that was 91 percent of the California average, considerably higher than the 79 percent for all school employees. For New York, the percentages were virtually the same for teachers and for all employees. Our analysis thus far has emphasized three points. First, expenditures per capita on K–12 education are about the same in California as in other states. Second, California has more public school students per capita than other states, resulting in lower expenditures per student. Third, public school employees in California are compensated more highly than public school employees in other states, resulting in fewer staff per student. 49 Another way to understand the difference between California and other states is to ask how much spending would have to be increased in California for its schools to have the same resources as schools in other states. As a baseline, suppose California looked exactly like other states. In particular, suppose it had the same staff compensation, the same ratio of staff to students, the same level of other expenditures per student, and the same fraction of public school students in the population. Under those suppositions, California would have the same expenditures per capita as other states. Now, suppose that staff compensation were increased to California’s level and everything else remained the same. Expenditures per capita would increase by 29 percent. In addition, if the fraction of students in the population were also increased to California’s level, expenditures per capita would increase by 40 percent. In that sense, those two factors together, compensation per staff and students per capita, increased California’s cost of elementary and secondary education by 40 percent. Seventy-three percent of that increase was due to higher compensation. An alternative calculation of this cost starts with California’s compensation levels and ratios of staff per pupil and pupils per capita. From that baseline, if compensation per staff in California were reduced to the level of other states, California expenditures per capita would fall by 21 percent. If pupils per capita were also reduced to the level of other states, expenditures per capita would fall by another 6 percent for a total decline of 27 percent. Accounted for in this way, California’s high staff compensation and ratio of pupils per capita increased the cost of elementary and secondary education by 27 percent. Capital Expenditures As noted above, K–12 education capital expenditures per capita were higher in California than in the rest of the nation in 2002. However, as Figure 4.4 shows, this situation was without precedent in the previous 25 years. Throughout the 1980s and 1990s, capital expenditures per capita were substantially lower in California than in the rest of the country. 50 Expenditures (2002 $) 250 California 200 U.S.–California 150 100 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.4—Capital Expenditures on Elementary and Secondary Education per Student, 1977–2002 An important part of capital expenditures is the construction of new school buildings. Because school buildings are long-lasting, the need for new construction is better measured by the increase in student enrollment than by the level of enrollment. The need for additional school buildings is less in a state with high but stable enrollment than in a state where enrollment is lower initially but growing rapidly. In fact, enrollment growth rates varied considerably across states. From 1992 to 2002, enrollment grew by 20 percent in Texas, 22 percent in California, and 29 percent in Florida. In contrast, enrollment grew by only 12 percent in Illinois and 8 percent in New York. For all states except California, enrollment grew by 12 percent from 1992 to 2002. Table 4.10 compares the sum of annual capital outlays from 1992 to 2002 with the increase in enrollment during that period. Capital expenditures per additional student were substantially lower in California than in the rest of the country. Part of this difference may be due to the renovation of older school buildings. Even if a state is barely growing in enrollment, it must undertake some capital expenditures for renovation, and so its capital outlays per additional student will be very high. This possibility is consistent with the very high capital outlays per additional 51 Table 4.10 Capital Outlay per Additional Student, 1992–2002 California U.S.–California Florida Illinois New York Texas Increase in Enrollment, 1992 to 2002 1,140,581 4,530,953 568,347 223,225 228,139 699,076 Real Capital Outlay, 1992 to 2002 (1,000s of 2002 $) 45,348,972 366,246,776 24,971,202 20,497,211 32,138,075 40,025,032 Real Capital Outlay per Additional Student (2002 $) 39,760 80,832 43,937 91,823 140,871 57,254 student in Illinois and New York. On the other hand, enrollment in Texas grew about as rapidly as in California and Florida’s enrollment grew more rapidly in proportionate terms, yet capital outlays per additional student in both Texas and Florida were also significantly greater than in California. California has not responded as vigorously to the need for new school buildings as have Texas and Florida. As Figure 4.4 demonstrates, however, California did substantially increase its capital expenditures during the latter part of this ten-year period, an increase partly reflecting two voter-approved propositions. Proposition 1A, passed in 1998, authorized the state to issue over $9 billion in bonds for the construction and modernization of buildings in public schools, colleges, and universities. Proposition 39, passed in 2000, made it easier for school districts to issue their own bonds for construction and modernization. Before Proposition 39, school districts had to secure a two-thirds approval of their voters before issuing such bonds. The proposition lowered that requirement to 55 percent. From the passage of Proposition 39 through 2004, voters approved local school bonds totaling $20 billion. Nearly half of that total would not have been approved under the old requirement (Rueben and de Alth, 2005). This trend of increasing capital expenditures has continued beyond 2002. In November 2002, the voters of California passed Proposition 47, which authorized the state to sell bonds worth $13 billion for building and remodeling public schools and also facilities of public colleges and universities. In March 2004, the voters passed Proposition 52 55, which authorized $21 billion for the same purpose. The capital expenditures resulting from these bonds do not show up in Figure 4.4, but they will in future years. Outcomes School resources are not the ultimate measure of a school system’s success. In fact, if a state’s students are learning the lessons it expects them to learn and its schools require less resources than schools in other states, its school system is very efficient, providing a good return on the public funds invested in it. Although California schools are leaner than schools in other states, we cannot say that they are particularly efficient because students in California schools do not appear to be learning as much as comparable students in other states. One measure of student outcomes is the Stanford Achievement Test, Ninth Edition (SAT9). The SAT9 was the first test administered as part of California’s new accountability system. From the spring of 1998 to the spring of 2002, California public school students in grades 2 through 11 were required to take the SAT9 reading and math exams. Because the SAT9 was first given to more than 500,000 students throughout the country, the test permits a straightforward comparison of California students with students in other states. The test reveals that California students lagged behind students in other states, particularly in reading. For example, in 1998 only 39 percent of California students exceeded the national median score in reading. The performance of California students improved each year, but more than half were still below the national median by 2002. In math, the performance of California students was noticeably better. Whereas only 42 percent exceeded the national median in 1998, that number had reached 55 percent by 2002. Another measure of student achievement across states is the National Assessment of Educational Progress (NAEP), a standardized test of reading, mathematics, and other subjects administered by the U.S. Department of Education to a random sample of fourth, eighth, and twelfth graders across the country. The number of twelfth graders taking the test is too small to provide meaningful comparisons for some 53 subgroups of students, so we focus on the scores of fourth and eighth graders in reading and mathematics. California students compare poorly to students in other states. In both reading and mathematics, the average scores of California fourth and eighth graders are lower than the national average and lower than the scores in any of our four comparison states (Table 4.11). In each of these comparisons, the difference between California students and other students is statistically significant.1 The relatively poor results for California students may be partly explained by the large number of students who are recent immigrants and still not proficient in English. Indeed, 30 percent of California fourth graders who took the NAEP were classified by their schools as not fully proficient in English. In contrast, only 8 percent of fourth grade test-takers in the nation were not English proficient. Among our four comparison states, Texas, at 12 percent, had the highest percentage of test-takers who were not proficient in English. When test results are restricted to students who were proficient in English, California’s performance improves but still lags that of other states (Table 4.12). In reading, the average for California students in both fourth and eighth grade is lower than the national average and Table 4.11 NAEP Scores for All Students, 2003 Grade 4 Grade 8 California United States Florida Illinois New York Texas Reading 206 216 218 216 222 215 Mathematics 227 234 234 233 236 237 Reading 251 261 257 266 265 259 Mathematics 267 276 271 277 280 277 _____________ 1For fourth graders in the country as a whole, the standard errors of the average scores in reading and math are 0.3 and 0.2. For eighth graders, those standard errors are 0.2 and 0.3. For California students, the equivalent standard errors are 1.2, 0.9, 1.3, and 1.2. Standard errors for averages in other states are approximately the same as those for averages in California. 54 Table 4.12 NAEP Scores for Students Proficient in English, 2003 California United States Florida Illinois New York Texas Grade 4 Reading Mathematics 215 235 219 236 220 235 219 235 223 237 218 241 Grade 8 Reading Mathematics 258 274 265 278 259 273 268 279 267 282 261 279 the average of each of our four comparison states. In mathematics, California students do somewhat better. Fourth graders are slightly below the national average but are tied with fourth graders in Florida and Illinois. Eighth graders in California who are proficient in English have higher average mathematics scores than English-proficient eighth graders in Florida but fall short of all other comparison states and the national average. Because states may use different standards to determine whether a student is proficient in English, the results in Table 4.12 should be interpreted with caution. Another approach is to focus on the results of non-Hispanic white students because they are far less likely to be recent immigrants. The results for these students are similar to those for English-proficient students. The gaps between California white students and white students in other states are less than the gaps for all students, but California white students still performed poorly relative to other white students. In reading, California white students, both fourth and eighth graders, were below the national average for white students and the averages in each of our comparison states. In mathematics, California white students in the fourth grade equaled the national average for white students. In the eighth grade, however, they were below the national average and the averages of each of our comparison states. Similar comparisons could be made for other groups of students, an analysis undertaken in a recent RAND Corporation report (Carroll et al., 2005). The results of that analysis are similar to the results shown in Tables 4.11 and 4.12. When the NAEP scores of any group of 55 California students are compared with the scores of similar students from other states, California students tend to score lower on average than students in other states. The RAND study also adjusted state NAEP scores for differences among states in family characteristics such as income and parental education. Even after these adjustments, California students scored lower than students in other states. Conclusions The largest category of state and local public expenditures is elementary and secondary education. Although California’s expenditures per capita in this area were higher than the average for all other states in 2002, its schools had substantially fewer staff per student than schools in other states. There are two reasons for this outcome. First, a higher fraction of California’s total population is enrolled in public schools than in other states, spreading California’s expenditures in this area more thinly than in other states. Second, California school districts compensate their employees more generously than in other states. About half of this higher compensation can be explained by California’s labor market, where college-educated workers are paid higher salaries than in other states. Student achievement is also lower in California than in the rest of the nation. On average, California students score lower on the NAEP than students in other states. This result persists even after attempts to account for the higher percentage of California students who are recent immigrants and not yet proficient in English. A logical inference is that the relatively low resources in California schools explain the relatively low achievement of its students. The link between achievement and resources has received more attention in California recently because of the state’s new standards and accountability system. At the heart of that system is an achievement goal for every school, a goal that implies that each school’s students are expected to perform substantially better on standardized tests than students in other states (see Rose et al., 2003, and Rose, Sonstelie, and Richardson, 2004.) Although there is much academic debate about the link between resources and achievement (e.g., Hanushek, 1997, and Krueger, 2003), it is fair to ask whether California’s goals for its schools match its financial commitment to them. 56 5. Higher Education Higher education expenditures accounted for 8 percent of total state and local expenditures in California in 2002, the same share as in the rest of the nation. This chapter examines total spending for higher education in California, comparing it across states and over time. Given the significant contribution of fees for tuition, housing, and other services in the total funding picture, the chapter also includes revenues in the analysis to calculate net support (total expenditures less revenues) for higher education. Overall, California’s net support for higher education is above the average for the rest of the nation on a per capita basis. Enrollment in public colleges and universities in the state, however, is also significantly higher than in the rest of the country. The result is that the state’s net support on a per student basis is lower than in other states. The state does manage to produce more undergraduate degrees on a per capita basis than the average state, but because of its larger enrollments, it produces fewer degrees per full-time student. Total Expenditures Higher education refers to education and education services provided in degree-granting institutions beyond the high school level. The vast majority of the expenditures associated with public higher education institutions (e.g., instruction, research, public service, administration, libraries, and student services) falls under a category labeled by the U.S. Census Bureau as “other higher education” spending. The Census Bureau differentiates these expenditures from “auxiliary services” that provide supplementary services to students beyond the direct provision of education, such as dormitories, cafeterias, and bookstores. Separate designations also exist within each of these categories for capital expenditures. In addition to direct expenditures, 57 states also provide higher education subsidies to their residents in the form of student assistance. In California, relatively few programs and agencies fall under the heading of higher education. The major institutions providing higher education services consist of the University of California (UC) system, the California State University (CSU) system, and the California Community College (CCC) system. Funds to support small agencies (the California Postsecondary Education Commission, the ScholarShare Investment Board, and the California Student Aid Commission), as well as resources for state grants (Cal Grants) to California residents attending college are also included in this discussion. The expenditures and revenues associated with university-affiliated hospitals are not included. Overall, California’s higher education spending per capita in 2002 was 8 percent higher than such spending in the rest of the nation (Table 5.1). This total masks variation across individual categories. Expenditures in California for instruction, research, and administration (the Census Bureau’s “other higher education” category) were 11 percent above average spending in all other states. California spent less on auxiliary services while providing slightly more in financial aid. In contrast, student assistance in Texas was well below the national average, representing less than 5 percent of that state’s total higher education spending. In California, financial aid accounted for 8 percent of the total.1 Table 5.1 Expenditures on Higher Education per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 634 586 394 577 472 601 Other Higher Education 543 488 321 476 392 538 Auxiliary Services 39 51 26 41 25 37 Student Assistance 51 47 47 60 55 27 _____________ 1Note that not all financial aid provided by the state is spent at public institutions. The Cal Grant program, for example, provides assistance to state residents who attend a higher education institution in the state, public or private. 58 Within these categories, the Census Bureau also reports capital expenditures (not shown). Spending on capital projects in California represented less than 6 percent ($35 per capita) of total higher education resources in 2002, a level nearly one-half the size of the figure for the rest of the country ($65 per capita). The nature of capital projects, however, can make single-year comparisons of capital spending less meaningful. For this reason, a multiyear analysis of capital spending appears later in this chapter. Relative spending for higher education has changed over time. Per capita spending was 29 percent higher on average in California than in the rest of the nation between 1977 and 1987 (Figure 5.1). From the late 1980s until the early 1990s, the gap between the state and the rest of the country began to narrow. After dropping below the average for all other states in the mid-1990s, California increased its spending in the remainder of the decade, keeping pace with, and eventually outpacing, a nationwide growth in resources directed to higher education. 700 600 500 400 300 California U.S.–California 200 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.1—Higher Education Total Expenditures per Capita, 1977–2002 Higher Education Revenue Higher education differs from many of the other functional areas discussed in this report in that it has user fees, primarily in the form of 59 Expenditures (2002 $) tuition, which directly offset program expenditures. The Census Bureau provides data on two types of higher education revenue: income from auxiliary services and income from all other higher education charges, including tuition. California collected slightly less higher education revenue per capita than the rest of the nation in 2002 (Table 5.2). However, per capita revenues in Florida and New York were considerably lower than in California. Most of this difference was accounted for by significantly less money being generated from sources other than auxiliary services. These revenues, primarily tuition charges, were 40 percent lower in New York and 47 percent lower in Florida than in California. Historically, the trend for California and the nation (as well as for the other states in our comparison group) in higher education revenues has been one of general increases in real terms over the past two decades (Figure 5.2). As a group, California and the other large states considered in this report rank relatively low in higher education revenues, all generating less revenue per capita than the national average excluding California. Since the mid-1980s, Florida has consistently raised the least amount of higher education revenues. On the other hand, Texas has produced the most revenue per capita within this group for much of the past 20 years. In 2002, tuition increases across all of California’s higher education systems moved the state toward the top of our comparison group, although it still remained below the average for the rest of the nation. Table 5.2 Higher Education Revenues per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 205 215 114 196 128 196 Other Higher Education 167 163 91 155 100 166 Auxiliary Services 38 51 22 41 28 30 60 Expenditures (2002 $) 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.2—Higher Education Total Revenues per Capita, 1977–2002 Net Support for Higher Education The significant role that tuition and other revenues play in supporting higher education suggests that calculating the net support (total expenditures less revenues) allows a more meaningful comparison than simply examining total expenditures. Calculating a state’s net support for higher education rather than expenditures per capita changes the picture slightly. Net support for higher education in California was 15 percent above the average for the rest of the nation in 2002, the highest among our comparison states (Table 5.3). This adjustment has an even greater effect on relative rankings for other states. For example, Illinois shifts from a position slightly below the national average in expenditures per capita to one slightly above average in net support per capita. Similarly, New York’s total expenditures per capita were 20 percent below the average for the United States excluding California in 2002. However, after adjusting for New York’s relatively low rate of revenues generated by tuition and fees, the state rises to within 8 percent of that figure. 61 Table 5.3 Higher Education Net Support per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 429 372 280 380 343 405 Other Higher Education 428 372 277 381 347 399 Auxiliary Services 1 — 4 (1) (4) 7 Although state governments do not “make money” on higher education services (i.e., they do not generate more revenue than they spend), some appear to be able to use net income from their auxiliary services to offset other expenditures. About half of the states, including Illinois and New York, reported receiving more revenue from auxiliary services than they spent on those activities. It is also worth noting that this idea of cross-subsidizing extends to different types of instruction and students. Graduate students, in general, consume a greater share of resources than undergraduates do. Similarly, the actual cost of instructing nursing and biology majors is likely to be higher than the expenditures associated with students studying literature or philosophy. This study, which aggregates data to make comparisons, necessarily masks such differences. Net Support per Student As in all functional areas, comparing net support relative to a state’s population provides some sense of the commitment the state’s taxpayers have made to the program. An examination of expenditures relative to the total number of beneficiaries gives some indication of how those resources are being distributed. Similar to enrollment in K–12 discussed in Chapter 4, enrollment in state colleges and universities represents the caseload for higher education spending. Unlike K–12 education, however, many higher education students participate part-time. Thus, it becomes necessary to convert the number of students enrolled to full-time-equivalent students 62 (FTES) to compare the level of resources committed to higher education relative to the number of students. Using the approach outlined in Chapter 2, the relationship between expenditures, population, and enrollment (caseload) is described by Eq. (5.1) Expenditures = FTES  Expenditures Population Population FTES (5.1) Using enrollment data from the National Center for Education Statistics (NCES), Table 5.4 compares net support for higher education for each full-time-equivalent student in public institutions. Relative to comparisons on a per capita basis, normalizing support for higher education on a per student basis presents a different picture. Net support for students in California was 7 percent below the average for the rest of the country in 2002, whereas three of our four comparison states supported students at a higher level than this average. California’s relatively high participation rate in higher education explains much of this shift. Although the state’s spending per capita is higher than average, the state’s relatively high rate of student participation offsets that difference. Illinois, for example, spends less than California per capita but 14 percent more than California per full-time student. This change is a result of the fact that Illinois enrolls about three-quarters as many students as California. An alternative way of looking at this would be to assume that enrollment rates in public higher education were the same in California Table 5.4 Net Support per Full-Time-Equivalent Student, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures Revenue Net Support Net Support per Capita per Capita per Capita 11,329 634 205 429 12,177 586 215 372 11,268 394 114 280 12,958 577 196 380 14,276 472 128 343 12,550 601 196 405 FTES per Capita 0.038 0.031 0.025 0.029 0.024 0.032 63 as in the rest of the nation (i.e., 31 rather than 38 FTES per 1,000 residents). At this lower rate of participation, net support per FTES would increase to $13,836, or 14 percent higher than the average for all other states. California’s high participation rate in higher education has been consistent through the years, dating back more than three decades. By contrast, the funding picture has changed in recent years. Net support for higher education per full-time student in California has been near or above the average for all other states during much of the past decade (Figure 5.3). It is only in the last year for which comparable data are available, 2002, that a significant gap emerged between this state and the rest of the country. The appearance of this gap can be explained by modest expenditure growth, which did not keep pace with growing enrollments. The end result was declining net support per full-time student. Aggregating the data for all California public higher education institutions masks significant differences in the average expenditures per student across the three California systems. Although the Census data do not permit further distinctions by system, figures from the California Postsecondary Education Commission (CPEC) provide some sense of 15,000 Net support per FTES (2002 $) 10,000 5,000 California U.S.–California 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.3—Higher Education Net Support per Full-Time-Equivalent Student, 1991–2002 64 this variation. In 2002, CPEC reported that general purpose expenditures per FTES were $20,241 in the UC system, $10,407 in the CSU system, and $4,629 in the CCC System (California Postsecondary Education Commission, 2004). Participation and Student Fees One explanation for California’s relatively high rate of participation in higher education could be that this state includes a disproportionately high number of residents of traditional college age (i.e., ages 18–24) relative to other states. Within our comparison group, California does include more residents of traditional college age than any state except Texas. However, this is not the whole story. Instead, more older as well as traditional-age students are enrolled in higher education in California. For example, in 2002, 44 percent of postsecondary students in California were over age 25, compared to 37 percent in the rest of the nation and similar shares in all of our comparison states. The exception was Illinois, where older students represented 43 percent of total enrollment. Illinois, not coincidently, was also second only to California within our comparison group in its community college enrollment (see Figure 5.7 later in this chapter). California’s relatively high enrollment levels are not entirely explained by the rate of participation by older students. So-called traditional age students, ages 18 to 24, appear to enroll in higher education at a much higher rate than one finds in other parts of the country. Although the data do not allow a perfect comparison, it is possible to get an estimate of the level of participation by comparing the number of students enrolled in public education who are under age 25 with the Census Bureau’s population estimate for 18 to 24 year olds in each state. Using this approximation, one out of three residents 18 to 24 years old is enrolled in a public higher education institution in California. This number compares with an estimate of one out of five 18 to 24 year olds participating in public higher education in the rest of the country (Figure 5.4). Estimates for our comparison states fall between these two numbers. 65 0.05 0.04 Enrollment rate 0.03 0.02 0.01 0 California U.S.– California Florida Illinois New York Texas Figure 5.4—Percentage of Population Ages 18 to 24 Enrolled in Public Higher Education, 2002 Although the estimates in Figure 5.4 do not account for nonresident enrollments, Californians clearly prefer to “stay at home” when it comes to higher education, and residents represent the vast majority of students enrolled at the public institutions. First-year California residents who are enrolling in a college or university are more likely than first-year students in the rest of the country to stay within the state (Table 5.5). Table 5.5 Residence and Migration of First-Year Undergraduate Students, 2002 (percent) California U.S.–California Florida Illinois New York Texas Percentage of Students Who Are State Residents 0.92 0.81 0.81 0.90 0.81 0.93 Percentage of State Residents Enrolled in State Institutions 0.92 0.83 0.90 0.83 0.83 0.92 66 Higher education institutions within California enroll a relatively higher share of state residents (92%) than the average for all other states (81%). Texas exhibits a similar pattern. However, there appears to be no consistent pattern with regard to migration among the other comparison states. Illinois’ institutions enroll a higher than average number of state residents, but students from that state are just as likely as students in the rest of the country to look beyond their borders to attend a college or university. Florida exhibits the opposite pattern. Students in Florida are more likely to stay at home for their undergraduate education, but the capacity of those schools enables that state to enroll a larger share of nonresident students. California also enrolls a relatively large share of students at public schools rather than private ones. In this state, public two- and four-year schools account for 81 percent of the total number of students in higher education. The national average in academic year 2002–2003 was 72 percent. Again, Texas was similar to California, with 85 percent of its students in public schools. New York, in contrast, enrolled only 52 percent of students in public institutions. Florida (74%) and Illinois (67%) fell between these two extremes, on either side of the national average. In sum, more Californians than residents of other states attend higher education institutions, and when they do, they are more likely to attend public schools within their home state’s borders. Relatively low fees likely contribute to this higher rate of participation because they make higher education more attractive and more accessible. From one perspective, this increased accessibility is a positive development. Low fees, however, also generate lower revenues, which must be spread across a larger student population. In 2002, for example, higher education revenues per FTES (including both tuition and auxiliary revenues) were $5,404 in California, more than 20 percent lower than the national figure of $7,027 per FTES. Had California generated revenue per FTES at the same rate as the rest of the country, net support per student would have been $12,952 or 6 percent higher than the average for all other states. Generating less revenue from student fees is hardly a new development. California has had a tradition of low tuition for its 67 colleges and universities. In fact, enrollment in a California community college was free as recently as 1985. In the latter half of the 1990s fees were held constant, resulting in a real decline in revenues per student during the early part of that period (Figure 5.5). Despite shifts in fee policies, the difference between the national figure and the state’s revenues per FTES has been relatively stable, averaging 26 percent. Differing levels of support from state and local governments, then, have driven much of the fluctuation in net support for higher education per student relative to the average for all other states. Figure 5.6 compares the variability in higher education expenditures per FTES in California during the 1990s to the relative stability of revenues per FTES over this period. Again, using aggregate data does obscure differences across the three California systems. CPEC reports average revenue per FTES for instruction-related activities, which provides some sense of the variation across the three systems. For UC campuses, this figure was $3,706 in 2002. Comparable figures for the CSU and CCC systems were $1,610 and $149, respectively (California Postsecondary Education Commission, 2004). The pressure of budget deficits led to significant 8,000 Revenues per FTES (2002 $) 6,000 4,000 California U.S.–California 2,000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.5—Higher Education Revenues per Full-Time-Equivalent Student, 1991–2002 68 Revenues and expenditures per FTES (2002 $) 20,000 15,000 10,000 Revenues Expenditures 5,000 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.6—Higher Education Revenues and Expenditures per Full-TimeEquivalent Student, 1991–2002 fee increases in the years after 2002. CPEC estimated that by 2004, average revenue per FTES for instruction-related activities would have risen by approximately 30 percent relative to the 2002 figures. Staffing and Salaries Most of the cost of providing higher education services stems from the expenses associated with personnel. Using employment and payroll figures reported to the Census Bureau, it is possible to calculate ratios similar to those in Eq. (2.3) in Chapter 2. Total Payroll = Total Payroll  Total Employees FTES Total Employees FTES (5.2) The story in Table 5.6 is similar to the decomposition presented in Chapter 4 for primary and secondary education. In California, payroll expenditures per employee are above the average for the rest of the country but compensation per FTES is near average. This implies that the ratio of employees per full-time-equivalent student is lower in California’s public colleges and universities than in the rest of the country. Although below the average for the nation excluding 69 Table 5.6 Total Payroll, Costs per Employee, and Number of Employees per Full-Time-Equivalent Student, 2002 California U.S.–California Florida Illinois New York Texas Total Payroll Costs per FTES ($) 8,513 8,786 7,888 8,575 7,295 8,010 Payroll Costs per Employee ($) 31,148 27,970 28,869 26,363 30,665 29,858 Total Employees per FTES 0.273 0.314 0.273 0.325 0.238 0.268 California, this ratio is similar to that found in the other comparison states, with the exception of Illinois. Higher compensation costs per employee also translate into payroll accounting for a larger share of the total cost of higher education. In California, payroll costs represent 75 percent of net support per FTES (Tables 5.4 and 5.6). This figure is slightly higher than the average of other states (72%) and significantly higher than comparable percentages in most of our comparison states (70% of net support in Florida, 66% in Illinois, 64% in Texas, and 51% in New York). Comparing total payroll costs per employee could conceal differences in the use of part-time and full-time employees. Different shares of employees may be involved in noninstructional activities, also affecting payroll costs. In California, however, regardless of how the payroll figures are disaggregated, the average wages for both full- and part-time employees are higher than in the rest of the country (Table 5.7). The average salary for full-time instructional employees provides an approximation of what different states pay their full-time faculty. Table 5.8 provides salary comparison figures that suggest that the relatively high full-time faculty salaries are driven by the state’s labor market. In occupational Zone 5, where employees are typically required to hold an advanced degree, average salaries in California are 12 percent higher (see Chapter 2). Assuming that this is the comparable group for full-time 70 Table 5.7 Wages for Full- and Part-Time Instructional and Noninstructional Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Full-Time Full-Time Instructional Noninstructional Employee Salary Employee Salary 73,393 51,793 65,428 37,807 69,607 34,531 66,074 38,431 65,946 42,489 68,910 37,941 Part-Time Part-Time Instructional Noninstructional Hourly Wage Hourly Wage 26 15 20 10 21 10 20 8 21 14 22 10 Table 5.8 Full-Time Instructional Employee Salary Comparisons, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Payroll per Full-Time Employee 73,393 65,428 69,607 66,074 65,946 68,910 Adjusted Payroll per Full-Time Employee 65,502 66,437 70,556 68,048 57,260 69,693 public college and university instructional employees, labor market conditions would explain the variation in faculty salaries in the state. In fact, when full-time wages are adjusted using the location premium for Zone 5, salaries in California higher education institutions actually fall slightly below the average for the rest of the nation and most of our comparison states. Only New York reports lower adjusted payroll figures. In addition to the number of employees available to teach and assist students, different varieties of educational experiences may be provided by different types of institutions. A second indicator of the type of services provided in California relative to other states is the mix of students enrolled at two- and four-year institutions. Of the more than 1.3 million full-time-equivalent students attending public higher 71 education institutions in the state, only 40 percent of them were enrolled in four-year institutions in 2002 (Figure 5.7). This figure was the lowest in the nation. For the rest of the country, the average share of FTES at four-year schools was 64 percent. California, therefore, relies heavily on its community colleges for the delivery of its higher education services. In fact, it is one of only four states in the country where a majority of FTES was served by two-year institutions. (Wyoming, Illinois, and Washington are the other three.) This observation is not intended to suggest that one model is preferred to another. Nevertheless, it is the case that the higher education activities at four-year institutions are not the same as those at two-year colleges. Community colleges, for example, typically are not engaged in research to the degree found on the University of California campuses. Alternatively, lower division classes in the CCC system often are smaller than those found on a UC campus. What is clear is that public support for higher education is providing a different mix of services. In Florida, for example, net support for higher education was slightly lower than in California ($11,268 versus $11,329 per FTES) in 2002. Over three-fifths of the full-time- 0.6 FTES share at two-year institutions 0.5 0.4 0.3 0.2 0.1 0 California U.S.– California Florida Illinois New York Texas Figure 5.7—Percentage of Higher Education Enrollment at Two-Year Institutions, 2002 72 equivalent students in Florida, however, were attending four-year universities as opposed to only two out of five FTES in California. It is difficult to disentangle the cause and effect of participation rates and the structure of higher education in the state. One consequence of having an extensive community college system in California may be that the state ranks among the highest in the country in terms of participation in higher education, as noted above. On the other hand, one could argue that the high demand for higher education services requires that the state maintain and expand all of its colleges and universities. Capital Expenditures As noted already, California capital expenditures per capita on higher education were significantly lower than the average for the rest of the country in 2002. Historically, this has been the case since the mid1980s (Figure 5.8). When one considers that California tends to enroll students at a higher rate than the average for the rest of the nation, the comparison of capital investment in any one year becomes even less favorable. For example, expenditures per capita in 2002 were 46 percent below the average for all other states. Capital expenditures per FTES, however, were 57 percent below the average for the rest of the nation ($924 versus $2,127 per FTES). The uneven nature of capital investment, however, renders singleyear comparisons of expenditures less useful. Assuming that a portion of new capital spending is in response to increased demand for services, it also makes sense to examine expenditures relative to marginal increases in enrollment. Table 5.9 compares capital expenditures and higher education enrollment changes over the period 1992 to 2002. Illinois and New York, two states with modest growth in their higher education systems over this period, demonstrated the highest rate of investment in capital projects relative to the number of new students. These resources, most likely, were devoted to maintaining and upgrading existing facilities. Public higher education enrollment in the other three comparison states grew much more rapidly. California (22%), Texas (24%), and Florida (29%) all experienced significant growth. Of these three, California’s expenditures per marginal student were the lowest. 73 Expenditures (2002 $) 80 70 60 50 40 30 20 California 10 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.8—Higher Education Capital Expenditures per Capita, 1977–2002 Table 5.9 Increases in Enrollment and Real Capital Outlays, 1992–2002 California U.S.–California Florida Illinois New York Texas Increases in Enrollment (FTES) 235,407 949,140 92,880 5,043 15,022 138,674 Real Capital Outlay (2002 $) 15,709,674,024 134,099,860,961 7,337,850,760 6,139,377,670 8,489,813,146 10,392,607,350 Real Capital Outlay per Additional FTES (2002 $) 66,734 141,286 79,004 1,217,406 565,159 74,943 Outcomes Identifying appropriate outcome measures in education is difficult. The comparison of standardized test scores as a proxy for educational achievement, as presented in Chapter 4, provides a qualified measure at best. In higher education, however, the equivalent of standardized 74 testing does not even exist. Output measures such as degrees conferred, then, serve as a rough approximation for assessing outcomes. California’s investment in higher education produces undergraduate degrees at a higher rate than our comparison states in most cases but a lower level of graduate degrees (Table 5.10). Only Florida’s colleges and universities awarded more associate’s degrees on a per capita basis in academic year 2002–2003. For its population, California also produces more bachelor’s degrees than all of our comparison states and it is only slightly below the average for the rest of the nation. The comparison becomes less favorable when one examines the number of graduate and professional degrees produced by higher education institutions in California. California, given its population, produced a low number of graduates with master’s degrees, Ph.D.s, and professional degrees when compared to the average for all other states. Completion of degrees is a crude measure of output and does not capture other contributions of higher education. It is possible, however, to combine data on degrees, enrollments, and net support to capture a rough approximation of efficiency. By these measures, California produces a relatively low level of degrees given its enrollment (Table 5.11). The figure of 0.15 per FTES is below the average for the rest of the country and all of our comparison states. When one examines the cost of conferring these degrees, California appears to be particularly Table 5.10 Degrees Conferred at Public Colleges and Universities per 1,000 State Residents, Academic Year 2002–2003 California U.S.–California Florida Illinois New York Texas Associate’s Bachelor’s Degrees Degrees per 1,000 per 1,000 Residents Residents 2.10 2.97 1.67 3.04 2.46 2.83 1.83 2.58 2.01 2.29 1.40 2.93 Master’s Degrees per 1,000 Residents 0.63 0.96 0.70 0.89 0.85 0.97 Ph.D.s per 1,000 Residents 0.08 0.10 0.08 0.09 0.06 0.10 Professional Degrees per 1,000 Residents 0.07 0.12 0.08 0.09 0.06 0.14 75 Table 5.11 Degrees Conferred per Full-Time-Equivalent Student and Net Support per Degree, 2002 California U.S.–California Florida Illinois New York Texas Total Degrees Conferred Total FTES 204,711 1,324,686 1,491,827 7,722,706 102,648 415,071 68,886 369,251 101,085 460,507 120,443 701,793 Degrees per FTES 0.15 0.19 0.25 0.19 0.22 0.17 Net Support per FTES 11,329 12,177 11,268 12,958 14,276 12,550 Net Support per Degree Conferred 73,309 63,034 45,563 69,458 65,036 73,127 unimpressive. By this calculation, the state provided over $73,000 for each degree produced in 2002. This figure is 16 percent higher than the average for the nation as a whole excluding California and the highest of our comparison states. In terms of both measures, the ideal analysis would disaggregate production and costs associated with associate’s, bachelor’s, and graduate degrees. Unfortunately, the data do not permit such a decomposition. However, it is possible to compare the number of associate’s degrees conferred relative to total enrollment in two-year institutions and the sum of bachelor’s and graduate degrees divided by enrollment in fouryear schools. Such a calculation suggests that California’s four-year universities produce bachelor’s and graduate degrees at a rate that exceeds the average for all other states and most of our comparison states (Table 5.12). The community colleges, however, lag behind the rest of the nation by this measure. California is significantly lower than the average for the rest of the country (by 38%) in terms of the number of associate’s degrees conferred per FTES enrolled in the community college system. Definitive interpretation of the data in Table 5.12 is difficult. For example, the mission of the community college system in California includes activities well beyond just conferring associate’s degrees. Proficiency in preparing students for transfer to four-year universities 76 Table 5.12 Estimated Degrees Conferred per Full-Time-Equivalent Student at Two- and Four-Year Institutions, 2002 California U.S.–California Florida Illinois New York Texas Associate’s Degrees per FTES at Two-Year Schools 0.09 0.15 0.26 0.12 0.21 0.10 Bachelor’s and Graduate Degrees per FTES at Four- Year Schools 0.25 0.22 0.24 0.27 0.22 0.23 would not be captured by such a measure. Two-year colleges in other states, however, presumably also are tasked with multifaceted missions. An additional qualification of these estimates stems from the fact that they do not take into account student preparation. If students enter colleges and universities at different levels of proficiency, the subsequent degree “production” could be affected. Chapter 4 notes that there is variation between the states in the level of achievement on standardized tests. Such disparities could affect the success of students once they enter higher education. Conclusions Participation in higher education in California is higher than the average for the rest of the nation and higher than in the other large states used in this analysis. This fact drives the conclusions regarding support for higher education on a number of fronts. On a per capita basis, the state’s total higher education expenditures are higher than average. Because of California’s relatively low student fees and other revenues to higher education institutions, net support per capita for higher education is even higher (15%) than the average for the rest of the country. California’s colleges and universities also confer more associate’s and bachelor’s degrees per capita than the rest of the nation. 77 Higher relative enrollments change the picture. Net support per student in California is 7 percent below the average for all other states. The production of graduate and professional degrees is also below average relative to the state’s student population. And the costs associated with producing any type of degree are relatively high. One element that does not change with regard to comparisons on a per capita or enrollment basis is California’s expenditures for higher education capital projects. Regardless of the basis, the state has consistently spent less in this area. Finally, California’s labor market contributes to higher salaries for employees in the state’s colleges and universities. This factor, combined with less net support per student, results in a higher ratio of students to employees in the state’s schools. In sum, higher education in the state presents a mixed picture. Californians invest more than residents of other states in higher education per capita. They also charge relatively low fees to students, making higher education more accessible and leading to higher than average rates of participation. In return, they get more undergraduate degrees than other states per capita. On the other hand, because of high participation rates, California’s higher education system has fewer resources per student, lagging capital investment, and fewer employees on campuses and it produces fewer graduate and professional degrees relative to the state’s population. Finally, it appears that the state invests more resources per degree produced than states in the rest of the country. 78 6. Health Services Health services is the second-largest category of state and local government expenditures. In 2002, health services constituted 14 percent of California’s spending, compared to 17 percent in the rest of the nation. This category covers a wide array of health-related services, including the provision of medical care for low-income individuals, the construction and operation of public hospitals, and public health services such as mental health and alcohol or drug rehabilitation treatment as well as some environmental protection activities. This chapter focuses on the largest state-administered health program, Medicaid (known as Medi-Cal in California) and examines spending differences between California and other states on this program. It also analyzes expenditures on public hospitals and other health services. California is similar to the rest of the nation in total governmentfinanced health expenditures per capita. However, it allocates these expenditures differently, spending less on Medi-Cal and more on other health services and subsidies to public hospitals. California is unusual in that it enrolls a higher percentage of the population in Medi-Cal and spends considerably less per recipient than the other large states in our comparison group. Several factors contribute to this outcome. First, eligibility rules are more generous in California than in other states and, partly as a result of these rules, the Medicaid caseload mix is younger in California than in other states. Second, Medi-Cal reimbursement rates under the traditional fee-for-service (FFS) system are lower than the national average excluding California. Third, more Medi-Cal recipients are enrolled in managed care plans, and capitated payments per enrollee under these plans are substantially lower in California than in other states. Although it is difficult to link public spending on health to specific outcomes, it is interesting to note that despite lower Medi-Cal spending 79 per capita, fewer Californians are hospitalized as a result of delayed medical intervention than are the residents of other large states. However, more low-income Californians than low-income residents in other large states report poor access to health care. Total Expenditures State and local governments engage in various activities to conserve and improve public health and make basic medical care available to lowincome or medically needy residents. The U.S. Census Bureau summarizes state and local expenditures on health services in four government functions: vendor payments for medical care, own public hospitals, other hospitals, and other health. Vendor payments for medical care include payments from state and local governments to private hospitals and health care providers under federal or state welfare programs such as Medicaid, general relief, and public assistance. In California, this function includes payments made on behalf of programs such as Medi-Cal, Healthy Families, Access for Infants and Mothers, and Major Risk Medical Insurance. The second function, own public hospitals, includes all state and local expenditures on the acquisition, construction, and operation of government-administered hospitals. It also comprises payments from the state to these hospitals on behalf of public medical assistance programs such as Medi-Cal. In California, these hospitals include county hospitals, developmental centers for persons with developmental disabilities, state hospitals for the mentally disabled, and hospitals affiliated with public universities, such as the UCLA Medical Center. They also include public children’s, maternity, and orthopedic hospitals that are operated by counties or public universities. The other hospitals function comprises state and local expenditures on private hospitals. Examples include government payments to private corporations that lease and operate government-owned hospitals, and government payments for the construction of hospitals to be leased or turned over to private operators. In California, as in many of our comparison states, these expenditures are negligible and will not be considered here. 80 The fourth expenditure function, other health, includes all state and local health activities unrelated to public medical assistance programs or the provision of hospital care. In California, this function is primarily composed of local expenditures (including those out of state-local realignment revenues) on county health services (including Healthy Kids), primary care and family health services, and drug and alcohol abuse services. This category also covers state and local spending on public health, broadly construed to include programs outside the Department of Health Services such as air quality control, pesticide regulation, and toxic substance control. California spends about the same amount per capita on total health services as the rest of the country, although it allocates these expenditures differently across the four categories described above. In 2002, total health expenditures per capita were $1,173 in California, on par with the national average excluding California and slightly more than Florida or Texas (Table 6.1). A little more than one-third of these expenditures were local (including local expenditures of state funds). New York spent 59 percent more and Illinois spent 27 percent less than California. Compared to the rest of the country, California spent less per capita on vendor payments for medical care, more on own public hospitals, and more on other health activities. Expenditures on other hospitals were negligible in California and in all of our comparison states except New York, which spent $26 per capita on this category. Table 6.1 Health Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 1,173 1,175 1,044 856 1,863 1,009 Medical Vendor Payments 574 679 594 443 1,184 544 Own Public Hospitals 319 298 257 169 454 341 Other Hospitals 0 3 1 7 26 0 Other Health 280 195 191 238 199 124 81 California has been on par with the rest of the nation in health expenditures throughout recent years. This picture stands in contrast to the late 1970s, when California spent more per capita on health services than the rest of the nation (Figure 6.1). However, as the rest of the country experienced more rapid growth in health spending, the expenditure gap between California and other states narrowed in the 1980s and disappeared by 1990. Health expenditures grew rapidly in both California and the rest of the nation in the early 1990s due mainly to an escalation of payments to Medicaid Disproportionate Share Hospitals (DSH) serving a large number of Medicaid patients and uninsured individuals. This growth flattened out in the late 1990s as a result of federal actions. Health expenditures rose again after 2000 in both California and the rest of the nation. State Medicaid programs constitute more than half of state and local health expenditures in California and the rest of the country. We therefore devote the next section to an analysis of spending on this program in California and other large states. We then briefly examine government subsidies to publicly owned hospitals and state expenditures on other health activities. Because Census of Governments data are not 1,400 1,200 1,000 800 600 400 California 200 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.1—Total Health Expenditures per Capita, 1977–2002 82 Expenditures (2002 $) designed to study individual programs, our analysis of Medicaid expenditures relies on data obtained from the Centers for Medicare and Medicaid Services (CMS). Medicaid Medicaid is the largest state-run health care program in every state. It pays for health care and long-term care for certain members of the low-income population, including families with children, pregnant women, individuals with disabilities, and the elderly. Medicaid is distinct from Medicare, the federal health insurance program for elderly or disabled persons. However, certain low-income Medicare beneficiaries may be entitled to Medicaid benefits for services not covered under Medicare (e.g., nursing home services and prescription drugs before January 2006) or payments of insurance premiums, deductibles, and co-payments under Medicare. Medicaid is jointly funded with federal and state revenues. The federal government matches state expenditures on Medicaid according to a formula that is inversely related to state income per capita. In 2002, California received an average of $1.06 from the federal government for every dollar it spent on Medi-Cal, similar to Florida and New York, which each received one federal dollar for each state dollar spent on Medicaid. By contrast, Florida received $1.30 for every dollar spent on Medicaid and Texas received $1.51. The comparisons that follow include both state and federal funds. To receive matching federal funds under Medicaid, states must provide a minimum set of “mandatory” services, including physician services, inpatient and outpatient hospital care, laboratory and x-ray services, nursing facilities, and home health care. Although these services are called “mandatory,” states have discretion over the amount, duration, and scope of services covered. For example, some states limit the number of nonemergency hospital visits per year. Beyond these mandated benefits, states can also provide “optional services” such as prescription drugs, clinic services, prosthetic devices, hearing aids, dental care, and long-term and personal care services for the elderly or disabled. In addition, states may apply for waivers from federal program requirements to provide services to individuals not traditionally 83 eligible for Medicaid or to provide additional benefits to certain groups. For example, the home- and community-based services (HCBS) waiver enables states to provide noninstitutional long-term care services to frail elderly and disabled individuals.1 To be eligible for federal funds, states must provide Medicaid coverage to certain “categorically eligible” groups. These groups include recipients of aid from federal programs such as Supplemental Security Income (SSI), those who would have formerly been eligible for Aid to Families with Dependent Children (AFDC), low-income children, pregnant women, parents of dependent children, and certain elderly individuals. In addition, states can expand coverage to the “medically needy” and other special groups who are not otherwise eligible for Medicaid. Income criteria for these groups can vary substantially by state and these differences affect the proportion of the low-income population enrolled in Medicaid in each state. Since 1997, states have had the option of expanding health insurance for children not otherwise eligible for Medicare under the State Children’s Health Insurance Program (SCHIP). As of 2002, 20 states (including New York and Texas) did so through a separate program, 15 states expanded eligibility for these groups within Medicaid, and the remaining states combined both options. Among our comparison states, California, Florida, and Illinois were in the latter group. Medicaid is a “vendor payment” program that makes payment to health care providers on behalf of program recipients. States have the flexibility to purchase covered services from health care providers on a fee-for-service basis or from managed care plans on a per enrollee, or capitation, basis. Under both systems, states negotiate reimbursement rates with providers. Federal Medicaid law does not establish specific floors or ceilings for provider payments. Reimbursement rates vary widely across states, although most Medicaid programs reimburse providers at a rate less than the full rate allowed by the federal government. _____________ 1Some HCBS services overlap with existing “optional services” such as personal care services. Other HCBS services, such as adult day health services and respite care, are exclusively available under the waiver program. 84 Compared to the average for all other states, California has spent consistently less per capita on Medicaid since the 1990s. In 2002, MediCal expenditures were $667 per capita, 8 percent less than the average of all other states (Table 6.2). New York spent more than twice as much per capita as California, whereas Illinois was very close to the average for the nation excluding California. Florida and Texas spent 12 and 23 percent less than California, respectively. California’s nearly average Medicaid expenditures per capita mask dramatic differences compared to the rest of the nation. Using the general formula from Chapter 2, we can decompose Medicaid expenditures per capita for a given year into the product of the number of individuals who received services in that year and spending per recipient: Medicaid Expenditures = Medcaid Recipients Population Population  Medicaid Expenditures Medcaid Recipient (6.1) Medicaid expenditures per recipient are substantially lower in California than in the rest of the United States or in any of our comparison states. In 2002, California spent $3,113 per Medicaid recipient, 36 percent less than the average for the rest of the nation (Table 6.2). Each of our comparison states spent at least 20 percent Table 6.2 Components of Medicaid Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 667 726 589 721 1,644 512 Expenditures per Recipient ($) 3,113 4,878 3,726 5,393 8,101 3,774 Recipients per Capita 0.21 0.15 0.16 0.13 0.20 0.14 85 more per recipient than California, with New York spending two and a half times as much. At the same time, Medicaid enrollment was higher in California compared to the average for all other states. Twenty-one percent of Californians were enrolled in Medicaid in 2002, compared to an average enrollment rate of 15 percent in the rest of the country. New York enrolled 20 percent of the population in Medicaid in 2002, the secondhighest enrollment within our comparison group. The high Medicaid enrollment rate in California relative to other large states could result from differences in the generosity of eligibility policies and behavioral responses to these policies or to differences in the economic and demographic characteristics of each state. For example, relative to some other states in our comparison group, California includes more children and fewer disabled residents as a share of the total population (Table 6.3). To understand the interaction of these factors, we can analyze the participation rate for each eligibility group separately. For each group, the participation rate is the poverty rate for that group multiplied by the rate at which low-income members of the group enroll in Medicaid: Medicaid Recipients = Low  Income Population Population Population  Medicaid Recipients Low  Income Population (6.2) Table 6.3 Composition of State Population, by Eligibility Groups, 2002 (percent) California U.S.–California Florida Illinois New York Texas Child 30 28 26 29 27 31 Nonelderly Adult 51 50 48 52 52 50 Elderly Adult 6 7 10 7 8 5 Disabled Adult 13 15 16 13 14 13 86 In general, financial eligibility criteria for Medicaid are more generous for children than for nonelderly, elderly, and disabled adults. However, there is variation among states in these criteria. We measure the low-income population as the number of residents living below 200 percent of FPL ($30,040 for a family of three in 2002) for nonelderly adults as well as for elderly and disabled adults, and 300 percent of FPL ($45,060 in 2002) for children. For most states, these are the most generous income criteria under which individuals in these eligibility groups may qualify for Medicaid. These measures therefore ensure that all who may be potentially eligible for Medicaid are included in the low-income population estimates for each group. In 2002, California enrolled 11 percent of nonelderly adults into Medicaid, about twice the average enrollment rate for this group in the rest of the country. California’s enrollment rate for nonelderly adults was also more than twice that of Florida, Illinois, or Texas. New York was a close second among our comparison states, with an enrollment rate of 10 percent for this eligibility group (Table 6.4). California’s high enrollment rate for nonelderly adults in part reflects that California includes more low-income residents in this eligibility group than other states. More than one-quarter of nonelderly adults in California had family incomes below 200 percent of FPL in 2002, compared to 22 percent of adults in the rest of the nation and similar shares in New York and Illinois (Table 6.4). On the other hand, the Table 6.4 Components of Medicaid Recipients per Nonelderly Adult, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Nonelderly Adults 11 6 5 5 10 4 Share of LowIncome Enrollment Rate Nonelderly for Low-Income Adults Nonelderly Adults 26 45 22 26 28 19 20 26 22 47 30 15 87 proportion of adults with incomes below 200 percent of FPL was even higher in Florida and Texas than in California in 2002. Although California is in the middle of our comparison group in terms of the percentage of nonelderly adults below 200 percent of FPL, it had many more Medicaid recipients within this group. Forty-five percent of low-income nonelderly adults were enrolled in Medicaid in 2002, second only to New York with 47 percent. By contrast, the enrollment rate among low-income adults was 26 percent or less in the rest of the nation and even lower in our other comparison states. In general, higher participation rates within a target group could result from more generous eligibility rules or higher participation or “take-up” rates. However, there is reason to believe that in California and New York it stems from more generous eligibility criteria. For example, in 2002 pregnant women with family incomes up to 200 percent of FPL were eligible for Medicaid in California, New York, and Illinois, whereas the income cutoffs were 185 percent of FPL in Florida and Texas. In addition, California covered working parents with family incomes up to 107 percent of FPL under Medi-Cal, although the federal government mandated coverage only of single-parent or two-parent families in which the principal earner was unemployed. New York was even more generous, covering all adults with children and family incomes up to 150 percent of FPL as well as adults without children and family incomes up to 100 percent of FPL.2 The story is somewhat different for elderly and disabled Medicaid recipients. Like New York, California enrolls many elderly and disabled individuals (Table 6.5). However, its poverty rate for this group is lower than in other states. These two facts imply that California enrolls a higher proportion of low-income elderly and disabled adults in Medicaid than does the rest of the nation. The high enrollment rate for this target group is again largely explained by generous eligibility rules. Medi-Cal provided full coverage _____________ 2We also analyzed the enrollment rates of adults with and without children separately. The basic findings on the effects of the income distribution and eligibility rules remain the same. 88 Table 6.5 Components of Medicaid Recipients per Elderly or Disabled Adult, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Elderly or Disabled Adults 24 18 15 16 25 15 Share of LowIncome Elderly or Disabled Adults 35 38 37 35 38 41 Enrollment Rate for Low-Income Elderly or Disabled Adults 70 47 42 45 65 36 for these individuals with family income up to 100 percent of FPL, whereas other states applied the income criteria for the federal SSI program. California also set income criteria for Medicaid and Medicare “dual eligibles” at the upper limit of federal mandates.3 As described above, low-income children may receive health insurance coverage through Medicaid, SCHIP, or some combination of these programs. For example, California operates both a Medicaid expansion SCHIP and a separate SCHIP program (Healthy Families) for low-income children not covered by Medicaid. Thus, we analyze the enrollment rate for children in both of these programs. In 2002, California’s Medicaid enrollment rate for all children was slightly higher than the average for the rest of the country but lower than that of Florida, Texas, or New York (Table 6.6). Compared to Florida and Texas, California had fewer children living in families with incomes below 300 percent of FPL. However, its proportion was higher than in New York or Illinois and higher than the average for the rest of the nation. _____________ 3Elderly and disabled groups are combined in this analysis mainly because of data restrictions. In general, individuals who could qualify for Medicaid through either category are more likely to be assigned to the elderly category because it is less complicated to verify this status than disability. This practice could artificially raise the enrollment rate of the elderly while suppressing that of disabled individuals. 89 Table 6.6 Components of Medicaid Recipients per Child, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Children 35 33 39 28 47 37 Share of Low-Income Children 58 56 61 52 54 62 Enrollment Rate for Low-Income Children 61 58 63 53 87 59 California also enrolled more low-income children in Medi-Cal and SCHIP than did the rest of the nation, although it enrolled a similar proportion as Florida and Texas. Within our comparison group, New York had the highest enrollment rate as a proportion of the target group, with 87 percent of low-income children covered under Medicaid or SCHIP in 2002. As with other eligibility groups, high enrollment rates for lowincome children in California may result from eligibility rules or take-up rates. A review of eligibility rules suggests that these policies are at work. Both New York and California covered children in families with incomes up to 250 percent of FPL whereas income cutoffs were 200 percent of FPL in Florida and Texas and 185 percent of FPL in Illinois. Not all Medicaid enrollees are eligible for the full range of Medicaid benefits. Individuals not fully insured by Medicaid include undocumented immigrants receiving emergency aid and elderly people for whom Medicaid pays a Medicare premium. These individuals accounted for a nontrivial proportion of Medicaid enrollment in California. Excluding them, the gap in enrollment rates between California and other states is not as stark. In an average month of 2001, 15 percent of Californians were enrolled in Medicaid or SCHIP and eligible for full benefits, lower than the 17 percent enrollment rate in New York although still higher than 90 Table 6.7 Components of Medicaid and SCHIP Enrollment, 2001 (rate per 100 residents) California U.S.–California Florida Illinois New York Texas Average Monthly Enrollment 15 12 11 11 17 10 Monthly Take-Up Rate 78 64 68 85 70 56 Average Monthly Eligible 20 18 17 13 24 18 the 12 percent average in the rest of the county, 10 percent in Texas, and 11 percent in Florida and Illinois (Table 6.7). For all Medicaid and SCHIP enrollees as a group, we can separate the effects of eligibility rules and demographics from individual participation decisions, or take-up rates. In the following equation, the first term measures the take-up rate of Medicaid or SCHIP among those who are eligible for the programs; the second term reflects the share of population who are eligible for these programs based on eligibility criteria set forth by states and each state’s demographic and economic characteristics: Medicaid and SCHIP Recipients = Medicaid and SCHIP Recipients Population Medicaid and SCHIP Eligibles  Medicaid and SCHIP Eligibles Population (6.3) The results of this decomposition presented in Table 6.7 show that, in an average month of 2001, 20 percent of Californians were potentially eligible for the full benefits of Medicaid or SCHIP, and 78 percent of them took up the services.4 Thus, both the eligibility rate and the take_____________ 4Estimates of potential eligible population and participants were obtained from the Urban Institute’s TRIM3 model. 91 up rate were higher in California than the average in the rest of the country. In sum, enrollment rates are higher for all Medicaid eligibility groups in California than in other states. California enrolled 6 percent more children, 33 percent more elderly and disabled, and about twice as many nonelderly adults into Medicaid than the rest of the country in 2002. These high enrollment rates do not result from substantially higher poverty rates among potentially eligible groups but rather from a combination of generous eligibility rules and high take-up rates. Although high enrollment rates are in evidence for all eligibility groups, the elderly constitute a smaller share of the overall population in California. Thus, the state’s overall high enrollment rate is primarily driven by the nonelderly population. Higher nonelderly enrollment also helps to suppress average expenditures per Medicaid recipient in California. Nonelderly beneficiaries tend to consume fewer services, and the services they consume are relatively less expensive—for example, clinic services as opposed to skilled nursing facilities (e.g., MaCurdy et al., 2005). In 2002, for example, Medi-Cal expenditures for children and nonelderly adults were about one-fifth those for the elderly or disabled (Table 6.8). Although this gap was similar in other states in our comparison group, California includes more children and nonelderly recipients than other states. Medi-Cal’s low spending per recipient can also be explained by higher managed care penetration in this state. In 2002, 85 percent of Table 6.8 Medicaid Expenditures per Recipient, by Eligibility Group, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 3,113 4,878 3,726 5,393 8,101 3,774 Children 1,715 2,447 1,189 1,503 5,771 1,702 Nonelderly Adults 1,549 2,505 1,988 2,769 3,951 2,694 Elderly or Disabled 9,530 11,592 10,103 15,161 18,669 11,403 92 Medi-Cal recipients were enrolled in some type of managed care plan that provided them with either a comprehensive set of services or a limited set of benefits such as mental health, dental, behavioral health, prenatal, or long-term care. The share of Medicaid recipients enrolled in managed care was 45 percent in the rest of the country and similar in our other comparison states (Table 6.9). In addition, California paid managed care organizations less per enrollee than other states. At $780 per enrollee annually, managed care reimbursements were 41 percent lower than in the rest of the county, 47 percent lower than in New York, 32 percent lower than in Illinois, 29 percent lower than in Texas, and 24 percent lower than in Florida. For recipients not enrolled in managed care plans or those enrolled in plans with limited benefits, Medicaid services are directly purchased from providers on a FFS basis. As the last column of Table 6.9 implies, the bulk of Medicaid expenditures occur under the FFS system. Overall, California reimburses these providers at a rate that is 91 percent of the national average, although there are some differences by type of service (Table 6.10). Variation in Medicaid reimbursement rates may reflect the costs of providing medical services in different states. However, California is a high-cost state in which to practice medicine. According to a 2001 study, it costs 7 percent more to provide physician services in California than the national average. Among the four comparison states, only New York is more expensive than California (Lewin Group, 2001). Table 6.9 Components of Medicaid Expenditures per Recipient, 2002 California U.S.–California Florida Illinois New York Texas Total Managed Care Managed Care Expenditure Expenditures Payment per Enrollment per Share of per Participant Participant Participant Managed Care 3,113 780 0.85 0.21 4,878 1,332 0.45 0.12 3,726 1,023 0.41 0.11 5,393 1,028 0.09 0.02 8,101 1,459 0.42 0.08 3,774 1,099 0.35 0.10 93 California United States Florida Illinois New York Texas Table 6.10 Medicaid Fee Index, 2003 Overall 91 100 95 92 70 99 Primary Care 87 100 96 89 71 96 Obstetric Care 83 100 104 103 88 93 Other Services 109 100 83 93 46 109 One way to adjust for geographic variation in the cost of providing medical services is to compare Medicaid reimbursement rates with the corresponding Medicare fee schedules for similar services in each state. Because Medicare fee schedules are supposed to reflect differences in the price of inputs used in furnishing medical services in each state and not case mix, the ratio of Medicaid to Medicare fees would be the same in California and the other states if the differences in Medicaid reimbursement rates represented solely the difference in cost of providing these services in California and the other states. Although Medicaid fees lagged well behind Medicare fees in all the states, the gap was larger in California. Overall Medi-Cal fees were 59 percent of Medicare fees, less than the national average of 69 percent (Table 6.11). This suggests that California reimbursed Medicaid providers at a lower rate than the national average even after adjusting for Table 6.11 Medicaid-to-Medicare Fee Indexes, 2003 California United States Florida Illinois New York Texas Overall 59 69 65 63 45 69 Primary Care 51 62 60 54 40 62 Obstetric Care 65 84 82 84 65 82 Other Services 74 73 58 68 31 82 94 geographic differences in the cost of medical services. Specifically, California paid significantly less for primary care and obstetrical care than the national average, while it reimbursed hospital services slightly more than the national average. In sum, Medi-Cal is distinct from other Medicaid programs because of its large enrollment and low expenditures per recipient. California has generous eligibility criteria for all eligible groups and high take-up rates among eligible individuals compared to the other large states we have considered. A younger low-income population in California as well as a generous enrollment policy has also helped suppress Medi-Cal expenditures per recipient because children and nonelderly adults have relatively low average medical costs. Extensive use of managed care and low rates of reimbursements for medical services provided on a fee-forservices basis brought expenditures per recipient further down in California than in other states. Own Public Hospitals Public hospitals administered by state and local governments serve a large proportion of Medicaid and uninsured patients and play an important role in the health care safety net. Expenditures per capita by these hospitals have been a bit higher in California than in the rest of the country since 1977 (Figure 6.2). A relative increase in California during the early 1990s reflects higher federal DSH payments to public hospitals in these years. Public hospitals receive revenues from patient charges (including those paid by private and public insurance programs), auxiliary services such as hospital cafeterias and gift shops, and government grants. To determine how much these hospitals rely on government subsidies for daily operations, we subtract these revenues from total hospital expenditures. The net amount measures subsidies that hospitals receive from state, local, and federal governments. Compared to the average for the rest of the country, California subsidized publicly administered hospitals at a higher rate in 2002. Out of public hospital expenditures of $319 per capita, 71 percent was financed with hospital charges and receipts and 29 percent reflected government subsidies (Table 6.12). In the rest of the country, the 95 Expenditures (2002 $) 400 350 300 250 200 150 100 California U.S.–California 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.2—Own Public Hospital Expenditures per Capita, 1977–2002 Table 6.12 Components of Own Public Hospital Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 319 298 257 169 454 341 Charges 227 228 218 72 268 205 Subsidy 92 70 39 97 186 136 average subsidy to public hospitals amounted to $70 per capita, or 23 percent of total spending. Among our comparison states, Florida spent 19 percent less per capita than California on public hospitals and subsidized 15 percent of the total hospital spending. New York and Texas spent more per capita than California and provided higher subsidies (40% of total expenditures). Illinois spent about half as much per capita as California on public hospitals, but 58 percent of this spending came from government subsidies. 96 The ratio of government subsidies to total hospital expenditures does not necessarily reflect the self-sufficiency of state hospitals. Several factors can affect the amount of subsidies hospitals receive. For instance, county hospitals may shoulder more responsibility in providing care to the medically indigent population and as a result may obtain more government funding. It is also important to note that government subsidies include subsidies to general hospitals and institutions for the developmentally disabled. Since the latter institutions are likely to receive more government funding, states with a larger share of such institutions may exhibit higher government subsidies. A rigorous study of the self-sufficiency of public hospitals is beyond the scope of this report. Other Health Services California has always spent more than the average for all other states on the category known as other health services, which covers a broad array of activities including mental health and developmental services, public health, and some environmental protection (Figure 6.3). In 1977, California spent 28 percent more per capita than the average of the rest of the nation on this function. Since then, expenditures per 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.3—Other Health Services Expenditures per Capita, 1977–2002 97 Expenditures (2002 $) capita on other health services have grown consistently relative to expenditures in the rest of the country. The early 1990s in particular witnessed a sharp rise in other health expenditures relative to the rest of the nation. This sharp increase coincided with a realignment of health and social services programs in California that transferred responsibilities for certain health care and mental health programs from the state to counties. In exchange, counties received additional revenues from a state-levied halfcent sales tax and increased vehicle license fees. In 1992, these revenues amounted to $1.5 billion. By 2002, they grew to $2.6 billion. The gap between expenditures on other health services between California and the rest of the nation further widened after 1996, when the state increased funding for developmental services related to health. It would be instructive to identify the specific health activities that drive California’s higher spending in this area. However, in California as in most other states, the bulk of activities included in this function are administered at the county level and comparable information across states on these local functions does not exist. Outcomes It is notoriously difficult to link public spending to health outcomes. However, Californians fare well in terms of two commonly accepted measures of quality of care—preventable hospitalizations and birth outcomes. They fare less well on a third measure—access to care. The preventable hospitalization rate reflects whether patients have access to timely and effective ambulatory care. Such care can help prevent hospitalizations for chronic conditions such as diabetes, asthma, and congestive heart failure, or for the deterioration of acute conditions such as infections and cellulitis or preventable diseases such as tetanus and rheumatic fever. A high rate of preventable hospitalizations therefore suggests access problems or poor performance of the health system. According to a study by the Agency for Healthcare Research and Quality (AHRQ), California had lower preventable hospitalizations for all age groups in 1999, the latest year for which complete data are available. Its rate of preventable hospitalizations was particularly low for 98 the nonelderly population compared to the nation overall and the three comparison states for which data are available (Table 6.13). We should be careful in attributing the low preventable hospitalization rate in California to government expenditures on health services. Other factors such as physician practice style and the local health care delivery system in private health care markets can also affect hospitalizations. For example, physicians have discretion over whether to admit a patient to a hospital or to manage the case on an ambulatory basis. The West Coast in general exhibits lower use of hospital services in studies of regional differences in physician practice style (Billings and Weinick, 2003). Similarly, the prevalence of HMO-type managed care plans in the private health insurance market in California could reduce hospital use among those who are insured, which in turn contributes to a low rate of preventable hospitalizations (Glied, 2000). As noted above, birth outcomes provide another measure of quality of care. Common measures include deliveries to mothers who received no or late prenatal care (in the third trimester), low-weight newborns (weighing less than 2,500 grams after a gestation period of 37 weeks or longer), and preterm births (deliveries before 37 weeks gestation). Compared to the national average and the other three large states for which data are available, California reported the lowest rates for all three indicators in 1999 (Table 6.14). On another measure, self-reported access to care, California did not fare as well. In major U.S. metropolitan areas, two of every ten persons with income up to 200 percent of FPL reported that they had no usual Table 6.13 Average Preventable Hospitalizations per 1,000 Persons, 1999 California United States Florida Illinois New York Ages 0–17 8 10 12 11 13 Ages 18–39 5 7 8 8 8 Ages 40–64 16 19 20 22 21 99 Table 6.14 Birth Outcome Indicators per 100 Births, 1999 California United States Florida Illinois New York Late or No Prenatal Care 3 4 3 4 5 Low Birth Weight 2 3 3 3 3 Preterm Births 10 12 13 12 11 place of care if they were sick or in need of health advice. This ratio was higher in certain regions of California (Table 6.15). Four out of ten individuals with incomes up to 200 percent of FPL reported having no usual source of care in the Ventura area, and about three out of ten Table 6.15 Access to Care Measures: Percentage of Population Below 200 Percent FPL, 1999–2000 California Los Angeles–Long Beach PMSA Orange County PMSA Riverside–San Bernardino PMSA San Diego MSA Ventura PMSA Oakland PMSA Sacramento PMSA San Francisco PMSA No Usual Source of Care No Doctor Visit in Past Year 25 26 30 33 28 27 27 21 39 35 12 16 18 23 19 23 Other Areas U.S. national average Miami PMSA Chicago PMSA New York PMSA 21 21 31 27 25 20 15 13 NOTES: MSAs are Metropolitan Statistical Areas as defined by United States Office of Management and Budget. PMSAs are Primary MSAs, or components of larger Consolidated MSAs (CMSAs). 100 responded similarly in the Orange County, Riverside–San Bernardino, San Diego, and Los Angeles–Long Beach areas. By comparison, Miami and Chicago also reported higher rates of no usual source of care than the national average, and New York reported a lower rate. Low-income individuals from areas with high rates of no usual source of care were also more likely to report not having visited a doctor within the last year. Various factors affect the self-reported measure of access to care. For example, the uninsured population is more likely than those who are insured to report problems with access to health care, and the uninsured rate is higher in California than the national average. Some studies suggest that higher Medicaid physician fees increase physicians’ participation in Medicaid (Zuckerman et al., 2004), thereby improving access to care by Medicaid patients. Although this link is not definite, low rates of reimbursement for primary care services of Medi-Cal could contribute to the relatively poor self-reported access to care by lowincome individuals in California. Conclusions California has the largest state and local health budget in the nation, with Medicaid program expenditures accounting for over half of this spending. Although the state’s total health expenditures per capita in 2002 were in line with those in the rest of the nation, per capita Medicaid expenditures were below this average. Medicaid expenditures per recipient were significantly lower in California than in the rest of the country, whereas Medicaid enrollment rates were higher. To a large extent, this pattern is explained by generous eligibility rules for children and adults under Medi-Cal and a younger low-income population in California. Other factors affecting expenditures per recipient in California are high managed care enrollments, low payments per managed care enrollee, and low reimbursements under the fee-forservice system. A quarter of California’s health budget was spent on hospitals, with governments subsidizing publicly administered hospitals at a rate higher than that in the rest of the nation although at a comparable rate to other large states. The remaining quarter of the state and local health budget 101 was spent on miscellaneous health services, most of which were administered at the local level, and California spent more per capita on these services than other states did. 102 7. Social Services In 2002, California state and local governments spent $18.7 billion on support for low-income households, including both cash payments and subsidized services. Expenditures in this area represented 6 percent of total state and local spending, compared to an average of 4 percent in the rest of the country. This chapter analyzes differences in social services expenditures between California and other states, focusing on major assistance programs such as Temporary Assistance for Needy Families (TANF), community services (e.g., child care, foster care, adoption assistance, and supportive services for the elderly and disabled), and the State Supplementary Program to Supplemental Security Income (SSI/SSP). California spends more per capita than the rest of the country on most types of assistance for low-income households. However, there are differences by program. California has higher maximum allowable benefits and serves more recipients per capita under its TANF program than does the rest of the country, but it spends less on average per recipient. By contrast, expenditures per recipient for community services programs and aid to the low-income aged, blind, and disabled are higher in California than in comparable states. Total Expenditures State and local governments in California provide services to the state’s low-income population through several programs. Perhaps the best known program is California Work Opportunity and Responsibility to Kids (CalWORKs), which provides cash aid, child care, and job training assistance to low-income families. CalWORKs implements the federal TANF program, a block grant program created in 1996 to replace the entitlement program AFDC. Under TANF, states receive fixed block grants to operate programs of their own design within federal limits on maximum lifetime participation and work requirements. In 103 exchange for these grants, states must provide matching funds, often referred to as “Maintenance of Effort” (MOE). In addition, states may choose to supplement these expenditures with additional resources. Throughout this chapter, we refer to state expenditures of federal funds and state matching funds as “federally mandated” expenditures and to the state supplements as “additional state spending.” The 1996 federal welfare reform effort also expanded funds for child care services, which are available to TANF recipients under different rules than those applicable to child care for other low-income working families. States can fund child care services for TANF recipients or other low-income Californians with grants from the federal Child Care and Development Fund (CCDF) and the Social Services Block Grant (SSBG) in addition to their own resources. As in welfare, some child care expenditures are mandated under the terms of participating in federal programs. In addition to welfare and child care programs, state and local governments operate a variety of community-based assistance programs. These include child welfare services such as the protection of abused or neglected children, foster care, and adoption assistance. States and localities may also provide supportive services for the elderly and disabled, including the developmentally disabled. States and localities have substantial flexibility in establishing these programs, often with support from federal SSBG funds. However, there are no federally mandated expenditures in this area, and states may supplement federal funds with their own resources. In the case of California, this state supplement is substantial. The final major assistance program is the State Supplementary Program to the federal Supplemental Security Income program, which provides cash transfers to low-income elderly, blind, and disabled individuals. Together, these programs are known as SSI/SSP. Under SSI, the federal government establishes eligibility rules and mandates minimum benefit levels to be funded by both the federal government and the states. States also have the option of providing additional aid. California is one state that chooses to supplement federally mandated aid. It also contracts with the federal Social Security Administration to administer SSI/SSP. 104 The U.S. Census Bureau classifies social services expenditures not by program but by type of aid: federal categorical assistance, other cash assistance, other welfare, vendor payments, welfare institutions, veterans’ services, and social insurance administration.1 The cash portions of CalWORKs and SSI/SSP are included in categorical and other cash assistance, hereafter referred to as cash assistance.2 Other welfare includes in-kind aid such as job preparedness assistance and training provided through CalWORKs, child care for welfare recipients and the working poor, foster care, adoption assistance, and community services. Vendor payments mostly include payments to energy providers under the Low-Income Home Energy Assistance Program (LIHEAP). This program provides heating and weatherization support to low-income individuals through cash transfers or subsidies of energy services. The remaining three Census categories consist of administrative functions of nursing facilities, veterans’ financial benefits programs, and unemployment and disability insurance. In 2002, California spent 70 percent more per capita than the rest of the country on social services, including cash assistance and in-kind services as well as administrative support (Table 7.1). Only New York spent more than California, although it provided relatively less aid in cash and more through services. Vendor payments were notably low in mild-climate states such as California and higher in states facing harsher weather conditions such as Illinois and New York. In 2002, aid to individuals and families as either cash or in-kind assistance (other welfare) accounted for 97 percent of social services spending in California. This share ranged from 88 percent in Illinois to 92 percent in Florida, and was 90 percent in the rest of the nation. Vendor payments and administration expenditures represented a very small fraction of total social services spending in California and the rest _____________ 1The Census Bureau includes an eighth category—veterans’ bonuses—which does not apply to California. Since this category constitutes a small share of social services expenditures (less than one-hundredth of a percentage point), it is dropped from the analysis. 2Categorical cash assistance also incorporates some state intergovernmental aid to local governments through the Medicaid program. These transfers are excluded from the analysis that follows (see Chapter 6). 105 Table 7.1 Social Services Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 533 311 164 363 563 157 Cash Assistance 240 62 21 45 132 38 Other Welfare 276 217 130 273 381 104 Vendor Payments Administration 2 16 8 25 49 18 27 23 27 1 15 of the country. Thus, the remainder of this chapter focuses on the programs included in cash assistance and other welfare. California has dedicated more resources per capita to social services than the rest of the country throughout recent history (Figure 7.1). Only New York has provided more total assistance per capita since 1977. In terms of specific types of assistance, California’s cash payments per capita have also exceeded those in the rest of the country, whereas inkind aid has remained close to the average for the rest of the nation until recently. More recently, California has relied less on cash transfers and more on in-kind assistance and services. The rest of the country 600 550 500 450 400 350 300 250 California 200 U.S.–California 150 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 7.1—Social Services Expenditures per Capita, 1977–2002 Expenditures (2002 $) 106 experienced a similar switch in the importance of these two types of aid as a consequence of the implementation of the 1996 federal welfare reform. California differs from other states in our comparison group in that its welfare programs are highly decentralized. This pattern is more pronounced for cash transfers than for other welfare services. Local governments in California spend 71 percent of total social services expenditures, whereas the average local government share of total social services spending is 40 percent in the rest of the country and as low as 9 percent in Illinois. Only New York exhibits a higher level of decentralization than California, with local governments spending 91 percent of total social assistance funds. The Census Bureau’s categorization of social services expenditures is not ideal for understanding individual programs. For example, TANF funds may be classified under categorical cash assistance or other welfare depending on state practices. Further difficulties arise with in-kind assistance and subsidized services. For instance, vouchers for child care are technically noncash social services expenditures (other welfare) from state and federal funds. However, some of these programs are administered by the Department of Education and the California Community Colleges. The Census Bureau thus classifies these expenditures under its education categories. For these reasons, the remainder of this chapter uses data from alternative sources that allow for more meaningful comparisons of programs across states. The analysis of welfare and community services uses reports to the Department of Health and Human Services’ Administration for Children and Families (ACF) on the use of TANF, CCDF, and SSBG funds, as well as data from the U.S. House of Representatives (2004), commonly known as the Green Book. The analysis of SSI/SSP relies on the Social Security Administration’s annual reports. These alternative data sources have some drawbacks. First, they restrict the analysis to expenditures on three major sets of programs: welfare services, community services, and SSI/SSP. Second, they provide only aggregate information about welfare caseload. This hinders an analysis by type of recipient, an important distinction in the 107 implementation of these assistance programs. Third, there are insufficient data on child care programs across states to permit comparisons of child care spending per recipient.3 Fourth, SSA reports on SSI/SSP include federal direct spending as well as state and local expenditures of federal funds. As a result, the sum of social services expenditures from these alternative sources will necessarily exceed Census Bureau state and local government expenditure totals. Despite these caveats, the data sources described above allow a better analysis of social services spending across states beyond what is permitted by the Census Bureau classification scheme. Throughout this chapter, the low-income population is defined as individuals with income below 200 percent of the FPL, $30,040 for a family of three in 2002. Although this benchmark does not necessarily coincide with eligibility criteria in each state for programs analyzed in this chapter, it encompasses the most generous eligibility criterion. Thus, this threshold accounts for all potentially eligible individuals and establishes a common reference for all states. Welfare (TANF/CalWORKs) In California, 32 percent of social assistance expenditures in 2002 were through CalWORKs, the state’s TANF program targeting lowincome families with children. This program provides cash aid, child care subsidies, employment training, and counseling services to reintegrate parents into the labor force. These activities are predominantly financed with federal expenditures and state maintenance of effort funds. Federal funds are based on state expenditures before the federal welfare reform of 1996. This decision locks in California’s TANF federal grant and mandated state expenditures at the highest level in the nation. In addition, California supplements these mandated funds more generously than other states. As a result, California’s expenditures on welfare programs were nearly twice the average for the rest of the nation _____________ 3For a detailed overview of child care expenditures and policies in California, see O’Brien-Strain, Moyé, and Sonenstein (2003), and Marrufo, O’Brien-Strain, and Oliver (2003). 108 in 2002 (Table 7.2). Compared to other large states, only New York had higher welfare expenditures per capita than California, in terms of both total spending and voluntary state spending. To understand the determinants of spending in this area, it is useful to separate expenditures per capita into caseload, or recipients per capita, and expenditures per recipient. We define caseload as the number of persons (adults plus children) receiving assistance under TANF. This definition differs from the commonly used metric of the number of families receiving assistance. Under the latter definition, a family of three (e.g., one parent and two children) receiving cash or services counts as one case, as does a child living with nonparent relatives, even if those relatives do not receive any assistance. Under our definition of caseload, the first family would include three recipients and the second would include only one. This definition is partly due to data limitations. At the same time, it ensures comparability across types of programs and across other chapters in this report. Following the decomposition method outlined in Chapter 2: TANF Expenditures = TANF Recipients  Expenditures Population Population Recipient (7.1) By this metric, California’s average annual expenditure of $4,700 per recipient was below the national average (Table 7.3). California’s yearly expenditures per case exceeded those of Texas by $2,500 but were the Table 7.2 Welfare Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita 159 81 61 79 205 35 State Separate Expenditures per Capita 14 5 1 0 36 1 109 Table 7.3 Components of Welfare Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 159 71 61 79 205 35 Caseloads per Capita 0.034 0.016 0.008 0.011 0.022 0.016 Expenditures per Case ($) 4,718 5,105 8,053 7,264 9,336 2,236 State Separate Expenditures per Case ($) 409 307 166 25 1,637 66 second lowest in our comparison group. New York spent nearly twice as much per case as California in 2002. By contrast, California’s additional expenditures beyond mandated spending were 33 percent higher than the average for the rest of the nation, although just one-quarter of New York’s. The finding of lower expenditures per case may seem at odds with California’s generous maximum allowable benefits per family. For instance, in 2002, California’s maximum cash grant per month for a family of three (one adult and two children, with no income) was $679, the third highest in the nation behind Alaska ($923) and Hawaii ($712). The gap between the maximum benefit per family and the average expenditures per case can arise from a number of factors, including shorter average spells on welfare in California, larger family sizes, or a different allocation of cash assistance versus in-kind benefits. California’s lower average expenditure per case is offset by the state’s higher caseload. In 2002, the number of welfare recipients per capita was 54 percent higher in California than in New York, and more than twice as high as in Texas or the average for the rest of the country (Table 7.3). California’s high caseload in CalWORKs may be a consequence of factors over which the state has little direct control, such as the poverty rate. Alternatively, differences across states may arise from policy choices, including eligibility rules, benefit levels, lifetime participation allowances within the five-year federal limit, and workforce participation requirements. 110 For these reasons, it is useful to separate caseload into the potentially eligible population and the participation rate within this group: TANF Recipients = Potential Eligibles  Recipients Population Population Potential Eligibles (7.2) The first component of this equation captures a state’s demographic and economic characteristics. The second component reflects both state eligibility rules or benefit levels and individual participation decisions (i.e., the take-up rate). For the purpose of this analysis, we broadly define potential eligibles as the number of individuals in a low-income family that includes at least one child. In 2002, the fraction of individuals in low-income families with children was 27 percent higher in California than in the rest of the United States. Among these individuals potentially eligible for welfare benefits, the recipiency rate was also higher, by almost 70 percent (Table 7.4). Of our comparison states, only Texas had more low-income individuals with children per capita than California. However, Texas also had a very low number of recipients within the low-income population with children (58 recipients per 1,000 potentially eligible state residents). Only New York’s recipiency rate among potential eligibles was, like California’s, higher than the rest of the country’s. However, New York had 28 fewer recipients per 1,000 low-income residents than California. Table 7.4 Components of Welfare Caseload, 2002 California U.S.–California Florida Illinois New York Texas Recipients per Capita 0.034 0.016 0.008 0.011 0.022 0.016 Low-Income Individuals with Children per Capita 0.243 0.192 0.196 0.169 0.200 0.273 Recipients per Low-Income Individuals with Children 0.138 0.082 0.038 0.064 0.110 0.058 111 Welfare eligibility policies span a number of dimensions. California’s rules are usually regarded as generous and are certainly part of the explanation for the high number of recipients in the state as a share of the potentially eligible population (MaCurdy, Mancuso, and O’Brien-Strain, 2002). Unfortunately, data limitations prevent us from separating the effect of these rules from individual participation decisions. The following thought experiments are useful to see what would happen if California’s eligibility policies and individual participation decisions mimicked those of other states. If California had the same economic and demographic characteristics as it does now with Florida’s recipiency rate, its caseload would drop from around 1.2 million to fewer than 319,000 recipients and spending would drop to only 28 percent of federally mandated funds. Assuming that total spending remained the same, this drop in recipients would allow California to increase annual expenditures per recipient by $12,000. Similarly, if California’s recipiency rate were like New York’s, it would experience a more modest reduction in its caseload, to just over 920,000 recipients, and potentially would spend only 80 percent of its mandatory funds. It would then hypothetically be able to increase yearly benefits by $1,200 per case or 26 percent. Of course, these changes are merely for illustrative purposes. They do not take into consideration adjustments to income eligibility thresholds and the corresponding take-up rate derived from a change in benefit levels.4 Nonetheless, they give a sense of the size and composition of California’s welfare program. One component of welfare expenditures that draws special attention is child care. California devoted $24 dollars per capita (15% of total welfare expenditures) to child care for welfare families in 2002 (Table 7.5). Of this total, $1 per capita was in addition to federally mandated expenditures. By contrast, the rest of the country devoted $11 per capita _____________ 4Specifically, higher benefits increase the income eligibility threshold. Such a policy would have two effects. First, a mechanical increase in the pool of potential eligible people would increase the number of recipients for a constant number of applicants. The second effect is a behavioral one, whereby people are more likely to apply for and receive assistance. The combination of both effects eventually increases caseload and erodes, at least partially, feasible increases in benefits per case. 112 Table 7.5 Welfare Child Care Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Welfare Child Care Expenditures per Capita 24 11 17 29 5 1 State Additional Welfare Child Care Expenditures per Capita <1 <1 0 0 0 0 to child care, with less than $1 per capita from additional state resources. Within our group of comparison states, Illinois spent the highest amount on child care at $29 dollars per capita. However, none of these resources were in addition to federally mandated spending. New York’s expenditures in this area were one-fifth of California’s, and Texas’s expenditures for welfare child care were 20 times smaller than California’s and the lowest in the nation. California’s welfare expenditures declined after the state’s full implementation of welfare reform in 1998. The most important factor behind this reduction—in California and in the rest of our comparison states—was caseload reduction (Figure 7.2). California’s number of TANF recipients per capita decreased 16 percent between 1998 and 2002, compared to 14 percent in the rest of the nation. This decline was accompanied by a 12 percent increase in benefits per recipient. The growth in benefits per capita in California exceeded the average rate of inflation over this period but remained below the 19 percent increase in benefits per recipient in the rest of the country. California’s reduction in caseload during this period lagged those in Florida (23%), Illinois (34%), and New York (22%) and thus its increase in benefits per case was also lower. 113 Caseload 0.09 0.08 California 0.07 U.S.–California 0.06 0.05 0.04 0.03 0.02 0.01 0 1998 1999 2000 2001 2002 Figure 7.2—TANF Caseload per Capita, 1998–2002 Community Services In 2002, 53 percent of California’s expenditures on social services were devoted to community-based programs supported at least in part with SSBG funds. Several of these programs are aimed exclusively at children: day care, foster care, and adoption assistance. Other programs provide day care for adults, home-based services (including meals for the elderly and supportive services for the disabled), and protective services for adults. In addition to SSBG, these programs are financed with supplemental state resources and federal funds, including CCDF and other matching federal grants.5 In recent years, California has consistently concentrated these funds on child day care, foster care, adoption assistance, day care for adults, home-based services including meals for the elderly, and services for the disabled. Other states in our comparison group provide additional services with these funds such as family planning and counseling, employment services, assistance for transitional and independent living, _____________ 5The latter are particularly important for foster care and adoption assistance (“Title IV” grants). 114 substance abuse, and a variety of other services including health, legal, and transportation assistance. As a whole, California’s per capita expenditures in 2002 on community-based services were 68 percent higher than in the rest of the nation but were 45 percent lower than those of Illinois (Table 7.6). In contrast to its welfare spending, Illinois spent the highest amount per capita on community services in our comparison group. Overall, these services are heavily funded by state-generated funds. California’s ownsource expenditures per capita were almost twice those in the rest of the country in 2002, and these funds represented a greater share of spending (70%) than in any of our comparison states except Illinois (73%). As in the case of welfare, the number of recipients of these services depends in large part on the demographic and economic characteristics of the state. However, the variety of programs in this area precludes using any one definition of the potentially eligible population in each state. For some services, individuals can be eligible even in the absence of children. For others, such as foster care and adoption services, there is no low-income requirement. Because of these distinctions among community services programs, we analyze three major types of services separately: child care, child welfare, and services targeted primarily toward adults and the disabled. In the case of child care, data limitations related to sources of funds and administration of programs prevent a meaningful comparison of Table 7.6 Total Community Services Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 237 141 78 344 140 94 State-Fund Expenditures 167 87 35 250 64 55 115 caseloads and expenditures per child across states. It is nevertheless useful to have a reference point. In 2002, California’s expenditures on child care for both welfare and nonwelfare recipients totaled $3.5 billion or $100 per capita. Of this total, almost $1.9 billion in spending was federally mandated and the remainder came from state funds. Child care programs served approximately 611,500 children in that year, averaging expenditures of $5,660 per case (Legislative Analyst’s Office, 2002, 2003). Our second group of community-based services is child welfare. The largest programs in this group are foster care and adoption assistance. The goal of these programs is to improve the living conditions of children and their families or to provide substitutes for parents who have difficulty performing their obligations. California’s per capita expenditures on foster care were twice as much as the average in the rest of the country in 2002, although only half those of Illinois (Table 7.7). Nearly half of these expenditures were covered by state and local funds, closely following the average matching rate required to gain federal funds. In contrast, Illinois funded almost 60 percent of these services through state and local funds, well above the federal matching rate. California’s caseload per capita for foster care was practically the same as in the rest of the country and in other large states in 2002. Higher expenditures on foster care therefore reflect higher expenditures per child (26% higher than in the rest of the nation and about twice those of Florida or Texas). On the other hand, California’s foster Table 7.7 Components of Foster Care Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita ($) 65 33 21 138 51 13 Cases per Capita 0.003 0.002 0.002 0.002 0.002 0.001 Total Expenditures per Case ($) 15,388 12,149 7,059 52,633 16,233 8,761 116 care expenditures per child were within 5 percent of those in New York and much lower (by one-third) of those in Illinois. Expenditures for adoptions assistance show a similar pattern in 2002. In that year, California spent 18 percent more per capita than the rest of the country but 70 percent more per case (Table 7.8). Expenditures per case were four times higher in California than in Texas (the state with the lowest expenditures per case) but were 20 percent lower than in Illinois. The level of expenditures per case in California was high despite having the lowest caseload per capita in our comparison group: 40 percent of New York’s caseload and just over 50 percent of the average for all other states. For both foster care and adoption assistance, California’s higher expenditures per case may result from a number of factors. In foster care, assistance payments to foster families constituted 35 percent of total foster care expenditures in California, compared to 45 percent in the rest of the nation in 2002. Among the states in our comparison group, California devoted the lowest share of foster care expenditures to assistance after Florida (20%) and considerably less than this share in Illinois (83%). Instead, a higher share of expenditures in California went toward administration, case management, and placement services. Expenditures for adoption assistance displayed a similar pattern. Direct adoption payments in California accounted for 79 percent of total spending, compared to 76 percent in the rest of the nation, 62 percent in Florida, and 98 percent in Illinois. Table 7.8 Components of Adoptions Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita ($) 13 11 7 31 21 4 Adoption Cases per Capita 0.00044 0.00066 0.00064 0.00084 0.00099 0.00051 Total Expenditures per Adoption Case ($) 28,560 16,725 10,524 36,204 20,871 6,835 117 The third group in our community services analysis encompasses all other expenditures of SSBG funds. In California, this group is primarily made up of protective services for adults, home-based care and meals for the elderly, and developmental services for the disabled. In other states, it may include counseling services, substance abuse treatment, transportation, promotion of independent living for adults, and family planning, among other services. Taken together, California’s per capita expenditures on other community services are almost three times those of the rest of the country and the highest among our comparison states (Table 7.9). Similar to Illinois, state and local own funds are the main source of funding for these services in California: The shares of expenditures financed with these sources were 95 percent in California and 92 percent in Illinois in 2002. As in the rest of this chapter, we analyze the roles of caseload and expenditures per case in determining total expenditures through Eq. (7.1). This decomposition shows that the number of community services beneficiaries relative to the population in California was close to New York’s, exceeding it by only 8 recipients per 1,000 residents in 2002. Caseload per capita in California was 44 percent lower than in the rest of the country, and 48 percent of Florida’s (Table 7.10). It was six times lower than the caseload in Illinois. Average expenditures per recipient in California were almost $6,000 higher than in the rest of the country and twice those in New York, Table 7.9 Expenditures on Other Community Services per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 127 46 22 99 35 30 State Separate Expenditures 121 41 15 92 22 23 118 Table 7.10 Components of Expenditures on Other Community Services per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Other Expenditures per Capita 127 46 22 99 35 30 Total Other Beneficiaries per Capita 0.018 0.041 0.037 0.110 0.010 0.075 Total Other Total Other State Additional Expenditures Expenditures per Beneficiary per Beneficiary 7,011 6,690 1,117 986 611 409 902 834 3,374 2,144 397 312 despite similar numbers of beneficiaries per capita in these two states. In contrast, Illinois provided less than one-seventh the benefits per recipient as California but had six times as many recipients. These differences are consistent with the fact that California spends nearly half of its community services expenditures other than child welfare on the disabled population. In fact, California’s community services other than child care and child welfare have consistently focused on developmental services for the disabled and home-based care in recent years. By comparison, this group accounts for 23 percent of expenditures from these funds in Florida and 7 percent in Illinois. How much would California spend on each of its beneficiaries if it adopted other states’ policies and had similar take-up rates? The relevant exercise is to consider other states’ choices of own-source expenditures. The results are quite dramatic. If California’s expenditures per recipient from its own funds were the same as those in Texas, it would reduce total expenditures to $12 per capita. This would imply a reduction from $6,700 to $662 in total expenditures per recipient. Adopting New York’s state supplementation per case would shrink California’s expenditures to $45 per capita, or $2,500 per case. There has been a wide disparity in the evolution of community services expenditures across states, mostly driven by the change in 119 additional state resources per case (Table 7.11). This is clearly the case for California, where the growth in expenditures from the state’s own sources has outpaced the growth in its caseload. As a consequence, expenditures per beneficiary increased threefold from 1998 to 2002. During this period, all our comparison states also expanded their coverage in terms of caseload per capita. However, all of them have spent less from their own resources on those additional cases almost every year. Florida and New York stand out in this respect, respectively reducing own-source spending by approximately 63 and 60 percent. By contrast, Illinois doubled its expenditures per case in the same period. Table 7.11 Other Community Services Expenditures per Beneficiary from State-Only Sources, 1998–2002 (dollars) California Florida Illinois New York Texas 1998 2,126 1,117 413 5,425 468 1999 2,392 555 127 1,482 322 2000 2,840 1,999 777 8,352 376 2001 4,265 526 136 1,032 683 2002 6,690 409 834 2,144 312 State Supplementary Program for Supplemental Security Income The SSI program provides cash transfers to low-income elderly, blind, and disabled people. Its goal is to provide a means of subsistence for members of these groups who are not eligible for other welfare programs. The federal government establishes eligibility rules for this program. Benefits consist of three components: a minimum level provided by the federal government, a mandated amount to be provided by states, and an optional state supplement. We cannot distinguish between the two types of state spending in our data. Even though states cannot affect the group of eligible individuals by establishing eligibility rules directly, they affect their numbers through their voluntary supplemental assistance. This happens because the maximum combined SSI/SSP benefit (i.e., federal aid plus state supplement) determines the income eligibility threshold. California’s 120 total supplementation to federal aid represented 38 percent of its total SSI/SSP spending in 2002. The state also contracts with the federal government for the administration and distribution of benefits under SSI. In 2002, California’s SSI/SSP expenditures per capita were almost twice those in the rest of the country, Florida, or Illinois and much higher than those in Texas (Table 7.12). Its spending on this program was similar to that of New York. Both states had higher-than-average state supplementation (including the mandated portion) in 2002. However, California’s supplement was more than twice that of New York’s and ten times beyond the average for the rest of the country. In California and other states, the largest share of expenditures was targeted at the low-income blind and disabled populations. California and New York stand out as states with large caseloads and higher expenditures per case (Table 7.13). California’s high expenditures per case result from its choice of a more generous state supplement. State aid per case in 2002 was highest in California at roughly $2,400 per recipient (38% of total assistance). Without state supplementation, California’s benefits per case ($3,927) would be the lowest within our comparison group, although only 15 percent below New York’s and $100 less per recipient than in the rest of the country. It is also important to notice that California is the only state that has converted food stamps into cash and that it includes these payments as part of the state supplement. This conversion further increases California’s Table 7.12 SSI Expenditures per Capita: Use and State Supplementation, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures Total Expenditures on the Blind Total State Expenditures on the Elderly and Disabled Supplementation 207 55 152 79 108 11 93 8 109 20 89 1 101 9 90 2 178 31 147 29 83 14 69 <1 121 Table 7.13 Components of SSI Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 207 108 109 101 178 83 Cases per Capita 0.032 0.022 0.024 0.020 0.033 0.020 Expenditures per Case ($) 6,366 4,406 4,425 4,474 5,450 4,080 expenditures beyond federal requirements and distinguishes it from other states. To examine the components of caseloads, we can use an analog to Eq. (7.3), with the low-income population restricted to the elderly, blind, and disabled. This analysis shows that even though there are fewer low-income elderly, blind, and disabled individuals per capita in California than in the rest of the country, more people within this pool of potential eligibles receive SSI/SSP assistance (Table 7.14). In particular, even though there were fewer potential eligibles in California in 2002 than in the rest of the country (by 14 per 1,000 residents), California’s recipiency rate was almost twice as high as in the rest of the country. New York had a similar situation, with an almost Table 7.14 Components of SSI Caseload, 2002 California U.S.–California Florida Illinois New York Texas Cases per Capita 0.032 0.022 0.024 0.020 0.033 0.020 Low-Income Elderly and Disabled per Capita 0.072 0.086 0.108 0.071 0.090 0.078 Cases per Low-Income Elderly and Disabled Population 0.453 0.258 0.220 0.281 0.361 0.257 122 identical pool of potential eligibles per capita as in the rest of the country but with a recipiency rate 40 percent higher. Both California and New York report high state supplementation per case. However, it is important to recall that the data do not allow separating state supplemental expenditures into their mandated and discretionary components. Despite the link between eligibility and total benefits, it is illustrative to see the effect of California’s higher recipiency rate on expenditures based on the following thought experiment. If California had Florida’s caseload but kept the same benefit levels, it would spend only $100 per capita on SSI/SSP, a 50 percent reduction from its 2002 levels. Alternatively, if it kept its current recipiency rate but provided Texas’s benefit levels, expenditures per capita would only drop to $132 per capita. Thus, California’s higher expenditures on supplementation to SSI are due to a combination of extending additional aid to more people through higher income thresholds and larger supplementation to those cases. Conclusions Social services represent a modest share of total state and local expenditures in California. However, these expenditures have been high relative to those in comparable states in recent years. The state’s welfare programs are a combination of direct cash payments to individuals and families and services that are either subsidized or provided by government agencies. An intricate combination of state policy choices and federal programs influences expenditures in all states. Overall, California’s high per capita expenditures on welfare indicate that the state has relatively generous assistance provisions for these programs. This generosity is reflected in the state’s broad coverage. From the potentially eligible population, more children and adults receive welfare assistance in California than in other states. As a consequence, average benefits per welfare recipient are spread more thinly. In the case of assistance to the low-income blind, elderly, and disabled population, both caseload and expenditures per case reflect California’s generous policies. 123 California has shifted the focus of community services in recent years. Its coverage of at-risk and disabled populations is on par with the rest of the nation although below that in other states. On the other hand, the benefits provided are high. For child welfare services such as foster care and adoptions, expenditures are heavily allocated to administration and management and less to direct assistance. 124 8. Transportation In 2002, transportation expenditures constituted 7 percent of state and local expenditures in California and 9 percent of expenditures in the rest of the country. This expenditure category includes highways, mass transit, airports, water transport, and parking facilities. Because the last three areas are largely enterprise activities and are relatively small, this chapter focuses on highways and mass transit. This focus leads naturally to the issues of transportation within urban areas, highway congestion, and mass transit as an alternative to the automobile. These issues are particularly relevant in California. Since 1977, the state has spent less per capita on highway construction than the rest of the country. As a result, the state’s highway capacity has grown more slowly than its population. On the other hand, California has invested relatively heavily in alternatives to the automobile. Since 1990, it has spent more per capita on mass transit than the rest of the country. Nevertheless, rates of highway travel in California rival those of Texas and Florida, states whose residents rely heavily on the automobile. High demands on California highways, coupled with their relatively low capacity, cause substantial traffic congestion in the state’s urban areas. According to estimates from the Texas Transportation Institute, the freeways of California’s cities are among the most congested in the nation. Total Expenditures Total transportation expenditures, as noted above, include expenditures on highways, mass transit, airports, water transport, and parking. Highway expenditures consist of the maintenance and construction of public roads, streets, highways, and freeways. Within this broad area, the U.S. Census Bureau also includes expenditures on street lighting, traffic signals, and highway storm drains. The main 125 exclusions are expenditures for the patrolling and policing of streets and highways, which are included in the category of police protection. The Census reports expenditures separately for toll and nontoll highways as part of total transportation expenditures. For California and its four comparison states, expenditures on toll highways are generally 10 percent or less of the expenditures of nontoll highways. The sole exception is New York, where this ratio is 14 percent. Because expenditures on toll highways are relatively small, we have combined them with expenditures on nontoll highways in the analysis that follows. California’s roadway is vast. In 2002, the state had 373,000 lane miles of public roadway.1 Of those lane miles, 57,000 were freeways or other principal arterials, 70 percent of which ran through the state’s urban areas. Mass transit expenditures include the expenditures of public transit agencies and also public subsidies for private transit companies. These subsidies are mainly for privately owned railroad companies providing commuter services. The main exclusions from mass transit are the bus systems transporting public school students. These expenditures are included under the category of elementary and secondary education. Although California has 77 transit agencies, most passenger trips are provided by a few agencies. In 2002, public transit agencies in California provided 1.4 billion passenger trips. The largest agency was the Los Angeles County Metropolitan Transportation Agency (LACMTA), which supplied 32 percent of those trips. The three next largest agencies all served the San Francisco Bay Area—San Francisco Municipal Railway (MUNI), the Bay Area Rapid Transit District (BART), and the Alameda–Contra Costa Transit District (AC Transit). These three agencies provided 28 percent of passenger trips in the state. Nearly 90 percent of passenger trips occurred in either the Los Angeles metropolitan area or the San Francisco Bay Area, with roughly the same number of trips in each. Airport expenditures consist of the operation and construction of public airport facilities. This total does not include the expenditures of _____________ 1Lane miles is the product of road length and the number of lanes. A ten-mile stretch of a four-lane highway has 40 lane miles. 126 airline companies leasing space at public airports. In 2002, California had 28 commercial airports offering scheduled airline service, as well as 250 airports serving general aviation. Two-thirds of airport operating expenditures were generated by two airports: Los Angeles and San Francisco. Expenditures on water transport are due mainly to commercial seaports and municipal marinas. As with airports, the Census Bureau excludes the expenditures of lessees such as shipping companies. In 2002, the Port of Los Angeles accounted for 25 percent of California’s water transport operating expenditures. Six ports—Los Angeles, San Diego, San Francisco, Oakland, Stockton, and Long Beach—accounted for 72 percent of these expenditures. The final component of transportation expenditures is parking. This area includes the construction and operation of public parking lots and garages and the purchase and maintenance of on-street parking meters. It excludes parking facilities connected to other public facilities such as sports stadiums and airports. Transportation expenditures differ dramatically across states. In 2002, California spent about the same amount per capita on transportation as the rest of the country (Table 8.1). However, New York spent two-thirds more than California and more than twice as much as Texas. These differences are largely due to mass transit expenditures. New York spent $465 more per capita on mass transit than Texas, accounting for nearly 90 percent of the difference between these two states in total transportation expenditures. Table 8.1 Expenditures on Transportation per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 611 599 586 706 1,004 482 Mass Transit 187 119 62 202 525 60 Water Highways Airports Transport 328 59 32 411 56 10 403 96 20 451 48 1 370 97 10 347 65 9 Parking 5 4 5 4 3 1 127 California stands out in these comparisons in the way it allocates total transportation expenditures between highways and mass transit. In 2002, highway spending per capita was lower in California than in the rest of the country. It was also lower in California than in each of the four comparison states. At the same time, spending per capita on mass transit was higher in California than in the rest of the country. It was also higher than in Texas and Florida, although less than in Illinois and New York, two states dominated by large cities in which fixed-rail transit systems were established before the automobile emerged as a major mode of commuting. Many transportation activities collect revenue from their users. Some highways have tolls for vehicles, buses and subways collect fares from their passengers, airports and seaports impose rents and fees on private companies using their facilities, and parking facilities charge parking rates. These charges defray some public expenditures on these activities. To better measure the effect of transportation activities on public funds, Table 8.2 presents net expenditures, defined as total expenditures minus user fees, charges, and tolls. The user fees, charges, and tolls netted out against highway expenditures in Table 8.2 do not include a variety of taxes and license fees closely related to highway use. These taxes and fees include state gasoline taxes, vehicle license fees, mileage and weight fees on motor carriers, and driver’s license fees. In total, these taxes and fees amounted to just over $153 per capita in California in 2002. Sixty-three Table 8.2 Net Expenditures on Transportation per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 478 484 430 551 615 405 Mass Transit 151 88 51 143 350 54 Highways 309 381 359 417 259 322 Airports 11 14 16 –8 8 28 Water Transport 8 2 5 0 3 1 Parking –2 –1 –1 –1 –5 0 128 percent of this total was the revenue from the state gasoline tax. In other states, highway-related taxes and fees amounted to $186 per capita. When expenditures are netted out against revenues, the smallest three expenditure areas—airports, water transport, and parking— become an even less significant part of transportation spending. For all states except Texas, airport revenues cover at least 75 percent of total expenditures. Ports and marinas also cover most of their expenditures. Parking activities consistently take in more revenue than their expenditures. Because of their small net effect on public funds, these areas are not analyzed further in this chapter. In 2002, net transportation expenditures in California were approximately equal in per capita terms to expenditures in the rest of the country. As Figure 8.1 shows, expenditures in 2002 were somewhat anomalous, although perhaps also the culmination of a long-term trend. From 1977 through 2000, net expenditures per capita were lower in California than in the rest of the country. In the 1970s and 1980s, expenditures per capita were consistently 20 to 30 percent lower in California than in the rest of the country. That relationship began to change in the 1990s as real expenditures per capita grew in California 600 500 400 300 200 California 100 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.1—Net Expenditures on Transportation per Capita, 1977–2002 129 Expenditures (2002 $) relative to the rest of the country, reaching the level of the rest of the country in 2002. Mass Transit Mass transit is an important reason why transportation expenditures per capita differ significantly across states. The five largest states fall into three broad classes (Table 8.3). Florida and Texas spend relatively little on mass transit. At the other extreme, New York spends seven times as much per capita as either Texas or Florida. California and Illinois lie between those two extremes with net expenditures per capita three times higher than in Florida and Texas but half of those in New York. Subsidies to private transit companies are small in all states. Public transit is most practical in urban areas, and thus differences in expenditures per capita across states are at least partly due to differences in the extent and density of a state’s urban areas. California is one of the nation’s most urbanized states. Eighty-eight percent of California residents live in urbanized areas as defined by the Census Bureau (areas with populations of at least 50,000 and densities of at least 1,000 persons per square mile), a higher percentage than in any of the four comparison states (Table 8.4). California’s urbanized areas are also more densely populated than those of all our comparison states except New York. In 2002, capital expenditures were a significant fraction of the total expenditures of public transit agencies (Table 8.5). In New York, capital Table 8.3 Net Expenditures on Mass Transit per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Net Expenditures 151 88 51 146 350 54 Total Expenditures on Public Transit 181 117 62 198 514 60 Fare Revenue from Public Transit 36 30 10 56 175 7 Subsidies to Private Transit 6 2 0 4 11 0 130 Table 8.4 Expenditures on Mass Transit and Urban Density, 2002 California U.S.–California Florida Illinois New York Texas 2002 Expenditures per Capita on Public Transit ($) 181 117 62 198 514 60 Percentage of Population in Urbanized Area 88 66 84 78 82 71 Residents per Square Mile in Urbanized Area 4,571 2,480 2,451 3,390 4,732 2,697 Table 8.5 Expenditures on Public Mass Transit per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 181 117 62 198 514 60 Capital Expenditures 38 40 22 50 172 22 Current Expenditures 142 77 39 148 342 38 expenditures represented half of current expenditures; in Illinois, they accounted for about one-third. In comparison, California’s capital expenditures were modest—26 percent of current expenditures—and slightly less than the average in the rest of the country. However, because capital expenditures vary considerably from year to year, it is more informative to focus on long-run trends. From a long-run perspective, it becomes clear that California has been steadily building its transit capacity (Figure 8.2). In the 1970s and 1980s, expenditures per capita were lower in California than in the rest of the country. That changed in the 1990s, however, with capital expenditures per capita substantially higher in California than in the rest of the country. 131 Expenditures (2002 $) 100 California 75 U.S.–California 50 25 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.2—Capital Expenditures on Mass Transit per Capita, 1977–2002 Current expenditures have followed the same general trend (Figure 8.3). Throughout the 1970s and 1980s, current expenditures per capita on mass transit were slightly higher in California than in the rest of the nation. Throughout this period, real expenditures per capita were virtually unchanged from year to year in both California and the rest of 200 California 150 U.S.–California 100 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.3—Current Expenditures on Mass Transit per Capita, 1977–2002 132 Expenditures (2002 $) the nation. In the 1990s, however, real current expenditures per capita rose sharply in California while remaining flat in the rest of the country. By 2002, current expenditures per capita were twice as high in California as in the rest of the country. To further analyze current expenditures, we turn to the National Transit Database (NTD) compiled by the Federal Transit Administration. The database includes information about ridership and expenditures for all transit agencies receiving federal support under the Urbanized Area Formula Program. Although the NTD uses slightly different definitions of transit expenditures and does not include all transit agencies, in the aggregate its data match the Census of Governments figures very well. The operating expenditures of transit agencies included in the NTD in 2002 were 93 percent of the current expenditures reported for transit agencies in the Census of Governments for that year. For California, however, that figure was only 75 percent. Three-quarters of the gap was due to differences in current expenditures for the Los Angeles Metropolitan Transit Authority (LAMTA). NTD includes only those expenses directed to the operation of the authority’s public transit system. The Census also includes expenses associated with the authority’s role as a regional transportation planning and coordinating agency. We thus believe that the NTD is a reliable source of information about transit operating costs for both California and the rest of the country. In terms of both costs and service, rail transit differs considerably from bus transit. Rail transit primarily services commuters traveling from locations throughout an urban area to its central city. Although bus transit may play this role too, it also provides transit between noncentral-city locations. As a consequence, bus trips are shorter on average (Table 8.6). States also differ considerably in the extent to which they depend on rail transit versus other forms of transit. In California, 21 percent of mass transit trips in 2002 were by rail (Table 8.6). On the other hand, in New York, 60 percent of mass transit trips were by rail, mostly as a result of the extensive commuter rail and subway system serving New York City. At the other extreme are Texas and Florida, with rail accounting for less than 7 percent of transit trips. 133 Table 8.6 Passenger Miles per Trip, by Transit Mode, 2002 California U.S.–California Florida Illinois New York Texas Miles per Trip Rail Nonrail Total 7.5 3.6 6.9 3.8 7.8 4.8 10.1 3.1 6.5 2.5 6.5 5.0 4.4 5.2 5.0 6.0 4.9 5.1 Percentage of Trips by Rail 21.1 43.0 6.9 41.3 62.0 5.5 Rail and bus transit involve the classic tradeoff between fixed cost and variable cost. Rail transit requires high capital costs, particularly for underground systems. Once those construction costs have been incurred, however, the operating cost of transporting a passenger for a given distance is generally cheaper by rail than by bus. In Table 8.7, operating cost per passenger mile is the total miles traveled by all passengers during a year divided by operating costs during that year. For California, Illinois, and New York, operating cost per mile for rail transit is 45 to 60 percent of the cost of nonrail transit. Within commuting modes, operating costs per mile do not differ significantly among states. For rail transit, cost per mile in California is slightly higher than the average in the rest of the nation. For nonrail transit, the cost per mile in California is slightly lower than the average for all other states. Table 8.7 Operating Cost per Passenger Mile, by Mode, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Rail Nonrail Total 0.41 0.67 0.58 0.36 0.72 0.51 0.57 0.67 0.66 0.32 0.71 0.44 0.37 0.79 0.45 0.69 0.64 0.64 134 In the case of both rail and nonrail transit, the largest component of operating cost is employee compensation (Table 8.8). In California, employee compensation represents 70 percent of rail operating costs and 72 percent of nonrail operating costs. These percentages are similar in other states. Moreover, within a transit mode, employee compensation per passenger mile does not differ much across states. Employee compensation per passenger mile can be decomposed into average hourly employee compensation and employee hours per passenger mile. This decomposition follows the general form of Eq. (2.3) in Chapter 2, with passenger miles as the unit of service, employee hours as the unit of input, and compensation per hour as the price of a unit of that input. In the present case, this equation is: Employee Compensation = Employee Hours Passenger Mile Passenger Mile  Employee Compensation Employee Hour (8.1) Table 8.8 Components of Operating Cost per Passenger Mile, by Mode, 2002 (dollars) California U.S.–California Florida Illinois New York Texas California U.S.–California Florida Illinois New York Texas Employee Fuel and Compensation Utilities Rail 0.29 0.05 0.27 0.05 0.30 0.12 0.22 0.05 0.30 0.05 0.31 0.09 Nonrail 0.48 0.07 0.51 0.09 0.43 0.08 0.50 0.08 0.62 0.09 0.46 0.08 Other 0.06 0.04 0.15 0.04 0.03 0.29 0.12 0.12 0.16 0.13 0.09 0.10 Total 0.41 0.36 0.57 0.32 0.37 0.69 0.67 0.72 0.67 0.71 0.79 0.64 135 Table 8.9 presents the components of this equation. For both rail and nonrail modes, average compensation per hour is slightly higher in California than in the rest of the nation. However, for nonrail transit, employee hours per passenger mile are lower in California than the average for all other states, resulting in a lower compensation per passenger mile than in other states. California’s higher average compensation for transit workers is consistent with general salary differences between California and other states (Table 8.10). In California and in other states, salary ranged from 63 to 66 percent of average compensation. Following the procedures described in Chapter 2, average salary per hour in California and other states was adjusted for location differences using the average salary of Zone 3 workers. After that adjustment, the average salary of transit workers in California is slightly lower than that of transit workers in other states. In terms of operating costs, California’s public transit system appears to be relatively efficient. Despite higher employee compensation Table 8.9 Components of Employee Compensation per Passenger Mile, 2002 California U.S.–California Florida Illinois New York Texas California U.S.–California Florida Illinois New York Texas Employee Average Compensation Employee per Passenger Compensation Mile ($) ($/hr) Rail 0.29 45 0.27 44 0.30 34 0.22 39 0.30 50 0.31 29 Nonrail 0.48 34 0.51 30 0.43 26 0.50 31 0.62 36 0.46 27 Employee Hours per Passenger Mile ($) 0.007 0.006 0.009 0.006 0.006 0.011 0.014 0.017 0.017 0.016 0.017 0.017 136 Table 8.10 Mass Transit Salary Adjusted for Location, 2002 (dollars/hour) California U.S.–California Florida Illinois New York Texas Rail Unadjusted Average Salary Average Salary Adjusted for Location 29 25 28 28 23 25 24 23 31 27 19 20 Nonrail Average Salary Unadjusted Adjusted for Average Salary Location 21 19 20 20 18 19 20 19 23 20 18 19 necessitated by California’s labor market, operating costs per passenger mile are comparable to costs per mile in other states. For nonrail transit in particular, these relatively low costs are due to relatively low employee hours per passenger mile, which suggests that California’s bus systems are operating close to full capacity. Highways The dominant mode of commuting is the automobile. The public sector facilitates this essentially private mode of transportation by constructing and maintaining the highways that make automobile travel possible. Revenue from users is not as significant for highways as for public mass transit (Table 8.11). User revenue includes revenue from toll highways and various reimbursements that cities, counties, and states may receive for constructing and maintaining streets. These reimbursements also include maintenance assessments for street lighting and snow plowing. User revenue for highways does not include fuel taxes although these funds may be earmarked for highway construction. In fact, California governments received more user revenue from nontoll highways ($11 per capita) than from toll highways ($8 per capita) in 2002. In every other state, toll revenues were much higher than reimbursements. 137 Table 8.11 Expenditures on Highways per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Net Expenditures 309 381 359 417 259 322 Total Expenditures 328 411 403 451 370 347 User Revenue 19 30 44 34 111 24 California stood out in a more important regard. Although its current expenditures per capita were roughly equivalent to those of other states, its capital expenditures per capita were much lower (Table 8.12). As Figure 8.4 shows, 2002 was not an exception. Since 1977, capital expenditures per capita in California have been between 40 and 60 percent of expenditures per capita in the rest of the country. In fact, as a percentage of expenditures in the rest of the country, 2002 was nearly a high point, with spending per capita in California at 64 percent of the spending in the rest of the country. Only in 1994 did California reach a higher percentage. In that year, the combination of a dip in real spending per capita in the rest of the country and a small increase in real spending per capita in California produced a ratio of 65 percent. The cumulative effect of California’s relatively low capital spending has been a steady decline of highway capacity relative to population (Figure 8.5). In 1980, California had approximately 2,000 lane miles Table 8.12 Current and Capital Expenditures on Highways per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 328 411 403 451 370 347 Current Expenditures 174 171 129 180 154 128 Capital Expenditures 154 240 273 271 215 218 138 Expenditures (2002 $) 250 200 150 100 50 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.4—Capital Expenditures on Highways per Capita, 1977–2002 3,000 Lane miles 2,500 2,000 California U.S.–California 1,500 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Figure 8.5—Freeway and Other Principal Arterial Lane Miles per Million Residents, 1980–2002 of freeway and principal arterials for every million residents. That ratio has fallen ever since, resulting in 1,600 lane miles per million residents in 2002. In contrast, the rest of the nation had 2,500 lane miles per million 139 people in 1980 and maintained that ratio through the 1980s and 1990s. By 2002, lane miles per capita in California were only 65 percent of the ratio in the rest of the country. This decline does not square with California’s reputation as a pioneer of the modern freeway. In fact, in the 1950s and 1960s, the state was spending more per capita on highway construction than the rest of the country (Table 8.13). Real spending per capita was also rising each year. This pattern changed in 1972, when real spending per capita fell to the level of 1957. It was also less than in all of our comparison states except New York. California’s real spending per capita declined even further in 1977 to a level less than half of 1957. As Figure 8.4 shows, real spending stayed at about this level throughout the 1980s. In highway construction as in all public services, the need for capital spending is more closely related to the growth in population than to the level of population. In states with low population growth, a low level of spending per capita can maintain the existing ratio of highway capacity to population. New York is a good example. From 1980 to 2002, New York’s population increased by 9 percent, its freeway lane miles increased by 7 percent, and it maintained its ratio of lane miles to population (Table 8.14). New York’s low level of capital spending per person was consistent with its low population growth rate. From 1980 to 2002, California’s lane miles of freeways and arterials increased by 19 percent, an increase larger than New York’s but smaller than that in the rest of the nation or in any of our comparison states. In contrast, its population increased by 48 percent during this period, an increase much larger than in New York, Illinois, and all states other than Table 8.13 Capital Expenditures on Highways per Capita, 1957–1977 (2002 $) California U.S.–California Florida Illinois New York Texas 1957 221 192 220 205 149 196 1962 248 221 184 210 206 242 1967 275 256 229 187 203 305 1972 224 257 249 279 203 286 1977 89 179 153 229 94 151 140 Table 8.14 Freeways and Other Principal Arterial Growth in Lane Miles and Population, 1980–2002 California U.S.–California Florida Illinois New York Texas Lane Miles per Million People 1980 2002 2,038 2,484 2,151 1,287 1,260 3,496 1,639 2,528 1,859 1,955 1,236 2,759 Increase, 1980–2002 (%) Population Lane Miles 48 19 25 27 71 48 10 67 97 53 20 California and almost as large as in Florida and Texas. This combination of a relatively modest increase in lane miles and a rapid increase in population decreased the state’s lane miles per capita by 20 percent. Its ratio of lane miles to population is now the lowest of any of our comparison states except New York. Illinois has followed a different course. In 1980, its ratio of lane miles to population was as low as New York’s. Since then, it has increased its lane miles by 67 percent, but its population has grown by only 10 percent. It now has more lane miles per capita than New York, California, or Florida. The ratio of lane miles to population is a proxy for highway capacity relative to the demand for highway travel. However, population is an imperfect measure of the demand for highway travel. Factors such as the locations of jobs and residences can also have a significant effect on demand. In turn, highway routes can affect settlement patterns, creating a feedback between highway supply and demand. This feedback makes highway demand a complex issue, but there is no doubt that demands differ among states. For example, in 2002, residents of New York averaged a little over 3,000 vehicle miles per person, a little more than half of the average vehicle miles traveled by Californians (Table 8.15). Thus, although residents of California had 30 percent more lane miles per capita than residents of New York, freeways and arterials were used 141 Table 8.15 Traffic on Freeways and Other Principal Arterials, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles per Million Persons 1,639 2,528 1,859 1,955 1,236 2,759 Annual Vehicle Annual Vehicle Miles per Miles per Capita Lane Mile 5,938 3,622,515 5,283 2,090,087 5,191 2,791,882 4,532 2,318,457 3,308 2,676,524 6,010 2,178,663 more intensively in California. The ratio of vehicle miles traveled to lane miles was more than 30 percent higher in California than in New York. This high demand for vehicle travel is California’s main transportation challenge. The state has invested in mass transit, and the use of mass transit is higher in California than in many other states. Nevertheless, the ratio of vehicle miles to population is nearly as high in California as in Texas, a state that has invested relatively little in mass transit. Yet, despite this high demand, California has relatively few lane miles of freeways and arterials. The result is very high levels of travel per freeway mile, almost 30 percent higher than in any of California’s four comparison states. Urban highways are generally used more intensively than rural highways. Thus, combining miles traveled and lane miles in urban and rural areas may be misleading, particularly when comparing states with different percentages of urban and rural highways. For example, in Texas, 72 percent of miles traveled are in urban areas, but only 51 percent of freeway lane miles are in those areas (Table 8.16). In contrast, 82 percent of California’s vehicle miles occur in urban areas, and 71 percent of its lane miles are in urban areas, a closer match between capacity and demand. That closer match notwithstanding, freeways and arterials in both urban and rural areas are used more intensively in California than in any other comparison state. In urban areas, travel per lane mile is at least 18 142 Table 8.16 Traffic on Freeways and Other Principal Arterials in Urban and Rural Areas, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles in Urban Areas (%) 71 41 54 52 64 51 Vehicle Miles in Urban Areas (%) 82 63 69 73 80 72 Annual Vehicle Miles per Lane Mile Urban Areas Rural Areas 4,209,465 3,148,986 3,560,831 3,282,720 3,338,413 3,088,842 2,192,907 1,339,106 1,894,649 1,283,415 1,497,768 1,242,009 percent higher in California than in any of our comparison states. In rural areas, travel per lane mile is at least 15 percent higher. The heavy usage of California highways surely affects the cost of maintaining them. Some roads are owned and maintained by state governments (Caltrans in the case of California), and the remainder are owned and maintained by local governments. In either case, the cost of maintenance is reflected in current highway expenditures. Table 8.17 shows the division of current expenditures between state highway departments and local governments in 2002. With the exception of Texas, state highway departments spent more than local governments on highways. However, most roads are owned and maintained by local governments. In California, for example, the state maintained 50,000 lane miles of roadway and localities maintained 320,000 lane miles. Table 8.17 Current Highway Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas All Roads and Highways 174 171 129 180 154 128 Local Roads and Highways 77 76 61 63 42 65 State Roads and Highways 97 95 69 118 112 64 143 Current highway expenditures per capita can be decomposed into lane miles per capita and expenditures per lane mile. This decomposition is an application of Eq. (2.1) of Chapter 2, although it has a slightly different interpretation from that equation because lane miles play the role of caseload. That unusual interpretation notwithstanding, the equation is: Expenditures = Lane Miles  Expenditures Population Population Lane Mile (8.2) The components of this equation are listed in Table 8.18 for local roads and highways. In 2002, California had less than half of the lane miles per capita of the rest of the country. It also spent more than twice as much per lane mile as the rest of the country. The same general relationship holds between California and each of our four comparison states: fewer lane miles per capita and higher current expenditures per lane mile. A similar pattern holds for current expenditures on state-owned highways (Table 8.19). California had the fewest lane miles per capita of any of our comparison states. It also spent the most per lane mile. In all states, current expenditures per lane mile were substantially higher for state roads and highways than for local roads and highways, a difference reflecting the heavier use of state highways. The heavy use of California highways may partly explain its high current expenditures per lane mile, but another factor is California’s Table 8.18 Components of Current Highway Expenditures per Capita on Local Roads and Highways, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 77 76 61 63 42 65 Lane Miles per 1,000 Residents 9.2 24.3 13.2 19.7 10.6 20.8 Expenditures per Lane Mile ($) 8,376 3,133 4,619 3,200 3,955 3,107 144 Table 8.19 Components of Current Highway Expenditures per Capita on State Roads and Highways, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 97 95 69 118 112 64 Lane Miles per 1,000 Residents 1.4 7.0 2.4 3.3 2.0 8.7 Expenditures per Lane Mile ($) 67,023 13,514 28,173 35,395 56,348 7,311 relatively high payroll expenditures per highway employee. In 2002, California highway employees were paid 52 percent more than equivalent employees in other states (Table 8.20). They were also paid more than equivalent employees in any of the four comparison states. Following the procedure described in Chapter 2, payroll for employee is adjusted for location differences using the average salary of Zone 3 workers. Even after that adjustment, highway employees in California were paid considerably more than similar employees in other states. The Census Bureau does not distinguish between payroll expenditures for capital projects and payroll expenditures for maintenance and other current activities. It seems reasonable to assume, however, that payroll expenditures are a higher proportion of current Table 8.20 Annual Highway Payroll Adjusted for Location, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Payroll per Employee 56,163 36,812 34,831 43,832 45,406 31,856 Payroll per Employee Adjusted for Location 49,280 37,486 38,054 42,474 39,642 34,493 145 expenditures than they are of capital expenditures. With that assumption, the percentage of current highway expenditures allocated to payroll must be larger than total payroll expenditures as a percentage of total expenditures (capital plus current). On the other hand, the percentage of current expenditures allocated to payroll cannot be larger than total payroll expenditures as a percentage of current expenditures. For California, that lower limit (total payroll expenditures as a percentage of total expenditures) is 22 percent, and the upper limit (total payroll expenditures as a percentage of current expenditures) is 41 percent. With current payroll expenditures of that magnitude, California’s substantially higher payroll expenditures per employee is another factor in addition to heavy use explaining its higher current expenditures per lane mile. Outcomes Highway use and capacity is a delicate balance. It is a waste of public funds to add lanes to an uncongested freeway. However, as traffic on a highway grows, at some point travel speeds begin to slow, and the commute from home to work becomes more time-consuming. California freeways and arterials are used intensively, but does the level of use significantly extend the commuting time of the average California resident? Traffic delay is certainly significant on the freeways surrounding California cities. On 40 percent of the state’s urban freeways, traffic moved at 35 miles per hour or less during peak periods of a typical workday in 1998 (Legislative Analyst’s Office, 2000). Traffic delay amounted to over 400,000 hours per day. According to estimates from the Texas Transportation Institute (Schrank and Lomax, 2005), traffic delay in California also tends to be higher than delay in other states. It is important to note that the institute does not directly measure the speed at which cars actually travel on urban highways. Instead, it estimates travel delay with a numerical algorithm based on the ratio of traffic volume to the supply of roadway. Using observations of roadways in many situations, the institute has estimated a relationship between that ratio and the average speed that cars travel. That relationship is flat until the ratio reaches a level at 146 which congestion starts to occur. As the ratio increases beyond that point, travel speed falls steeply. The institute is able to use this relationship and data on traffic volume to estimate average speeds on roadway segments within urban areas. It then compares these estimates to speeds if there were no congestion. The difference is then aggregated to the level of the urban area for 85 urban areas throughout the country. The institute makes other assumptions in its algorithm. For example, it assumes a relationship between traffic volume and the percentage of travel that is congested.2 All of these assumptions can be criticized and surely apply better in some areas than in others. On the other hand, the institute is the only source of information about traffic delay that applies a consistent set of measurements and estimates to a large number of American cities. Its methods are transparent and its data readily available. Appendix Table A.1 lists the urban areas included in California and our comparison states that are included in the institute’s study. Nine of the institute’s 85 urban areas are in California. Residents of these areas tend to use mass transit more than residents of Texas and Florida cities but not as much as residents of Illinois and New York cities (Table 8.21). Despite their more frequent mass transit use, residents of California cities travel as many vehicle miles on average as residents of Florida and Texas cities. Californians have fewer lane miles per resident, however, so the ratio of vehicle miles to lane miles is higher in California cities than in Florida and Texas cities. It is also higher than in Illinois and New York cities. Cities in those two states have fewer lane miles per capita than California cities, but their residents travel less by automobile, presumably because of the extensive mass transit systems in Chicago and New York. According to the institute’s estimates, California’s high ratio of highway usage to capacity leads to substantially slower speeds and extensive travel delay. For the nine California cities, annual delay per _____________ 2For example, the volume of traffic per lane mile in Los Angeles is 52 percent higher than in Jacksonville, Florida. The institute’s algorithm assumes that this difference in volume increases the percentage of traffic incurring congestion from 42 percent to 50 percent. 147 Table 8.21 Road Congestion and Delay in 85 Urban Areas, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles of Freeways and Daily Annual Annual Mass Daily Vehicle Arterials Vehicle Hours of Transit Trips Miles Traveled per Million Miles per Delay per Capita per Capita Residents Lane Mile per Capita 51 16 1,388 11,803 39 59 14 1,327 10,293 22 18 15 1,584 9,355 25 73 12 1,033 11,362 29 169 10 923 10,544 27 24 17 1,576 10,582 30 capita is estimated to be 40 hours, one-third higher than in Texas and Florida cities. Annual delay in California cities is also significantly higher than in New York and Illinois cities, although those comparisons may be misleading. The institute estimates only highway delays. It does not estimate the lengthy travel time that sometimes accompanies travel by mass transit. Traffic delay can vary considerably across cities within the same state. For urban areas with more than one million residents, Figure 8.6 plots hours of delay per capita against the ratio of miles traveled to lane miles. Los Angeles and San Jose experienced the most delay. San Francisco ranked with Dallas and Houston. New York and Chicago have less delay, but, to repeat, the estimates do not include travel time for mass transit commuters. Conclusions The current congestion on California’s freeways can be traced to policy decisions made during the 1970s. During that period, California began to invest less than other states on highways. This trend has continued through the 1980s and 1990s. California’s lower investment in highways and its rapid population growth have led to a decline in highway capacity relative to population. 148 60 • San Jose •50 Los Angeles Hours 40 •Dallas Riverside– • San Francisco • Houston • •30 San Bernardino Orlando •Tampa Chicago • San Diego • •New York • •20 San Antonio Sacramento 10 0 6,000 • Buffalo 8,000 • Miami • Jacksonville 10,000 12,000 14,000 16,000 Daily vehicle miles traveled per lane mile Figure 8.6—Annual Hours of Delay per Capita in the Largest Urban Areas, 2002 Over the same period of time, California has steadily increased real spending per capita on mass transit. Perhaps as a result, mass transit use is somewhat higher in California than in other states. However, despite this mass transit use, automobile travel is high in California. The average Californian travels as many automobile miles as the average Texan or Floridian. This high demand for automobile travel has run up against lower and lower highway capacity per resident. The result is relatively high levels of congestion and traffic delay. The heavy use of California highways may also partly explain its relatively high current expenditures per lane mile. 149 9. Public Safety California spent $27.2 billion on public safety in 2002. This represented 9 percent of total state and local government expenditures, compared to 7 percent in the rest of the country. This chapter analyzes expenditure differences between California and other states in this area. It finds that California’s total expenditures per capita on public safety are higher than in the rest of the country. Underlying conditions are partially responsible for the state’s higher expenditures, but service provision costs—in particular labor costs—also contribute importantly to California’s spending levels. Staffing levels in police and fire are also lower in California than in comparable states. Expenditures on inspection and regulation activities are considerably higher in California than in the rest of the country. Total Expenditures The U.S. Census Bureau classifies public safety expenditures under four functions: corrections, fire protection, police protection, and protective inspection and regulation. Spending on corrections includes the operation and administration of state prisons, local jails, youth corrections, and probation and parole programs. Fire protection includes only the activities of local governments (cities, counties, and special fire districts). The Census Bureau classifies state fire operations, which focus on forest fire protection, under natural resources (see Chapter 10). Police protection covers all police departments and law enforcement agencies with the exception of special forces, park rangers, fish and game wardens, and enforcement or police who serve specific purposes such as judicial services or correctional functions. It includes activities by state governments as well as a variety of local jurisdictions. The protective inspection and regulation category includes expenditures related to a variety of regulatory activities of state and local governments. Some of these activities include the regulation of utilities, 151 vehicle registration, and consumer protection at the state level, and construction regulation and building inspection at the local level. The components of this category are likely to differ from state to state, depending on the scope of their regulatory needs and activities. Thus, we keep comparisons between California and other states to a minimum. As in other areas, California’s per capita expenditures on public safety were higher than in the rest of the country in 2002 (Table 9.1). Differences were pronounced for corrections and protective inspection and regulation. Fire and police protection expenditures in California were lower than in New York and did not differ much from those in Florida or Illinois. With the exception of fire protection, state and local authorities share public safety responsibilities. Local governments spent roughly 66 percent of the total for this category in California in 2002, the same as in the rest of the nation but less than the local share of public safety expenditures in New York (77%) or Florida (71%). Compared to the rest of the country, California’s public safety expenditures have been high through most of recent history (Figure 9.1). Only New York spent as much per capita during the years 1977–2002. Nationwide, public safety expenditures have been increasing, but California experienced a relative acceleration in the mid-1980s. The fall and rise in total expenditures in the late 1980s resulted mainly from fluctuations in corrections capital expenditures and fluctuations in fire protection spending at the fire district level. The increase in public safety Table 9.1 Public Safety Expenditures per Capita, by Type of Service, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 778 512 610 538 761 464 Corrections 257 181 202 149 244 202 Fire Protection 119 86 111 108 128 69 Police Protection 290 215 263 256 351 191 Protective Inspection and Regulation 111 31 34 24 38 21 152 Expenditures (2002 $) 900 800 700 600 500 400 California 300 U.S.–California 200 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.1—Public Safety Expenditures per Capita, 1977–2002 expenditures in 2002 was also a consequence of several factors: substantially higher inspection and regulation expenditures, growth in nonpayroll costs in corrections and police protection, and increases in payroll costs for fire protection. Relative to total public safety expenditures, corrections have increased in importance both in California and in the rest of the nation. In 1977, 23 percent of public safety expenditures were devoted to corrections in California, compared to 20 percent nationwide. By 2002, that share was 33 percent in California and 35 percent nationwide. Over the same period, California’s police protection expenditures fell from 46 to 42 percent of total public safety spending, whereas the share of public safety expenditures allocated to fire protection fell from 21 to 15 percent. Similarly, in the rest of the nation, shares allocated to police and fire dropped from 21 to 17 percent and 51 to 42 percent, respectively. Corrections State and local expenditures per capita on corrections were higher in California than in the rest of the nation in 2002 (Table 9.2). In California as in other states, state and local authorities share the operational and financial responsibilities of running correctional 153 Table 9.2 Corrections Expenditures per Capita, by Type, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 257 181 202 149 244 202 Current 247 169 193 138 225 190 Capital 10 12 8 12 19 12 institutions. Local jurisdictions run jails and detention centers, and the state administers adult prisons. Both levels of government run juvenile centers and parole and probation programs. In California, the state government spent 59 percent of total corrections expenditures in 2002. The remaining expenditures were fairly evenly distributed across cities and counties. The growth in per capita expenditures on corrections in California outpaced that of the rest of the nation in the mid-1980s (Figure 9.2). Some of our comparison states, notably New York and Florida, had similar experiences. Texas’s expenditures rose above the national average after a particularly strong increase in the 1990s. Fluctuations in California’s spending on corrections between 1977 and 2002 stemmed mainly from capital expenditures. However, their average share of total corrections spending over this period was only 10 percent. Thus, current expenditures explain most of the long-run growth in total expenditures. A basic way to analyze the evolution of total expenditures is through the following decomposition: Expenditures = Inmates  Expenditures Population Population Inmate (9.1) A natural question to ask is whether more sentencing and longer confinement periods explain the growth in per capita corrections spending in California (Mackenzie, 2001; Legislative Analyst’s Office, 2004). The growth of the inmate population—defined as people in prison, youth corrections, probation, and parole—throughout the 154 Expenditures (2002 $) 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.2—Corrections Expenditures per Capita, 1977–2002 country is consistent with this conjecture.1 Nationwide, the number of inmates increased from 2.4 to 4.2 million over the 1983–1990 period, a 75 percent increase. From 1991 to 2002, it increased from 4.4 million to 6.6 million, or by 50 percent. Although California experienced a similar trend, its caseload, or number of inmates per capita, does not differ substantially from that of most other large states (Table 9.3). Texas holds the largest inmate population in our comparison group, but it also has the lowest expenditures per inmate. California’s inmate-to-population ratio was at or above the average in the rest of the nation until the early 1990s, when it experienced a decline. Even though the state’s inmate population increased slightly after the approval of Proposition 184 in 1994, it has remained lower than in the rest of the country in recent years (Figure 9.3).2 _____________ 1We use this definition of inmate population because of limited data on jail inmates for all states during our study period. In California, jail inmates made up only 13 percent of the total correctional population, similar to the nation as a whole at 11 percent, as of 2001. 2Proposition 184, also known as the “three strikes and you’re out” initiative, increased penalties for felons convicted of a third crime. In particular, if a criminal has had one previous serious or violent felony conviction, the mandatory sentence for a 155 Table 9.3 Components of Corrections Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 257 181 202 149 244 202 Inmates per Capita 0.018 0.021 0.023 0.017 0.017 0.032 Expenditures per Inmate ($) 14,222 8,655 8,619 8,555 14,521 6,291 0.025 0.020 Inmates per capita 0.015 0.010 0.005 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.3—Inmates Relative to Population, 1977–2002 Consequently, California’s per capita expenditures are high because of above-average spending per inmate. Since expenditures per inmate drive most of California’s spending on corrections, we will analyze their components in the following way: ______________________________________________________________ second such conviction is doubled. After two violent or serious felony convictions, any further felony, nonviolent or not, will trigger a third strike. The mandatory sentence will then be the greater of: three times the term ordinarily required, 25 years, or a term determined by the court. 156 Expenditures Inmate =  Employees   Inmate  Payroll Employee    (9.2) + Nonpayroll Expenditures Inmate The first component in this equation separates labor inputs and their average costs. Payroll expenditures account for nearly half of all corrections expenditures, except in Illinois and New York, where they constitute 61 and 73 percent of total spending, respectively. The second component—nonpayroll expenditures per inmate—consists of total expenses for the provision of inmate services such as dining, supplies, health care, and job training. Governments might incur these costs directly through their own provision or indirectly through contracts with private enterprises. Nonpayroll costs also include some employee benefits but exclude pension plans.3 California’s corrections staffing ratio is very close to the national average, but its expenditures per employee are significantly higher than in comparable states (Table 9.4). By contrast, in New York, payroll costs are higher than in California, and the staffing ratio is much higher. On the other hand, New York’s nonpayroll costs are much lower. The gap between wages in California and in other states stems in part from the need for all employers in California to compensate workers more in this state. For this group of workers (Zone 3 as described in Chapter 2), employers must pay 14 percent more than their counterparts in other states. However, California governments pay their corrections employees nearly 40 percent more than the average for all other states (Table 9.5). Thus, more than half the difference in corrections payroll between California and other states remains unexplained by general labor market conditions. _____________ 3The Census Bureau tracks these expenditures in a separate retirement category in the insurance trusts sector. It does not distinguish among government functions (such as corrections) in these data. 157 Table 9.4 Components of Corrections Expenditures per Inmate, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Inmate ($) 14,222 8,655 8,619 8,555 14,521 6,291 Payroll Expenditures Employees per Employee per Inmate ($) 0.12 52,279 0.12 37,707 0.11 34,679 0.12 42,302 0.19 55,228 0.10 30,281 Nonpayroll Expenditures per Inmate ($) 7,777 4,151 4,929 3,379 3,893 3,224 Table 9.5 Average Yearly Payroll in the Correctional System, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 52,279 37,707 34,679 42,302 55,228 30,281 Location Adjusted 46,267 38,397 37,889 40,991 48,217 32,788 As mentioned above, nonpayroll costs include employee benefits as well as services to inmates and other operational expenditures. Examples of inmate services are rehabilitation and vocational training programs and health care services (e.g., drug treatment). Other operational expenditures include utilities, leases, and payments to outside contractors. Although our data do not allow us to separate these individual components, nonpayroll costs as a group are considerably higher in California than in the rest of the country. They are even higher than in New York, typically considered a high-cost state. Differences across states in nonpayroll costs may be due to different choices of which supplemental services to provide and how to provide them (e.g., different practices of contracting out). 158 The evidence presented so far suggests that California’s higher expenditures on corrections result mainly from higher operating costs. However, it is undeniable that the correctional system has also faced an increasing workload over the past 30 years (Figure 9.3). This trend raises the question of whether California’s correctional system has sufficient infrastructure to support a very large number of prisoners (e.g., Warren and Reiterman, 2005). The answer to this question depends on which measure of crowding is used. The “rated capacity,” or the number of beds judged appropriate for prison facilities, does not show that California’s large inmate population has induced an overcrowding of facilities. By this measure, California state prisons were slightly under full capacity (98%) in 2002 (Table 9.6). The ratio of inmates to beds officially assigned at each facility was also below that of Illinois or New York. At the local level, California’s jails were more crowded than those in the rest of the nation and all of our comparison states, but they were only slightly above full capacity. These data suggest that anecdotal evidence of overcrowding might be due not only to overall capacity constraints but also to the alternative use of spaces originally conceived for rehabilitation or recreation to house inmates. In addition, there may be limits on availability in certain types of facilities (e.g., maximum security). The ratio of inmates to design capacity, or the original number of beds planned for prison facilities, presents a different picture for state prisons. Here, California state prisons are more crowded than those in other large states (Table 9.6). (Data on design capacity are not available for local facilities.) (See Bailey and Hayes, 2006.) The perception of overcrowding could also reflect an ongoing mismatch between the growth in inmate population and the allocation of resources for capital improvements. A way to assess this conjecture is by comparing cumulative capital outlays on corrections with the growth of the inmate population (Table 9.7). Between 1992 and 2002, California’s investment per additional inmate was not very different from that in the rest of the country. It was far ahead of investment per additional inmate in Texas but below expenditures in Illinois or New York. 159 Table 9.6 Average Occupancy of Correctional Facilities, 2002 California U.S.–California Florida Illinois New York Texas Percentage of Rated Capacity State Prisons Local Jails 98 103 — 102 95 95 143 98 110 75 91 94 Percentage of Design Capacity State Prisons 201 113 126 164 124 91 Table 9.7 Real Capital Outlay per Additional Inmate in State Prisons, 1992–2002 California U.S.–California Florida Illinois New York Texas Increase in State Prison Inmate Population 52,821 422,065 26,908 11,053 5,329 101,536 Real Capital Outlay ($) 4,612,590,963 38,490,431,059 2,532,558,435 1,801,186,458 4,797,594,035 4,270,701,794 Real Capital Outlay per Additional Inmate ($) 87,325 91,196 94,119 162,959 900,280 42,061 Police Protection California spent more than the average of all other states on police protection in 2002 ($290 per capita compared to $215 per capita). However, its spending was on par with other large states—including Florida ($263 per capita) and Illinois ($256 per capita)—and less than New York’s ($351 per capita) (Table 9.8). These expenditures occurred mostly at the local level. In the last 35 years, California has spent more per capita on police protection than the average in the rest of the nation and any other state in our comparison group, with the exception of New York (Figure 9.4). 160 Table 9.8 Police Protection Expenditures per Capita, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 290 215 263 256 351 171 State Local 32 258 33 182 25 238 29 228 30 321 18 154 300 250 Expenditures (2002 $) 200 150 California U.S.–California 100 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.4—Police Protection Expenditures per Capita, 1977–2002 As in the case of corrections, capital expenditures are a modest share of total police protection spending. Therefore, this section focuses on operating costs. To gain some insight into police services, it is useful to look at a simple decomposition of expenditures per capita: 161 Expenditures Population =  Crimes   Population  Employees Crime  Payroll Employee    + Nonpayroll Population (9.3) California’s crime rate and staffing level per crime were close to the average in the rest of the nation in 2002 (Table 9.9). California’s ratio of police staff to the number of crimes contrasts with New York’s, the highest in our comparison group. In California and most comparison states, payroll expenditures represent roughly two-thirds of expenditures per employee. The exception is New York, where payroll takes up 95 percent of total expenditures per employee. Operating costs other than wages, which include employee benefits, have also been high in California. It is unclear whether this is a consequence of state-specific factors. For example, New York’s nonpayroll expenditures per capita have historically been quite low whereas Florida’s are large and closer to those in California. As shown in Table 9.9, wages for police officers and other police protection staff members are higher in California than in the rest of the nation and in all of our comparison states except New York. As in the other professions analyzed in this chapter, this is partially due to a location premium. However, even after taking this premium into account, wages remain higher in New York and California (Table 9.10). In recent years, California’s per capita expenditures for police protection have not grown as fast as in the rest of the nation (1.6% compared to 2.8%), outpacing only Florida’s (0.9%) and Texas’s (1%). This trend has mirrored the decrease in the number of crimes per capita in California, especially in recent years (Figure 9.5). Taken together, Table 9.9 and Figure 9.5 raise questions about California’s resource allocation in police protection. As a consequence of the reduction in the crime rate and a moderate growth in police staffing per capita, the number of police staff per crime has increased. To be sure, staffing levels are not determined in isolation. For most of the past 162 Table 9.9 Components of Police Protection Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Total per Crimes per Employees Employee Capita ($) Capita per Crime ($) 268 0.040 0.072 64,456 203 0.041 0.075 47,455 244 0.054 0.065 45,698 257 0.040 0.089 49,194 315 0.028 0.165 71,942 161 0.052 0.054 40,806 Nonpayroll per Capita ($) 105 67 101 81 17 58 Table 9.10 Average Yearly Payroll for Police Protection, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 64,456 47,455 45,698 49,194 71,942 40,806 Location Adjusted 56,557 48,324 49,927 47,670 62,809 44,184 35 years, California’s crime rate exceeded that in the rest of the country. Like many states, California has experienced fewer reported crimes per capita since the mid-1990s. This trend could be due to any number of factors including a better trained police force as well as changing demographics, longer prison sentences, or a strong economy. In any event, this may be an opportune time to reexamine resources dedicated to staffing versus other priorities within police protection and across state and local government budgets. Fire Protection California’s fire protection expenditures were high relative to those in the rest of the nation in 2002 (Table 9.11). With the exception of 163 Reported offenses 0.070 0.065 0.060 California U.S.–California 0.055 0.050 0.045 0.040 0.035 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Figure 9.5—Reported Offenses per Capita, 1990–2002 Table 9.11 Fire Protection Expenditures per Capita, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures per Capita 119 86 111 108 128 69 Counties 28 11 47 <1 2 1 Cities 73 64 55 79 102 65 Special Districts 18 10 9 29 24 2 Texas, other comparison states also displayed high spending levels in this area. Although all types of local governments (counties, cities, and special districts) provide these services, responsibilities for fire protection are concentrated at the city level. However, the role of cities is less dominant in California than in other states. 164 Expenditures (2002 $) 130 120 110 100 90 80 70 60 California 50 U.S.–California 40 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.6—Fire Protection Expenditures per Capita, 1977–2002 California’s expenditures on fire protection have consistently exceeded spending in the rest of the nation (Figure 9.6).4 Operating costs drive most fire protection expenditures. As a result, we can gain a better understanding of expenditures per capita through their decomposition into staffing levels, payroll, and nonpayroll expenditures (Table 9.12). California has the lowest fire protection staffing ratio within our comparison group, even below that of Texas, where per capita spending is much lower. Higher expenditures per capita in California mainly result from higher expenditures per employee, which mainly consist of payroll. This contrasts with Florida, where expenditures per employee are moderate and staffing ratios are high. As a result, its per capita expenditures are high not only because of the costs of labor but also because Florida maintains a higher level of labor inputs. As in the case of corrections and police protection, the differences in payroll costs are not fully explained by location factors (Table 9.13). _____________ 4The unusually large variation in fire protection expenditures for California between 1988 and 1991 is due to fluctuations in special districts’ expenditures. Most of this variability is attributable to reduced precision in the Census Bureau’s estimates stemming from the nature and size of the sample in those years. 165 Table 9.12 Components of Fire Protection Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Employees Expenditures per 1,000 per Capita ($) Residents 119 0.9 86 1.1 111 1.4 108 1.3 128 1.2 69 1.0 Payroll Expenditures per Employee ($) 80,372 50,324 50,443 55,019 74,590 47,978 Nonpayroll Expenditures per Capita ($) 46 32 41 38 36 22 Table 9.13 Average Yearly Payroll for Fire Protection, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 80,372 50,324 50,443 55,019 74,590 47,978 Location Adjusted 70,522 51,245 55,112 53,314 65,121 51,950 The low-staff/high-wage situation in California is not new. The state’s per capita expenditures exhibited only moderate real growth from 1993 to 2002 (6%).5 This was the slowest pace within our group of states and well below the growth in the rest of the nation. California is also the only state with a decline in the number of employees per capita from 1993 to 2002 (Figure 9.7). On the other hand, expenditures per employee have been higher than in the rest of the nation, primarily driven by increases in payroll. However, their growth has been modest in recent years, averaging 1 percent per year in California. By contrast, in New York, payroll _____________ 5The break in Figure 9.7 occurs because the Census Bureau did not report employment data for fire protection in 1996. 166 Fire protection employees 0.12 California U.S.–California 0.11 0.10 0.09 0.08 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Figure 9.7—Fire Protection Employees per 10,000 Residents, 1993–2002 expenditures have grown by an average of almost 2 percent per year. This trend is similar to the national average excluding California, although above the growth rate in some of our comparison states. Protective Inspection and Regulation This category encompasses several regulatory, oversight, and enforcement activities that do not entail a direct provision of protective services. In California, the largest expenditures in this category are related to regulation of public utilities, licensing and expedition of identifications, and consumer protection. Smaller functions include fraud control, regulation and inspection of medical boards and insurance and financial institutions, and supervision of industrial relations. It is important to realize that because of the residual nature of this Census category, the range of activities it includes for other states is likely to include similar functions as in California as well as many others that differ from one state to another. California’s state government incurred 74 percent of total protection and inspection expenditures in 2002, leaving the rest to cities and counties (Table 9.14). The historical averages for those shares are 61 and 39 percent, respectively. The higher share of state expenditures in 2002 167 Table 9.14 Protective Inspection and Regulation Expenditures, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures per Capita 111 31 34 24 38 21 State Expenditures per Capita 82 20 14 16 25 14 Local Expenditures per Capita 29 11 20 9 13 7 may stem from one-time spikes in the costs of regulating utilities after the energy crisis in 2000–2001. Despite the small share of local expenditures in California in 2002 (26%), the state has had a historically high degree of decentralization of inspection and regulation services relative to the rest of the country. In California, state expenditures in this category consisted of 49 programs distributed among 18 agencies in 2002. Within state activities, the Public Utilities Commission, the Department of Motor Vehicles, and some programs in the Department of Consumer Affairs spent the largest share of expenditures (82%). Smaller state oversight activities were aimed at labor relations, insurance and financial institutions, and alcohol, games, and gambling. Regulatory activities by California’s local governments were mostly under control of city governments. Construction regulation and enforcement constituted 83 percent of total local expenditures. The rest was aimed at building inspection and agricultural regulation and was under the responsibility of county governments. California’s spending in such activities has always been considerably higher than in the rest of the country (Figure 9.8). Expenditures in 2002 were exceptionally high, perhaps as a consequence of additional regulatory burdens stemming from the 2000–2001 energy crisis. In 2000, protective inspection and regulation expenditures amounted to 168 Expenditures (2002 $) 120 California 100 U.S.–California 80 60 40 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.8—Protective Inspection and Regulation per Capita, 1977–2002 $73 per capita, an amount much closer to the state’s historical average expenditures on these activities. It is difficult to determine what has caused these high expenditures in California. One reason could be the higher costs of enforcement, regulation, or oversight. Another possibility is a large number of programs that need to be monitored. If the former reason dominates, California’s spending level could indicate high compliance burdens. Alternatively, the state’s size and economic diversity make a case for a need to oversee a wider set of activities in California. Both explanations are likely to carry some weight. Conclusions California devotes more resources to public safety than many comparable states do. The size of the state’s population and the underlying need for safety services explain some of these expenditures. A closer examination reveals that provision costs are also higher in California than in most of our comparison states. Labor costs are the most important component of operating expenses. The observed gap between wages in California and in other states cannot be completely 169 explained by the need to compensate employees for living in a high-cost state. California’s spending on corrections is high relative to that in the rest of the nation and to our comparison group, even though the number of inmates per capita is among the lowest. California’s high correctional payroll is consistent with high overall operations expenditures. California does not exhibit overcrowding of state prisons in terms of the number of inmates to beds officially assigned at each facility. However, local jails are operating above capacity according to this measure. California’s state prison occupancy also looks much worse compared to other states as a percentage of design capacity. In addition, California’s capital outlay per additional inmate has been slightly below the average for the rest of the nation in recent years. It has also been lower than the infrastructure investment of other states with large confined populations, such as Illinois and New York. Fire and police protection are also labor-intensive activities where California’s labor costs remain higher than our comparison group, except for police officer wages in New York. Payroll expenditures have been the primary source of recent growth in per capita expenditures on these functions. Finally, California’s expenditures in government activities aimed at inspection and regulation are substantially higher than those in the rest of the country. Possible explanations for these high levels of spending include a wider set of activities that need regulation because of the size and diversity of California’s economy, higher regulatory costs, and an inherently stronger emphasis on regulation than in other states. 170 10. Environment and Housing In 2002, California state and local governments allocated 7 percent of total expenditures to environment and housing, compared to 6 percent in the rest of the nation. Environment and housing is a broad category as defined by the U.S. Census Bureau, combining two areas not typically considered together. The environment component includes public spending on agriculture, fish and game, forestry, and other natural resources as well as on parks and recreational facilities. It also includes spending on public services that can affect environmental quality, such as sewerage and solid waste management.1 For similar reasons, we consider these enterprises together with publicly owned water, gas, and electric utilities at the end of the chapter. In 2002, California’s state and local governments spent one-third more than the average for all other states on environment and housing. They spent 50 percent more on housing and community development, even though fewer households received direct housing assistance than in the rest of the nation. California’s higher spending on this function may be due to the higher costs of providing housing assistance in this state or to a greater emphasis on redevelopment rather than housing assistance compared to other states. California spent 80 percent more than the rest of the country on natural resources in 2002. Its total expenditures were comparable to those of Florida and other western states. Natural resources spending net of revenues (including fees from local water agencies in California) was only 30 percent greater than the average for all other states, and it was 40 percent less than in Florida. Per capita expenditures on parks and recreation and enterprise activities including public water, gas, and _____________ 1This category does not include state and local spending on environmental protection, such as monitoring and regulation of air and water quality. These expenditures are captured in the “other health” category discussed in Chapter 6. 171 electric utilities did not differ markedly from those of other states once user fees were subtracted from the total. Total Expenditures Environment and housing is a diverse category as defined by the Census Bureau. The housing and community development function itself comprises several types of programs, including those that build and maintain affordable housing, provide assistance to low-income renters and homebuyers, and revitalize urban and rural areas. Expenditures in this category are predominantly local, although they often include passthroughs of state and federal funds. In particular, this category includes the expenditures of local public housing authorities (PHAs), locally chartered and governed agencies responsible for administering federal housing programs such as the Section 8 rental subsidy program and public housing projects overseen by the U.S. Department of Housing and Urban Development (HUD). It also includes local spending from federal Community Development Block Grants (CDBGs) for neighborhood revitalization. It is worth noting that the largest active form of housing assistance in the United States, the low-income housing tax credit (LIHTC), is not reflected in these data because it represents forgone revenue rather than a direct expenditure.2 Similarly, Census Bureau data do not reflect the proceeds of tax-exempt private activity bonds (PABs), including bonds for the construction of multifamily housing as well as bonds that finance below-market interest rate mortgages or reduced federal tax liability for eligible homeowners (i.e., mortgage revenue bonds and mortgage credit certificates).3 _____________ 2In 2002, the federal government authorized $60 million in tax credits for California. California also runs its own tax credit program, with $70 million authorized in 2002 (California Tax Credit Allocation Committee, 2006). 3In 1999, the federal government authorized $1.6 billion in tax-exempt bonds for private purposes in California, of which the state allocated $1.4 billion to housing. California accounted for nearly 30 percent of the value of mortgage credit certificates issued in 2000 and more than half of the value of these bonds issued from 1992 to 2000 (Greulich and Quigley, 2003). 172 The natural resources component of environment and housing describes government conservation, improvement, and promotion efforts in agriculture, fish and game, forestry, and resources “not elsewhere classified” (NEC). In California, the NEC category corresponds mainly to water resources. Roughly half of this spending occurs at the state level through the Department of Water Resources and several state conservancies (e.g., the State Coastal Conservancy, Santa Monica Mountains Conservancy, Tahoe Conservancy). The remainder is spent by counties and special districts on flood control, drainage, land reclamation, and levee maintenance. In other states, this subcategory may correspond to the protection of water, soil, energy, or mineral resources. Parks and recreation is the umbrella term for recreational, cultural, or scientific facilities. The bulk of spending in this area comes from local parks, golf courses, public beaches, museums, zoos, and convention centers. At the state level, this category includes expenditures of the state Department of Parks and Recreation as well as the California Science Center and Arts Council. Sewerage refers to government financing of sanitary sewers and storm water facilities. Sanitary sewers dispose of wastewater from buildings, whereas storm sewer systems channel runoff from storm drains. Solid waste management includes garbage collection and disposal as well as recycling, toxic cleanup, and street cleaning. The Census Bureau records expenditures on these activities whether they are performed by the government itself or through a private contractor. Finally, the water, gas, and electric utility functions comprise all government activities relating to the production, acquisition, and distribution of these resources to the general public or to other public or private utilities. In 2002, California state and local governments spent $20.6 billion on environment and housing combined, translating into $590 per capita (Table 10.1). The bulk of these expenditures (97%) occurred at the local level, although they often involved state and federal funds. California spent one-third more than the rest of the nation on this category but a similar amount to other large states including Florida ($539), Illinois 173 Table 10.1 Environment and Housing Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 590 448 539 548 566 336 Housing and Community Natural Parks and Solid Development Resources Recreation Sewerage Waste 154 126 119 113 77 104 69 102 108 65 71 144 110 103 111 142 51 201 118 37 197 24 105 117 124 72 46 94 82 42 ($548), and New York ($566). By contrast, California spent 75 percent more than Texas ($336). The pattern has persisted in California since 1977. Expenditures increased in relative terms during the late 1980s and mid-1990s, mainly because of increases in housing and community development expenditures. Among other large states, California’s spending over time has been similar to that of New York, whereas Florida, Texas, and Illinois more closely resemble the rest of the nation. The diversity of programs included in this category limits the usefulness of further comparisons of total expenditures across states. We therefore move to a discussion of the major components of this category, starting with housing and community development and then turning to natural resources, parks and recreation, and enterprise activities including sewerage, solid waste, and other utilities. Housing and Community Development In 2002, California state and local governments allocated $5.4 billion to housing and community development, equivalent to $154 per capita or about 50 percent more than the average for all other states (Table 10.2). Compared to other large states, California spent roughly the same amount per capita as Illinois, about 20 percent less than New York, and more than twice as much as Florida or Texas. Capital spending represented one-third of total spending on housing and community development, a similar proportion as in other states except 174 Table 10.2 Housing and Community Development Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 154 104 71 142 197 72 Current 103 83 64 92 153 50 Capital 51 20 7 50 44 22 Florida. However, these amounts fluctuate over time, as discussed below. California has not always been a “high-spending state” on housing and community development. It was at or below the average for all other states during the 1970s and early 1980s (Figure 10.1). Spending doubled during the 1980s and grew another 30 percent in the early 1990s before dropping sharply in 1997. Among our comparison states, New York exhibited a similar pattern, although accelerated by a few years, with expenditures peaking in the early 1990s. 200 180 160 140 120 100 80 60 California 40 U.S.–California 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.1—Housing and Community Development Expenditures per Capita, 1977–2002 Expenditures (2002 $) 175 As with all functional categories defined by the Census Bureau, it is not possible to further disaggregate spending on housing and community development by program. It is clear from Table 10.3, however, that total spending in this area is driven largely by federal grants. In 2002, for example, federal intergovernmental revenue accounted for over half of all California spending on housing and community development and an even greater share of spending in all of our comparison states, reaching a high of 95 percent in New York. To better understand the drivers of housing and community development expenditures in California and other states, we can apply our standard decomposition framework from Chapter 2: Expenditures = Caseload  Expenditures Population Population Case (10.1) and Caseload = Participants  Eligibles Population Eligibles Population (10.2) Identifying the caseload for spending on housing and community development is complicated by the variety of programs in this area. Cases may include families receiving rent subsidies under federal housing programs as well as those benefiting from local planning or redevelopment efforts. Table 10.3 Housing and Community Development Federal Intergovernmental Revenue per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 154 104 71 142 197 72 Federal Intergovernmental Revenue 84 78 51 109 188 45 176 For major federal housing assistance programs, eligibility is restricted to “low-income” households, defined as those with earnings at 80 percent of the local median income. There are also adjustments for family size and incentives for local authorities to target “very” and “extremely” low-income families at 50 and 30 percent of the area median income, respectively. For local redevelopment efforts supported by federal funds, the Community Development Block Grant program requires that funded activities benefit areas with a concentration of lowincome individuals.4 Table 10.4 shows that there is little variation by state in the proportion of households meeting these eligibility criteria. Where states do differ is in the proportion that actually receives federal, state, or local rental subsidies or resides in federally supported public housing. Although participation rates are generally quite low, they are particularly low in California. In 2002, 8.3 percent of all eligible households in California received rent subsidies or lived in public housing, compared to 11.5 percent in the rest of the nation. Moreover, California’s participation rate was lower than that of all other large states in our comparison group except Texas. Although figures differ somewhat from year to year, California was below the average for all other states in the Table 10.4 Housing Assistance Caseloads and Components, 2002 Total Households Recipients of Low-Income Housing Families as Assistance as Percentage of Percentage of Total Total Participation in Housing Assistance as Percentage of Eligible California 11,935,960 41.8 3.5 8.3 U.S.–California 97,452,805 41.4 4.8 11.5 Florida 6,683,618 39.6 3.8 9.5 Illinois 4,836,880 42.0 4.7 11.2 New York 7,294,127 43.4 7.9 18.3 Texas 7,493,242 43.5 2.6 6.0 _____________ 4Similarly, in California, Redevelopment Agencies (RDAs) must spend 20 percent of their tax increment financing (i.e., increases in property tax revenues above inflation) on low- to moderate-income housing. 177 share of all households receiving housing assistance from 1977 to 2002 (Figure 10.2). Unlike other forms of public support, housing assistance is not an entitlement. Rather, households must apply for aid and there are generally many more applicants than subsidies available. Local PHAs determine who receives federal assistance within HUD guidelines. Table 10.5 suggests that it is unlikely that California’s lower participation rate in rent subsidy and public housing programs stems from a lack of demand. In fact, California includes more households with “moderate” to “severe” rent burdens (defined by HUD as households paying rents in excess of 30 and 50 percent of their income) than in the rest of the nation and in all other large states except New York. Moreover, according to the most recent data available, California residents who are eventually accepted into federal programs tend to spend a longer time on waiting lists than the average for the rest of the nation and in many of our comparison states (Table 10.6). California’s higher rent burdens also suggest that the cost of subsidizing each household is higher. This is evident in the Fair Market 6.0 5.0 4.0 3.0 2.0 California 1.0 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Figure 10.2—Percentage of All Households Receiving Rent Subsidies or Public Housing, 1977–2002 178 Percentage Table 10.5 Rent Burdens as a Percentage of Household Income, 2000 California U.S.–California Florida Illinois New York Texas Renters Paying 30 to 50 Percent of Income 22.1 18.7 21.3 18.0 18.9 18.6 Renters Paying More Than 50 Percent of Income 20.2 17.2 19.6 17.3 21.6 15.7 Table 10.6 Average Months on Waiting List for Housing Program Participants, 1998 California U.S.–California Florida Illinois New York Texas All Programs 33 22 28 40 29 22 Section 8 Certificates and Vouchers 37 28 36 63 34 31 Public Housing 18 11 16 9 10 11 Rents (FMRs) used by HUD in calculating rent subsidies under the Section 8 program.5 Table 10.7 confirms that FMRs are higher for nearly all apartment sizes in California than in other large states. The average rent for a two-bedroom unit in 2002 was $934 in California, compared to $673 in the rest of the nation and $903 in the next highest rental market, New York. It is worth noting that these higher rents may reflect underlying differences in the quality of the housing stock apart from apartment size. _____________ 5FMRs are defined as the 40th percentile of rents in a metropolitan area or nonmetropolitan county (or the 50th percentile in a few metropolitan areas). 179 Table 10.7 Statewide Average Fair Market Rents Weighted by Population in Each Market Area, 2002 (dollars) California U.S.–California Florida Illinois New York Texas 01 2 3 4 Bedrooms Bedroom Bedrooms Bedrooms Bedrooms 639 746 934 1,283 1,484 459 547 673 886 1,021 478 560 686 916 1,068 535 639 774 978 1,099 661 768 903 1,161 1,296 434 499 635 874 1,023 In 2002, capital outlay on housing and community development amounted to $51 per capita in California and $20 per capita in the rest of the nation. However, because capital expenditures differ significantly from year to year and contribute to a preexisting stock, a more appropriate measure would consider spending over a longer period as a function of growth in demand, measured here as population growth. Capital spending also depends on the age of the housing stock. Presumably, states with an older stock will need to spend more because of depreciation regardless of the population growth rate. From 1992 to 2002, California spent $14.5 billion on capital outlay for housing and community development.6 This translated to $3,610 per additional state resident, more than twice that in the rest of the nation but less than Illinois ($4,460 per additional person) and New York ($8,049 per additional person) (Table 10.8). In terms of the age of the housing stock, California is at the midpoint of our comparison group with the median housing unit being built in 1970, compared to 1962 in Illinois, 1954 in New York, 1977 in Texas, and 1980 in Florida. In sum, California spent 50 percent more than the average for all other states on housing and community development, although a similar amount as other large states including Illinois and New York. California served fewer households than other states, however, and this may be due to higher costs per subsidized household. However, this spending _____________ 6These expenditures do not reflect the proceeds of a $2.1 general obligation bond, authorized with the passage of Proposition 46 in November 2002. 180 Table 10.8 Housing and Community Development Real Capital Outlay, 1992–2002 California U.S.–California Florida Illinois New York Texas Cumulative Real Capital Outlay (2002 $) 14,485,723,956 44,854,701,852 1,124,716,237 3,973,465,154 7,276,633,908 2,532,155,424 Change in Population 4,013,000 27,411,000 3,030,000 891,000 904,000 3,963,000 Real Capital Outlay per Additional Person (2002 $) 3,610 1,636 371 4,460 8,049 639 category also encompasses local redevelopment and planning efforts, and there may be a greater emphasis on these efforts in California. Unfortunately, there are not sufficient data on redevelopment spending to distinguish between these two explanations. Natural Resources In 2002, California spent $4.4 billion overseeing its natural resources, including agriculture, fish and game, forestry, and “other” resources not elsewhere classified, which in California corresponds mainly to water resources. At $126 per capita, these expenditures were 80 percent higher than the average for the rest of the nation and significantly greater than expenditures in all of our comparison states except Florida, which spent 15 percent more than California in 2002 (Table 10.9). Spending in this category clearly depends on a state’s endowment of natural resources. For example, it is not surprising that California spent more in 2002 than did any of our comparison states on forestry, which includes spending on forest fire prevention, control, and suppression efforts. Sparsely populated states with rich oil and mineral resources as well as higher fish and game, forestry, and agriculture expenditures such as Alaska, Wyoming, Montana, and North Dakota tend to dominate 181 Table 10.9 Natural Resources Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 126 69 144 51 24 46 Agriculture 27 21 30 16 6 13 Fish and Game 10 10 9 4 2 7 Forestry 14 4 4 1 1 1 Other Resources 75 35 100 30 14 25 expenditures per capita in this category. California and Florida occupy a second tier along with other western states, including Arizona ($76 per capita in 2002), Nevada ($99), Utah ($93), and Washington ($118). The variety of programs included in the “other” category makes further comparisons difficult. For example, in California, roughly onethird of all “other” natural resources spending corresponds to the State Water Project (SWP), or the system of reservoirs, aqueducts, and pumps that transport water from the northern to the southern regions of the state. The Census Bureau categorizes the SWP under water resources because of its role in irrigation, drainage, flood control, soil conservation, and reclamation, even though the SWP also acts as a wholesaler to local water agencies, whose activities are recorded in the “water supply” category discussed below.7 In other western states, the “other” category probably also corresponds to water resources management activities, but we have no way to know for certain. In Florida, this category encompasses spending on an ambitious state and federal Everglades restoration program begun in the mid-1990s. To properly account for the public subsidy to natural resources, we must calculate spending net of user charges, such as fees paid to the State Water Project or local flood control districts. After subtracting these _____________ 7Apart from the SWP, two major federal water conveyance and storage systems— the Central Valley Project (CVP) and Colorado River Project—and several smaller Army Corps of Engineers projects provide water to local utilities in California. Expenditures of these projects are not reflected in Census of Governments data. However, the presence of federal projects may have the effect of lowering state and local government outlays for water in California and other western states. 182 revenues, California’s net spending on natural resources was only 30 percent higher than the average for all other states in 2002. California spent 2.5 times as much as New York, nearly twice as much as Texas, and two-thirds more than Illinois. However, California spent 40 percent less than Florida in 2002 (Table 10.10). The state spent one-third more than Arizona and about 90 percent more than Colorado on a net expenditure basis, although a comparable amount to Nevada, Utah, and Washington in 2002. Natural resources expenditures may differ considerably from year to year (Figure 10.3). For instance, both gross and net expenditures dropped by 40 percent during the early 1990s, perhaps as a result of declining water resource management responsibilities following the drought of 1987 to 1992. Nevertheless, California’s net spending on natural resources has consistently exceeded the average for all other states and that of all of our comparison states except Florida since 1977. It has been within the range of other western states over this period. Compared to its own historical record, California used to spend more on natural resources. Notably, in 1960, voters approved a $1.75 billion general obligation bond ($8.2 billion in 2003 dollars) to finance construction of the SWP (de Alth and Rueben, 2005). More recently, state funds for water resources management have increased with the passage of Propositions 40 and 50, which provide funds for ecological restoration and water supply management among other priorities. Table 10.10 Natural Resources Net Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures 126 69 144 51 24 46 Fees and Charges 44 6 2 1 0 3 Net Expenditures 83 64 142 50 23 44 183 Expenditures (2002 $) 100 90 80 70 60 50 40 30 20 California U.S.–California 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.3—Natural Resources Net Expenditures per Capita, 1977–2002 Parks and Recreation California spent $4.2 billion on parks and recreation in 2002. Nearly two-thirds of this spending occurred at the city level. The remainder was equally distributed between other types of local governments (counties and special districts) and the state (including local assistance). At nearly $120 per capita, California’s expenditures on parks and recreation were 16 percent higher than the average for all other states but similar to other large states, including Florida (whose expenditures were 8% lower than California’s), New York (12% lower), and Texas (26% lower) (Table 10.11). In contrast, Illinois spent nearly 70 percent more than California on parks and recreation in 2002. Capital expenditures represented approximately one-quarter of this total in California as in the rest of the nation and our other comparison states except Texas. In addition to spending more than the average for all other states, California also brings in more revenue from its parks and recreational facilities. In 2002, California state and local governments recouped nearly 30 percent of total expenditures from park entrance fees, parking 184 Table 10.11 Parks and Recreation Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 119 102 110 201 105 94 Current 91 70 80 132 80 49 Capital 28 32 30 69 25 46 fees, campground fees, and other user charges.8 Illinois captured a similar share of expenditures in 2002, whereas in other large states, revenues offset expenditures by around 15 to 20 percent (Table 10.12). Subtracting these fees and charges, California’s net expenditures on parks and recreation were $86 per capita in 2002, only 10 percent higher than in the rest of the nation. California used to spend as much as 80 percent more per capita than the average for all other states, but this gap has been generally narrowing over time (Figure 10.4). California’s higher net spending on parks and recreation after 2001 reflects an influx of new bond funds. For example, the passage of Proposition 12 in Table 10.12 Parks and Recreation Net Expenditures per Capita, 2002 (dollars) Fees and Net Expenditures Charges Expenditures California U.S.–California Florida Illinois New York Texas 119 102 110 201 105 94 33 86 23 79 23 87 61 139 23 81 13 81 _____________ 8This figure does not include revenues from hunting and fishing licenses, which contributed $2 per capita in 2002 in California and similar amounts in other states. 185 Expenditures (2002 $) 100 90 80 70 60 50 40 30 20 California U.S.–California 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.4—Parks and Recreation Net Expenditures per Capita, 1977–2002 March 2000 provided funds for land acquisition and local assistance grants. California’s roughly average expenditures per capita in this category are perhaps surprising in light of the state’s reputation as home to several landmark parks and recreation areas. One way to evaluate these expenditures is as a function of acres provided by state and local governments, similar to the decomposition of expenditures per capita into caseload and expenditures per case in Chapter 2: Expenditures = Acres  Expenditures Population Population Acre (10.3) California leads our comparison group in absolute state and local park acreage, although New York’s parks represent a larger share of nonfederal land (Table 10.13). After adjusting expenditures for land area in state and local parks, California spent one-third more than New York, a comparable amount to Florida and Texas, and about half as much as Illinois (Table 10.14). Illinois’ higher expenditures per acre may be due to the inclusion of museums in this category or to a greater presence of high-amenity parks such as golf courses. 186 Table 10.13 Federal, State, and Local Parks and Natural Resource Lands, 2002 (thousands of acres) California Florida Illinois New York Texas Local Parks 600 324 279 186 340 State Parks and Federal Parks, Natural Resource Forests, and Lands Wildlife Refuges 1,457 31,691 592 4,982 306 3,764 1,159 70 668 970 Total Land Area in State 99,814 34,513 35,573 30,217 167,550 In addition to state parks, California is home to over 31 million acres of federal parks, forests, and wildlife refuges, more by far than any of our comparison states. In 2002, California received more visitors per capita to its national parks than any other state. California also includes private alternatives to public recreation areas, such as private land trusts or land protected by conservation easements. According to the Land Trust Alliance (2004), California led the nation with 1.5 million acres covered through these arrangements in 2003. Within our comparison group, New York had 571,000 privately protected acres, Texas had 194,000, Florida 79,000, and Illinois 52,000 in 2003. Perhaps as a result of these alternatives, visits per capita to California state parks (including both state residents and nonresidents) lag those of Illinois and New York (Table 10.15). Table 10.14 Parks and Recreation Expenditures per Acre, 2002 California Florida Illinois New York Texas Total Expenditures ($ 1,000s) 4,172,705 1,836,837 2,523,951 2,006,117 2,050,467 State and Local Parks (1,000 acres) 2,057 916 585 1,345 1,008 Expenditures per Acre ($) 2,029 2,005 4,312 1,491 2,034 187 Table 10.15 State and National Park Visits per Capita, 2002 California U.S.–California Florida Illinois New York Texas State Parks 2.53 2.72 0.93 3.51 2.73 1.07 National Parks 0.98 0.85 0.29 0.03 0.25 0.32 Sewerage, Solid Waste, and Utilities California’s per capita expenditures on sewerage and solid waste public enterprises were comparable to the rest of the nation and many other large states in 2002 (Table 10.16). Its public water, gas, and electric utilities spent nearly three times more than the rest of the nation as a group. California’s public electric utility expenditures were particularly high in 2002 because of the state energy crisis.9 Table 10.16 Utility Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Sewerage 113 108 103 118 117 82 Solid Waste 77 65 111 37 124 42 Water 231 128 168 140 98 173 Electric 401 166 188 36 251 171 Gas 3 22 13 5 1 14 _____________ 9In January 2001, the state Department of Water Resources (DWR) purchased roughly $10 billion of electricity on the wholesale market and later resold it to investorowned utilities for approximately $3 billion. These purchases were underwritten by a loan of approximately $6 billion from the state General Fund and a $4 billion interim loan. Although the DWR eventually repaid this borrowing with proceeds from a bond secured by future rate payments, it appeared in the fiscal year 2002–2003 budget as a $6.9 billion expenditure from “unclassified state funds.” 188 Excluding electric utilities, expenditures net of user charges for these activities are quite low. For example, California state and local governments spent only $7 per capita on sewerage beyond user fees in 2002, compared to $13.5 nationally not including California (Table 10.17). Table 10.17 Net Enterprise and Utility Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Sewerage 7 14 6 52 40 –9 Solid Waste 25 28 21 27 101 9 Water 31 23 38 24 19 31 Electric 38 1 2 –6 12 0 Gas 0 0 –1 0 0 1 Conclusions California’s higher than average spending on environment and housing is evident in each of its major subcategories. California resembles Illinois and New York in its housing and community development expenditures per capita and is close to but lower than Florida and other western states in its natural resources spending. Expenditures on parks and recreation were comparable to those of other states once user fees were subtracted from the total. Similarly, net spending on sewerage and solid waste enterprise activities were relatively small and similar to other states. 189 11. Governmental Administration The previous seven chapters have analyzed specific government services, such as police protection, transportation, and housing. In the typical case, these services are the responsibility of a department or agency of a general government such as a city, county, or state. This government then has the responsibility of overseeing and supporting these departments and agencies. In the case of the city, for example, a city manager may supervise all the city’s departments, and these departments may receive support services from the city’s personnel and finance departments. The U.S Census Bureau classifies these supervising and supporting activities as governmental administration. It also includes the legal and judicial activities of government in this category. In 2002, governmental administration accounted for 5 percent of state and local expenditures in California and 4 percent in the rest of the country. California’s relatively high spending per capita in this area was due to both high nonpayroll expenditures and high payroll expenditures. California’s high payroll expenditures were not due to significantly higher employees per capita but rather to high average salaries, even after adjusting for the generally higher level of salaries and wages in the California labor market. Total Expenditures Governmental administration includes five Census Bureau functions: financial administration, central staff services, judicial and legal, general public buildings, and legislative. Because financial administration and central staff services are similar in nature, we combined them into an area labeled “finance and central staff.” At the state level in California, the five departments with the largest expenditures in this area are the Franchise Tax Board, the Department of 191 Motor Vehicles (DMV), the Department of General Services, the Public Employees’ Retirement System, and the State Board of Equalization.1 At the local level, the finance and central staff function encompasses the offices of county administrators, auditors, assessors, recorders, supervisors, and treasurers. At the city level in particular, this area includes the offices of city mayors, managers, councils, and clerks. It also includes offices of planning and zoning. However, administrative activities particular to a specific government service are excluded from this area. For example, the administration of a city’s police department would be included in public safety, not finance and central staff. For similar reasons, this area excludes administrative functions of singlepurpose special districts, such as water, sewer, fire, and mosquito abatement districts. The “judicial and legal” area comprises the courts, public prosecutors, public defenders, and legal counselors to city, county, and state government. In California, this area encompasses the 58 state trial courts located in each county, the six Courts of Appeal, the Supreme Court, the Attorney General’s Office, and the state Department of Justice. At the city and county level, it includes district attorneys, public defenders, and grand juries. The “general public buildings” category includes the construction and operation of buildings not related to specific functions and agencies. Examples are courthouses, city halls, and city and county administrative buildings. Excluded from this category are schools, police stations, firehouses, libraries, jails, and hospitals—buildings associated with specific government services. The legislative function encompasses state legislatures and related research and investigative agencies. In California, it includes the Assembly and Senate as well as the Office of Legislative Counsel. The category excludes similar activities at the city and county level, such as city councils and county boards of supervisors. These activities are _____________ 1Drivers’ licensing and other public safety aspects of the DMV are included in Chapter 8, whereas general government support such as Vehicle/Vessel ID and Compliance is included here. 192 included in central staff services, which we have combined with financial administration as previously discussed. In 2002, per capita spending on governmental administration was significantly higher in California than in other states (Table 11.1). Spending per capita in California was 47 percent higher than in the rest of the country. It was more than twice as high as in Texas, and even 15 percent higher than in New York, the next highest spending state in this area. Administrative spending per capita in California has exceeded spending per capita in other states for every year from 1977 to 2002 (Figure 11.1). Even at its lowest level in relative terms, in 1983, California’s spending per capita was 25 percent higher than the average for all other states. In California and its comparison states, finance and central staff represent the largest area of administrative spending. Legal and judicial spending is also significant in every state. On the other hand, legislative expenditures and expenditures on general public buildings are relatively small in every state—the two areas together constitute less than 20 percent of administrative spending. As a consequence, these two areas cannot be an important explanation of why administrative expenditures differ so widely across states, and we do not consider them further in this chapter. Table 11.1 Governmental Administration Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 447 305 293 330 390 208 Finance and Central Judicial Staff and Legal 216 201 161 96 160 99 198 88 178 143 97 75 General Public Buildings 21 40 26 36 59 30 Legislative 8 8 9 7 11 5 193 Expenditures (2002 $) 500 450 400 350 300 250 200 150 California 100 U.S.–California 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.1—Governmental Administration Expenditures per Capita, 1977–2002 Finance and Central Staff Spending per capita on finance and central staff has been higher in California than in the rest of the country for every year from 1977 to 2002 (Figure 11.2). The lowest margin occurred in 1998, when spending per capita in California was 17 percent higher than in other states. The highest, 48 percent, occurred in 1978. In 2002, expenditures per capita on finance and central staff were 35 percent higher in California than in other states. Expenditures in this area are primarily current rather than capital expenditures. In 2002, 95 percent of California’s total finance and central staff expenditures were current expenditures, a similar proportion as in other states. Although the mix between local and state expenditures differs widely among states, in most states at least half of finance and central staff expenditures occurred at the local level. In California, this percentage was 54 percent. Among our comparison states, Texas had the highest percentage at 71 percent, and New York had the lowest at 44 percent. 194 250 200 Expenditures (2002 $) 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.2—Finance and Central Staff Expenditures per Capita, 1977–2002 The activities in the area of finance and central staff are laborintensive, which implies that payroll expenditures are a large portion of total expenditures (Table 11.2). In California, payroll expenditures were 52 percent of the total. This percentage was even higher in Florida, New York, and Texas. In contrast, payroll expenditures were only 44 percent of total expenditures in Illinois. In general, payroll expenditures understate personnel costs because they do not include the cost of employee benefits, which are included in nonpayroll expenditures. California’s relatively high spending in the area of finance and central staff is due to both high payroll expenditures per capita and high Table 11.2 Components of Total Expenditures per Capita on Finance and Central Staff, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 216 161 160 198 178 97 Payroll 113 90 88 85 111 67 Nonpayroll 103 71 72 113 67 30 195 nonpayroll expenditures. Payroll expenditures per capita in California were 25 percent higher than in other states. They were also higher than in the four comparison states. Nonpayroll expenditures per capita were 45 percent higher in California than in other states and higher than in all of our comparison states except Illinois. Following the general approach outlined in Chapter 2, payroll expenditures per capita can be decomposed into two components: payroll per employee and employees per capita. Specifically, applying Eq. (2.3): Payroll = Payroll  Employees Population Employee Population (11.1) California had slightly fewer employees per capita than the rest of the country, but it paid its employees much more generously than did other states (Table 11.3). Equation (2.4) in Chapter 2 demonstrates how the difference in payroll expenditures per capita can be attributed to each of these two components, where  represents the percentage difference in a variable between two states or between California and all other states:  Payroll   Payroll +  Employees Population Employee Population (11.2) Table 11.3 Components of Finance and Central Staff Payroll Expenditures, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Capita ($) 113 90 88 85 111 67 Employees per Payroll per 10,000 Employee ($) Residents 51,450 22 39,451 23 40,528 22 41,853 20 45,392 24 37,332 18 196 Applying this equation, in 2002, payroll expenditures per employee were 30 percent higher in California than in other states (the second term in Eq. (11.2)). These higher expenditures per employee were partially offset by a lower ratio of employees to population. Employees per capita (the third term in Eq. (11.2)) were 9 percent lower in California than in other states. As a consequence, payroll expenditures per capita (the first term) were 25 percent higher in California than in other states. California’s high average payroll for employees in finance and central staff is partly due to the generally high level of salaries and wages in the California labor market. Even after adjusting for these labor market conditions, however, California’s average payroll in this category is higher than the average in other states. The first column of Table 11.4 lists the average payroll for employees in finance and central staff. In the second column, average salaries are adjusted for labor market conditions using the method described in Chapter 2. In this particular case, these adjustments were made using the average salaries of Zone 3 workers, who have the education and skills typical of administrative employees in business and government. Before adjusting for labor market conditions, average payroll in California was 30 percent higher than in the rest of the country. After those adjustments, it is 12 percent higher. Thus, labor market conditions account for a little more than half of California’s relatively high average payroll for finance and central staff employees. Table 11.4 Average Payroll for Finance and Central Staff Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Payroll 51,450 39,451 40,528 41,853 45,392 37,332 Average Payroll Adjusted for Location 45,145 40,173 44,279 40,556 39,629 40,423 197 We do not have specific measures of outcomes or performance in the area of finance and central staff. In large part, this is due to the indirect nature of many of the activities in this area. An efficient human resources department is a key part of a well-run city, but it is inherently difficult to measure the contribution of that department to the effectiveness of any particular government service. Judicial and Legal In 2002, judicial and legal expenditures per capita were twice as high in California as in other states. As Figure 11.3 demonstrates, they have also been higher in California in every year since 1977. In that year, judicial and legal expenditures per capita were $12 per capita higher in California than in the rest of the nation. That margin remained roughly constant through 1981, then increased steadily until 1991, when California spent $112 per capita more than the rest of the country. From 1991 to 1997, this margin increased only slightly. From 1997 on, however, it increased dramatically. In 2002, California spent $181 more per capita than the rest of the country. As in the case of finance and central staff expenditures, judicial and legal expenditures are overwhelmingly current expenditures. In 250 200 150 100 50 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.3—Total Judicial and Legal Expenditures per Capita, 1977–2002 198 Expenditures (2002 $) California in 2002, 97 percent of total expenditures in this area were current expenditures. Other states had similar percentages. Our analysis of judicial and legal expenditures follows the same approach as our analysis of finance and central staff expenditures. Total expenditures are first broken down into payroll and nonpayroll expenditures. Payroll expenditures are then decomposed into payroll per employee and employees per capita. The final step is to adjust average payroll expenses for labor market conditions. Both payroll and nonpayroll expenditures per capita are higher in California than in the average in the rest of the country (Table 11.5). California’s nonpayroll expenditures are particularly remarkable. In 2002, nonpayroll expenditures per capita in California were triple the average in all other states. Nonpayroll expenditures include some personnel-related costs, such as health benefits. These expenditures also include payment for court-appointed lawyers, witness and jury fees, and private attorneys on retainer. We do not know, however, what specific factors account for California’s extraordinarily high nonpayroll expenditures. Our lack of this knowledge is particularly frustrating because nonpayroll expenditures account for much of the widening in the margin between expenditures per capita in California and those in the rest of the nation. In 1982, nonpayroll expenditures per capita were 13 percent higher in California than in the rest of the nation. By 2002, they were three times higher. In contrast, in 1982 payroll expenditures per capita were 33 percent higher in California than in the rest of the nation. By Table 11.5 Components of Total Judicial and Legal Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 201 96 99 88 143 75 Payroll 93 61 75 65 102 45 Nonpayroll 108 35 24 23 31 30 199 2002, that margin had grown, but only modestly—it was 52 percent higher. Although less dramatic, payroll expenditures per capita are also higher in California than in the rest of the nation. In 2002, California exceeded other states in both components of judicial and legal payroll per capita (Table 11.6). It had higher average payroll per employee (34% higher) and more employees per resident (14% more) than all other states. The result was payroll expenditures per capita that were approximately 50 percent higher than in other states. New York followed a similar pattern—higher average payroll and more employees. Its payroll expenditures per capita were even higher than California’s. Labor market conditions in California are partly responsible for the high average payroll of California’s judicial and legal employees. In 2002, average payroll in other states was 26 percent lower than in California (Table 11.7). After adjustments for labor market conditions, this difference was 14 percent. Because many employees in the judicial and legal area are highly educated, we used the average salary of Zone 4 workers in making these adjustments. For the most part, Zone 4 workers have college degrees. As in the case of finance and central staff, we do not have performance or outcome measures in the judicial and legal area. This area also is a difficult one for defining and measuring output. Table 11.6 Components of Judicial and Legal Payroll Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Capita ($) 93 61 75 65 102 45 Employees per Payroll per 10,000 Employee ($) Residents 59,104 16 43,993 14 40,422 18 45,240 14 61,679 17 40,259 11 200 Table 11.7 Average Payroll for Judicial and Legal Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Payroll 59,104 43,993 40,422 45,240 61,679 40,259 Average Payroll Adjusted for Location 52,105 44,796 43,010 45,453 50,450 42,763 Conclusions The activities in governmental administration are among the most important in the public sector. No other area of government has the power of the courts to affect the welfare of individual citizens. The laws and regulations enacted by city councils and state legislators influence all government services. Partly because of their wide reach and indirect consequences, it is inherently difficult to measure performance in this category of governmental administration. We are left, therefore, with some observations about the costs of these activities. Per capita expenditures on governmental administration were significantly higher in California than in other states, and this condition has persisted for many years. Part of the explanation for California’s high expenditures is generous compensation for employees—more generous than dictated by general labor market conditions. On the other hand, California does not employ significantly more workers in this category than do other states. Nonpayroll expenditures are also part of the explanation. These expenditures are significantly higher in California than in other states. 201 12. California’s Fiscal Tradeoffs Previous chapters have examined California’s revenue system and spending patterns in each major functional area, such as education, health, and public safety. This chapter draws this information together in one place to ask: What tradeoffs has California made, either explicitly or implicitly, across budget categories? What might other tradeoffs look like? The chapter begins by reviewing what we know about Californians’ fiscal preferences based on public opinion data. Where expressed preferences differ from actual choices, it then poses some alternative scenarios. These scenarios are not policy prescriptions but illustrations of past fiscal decisions and the kinds of changes that would be necessary to undo these tradeoffs. Californians’ Fiscal Preferences Public opinion data provide a window to Californians’ thoughts about where the state should be in its revenues and public spending. The data suggest that Californians have a preference for a larger public sector. In January 2006, 61 percent of state residents said that they would rather pay higher taxes and have a state government that provided more in services, compared to 31 percent who preferred lower taxes and a smaller public sector (Baldassare, 2006a). In previous years, however, Californians were more divided on this question, at 49 to 44 percent in 2005 and 51 to 44 percent in 1998 (Baldassare, 2005a; 1998).1 Moreover, statewide average responses often mask divisions by political party and race or ethnicity. Democrats are much more likely _____________ 1In 1998, the question was worded differently: “If the government had a choice between reducing taxes or spending more on social programs like health care, social security, and unemployment benefits, which do you think it should do? (a) reduce taxes, even if this means spending less on social programs. (b) spend more on social programs, even if this means higher taxes.” 203 than Republicans to be in favor of a larger state government (76% compared to 53% in 2006). Independents, or those declining to state a party affiliation, fall between (61 percent). Also, white respondents are less likely than Latinos (52% versus 79%) to select the option of higher taxes and more services (Baldassare, 2006a). In terms of specific spending priorities, Californians nearly always single out K–12 education as their top concern. In January 2006, 71 percent of state residents said that they thought the state should spend more on public schools (Table 12.1). Similarly, when asked to name the most important problem facing the governor and state legislature, more respondents (25%) chose education and schools than any other issue (Baldassare, 2006a). Although top concerns often fluctuate with issues of the day, K–12 education is usually mentioned along with jobs and the economy, the state budget, and immigration. Health and human services is typically the second area where Californians would most like to increase state funds. In 2006, 61 percent of state residents indicated that they would like the state to spend more on this area (Table 12.1). Health and human services is also the area after K–12 education that Californians would most like to protect from spending cuts (Baldassare, 2003). Attitudes differ significantly Table 12.1 California Preferences for Government Spending, January 2006 (percent) K–12 public education Health and human services Roads and other infrastructure projects Colleges and universities Corrections, including prisons “For each area, please tell me if you think that the state government should spend more money than it does now, the same amount as now, or less money than now.” More Same Don’t Less Know 71 19 82 61 25 11 3 58 31 83 56 30 95 24 37 32 7 204 across party lines, however, with majorities of Democrats and Independents favoring additional spending on health and human services and Republicans as likely to want to maintain current spending levels as to increase funds in this area (Baldassare, 2006a). Usually vying for third place among Californians’ spending priorities are higher education and roads or infrastructure. In January 2006, 58 percent of California residents and 59 percent of likely voters voiced support for increased funding for roads and infrastructure. At the same time, 56 percent of state residents and 51 percent of likely voters said that they wanted the state to spend more on public colleges and universities. Support for higher education depends somewhat on party affiliation, with a plurality of Republicans (46%) favoring current funding levels. Similarly, support for roads and infrastructure varies by region, with residents of Los Angeles and the rest of Southern California more likely than other Californians to favor additional state funds in this area (Baldassare, 2006a). The only area where many Californians would like to decrease state spending is the correctional system. In January 2006, 32 percent of state residents said that they would like to spend less on corrections, including prisons (Table 12.1). The state corrections system is also the only area where majorities from both political parties support the same or less funding (Baldassare, 2006a). In prior years, even more Californians (42% in May 2005) expressed support for reducing state spending in this area (Baldassare, 2005a). Many Californians (e.g., 57% in February 2003) say that they would be willing to pay higher taxes to fund their budget priorities. Support for tax increases is higher if revenues are to be earmarked for specific purposes or if tax increases are to be limited in size (e.g., Baldassare, 2003). However, when asked about the level of taxation apart from spending priorities, Californians are often evenly split on whether they pay “about the right amount” or “somewhat” to “much more” in taxes than they should (e.g., Baldassare, 2004a). In addition, a significant proportion of Californians (nearly two-thirds in January 2005) believes that there is sufficient waste in state government to spend less without cutting services. Among these respondents, 70 percent believe the state wastes 10 percent or more of total expenditures (Baldassare, 2005a). 205 The latter responses suggest an underlying distrust of government. When asked in 2005 how often they could trust statewide elected officials “to do what’s right,” a large majority (73%) suggested “only some of the time” (Baldassare, 2005b). Californians are only slightly more confident in their local government officials (e.g., Baldassare, 2005c). In contrast, many Californians (40%) have a “fair amount” of confidence in their own ability to decide fiscal and other issues at the ballot box (Baldassare, 2005b). Although regional differences in values or culture are often invoked as explanations for divergent fiscal choices across the nation, Californians do not differ from other Americans in their stated fiscal preferences. For example, in a national Gallup poll, nearly 80 percent of Americans were unwilling to cut spending on K–12 education or health and human services to balance their state budgets (Table 12.2). On the other hand, 56 percent of Americans were willing to reduce funding for their state correctional systems (The Gallup Organization, 2003). Similarly, survey respondents in other large states including Florida, Illinois, and Texas also typically rank K–12 education, health, social services, and higher education as their top budget priorities (Table 12.2). Residents of these states also voice doubts about their elected officials. For example, majorities of New Yorkers have disapproved of their state legislature’s job performance since 2004 (Quinnipiac University, 2006). In Florida, most residents (e.g., 57% in 2004) trust their state government to do what is right only “some of the time” (McDonald, 2004). Californians’ desires for enhanced public spending, mixed sentiments toward taxes, and distrust of elected officials are far from unique but present challenges to state and local decisionmakers. Alternative Tradeoffs Results presented in foregoing chapters suggest that there are gaps between Californians’ fiscal preferences and budget realities. These gaps may be due to difficulties that state and local governments face in reconciling higher costs, larger service populations, and federal program requirements, among other pressures. Alternatively, they may reflect unrealistic preferences among California voters. In any event, it is 206 Table 12.2 Budget Priorities in Other States, 2002 and 2003 (percent) Texas, Florida, 2002 Illinois, 2003 2003 New York, 2003 United States, 2003 “What areas “Would you “Bad idea to of state be willing to “Should balance the government “Should state pay $25 to state state budget by “Would you be would be “Please tell me . . . whether you spending in maintain spending in cutting willing to cut most think the amount now being this area be current service this area be spending in state funding appropriate spent should be . . .” increased?” levels?” cut . . . ?” this area?” for this area?” to cut?” Yes, No, Not Increased Maintained Decreased Yes Yes Not at All Yes Willing Willing 207 K–12 education 85 13 2 77 78 Health services 74 23 4 61 75 Social services 62 32 7 61 63 Higher education 58 38 4 54 58 Protecting the environment 57 36 7 48 State highways and road systems 41 50 8 37 State prisons and correctional facilities 34 48 18 23 77 69 64 59 34 25 84 22 77 8 85 21 78 16 56 useful to explore these gaps further by illustrating the kinds of changes that would be required to undo California’s current fiscal tradeoffs. These scenarios are not meant to be exhaustive. There are many possible alternatives, depending on choices of policy levers, comparison states, and base year. Readers may also look across scenarios and juxtapose different options. More important, these scenarios are not prescriptive. We do not explore how to implement these changes and, indeed, there may be substantial barriers to implementation, including federal requirements or state constitutional restrictions on government tax and spending authority. On the other hand, institutional barriers also represent policy choices. For example, the state could apply for waivers from rules to participate in federal programs, and voters could amend constitutional limits on state and local revenues and spending. Raising Staffing Levels in K–12 Schools As we have seen, more than 70 percent of Californians would like to spend more on K–12 education. In particular, they would like to put more teachers in the classroom. Class size tops the list when Californians are asked what feature of California’s public schools most needs improvement (Baldassare, 2005d). Achieving this goal would be expensive. This is partly because the state is already at a disadvantage compared to the rest of the nation when it comes to staff resources per student. In 2002, California’s public schools employed 92 people per 1,000 students, compared to 129 per 1,000 in the rest of the nation and even higher ratios in Texas and New York (140 and 147 employees per 1,000 students, respectively). In addition, average compensation per school employee is high in California at $65,710 in 2002, compared to $63,456 in New York, the next highest-paying state in our comparison group, and $38,269 in Texas, the lowest-paying state in our group. If California were to increase school staffing ratios to the average for all other states, assuming that new teachers would command the same average salary as current teachers, total state and local government 208 expenditures would increase by $15 billion, or $430 per capita.2 This change is equivalent to a 30 percent increase in K–12 education expenditures or a 5 percent increase in total state and local spending as of 2002 (Table 12.3). Where might these additional resources come from? As we have seen, many Californians would like to reduce spending on corrections (Table 12.1). In 2002, California’s state and local governments spent more per capita in this area than the average for all other states ($257 versus $181). They also spent more than any of our comparison states, although the difference with respect to New York was not large. This higher spending came about despite California’s having fewer inmates per capita than the rest of the nation (0.018 versus 0.021) as well as fewer inmates than in some of our comparison states (Florida and Texas). As a result, expenditures per inmate were two-thirds higher in 2002 for California than the average for the rest of the nation ($14,222 compared to $8,655 per inmate). Table 12.3 Alternative Scenario #1: Raising Staffing Levels in K–12 Schools Expected Change in State and Local Expenditures Raise number of K–12 education staff per student to U.S. average excluding California Reduce corrections payroll per employee to U.S. average excluding California Reduce corrections nonpayroll costs per inmate to U.S. average excluding California Reduce K–12 education payroll per employee to U.S. excluding California Increase property taxes to U.S. average excluding California Increase revenues per higher education student to U.S. average excluding California $ per Capita 430 –24 –70 –150 –192 –123 % Total 5 –0.3 –0.8 –2 –2 –1 _____________ 2This assumption may be violated if new teachers draw lower salaries (e.g., Reed, Rueben, and Barbour, 2006). 209 Roughly half of all corrections expenditures per inmate are payroll costs, including salaries, wages, overtime, and bonuses but excluding employer contributions to retirement saving plans or health insurance. As explained in Chapter 9, some of these payroll expenditures certainly reflect a location premium, or the higher labor costs for all California employers seeking to attract workers with a given background or level of experience. However, the unexplained component of payroll costs accounts for about half the difference between California and other states. If California were to reduce corrections payroll expenses to the national average while continuing to pay workers the same location premium, payroll costs would drop from about $52,000 to $44,000 per employee. However, because California also employs fewer corrections personnel per capita than other states, total savings from this change would be only $800 million, or $24 per capita (Table 12.3). Reducing nonpayroll costs per inmate to the average for all other states ($4,151 instead of $7,777 per inmate) would provide an additional $2.4 billion, or $70 per capita (Table 12.3). Thus, total savings in corrections would be $94 per capita, less than one-quarter of the savings necessary to reach national staffing levels in K–12 education. Note that California could enact the same payroll adjustment for K–12 education as discussed above for corrections. That is, it could reduce salaries to the average for all other states while maintaining the location premium that all employers in California must pay for collegeeducated workers. This change would be analogous to the class size reduction (CSR) policy enacted in 1996, which earmarked funds for smaller class sizes rather than higher salaries.3 If this change were implemented, compensation per public school employee would fall from an average of about $66,000 to $57,000 per year, providing an annual $5.2 billion or $150 per capita in savings (Table 12.3). The substantial savings from this change in K–12 education as opposed to corrections result from the higher number of employees in elementary and secondary education (669,238 full-time_____________ 3However, demand for more teachers may also drive up salaries in the longer term (Jepsen and Rivkin, 2002). 210 equivalent employees, compared to 77,455 in corrections as of March 2002). California’s school finance system was transformed with the passage of Proposition 13 in 1978. In addition to capping property tax rates and limiting the growth in assessed values, Proposition 13 made the state responsible for allocating property tax revenues among schools and local governments (e.g., Sonstelie, Brunner, and Ardon, 2000). The effects of lifting Proposition 13 restrictions on local property tax rates and assessment practices are unclear. Sources of uncertainty include court mandates to equalize school funding across districts and feedback effects on real estate markets. Nevertheless, the experiences of other states provide a basis of comparison. As a result of Proposition 13, California state and local governments collect less in property taxes than the average state—$26 per $1,000 of personal income in 2002, compared to $32 per $1,000 in the rest of the nation—and even less than in other large states such as New York, Illinois, and Texas. If California were to raise the same amount of property tax revenues per $1,000 of personal income as in the rest of the nation, it would receive an additional $192 per capita in revenues (Table 12.3).4 Finally, California’s K–12 and higher education systems are clearly interdependent. The state’s public schools depend on its colleges and universities to provide skilled teachers. The higher education system, in turn, relies on K–12 schools for adequate student preparation. Institutions of higher learning also have potential social benefits as centers of innovation, public service, and regional economic growth. On the other hand, it is inescapable that average net support per fulltime-equivalent student in California’s colleges and universities is 50 percent higher than in its K–12 public schools ($11,300 per student, compared to $7,400 per student). This result stems in part from higher labor costs and in part from a tradition of low tuition and fees in _____________ 4Another proposal is to move to a “split roll,” under which commercial and industrial property would be assessed at market value and residential property would continue to be assessed at acquisition value. Revenue estimates for these proposals range from $1.9 to $3.3 billion, or roughly $54 to $94 per capita (Sexton and Sheffrin, 2002; Auerbach, 2003; California State Board of Equalization, 2002). 211 California. For example, enrollment in California’s community colleges was free as recently as 1985. Similarly, in the latter half of the 1990s, higher education fees were held constant in nominal terms, resulting in a real decline in revenues per student over that period. Despite more recent policy shifts, average fees for the UC and CSU systems remain below those of comparable institutions and CCC fees are the lowest in the nation (Legislative Analyst’s Office, 2006a). In 2002, higher education revenues (including both tuition and auxiliary revenues) were $5,404 per full-time-equivalent student in California, compared to $7,027 per FTES in the rest of the nation. If California were to raise revenues per FTES to the average for all other states, it would gain $123 per capita (Table 12.3). This change would also result in a significant reduction (more than $3,000) in net support per FTES, part of which could be offset by an increase in student assistance or a greater reliance on federal sources. (California is currently slightly below average in direct student aid, at $1,351 versus $1,537 per student.) It is important to note that this scenario assumes no change in overall higher education participation rates. In sum, achieving one of Californians’ most desired policy goals would be expensive. Potential tradeoffs with other budget priorities often involve difficult choices and may reap less in savings than would be expected. On the other hand, Californians may wish to implement some variant of these changes, particularly if low school staffing ratios harm the least advantaged in our society. For example, in April 2006, more than 70 percent of Californians indicated that they would favor adding support staff (counselors, mentors, and social workers) in low-income schools to reduce dropout rates, even if this change cost the state more money (Baldassare, 2006b). Expanding Coverage in Health and Human Services As noted above, a majority of Californians (52%) would like to dedicate more resources to health and human services (Table 12.1). This category encompasses several types of programs, ranging from cash assistance for low-income families to subsidized medical care for children and their parents to meals and transportation for elderly and disabled individuals, and it is difficult to know which programs in particular are 212 drawing popular support. Moreover, support for expanding health and human services is not necessarily shared across political parties, regions, or other breakdowns of survey respondents. Nevertheless, to understand California’s current fiscal tradeoffs, it is useful to consider an alternative scenario whereby California expands its two primary health and human services programs, Medi-Cal and CalWORKs. Both programs are already characterized by broader coverage than in other states. In 2002, Medi-Cal enrolled 21 percent of the state population and 46 percent of all Californians with incomes below 300 percent of the FPL, compared to shares of 15 and 32 percent, respectively, in the rest of the nation. Similarly, CalWORKs recipients represented 3.4 percent of the overall population and 13.8 percent of families with children and incomes below 200 percent of the FPL, compared to 1.6 and 8.2 percent, respectively, in the rest of the nation. California’s higher enrollment rates may be driven by broader rules about who is eligible for these programs or by individual decisions about whether or not to participate. Although we cannot always separate these two influences, under both CalWORKs and Medi-Cal the state exceeds federally mandated expenditures. For example, in 2002, the maximum allowable CalWORKs grant for a family of three (one adult and two children) and no earnings was $679 per month, the third highest in the nation after Alaska ($923) and Hawaii ($712).5 Higher maximum cash grants increase the pool of potential TANF recipients both because they raise the level of income at which individuals can start receiving benefits and because more individuals may apply for and receive benefits. Similarly, Medi-Cal covers individuals at higher income and resource levels than Medicaid programs in several other large states and provides more benefits that are “optional” under federal program rules. In addition, California has extended coverage to groups who are not traditionally covered under Medicaid, such as working parents with family incomes up to 107 percent of the FPL who were not otherwise eligible. _____________ 5These grants reflect in part higher spending in California before the 1996 federal welfare reform, which “locked in” a certain floor of public spending in the states. 213 For both Medi-Cal and CalWORKs, the per capita costs of expanding coverage even further would be modest relative to some of the changes considered in the previous scenario, for example. The reason is that, despite higher enrollments in each of these programs, California has a lower cost per beneficiary relative to other states. For CalWORKs in 2002, for example, the average cost per recipient was roughly $4,700, compared to $5,100 in the rest of the nation and higher amounts in all of our comparison states except Texas.6 The differences between average Medi-Cal expenditures in California and the rest of the country were even more dramatic, at about $3,100 versus $4,900, respectively. Thus, if California were to adopt a policy like that of New York, offering Medicaid to all working parents with family incomes up to 150 percent of the FPL and adults without children below the poverty line, the state would gain roughly 800,000 enrollees, a 40 percent increase for this group.7 However, because of California’s lower cost per recipient and particularly low relative costs for nonelderly adults ($1,549 in 2002, compared to $3,951 in New York), the cost would be only $1.2 billion, or $35 per capita. Similarly, if California were to expand CalWORKs coverage by 10 percent (it already has the maximum coverage among states in our comparison group), the additional cost would be $660 million, or $19 per capita (Table 12.4). Where might these resources come from? States typically expand Medicaid eligibility to nontraditional groups through Section 1115 waivers from federal program requirements. However, the federal government requires that these demonstration projects be “budget neutral.” States thus often use savings from other program areas such as enhanced managed care enrollment, limited benefits for some groups, or requiring greater cost-sharing for recipients. Given Medi-Cal’s already _____________ 6Recipients are defined here as individuals rather than families. Thus, a family of three with one parent and two children all receiving cash assistance would include three recipients, whereas one including a child and two nonparent relatives who do not receive assistance would include only one recipient. 7If only a fraction of potentially eligible adults applied for and received benefits, the enrollment gain would obviously be smaller. As noted in Chapter 6, the overall take-up rate in California for both Medi-Cal and SCHIP is 78 percent, higher than the average rate for all other states but somewhere between that for Illinois and New York. 214 Table 12.4 Alternative Scenario #2: Expanding Medi-Cal and CalWORKs Raise Medi-Cal maximum allowable incomes for working parents, extend coverage to all adults without children below FPL Increase CalWORKs caseloads per capita by 10 percent Increase selective sales taxes to U.S. average excluding California Reduce central finance and staff to Illinois level Reduce judicial and legal to New York level Expected Change in State and Local Expenditures $ per Capita % Total 35 0.4 19 0.2 –94 –1.1 –18 –0.2 –59 –0.7 high managed care enrollments and cost increases on par with those of traditional fee-for-service (FFS), these savings may be limited for California. However, it is also possible that California would fund such an expansion using new resources. A commonly suggested way to finance health program expansions is by increases in “sin” taxes such as levies on alcoholic beverages and tobacco sales. Although California voters rejected a proposal on the November 2006 ballot to increase cigarette taxes by $2.60 per pack, survey respondents generally favor these taxes (e.g., Baldassare, 2006a, 2005a, 2004a). Chapter 3 shows that, as a group, selective sales tax revenues are lower in California than in other states ($9 versus $12 per $1,000 of personal income). If selective sales tax collections were the same in California as in other states, revenues per capita would climb by $3.3 billion, or $94 per capita (Table 12.4). It is worth noting that California’s taxes on tobacco products in particular are already relatively high ($0.87 per pack, compared to $1.50 in New York, $0.98 in Illinois, $0.41 in Texas, and $0.34 in Florida as of 2002). Moreover, cigarette taxes can discourage consumption and have negative distributional effects. More generally, economists are already critical of California’s current sales and use tax because of its high rate (7.25% statewide, plus local levies) and narrow base (e.g., exemptions for most services as well as utilities and food products). 215 An unpopular area of public spending and one where California differs markedly from other states is government administration. In 2002, California’s per capita spending in this area was $447. California’s expenditures on administration exceeded the average for all other states by nearly 50 percent and were higher than in any of our comparison states including New York (by 15%). Administrative services are necessary to keep government functioning. They are also labor-intensive, suggesting that payroll costs could explain much of the variation between California and other states. However, nonpayroll costs are also high in this area, and differences in salaries are not explained by general labor market conditions or by the location premiums that all employers in California must pay to attract qualified workers. The largest components of administrative expenditures are finance and central staff (e.g., general government support, revenue collection, and the administration of public employee retirement funds) and judicial and legal functions. In 2002, state and local governments in California spent $216 and $201 per capita on these areas, respectively. If California were to reduce per capita expenditures in each category to the level of the next highest state in our comparison group (Illinois for central finance and staff and New York for judicial and legal), it would save $632 million (or $18 per capita) on central finance and staff and $2 billion ($59 per capita) on judicial and legal costs. The breadth of activities included in each of these categories makes it difficult to understand where these cuts might come from and how they might affect government outputs. In sum, health and social services is a complicated area in which to consider alternative tradeoffs because of complex and often interlocking federal program requirements. However, the above scenario reinforces a key finding of this report, namely, that California enrolls larger shares of its population in Medicaid and TANF than other states do and it incurs lower expenditures per case. As a result, hypothetical program expansions would not be very costly in per capita terms relative to some of the other scenarios presented here. These calculations do not take into account differences between new enrollees and current beneficiaries of these programs. Differences in 216 health care use patterns among these groups, for example, could alter the conclusions presented here (e.g., MaCurdy et al., 2005). In addition, we have not addressed interactions between public and private health care markets, including potential cross-subsidies, which might be affected by changes in public programs. Investing in Roads and Infrastructure Twenty-seven percent of Californians would like the state to spend more on roads and infrastructure (Table 12.1). When asked what types of infrastructure should receive funding priority, survey respondents typically choose school facilities (e.g., 40% in August 2004), surface transportation (24%), and water systems (21%) (Baldassare, 2004b). These three sectors represent the bulk (85%) of $220 billion in proposed spending under the governor’s ten-year strategic growth plan (Office of the Governor, 2006) and more than 90 percent of borrowing approved by California voters in the November 2006 election. We find that California’s capital outlays in these areas have lagged those of other states in recent years. For example, in 2002, California spent roughly half as much as the average for all other states on higher education facilities ($35, compared to $65 per capita). Although it spent more than other states on K–12 educational facilities ($201 versus $182 per capita), this situation was without precedent since the 1970s. Similarly, capital outlays on mass transit were similar to the national average excluding California but those on highways lagged the average for other states ($154, compared to $240 per capita) and expenditures in every state within our comparison group. Capital spending on water supply and quality was higher than average ($77, compared to $53 per capita) but lower than spending in Florida ($97 per capita) and other western states. More important, California’s capital investments have often failed to keep pace with growing demand. From 1992 to 2002, California experienced increases in K–12 and higher education enrollment comparable to those of Texas and Florida. However, its capital investment per additional student was lower than in either of these states 217 or the national average excluding California.8 Similarly, in transportation, California’s recent population growth and trends in vehicle miles traveled mirror those of Florida and Texas, yet its addition of freeway lane miles has been modest, more akin to that of New York. As a result, California’s freeways have been ranked among the most congested in the nation (Schrank and Lomax, 2005). The experiences of other states provide a range of alternative scenarios. If California were to fund highway construction at the level of another fast-growing state with high automobile use, such as Florida, state and local annual expenditures would rise by $4.2 billion, or $120 per capita (Table 12.5). If it were to invest in higher education facilities at a level comparable to another state with a large public college and university system and rising enrollments, such as Texas, spending would be $42 per capita higher. If California’s capital investments in water supply and other natural resources were equivalent to those of Florida in 2002, its annual expenditures would increase by $19 per capita. How big are these changes? California state and local governments might offset increases in highway capital spending with reductions in mass transit. If California reduced these expenditures to the level of another state with high annual vehicle miles traveled, Texas, it would produce savings of $4.4 billion, or $127 per capita (Table 12.5). California could also charge user fees for mass transit comparable to those in Illinois ($59 per capita) rather than the current $36 per capita in fares. If it followed this course, California would raise an additional $805 million ($23 per capita) (Table 12.5). Yet another alternative would be to increase highway-related taxes and fees, including state _____________ 8The funding picture for educational facilities is changing because of several recent ballot propositions. In 1998, voters authorized $9 billion in general obligation bonds for the construction and modernization of buildings in public schools, colleges, and universities (Proposition 1A). In 2000, they passed Proposition 39, lowering the vote requirement for local bonds for these purposes from two-thirds to 55 percent. In 2002, voters passed Proposition 47, which authorized the state to sell bonds worth $13 billion for capital improvements to public schools and public colleges and universities. In 2004, they passed Proposition 55, which authorized $21 billion for the same purposes, and in 2006, voters approved Proposition 1D, authorizing $10.4 billion in bonds for K–12 and higher education facilities. 218 Table 12.5 Alternative Scenario #3: Investing in Roads and Other Infrastructure Increase highway capital expenditures to level in Florida Increase higher education capital spending to level in Texas Invest in natural resources and water supply at level of Florida Reduce expenditures per capita on mass transit to level in Texas Increase transportation user fees to level in Illinois Increase fuel tax to average for U.S.–California Expected Change in State and Local Expenditures $ per Capita % Total 120 1.4 42 0.5 19 0.2 –127 –23 –33 –1.5 –0.3 –0.4 gasoline taxes, vehicle license fees, mileage and weight fees on motor carriers, and driver’s license fees. In 2002, these taxes and fees amounted to roughly $153 per capita in California, compared to $186 per capita on average in all other states. If California had the same such tax burden as in the rest of the nation, it would collect an additional $33 per capita, or $1.2 billion (Table 12.5). In sum, raising annual capital expenditures on California’s infrastructure would not require a large infusion of spending, although eradicating years of lower-than-average spending may require additional resources. California’s capital spending, although uneven across categories, is not widely off the mark of the average for other states. Conclusions These tradeoffs have served to reinforce certain themes of our report. In many areas, the state faces higher-than-average caseloads. This may be due to demographics (as in the case of K–12 education, for example), policy decisions (as in Medi-Cal), or both (CalWORKs and higher education, for example). Like all employers in California, state and local governments must pay higher wages to attract employees who can afford to live in the state. However, governments often pay more than the 219 premium required at comparable skill and experience levels. Higher payroll expenditures translate into lower staffing ratios in K–12 education, police, and fire protection. The state also has relatively low user fees in certain areas (higher education and mass transit) and higher administration costs overall. This is an opportune time for state residents to reexamine these choices and to consider alternative tradeoffs. 220 Appendix Data Sources Chapter 2 State Rankings State geography, population, and economy data and rankings reported in Table 2.2 come from the U.S. Census Bureau, Statistical Abstract of the United States: 2006, available at http://www. census. gov/prod/www/statistical-abstract.html. Wage Data Wage data come from the “Occupation Employment Statistics” (OES) survey, available at www.bls.gov/OES/. The United States Bureau of Labor Statistics administers this survey semi-annually to every employer in the United States with more than 250 employees and to a random sample of smaller firms. Wage estimates for 2002 are based on responses from 1999 through 2002 for a total sample size of approximately 1.4 million establishments. The nationwide response rate in 2002 was 78 percent, covering 71 percent of employment. Job Zones Job Zones (Tables 2.3–2.5) come from the O*NET or Occupational Information Network database, version 7.0, available at http://www. onetcenter.org/database.html. This database combines information on worker and occupation characteristics from occupational analysts and surveys of employees, their supervisors, and co-workers. We weighted average salaries in each zone by the national employment shares in each occupation. We use national rather than state employment shares as weights so that our location premium reflects only price differences and not the composition of labor markets in each state. 221 Payroll Public employment and payroll expenditures used throughout the report come from the Census of Governments Employment Phase in 2002, available at www.census.gov/govs/www/apcs.html. The Census Bureau canvasses all U.S. state and local governments in Census years (ending in 2 or 7) and all states and a sample of local governments annually. Statistics for part- and full-time employees are reported for the same functional categories used in the Finance Phase (de Alth and Haskel, 2007). Gross payrolls include all salaries, wages, fees, overtime, bonuses, and awards paid to employees as of the March 12 pay period. To convert these monthly payrolls to yearly payrolls, the Census Bureau payroll total was multiplied by 11.77 (365 days/31 days). Fringe benefits including health and life insurance, Social Security, and retirement contributions are not included. Chapter 3 Personal Income Personal income data (Tables 3.3, 3.4, and 3.7) come from the Bureau of Economic Analysis, SA04 current series, available at http:// www.bea.gov/bea/regional/spi/default.cfm?satable=SA04&series=ancillary. As noted in the chapter, this series excludes some types of income that are part of the tax base, such as capital gains, pension benefit payments, and personal contributions to social insurance programs. It includes other sources of nontaxable income, such as the nontaxable portion of Social Security benefits and certain investment income. It is not clear how adjusting for these factors would affect California’s relative position. Tax Rates and Incidence State tax rates and provisions (Tables 3.5 and 3.6) are from the Federation of Tax Administrators (FTA), available at http://www. taxadmin.org/fta/rate/tax_stru.html, and the Urban-Brookings Tax Policy Center, available at http://www.taxpolicycenter.org/TaxFacts/ tfdb/TFTemplate.cfm? topic2id=90. In the case of sales tax rates, the highest local rate is that known to be actually levied by at least one 222 jurisdiction. Average tax rates (e.g., in Table 3.6) are not weighted by population or sales. Income thresholds for state income taxes are from Johnson, Llobrera, and Zahradnik (2003), Table 3, also available at http://www. taxpolicycenter.org/TaxFacts/tfdb/TFTemplate.cfm?topic2id=90. Information on income tax progressivity is based on reports to the California Franchise Tax Board and comparable bodies in New York and Illinois. Sales taxes on services are compiled from a survey undertaken in 2004 by the FTA, available at http://www.taxadmin.org/fta/pub/services/ services04.html#summary. Chapter 4 Children Eligible for Public School Children eligible for public school (Table 4.4) is from Census 2000, Table QT-P19, School Enrollment: 2000, available at http://factfinder. census.gov/servlet/QTTable. It is the sum of public and private school enrollment in kindergarten, elementary school, and high school plus the population ages 16 to 19 who are not enrolled in school and are not high school graduates. Employees Employees in elementary and secondary education (Tables 4.6 and 4.7) are from the National Center for Education Statistics, State Nonfiscal Survey of Public Elementary and Secondary Education, 2001–02, Common Core of Data, available at http://nces.ed.gov/ccd/ccddata.asp. Staff are measured in full-time-equivalents. Support staff in Table 4.7 consists of instructional aides, instructional coordinators, guidance counselors, librarians, library support staff, and student support staff. Administrative staff in Table 4.7 consists of local education agency (LEA) administrators, LEA administrative support staff, school administrators, and school administrative support staff. Other support staff in Table 4.7 includes the category “all other staff” in the NCES survey. 223 Salaries and Benefits of Public School Employees Compensation for public school employees (Tables 4.5, 4.6, and 4.8) is from the National Center for Education Statistics, National Public Education Financial Survey, School Year 2001–02, Common Core of Data, available at http://nces.ed.gov/ccd/ccddata.asp. These salaries and benefits are the sum of salaries and benefits for the following four areas: instructional expenditures, support expenditures, noninstructional food service expenditures, and noninstructional enterprise expenditures. School Enrollments Public school enrollments (Tables 4.3–4.7 and 4.10 and Figures 4.2–4.4) are from the National Center for Education Statistics, State Nonfiscal Survey of Public Elementary and Secondary Education, 2001– 2002, Common Core of Data, available at http://nces.ed.gov/ccd/ ccddata.asp. In Table 4.4, public and private school enrollment is from Census 2000, Table QT-P19, School Enrollment: 2000, available at http://factfinder.census.gov/servlet/QTTable. Stanford Achievement Test, Ninth Edition SAT9 scores are available at the web site of the California Department of Education for its Standardized Testing and Reporting Program (http://star.cde.ca.gov/). Average Scores on NAEP Average NAEP scores for different states and groups of students (Tables 4.11 and 4.12) are from the NAEP website, available at http://nces.ed.gov/nationsreportcard/naepdata/search.asp. Chapter 5 Higher Education Enrollment Data The National Center for Education Statistics provided enrollment figures. These data were extracted from the Integrated Postsecondary Education Data System (IPEDS) using the Dataset Cutting Tool accessed via the IPEDS website, available at http://nces.ed.gov/ipedspas. 224 For each state, enrollment figures were collected for public two- and four-year, degree-granting institutions for the fall 2002 period. Full-Time-Equivalent Students The IPEDS dataset derives full-time-equivalent student numbers (Tables 5.4, 5.6, 5.9, 5.11, and 5.12; Figures 5.3, 5.5, 5.6, and 5.7) from reported full- and part-time enrollment for the fall period. A formula is used to convert an institution’s part-time enrollment into full-timeequivalents and then that number is added to the reported full-time enrollment of the institution. The full-time-equivalent of part-time enrollment is estimated by multiplying reported part-time enrollment by factors that vary by control, level of institution, and level of student. The formula used is published annually in the Digest of Education Statistics. In the case of California, this formula produces a smaller number of FTES than those reported by the three state systems. Student Migration Figures for enrollment in home state versus out-of-state schools (Table 5.5) were derived from Table 203, “Residence and Migration of All Freshmen Students in Degree-Granting Institutions, by State or Jurisdiction, Fall, 2002” in the Digest of Education Statistics, 2004. Degrees Awarded IPEDS provides data on the number of degrees completed in a given academic year (in this case, July 1, 2002, and June 30, 2003). Data were collected for associate’s, bachelor’s, master’s, doctorates, and professional degrees (Tables 5.10, 5.11, and 5.12). Awards for programs shorter than two years were not included in the total. Chapter 6 Medicaid Expenditures, Recipients, Enrollment, and Service Use Medicaid expenditures, recipients, and enrollment data (Tables 6.2 and 6.4–6.9) are from the Medicaid Statistical Information System (MSIS) State Summary Datamart (SSD), an exclusive web-based 225 Medicaid data analysis tool from the Centers for Medicare & Medicaid Services, available at http://www.cms.hhs.gov/MSIS/. Medicaid participants are recipients of Medicaid services during the fiscal year. Medicaid enrollment is the total count of recipients of Medicaid during the fiscal year, excluding those enrolled in Medicaid expansion SCHIP programs and those enrolled in Family Planning Services programs under a section 1115 waiver. Medicaid expenditures measure payments for claims adjudicated during the fiscal year. Population with Income Below 300 Percent of Federal Poverty Level Estimates of the state population composition in Table 6.3, and estimates of state population groups with incomes below 200 and 300 percent of FPL (Tables 6.4–6.6) are the authors’ calculations using the Census Bureau’s American Community Survey 2002, available at www.census.gov/acs/www. Children ever enrolled in SCHIP in Table 6.6 (both Medicaid expansion SCHIP and Separate SCHIP) are obtained from the Centers for Medicare & Medicaid Services, SCHIP Enrollment Report FY2002, available at http://www.cms.hhs.gov/schip/enrollment/schip02.pdf. Take-Up Rate and Eligibility Rate Monthly take-up rate and eligibility rates (Table 6.7) are from the Office of Health Policy, U.S. Department of Health and Human Services, TRIM3’s 2001 Baseline Simulation of Medicaid and SCHIP Eligibility and Enrollment: Methods and Results, Table 4, available at http://aspe.hhs.gov/health/reports/05/medicaid-schip-simulation/. Managed Care Enrollment and Payment Managed care enrollment and capitated payments (Table 6.9) are from the Centers for Medicare and Medicaid Services, National MSIS Tables by States—2002, Tables 10 and 17, available at http://www.cms. hhs.gov/medicaid/msis/tables2002.asp. 226 Medicaid Fee Indexes and Medicaid-Medicare Fee Indexes Table 6.10 and Table 6.11 are extracted from “Exhibit 2: Medicaid Fee Indexes and Medicaid-to Medicare Fee Indexes 2003,” in Zuckerman et al. (2004). The computation of Medicare obstetrics fees uses the formulas (conversion factors and geographic adjusters) for Medicare fees based on relative value units (RVUs) in the Medicare Physician Fee Schedule. Although not many Medicare patients get obstetrics care, the RVUs are created to be service-specific, and do not take into account patient characteristics. Charges of Own Public Hospitals Charges of publicly administered hospitals (Table 6.12) include charges from patients, private insurance companies, public insurance programs (such as Medicare) of public hospitals and of institutions for care and treatment of the handicapped, and receipts from hospital canteens, cafeterias, gift shops, etc. Outcome Measures Performance measures of preventable hospitalizations, birth outcomes, and access to care (Tables 6.13–6.15) are extracted from Billings and Weinick (2003), available at http://www.ahrq.gov/data/ safetynet/netfact.htm. Chapter 7 Low-Income Population All low-income population figures (Tables 7.4 and 7.14) were estimated using the U.S. Bureau of the Census March 2002 Current Population Survey (CPS). Low income was defined at 200 percent of the FPL. SSI Caseload and Aid Payment and recipient data (Tables 7.12 and 7.14) come from the Social Security Administration’s Supplemental Security Income Annual Statistical Supplement and Annual Statistical Supplement, available at http://www.ssa.gov/policy/docs/statcomps/supplement/2003/index.html. 227 TANF, CCDF, and SSBG Caseload and Financial Information The analysis of welfare, child care, and community-based services (Tables 7.2–7.5, 7.6, 7.9–7.11) draws from several publications of the U.S. Department of Health and Human Services’ Administration for Children and Families (ACF): TANF Annual Reports to Congress, CCDF data tables, SSBG Annual Reports on Expenditures and Recipients, and SSBG Focus Reports. All these publications are available at http://www. acf.hhs.gov. Recipiency rates in Table 7.3 were calculated by adding the number of recipients from services funded with TANF and CCDF. This potentially induces some double-counting and underestimates benefits per recipient. Exclusive use of TANF expenditure and caseload data increases the variability in recipiency rates and benefits per case but leaves the qualitative conclusions unchanged. It is important to keep in mind that in Table 7.3, the number of recipients for state supplementation is assumed to be exactly the same as the total number of recipients. This is not necessarily true, but it is not possible to have a detailed breakdown of the respective recipient populations with the available information. Adoptions and Foster Care The Green Book reports only federal funds expended through the matching grants authorized under Title IV, Sections B and E, of the Social Security Act. However, it was possible to impute state expenditures from the matching requirements associated with each type of service and, with them, total spending. Caseload figures for these programs (Tables 7.7 and 7.8) were extracted from ACF’s Child Welfare Outcomes Annual Report. SSI/SSP The Social Security Administration’s data (Tables 7.12–7.14) include payments made by the federal government, which the Census Bureau does not count toward each state’s expenditures. As a consequence, the sum of expenditures in the three programs presented in 228 this chapter exceeds, by construction, the total of cash assistance, categorical assistance, and other welfare from the Census categories. Chapter 8 Lane Miles of Freeways and Other Principal Arterials Lane miles of freeways and other principal arterials (Tables 8.13–8.15 and Figure 8.5) are from Table HM-60, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm. These are the sum of lane miles for rural interstates, other rural principal arterials, urban interstates, other urban freeways and expressways, and other urban principal arterials. Lane Miles of Local and State Highways State highway lane miles (Table 8.18) are from Table HM-81, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/ hss/hsspubs.htm. Local highway lane miles (Table 8.17) are total lane miles from Table HM-60 less state highway lane miles from Table HM81. Operating Costs and Employee Compensation for Public Transit Operating costs and employee compensation for public transit (Tables 8.6–8.9) are from the 2002 National Transit Database of the Federal Transit Administration, available at http://www.ntdprogram. com. Passenger Trips and Passenger Miles Passenger trips and passenger miles (Tables 8.5–8.8) are from the 2002 National Transit Database of the Federal Transit Administration, available at http://www.ntdprogram.com. 229 Population Density of Urbanized Areas The percentage of a state’s population in urbanized areas and the density of urbanized areas are from the Census 2000 Summary File 1, Table GCT-PH1, Population, Housing Units, Area, and Density: 2000. Road Congestion and Delay Data on road congestion and delay in urban areas (Table 8.12 and Figure 8.6) are from The 2005 Urban Mobility Report, Texas Transportation Institute, May 2005, available at http://mobility.tamu. edu. Vehicle Miles of Travel Vehicle miles of travel (Tables 8.14 and 8.15) are from Table VM-2, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/hss/ hsspubs.htm. Chapter 9 Employment and Payroll Throughout the chapter, payroll costs per employee (Tables 9.4, 9.5, 9.9, 9.10, 9.12, and 9.13) are calculated by dividing total payroll by the number of full-time-equivalent employees in March of each year. The sources are described above. Confined Population and People Under Probation or Parole The U.S. Department of Justice’s Bureau of Justice Statistics provided a variety of documents for complementary data. Counts on people in jails, in prisons, and under probation or parole (Figure 9.2 and Tables 9.3 and 9.4) come from Bureau of Justice Statistics available at http://www.ojp.usdoj.gov/bjs/dtdata.htm#ncrp, for data before 1999, as well as from the Annual Survey of Jails and the Sourcebook of Criminal Justice Statistics for 1999–2002. 230 Correctional Facilities Information on the number of state prisons and local jails (Table 9.6) comes directly from each state’s Department of Corrections. The rated and design capacity of both state prisons and local jails was extracted from the Department of Justice, Prisoners in 2002, and Sourcebook of Criminal Justice Statistics, respectively. “Design capacity” refers to how many inmates the original design of a facility was intended in to house. “Rated capacity” is a more subjective measure determined by officials within each jurisdiction. Crimes The Federal Bureau of Investigation’s Uniform Crime Reports was the source for the number of crimes (Table 9.9). We used the number of criminal offenses for the analysis in this chapter. Criminal offenses encompass more incidents than those included in the FBI’s commonly used crime index. The latter uses information of offenses cleared by arrest or by exceptional means and which generally entail higher levels of seriousness. Examples include criminal homicide, aggravated assault, forcible rape, and arson, among others. In addition to the number of incidents the crime index comprises, the number of criminal offenses includes nonviolent crimes, such as assaults without weapons, fraud, vandalism, drug abuse violations, driving under the influence, etc. Chapter 10 HUD Income Limits, Rent Burdens, Wait Lists, and Fair Market Rents Income limits for public housing and other housing assistance (Table 10.4) come from the U.S. Department of Housing and Urban Development, FY2002 HUD Income Limits Briefing Material, available at http://www.huduser.org/DATASETS/il/fmr02/index.html. These data were merged with income data by family characteristics from the 2002 March CPS. Data on the receipt of housing assistance also come from the CPS’s HPUBLIC and HRENT questions. Data on rent burdens (Table 10.5) come from the 2000 Census of Population and Housing. HUD Fair Market Rents are taken from http://www. 231 huduser.org/datasets/fmr.html. Data on average waiting times for housing programs (Table 10.6) are taken from “Picture of Susidized Households, 1998,” available at http://www.huduser.org/datasets/ assthsg/statedata98/. Fair market rents in 2002 (Table 10.7) come from the HUDuser website, available at http://www.huduser.org/datasets/ fmr/fmr2002/index.html. Park Acreage and Visits For Tables 10.13 and 10.14, the source for local acreage was State Comprehensive Outdoor Recreation Plans (SCORPs) reported by states to the National Park Service. For state park acreage, we drew on the 2003 Annual Information Exchange by the National Association of State Park Directors, available at http://www.naspd.org. National park data came from U.S. Fish and Wildlife Service, Land Areas Report, 2004, and data from the Land Resource Division of the U.S. National Park Service, available at http://ca.rand.org/stats/community/parkvisitsCA.html. Data on visits for state and national parks came from the 2005 Statistical Abstract, Table 1250. Chapter 11 Expenditures Expenditures for finance and central staff, judicial and legal, general public buildings, and legislative (Tables 11.1–11.3, and 11.5 and Figures 11.1–11.3) are from the Census of Governments. Expenditures on finance and central staff combine the Census categories of financial administration (the Census Bureau series E23, F23, and G23) and central staff services (E29, F29, and G29). Legal and judicial expenditures are the sum of Census categories E25, F25, and G25. Legislative expenditures are the sum of E26, F26, and G26. Expenditures on general public buildings are the sum of E31, F31, and G31. Employees and Payroll Employees and payroll expenditures (Tables 11.2–11.7) come from the employment portion of the 2002 Census of Governments, available 232 at http://www.census.gov/govs/www/apes.html. The Census Bureau measures the number of employees in full-time-equivalents and the gross payroll for the month of March. Population When expenditures are expressed in per capita terms (Tables 11.1–11.3, 11.5, and 11.6 and Figures 11.1–11.3), the denominator is the population estimate by states released by the U.S. Census Bureau on December 22, 2004. 233 Table A.1 Urban Areas in Texas Transportation Institute (TTI) Study, 2002 Population (1,000s) % of State Population California Bakersfield Oxnard-Ventura Fresno Sacramento Riverside–San Bernardino San Jose San Diego San Francisco–Oakland Los Angeles–Long Beach–Santa Ana All urban areas in TTI study Florida Pensacola Cape Coral Sarasota-Bradenton Jacksonville Orlando Tampa–St. Petersburg Miami All urban areas in TTI study Illinois Chicago New York Albany-Schenectady Rochester Buffalo New York–Long Islanda All urban areas in TTI study Texas Beaumont Brownsville Laredo Corpus Christi El Paso Austin San Antonio Houston Dallas–Fort Worth–Arlington All urban areas in TTI study 425 560 585 1,560 1,610 1,675 2,825 12,400 12,400 25,760 310 315 560 905 1,255 2,025 5,000 10,370 8,120 525 660 1,115 12,736 15,036 145 160 190 320 665 835 1,280 3,720 4,200 11,515 1.3 1.7 1.7 4.6 4.8 4.8 8.3 36.6 76.1 76.1 1.9 2.0 3.5 5.7 7.9 12.7 31.3 64.9 65.4 2.8 3.5 5.9 67.1 9.2 0.7 0.8 0.9 1.5 3.2 4.0 6.1 17.8 20.1 55.2 aStatistics for New York CMSA prorated to portion in New York State. 234 References Auerbach, Richard, “Estimates of Los Angeles County’s Commercial Property Values Too High in Proposition 13–Related Study,” press release, Los Angeles County Assessor, Los Angeles, California, February 25, 2003. Bailey, Amanda, and Joseph P. Hayes, “Who’s in Prison? The Changing Demographics of Incarceration,” California Counts, Vol. 8, No. 1, Public Policy Institute of California, San Francisco, California, August 2006. Baldassare, Mark, PPIC Statewide Survey: May 1998, Public Policy Institute of California, San Francisco, California, May 1998. Baldassare, Mark, PPIC Statewide Survey: Californians and Their Government, Public Policy Institute of California, San Francisco, California, February 2003. Baldassare, Mark, PPIC Statewide Survey: Special Survey on California’s Fiscal System, Public Policy Institute of California, San Francisco, California, January 2004a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on California’s Fiscal System, Public Policy Institute of California, San Francisco, California, August 2004b. Baldassare, Mark, PPIC Statewide Survey: Special Survey on the California State Budget, Public Policy Institute of California, San Francisco, California, May 2005a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Californians and the Initiative Process, Public Policy Institute of California, San Francisco, California, November 2005b. Baldassare, Mark, PPIC Statewide Survey: Special Survey of Los Angeles, Public Policy Institute of California, San Francisco, California, March 2005c. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Education, Public Policy Institute of California, San Francisco, California, April 2005d. 235 Baldassare, Mark, PPIC Statewide Survey: Special Survey on the California State Budget, Public Policy Institute of California, San Francisco, California, January 2006a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Education, Public Policy Institute of California, San Francisco, California, April 2006b. Billings, John, and Robin M. Weinick, eds., Monitoring the Health Care Safety Net—Book I: A Data Book for Metropolitan Area, Agency for Healthcare Research and Quality, AHRQ Publication No, 030025, 2003, available at http://www.ahrq.gov/data/safetynet/ netfact.htm. California Postsecondary Education Commission, Fiscal Profiles, 2004, Commission Report 04-20, Sacramento, California, December 2004. California State Board of Equalization, “Staff Legislative Bill Analysis— SB 1662 (Peace),” Sacramento, California, 2002. California Tax Credit Allocation Committee (CTAC), “A Description of California Tax Credit Allocation Committee Programs,” California Office of the Treasurer, Sacramento, California, 2006, available at http://www.treasurer.ca.gov/ctcac/program.pdf. Carroll, Stephen J., Cathy S. Krop, Jeremy Arkes, Peter A. Morrison, and Ann Flanagan, California’s K–12 Schools: How Are They Doing? RAND Corporation, Santa Monica, California, 2005. Centers for Medicare & Medicaid Services, Medicaid At-a-Glance 2003, available at http://www.cms.hhs.gov/states/maag2003.pdf. de Alth, Shelley, and David Haskel, Data Appendix, Public Policy Institute of California, San Francisco, California, 2007. Gallup Organization, “Gallup Poll: April Wave 3” and “May Wave 2,” 2003, available at http://brain.gallup.com/. General Accountability Office, Medicaid: Spending Pressures Drive States toward Program Reinvention, GAO/HEHS-95-122, 1995, available at http://www.gao.gov/archive/1995/he95122.pdf. General Accountability Office, Sales Taxes: Electronic Commerce Growth Presents Challenges; Revenue Losses Are Uncertain, GGD/OCE-00165, 2000, available at http://www.gao.gov/archive/2000/g600165. pdf. Glied, Sherry, “Managed Care,” Chapter 13, Volume 1A, in Anthony J. Culyer and Joseph P. Newhouse, eds., The Handbook of Health Economics, North-Holland, Amsterdam, 2000. 236 Greulich, Erica, and John M. Quigley, “Housing Subsidies and the Tax Code: The Case of Mortgage Credit Certificates,” unpublished manuscript, April 2003. Hanak, Ellen, and Elisa Barbour, “Sizing Up the Challenge: California’s Infrastructure Needs and Tradeoffs,” in Ellen Hanak and Mark Baldassare, eds., California 2025: Taking on the Future, Public Policy Institute of California, San Francisco, California, 2005. Hanushek, Eric A., “Assessing the Effects of School Resources on Student Performance: An Update,” Education Evaluation and Policy Analysis, Vol. 19, No. 2, 1997, pp. 141–164. Jepsen, Christopher, and Steven Rivkin, Class Size Reduction, Teacher Quality, and Academic Achievement in California Public Elementary Schools, Public Policy Institute of California, San Francisco, California, 2002. Johnson, Nicholas, Joseph Llobrera, and Bob Zahradnik, A Hand Up, How State Earned Income Tax Credits Help Working Families Escape Poverty in 2003, Center on Budget and Policy Priorities, Washington, D.C., March 3, 2003. Jones, Martha, The Streamlined Sales and Use Tax Agreement: A California Perspective, California Research Bureau, Sacramento, California, February 2005. The Kaiser Commission on Medicaid and the Uninsured, Medicaid Benefit, available at http://www.kff.org/medicaidbenefits/index.cfm. The Kaiser Commission on Medicaid and the Uninsured, The Medicaid Resource Book, July 2002, available at http://www.kff.org/medicaid/ 2236-index.cfm. Krueger, Alan B., “Economic Considerations and Class Size,” Economic Journal, Vol. 113, 2003, pp. 34–63. Land Trust Alliance, “National Land Trust Census,” available at http://www.lta.org/aboutlt/census.shtml. Legislative Analyst’s Office, California Travels: Financing Our Transportation, Sacramento, California, May 2000. Legislative Analyst’s Office, Analysis of the 2001–02 Budget Bill, Sacramento, California, 2002. Legislative Analyst’s Office, Analysis of the 2002–03 Budget Bill, Sacramento, California, 2003. Legislative Analyst’s Office, Cal Facts: California’s Economy and Budget in Perspective, Sacramento, California, December 2004. 237 Legislative Analyst’s Office, Revenue Volatility in California, Sacramento, California, January 2005. Legislative Analyst’s Office, A Primer: Funding Higher Education, Sacramento, California, May 15, 2006a. Legislative Analyst’s Office, Overview of the 2006–2007 May Revision, Sacramento, California, May 15, 2006b. Lewin Group, Comparing Physician and Dentist Fees among Medicaid Programs, Fairfax, Virginia, 2001. Mackenzie, Doris Layton, Sentencing and Corrections in the 21st Century: Setting the Stage for the Future, U.S. Department of Justice Report 189089, July 2001. MaCurdy, Thomas, David C. Mancuso, and Margaret O’Brien-Strain, Does California’s Welfare Policy Explain the Slower Decline of Its Caseload? Public Policy Institute of California, San Francisco, California, January 2002. MaCurdy, Thomas, Raymond Chan, Rodney Chun, Jans Johnson, and Margaret O’Brien-Strain, Medi-Cal Expenditures: Historical Growth and Long Term Forecasts, Public Policy Institute of California, San Francisco, California, 2005. Marrufo, Grecia, Margaret O’Brien-Strain, and Helen Oliver, Child Care Price Dynamics in California, Public Policy Institute of California, San Francisco, California, December 2003. Mazerov, Michael, Developing the Capacity to Analyze the Distributional Impact of State and Local Taxes: Issues and Options for States, Center for Budget and Policy Priorities, Washington, D.C., January 2002. McDonald, Steve, 2003–2004 Florida Annual Policy Survey Report, Survey Research Laboratory, Talahassee, Florida, 2004, available at http://www.fsu.edu/~survey/FAPS/#Reports. National Association of Realtors, “Median Sales Price of Existing SingleFamily Homes for Metropolitan Areas,” August 2005, available at http://www.realtor.org/Research.nsf/files/REL05Q1T.pdf/$FILE/ REL05Q1T.pdf (as of August 3, 2005). National Center for Education Statistics, Digest of Education Statistics, Table 78, Washington, D.C., 2003. National Conference of State Legislatures, Principles of a High-Quality State Revenue System, 4th ed., Washington, D.C., June 2001. 238 NEA Research, Rankings & Estimates: Rankings of the States 2002 and Estimates of School Statistics 2003, National Education Association, May 2003. O’Brien-Strain, Margaret, Laura Moyé, and Freya Lund Sonenstein, Arranging and Paying for Child Care, Public Policy Institute of California, San Francisco, California, December 2003. Office of the Governor, “Strategic Growth Plan Briefing Packet,” 2006, available at http://www.governor.ca.gov/govsite/pdf/press_release_ 2006/SGP_Overview.pdf (as of April 31, 2006). Orzechowski, William, and Robert Walker, The Tax Burden on Tobacco. Orzechowski & Walker, Arlington, Virginia, 2003. Quinnipiac University, “Polling Results: Spitzer Tops Suozzi 5 – 1 Among Democrats, Quinnipiac University New York State Poll Finds; Upstate Economic Outlook Is Bleak,” March 29, 2006, available at http://www.quinnipiac.edu/x1318.xml?ReleaseID-896. Ransdell, Tim, and Shervin Boloorian, Federal Formula Grants: TANF and Welfare Programs, Public Policy Institute of California, San Francisco, California, December 2002. Ransdell, Tim, and Shervin Boloorian, Federal Formula Grants: Federal Child Care Programs, Public Policy Institute of California, San Francisco, California, June 2005. Reed, Deborah, Kim S. Rueben, and Elisa Barbour, Retention of New Teachers in California, Public Policy Institute of California, San Francisco, California, February 2006. Rose, Heather, Jon Sonstelie, and Peter Richardson, School Budgets and Student Achievement in California: The Principal’s Perspective, Public Policy Institute of California, San Francisco, California, 2004. Rose, Heather, Jon Sonstelie, Ray Reinhard, and Sharmaine Heng, High Expectations, Modest Means: The Challenge Facing California’s Public Schools, Public Policy Institute of California, San Francisco, California, 2003. Rueben, Kim, and Shelley de Alth, “Infrastructure Financing in California,” in Ellen Hanak and Mark Baldassare, eds., California 2025: Taking on the Future, Public Policy Institute of California, San Francisco, California, 2005. Schrank, David, and Tim Lomax, The 2005 Urban Mobility Report, Texas Transportation Institute, May 2005, available at http:// mobility.tamu.edu. 239 Sexton, Terri A., and Steven M. Sheffrin, The Market Value of Commercial Real Property in Los Angeles County in 2002, Center for State and Local Taxation, University of California, Davis, California, 2002. Sheffrin, Steven M., and Terri A. Sexton, Proposition 13 in Recession and Recovery, Public Policy Institute of California, San Francisco, California, 1998. Sheffrin, Steven M., and Terri A. Sexton, The Market Value of Commercial Real Property in Los Angeles County in 2002, Center for State and Local Taxation, University of California, Davis, December 2002. Sonstelie, Jon, Eric Brunner, and Kenneth Ardon, For Better or For Worse? School Finance Reform in California, Public Policy Institute of California, San Francisco, California, 2000. Taylor, L. L., Adjusting for Geographic Variations in Teacher Compensation: Updating the Texas Cost-of-Education Index, Joint Select Committee on Public School Finance, Austin, Texas, 2004. U.S. Census Bureau, Statistical Abstract of the United States: 2006, available at http://www.census.gov/prod/www/statistical-abstract. html. U.S. House of Representatives Committee on Ways and Means, Background Material and Data on Programs Within the Jurisdiction of the House Committee on Ways and Means (the Green Book), Government Printing Office, Washington, D.C., 2004. Warren, Jenifer, and Tim Reiterman, “Crowding in Prisons Has State in a Jam,” Los Angeles Times, March 13, 2005. Wear Simmons, Charlene, A Summary of Recommendations for Reforms to the State Budget Process, California Research Bureau, Sacramento, California, 2002. Yelowitz, Aaron S., “Public Policy and Health Insurance Choice of the Elderly: Evidence from the Medicare Buy-in Program,” Journal of Public Economics, Vol. 78, 2000, pp. 301–324. Zuckerman, Stephen, Joshua McFeeters, Peter Cunningham, and Len Nichols, “Changes in Medicaid Physician Fees, 1998–2003: Implications for Physician Participation,” Health Affairs, Web Exclusive, W4-374–384, June 23, 2004. 240 About the Authors TRACY M. GORDON Tracy M. Gordon is a research fellow at the Public Policy Institute of California. Her research interests include state and local public finance, political economy, and urban economics. She has authored reports and journal articles on state and local budgeting, local property taxes, the local initiative process, and so-called “private governments” or common interest developments. She holds a B.A. in human biology from Stanford University and a Ph.D. in public policy with a concurrent M.A. in economics from the University of California, Berkeley. JAIME CALLEJA ALDERETE Jaime Calleja Alderete is a research fellow at PPIC. His research interests include state and local public finance, intergovernmental fiscal relations, and political economy. He has written on the short- and longterm response of local expenditures to intergovernmental grants and his current research examines the role of budget rules and practices on state expenditures. He holds a B.A. in economics from the Instituto Tecnológico Autónomo de México and a Ph.D. in economics from Stanford University. JON SONSTELIE Jon Sonstelie is a professor of economics at the University of California, Santa Barbara. His research interests include several areas in public finance and urban economics, including the effect of public school quality on private school enrollment, the incidence of the property tax, the demand for public school spending, the economics of rationing by waiting, and the effect of transportation innovations on residential locations. He was previously a research fellow at Resources for the Future. He holds a B.A. from Washington State University and a Ph.D. in economics from Northwestern University. 241 PATRICK J. MURPHY Patrick J. Murphy is an adjunct fellow at PPIC and director of the Leo T. McCarthy Center for Public Service & the Common Good at the University of San Francisco, where he is also associate professor in the department of politics. He received his Ph.D. from the University of Wisconsin–Madison, a Master’s of Public Affairs degree from the University of Texas–Austin, and a B.A. from the University of Notre Dame. His research focuses on public management and public policy issues, and he has co-authored several reports and articles on the problems of illicit drug use, the management of drug policy, and the economics of drug selling. In recent years, Murphy’s research has focused on education policy. PING ZHANG Ping Zhang, a research fellow at PPIC, received her Ph.D. in economics from the University of Maryland, College Park, and her master’s and bachelor’s degrees in international finance and international trade, respectively, from Fudan University in Shanghai. Her research interests center on state and local public finance, applied microeconomics, and health economics. She has held positions at the University of Maryland, the International Monetary Fund, and the Agricultural Bank of China. 242 Related PPIC Publications Has Proposition 13 Delivered? The Changing Tax Burden in California (1998) Michael A. Shires, John Ellwood, Mary Sprague Fiscal Rules and State Borrowing Costs: Evidence from California and Other States (1999) James M. Poterba, Kim S. Rueben Patterns in California Government Revenues Since Proposition 13 (1999) Michael A. Shires The State-Local Fiscal Relationship in California: A Changing Balance of Power (1999) J. Fred Silva, Elisa Barbour California’s Natural Resource Programs: Where Does the Money Come From and Where Does It Go? (2002) J. Fred Silva Fiscal Effects of Voter Approval Requirements on Local Governments (2003) Kim S. Rueben, Pedro Cerdán PPIC publications may be ordered by phone or from our website (800) 232-5343 (mainland U.S.) (415) 291-4400 (outside mainland U.S.) www.ppic.org 243" } ["___content":protected]=> string(102) "

R 107TGR

" ["_permalink":protected]=> string(101) "https://www.ppic.org/publication/fiscal-realities-budget-tradeoffs-in-california-government/r_107tgr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8454) ["ID"]=> int(8454) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:37:43" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3658) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 107TGR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_107tgr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_107TGR.pdf" ["wpmf_size"]=> string(7) "2284892" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(472904) "Public Policy Institute of California Fiscal Realities: Budget Tradeoffs in California Government ••• Tracy M. Gordon Jaime Calleja Alderete Patrick J. Murphy Jon Sonstelie Ping Zhang Fiscal Realities: Budget Tradeoffs in California Government ••• Tracy M. Gordon Jaime Calleja Alderete Patrick J. Murphy Jon Sonstelie Ping Zhang 2007 Library of Congress Cataloging-in-Publication Data Fiscal realities: budget tradeoffs in California government / Tracy M. Gordon ... [et al.]. p. cm. Includes bibliographical references. ISBN 978-58213-118-4 1. Budget—California. 2. Fiscal policy—California. 3. Intergovernmental fiscal relations—California. I. Gordon, Tracy M., 1970– HJ325.F57 2007 336.794—dc22 2006102045 Copyright © 2007 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword Issues of state and local governance and public finance took center stage in California’s most recent election. Voters approved five bond issues worth $43 billion. Still, a recent report on California’s fiscal outlook by the Legislative Analyst’s Office warns of “operating shortfalls in excess of $5 billion in both FY 2007–08 and 2008–09, which will require significant actions to eliminate.” California’s growing population is putting considerable strain on the state’s health, education, and social services programs as well as in many other areas. With the persistent shortfalls in operating budgets and the uncertainty about who and what will receive benefits from the bond issues, planning for and financing future programs and projects is increasingly difficult. “What kind of California do you want?” is a difficult question to answer when residents are not sure that there is enough money for even the most modest visions of the future. Ten years ago, PPIC published its first study on California public finance. That study evaluated the accuracy and timeliness of the revenue data reported by more than 7,000 local government entities. Since then, the institute has published nearly two dozen reports examining a wide range of governance and fiscal policy issues, including the effects of Proposition 13, the relationship between local sales taxes and land-use decisions, the changing balance of fiscal power between state and local governments, and the challenges that future growth and changing demography present for future planning. In this volume, five research fellows analyze and compare California’s revenue and spending patterns to those of several other large states and to those of the nation as a whole. Their objective is to give Californians a better sense of how they are doing compared to other states facing similar public spending challenges. California voters have a strong say in how much they are willing to pay for services and the kind and degree of services they expect to receive in return. However, the authors find a considerable gap between the iii services state residents would like their government to provide and the taxes they are willing to pay to receive those services. The authors find that California collects more taxes, fees, and miscellaneous revenues per capita than the average state in the rest of the nation. They also find that California differs markedly in its mix of tax revenues: Whereas the average state in the rest of the nation relies on property taxes for 32 percent of its tax revenues and on income taxes for 25 percent, these proportions are reversed in California as a result of its property tax restrictions and highly progressive income tax. The authors also find that some of Californians’ strongest preferences for policy change are very expensive. And although many observers believe that Proposition 13 has severely limited California’s revenues and expenditures, the authors find that the state’s total spending on public services per capita is higher than the average in the rest of the nation. Like all employers in California, state and local governments must pay higher wages to attract employees who can afford to live in the state. Yet, comparative analyses show that labor market conditions are responsible for only a portion of the differences in government payroll expenditures between California and other states. Nonpayroll expenditures also figure into the equation. So higher spending levels in California do not necessarily translate into more services per capita or higher staffing ratios in K–12 education or in police and fire protection. For example, California spends 50 percent more per capita on public safety than the average in all other states, but it employs proportionately fewer police and fire protection personnel. This report provides a unique explanation of these and other factors involved in California finance—benchmarking California against other states. The authors’ goal was to provide a clearer picture of where California spends its public revenues and how much it takes in. They have met this goal most effectively and have set the stage for a more informed discussion of the state’s budget policies and priorities. David W. Lyon President and CEO Public Policy Institute of California iv Summary California’s recent fiscal tumult has prompted calls for a frank public discussion of state and local budget priorities. In short, what do Californians want from their governments and what are they willing to pay for? Despite exhortations from many quarters, this conversation has yet to occur. This omission may have contributed to delays in the passage of state and local budgets in recent years and to an increasing detachment of citizens from their governments. We argue that deciding what California should raise in revenues and spend on services requires first understanding what it currently does raise and spend and why. The usual framing of state and local budget choices can be an obstacle to this understanding. For example, focusing on the state budget alone, and the General Fund within this budget, obscures the activities of other layers of government and their contributions to overall revenues, expenditures, and service delivery. More important, limiting considerations of public spending to one category at a time (such as K–12 education, health and human services, or public safety) prevents a recognition of the explicit or implicit tradeoffs California has made across spending categories or between revenues and public spending. This report takes a different approach. It considers state and local government finances together using a common framework. This framework is comparative, evaluating California’s fiscal choices against those of other states and of California in earlier periods. The purpose of these comparisons is not to make the case that California should be more like one state or another. Rather, it is to provide a benchmark of comparison. These comparisons also provide insights into the drivers of revenue and public spending. Although this framework ignores certain institutions of California budgeting—for example, the two-thirds voting requirement for enacting a budget and constitutional spending mandates v in K–12 education, transportation, and other areas—these rules are also ultimately policy choices. Throughout the report, we compare California’s fiscal choices to those of the United States excluding California and to those of other large states including New York, Illinois, Florida, and Texas. Although no set of states provides a perfect comparison group, this group shares certain features with California, such as a large and diverse population, high median incomes, and a substantial proportion of the population living in poverty. We rely on the most timely, complete, and accurate source of information on state and local budgets in all states over time. Although the most recent data available for all states at the time of our study were for fiscal year 2001–2002, we feel that this base year provides a reliable indicator of where California stands in relative terms today. Our principal findings are presented below. Revenues On a per capita basis, California’s total revenues were $7,166 in 2002, 16 percent higher than the average in the rest of the nation. California also collected more taxes, fees, and miscellaneous revenues than the average state. In 2002, California state and local governments raised $5,099 per state resident, compared to an average of $4,530 in the United States excluding California, $6,232 in New York, $4,480 in Illinois, $4,299 in Florida, and $3,969 in Texas.1 These differences narrow once we adjust for higher personal incomes in this state. In 2002, California’s own-source revenues represented $155 per $1,000 of personal income, 5 percent higher than the average in the rest of the nation, 7 percent higher than in Florida, 12 percent higher than in Texas, 13 percent higher than in Illinois, and 12 percent lower than in New York. California also relied on roughly the same composition of taxes, charges, miscellaneous revenues, and federal income as other states. Where California differs more dramatically from other states is in its tax mix. The average state in the rest of the nation relies on the property _____________ 1In this report, we refer to fiscal years by their terminal years. For example, fiscal year 2001–2002 is referred to as 2002. vi tax for 32 percent of tax revenues and on income taxes for 25 percent. In California, these proportions are reversed as a result of both limits on the property tax and the state’s highly progressive personal income tax. Spending California’s total per capita spending is higher than that of other states ($8,386 versus $6,939 for the rest of the nation in 2002). Focusing on the core activities of government (excluding public sector utilities and insurance trusts), California’s general expenditures are closer to the average for the rest of the nation but still 15 percent higher.2 Spending differences between California and the rest of the nation are particularly large in public safety ($266 per capita in 2002), social services ($222 per capita), environment and housing ($142), and government administration ($142 per capita) (Figure S.1). What are the sources of these differences? Do California’s higher per capita expenditures reflect the particular challenges California faces, such as a large and diverse population, high labor costs, and stewardship over a rich endowment of natural resources? Or do they reflect policy choices—such as decisions about what level of services to provide and who should receive these services? K–12 Education Contrary to perceptions that California is near the bottom in K–12 education spending, it spends more per capita than the average for all other states, at $1,592 compared to $1,489 in 2002. However, California is also home to more students. As a result, spending per pupil in California is slightly below the average for the rest of the nation, at $7,391 compared to $7,523 per student in 2002. Another difference between California and other states is that California pays higher salaries to school district and state education employees. In 2002, California teachers earned an average of $54,348, _____________ 2Throughout this report, we include mass transit in the transportation category, which is part of the U.S. Census Bureau’s general government sector. In contrast, the bureau considers transit part of the utility sector along with water, gas, and electricity supply. Our concept of general expenditures therefore differs slightly from theirs. vii Expenditures (2002 $) 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 California U.S.–California aHidKHagST–enrhmoi1PadaecElGin2rtiunnoahslhsbveetvlioirprddsseucaroouueetsriirrnncstcvvonaaaaiimmtfttngiiicceeeoooetennnsytstnn Figure S.1—General Expenditures per Capita, 2002 the highest salary for any state in the country according to the National Center for Education Statistics (2003). These salaries partly reflect the premium that all California employers must pay to attract collegeeducated workers from other states. However, about half of the compensation difference between California and other states is not explained by labor market conditions. Higher compensation for public school employees and more K–12 students per capita imply lower staffing ratios per student in California than in other states, all else being equal, and this is the case in California. In 2002, there were 49 teachers per 1,000 students in California, compared to 65 per 1,000 in the rest of the nation. Lower staffing ratios were also evident for other types of school employees, including instructional aides, librarians, principals, and administrative assistants. It is also true that California’s student achievement scores lag those of other states. These differences persist even after taking into account differences in English language proficiency, race or ethnicity, and family characteristics, such as income and parental education (Carroll et al., viii 2005). Although the link between resources and outcomes is not straightforward (see, for example, Hanushek, 1997, and Krueger, 2003), California’s recent focus on accountability standards raises the question of whether schools have adequate resources to meet these goals. Higher Education California spends about 8 percent more per capita than the average for all other states on its public colleges and universities, including the University of California, California State University, and California Community College systems ($634 compared to $586). Moreover, California has a tradition of low tuition and fees and a tendency to collect less in revenue from auxiliary services, such as dormitories, cafeterias, and bookstores. As a result, its expenditures net of revenues exceeded the average for other states by an even greater margin (15%) in 2002. As in the case of elementary and secondary education, California has more college and university students per capita than other states. As a result of higher enrollments, net support per full-time-equivalent student (FTES) is lower in California than in other states. In 2002, California spent 7 percent less per student than the average for all other states ($11,329 versus $12,177). Compared to other large states, California was similar to Florida in net support per FTES but lagged Texas by 10 percent, Illinois by 13 percent, and New York by 21 percent. California’s broad participation in higher education stems from its lower tuition and fees as well as its reliance on the California Community College system, which enrolls approximately two million full-time-equivalent students. In fact, California is one of only a handful of states where a majority of higher education students (60%) are enrolled in two-year institutions. Lower costs per FTES in community colleges could bring down average spending for all students. However, California also lags other states in a key measure of productivity in higher education—the number of degrees conferred per student. In 2002, California’s colleges and universities as a group conferred fewer degrees per student than other states did (an average of 0.15 compared to 0.19 degree per FTES). The gap relative to the rest of ix the nation was particularly large (38%) for associate degrees granted relative to enrollment in two-year institutions. Health Services California spends virtually the same amount per capita as the rest of the nation on health services, including medical care for low-income residents, publicly administered hospitals, and public health ($1,173 versus $1,175 per capita). However, California spends less per capita on the largest individual program in this category, Medicaid (known as Medi-Cal in California), and even less per Medicaid recipient relative to the rest of the nation ($3,113 compared to $4,878). California’s lower Medicaid expenditures stem in part from its younger population. California enrolls more nonelderly adults and children in its Medicaid program than other states do and health care costs for these enrollees are generally about one-fifth of those for the elderly and disabled. California also enrolls a higher proportion of MediCal beneficiaries in managed care plans; and it reimburses these and other health care providers at relatively low rates, despite the higher costs of practicing medicine in this state. As in any public program, high enrollment rates can reflect demographics, eligibility rules, or the rate at which individuals apply for and receive benefits (also referred to as the “take-up” rate). A review of these factors suggests that both generous eligibility rules and high take-up rates are at work. For all groups—nonelderly adults, elderly and disabled individuals, and children—California’s income criteria for receiving Medi-Cal benefits were higher than those of other states in 2002. California also extended Medicaid eligibility to groups for whom the federal government did not mandate coverage, such as working parents. Social Services Beyond medical care, state and local governments provide direct cash payments and other subsidized services to low-income individuals. California spent $533 per capita on these services in 2002, roughly 70 percent more than the average for all other states although only a modest share (6%) of total state and local government spending. x The largest of California’s social services programs is Temporary Assistance for Needy Families (TANF), more commonly known as welfare, or CalWORKs in California. CalWORKs is more generous than TANF programs in other states along several dimensions, including higher maximum allowable benefits per family, milder sanctions for noncompliance, and longer time limits before termination of benefits (MaCurdy, Mancuso, and O’Brien-Strain, 2002). Nevertheless, actual expenditures per individual recipient were lower in California than in other states in 2002 ($4,718 compared to $5,105). This discrepancy may be due to shorter average welfare spells, larger family sizes, or a different allocation of cash assistance versus in-kind benefits in California than in other states. Apart from CalWORKs, state and local governments in California support a variety of community-based services, such as child care, child welfare (including foster care and adoption assistance), and programs geared toward low-income elderly and disabled adults. These services are often characterized by low caseloads and high costs per recipient. For example, California spent roughly $15,000 and $29,000 per case for foster care and adoptions, respectively, in 2002. Although exceeding the average for the rest of the nation, these expenditures were not out of line with those of New York or Illinois. In contrast, California’s expenditures on other community services apart from child care and child welfare were $6,690 per case in 2002, much higher than the average for the rest of the nation or for any of our comparison states. Finally, California spends much more than the average state on supplements to the federal Supplemental Security Income (SSI) program, although the per capita amount ($79) is low in absolute terms. California also has higher-than-average caseloads in this program. These caseloads do not stem from a greater concentration of low-income elderly and disabled individuals in the state but rather from a higher rate of participation within this target group. Transportation In 2002, California’s per capita spending on transportation was roughly the same as the average for all other states. However, in every other year since 1977, California’s spending lagged that of the rest of the xi nation. California is also distinctive in the way it allocates transportation expenditures, with less money going toward highway construction ($154 versus $240 in 2002) and more toward mass transit ($181 compared to $117 in 2002). Mass transit use is higher in California than in many other states, comparable to levels in Illinois. However, automobile use is also high, rivaling that of Texas or Florida. This intensive highway use, coupled with lower capacity, contributes to road congestion and delays. Public Safety California’s third-largest area of government spending is public safety. Although California spends 50 percent more per capita on this function than the average for all other states, it employs fewer police and fire protection personnel. California lags even further behind other large states in police protection (with 72 employees per 1,000 crimes in 2002, compared to 89 and 165 per 1,000 in Illinois and New York, respectively) and fire protection (with 0.9 staff per 1,000 residents in 2002, compared to 1.4 per 1,000 in Florida). As in K–12 education, low staffing ratios in public safety go hand in hand with high operating costs, including payroll. Payroll cost differences are only partly explained by the salary premium that all employers in California must pay. Differences in nonpayroll expenses were even wider, perhaps due to more supplemental services provided in California or to different practices of contracting out for services. Higher payroll costs may reflect differences in quality or other personnel attributes. For example, crime rates have fallen more rapidly in California than in the rest of the nation since the mid-1990s. This trend may be due to the effectiveness of law enforcement in this state as well as to the improving economy, changing demographics, and stricter sentencing policies. On the other hand, falling crime rates and relatively low numbers of inmates per capita may also signal an opportunity to reexamine resource allocation in this area. Housing and Community Development In 2002, California spent roughly 50 percent more on housing and community development than the average for all other states. The majority of these expenditures were pass-throughs of federal funds for xii programs administered by local public housing authorities on behalf of the U.S. Department of Housing and Urban Development. Despite these higher expenditures per capita, fewer households report receiving direct housing aid in California. This result may be due to higher costs per recipient or a greater emphasis on redevelopment in California than in other states. Waiting lists for housing assistance are also longer in California than in other states. Environment California’s per capita expenditures on natural resources were 80 percent higher than the average for all other states in 2002. However, they were less than those of other western states or Florida, which face similar water management issues. Water supply is the largest component of natural resources spending in California. Government Administration In 2002, nearly 5 percent of California’s state and local government expenditures went toward administration. At $447 per capita, California’s spending on administration exceeded the average for all other states in that year and even surpassed that of New York, which spent $390 per capita. Part of the explanation for California’s high administrative costs is that these functions are labor-intensive and labor costs are high in this state. However, as in other areas we have considered, only part of the difference between California and other states is attributable to labor market conditions. Nonpayroll expenditures are also part of the explanation. Unfortunately, it is unclear what these expenditures include and whether they are excessive in the absence of appropriate performance measures. Alternative Tradeoffs The final chapter of this report draws these findings together and identifies the implicit or explicit choices California has made over time, either between revenues and public spending or across spending categories. It also compares California’s budget tradeoffs to the expressed xiii preferences of its residents and, where actual choices deviate from public opinion, it poses some alternative scenarios. These alternative tradeoffs are intended not as policy prescriptions but as illustrations of the kinds of changes necessary to achieve desired policy goals. Perhaps the most important theme of this chapter is that some of Californians’ most desired policy goals are costly. For example, increasing K–12 staffing levels to national levels (129 employees per 1,000 students versus 92 per 1,000) would cost five times as much as reducing corrections costs per inmate to the average for all other states and twice as much as raising property taxes per $1,000 of personal income to the average for all other states. These and other tradeoffs provide an analytical foundation for a far-reaching public conversation on what kind of state California should be. xiv Contents Foreword......................................... iii Summary......................................... v Figures .......................................... xix Tables ........................................... xxiii Acknowledgments................................... xxix Acronyms ........................................ xxxi 1. INTRODUCTION .............................. 1 2. ANALYTICAL TOOLS ........................... Overview ...................................... Comparative Approach ............................ Data Sources ................................... Decompositions and Wages ......................... Treatment of Capital.............................. 3. CALIFORNIA’S REVENUE SYSTEM ................ Total Revenues.................................. General Revenues ................................ Own-Source Revenues ............................ Taxes......................................... Conclusions .................................... 4. ELEMENTARY AND SECONDARY EDUCATION ..... Total Expenditures ............................... Current Expenditures ............................. Capital Expenditures.............................. Outcomes ..................................... Conclusions .................................... 5 5 7 8 10 13 19 19 22 24 25 33 35 35 39 50 53 56 5. HIGHER EDUCATION .......................... Total Expenditures ............................... Higher Education Revenue ......................... Net Support for Higher Education .................... Net Support per Student ........................... 57 57 59 61 62 xv Participation and Student Fees ....................... Staffing and Salaries .............................. Capital Expenditures.............................. Outcomes ..................................... Conclusions .................................... 65 69 73 74 77 6. HEALTH SERVICES ............................ 79 Total Expenditures ............................... 80 Medicaid ...................................... 83 Own Public Hospitals ............................. 95 Other Health Services ............................. 97 Outcomes ..................................... 98 Conclusions .................................... 101 7. SOCIAL SERVICES ............................. 103 Total Expenditures ............................... 103 Welfare (TANF/CalWORKs) ....................... 108 Community Services .............................. 114 State Supplementary Program for Supplemental Security Income.................................... 120 Conclusions .................................... 123 8. TRANSPORTATION ............................ 125 Total Expenditures ............................... 125 Mass Transit ................................... 130 Highways...................................... 137 Outcomes ..................................... 146 Conclusions .................................... 148 9. PUBLIC SAFETY ............................... 151 Total Expenditures ............................... 151 Corrections .................................... 153 Police Protection ................................ 160 Fire Protection .................................. 163 Protective Inspection and Regulation .................. 167 Conclusions .................................... 169 10. ENVIRONMENT AND HOUSING ................. 171 Total Expenditures ............................... 172 Housing and Community Development ................ 174 Natural Resources................................ 181 xvi Parks and Recreation.............................. 184 Sewerage, Solid Waste, and Utilities ................... 188 Conclusions .................................... 189 11. GOVERNMENTAL ADMINISTRATION............. 191 Total Expenditures ............................... 191 Finance and Central Staff .......................... 194 Judicial and Legal ................................ 198 Conclusions .................................... 201 12. CALIFORNIA’S FISCAL TRADEOFFS ............... 203 Californians’ Fiscal Preferences....................... 203 Alternative Tradeoffs.............................. 206 Raising Staffing Levels in K–12 Schools............... 208 Expanding Coverage in Health and Human Services ...... 212 Investing in Roads and Infrastructure ................ 217 Conclusions .................................... 219 Appendix: Data Sources .............................. 221 References ........................................ 235 About the Authors .................................. 241 Related PPIC Publications............................. 243 xvii Figures S.1. General Expenditures per Capita, 2002.............. viii 3.1. General Revenues per Capita, 1977–2002 ............ 23 3.2. General Own-Source Revenues per $1,000 of Personal Income, 1977–2002 ........................... 25 3.3. California State and Local Major Taxes per $1,000 of Personal Income, 1977–2002..................... 27 4.1. Total Expenditures on Elementary and Secondary Education per Capita, 1977–2002 ................. 37 4.2. Number of Public School Students per Capita, 1977–2002 ................................. 43 4.3. Current Expenditures on Elementary and Secondary Education per Student, 1977–2002 ................ 44 4.4. Capital Expenditures on Elementary and Secondary Education per Student, 1977–2002 ................ 51 5.1. Higher Education Total Expenditures per Capita, 1977–2002 ................................. 59 5.2. Higher Education Total Revenues per Capita, 1977–2002 ................................. 61 5.3. Higher Education Net Support per Full-Time- Equivalent Student, 1991–2002................... 64 5.4. Percentage of Population Ages 18 to 24 Enrolled in Public Higher Education, 2002 ................... 66 5.5. Higher Education Revenues per Full-Time-Equivalent Student, 1991–2002 ........................... 68 5.6. Higher Education Revenues and Expenditures per Full- Time-Equivalent Student, 1991–2002 .............. 69 5.7. Percentage of Higher Education Enrollment at Two-Year Institutions, 2002 ............................. 72 5.8. Higher Education Capital Expenditures per Capita, 1977–2002 ................................. 74 6.1. Total Health Expenditures per Capita, 1977–2002 ..... 82 xix 6.2. Own Public Hospital Expenditures per Capita, 1977–2002 ................................. 96 6.3. Other Health Services Expenditures per Capita, 1977–2002 ................................. 97 7.1. Social Services Expenditures per Capita, 1977–2002..... 106 7.2. TANF Caseload per Capita, 1998–2002 ............. 114 8.1. Net Expenditures on Transportation per Capita, 1977–2002 ................................. 129 8.2. Capital Expenditures on Mass Transit per Capita, 1977–2002 ................................. 132 8.3. Current Expenditures on Mass Transit per Capita, 1977–2002 ................................. 132 8.4. Capital Expenditures on Highways per Capita, 1977–2002 ................................. 139 8.5. Freeway and Other Principal Arterial Lane Miles per Million Residents, 1980–2002 .................... 139 8.6. Annual Hours of Delay per Capita in the Largest Urban Areas, 2002 ................................. 149 9.1. Public Safety Expenditures per Capita, 1977–2002 ..... 153 9.2. Corrections Expenditures per Capita, 1977–2002 ...... 155 9.3. Inmates Relative to Population, 1977–2002 .......... 156 9.4. Police Protection Expenditures per Capita, 1977–2002... 161 9.5. Reported Offenses per Capita, 1990–2002 ........... 164 9.6. Fire Protection Expenditures per Capita, 1977–2002 .... 165 9.7. Fire Protection Employees per 10,000 Residents, 1993–2002 ................................. 167 9.8. Protective Inspection and Regulation per Capita, 1977–2002 ................................. 169 10.1. Housing and Community Development Expenditures per Capita, 1977–2002 ......................... 175 10.2. Percentage of All Households Receiving Rent Subsidies or Public Housing, 1977–2002 ................... 178 10.3. Natural Resources Net Expenditures per Capita, 1977–2002 ................................. 184 10.4. Parks and Recreation Net Expenditures per Capita, 1977–2002 ................................. 186 11.1. Governmental Administration Expenditures per Capita, 1977–2002 ................................. 194 xx 11.2. Finance and Central Staff Expenditures per Capita, 1977–2002 ................................. 195 11.3. Total Judicial and Legal Expenditures per Capita, 1977–2002 ................................. 198 xxi Tables 2.1. 2.2. 2.3. 2.4. 2.5. 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 3.7. 4.1. 4.2. 4.3. 4.4. 4.5. 4.6. 4.7. 4.8. 4.9. 4.10. 4.11. State and Local Government Expenditures per Capita, 2002 ...................................... Geographic and Economic Features of California and Comparison States ............................ Job Zones from the O*NET Database .............. Average Wages, by State and Job Zone, 2002.......... Wage Indexes Relative to the United States, by Job Zone, 2002 ................................. State and Local Total Revenues per Capita, 2002 ....... State and Local General Revenues per Capita, 2002 ..... State and Local Own-Source Revenues, 2002.......... State and Local Taxes per $1,000 of Personal Income, 2002 ...................................... State and Local Retail Sales Tax Rates, 2004 .......... Selective Sales Tax Rates, 2005.................... Local Revenues per $1,000 of Personal Income, by Source, 2002 ................................ Education Expenditures per Capita, 2002 ............ Total Expenditures on Elementary and Secondary Education per Capita, 2002 ...................... Components of Current Expenditures per Capita, 2002 ...................................... Components of Public School Enrollment, 2000 ....... Components of Current Expenditures per Student, 2002 ...................................... Components of Compensation per Student, 2002 ...... Public School Staff per 1,000 Students, 2002.......... Components of Average Compensation, 2002 ......... Average Public School Salary Adjusted for Location, 2002 ...................................... Capital Outlay per Additional Student, 1992–2002 ..... NAEP Scores for All Students, 2003 ................ 6 8 14 15 15 21 23 24 26 29 30 32 36 38 40 42 46 46 47 48 49 52 54 xxiii 4.12. NAEP Scores for Students Proficient in English, 2003 ... 5.1. Expenditures on Higher Education per Capita, 2002 .... 5.2. Higher Education Revenues per Capita, 2002 ......... 5.3. Higher Education Net Support per Capita, 2002 ....... 5.4. Net Support per Full-Time-Equivalent Student, 2002 ... 5.5. Residence and Migration of First-Year Undergraduate Students, 2002 ............................... 5.6. Total Payroll, Costs per Employee, and Number of Employees per Full-Time-Equivalent Student, 2002..... 5.7. Wages for Full- and Part-Time Instructional and Noninstructional Employees, 2002................. 5.8. Full-Time Instructional Employee Salary Comparisons, 2002 ...................................... 5.9. Increases in Enrollment and Real Capital Outlays, 1992–2002 ................................. 5.10. Degrees Conferred at Public Colleges and Universities per 1,000 State Residents, Academic Year 2002–2003 ... 5.11. Degrees Conferred per Full-Time-Equivalent Student and Net Support per Degree, 2002................. 5.12. Estimated Degrees Conferred per Full-Time-Equivalent Student at Two- and Four-Year Institutions, 2002 ...... 6.1. Health Expenditures per Capita, 2002 .............. 6.2. Components of Medicaid Expenditures per Capita, 2002 ...................................... 6.3. Composition of State Population, by Eligibility Groups, 2002 ..................................... 6.4. Components of Medicaid Recipients per Nonelderly Adult, 2002 ................................. 6.5. Components of Medicaid Recipients per Elderly or Disabled Adult, 2002 .......................... 6.6. Components of Medicaid Recipients per Child, 2002.... 6.7. Components of Medicaid and SCHIP Enrollment, 2001 ...................................... 6.8. Medicaid Expenditures per Recipient, by Eligibility Group, 2002 ................................ 6.9. Components of Medicaid Expenditures per Recipient, 2002 ...................................... 6.10. Medicaid Fee Index, 2003 ....................... 55 58 60 62 63 66 70 71 71 74 75 76 77 81 85 86 87 89 90 91 92 93 94 xxiv 6.11. Medicaid-to-Medicare Fee Indexes, 2003 ............ 94 6.12. Components of Own Public Hospital Expenditures per Capita, 2002 ................................ 96 6.13. Average Preventable Hospitalizations per 1,000 Persons, 1999 ...................................... 99 6.14. Birth Outcome Indicators per 100 Births, 1999 ........ 100 6.15. Access to Care Measures: Percentage of Population Below 200 Percent FPL, 1999–2000................ 100 7.1. Social Services Expenditures per Capita, 2002 ......... 106 7.2. Welfare Expenditures per Capita, by Source of Funding, 2002 ...................................... 109 7.3. Components of Welfare Expenditures per Capita, 2002 .. 110 7.4. Components of Welfare Caseload, 2002 ............. 111 7.5. Welfare Child Care Expenditures per Capita, by Source of Funding, 2002 ............................. 113 7.6. Total Community Services Expenditures per Capita, by Source of Funding, 2002 ........................ 115 7.7. Components of Foster Care Expenditures per Capita, 2002 ...................................... 116 7.8. Components of Adoptions Expenditures per Capita, 2002 ...................................... 117 7.9. Expenditures on Other Community Services per Capita, by Source of Funding, 2002...................... 118 7.10. Components of Expenditures on Other Community Services per Capita, 2002........................ 119 7.11. Other Community Services Expenditures per Beneficiary from State-Only Sources, 1998–2002 ............... 120 7.12. SSI Expenditures per Capita: Use and State Supplementation, 2002 ......................... 121 7.13. Components of SSI Expenditures per Capita, 2002 ..... 122 7.14. Components of SSI Caseload, 2002 ................ 122 8.1. Expenditure on Transportation per Capita, 2002 ....... 127 8.2. Net Expenditures on Transportation per Capita, 2002 ... 128 8.3. Net Expenditures on Mass Transit per Capita, 2002..... 130 8.4. Expenditures on Mass Transit and Urban Density, 2002 ...................................... 131 8.5. Expenditures on Public Mass Transit per Capita, 2002 ...................................... 131 xxv 8.6. Passenger Miles per Trip, by Transit Mode, 2002 ...... 134 8.7. Operating Cost per Passenger Mile, by Mode, 2002 ..... 134 8.8. Components of Operating Cost per Passenger Mile, by Mode, 2002 ................................. 135 8.9. Components of Employee Compensation per Passenger Mile, 2002.................................. 136 8.10. Mass Transit Salary Adjusted for Location, 2002 ....... 137 8.11. Expenditures on Highways per Capita, 2002 .......... 138 8.12. Current and Capital Expenditures on Highways per Capita, 2002 ................................ 138 8.13. Capital Expenditures on Highways per Capita, 1957–1997 ................................. 140 8.14. Freeways and Other Principal Arterial Growth in Lane Miles and Population, 1980–2002 ................. 141 8.15. Traffic on Freeways and Other Principal Arterials, 2002 .. 142 8.16. Traffic on Freeways and Other Principal Arterials in Urban and Rural Areas, 2002..................... 143 8.17. Current Highway Expenditures per Capita, 2002....... 143 8.18. Components of Current Highway Expenditures per Capita on Local Roads and Highways, 2002 .......... 144 8.19. Components of Current Highway Expenditures per Capita on State Roads and Highways, 2002........... 145 8.20. Annual Highway Payroll Adjusted for Location, 2002 ... 145 8.21. Road Congestion and Delay in 85 Urban Areas, 2002 ... 148 9.1. Public Safety Expenditures per Capita, by Type of Service, 2002 ................................ 152 9.2. Corrections Expenditures per Capita, by Type, 2002 .... 154 9.3. Components of Corrections Expenditures per Capita, 2002 ...................................... 156 9.4. Components of Corrections Expenditures per Inmate, 2002 ...................................... 158 9.5. Average Yearly Payroll in the Correctional System, 2002.. 158 9.6. Average Occupancy of Correctional Facilities, 2002 ..... 160 9.7. Real Capital Outlay per Additional Inmate in State Prisons, 1992–2002 ........................... 160 9.8. Police Protection Expenditures per Capita, by Type of Government, 2002 ............................ 161 xxvi 9.9. Components of Police Protection Expenditures per Capita, 2002 ................................ 163 9.10. Average Yearly Payroll for Police Protection, 2002 ...... 163 9.11. Fire Protection Expenditures per Capita, by Type of Government, 2002 ............................ 164 9.12. Components of Fire Protection Expenditures per Capita, 2002 ...................................... 166 9.13. Average Yearly Payroll for Fire Protection, 2002........ 166 9.14. Protective Inspection and Regulation Expenditures, by Type of Government, 2002 ...................... 168 10.1. Environment and Housing Expenditures per Capita, 2002 ...................................... 174 10.2. Housing and Community Development Expenditures per Capita, 2002.............................. 175 10.3. Housing and Community Development Federal Intergovernmental Revenue per Capita, 2002 ......... 176 10.4. Housing Assistance Caseloads and Components, 2002 ... 177 10.5. Rent Burdens as a Percentage of Household Income, 2002 ...................................... 179 10.6. Average Months on Waiting List for Housing Program Participants, 2002............................. 179 10.7. Statewide Average Fair Market Rents Weighted by Population in Each Market Area, 2002 .............. 180 10.8. Housing and Community Development Real Capital Outlay, 1992–2002 ........................... 181 10.9. Natural Resources Expenditures per Capita, 2002 ...... 182 10.10. Natural Resources Net Expenditures per Capita, 2002 ... 183 10.11. Parks and Recreation Expenditures per Capita, 2002 .... 185 10.12. Parks and Recreation Net Expenditures per Capita, 2002 ...................................... 185 10.13. Federal, State, and Local Parks and Natural Resource Lands, 2002................................. 187 10.14. Parks and Recreation Expenditures per Acre, 2002 ...... 187 10.15. State and National Park Visits per Capita, 2002........ 188 10.16. Utility Expenditures per Capita, 2002............... 188 10.17. Net Enterprise and Utility Expenditures per Capita, 2002 ...................................... 189 xxvii 11.1. Governmental Administration Expenditures per Capita, 2002 ...................................... 193 11.2. Components of Total Expenditures per Capita on Finance and Central Staff, 2002 ................... 195 11.3. Components of Finance and Central Staff Payroll Expenditures, 2002............................ 196 11.4. Average Payroll for Finance and Central Staff Employees, 2002 ...................................... 197 11.5. Components of Total Judicial and Legal Expenditures per Capita, 2002.............................. 199 11.6. Components of Judicial and Legal Payroll Expenditures per Capita, 2002.............................. 200 11.7. Average Payroll for Judicial and Legal Employees, 2002 .. 201 12.1. California Preferences for Governmental Spending, January 2006 ................................ 204 12.2. Budget Priorities in Other States, 2002 and 2003....... 207 12.3. Alternative Scenario #1: Raising Staffing Levels in K–12 Schools .................................... 209 12.4. Alternative Scenario #2: Expanding Medi-Cal and CalWORKs ................................. 215 12.5. Alternative Scenario #3: Investing in Roads and Other Infrastructure ................................ 219 A.1. Urban Areas in Texas Transportation Institute Study, 2002 ...................................... 234 xxviii Acknowledgments This report reflects the contributions of several individuals and institutions. We are grateful to The James Irvine Foundation for providing a generous planning grant that enabled us to think broadly about the goals of the report and to consult with experts in government, academia, journalism, business, and the nonprofit sector. We are also indebted to these experts for providing us with their counsel. This report would not have been possible without the comprehensive, high-quality data produced by the U.S. Census Bureau. We spent several days on site at the Governments Division and engaged in frequent consultations with Census Bureau staff by telephone and email. We are especially indebted to Henry Wulf, Chief of the Governments Division, as well as David Kellerman, Christopher Pece, Steven Poyta, Jeffrey Little, Jacques C. Kusseling, and Joseph Bruzzese. We are also indebted to those who reviewed this report in its entirety: Donald Boyd of the Rockefeller Institute of Government at SUNY, Albany; Heather Rose of PPIC; Kim Rueben of the Tax Policy Center and the Urban Institute; Peter Schaafsma of the California State Assembly Republican Caucus; and Fred Silva formerly of PPIC and now consultant to the New California Network. In addition, several individuals provided comments on specific subject areas: California Legislative Analyst, Elizabeth Hill, as well as several members of her staff including Brad Williams, David Vasché, Robert Manwaring, Steve Boilard, Daniel Carson, Todd Bland, Dana Curry, Gregory Jolivette, Michael Cohen, Mark Newton, and Brendan McCarthy; Margaret O’Brien-Strain of the SPHERE Institute; Ellen Hanak and Marianne Bitler of PPIC; and Phyllis Myers of State Resource Strategies. Mark Baldassare and Max Neiman of PPIC and Paul Lewis formerly of PPIC and now at Arizona State University offered guidance on this project from the initial stages to its completion. PPIC’s communications group improved the report’s clarity and exposition. Finally, this report xxix would not have been possible without the contributions of several excellent research associates at PPIC: Shelley de Alth, David Haskel, Supreet Kaur, and Kate Jordan. Although we are pleased to thank each of these persons and organizations, all findings, conclusions, errors, and omissions in this report are our own. xxx Acronyms AC Transit ACF AFDC AHRQ AMT BART BEA CalPERS CalSTRS CalWORKs CCC CCDF CDBG CMS CMSA CPEC CPI CPS CSR CSU CVP DMV DSH DWR FFS FMR FPL FTA FTES GAO Alameda–Contra Costa Administration for Children and Families Aid to Families with Dependent Children Agency for Healthcare Research and Quality Alternative Minimum Tax Bay Area Rapid Transit Bureau of Economic Analysis California Public Employees’ Retirement System California State Teachers’ Retirement System California Work Opportunity and Responsibility to Kids California Community Colleges Child Care and Development Fund Community Development Block Grant Centers for Medicare and Medicaid Services Consolidated MSA California Postsecondary Education Commission Consumer Price Index Current Population Survey Class Size Reduction California State University Central Valley Project Department of Motor Vehicles Disproportionate Share Hospital Department of Water Resources Fee for Service Fair Market Rent Federal Poverty Line Federation of Tax Administrators Full-Time-Equivalent Student Government Accountability Office xxxi HCBS HUD IPEDS LACMTA LAMTA LAO LEA LIHEAP LIHTC MIC MOE MSA MSIS MUNI NAEP NCES NEA NEC NOL NTD O*NET OES PAB PHA PMSA RDA RVU SAT9 SCHIP SCORP SSBG SSD SSI SSP SSTP SWP Home- and Community-Based Services Housing and Urban Development Integrated Postsecondary Education Data System Los Angeles County Metropolitan Transportation Agency Los Angeles Metropolitan Transit Authority Legislative Analyst’s Office Local Education Agency Low-Income Home Energy Assistance Program Low-Income Housing Tax Credit Manufacturer’s Investment Credit Maintenance of Effort Metropolitan Statistical Area Medicaid Statistical Information System Municipal Railway National Assessment of Educational Progress National Center for Education Statistics National Education Association Not Elsewhere Classified Net Operating Loss National Transit Database Occupational Information Network Occupation Employment Statistics Private Activity Bond Public Housing Authority Primary Metropolitan Statistical Area Redevelopment Agency Relative Value Unit Stanford Achievement Test, Ninth Edition State Children’s Health Insurance Program State Comprehensive Outdoor Recreation Plan Social Services Block Grant State Summary Datamart Supplemental Security Income State Supplemental Program Streamlined Sales Tax Project State Water Project xxxii TANF TTI UC Temporary Assistance for Families with Dependent Children Texas Transportation Institute University of California xxxiii 1. Introduction California has been on a fiscal rollercoaster in recent decades. The 1990s started with a recession that was mild by national standards but strongly felt in this state. At the time, many observers expressed alarm about a “structural deficit,” or a permanent mismatch between state revenues and expenditures, apart from effects of the business cycle. They also pointed out that local governments were still struggling with uncertain revenues after the passage of Proposition 13. These concerns diminished as government revenues soared in the late 1990s, fueled by rising income from stock options and capital gains. Revenue increases funded popular program expansions in education, health, social services, and tax relief. In hindsight, of course, these gains were short-lived. The collapse of the stock market and economic downturn of 2001 generated massive losses for the state ($12 billion out of a roughly $100 billion budget). Budget shortfalls widened as economic conditions worsened and corrective actions yielded disappointing results. By May 2003, the Department of Finance projected a $38 billion gap between revenues and expenditures. More recently, state coffers have rebounded thanks to stronger than expected personal and corporate income tax receipts. Local governments are also in a more secure fiscal position, having won passage of Proposition 1A, which prohibits future reductions in allocations of state funds. Nevertheless, concerns persist about an underlying imbalance between state revenues and expenditures. For example, the Legislative Analyst’s Office (LAO) has warned that higher energy costs and a steeper-than-expected slowdown in real estate markets could expose the state to more budget risk. The LAO has also noted that the state will face substantial borrowing costs and large unfunded liabilities for state government retiree health plans (Legislative Analyst’s Office, 2006b). Throughout these budgetary swings, there have been numerous calls for reform. A report by the California Research Bureau tabulates 1 recommendations from nearly a dozen state commissions, business groups, civic task forces, and academic studies since 1991 (Wear Simmons, 2002). Although there appears to be a consensus that California’s budget process is “broken,” there is less agreement about what should be done to fix it. An oft-cited reason for inaction on budget reform is that Californians have not confronted a fundamental disconnect in preferences for both lower taxes and higher government spending. By this argument, a “something for nothing” mentality dating to the 1970s tax revolt pervades budget discussions and precludes consideration of revenue increases needed to support publicly favored programs. On the other hand, an equally vociferous group argues that there is no disconnect but instead an inefficient public sector containing enough waste to allow spending reductions without harming the quality of services. Interestingly, a common refrain in both arguments is that California needs to have a frank public conversation about its budget priorities. In short, what kind of a state do Californians want and what are they willing to pay for? Yet, despite these exhortations on both sides of the political spectrum, such a conversation has not occurred. One obstacle to this kind of far-reaching budget conversation is insufficient information. Given the current framework for budget decisions in California, it is not clear how to assess what state and local governments raise, spend, and produce. For example, budget decisions are typically framed within a category of spending (e.g., categorical versus unrestricted aid for schools), by level of government (e.g., cities and counties versus school districts), and by source of funds (e.g., the state General Fund versus special funds). These distinctions are not helpful for understanding broader tradeoffs between spending categories (e.g., K–12 education versus health services) or between revenues and public spending. Moreover, they do not permit a discussion of whether to continue these tradeoffs. This report takes a different approach. It analyzes revenues and spending in every major category (e.g., education, health, social services, and so on) using a common framework that ignores distinctions between levels of government and sources of funds. The framework is 2 comparative, evaluating California’s fiscal tradeoffs against those of other states. The purpose of these comparisons is not to argue that California should be more like one state or another. Rather, it is to provide a basis of comparison. These comparisons further allow us to identify common drivers of spending—such as program caseloads or workload, input costs, and the level or quality of services provided—and to highlight which tradeoffs reflect the particular challenges this state faces and which stem from policy decisions. According to this framework, institutions such as voter-approved limits on revenues and spending are also policy choices. Finally, the report assesses current tradeoffs against public opinion. Where actual tradeoffs deviate from these expressed preferences, it presents alternative scenarios. These scenarios are not policy prescriptions. Indeed, some of them may be impractical or infeasible. Instead, these scenarios are intended to reinforce our findings and to illustrate the magnitudes of past tradeoffs in California finances as well as the difficulty in some cases of undoing these tradeoffs. Taken together, the findings in this report provide an analytical foundation for a public conversation about California’s budget priorities. As in other previous studies, we identify some potentially conflicting goals and tough choices. However, there is no reason to expect that these vexing questions about the goals of government should not be amenable to the same creativity and innovation for which California is known in other sectors. The remainder of this report is organized as follows. The next chapter explains the overarching approach, concepts, and analytical tools used throughout the report. Chapter 3 analyzes California’s revenue system, including taxes, fees, charges, other state and locally generated revenues, and federal contributions. Subsequent chapters examine state and local government spending in each major category: elementary and secondary education, higher education, health, social services, transportation, public safety, environment and housing, and government administration. The report concludes with a review of findings from each chapter and an exploration of alternative budget tradeoffs. 3 2. Analytical Tools This chapter reviews the analytical approach, data, and methods used in this report. It begins with a snapshot of California’s government finances in fiscal year 2001–2002, the most recent year for which our primary data were available.1 The discussion in this chapter is not a summary but a point of departure for the chapters that follow. The remainder of this chapter explains how we will analyze California’s revenues and expenditures in greater detail using the experiences of other states. Overview In 2002, California’s state and local governments collected nearly $251 billion in revenues and administered over $290 billion in spending.2 The gap between these two numbers arises from several features of that particular year. First, as discussed in the previous chapter, a collapse of the stock market and economic downturn in 2001 led to a plunge in state revenues of about $12 billion in 2002. It also generated investment losses of approximately $23 billion for California’s insurance trust funds, such as the public employee retirement plans California Public Employees’ Retirement System (CalPERS) and California State Teachers’ Retirement System (CalSTRS). An added complication in California was the state energy crisis of 2001, during which the state purchased electricity on the wholesale market on behalf of investor-owned utilities. Although later repaid from a bond secured by consumer rate payments, this loan appeared in 2002 as a state expenditure of approximately $7 billion. _____________ 1These data, through 2004, are available on PPIC’s website (www.ppic.org) in a user-friendly format. They are also available from the Census Bureau (ftp://ftp2.census. gov/pub/outgoing/govs/special60/). 2In the remainder of this report, we refer to fiscal years by their terminal year. For example, fiscal year 2001–2002 is referred to as 2002. 5 On a per capita basis, California’s total expenditures were $8,386 and its revenues were $7,166. Other analysts have used personal income rather than population as a benchmark for expenditure comparisons. We use per capita figures here because controlling for income incorporates several factors that we will investigate separately below (e.g., higher labor costs). We use the average for the rest of the nation rather than the United States overall because California is such a large state that it can skew national averages. Rather than total expenditures, we can also focus on the core activities of government by excluding public sector utilities and insurance trusts. By this measure, California’s general expenditures were $6,735 per capita, roughly 15 percent higher than the average for all other states. Per capita spending differences between California and the rest of the nation were particularly large in public safety ($266), social services ($222), environment and housing ($142), and government administration ($142) (Table 2.1). Table 2.1 State and Local Government Total Expenditures per Capita, 2002 K–12 education Health Public safety Higher education Environment and housing Social services Government administration Transportation Utilities Insurance trusts Interest on general debt Other Liquor Total California ($) 1,592 1,173 778 634 590 533 447 430 824 746 248 391 — 8,386 U.S.–California ($) 1,489 1,175 512 586 448 311 305 482 438 568 263 344 17 6,939 Difference $% 102 6 –2 0 266 34 47 7 142 24 222 42 142 32 –52 –12 385 47 179 24 –15 –6 47 12 –17 — NOTES: Throughout this report, numbers in the tables may not sum to totals because of rounding. Also, we often refer to percentage differences in the text although we do not report these percentages in the tables for the sake of brevity. 6 Comparative Approach These figures raise a series of questions. Is California a high-tax and high-spending state? Has it always been so? What is high or low? California’s higher per capita expenditures may reflect policy choices— such as decisions to collect more in revenues and provide more in services—or background conditions—such as higher costs of living and thus higher wages for both public and private sector workers in the state. California’s public finances are also undoubtedly shaped by its population. More school-age children place greater demands on the K–12 education system, and more low-income families and elderly or disabled individuals can strain state and federally funded social programs. Answering these questions requires a basis for comparison. One benchmark is historical experience. Is California raising and spending more or less now than in 1977? However, intervening events— fluctuations in the economy, demographic shifts, and technological improvements—may lead to observed differences between now and the past. Disentangling which events matter for revenues and public spending is complicated in practice. We can better understand which factors affect government finances by broadening our analysis to include the experiences of other states, which are presumably affected by the same general trends. These comparisons are even more meaningful if these states are similar to California. California is a large and complex state, which many claim defies comparison with other states, if not with other countries. Yet, California shares certain features with other states, particularly other large states including Florida, Texas, Illinois, and New York (Table 2.2). Like California, all of these states except Texas are urbanized and densely populated. Similar to New York and Illinois, California is also a relatively high-income state. At the same time, a substantial share of Californians lives in poverty, as in Florida, New York, and Texas. Apart from its size, California’s population is notable for its diversity. Like Texas, California is a “majority minority” state, meaning that more than half of its residents are nonwhite. Both states are also home to large immigrant populations, which tend to be predominantly younger. In the remainder of this report, we will compare California to the rest of the 7 Table 2.2 Geographic and Economic Features of California and Comparison States California U.S. total Florida Illinois New York Texas Population 36,132,147 296,410,404 17,789,864 12,763,371 19,254,630 22,859,968 Median Below NonTotal Area Household Poverty Hispanic Foreign(sq. miles) Income ($) Level (%) White (%) Born (%) 163,696 50,220 13.4 44.5 26.5 3,794,083 43,564 12.7 67.4 11.9 65,755 39,871 13.1 62.8 17.6 57,914 47,977 11.3 66.2 13.3 54,556 46,195 13.5 61.1 20.8 268,581 40,674 16.3 49.8 15.6 nation (the United States excluding California) and to these other large states, often referring to them as our “comparison group.” Data Sources This report relies primarily on government finance and payroll data collected through the U.S. Census Bureau’s Census of Governments and Annual Government Finance series. These data are the most complete, accurate, and timely source of information available on state and local government finances. The Census Bureau canvasses all state and local governments in the nation in years ending in 2 and 7 and all states and a sample of local governments annually.3 Data are available for all states in the same basic format going back to the 1950s. To maximize the consistency of the data over time, we rely mainly on a series starting in 1977. Although every state has its own reporting system, the Census Bureau collects and recasts this information using its own standard format. This format ignores distinctions between funds (including borrowed funds and prior year balances) although it does track intergovernmental revenues and capital expenditures separately.4 The format classifies revenues and expenditures according to functions rather _____________ 3Local government totals in the annual survey are generally subject to a sampling variability of 1 percent or less. 4These expenditures do not include debt service payments, which have their own category. 8 than programs. In what the Census Bureau refers to as the general government sector, these functions are elementary and secondary education, higher education, health, social services, transportation, public safety, environment and housing, and government administration. In addition, the Census Bureau tracks revenues and expenditures of public water, gas, electric, and transit utilities separately. In our discussion we include the first three utilities with the environment and housing function and transit with transportation. The Census Bureau also gathers data on the finances of insurance trusts such as public employee retirement, unemployment compensation, workers’ compensation, and disability insurance systems. We do not analyze these systems here. These data have their limitations. The most recent year for which Census Bureau data were available at the time of this writing for both states and localities was 2002. In some sense, this year represents a highwater mark for government finances because it reflects higher spending during the economic boom of the 1990s. However, as the most recent year available, it was also the best predictor of government finances today. The Census Bureau’s standard reporting format also means that these data may differ from totals in state and local government reports. For example, some observers have expressed concern that Census Bureau data on state and local spending separately count both state assistance to local governments and local expenditures of these funds. Moreover, relationships between functions and programs are not always clear. For example, Medicaid expenditures are split between several categories, including vendor payments, health, hospitals, and cash assistance. To address these concerns, we performed an extensive crosswalk between the Census Bureau data and California state and local sources, including the Governor’s Budget and Annual Financial Reports from the State Controller’s Office. The results of this crosswalk (outlined in de Alth and Haskel, 2007) explain the relationship between government programs in California and Census Bureau functional categories. Through this exercise, we were also able to reproduce Census Bureau figures for expenditures and revenues within 2 and 3 percent, respectively. Together with the reputation of the Census of 9 Governments for producing high-quality data for forecasting and research, this result makes us reasonably confident in the data for all states and previous years. Using the Census Bureau data, we consider state and local government finances together in each major functional category. We do so mainly because we are interested in the overall activities of the public sector in California and not in the state-local fiscal relationship. There is also a more practical reason. States have different traditions of allocating revenues and program responsibility, and ignoring these differences by focusing on the state alone can lead to mistaken inferences. For example, California’s state expenditures on K–12 education are higher than most other states’ expenditures, although state and local expenditures are near the national average. This pattern is an artifact of the centralization of education finance in this state after passage of Proposition 13 in 1978. Finally, the Census of Governments collects data on the government payroll and employment using the same classification scheme used for tracking government finances. Payroll data do not include information on benefits such as paid leave, overtime, and employer contributions to health and disability insurance, savings and retirement plans, workers’ compensation, and unemployment funds. Where possible, we discuss differences in nonwage compensation in each functional area. Decompositions and Wages We analyze public spending in this report using a series of basic accounting relationships. First, for any area of public spending, expenditures per capita depend on caseloads—the number of recipients or users of public goods and services—and the costs per recipient: Expenditures = Caseload  Expenditures Population Population Case (2.1) Second, as discussed above, the workload for any public program depends on a state’s demographic characteristics, such as the size of its population and its age or income distribution, as well as on the number of service users within this target group. We can separate the influence of these two factors using the following relationship: 10 Caseload = Participants  Potentially Eligible Population Potentially Eligible Population (2.2) Here, the second term reflects the demographic features of a state whereas the first term represents the combined effect of policymaker decisions and individual behaviors. For example, policymakers may decide who is eligible for a program, for what services they are eligible, and for how long they may receive benefits. These rules may also be determined at the federal level as a condition for states to receive matching funds. Eligible individuals then decide whether to use or “take up” services. When demand for services outstrips capacity, an administrative process may determine who receives benefits, or rationing may occur based on wait times or congestion. In some cases, we are able to further decompose the first term of Eq. (2.2) into take-up rates and program rules, but often there are insufficient data for this task. Finally, as noted above, expenditures per capita also depend on the costs of labor and other inputs such as energy or land prices. Expenditures per case can be represented as the product of these costs and the average quantity of inputs used: Expenditures = Units  Price Case Case Unit (2.3) In the same way that the above relationships shed light on the determinants of expenditures per capita within a state, they can also help explain spending differences between states. For example, where  represents the difference between California and another state or the average for all other states:  Expenditures   Caseload +  Expenditures Population Population Case (2.4) 11 That is, the difference in expenditures per capita between two states is equal to the difference in cases per capita plus the difference in expenditures per case.5 Labor costs deserve special attention in any analysis of government expenditures because public services tend to be labor-intensive. Indeed, perceptions of quality in the public sector often depend on how much labor is involved (e.g., smaller class sizes in public schools). In California, public sector wages may be particularly high because of the higher costs of living in the state. At the same time, California also offers certain amenities, such as its climate, geography, and culture. Market-basket cost adjustments, such as the U.S. Bureau of Labor Statistics Consumer Price Index (CPI), fail to account for how these amenities might affect prices. Although this point is well understood in real estate, where similar houses with additional rooms, swimming pools, or larger lots command a price premium, it is frequently overlooked in discussions of regional cost differences. To capture both the costs and benefits of living in California, we rely on the concept of a “location premium.” This premium is the difference in wages that employers in California must pay to compete in a national labor market. In principle, it may be either positive (i.e., if employers have to pay more to attract and retain workers) or negative (i.e., if workers have to give up some of their earnings to live in a desirable place). For any occupation, the location premium is based on the average salary difference for a range of occupations requiring comparable education, skills, and experience. For example, the location premium for teachers is based on other occupations that typically require a four-year college degree. We use a range rather than simply comparing average public school teacher salaries across states to avoid conflating labor costs with policy decisions about teacher pay.6 _____________ 5In fact, this is an approximation because of an arbitrary weight that is introduced when differences are taken in discrete terms rather than continuously (e.g., a Taylor expansion). 6This is equivalent to the Comparable Wage Index approach (e.g., Taylor, 2004). 12 In this report, we use labor market zones as defined by the Occupational Information Network database, or O*NET (see the appendix). This database uses surveys of employers and employees to define ranges of skill, knowledge, and experience known as labor market zones. Zone 1 corresponds to occupations requiring no previous workrelated skills, knowledge, or experience, whereas Zone 5 refers to jobs that require extensive preparation, often including an advanced degree. Table 2.3 describes these and intermediate zones more fully. In all labor market zones, California employers must offer higher wages than the U.S. average (Table 2.4). The gap is largest in Zones 3 through 5, where more education and experience are required. In New York, the gaps are even larger, whereas in Texas and Florida, the gaps are negative, suggesting that comparable workers earn less than the national average in these states. Table 2.5 summarizes the percentage difference in average salaries between each state and the U.S. average for each labor market zone. To illustrate what this means for understanding California’s expenditures per capita, consider a government agency hiring a worker from Zone 3. It would have to pay an annual salary of nearly $43,000 for the same worker who would earn $37,000 per year in the rest of the country, a 14 percent difference as shown in the third column of Table 2.5. Suppose we actually observe the California agency paying salaries at 25 percent above the average for a similar occupation in all other states. The remaining 11 percent is unexplained by labor market conditions. Treatment of Capital State and local budgeting practices are usually based on cash accounting, or measuring current revenues against current expenditures. This method captures debt service and rental payments but otherwise ignores the value of long-lived investments such as buildings, roads, or land and mineral rights. A capital budget would take into account these assets as well as their depreciation. Capital budgeting is not widespread, however. In this report, we focus on current revenues and expenditures, adjusting historical figures for inflation using 2002 dollars (CPI-U). We 13 Table 2.3 Job Zones from the O*NET Database Zone 1 Description No previous work-related skill, knowledge, or experience. May require high school diploma or GED or formal training course to obtain license. Training on the job from experienced co-worker for a few days to months. Skills involve following instructions and helping others. Examples include bus drivers, forest and conservation workers, general office clerks, home health aides, and waiters/waitresses. 2 Some previous work-related skill, knowledge, or experience. Usually requires a high school diploma and may require vocational training or job-related coursework. Training from a few months to one year with experienced employees. Skills involve using knowledge and skills to help others. Examples include drywall installers, fire inspectors, flight attendants, pharmacy technicians, salespersons (retail), and tellers. 3 Previous work-related skill, knowledge, or experience. Usually requires training in vocational schools, related on-the-job experience, or an associate’s degree. One or two years of training involving on-the-job experience and informal training with experienced workers. Skills involve using communication and organizational skills to coordinate, supervise, manage, or train others. Examples include dental assistants, electricians, fish and game wardens, legal secretaries, personnel recruiters, and recreation workers. 4 Minimum of two to four years of work-related skill, knowledge, or experience. Most require four-year bachelor’s degree. Usually need several years of workrelated experience, on-the-job training, or vocational training. Skills involve coordinating, supervising, managing, or training others. Examples include accountants, chefs and head cooks, computer programmers, historians, pharmacists, and police detectives. 5 Extensive skill, knowledge, and experience. Bachelor’s degree is minimum formal education required, but many require graduate school (master’s degree, Ph.D., M.D., or J.D.). Usually, a person will already have required skills, knowledge, work-related experience, or training. Very advanced communication and organizational skills. Examples include athletic trainers, lawyers, managing editors, physicists, social psychologists, and surgeons. do not discuss the value of assets, including trust funds set aside to cover future liabilities (e.g., public employee pension systems such as CalPERS and CalSTRS) or outstanding debt. As mentioned above, we ignore 14 Table 2.4 Average Wages, by State and Job Zone, 2002 (dollars) California U.S.–California Florida Illinois New York Texas United States 1 22,067 20,681 19,139 21,162 22,446 19,336 20,831 2 29,771 26,896 25,000 28,283 30,055 24,894 27,223 Zone 3 42,660 36,759 34,261 38,629 42,875 34,570 37,432 4 62,711 54,294 51,958 55,026 67,591 52,048 55,285 5 83,331 73,242 73,370 72,214 85,652 73,535 74,371 Table 2.5 Wage Indexes Relative to the United States, by Job Zone, 2002 California U.S.–California Florida Illinois New York Texas Zone 1 2 3 45 106 109 114 113 112 99 99 98 98 98 92 92 92 94 99 102 104 103 100 97 108 110 115 122 115 93 91 92 94 99 distinctions between funds, including so-called “nongovernmental cost funds” such as bond funds. We do this because we are trying to capture total spending in each major area and, although these funds may be predetermined or “off-budget” in a given year, they are subject to change over time. Nevertheless, the uneven nature of capital spending and the durable nature of these investments require that we treat these expenditures differently. Thus, in addition to examining capital outlays in a given year, we evaluate trends in capital spending per additional user over a longer period.7 This method is not perfect: It ignores differences in the _____________ 7These comparisons are usually from 1992 to 2002. Because the Census Bureau does not report expenditures for state and local governments separately by state for 2001, they include cumulative spending over nine years and population growth over ten years, except in Chapters 4 and 8, which interpolate the data for 2001. We could also consider 15 age of a capital stock at the start of the period, which may affect the need for additional spending (e.g., retrofitting). It also does not consider the efficiency with which resources are used and how the intensity of use might affect the need for additional spending. Wherever possible, we note these factors and how they might affect relative spending differences. Finally, we refer throughout this report to “fiscal tradeoffs,” or choices across budget categories. The concluding chapter goes even further, asking what alternative tradeoffs might look like. Of course, these choices may be implicit rather than explicit. That is, individual government revenue or spending decisions may be made in isolation, or without regard for “spillover” effects on other areas. However, at the state, local, or even household level, these choices are necessarily interdependent. If state and local governments choose to collect less in revenue or spend more on a particular program, or are compelled to do so because of demographic and economic pressures, there will be fewer resources left over for other purposes. Having budget information cast in these terms is therefore essential to evaluate the consequences of current budget decisions and to decide whether the state ought to do things differently. ______________________________________________________________ debt service payments per year as a way to smooth capital expenditures over time. However, these payments reflect interest rates and terms of borrowing as well as capital outlays. 16 A Note on Budget Concepts Discussions of budget concepts can be confusing. To clarify, we define a few terms. For federal, state, and local governments, the debt is the total amount of resources owed, just as it is for households. At the federal level, the deficit is the difference between revenues and spending in any given year. It is also the incremental addition to the debt. When state revenues are insufficient to cover expenditures, this is referred to as a budget shortfall or gap. Some claim that states cannot run deficits because, unlike the federal government, they face constitutional balanced budget rules or limits on borrowing. In fact, the picture is more complicated. In about onethird of all states, including California, budgets must be balanced at the start but not at the end of a fiscal year. Also, although state constitutions often require voter and legislative approval for the issuance of “full faith and credit” debt, or general obligation bonds, other types of “nonguaranteed” borrowing are available as long as capital markets are willing to supply the funds. Nevertheless, there are limits to how much spending can outpace revenues. These limits are captured by what is known in economics as the intertemporal budget constraint. In brief, this concept states that the value in today’s dollars of all current and future revenue streams must be sufficient to cover all current and future spending obligations and indebtedness net of assets such as public buildings or land. In a two period example, the difference between taxes net of transfers, T1, and government purchases net of debt service, G1, is the deficit, d1. The next year, d1 becomes the accumulated debt D, which must be paid back with interest, r. Therefore, in period 2, T2 = (1 + r) D + G2. Recall that D came about because of the gap between G1 and T1. Therefore, T2 = (1 + r) (G1 – T1) + G2/(1 + r). In other words, a tax cut in period 1 will raise taxes in period 2, all else being equal. A gap between the net present value of all future outlays on one hand and revenues plus debt on the other is a fiscal imbalance. This imbalance may have negative consequences for social equity if future generations are compelled to pay higher taxes or receive less in public services and current investments do not yield benefits to them. This scenario further assumes that current generations will not offset this imbalance by leaving higher bequests, the so-called Ricardian equivalence result. 17 3. California’s Revenue System State and local governments depend on a variety of revenues to fund public programs. Major sources include taxes, fees or charges for services rendered, and contributions from the federal government. This chapter compares California’s overall level of revenues and its reliance on each of these sources to other states. As shown in this chapter, whether California is considered a “highrevenue” state depends on measurement choices. On a per capita basis, California collected more total and general revenues than the average for all other states in 2002. However, its state and locally generated revenues (i.e., taxes, fees or charges, and miscellaneous sources of income) were close to the average for the rest of the nation as a share of personal income. California relied on taxes, charges, miscellaneous revenues, and federal intergovernmental revenues in roughly the same proportions as the rest of the nation. California’s tax mix differs from that of other states because of its greater reliance on income taxes and lower reliance on property taxes. Total Revenues The U.S. Census Bureau classifies state and local government revenue sources using four major categories: taxes, charges, miscellaneous revenues, and federal intergovernmental revenue. Taxes are the predominant source of government income in all states except Alaska, where oil and mineral sales and investments provide more funds. Most states rely on a combination of income, sales, and property taxes. Income taxes are typically assessed at the state level. Personal income taxes generally follow the federal tax code, using federal adjusted gross income or taxable income as the basis for calculating tax liability and allowing federal itemized deductions. Beyond these provisions, states may offer additional tax credits and deductions as well as exempting certain types of income such as Social Security benefits, capital gains, and 19 dividends or interest payments. As discussed below, states define their own marginal tax rates and brackets, including the level at which income is untaxed. Corporate income taxes operate analogously to the individual income tax, applying to corporate profits less credits, exemptions, and deductions. Certain types of businesses (sole proprietorships, partnerships, and “S” corporations) may pay individual rather than corporate income taxes. Provisions for taxing corporate income vary widely by state. The most important sources of variation are the connection or “nexus” that must exist between a corporation and a state for the corporation to pay taxes in that state, which sources of corporate income are included in the tax base, and how much of this income will be apportioned to the state. The Census Bureau also includes in this category levies on specific industries—such as banks and other financial institutions—as well as franchise or license taxes that apply to corporate net income. Sales and use taxes are levied at both the state and local level. General sales taxes apply to the value of goods or services sold within a jurisdiction, with exemptions for specific items such as food or prescription medicines. Selective sales taxes, also known as excise taxes, apply only to specific commodities, such as motor fuels, alcoholic beverages, and tobacco products. The amount owed may depend on price or volume of a good that is sold. Use taxes are analogous to sales taxes except that they are imposed on items purchased out of state for use in the state of residence. Property taxes are levies on real property, including land and structures, as well as on personal property, including vehicles, equipment, and “intangibles” such as bank accounts. They are assessed and collected mainly at the local level, although states may be responsible for distributing these revenues as is the case in California. Property taxes do not include levies based on characteristics other than a property’s value, such as its area or frontage. These are included in special assessments or one-time levies to fund public capital improvements. In addition to taxes, state and local governments rely on income from user fees or charges. Charges are distinct from taxes because they must be reasonably related to the costs of providing a service or product. 20 They differ from special assessments because they vary with use of a service or good. Examples include fees for higher education and charges for patient care in public hospitals. Miscellaneous revenues include special assessments, interest earnings, lottery revenues, private donations, fines and forfeitures, rents, royalties, and other unallocated sums. This category does not include license taxes, permit fees, or other payments to offset the costs of regulation, which have their own category under taxes. Finally, state and local governments receive intergovernmental revenue from the federal government in the form of grants for specific purposes, such as building or maintaining highways and mass transit systems, or reimbursements for administering federal or joint federalstate programs, such as Medicaid. State governments also provide assistance to local governments. To avoid duplication, the Census Bureau subtracts the value of the latter transactions from combined state and local expenditure totals. In 2002, California state and local governments collected $251 billion in revenues, equivalent to $7,166 per capita (Table 3.1). By comparison, the average state not including California took in $6,155 per capita, or about 14 percent less than California. Other states in our comparison group had lower revenues per capita than California, ranging from 18 percent lower in Illinois to 27 percent lower in Texas. On the other hand, New York collected 25 percent more in total revenues than California. Table 3.1 State and Local Total Revenues per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 7,166 6,155 5,553 5,847 8,928 5,249 General 6,478 5,764 5,195 5,491 8,122 5,024 21 General Revenues Total revenues include payments to public water, gas, electric, and transit utilities as well as contributions to insurance trusts such as public employee retirement plans and investment earnings from these funds. Excluding these sources of funds leaves what the Census Bureau refers to as the general government sector.1 Compared to total revenues, California’s general revenues per capita are closer to those of other states. In 2002, general revenues were 12 percent higher than the average for all other states, or $6,478 per capita, compared to $5,764. Similarly, the gap between California and certain other large states narrows to 15 percent for Illinois and 22 percent for Texas. New York, however, still collected 25 percent more than California in 2002 by this measure. Figure 3.1 shows that the gap between California and the rest of the nation in general revenues per capita has been fairly consistent over time with a few notable exceptions. Prior to the passage of Proposition 13 in 1978, California collected nearly one-third more in general revenues per capita than the average state. The following year, this gap narrowed to 13 percent as Proposition 13 cut property taxes in half by rolling back assessed values to 1976 levels, among other provisions. (Ongoing revenue effects of Proposition 13 are discussed below.) For the most part, California maintained this distance from the average for all other states until the recession of the early 1990s. Because California was more deeply affected by this recession than most other states, its general revenues per capita fell from 15 percent above average in 1990 to 8 percent above average in 1993. As economic growth accelerated during the late 1990s, California’s general revenues per capita rose from a low of 6 percent above the average for all other states in 1997 to 14 percent above average in 2000. Although its collections exceed those of other states, California is similar to the rest of the nation in the composition of these revenues. In 2002, more than half of California’s general revenues (53%) came from _____________ 1The general government sector also excludes publicly owned liquor stores, which do not exist in California and which provide only a small share of revenues in all other states. 22 7,000 6,000 5,000 Revenues ($) 4,000 3,000 2,000 1,000 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.1—General Revenues per Capita, 1977–2002 taxes, one-quarter were derived from charges and miscellaneous revenues, and the remaining 20 percent were revenues from the federal government (Table 3.2). The contributions from each of these sources were nearly identical in the rest of the nation on average. Within our comparison group, Illinois and New York collected a greater share from taxes (60% and 57%) and less from charges (11% and 12%). Florida gained more income from charges and miscellaneous Table 3.2 State and Local General Revenues per Capita, 2002 (dollars) California U.S.–Californi a Florida Illinois New York Texas Total 6,478 5,764 5,195 5,491 8,122 5,024 Taxes 3,442 Charges 1,029 Federal Miscellaneous Intergovernmental Revenue Revenue 628 1,379 3,102 2,688 3,303 4,641 2,715 859 901 627 949 744 570 710 550 642 510 1,235 897 1,010 1,889 1,055 23 revenues (31%), and New York received more federal aid as a share of general revenues (23%). Own-Source Revenues Focusing on state and locally generated revenues, California state and local governments raised $5,099 per capita from their own sources (i.e., excluding federal contributions) in 2002, 13 percent more than the average for all other states. One way to understand the burden of these collections is to adjust for personal income. States with higher incomes also include more resources and economic activity to tax.2 Performing this adjustment narrows the gap between California and other states considerably. In 2002, California’s own-source revenues represented $155 per $1,000 of personal income, 5 percent higher than the $148 average for the rest of the nation (Table 3.3). Within our comparison group, California’s revenues were 7 percent higher than in Florida, 12 percent higher than in Texas, 13 percent higher than in Illinois, and 12 percent lower than in New York. Nationally, California ranked 19th in its revenue collections per $1,000, immediately below Montana and the District of Columbia and above Ohio and Alabama. Table 3.3 State and Local Own-Source Revenues, 2002 Per Capita ($) U.S. Rank Per $1,000 of Personal Income ($) U.S. Rank California 5,099 10 155 19 U.S.–California 4,530 — 148 — Florida 4,299 30 145 37 Illinois 4,480 26 138 42 New York 6,232 3 176 3 Texas 3,969 38 138 41 _____________ 2The widely used personal income series from the Bureau of Economic Analysis (BEA) provides an imperfect adjustment because it excludes certain types of taxable income and includes other nontaxable sources. It is not clear how adjusting for these differences would affect California’s relative position. 24 Revenues ($) Adjusting for income also changes California’s relative position over time (Figure 3.2). In 1978, California’s general own-source revenues per $1,000 were 15 percent higher than the average for all other states; by 1979, they were 5 percent below this average. Revenues per $1,000 remained below the U.S. average excluding California throughout most of the 1980s. California kept pace with the rest of the nation during the late 1980s and 1990s, surpassing this average in 2000 and remaining at this level in 2002. 180 160 140 120 100 80 60 40 California U.S.–California 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.2—General Own-Source Revenues per $1,000 of Personal Income, 1977–2002 Taxes Tax systems vary from state to state. Four states (Delaware, Montana, New Hampshire, and Oregon) do not levy a sales tax and one state (Alaska) does not have a state sales tax but allows cities and counties to have one. Seven states (including two in our comparison group, Florida and Texas) do not have a tax on personal income. Another two states have a personal income tax but apply it only to unearned income such as interest and dividends. Although the property tax exists in all U.S. states, several states have limits on tax rates, assessment increases, or total revenues from this source. 25 Most states rely on a “three-legged stool” of income, sales, and property taxes. Table 3.4 shows these collections per $1,000 of personal income in 2002. In percentage terms, the average U.S. state excluding California derived 25 percent of total tax revenue from income taxes, 36 percent from sales and use taxes, and 32 percent from the property tax (Table 3.4). California derived a similar share of tax revenues from sales taxes overall, although it collected less from selective sales taxes or taxes on particular goods (9% of total taxes). Compared to the average for all other states, income and property tax shares were reversed in California, with property taxes contributing 25 percent of total taxes and income taxes providing 32 percent. Among the states in our comparison group, Florida and Texas stand out because they do not have a personal income tax. As a result, these states depend more on collections from sales taxes (51% and 49% of total taxes, respectively) and property taxes (35% and 42% of total taxes, respectively). California collects a greater share of tax revenues from income taxes than any other state in our comparison group except New York, which takes in 40 percent of tax revenue from this source. At the same time, California receives less in property tax revenues than any of the other large states considered here. This has not always been the case. As noted above, property tax collections in California fell by half immediately after the passage of Proposition 13 in 1978. Since then, personal income and sales taxes have provided a greater share of state and local tax revenues (Figure 3.3). In particular, the share of state and local tax revenues from the personal Table 3.4 State and Local Taxes per $1,000 of Personal Income, 2002 (dollars) California U.S.–California Florida Illinois New York Texas License Individual Corporate General Selective and Total Income Income Sales Sales Property Other 105 29 5 27 9 26 9 102 22 3 25 12 32 8 91 — 2 31 16 32 10 102 18 3 18 16 39 7 131 45 8 25 9 40 6 95 — — 29 17 39 9 26 Revenues per $1,000 of personal income ($) 60 Property General sales 50 Selective sales Individual income Corporate income 40 License and other 30 20 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 3.3—California State and Local Major Taxes per $1,000 of Personal Income, 1977–2002 income tax grew rapidly during the late 1990s, from 26 percent in 1997 to 33 percent in 2000, only to fall again to 27 percent in 2002.3 The rising importance of the personal income tax stems in part from its progressivity. As incomes grew in the state during the late 1990s, the state raised more funds from this source. California has one of the highest top marginal tax rates in the nation at 9.3 percent, compared to 7.7 percent in New York as of 2004 (6.85% in 2002) and 3 percent in Illinois.4 California’s personal income tax also has more rate brackets (six) than other large states in our comparison group, with the exception _____________ 3Because they are of state and local taxes, these shares differ from widely reported figures for the personal income tax as a share of state General Fund revenues: roughly 50 percent, compared to less than 20 percent in 1964 (Legislative Analyst’s Office, 2004). 4In addition, high-income taxpayers in California may be subject to an Alternative Minimum Tax analogous to the AMT at the federal level. Unlike its federal counterpart, however, California’s AMT exemptions are indexed to inflation, preventing “bracket creep.” Also, in November 2004, voters approved Proposition 63, which imposed an additional 1 percent tax on the portion of a taxpayer’s income in excess of $1 million to fund mental health programs. 27 of New York. For example, Illinois has only one tax bracket (i.e., a flat tax) on income. Beyond its high marginal tax rates and graduated structure, California’s personal income tax is also more progressive than in the average state because of credits, exemptions, and deductions that preclude many households from paying state income taxes. For example, in 2000, two-parent families of four earning less than $37,000 were exempt from state income taxes in California. In New York and Illinois, comparable income thresholds were $23,800 and $14,000, respectively. As a result of this progressivity, in 2001 the top 1 percent of income taxpayers in California contributed 34 percent of all income tax receipts, whereas the bottom 40 percent provided 1 percent of these revenues. By comparison, in New York the top 1 percent of tax filers paid 30 percent of all individual income taxes and the bottom 40 percent paid 7 percent. The difference between California and Illinois was even more dramatic because of that state’s flat rate structure. There, the top 1 percent paid only 17 percent of all personal income taxes. In contrast to the personal income tax, sales and use taxes are generally viewed as regressive because consumption represents a larger share of income for less affluent households. The collections and incidence of sales and use taxes depend on the tax rate and tax base. Like all states in our comparison group, California has a relatively high combined state and local sales tax rate compared to the rest of the nation. In 2004, the maximum combined state and local sales tax rate was 8.75 percent in California, lower than that of Illinois but higher than that of Florida or Texas (Table 3.5).5 _____________ 5California’s combined rate includes one state tax at 5 percent with revenues dedicated to the state General Fund. There are also two state taxes at one-half of one percent with proceeds going to special funds for local government. In addition, there is a statewide local tax of 1.25 percent (the so-called Bradley-Burns rate). Finally, localities may levy their own sales taxes with local voter approval. Proposition 57, approved by California voters in March 2004, diverts a portion of the uniform Bradley-Burns rate to a special fund to repay the state’s deficit financing bonds. While this repayment continues, the state will replace local sales tax revenues with property taxes from schools and provide schools with a greater infusion of resources from the state General Fund. This is the socalled “triple flip.” 28 Table 3.5 State and Local Retail Sales Tax Rates, 2004 (percent) California U.S.–California Florida Illinois New York Texas Maximum Combined 8.75 7.37 7.5 9.25 8.75 8.25 State 6 5.31 6 6.25 4.25 6.25 Maximum Local 2.75 2.83 1.5 3 4.5 2 In California, as in all of our comparison states, general sales taxes exempt certain purchases. For example, California and all comparison states except Illinois exempt food for home consumption. (Illinois taxes these purchases at a lower rate.) California also exempts utility payments, prescription drugs, custom computer programs, snacks, animal feed, bottled water, and free newspapers or periodicals (California State Board of Equalization, 2002). Some states in our comparison group also exempt prescribed and nonprescribed medications or tax these purchases at a lower rate (Illinois). More significantly, California is one of several states that do not tax most services, even though these purchases constitute the bulk of consumer expenditures. According to a 2004 survey by the Federation of Tax Administrators (FTA), California taxes 23 out of 168 potential services, compared to 17 in Illinois, 56 in New York, 62 in Florida, and 81 in Texas. Similarly, the Center for Budget and Policy Priorities notes that California currently taxes only two out of 40 “readily taxable services” other than health care, housing, education, legal, banking, public transit, insurance, and funeral services. By comparison, Illinois taxes three out of 40 such services, whereas Florida taxes 22, New York taxes 23, and Texas taxes 24 (Mazerov, 2002).6 _____________ 6A related issue is enforcement of the “use” tax on remote sales by catalogue or through the Internet. The Government Accountability Office (GAO) estimates that California loses more revenues to untaxed remote sales than any of our comparison states (Government Accountability Office, 2000, pp. 59–61). California has joined the national Streamlined Sales Tax Project (SSTP), in part to raise congressional attention to 29 The more items that are exempt from taxation, the narrower the base of the tax. At the extreme are selective sales taxes, including “sin” taxes designed to reduce consumption of particular goods such as alcoholic beverages, tobacco products, or motor fuels and perhaps also to fund programs counteracting harmful consequences of these commodities (e.g., smoking cessation programs). Although California derives less income from selective sales taxes as a group than other states (Table 3.4), “sin tax” rates are often close to the national median (Table 3.6).7 Like some of the other taxes we have considered, California’s corporate income tax rate is relatively high at 8.84 percent, compared to 5.5 percent in Florida, 7.3 percent in Illinois, and 7.5 percent in New York. In California, banks and other financial institutions also pay a surcharge of 2 percent, whereas subchapter S firms pay a lower rate (1.5%), and limited liability companies pay a flat fee.8 How much corporate income is subject to taxation depends on rules for determining the tax base and apportioning income to the state. California determines the amount of income subject to taxation using “combined reporting” for all holdings of a unitary business group rather than using separate reporting for each incorporated entity. Within Table 3.6 Selective Sales Tax Rates, 2005 California U.S. median Florida Illinois New York Texas $ per Pack of Cigarettes 0.87 0.70 0.34 0.98 1.50 0.41 Liquor 3.30 3.75 6.50 4.50 6.44 2.40 $ per Gallon of Wine Beer 0.20 0.20 0.64 0.19 2.25 0.48 0.73 0.19 0.19 0.11 0.20 0.19 Gas 0.18 0.20 0.40 0.19 0.80 0.20 ______________________________________________________________ this issue. It has not, however, conformed its sales and use tax laws to those of the Streamlined Sales and Use Tax Agreement (Jones, 2005). 7These rates are exclusive of federal taxes and general sales taxes on these goods. California’s Proposition 42 earmarks the latter revenues from gasoline sales to transportation purposes. 8Like individuals, corporations are also subject to an AMT at 6.64 percent of income. 30 our comparison group, Illinois also requires combined reporting, but New York and Florida do not. California apportions income from multistate and multinational corporations to the state based on a formula that weights property, payroll, and sales in the state. Like Florida and New York, California assigns double weight to sales to encourage businesses to invest capital and labor in the state. In addition, firms may elect to report all business activity or that which stops at the “water’s edge” (i.e., within the United States only). California’s corporate income tax also includes several exclusions, exemptions, deductions, and credits. Notably, firms can claim research and development credits and can carry forward net operating losses (NOLs) as deductions against future earnings.9 Understanding the net effects of these credits on revenue collections and economic competitiveness is complicated. However, it is interesting to note from Figure 3.3 that corporate tax earnings declined as a share of personal income in the 1990s, consistent with the promulgation of credits over this period, such as the Manufacturer’s Investment Credit (MIC) on intermediate purchases, as well as a trend toward corporate reorganizations. As noted throughout this chapter, the passage of Proposition 13 in 1978 dramatically altered the public finance landscape in California. This measure rolled back assessed property values to 1976 levels, capped property tax rates at 1 percent, and limited the growth in assessed values to 2 percent per year except for properties that are sold in that year. It also instituted a two-thirds popular vote requirement for all new state taxes and local special taxes. As a result, California has collected less in property tax revenues than the average for all other states since 1977. Local governments in particular have become more dependent on intergovernmental revenue, contrary to the trend in the rest of the nation and in all of our comparison states (Table 3.7). Although charges and miscellaneous _____________ 9After two years of suspension, the carry forward provision was reinstated in 2004, and in January 2005, the portion of losses eligible for deduction increased to 100 percent. 31 Table 3.7 Local Revenues per $1,000 of Personal Income, by Source, 2002 (dollars) California U.S.-California Florida Illinois New York Texas California U.S.-California Florida Illinois New York Texas Taxes 59 44 33 45 81 37 37 42 40 47 67 49 Charges and Miscellaneous Revenues 1977 17 17 25 12 21 17 2002 33 25 37 20 27 24 Intergovernmental Revenues 55 46 45 35 85 31 69 41 35 37 60 31 sources have become a more important source of revenue in all states since 1977, this growth is more pronounced in California. As in other states, California’s property tax includes many provisions designed to reduce the burdens on specific taxpayers. Like many states, California offers a homestead exemption that reduces the assessed value subject to taxation for an owner-occupied home. It also provides limited relief to elderly homeowners through “circuit-breakers,” or credits against property taxes when liabilities exceed a certain amount, and property tax deferrals. Whereas California’s property tax may fare as well as any state’s in terms of its “vertical equity,” it has been criticized on “horizontal equity” grounds for treating similarly situated homeowners differently based on the purchase date of a property. On the other hand, these effects have been diminishing over time with the turnover of properties and periods of lower price inflation (e.g., Sheffrin and Sexton, 1998). 32 Conclusions California collects more revenues than the average state in the rest of the nation. In 2002, it collected 16 percent more in total revenues, 12 percent more in general revenues, and 13 percent more in own-source (i.e., nonfederal) revenues than the average for all other states. Among other large states in our comparison group, Florida and Texas collected less per capita than California whereas New York had higher revenues per capita than California. California is also a high-income state. When we consider revenues as a proportion of personal income, California looks more like the rest of the nation. In 2002, own-source revenues per $1,000 in California were 5 percent higher than the average for all other states. Differences with respect to other large states are also narrower. California’s tax mix is relatively balanced, relying on roughly equal proportions of the income, sales, and property tax, although tilting more than other states toward income taxes and away from property taxes. California’s heightened reliance on the income tax stems from the progressivity of this tax, from rising incomes in the state in recent years, and from an increasing proportion of income from nonwage sources, such as stock options and capital gains. For these reasons, several analysts have pointed out that California’s tax system has also been more volatile than that of other states (Legislative Analyst’s Office, 2005). 33 4. Elementary and Secondary Education Elementary and secondary education, more commonly known as K–12 education, is the largest category of state and local government expenditures. In 2002, this category constituted 19 percent of all expenditures in California and 21 percent in the rest of the country. This chapter analyzes expenditure differences between California and other states in this important area and also briefly summarizes expenditures for public libraries and a variety of other activities related to K–12 education. We focus first on expenditures per capita because this is a better measure of the cost of public schools to taxpayers than is the commonly used measure of expenditures per pupil. We then consider expenditures per pupil and its components, including staffing levels and average employee compensation. As shown below, expenditures per capita and staff per pupil together often paint a more revealing picture than do expenditures per pupil alone. Per capita expenditures on K–12 education in California are similar to those in other states. However, public schools in California have far fewer staff resources per student than do schools in other states. This outcome stems primarily from two factors: California has more students per capita than other states, and school districts in California compensate their employees more highly than districts in other states. Although the link between resources and outcomes in K–12 education is not always clear, it is interesting to note that student achievement in California also lags that of other states. Total Expenditures The activities considered in this chapter are relatively easy to describe. Elementary and secondary education as defined by the U.S. Census Bureau includes the instructional activities of schools overseen by 35 public school districts and county offices of education. It also includes the administrative support for these activities and other support and auxiliary services such as school lunch programs, counseling services, and pupil transportation. As a practical matter, K–12 expenditures in California are the expenditures of its school districts, their joint powers agencies, and the county offices of education. Any expenditure of those three types of agencies is counted as an expenditure on elementary and secondary education. Public libraries are the libraries run by cities, counties, and states. This category excludes the libraries of public schools, colleges, and universities, which are included in the larger categories of elementary and secondary education or higher education. The Census Bureau defines a third area, “other education,” which includes services related to education that are not provided directly by libraries, public schools, and public colleges and universities. In California, this category includes the Office of the Secretary for Education and the operations of the State Department of Education. It also includes the vocational rehabilitation services provided by the Department of Rehabilitation Services. In per capita terms, library and other education expenditures are relatively small (Table 4.1). Although expenditures per capita in those two areas differ across states, the differences are not large enough to be an important factor in comparing educational expenditures in one state to Table 4.1 Education Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 1,592 1,490 1,175 1,531 2,059 1,518 Elementary and Secondary Education Libraries 1,521 1,415 1,116 1,426 1,999 1,476 29 29 23 45 39 16 Other Education 42 46 36 60 21 26 36 those in another. As a consequence, the remainder of this chapter focuses on the Census Bureau category of elementary and secondary education. In contrast, there are large differences across states in expenditures on elementary and secondary education. In 2002, California was roughly representative of other states. However, New York spent $478 more per capita than California (31% more), and Florida spent $405 less (27% less). In Texas and Illinois, public school spending per capita was approximately equal to that of California. California’s position relative to other states has changed significantly over time (Figure 4.1). In 1977, California was spending 13 percent more per capita than the rest of the nation; by the end of the decade, this margin had essentially disappeared. The 1970s was also the period during which California transformed its school finance system from one in which school districts levied their own property tax rates to the current system in which the property tax rate is capped statewide and the legislature decides how much property tax revenue each district receives (Sonstelie, Brunner, and Ardon, 2000). 1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 California 800 U.S.–California 700 600 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.1—Total Expenditures on Elementary and Secondary Education per Capita, 1977–2002 37 Expenditures (2002 $) After falling relative to other states in the 1970s, K–12 education expenditures per capita were roughly the same in California as in other states during the 1980s. This parity ended during the recession of the early 1990s, which was felt strongly in California. California’s real spending per capita fell by 10 percent between 1991 and 1994, while real spending per capita rose by 3 percent in the rest of the nation. From this low point, however, public school spending per capita has risen much faster in California than in other states. Between 1994 and 2002, real spending per capita rose by 45 percent in California, compared to 21 percent in the rest of the nation. Total expenditures on K–12 education can be partitioned into current and capital expenditures. Current expenditures include the salary and benefits of teachers, instructional aides, counselors, principals, and other staff. They also include the purchase of goods and services such as textbooks and utilities. Capital expenditures include outlays for constructing new schools and renovating existing schools. They also include purchases of equipment such as desks and computers. For California and its comparison states in 2002, capital outlays constituted 12 to 16 percent of total expenditures (Table 4.2). Because capital expenditures have long-lasting effects, however, outlays in one year are not particularly informative. Later in the chapter, we examine trends in these expenditures over many years. Table 4.2 Total Expenditures on Elementary and Secondary Education per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita 1,521 1,415 1,116 1,426 1,999 1,476 Current Expenditures per Capita 1,320 1,232 936 1,197 1,748 1,229 Capital Expenditures per Capita 201 182 180 229 251 247 38 Current Expenditures California’s current expenditures on K–12 education followed the same general trend as total K–12 expenditures: Real spending per capita declined during the 1970s and rose in each year after with the exception of the early 1990s. Compared to other states, California began the 1970s spending significantly higher amounts per capita, fell to parity in the 1980s, declined below spending in the rest of the nation in the 1990s, and returned to parity at the end of the decade. Underlying the trends in current expenditures per capita are fundamental trends in school enrollment. During the 1970s, for example, enrollments were falling throughout the country, so the decline in expenditures per capita did not necessarily translate into fewer resources for the typical school. Similarly, if one state has fewer students per capita than another, lower expenditures per capita in that state do not necessarily imply that its schools have fewer resources. In concept, expenditures per capita measure the cost of public schools to the average taxpayer, and expenditures per student measure the resources provided to the typical school. The two measures are tied together by Eq. (2.1) in Chapter 2. In the particular case of K–12 education, public school students play the role of caseload, yielding the following version of that equation: Expenditures = Public School Students  Expenditures (4.1) Population Population Public School Student Conceptually, the ratio of public school students to total population is the cost to the average taxpayer of increasing public school spending per student by one dollar. Raising expenditures per student is less expensive when there are fewer students. For example, doubling the number of public school students would double the cost in taxes to the average taxpayer of increasing expenditures per pupil by one dollar. Compared to other states in 2002, California had high student enrollments, making K–12 education per student relatively expensive (Table 4.3). Using Eq. (2.4) in Chapter 2, the effect of this higher cost is reflected in the following way: 39 Table 4.3 Components of Current Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 1,320 1,232 936 1,197 1,748 1,229 Students Expenditures per Capita per Student ($) 0.179 7,391 0.164 7,523 0.150 6,245 0.165 7,271 0.150 11,653 0.192 6,412  Expenditures   Public School Students +  Expenditures Population Population Public School Student (4.2) In this equation,  represents the percentage difference in a variable between two states or between California and all other states. Taking the first term in Eq. (4.2), in 2002 expenditures per capita in California were 7 percent higher than in other states. However, public school students per capita, the second term, were 9 percent higher in California than in other states. As a result, expenditures per student, the third term, were 2 percent less in California than in other states. California’s higher spending per capita was not quite enough to offset its higher public school enrollments per capita. Comparisons between California and Texas are also informative. In per capita terms, Texas spent 7 percent less than California. It also had 7 percent more students per capita, resulting in expenditures per student that were 14 percent less than in California. In this comparison with California, New York was the mirror image of Texas. It spent more per capita (32%) and had fewer students per capita (16%), resulting in expenditures per student that were 58 percent higher than in California. Each year, the National Education Association (NEA) ranks states according to their current expenditures per pupil on K–12 education (National Education Association, 2003). This well-known ranking is based on a slightly different definition of current expenditures than that used by the Census Bureau. The Census definition includes all current expenditures of school districts and county offices of education. The 40 NEA definition excludes some of those expenditures, particularly expenditures for summer schools, adult education, and community services. For 2002, the NEA measure of current expenditures per pupil in California was 4.5 percent less than the Census measure. There were also significant differences between the NEA and the Census measures for Illinois and Texas. Despite these differences, however, the NEA measure paints a similar picture to that of Table 4.3: Illinois spent about the same per pupil as California, Texas and Florida spent less, and New York spent much more. Equation (2.2) in Chapter 2 describes how the caseload for a public service can be decomposed into a take-up rate and an eligibility rate. In the case of K–12 education, this equation is: Public School Students = Public School Students Population Children Eligible for Public School  Children Eligible for Public School Population (4.3) For both the take-up rate and the eligibility rate, the difficulty lies in determining the number of children eligible for public school. Children are generally eligible for 13 years of schooling, kindergarten through high school, although they can also be retained for one or more grades, implying more years in school. In addition, at certain ages, compulsory attendance laws require that children attend either a public or a private school. In California, children must start school at age six and attend school until they either receive a high school diploma or reach the age of 18. Among other states, the starting age differs from age five to age seven, and the dropout age differs from age 16 to age 18. Accordingly, we define children eligible for public schools to be all children in public and private schools plus those who have dropped out of school before graduating as well as immigrant children who were never enrolled. The Census Bureau provides a working definition of the latter two groups in the number of children between ages 16 and 19 who do not have a high school diploma and are not enrolled in school. The cutoff at 41 age 19 is somewhat arbitrary; a twenty-year-old who has not finished high school is still eligible to enroll in school. However, a person without a high school diploma is less likely to enroll in school as his or her age increases, making some cutoff age a reasonable restriction on the definition of those eligible to enroll in school. Table 4.4 uses the Census restriction to define children eligible for school. The number of public school students per capita is relatively high in California, partly because of the state’s relatively high take-up rate. The take-up rate is the percentage of children eligible for a public school who actually attend such a school. This rate is the product of two factors: the percentage of eligible children who attend any school (public or private) and the percentage of those students who attend a public school. The first factor is presented in the third column of Table 4.4. Across all states, 96 to 98 percent of children eligible for public school are enrolled in a public or private school. This factor is not a significant source of difference among states in the take-up rate. However, the percentage of these students who attend a public school (the second column) varies from 86 percent to 93 percent. California has a relatively high public school enrollment rate of 90 percent, which is the primary source of its relatively high take-up rate. A more important source of California’s high number of public school students per capita is its high eligibility rate, the last column in Table 4.4. In California, 21.3 of every 100 residents were eligible for public school, as opposed to 19.6 in the rest of the country, a difference Table 4.4 Components of Public School Enrollment, 2000 California U.S.–California Florida Illinois New York Texas Public School Students per Capita 0.188 0.170 0.154 0.167 0.162 0.194 Students in Public School (%) 90 89 88 87 86 93 Eligible Children in School (%) 97 97 97 97 98 96 Children Eligible for Public School per Capita 0.213 0.196 0.180 0.198 0.194 0.217 42 of 9 percent. Among our comparison states, only Texas had a higher eligibility rate, with 21.7 of every 100 residents eligible. In fact, in both Texas and California, a relatively high eligibility rate combined with a relatively high take-up rate produce a high rate of public school students per capita. The relative size of a state’s public school enrollment may change significantly over time, with significant consequences for the cost of K–12 education (Figure 4.2). From 1977 through 1990, public school students per capita fell steadily outside California. Since 1990, that ratio has been relatively constant. In California, the ratio also fell in the 1970s and 1980s, but the decline was steeper and shorter. By 1985, the number of public school students per capita had reached its lowest level in California and had begun to rise. In the 1990s, while other states were benefiting from a relatively stable ratio of students to population, California was coping with an increase in that ratio and thus in the cost per capita of public elementary and secondary education. During the 1990s, three of our four comparison states also saw increases in students per capita. However, none experienced an increase as rapid as California’s. During the 1990s, the number of students per capita increased by 7 percent in California. In comparison, the ratio rose 0.21 0.20 California U.S.–California 0.19 0.18 0.17 0.16 0.15 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.2—Number of Public School Students per Capita, 1977–2002 43 Number of students by 5 percent in New York, 4 percent in Florida, and 2 percent in Illinois. Texas actually experienced a 4 percent decline in its ratio during this period. These demographic trends are evident when the focus shifts from expenditures per capita to expenditures per pupil (Figure 4.3). Throughout the country, real expenditures per capita fell from 1977 to 1985. However, expenditures per pupil actually rose during this period because of falling enrollments. In other words, the resources of public schools in the rest of the nation were improving during this period, even though relatively less tax revenue (on a per capita basis) was being allocated to them. On the other hand, California’s expenditures per pupil were flat from 1977 to 1985, and the steady increase in California’s studentpopulation ratio from 1985 onward magnified the decline in spending per capita during the early 1990s, resulting in a decrease in real expenditures per pupil of 13 percent from 1991 to 1994. This decline reversed in the late 1990s, however. Although pupils per capita continued to rise in California throughout the 1990s, real expenditures rose faster than enrollment from 1995 through 2002. In 8,000 7,500 7,000 Expenditures (2002 $) 6,500 6,000 5,500 5,000 4,500 4,000 3,500 California U.S.–California 3,000 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.3—Current Expenditures on Elementary and Secondary Education per Student, 1977–2002 44 fact, California’s real expenditures per pupil rose by 30 percent from 1995 to 2002. In contrast, in the rest of the country, real expenditures per pupil rose by only 15 percent during this period. Thus, by 2002, California had largely erased the gap in expenditures per pupil that had developed between itself and other states during the recession of the early 1990s. From 1977 to the present, expenditures per pupil in our comparison states rose significantly relative to California. In 1977, expenditures per pupil in Texas were 66 percent of expenditures per pupil in California. The ratios for Florida, Illinois, and New York were 82 percent, 90 percent, and 122 percent, respectively. By 2002, spending per pupil in Texas was 87 percent of that in California. For Florida, Illinois, and New York, the ratios were 85 percent, 98 percent, and 158 percent, respectively. Each of the four states had gained in expenditures per pupil relative to California. The most dramatic gains were in the mid-1990s. In 1997, for example, New York, Florida, and Illinois were spending more per student than California, and Texas was at 95 percent of California’s level. In addition to pupils per capita, another important element in the cost of education is the salaries and benefits paid to school district employees. For the nation as a whole in 2002, employee salaries and benefits constituted 81 percent of current expenditures on K–12 education. Employee benefits represented 20 percent of total compensation. The salaries and benefits of teachers and instructional aides accounted for only 69 percent of all employee compensation. Table 4.5 partitions current expenditures per student into staff compensation per student and other expenditures per student, including purchases of good and services such as textbooks, utilities, office supplies, and so on. In California, staff compensation per student was slightly less than in the rest of the nation and other expenditures were slightly higher (Table 4.5). Following Eq. (2.3) in Chapter 2, staff compensation per student can be decomposed in the following way: Compensation = Staff  Compensation Student Student Staff (4.4) 45 Table 4.5 Components of Current Expenditures per Student, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Current Expenditures per Student 7,391 7,523 6,245 7,271 11,653 6,412 Staff Compensation per Student 6,041 6,300 4,896 6,339 9,350 5,355 Other Expenditures per Student 1,350 1,223 1,349 932 2,303 1,058 Staff per student is a measure of the personnel resources schools have, and compensation per staff is a measure of the cost of each unit of resources. These two components are listed in Table 4.6. Despite nearly average staff compensation per pupil, employees in California’s K–12 were paid more on average, and thus California schools had relatively low staff-student ratios. California schools spent about 4 percent less per student on staff compensation than schools in the rest of the country. Where California differed more dramatically was in average staff compensation, which was 34 percent higher than in other states. With staff compensation per student approximately the same as in other states, these higher employee costs necessarily implied a lower ratio of staff to students. As the third column of Table 4.6 shows, the Table 4.6 Components of Compensation per Student, 2002 California U.S.–California Florida Illinois New York Texas Compensation per Student ($) 6,041 6,300 4,896 6,339 9,350 5,355 Compensation per Staff ($) 65,710 48,982 43,307 51,387 63,456 38,269 Staff per Student Ratio 0.092 0.129 0.113 0.123 0.147 0.140 46 ratio of staff to students in California was only 71 percent of the ratio in other states. Because of average staff compensation per pupil and above average employee compensation, California schools were thinly staffed relative to schools in other states. New York also had high compensation costs for public school employees, with an average nearly equal to the average in California. However, its very high expenditures per student (58% higher than California) meant that it could achieve a higher staff-student ratio than California (60% higher), despite its high compensation levels. Texas also had a much higher staff-student ratio than California (52% higher), but that ratio was not due to higher spending per pupil. Texas actually spent less per pupil than California. However, its average compensation was 58 percent of the average in California. In these comparisons, Illinois and Florida were less dramatic versions of New York and Texas, respectively. Both states had higher staff-student ratios than California’s ratio. California’s relatively low staffing ratio holds for all categories of employees (Table 4.7). On a per student basis, other states had 33 percent more teachers than California. In terms of support staff such as instructional aides, counselors, and librarians, other states had 43 percent more staff per student than California. Similar comparisons hold for administrators, such as superintendents, principals, and administrative assistants, and for other staff, such as janitors, bus drivers, and security guards. In the administrative category, other states had 19 percent more Table 4.7 Public School Staff per 1,000 Students, 2002 California U.S.–California Florida Illinois New York Texas Teachers 49 65 54 63 73 68 Support Staff 17 24 21 24 22 20 Administrative Staff 11 14 15 13 15 17 Other Staff 15 26 24 24 37 34 47 staff per student than California; in the “other” category, this margin was 73 percent. Employee compensation is the sum of salaries and benefits, and California’s relatively high compensation for public school employees extends to both components (Table 4.8). The ratio of benefits to total compensation is 20 percent in California, slightly lower than the average for all other states. Three of California’s four comparison states have similar benefit ratios. The only exception is Texas, in which the benefit ratio is only 13 percent. School districts must compete for most of their employees in the market for college-educated workers. In California, half of public school employees are teachers, who must have a college degree. Thirty percent are administrators, counselors, and other support staff, who generally have college degrees. As a consequence, the best comparison group for public school employees is Zone 4 workers (as described in Chapter 2), who are also employed in occupations requiring a college degree. In 2002, the average salary for these workers was 13 percent higher in California than the average in all other states. We used this difference to adjust the salaries of public school employees for differences in labor market conditions among states. The result is shown in the second column of Table 4.9. After adjustments for labor market conditions, the salaries of public school employees were about 17 percent higher in California than in other states. Before that adjustment, the differential was 35 percent. Thus, about half of the difference between the salaries of public Table 4.8 Components of Average Compensation, 2002 California U.S.–California Florida Illinois New York Texas Average Compensation ($) 65,710 48,982 43,307 51,387 63,456 38,269 Average Salary ($) 52,449 38,777 34,234 40,797 50,078 33,133 Average Benefit ($) 13,261 10,205 9,072 10,590 13,377 5,137 Benefit Ratio (%) 20 21 21 21 21 13 48 Table 4.9 Average Public School Salary Adjusted for Location, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Salary of Public School Employee 52,449 38,777 34,234 40,797 50,078 33,133 Average Salary Adjusted for Location 46,238 39,485 36,426 40,989 40,961 35,193 school employees in California and other states can be explained by differences in general labor market conditions in California and other states. The general compensation pattern portrayed in Table 4.9 holds in particular for public school teachers. According to the annual survey of the National Education Association, the average salary of California teachers in 2002 was $54,348, the highest average of any state in the country (National Center for Education Statistics, 2003). Teachers in Florida had an average salary that was 70 percent of the California average, a little higher than the 65 percent that we calculate for all school employees as reported in Table 4.9. For Texas, the comparisons are similar—71 percent for teachers versus 63 percent for all school employees. Illinois was somewhat different, however. Its teachers had an average salary that was 91 percent of the California average, considerably higher than the 79 percent for all school employees. For New York, the percentages were virtually the same for teachers and for all employees. Our analysis thus far has emphasized three points. First, expenditures per capita on K–12 education are about the same in California as in other states. Second, California has more public school students per capita than other states, resulting in lower expenditures per student. Third, public school employees in California are compensated more highly than public school employees in other states, resulting in fewer staff per student. 49 Another way to understand the difference between California and other states is to ask how much spending would have to be increased in California for its schools to have the same resources as schools in other states. As a baseline, suppose California looked exactly like other states. In particular, suppose it had the same staff compensation, the same ratio of staff to students, the same level of other expenditures per student, and the same fraction of public school students in the population. Under those suppositions, California would have the same expenditures per capita as other states. Now, suppose that staff compensation were increased to California’s level and everything else remained the same. Expenditures per capita would increase by 29 percent. In addition, if the fraction of students in the population were also increased to California’s level, expenditures per capita would increase by 40 percent. In that sense, those two factors together, compensation per staff and students per capita, increased California’s cost of elementary and secondary education by 40 percent. Seventy-three percent of that increase was due to higher compensation. An alternative calculation of this cost starts with California’s compensation levels and ratios of staff per pupil and pupils per capita. From that baseline, if compensation per staff in California were reduced to the level of other states, California expenditures per capita would fall by 21 percent. If pupils per capita were also reduced to the level of other states, expenditures per capita would fall by another 6 percent for a total decline of 27 percent. Accounted for in this way, California’s high staff compensation and ratio of pupils per capita increased the cost of elementary and secondary education by 27 percent. Capital Expenditures As noted above, K–12 education capital expenditures per capita were higher in California than in the rest of the nation in 2002. However, as Figure 4.4 shows, this situation was without precedent in the previous 25 years. Throughout the 1980s and 1990s, capital expenditures per capita were substantially lower in California than in the rest of the country. 50 Expenditures (2002 $) 250 California 200 U.S.–California 150 100 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 4.4—Capital Expenditures on Elementary and Secondary Education per Student, 1977–2002 An important part of capital expenditures is the construction of new school buildings. Because school buildings are long-lasting, the need for new construction is better measured by the increase in student enrollment than by the level of enrollment. The need for additional school buildings is less in a state with high but stable enrollment than in a state where enrollment is lower initially but growing rapidly. In fact, enrollment growth rates varied considerably across states. From 1992 to 2002, enrollment grew by 20 percent in Texas, 22 percent in California, and 29 percent in Florida. In contrast, enrollment grew by only 12 percent in Illinois and 8 percent in New York. For all states except California, enrollment grew by 12 percent from 1992 to 2002. Table 4.10 compares the sum of annual capital outlays from 1992 to 2002 with the increase in enrollment during that period. Capital expenditures per additional student were substantially lower in California than in the rest of the country. Part of this difference may be due to the renovation of older school buildings. Even if a state is barely growing in enrollment, it must undertake some capital expenditures for renovation, and so its capital outlays per additional student will be very high. This possibility is consistent with the very high capital outlays per additional 51 Table 4.10 Capital Outlay per Additional Student, 1992–2002 California U.S.–California Florida Illinois New York Texas Increase in Enrollment, 1992 to 2002 1,140,581 4,530,953 568,347 223,225 228,139 699,076 Real Capital Outlay, 1992 to 2002 (1,000s of 2002 $) 45,348,972 366,246,776 24,971,202 20,497,211 32,138,075 40,025,032 Real Capital Outlay per Additional Student (2002 $) 39,760 80,832 43,937 91,823 140,871 57,254 student in Illinois and New York. On the other hand, enrollment in Texas grew about as rapidly as in California and Florida’s enrollment grew more rapidly in proportionate terms, yet capital outlays per additional student in both Texas and Florida were also significantly greater than in California. California has not responded as vigorously to the need for new school buildings as have Texas and Florida. As Figure 4.4 demonstrates, however, California did substantially increase its capital expenditures during the latter part of this ten-year period, an increase partly reflecting two voter-approved propositions. Proposition 1A, passed in 1998, authorized the state to issue over $9 billion in bonds for the construction and modernization of buildings in public schools, colleges, and universities. Proposition 39, passed in 2000, made it easier for school districts to issue their own bonds for construction and modernization. Before Proposition 39, school districts had to secure a two-thirds approval of their voters before issuing such bonds. The proposition lowered that requirement to 55 percent. From the passage of Proposition 39 through 2004, voters approved local school bonds totaling $20 billion. Nearly half of that total would not have been approved under the old requirement (Rueben and de Alth, 2005). This trend of increasing capital expenditures has continued beyond 2002. In November 2002, the voters of California passed Proposition 47, which authorized the state to sell bonds worth $13 billion for building and remodeling public schools and also facilities of public colleges and universities. In March 2004, the voters passed Proposition 52 55, which authorized $21 billion for the same purpose. The capital expenditures resulting from these bonds do not show up in Figure 4.4, but they will in future years. Outcomes School resources are not the ultimate measure of a school system’s success. In fact, if a state’s students are learning the lessons it expects them to learn and its schools require less resources than schools in other states, its school system is very efficient, providing a good return on the public funds invested in it. Although California schools are leaner than schools in other states, we cannot say that they are particularly efficient because students in California schools do not appear to be learning as much as comparable students in other states. One measure of student outcomes is the Stanford Achievement Test, Ninth Edition (SAT9). The SAT9 was the first test administered as part of California’s new accountability system. From the spring of 1998 to the spring of 2002, California public school students in grades 2 through 11 were required to take the SAT9 reading and math exams. Because the SAT9 was first given to more than 500,000 students throughout the country, the test permits a straightforward comparison of California students with students in other states. The test reveals that California students lagged behind students in other states, particularly in reading. For example, in 1998 only 39 percent of California students exceeded the national median score in reading. The performance of California students improved each year, but more than half were still below the national median by 2002. In math, the performance of California students was noticeably better. Whereas only 42 percent exceeded the national median in 1998, that number had reached 55 percent by 2002. Another measure of student achievement across states is the National Assessment of Educational Progress (NAEP), a standardized test of reading, mathematics, and other subjects administered by the U.S. Department of Education to a random sample of fourth, eighth, and twelfth graders across the country. The number of twelfth graders taking the test is too small to provide meaningful comparisons for some 53 subgroups of students, so we focus on the scores of fourth and eighth graders in reading and mathematics. California students compare poorly to students in other states. In both reading and mathematics, the average scores of California fourth and eighth graders are lower than the national average and lower than the scores in any of our four comparison states (Table 4.11). In each of these comparisons, the difference between California students and other students is statistically significant.1 The relatively poor results for California students may be partly explained by the large number of students who are recent immigrants and still not proficient in English. Indeed, 30 percent of California fourth graders who took the NAEP were classified by their schools as not fully proficient in English. In contrast, only 8 percent of fourth grade test-takers in the nation were not English proficient. Among our four comparison states, Texas, at 12 percent, had the highest percentage of test-takers who were not proficient in English. When test results are restricted to students who were proficient in English, California’s performance improves but still lags that of other states (Table 4.12). In reading, the average for California students in both fourth and eighth grade is lower than the national average and Table 4.11 NAEP Scores for All Students, 2003 Grade 4 Grade 8 California United States Florida Illinois New York Texas Reading 206 216 218 216 222 215 Mathematics 227 234 234 233 236 237 Reading 251 261 257 266 265 259 Mathematics 267 276 271 277 280 277 _____________ 1For fourth graders in the country as a whole, the standard errors of the average scores in reading and math are 0.3 and 0.2. For eighth graders, those standard errors are 0.2 and 0.3. For California students, the equivalent standard errors are 1.2, 0.9, 1.3, and 1.2. Standard errors for averages in other states are approximately the same as those for averages in California. 54 Table 4.12 NAEP Scores for Students Proficient in English, 2003 California United States Florida Illinois New York Texas Grade 4 Reading Mathematics 215 235 219 236 220 235 219 235 223 237 218 241 Grade 8 Reading Mathematics 258 274 265 278 259 273 268 279 267 282 261 279 the average of each of our four comparison states. In mathematics, California students do somewhat better. Fourth graders are slightly below the national average but are tied with fourth graders in Florida and Illinois. Eighth graders in California who are proficient in English have higher average mathematics scores than English-proficient eighth graders in Florida but fall short of all other comparison states and the national average. Because states may use different standards to determine whether a student is proficient in English, the results in Table 4.12 should be interpreted with caution. Another approach is to focus on the results of non-Hispanic white students because they are far less likely to be recent immigrants. The results for these students are similar to those for English-proficient students. The gaps between California white students and white students in other states are less than the gaps for all students, but California white students still performed poorly relative to other white students. In reading, California white students, both fourth and eighth graders, were below the national average for white students and the averages in each of our comparison states. In mathematics, California white students in the fourth grade equaled the national average for white students. In the eighth grade, however, they were below the national average and the averages of each of our comparison states. Similar comparisons could be made for other groups of students, an analysis undertaken in a recent RAND Corporation report (Carroll et al., 2005). The results of that analysis are similar to the results shown in Tables 4.11 and 4.12. When the NAEP scores of any group of 55 California students are compared with the scores of similar students from other states, California students tend to score lower on average than students in other states. The RAND study also adjusted state NAEP scores for differences among states in family characteristics such as income and parental education. Even after these adjustments, California students scored lower than students in other states. Conclusions The largest category of state and local public expenditures is elementary and secondary education. Although California’s expenditures per capita in this area were higher than the average for all other states in 2002, its schools had substantially fewer staff per student than schools in other states. There are two reasons for this outcome. First, a higher fraction of California’s total population is enrolled in public schools than in other states, spreading California’s expenditures in this area more thinly than in other states. Second, California school districts compensate their employees more generously than in other states. About half of this higher compensation can be explained by California’s labor market, where college-educated workers are paid higher salaries than in other states. Student achievement is also lower in California than in the rest of the nation. On average, California students score lower on the NAEP than students in other states. This result persists even after attempts to account for the higher percentage of California students who are recent immigrants and not yet proficient in English. A logical inference is that the relatively low resources in California schools explain the relatively low achievement of its students. The link between achievement and resources has received more attention in California recently because of the state’s new standards and accountability system. At the heart of that system is an achievement goal for every school, a goal that implies that each school’s students are expected to perform substantially better on standardized tests than students in other states (see Rose et al., 2003, and Rose, Sonstelie, and Richardson, 2004.) Although there is much academic debate about the link between resources and achievement (e.g., Hanushek, 1997, and Krueger, 2003), it is fair to ask whether California’s goals for its schools match its financial commitment to them. 56 5. Higher Education Higher education expenditures accounted for 8 percent of total state and local expenditures in California in 2002, the same share as in the rest of the nation. This chapter examines total spending for higher education in California, comparing it across states and over time. Given the significant contribution of fees for tuition, housing, and other services in the total funding picture, the chapter also includes revenues in the analysis to calculate net support (total expenditures less revenues) for higher education. Overall, California’s net support for higher education is above the average for the rest of the nation on a per capita basis. Enrollment in public colleges and universities in the state, however, is also significantly higher than in the rest of the country. The result is that the state’s net support on a per student basis is lower than in other states. The state does manage to produce more undergraduate degrees on a per capita basis than the average state, but because of its larger enrollments, it produces fewer degrees per full-time student. Total Expenditures Higher education refers to education and education services provided in degree-granting institutions beyond the high school level. The vast majority of the expenditures associated with public higher education institutions (e.g., instruction, research, public service, administration, libraries, and student services) falls under a category labeled by the U.S. Census Bureau as “other higher education” spending. The Census Bureau differentiates these expenditures from “auxiliary services” that provide supplementary services to students beyond the direct provision of education, such as dormitories, cafeterias, and bookstores. Separate designations also exist within each of these categories for capital expenditures. In addition to direct expenditures, 57 states also provide higher education subsidies to their residents in the form of student assistance. In California, relatively few programs and agencies fall under the heading of higher education. The major institutions providing higher education services consist of the University of California (UC) system, the California State University (CSU) system, and the California Community College (CCC) system. Funds to support small agencies (the California Postsecondary Education Commission, the ScholarShare Investment Board, and the California Student Aid Commission), as well as resources for state grants (Cal Grants) to California residents attending college are also included in this discussion. The expenditures and revenues associated with university-affiliated hospitals are not included. Overall, California’s higher education spending per capita in 2002 was 8 percent higher than such spending in the rest of the nation (Table 5.1). This total masks variation across individual categories. Expenditures in California for instruction, research, and administration (the Census Bureau’s “other higher education” category) were 11 percent above average spending in all other states. California spent less on auxiliary services while providing slightly more in financial aid. In contrast, student assistance in Texas was well below the national average, representing less than 5 percent of that state’s total higher education spending. In California, financial aid accounted for 8 percent of the total.1 Table 5.1 Expenditures on Higher Education per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 634 586 394 577 472 601 Other Higher Education 543 488 321 476 392 538 Auxiliary Services 39 51 26 41 25 37 Student Assistance 51 47 47 60 55 27 _____________ 1Note that not all financial aid provided by the state is spent at public institutions. The Cal Grant program, for example, provides assistance to state residents who attend a higher education institution in the state, public or private. 58 Within these categories, the Census Bureau also reports capital expenditures (not shown). Spending on capital projects in California represented less than 6 percent ($35 per capita) of total higher education resources in 2002, a level nearly one-half the size of the figure for the rest of the country ($65 per capita). The nature of capital projects, however, can make single-year comparisons of capital spending less meaningful. For this reason, a multiyear analysis of capital spending appears later in this chapter. Relative spending for higher education has changed over time. Per capita spending was 29 percent higher on average in California than in the rest of the nation between 1977 and 1987 (Figure 5.1). From the late 1980s until the early 1990s, the gap between the state and the rest of the country began to narrow. After dropping below the average for all other states in the mid-1990s, California increased its spending in the remainder of the decade, keeping pace with, and eventually outpacing, a nationwide growth in resources directed to higher education. 700 600 500 400 300 California U.S.–California 200 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.1—Higher Education Total Expenditures per Capita, 1977–2002 Higher Education Revenue Higher education differs from many of the other functional areas discussed in this report in that it has user fees, primarily in the form of 59 Expenditures (2002 $) tuition, which directly offset program expenditures. The Census Bureau provides data on two types of higher education revenue: income from auxiliary services and income from all other higher education charges, including tuition. California collected slightly less higher education revenue per capita than the rest of the nation in 2002 (Table 5.2). However, per capita revenues in Florida and New York were considerably lower than in California. Most of this difference was accounted for by significantly less money being generated from sources other than auxiliary services. These revenues, primarily tuition charges, were 40 percent lower in New York and 47 percent lower in Florida than in California. Historically, the trend for California and the nation (as well as for the other states in our comparison group) in higher education revenues has been one of general increases in real terms over the past two decades (Figure 5.2). As a group, California and the other large states considered in this report rank relatively low in higher education revenues, all generating less revenue per capita than the national average excluding California. Since the mid-1980s, Florida has consistently raised the least amount of higher education revenues. On the other hand, Texas has produced the most revenue per capita within this group for much of the past 20 years. In 2002, tuition increases across all of California’s higher education systems moved the state toward the top of our comparison group, although it still remained below the average for the rest of the nation. Table 5.2 Higher Education Revenues per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 205 215 114 196 128 196 Other Higher Education 167 163 91 155 100 166 Auxiliary Services 38 51 22 41 28 30 60 Expenditures (2002 $) 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.2—Higher Education Total Revenues per Capita, 1977–2002 Net Support for Higher Education The significant role that tuition and other revenues play in supporting higher education suggests that calculating the net support (total expenditures less revenues) allows a more meaningful comparison than simply examining total expenditures. Calculating a state’s net support for higher education rather than expenditures per capita changes the picture slightly. Net support for higher education in California was 15 percent above the average for the rest of the nation in 2002, the highest among our comparison states (Table 5.3). This adjustment has an even greater effect on relative rankings for other states. For example, Illinois shifts from a position slightly below the national average in expenditures per capita to one slightly above average in net support per capita. Similarly, New York’s total expenditures per capita were 20 percent below the average for the United States excluding California in 2002. However, after adjusting for New York’s relatively low rate of revenues generated by tuition and fees, the state rises to within 8 percent of that figure. 61 Table 5.3 Higher Education Net Support per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 429 372 280 380 343 405 Other Higher Education 428 372 277 381 347 399 Auxiliary Services 1 — 4 (1) (4) 7 Although state governments do not “make money” on higher education services (i.e., they do not generate more revenue than they spend), some appear to be able to use net income from their auxiliary services to offset other expenditures. About half of the states, including Illinois and New York, reported receiving more revenue from auxiliary services than they spent on those activities. It is also worth noting that this idea of cross-subsidizing extends to different types of instruction and students. Graduate students, in general, consume a greater share of resources than undergraduates do. Similarly, the actual cost of instructing nursing and biology majors is likely to be higher than the expenditures associated with students studying literature or philosophy. This study, which aggregates data to make comparisons, necessarily masks such differences. Net Support per Student As in all functional areas, comparing net support relative to a state’s population provides some sense of the commitment the state’s taxpayers have made to the program. An examination of expenditures relative to the total number of beneficiaries gives some indication of how those resources are being distributed. Similar to enrollment in K–12 discussed in Chapter 4, enrollment in state colleges and universities represents the caseload for higher education spending. Unlike K–12 education, however, many higher education students participate part-time. Thus, it becomes necessary to convert the number of students enrolled to full-time-equivalent students 62 (FTES) to compare the level of resources committed to higher education relative to the number of students. Using the approach outlined in Chapter 2, the relationship between expenditures, population, and enrollment (caseload) is described by Eq. (5.1) Expenditures = FTES  Expenditures Population Population FTES (5.1) Using enrollment data from the National Center for Education Statistics (NCES), Table 5.4 compares net support for higher education for each full-time-equivalent student in public institutions. Relative to comparisons on a per capita basis, normalizing support for higher education on a per student basis presents a different picture. Net support for students in California was 7 percent below the average for the rest of the country in 2002, whereas three of our four comparison states supported students at a higher level than this average. California’s relatively high participation rate in higher education explains much of this shift. Although the state’s spending per capita is higher than average, the state’s relatively high rate of student participation offsets that difference. Illinois, for example, spends less than California per capita but 14 percent more than California per full-time student. This change is a result of the fact that Illinois enrolls about three-quarters as many students as California. An alternative way of looking at this would be to assume that enrollment rates in public higher education were the same in California Table 5.4 Net Support per Full-Time-Equivalent Student, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures Revenue Net Support Net Support per Capita per Capita per Capita 11,329 634 205 429 12,177 586 215 372 11,268 394 114 280 12,958 577 196 380 14,276 472 128 343 12,550 601 196 405 FTES per Capita 0.038 0.031 0.025 0.029 0.024 0.032 63 as in the rest of the nation (i.e., 31 rather than 38 FTES per 1,000 residents). At this lower rate of participation, net support per FTES would increase to $13,836, or 14 percent higher than the average for all other states. California’s high participation rate in higher education has been consistent through the years, dating back more than three decades. By contrast, the funding picture has changed in recent years. Net support for higher education per full-time student in California has been near or above the average for all other states during much of the past decade (Figure 5.3). It is only in the last year for which comparable data are available, 2002, that a significant gap emerged between this state and the rest of the country. The appearance of this gap can be explained by modest expenditure growth, which did not keep pace with growing enrollments. The end result was declining net support per full-time student. Aggregating the data for all California public higher education institutions masks significant differences in the average expenditures per student across the three California systems. Although the Census data do not permit further distinctions by system, figures from the California Postsecondary Education Commission (CPEC) provide some sense of 15,000 Net support per FTES (2002 $) 10,000 5,000 California U.S.–California 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.3—Higher Education Net Support per Full-Time-Equivalent Student, 1991–2002 64 this variation. In 2002, CPEC reported that general purpose expenditures per FTES were $20,241 in the UC system, $10,407 in the CSU system, and $4,629 in the CCC System (California Postsecondary Education Commission, 2004). Participation and Student Fees One explanation for California’s relatively high rate of participation in higher education could be that this state includes a disproportionately high number of residents of traditional college age (i.e., ages 18–24) relative to other states. Within our comparison group, California does include more residents of traditional college age than any state except Texas. However, this is not the whole story. Instead, more older as well as traditional-age students are enrolled in higher education in California. For example, in 2002, 44 percent of postsecondary students in California were over age 25, compared to 37 percent in the rest of the nation and similar shares in all of our comparison states. The exception was Illinois, where older students represented 43 percent of total enrollment. Illinois, not coincidently, was also second only to California within our comparison group in its community college enrollment (see Figure 5.7 later in this chapter). California’s relatively high enrollment levels are not entirely explained by the rate of participation by older students. So-called traditional age students, ages 18 to 24, appear to enroll in higher education at a much higher rate than one finds in other parts of the country. Although the data do not allow a perfect comparison, it is possible to get an estimate of the level of participation by comparing the number of students enrolled in public education who are under age 25 with the Census Bureau’s population estimate for 18 to 24 year olds in each state. Using this approximation, one out of three residents 18 to 24 years old is enrolled in a public higher education institution in California. This number compares with an estimate of one out of five 18 to 24 year olds participating in public higher education in the rest of the country (Figure 5.4). Estimates for our comparison states fall between these two numbers. 65 0.05 0.04 Enrollment rate 0.03 0.02 0.01 0 California U.S.– California Florida Illinois New York Texas Figure 5.4—Percentage of Population Ages 18 to 24 Enrolled in Public Higher Education, 2002 Although the estimates in Figure 5.4 do not account for nonresident enrollments, Californians clearly prefer to “stay at home” when it comes to higher education, and residents represent the vast majority of students enrolled at the public institutions. First-year California residents who are enrolling in a college or university are more likely than first-year students in the rest of the country to stay within the state (Table 5.5). Table 5.5 Residence and Migration of First-Year Undergraduate Students, 2002 (percent) California U.S.–California Florida Illinois New York Texas Percentage of Students Who Are State Residents 0.92 0.81 0.81 0.90 0.81 0.93 Percentage of State Residents Enrolled in State Institutions 0.92 0.83 0.90 0.83 0.83 0.92 66 Higher education institutions within California enroll a relatively higher share of state residents (92%) than the average for all other states (81%). Texas exhibits a similar pattern. However, there appears to be no consistent pattern with regard to migration among the other comparison states. Illinois’ institutions enroll a higher than average number of state residents, but students from that state are just as likely as students in the rest of the country to look beyond their borders to attend a college or university. Florida exhibits the opposite pattern. Students in Florida are more likely to stay at home for their undergraduate education, but the capacity of those schools enables that state to enroll a larger share of nonresident students. California also enrolls a relatively large share of students at public schools rather than private ones. In this state, public two- and four-year schools account for 81 percent of the total number of students in higher education. The national average in academic year 2002–2003 was 72 percent. Again, Texas was similar to California, with 85 percent of its students in public schools. New York, in contrast, enrolled only 52 percent of students in public institutions. Florida (74%) and Illinois (67%) fell between these two extremes, on either side of the national average. In sum, more Californians than residents of other states attend higher education institutions, and when they do, they are more likely to attend public schools within their home state’s borders. Relatively low fees likely contribute to this higher rate of participation because they make higher education more attractive and more accessible. From one perspective, this increased accessibility is a positive development. Low fees, however, also generate lower revenues, which must be spread across a larger student population. In 2002, for example, higher education revenues per FTES (including both tuition and auxiliary revenues) were $5,404 in California, more than 20 percent lower than the national figure of $7,027 per FTES. Had California generated revenue per FTES at the same rate as the rest of the country, net support per student would have been $12,952 or 6 percent higher than the average for all other states. Generating less revenue from student fees is hardly a new development. California has had a tradition of low tuition for its 67 colleges and universities. In fact, enrollment in a California community college was free as recently as 1985. In the latter half of the 1990s fees were held constant, resulting in a real decline in revenues per student during the early part of that period (Figure 5.5). Despite shifts in fee policies, the difference between the national figure and the state’s revenues per FTES has been relatively stable, averaging 26 percent. Differing levels of support from state and local governments, then, have driven much of the fluctuation in net support for higher education per student relative to the average for all other states. Figure 5.6 compares the variability in higher education expenditures per FTES in California during the 1990s to the relative stability of revenues per FTES over this period. Again, using aggregate data does obscure differences across the three California systems. CPEC reports average revenue per FTES for instruction-related activities, which provides some sense of the variation across the three systems. For UC campuses, this figure was $3,706 in 2002. Comparable figures for the CSU and CCC systems were $1,610 and $149, respectively (California Postsecondary Education Commission, 2004). The pressure of budget deficits led to significant 8,000 Revenues per FTES (2002 $) 6,000 4,000 California U.S.–California 2,000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.5—Higher Education Revenues per Full-Time-Equivalent Student, 1991–2002 68 Revenues and expenditures per FTES (2002 $) 20,000 15,000 10,000 Revenues Expenditures 5,000 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 Figure 5.6—Higher Education Revenues and Expenditures per Full-TimeEquivalent Student, 1991–2002 fee increases in the years after 2002. CPEC estimated that by 2004, average revenue per FTES for instruction-related activities would have risen by approximately 30 percent relative to the 2002 figures. Staffing and Salaries Most of the cost of providing higher education services stems from the expenses associated with personnel. Using employment and payroll figures reported to the Census Bureau, it is possible to calculate ratios similar to those in Eq. (2.3) in Chapter 2. Total Payroll = Total Payroll  Total Employees FTES Total Employees FTES (5.2) The story in Table 5.6 is similar to the decomposition presented in Chapter 4 for primary and secondary education. In California, payroll expenditures per employee are above the average for the rest of the country but compensation per FTES is near average. This implies that the ratio of employees per full-time-equivalent student is lower in California’s public colleges and universities than in the rest of the country. Although below the average for the nation excluding 69 Table 5.6 Total Payroll, Costs per Employee, and Number of Employees per Full-Time-Equivalent Student, 2002 California U.S.–California Florida Illinois New York Texas Total Payroll Costs per FTES ($) 8,513 8,786 7,888 8,575 7,295 8,010 Payroll Costs per Employee ($) 31,148 27,970 28,869 26,363 30,665 29,858 Total Employees per FTES 0.273 0.314 0.273 0.325 0.238 0.268 California, this ratio is similar to that found in the other comparison states, with the exception of Illinois. Higher compensation costs per employee also translate into payroll accounting for a larger share of the total cost of higher education. In California, payroll costs represent 75 percent of net support per FTES (Tables 5.4 and 5.6). This figure is slightly higher than the average of other states (72%) and significantly higher than comparable percentages in most of our comparison states (70% of net support in Florida, 66% in Illinois, 64% in Texas, and 51% in New York). Comparing total payroll costs per employee could conceal differences in the use of part-time and full-time employees. Different shares of employees may be involved in noninstructional activities, also affecting payroll costs. In California, however, regardless of how the payroll figures are disaggregated, the average wages for both full- and part-time employees are higher than in the rest of the country (Table 5.7). The average salary for full-time instructional employees provides an approximation of what different states pay their full-time faculty. Table 5.8 provides salary comparison figures that suggest that the relatively high full-time faculty salaries are driven by the state’s labor market. In occupational Zone 5, where employees are typically required to hold an advanced degree, average salaries in California are 12 percent higher (see Chapter 2). Assuming that this is the comparable group for full-time 70 Table 5.7 Wages for Full- and Part-Time Instructional and Noninstructional Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Full-Time Full-Time Instructional Noninstructional Employee Salary Employee Salary 73,393 51,793 65,428 37,807 69,607 34,531 66,074 38,431 65,946 42,489 68,910 37,941 Part-Time Part-Time Instructional Noninstructional Hourly Wage Hourly Wage 26 15 20 10 21 10 20 8 21 14 22 10 Table 5.8 Full-Time Instructional Employee Salary Comparisons, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Payroll per Full-Time Employee 73,393 65,428 69,607 66,074 65,946 68,910 Adjusted Payroll per Full-Time Employee 65,502 66,437 70,556 68,048 57,260 69,693 public college and university instructional employees, labor market conditions would explain the variation in faculty salaries in the state. In fact, when full-time wages are adjusted using the location premium for Zone 5, salaries in California higher education institutions actually fall slightly below the average for the rest of the nation and most of our comparison states. Only New York reports lower adjusted payroll figures. In addition to the number of employees available to teach and assist students, different varieties of educational experiences may be provided by different types of institutions. A second indicator of the type of services provided in California relative to other states is the mix of students enrolled at two- and four-year institutions. Of the more than 1.3 million full-time-equivalent students attending public higher 71 education institutions in the state, only 40 percent of them were enrolled in four-year institutions in 2002 (Figure 5.7). This figure was the lowest in the nation. For the rest of the country, the average share of FTES at four-year schools was 64 percent. California, therefore, relies heavily on its community colleges for the delivery of its higher education services. In fact, it is one of only four states in the country where a majority of FTES was served by two-year institutions. (Wyoming, Illinois, and Washington are the other three.) This observation is not intended to suggest that one model is preferred to another. Nevertheless, it is the case that the higher education activities at four-year institutions are not the same as those at two-year colleges. Community colleges, for example, typically are not engaged in research to the degree found on the University of California campuses. Alternatively, lower division classes in the CCC system often are smaller than those found on a UC campus. What is clear is that public support for higher education is providing a different mix of services. In Florida, for example, net support for higher education was slightly lower than in California ($11,268 versus $11,329 per FTES) in 2002. Over three-fifths of the full-time- 0.6 FTES share at two-year institutions 0.5 0.4 0.3 0.2 0.1 0 California U.S.– California Florida Illinois New York Texas Figure 5.7—Percentage of Higher Education Enrollment at Two-Year Institutions, 2002 72 equivalent students in Florida, however, were attending four-year universities as opposed to only two out of five FTES in California. It is difficult to disentangle the cause and effect of participation rates and the structure of higher education in the state. One consequence of having an extensive community college system in California may be that the state ranks among the highest in the country in terms of participation in higher education, as noted above. On the other hand, one could argue that the high demand for higher education services requires that the state maintain and expand all of its colleges and universities. Capital Expenditures As noted already, California capital expenditures per capita on higher education were significantly lower than the average for the rest of the country in 2002. Historically, this has been the case since the mid1980s (Figure 5.8). When one considers that California tends to enroll students at a higher rate than the average for the rest of the nation, the comparison of capital investment in any one year becomes even less favorable. For example, expenditures per capita in 2002 were 46 percent below the average for all other states. Capital expenditures per FTES, however, were 57 percent below the average for the rest of the nation ($924 versus $2,127 per FTES). The uneven nature of capital investment, however, renders singleyear comparisons of expenditures less useful. Assuming that a portion of new capital spending is in response to increased demand for services, it also makes sense to examine expenditures relative to marginal increases in enrollment. Table 5.9 compares capital expenditures and higher education enrollment changes over the period 1992 to 2002. Illinois and New York, two states with modest growth in their higher education systems over this period, demonstrated the highest rate of investment in capital projects relative to the number of new students. These resources, most likely, were devoted to maintaining and upgrading existing facilities. Public higher education enrollment in the other three comparison states grew much more rapidly. California (22%), Texas (24%), and Florida (29%) all experienced significant growth. Of these three, California’s expenditures per marginal student were the lowest. 73 Expenditures (2002 $) 80 70 60 50 40 30 20 California 10 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 5.8—Higher Education Capital Expenditures per Capita, 1977–2002 Table 5.9 Increases in Enrollment and Real Capital Outlays, 1992–2002 California U.S.–California Florida Illinois New York Texas Increases in Enrollment (FTES) 235,407 949,140 92,880 5,043 15,022 138,674 Real Capital Outlay (2002 $) 15,709,674,024 134,099,860,961 7,337,850,760 6,139,377,670 8,489,813,146 10,392,607,350 Real Capital Outlay per Additional FTES (2002 $) 66,734 141,286 79,004 1,217,406 565,159 74,943 Outcomes Identifying appropriate outcome measures in education is difficult. The comparison of standardized test scores as a proxy for educational achievement, as presented in Chapter 4, provides a qualified measure at best. In higher education, however, the equivalent of standardized 74 testing does not even exist. Output measures such as degrees conferred, then, serve as a rough approximation for assessing outcomes. California’s investment in higher education produces undergraduate degrees at a higher rate than our comparison states in most cases but a lower level of graduate degrees (Table 5.10). Only Florida’s colleges and universities awarded more associate’s degrees on a per capita basis in academic year 2002–2003. For its population, California also produces more bachelor’s degrees than all of our comparison states and it is only slightly below the average for the rest of the nation. The comparison becomes less favorable when one examines the number of graduate and professional degrees produced by higher education institutions in California. California, given its population, produced a low number of graduates with master’s degrees, Ph.D.s, and professional degrees when compared to the average for all other states. Completion of degrees is a crude measure of output and does not capture other contributions of higher education. It is possible, however, to combine data on degrees, enrollments, and net support to capture a rough approximation of efficiency. By these measures, California produces a relatively low level of degrees given its enrollment (Table 5.11). The figure of 0.15 per FTES is below the average for the rest of the country and all of our comparison states. When one examines the cost of conferring these degrees, California appears to be particularly Table 5.10 Degrees Conferred at Public Colleges and Universities per 1,000 State Residents, Academic Year 2002–2003 California U.S.–California Florida Illinois New York Texas Associate’s Bachelor’s Degrees Degrees per 1,000 per 1,000 Residents Residents 2.10 2.97 1.67 3.04 2.46 2.83 1.83 2.58 2.01 2.29 1.40 2.93 Master’s Degrees per 1,000 Residents 0.63 0.96 0.70 0.89 0.85 0.97 Ph.D.s per 1,000 Residents 0.08 0.10 0.08 0.09 0.06 0.10 Professional Degrees per 1,000 Residents 0.07 0.12 0.08 0.09 0.06 0.14 75 Table 5.11 Degrees Conferred per Full-Time-Equivalent Student and Net Support per Degree, 2002 California U.S.–California Florida Illinois New York Texas Total Degrees Conferred Total FTES 204,711 1,324,686 1,491,827 7,722,706 102,648 415,071 68,886 369,251 101,085 460,507 120,443 701,793 Degrees per FTES 0.15 0.19 0.25 0.19 0.22 0.17 Net Support per FTES 11,329 12,177 11,268 12,958 14,276 12,550 Net Support per Degree Conferred 73,309 63,034 45,563 69,458 65,036 73,127 unimpressive. By this calculation, the state provided over $73,000 for each degree produced in 2002. This figure is 16 percent higher than the average for the nation as a whole excluding California and the highest of our comparison states. In terms of both measures, the ideal analysis would disaggregate production and costs associated with associate’s, bachelor’s, and graduate degrees. Unfortunately, the data do not permit such a decomposition. However, it is possible to compare the number of associate’s degrees conferred relative to total enrollment in two-year institutions and the sum of bachelor’s and graduate degrees divided by enrollment in fouryear schools. Such a calculation suggests that California’s four-year universities produce bachelor’s and graduate degrees at a rate that exceeds the average for all other states and most of our comparison states (Table 5.12). The community colleges, however, lag behind the rest of the nation by this measure. California is significantly lower than the average for the rest of the country (by 38%) in terms of the number of associate’s degrees conferred per FTES enrolled in the community college system. Definitive interpretation of the data in Table 5.12 is difficult. For example, the mission of the community college system in California includes activities well beyond just conferring associate’s degrees. Proficiency in preparing students for transfer to four-year universities 76 Table 5.12 Estimated Degrees Conferred per Full-Time-Equivalent Student at Two- and Four-Year Institutions, 2002 California U.S.–California Florida Illinois New York Texas Associate’s Degrees per FTES at Two-Year Schools 0.09 0.15 0.26 0.12 0.21 0.10 Bachelor’s and Graduate Degrees per FTES at Four- Year Schools 0.25 0.22 0.24 0.27 0.22 0.23 would not be captured by such a measure. Two-year colleges in other states, however, presumably also are tasked with multifaceted missions. An additional qualification of these estimates stems from the fact that they do not take into account student preparation. If students enter colleges and universities at different levels of proficiency, the subsequent degree “production” could be affected. Chapter 4 notes that there is variation between the states in the level of achievement on standardized tests. Such disparities could affect the success of students once they enter higher education. Conclusions Participation in higher education in California is higher than the average for the rest of the nation and higher than in the other large states used in this analysis. This fact drives the conclusions regarding support for higher education on a number of fronts. On a per capita basis, the state’s total higher education expenditures are higher than average. Because of California’s relatively low student fees and other revenues to higher education institutions, net support per capita for higher education is even higher (15%) than the average for the rest of the country. California’s colleges and universities also confer more associate’s and bachelor’s degrees per capita than the rest of the nation. 77 Higher relative enrollments change the picture. Net support per student in California is 7 percent below the average for all other states. The production of graduate and professional degrees is also below average relative to the state’s student population. And the costs associated with producing any type of degree are relatively high. One element that does not change with regard to comparisons on a per capita or enrollment basis is California’s expenditures for higher education capital projects. Regardless of the basis, the state has consistently spent less in this area. Finally, California’s labor market contributes to higher salaries for employees in the state’s colleges and universities. This factor, combined with less net support per student, results in a higher ratio of students to employees in the state’s schools. In sum, higher education in the state presents a mixed picture. Californians invest more than residents of other states in higher education per capita. They also charge relatively low fees to students, making higher education more accessible and leading to higher than average rates of participation. In return, they get more undergraduate degrees than other states per capita. On the other hand, because of high participation rates, California’s higher education system has fewer resources per student, lagging capital investment, and fewer employees on campuses and it produces fewer graduate and professional degrees relative to the state’s population. Finally, it appears that the state invests more resources per degree produced than states in the rest of the country. 78 6. Health Services Health services is the second-largest category of state and local government expenditures. In 2002, health services constituted 14 percent of California’s spending, compared to 17 percent in the rest of the nation. This category covers a wide array of health-related services, including the provision of medical care for low-income individuals, the construction and operation of public hospitals, and public health services such as mental health and alcohol or drug rehabilitation treatment as well as some environmental protection activities. This chapter focuses on the largest state-administered health program, Medicaid (known as Medi-Cal in California) and examines spending differences between California and other states on this program. It also analyzes expenditures on public hospitals and other health services. California is similar to the rest of the nation in total governmentfinanced health expenditures per capita. However, it allocates these expenditures differently, spending less on Medi-Cal and more on other health services and subsidies to public hospitals. California is unusual in that it enrolls a higher percentage of the population in Medi-Cal and spends considerably less per recipient than the other large states in our comparison group. Several factors contribute to this outcome. First, eligibility rules are more generous in California than in other states and, partly as a result of these rules, the Medicaid caseload mix is younger in California than in other states. Second, Medi-Cal reimbursement rates under the traditional fee-for-service (FFS) system are lower than the national average excluding California. Third, more Medi-Cal recipients are enrolled in managed care plans, and capitated payments per enrollee under these plans are substantially lower in California than in other states. Although it is difficult to link public spending on health to specific outcomes, it is interesting to note that despite lower Medi-Cal spending 79 per capita, fewer Californians are hospitalized as a result of delayed medical intervention than are the residents of other large states. However, more low-income Californians than low-income residents in other large states report poor access to health care. Total Expenditures State and local governments engage in various activities to conserve and improve public health and make basic medical care available to lowincome or medically needy residents. The U.S. Census Bureau summarizes state and local expenditures on health services in four government functions: vendor payments for medical care, own public hospitals, other hospitals, and other health. Vendor payments for medical care include payments from state and local governments to private hospitals and health care providers under federal or state welfare programs such as Medicaid, general relief, and public assistance. In California, this function includes payments made on behalf of programs such as Medi-Cal, Healthy Families, Access for Infants and Mothers, and Major Risk Medical Insurance. The second function, own public hospitals, includes all state and local expenditures on the acquisition, construction, and operation of government-administered hospitals. It also comprises payments from the state to these hospitals on behalf of public medical assistance programs such as Medi-Cal. In California, these hospitals include county hospitals, developmental centers for persons with developmental disabilities, state hospitals for the mentally disabled, and hospitals affiliated with public universities, such as the UCLA Medical Center. They also include public children’s, maternity, and orthopedic hospitals that are operated by counties or public universities. The other hospitals function comprises state and local expenditures on private hospitals. Examples include government payments to private corporations that lease and operate government-owned hospitals, and government payments for the construction of hospitals to be leased or turned over to private operators. In California, as in many of our comparison states, these expenditures are negligible and will not be considered here. 80 The fourth expenditure function, other health, includes all state and local health activities unrelated to public medical assistance programs or the provision of hospital care. In California, this function is primarily composed of local expenditures (including those out of state-local realignment revenues) on county health services (including Healthy Kids), primary care and family health services, and drug and alcohol abuse services. This category also covers state and local spending on public health, broadly construed to include programs outside the Department of Health Services such as air quality control, pesticide regulation, and toxic substance control. California spends about the same amount per capita on total health services as the rest of the country, although it allocates these expenditures differently across the four categories described above. In 2002, total health expenditures per capita were $1,173 in California, on par with the national average excluding California and slightly more than Florida or Texas (Table 6.1). A little more than one-third of these expenditures were local (including local expenditures of state funds). New York spent 59 percent more and Illinois spent 27 percent less than California. Compared to the rest of the country, California spent less per capita on vendor payments for medical care, more on own public hospitals, and more on other health activities. Expenditures on other hospitals were negligible in California and in all of our comparison states except New York, which spent $26 per capita on this category. Table 6.1 Health Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 1,173 1,175 1,044 856 1,863 1,009 Medical Vendor Payments 574 679 594 443 1,184 544 Own Public Hospitals 319 298 257 169 454 341 Other Hospitals 0 3 1 7 26 0 Other Health 280 195 191 238 199 124 81 California has been on par with the rest of the nation in health expenditures throughout recent years. This picture stands in contrast to the late 1970s, when California spent more per capita on health services than the rest of the nation (Figure 6.1). However, as the rest of the country experienced more rapid growth in health spending, the expenditure gap between California and other states narrowed in the 1980s and disappeared by 1990. Health expenditures grew rapidly in both California and the rest of the nation in the early 1990s due mainly to an escalation of payments to Medicaid Disproportionate Share Hospitals (DSH) serving a large number of Medicaid patients and uninsured individuals. This growth flattened out in the late 1990s as a result of federal actions. Health expenditures rose again after 2000 in both California and the rest of the nation. State Medicaid programs constitute more than half of state and local health expenditures in California and the rest of the country. We therefore devote the next section to an analysis of spending on this program in California and other large states. We then briefly examine government subsidies to publicly owned hospitals and state expenditures on other health activities. Because Census of Governments data are not 1,400 1,200 1,000 800 600 400 California 200 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.1—Total Health Expenditures per Capita, 1977–2002 82 Expenditures (2002 $) designed to study individual programs, our analysis of Medicaid expenditures relies on data obtained from the Centers for Medicare and Medicaid Services (CMS). Medicaid Medicaid is the largest state-run health care program in every state. It pays for health care and long-term care for certain members of the low-income population, including families with children, pregnant women, individuals with disabilities, and the elderly. Medicaid is distinct from Medicare, the federal health insurance program for elderly or disabled persons. However, certain low-income Medicare beneficiaries may be entitled to Medicaid benefits for services not covered under Medicare (e.g., nursing home services and prescription drugs before January 2006) or payments of insurance premiums, deductibles, and co-payments under Medicare. Medicaid is jointly funded with federal and state revenues. The federal government matches state expenditures on Medicaid according to a formula that is inversely related to state income per capita. In 2002, California received an average of $1.06 from the federal government for every dollar it spent on Medi-Cal, similar to Florida and New York, which each received one federal dollar for each state dollar spent on Medicaid. By contrast, Florida received $1.30 for every dollar spent on Medicaid and Texas received $1.51. The comparisons that follow include both state and federal funds. To receive matching federal funds under Medicaid, states must provide a minimum set of “mandatory” services, including physician services, inpatient and outpatient hospital care, laboratory and x-ray services, nursing facilities, and home health care. Although these services are called “mandatory,” states have discretion over the amount, duration, and scope of services covered. For example, some states limit the number of nonemergency hospital visits per year. Beyond these mandated benefits, states can also provide “optional services” such as prescription drugs, clinic services, prosthetic devices, hearing aids, dental care, and long-term and personal care services for the elderly or disabled. In addition, states may apply for waivers from federal program requirements to provide services to individuals not traditionally 83 eligible for Medicaid or to provide additional benefits to certain groups. For example, the home- and community-based services (HCBS) waiver enables states to provide noninstitutional long-term care services to frail elderly and disabled individuals.1 To be eligible for federal funds, states must provide Medicaid coverage to certain “categorically eligible” groups. These groups include recipients of aid from federal programs such as Supplemental Security Income (SSI), those who would have formerly been eligible for Aid to Families with Dependent Children (AFDC), low-income children, pregnant women, parents of dependent children, and certain elderly individuals. In addition, states can expand coverage to the “medically needy” and other special groups who are not otherwise eligible for Medicaid. Income criteria for these groups can vary substantially by state and these differences affect the proportion of the low-income population enrolled in Medicaid in each state. Since 1997, states have had the option of expanding health insurance for children not otherwise eligible for Medicare under the State Children’s Health Insurance Program (SCHIP). As of 2002, 20 states (including New York and Texas) did so through a separate program, 15 states expanded eligibility for these groups within Medicaid, and the remaining states combined both options. Among our comparison states, California, Florida, and Illinois were in the latter group. Medicaid is a “vendor payment” program that makes payment to health care providers on behalf of program recipients. States have the flexibility to purchase covered services from health care providers on a fee-for-service basis or from managed care plans on a per enrollee, or capitation, basis. Under both systems, states negotiate reimbursement rates with providers. Federal Medicaid law does not establish specific floors or ceilings for provider payments. Reimbursement rates vary widely across states, although most Medicaid programs reimburse providers at a rate less than the full rate allowed by the federal government. _____________ 1Some HCBS services overlap with existing “optional services” such as personal care services. Other HCBS services, such as adult day health services and respite care, are exclusively available under the waiver program. 84 Compared to the average for all other states, California has spent consistently less per capita on Medicaid since the 1990s. In 2002, MediCal expenditures were $667 per capita, 8 percent less than the average of all other states (Table 6.2). New York spent more than twice as much per capita as California, whereas Illinois was very close to the average for the nation excluding California. Florida and Texas spent 12 and 23 percent less than California, respectively. California’s nearly average Medicaid expenditures per capita mask dramatic differences compared to the rest of the nation. Using the general formula from Chapter 2, we can decompose Medicaid expenditures per capita for a given year into the product of the number of individuals who received services in that year and spending per recipient: Medicaid Expenditures = Medcaid Recipients Population Population  Medicaid Expenditures Medcaid Recipient (6.1) Medicaid expenditures per recipient are substantially lower in California than in the rest of the United States or in any of our comparison states. In 2002, California spent $3,113 per Medicaid recipient, 36 percent less than the average for the rest of the nation (Table 6.2). Each of our comparison states spent at least 20 percent Table 6.2 Components of Medicaid Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 667 726 589 721 1,644 512 Expenditures per Recipient ($) 3,113 4,878 3,726 5,393 8,101 3,774 Recipients per Capita 0.21 0.15 0.16 0.13 0.20 0.14 85 more per recipient than California, with New York spending two and a half times as much. At the same time, Medicaid enrollment was higher in California compared to the average for all other states. Twenty-one percent of Californians were enrolled in Medicaid in 2002, compared to an average enrollment rate of 15 percent in the rest of the country. New York enrolled 20 percent of the population in Medicaid in 2002, the secondhighest enrollment within our comparison group. The high Medicaid enrollment rate in California relative to other large states could result from differences in the generosity of eligibility policies and behavioral responses to these policies or to differences in the economic and demographic characteristics of each state. For example, relative to some other states in our comparison group, California includes more children and fewer disabled residents as a share of the total population (Table 6.3). To understand the interaction of these factors, we can analyze the participation rate for each eligibility group separately. For each group, the participation rate is the poverty rate for that group multiplied by the rate at which low-income members of the group enroll in Medicaid: Medicaid Recipients = Low  Income Population Population Population  Medicaid Recipients Low  Income Population (6.2) Table 6.3 Composition of State Population, by Eligibility Groups, 2002 (percent) California U.S.–California Florida Illinois New York Texas Child 30 28 26 29 27 31 Nonelderly Adult 51 50 48 52 52 50 Elderly Adult 6 7 10 7 8 5 Disabled Adult 13 15 16 13 14 13 86 In general, financial eligibility criteria for Medicaid are more generous for children than for nonelderly, elderly, and disabled adults. However, there is variation among states in these criteria. We measure the low-income population as the number of residents living below 200 percent of FPL ($30,040 for a family of three in 2002) for nonelderly adults as well as for elderly and disabled adults, and 300 percent of FPL ($45,060 in 2002) for children. For most states, these are the most generous income criteria under which individuals in these eligibility groups may qualify for Medicaid. These measures therefore ensure that all who may be potentially eligible for Medicaid are included in the low-income population estimates for each group. In 2002, California enrolled 11 percent of nonelderly adults into Medicaid, about twice the average enrollment rate for this group in the rest of the country. California’s enrollment rate for nonelderly adults was also more than twice that of Florida, Illinois, or Texas. New York was a close second among our comparison states, with an enrollment rate of 10 percent for this eligibility group (Table 6.4). California’s high enrollment rate for nonelderly adults in part reflects that California includes more low-income residents in this eligibility group than other states. More than one-quarter of nonelderly adults in California had family incomes below 200 percent of FPL in 2002, compared to 22 percent of adults in the rest of the nation and similar shares in New York and Illinois (Table 6.4). On the other hand, the Table 6.4 Components of Medicaid Recipients per Nonelderly Adult, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Nonelderly Adults 11 6 5 5 10 4 Share of LowIncome Enrollment Rate Nonelderly for Low-Income Adults Nonelderly Adults 26 45 22 26 28 19 20 26 22 47 30 15 87 proportion of adults with incomes below 200 percent of FPL was even higher in Florida and Texas than in California in 2002. Although California is in the middle of our comparison group in terms of the percentage of nonelderly adults below 200 percent of FPL, it had many more Medicaid recipients within this group. Forty-five percent of low-income nonelderly adults were enrolled in Medicaid in 2002, second only to New York with 47 percent. By contrast, the enrollment rate among low-income adults was 26 percent or less in the rest of the nation and even lower in our other comparison states. In general, higher participation rates within a target group could result from more generous eligibility rules or higher participation or “take-up” rates. However, there is reason to believe that in California and New York it stems from more generous eligibility criteria. For example, in 2002 pregnant women with family incomes up to 200 percent of FPL were eligible for Medicaid in California, New York, and Illinois, whereas the income cutoffs were 185 percent of FPL in Florida and Texas. In addition, California covered working parents with family incomes up to 107 percent of FPL under Medi-Cal, although the federal government mandated coverage only of single-parent or two-parent families in which the principal earner was unemployed. New York was even more generous, covering all adults with children and family incomes up to 150 percent of FPL as well as adults without children and family incomes up to 100 percent of FPL.2 The story is somewhat different for elderly and disabled Medicaid recipients. Like New York, California enrolls many elderly and disabled individuals (Table 6.5). However, its poverty rate for this group is lower than in other states. These two facts imply that California enrolls a higher proportion of low-income elderly and disabled adults in Medicaid than does the rest of the nation. The high enrollment rate for this target group is again largely explained by generous eligibility rules. Medi-Cal provided full coverage _____________ 2We also analyzed the enrollment rates of adults with and without children separately. The basic findings on the effects of the income distribution and eligibility rules remain the same. 88 Table 6.5 Components of Medicaid Recipients per Elderly or Disabled Adult, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Elderly or Disabled Adults 24 18 15 16 25 15 Share of LowIncome Elderly or Disabled Adults 35 38 37 35 38 41 Enrollment Rate for Low-Income Elderly or Disabled Adults 70 47 42 45 65 36 for these individuals with family income up to 100 percent of FPL, whereas other states applied the income criteria for the federal SSI program. California also set income criteria for Medicaid and Medicare “dual eligibles” at the upper limit of federal mandates.3 As described above, low-income children may receive health insurance coverage through Medicaid, SCHIP, or some combination of these programs. For example, California operates both a Medicaid expansion SCHIP and a separate SCHIP program (Healthy Families) for low-income children not covered by Medicaid. Thus, we analyze the enrollment rate for children in both of these programs. In 2002, California’s Medicaid enrollment rate for all children was slightly higher than the average for the rest of the country but lower than that of Florida, Texas, or New York (Table 6.6). Compared to Florida and Texas, California had fewer children living in families with incomes below 300 percent of FPL. However, its proportion was higher than in New York or Illinois and higher than the average for the rest of the nation. _____________ 3Elderly and disabled groups are combined in this analysis mainly because of data restrictions. In general, individuals who could qualify for Medicaid through either category are more likely to be assigned to the elderly category because it is less complicated to verify this status than disability. This practice could artificially raise the enrollment rate of the elderly while suppressing that of disabled individuals. 89 Table 6.6 Components of Medicaid Recipients per Child, 2002 (per 100 persons) California U.S.–California Florida Illinois New York Texas Enrollment Rate for Children 35 33 39 28 47 37 Share of Low-Income Children 58 56 61 52 54 62 Enrollment Rate for Low-Income Children 61 58 63 53 87 59 California also enrolled more low-income children in Medi-Cal and SCHIP than did the rest of the nation, although it enrolled a similar proportion as Florida and Texas. Within our comparison group, New York had the highest enrollment rate as a proportion of the target group, with 87 percent of low-income children covered under Medicaid or SCHIP in 2002. As with other eligibility groups, high enrollment rates for lowincome children in California may result from eligibility rules or take-up rates. A review of eligibility rules suggests that these policies are at work. Both New York and California covered children in families with incomes up to 250 percent of FPL whereas income cutoffs were 200 percent of FPL in Florida and Texas and 185 percent of FPL in Illinois. Not all Medicaid enrollees are eligible for the full range of Medicaid benefits. Individuals not fully insured by Medicaid include undocumented immigrants receiving emergency aid and elderly people for whom Medicaid pays a Medicare premium. These individuals accounted for a nontrivial proportion of Medicaid enrollment in California. Excluding them, the gap in enrollment rates between California and other states is not as stark. In an average month of 2001, 15 percent of Californians were enrolled in Medicaid or SCHIP and eligible for full benefits, lower than the 17 percent enrollment rate in New York although still higher than 90 Table 6.7 Components of Medicaid and SCHIP Enrollment, 2001 (rate per 100 residents) California U.S.–California Florida Illinois New York Texas Average Monthly Enrollment 15 12 11 11 17 10 Monthly Take-Up Rate 78 64 68 85 70 56 Average Monthly Eligible 20 18 17 13 24 18 the 12 percent average in the rest of the county, 10 percent in Texas, and 11 percent in Florida and Illinois (Table 6.7). For all Medicaid and SCHIP enrollees as a group, we can separate the effects of eligibility rules and demographics from individual participation decisions, or take-up rates. In the following equation, the first term measures the take-up rate of Medicaid or SCHIP among those who are eligible for the programs; the second term reflects the share of population who are eligible for these programs based on eligibility criteria set forth by states and each state’s demographic and economic characteristics: Medicaid and SCHIP Recipients = Medicaid and SCHIP Recipients Population Medicaid and SCHIP Eligibles  Medicaid and SCHIP Eligibles Population (6.3) The results of this decomposition presented in Table 6.7 show that, in an average month of 2001, 20 percent of Californians were potentially eligible for the full benefits of Medicaid or SCHIP, and 78 percent of them took up the services.4 Thus, both the eligibility rate and the take_____________ 4Estimates of potential eligible population and participants were obtained from the Urban Institute’s TRIM3 model. 91 up rate were higher in California than the average in the rest of the country. In sum, enrollment rates are higher for all Medicaid eligibility groups in California than in other states. California enrolled 6 percent more children, 33 percent more elderly and disabled, and about twice as many nonelderly adults into Medicaid than the rest of the country in 2002. These high enrollment rates do not result from substantially higher poverty rates among potentially eligible groups but rather from a combination of generous eligibility rules and high take-up rates. Although high enrollment rates are in evidence for all eligibility groups, the elderly constitute a smaller share of the overall population in California. Thus, the state’s overall high enrollment rate is primarily driven by the nonelderly population. Higher nonelderly enrollment also helps to suppress average expenditures per Medicaid recipient in California. Nonelderly beneficiaries tend to consume fewer services, and the services they consume are relatively less expensive—for example, clinic services as opposed to skilled nursing facilities (e.g., MaCurdy et al., 2005). In 2002, for example, Medi-Cal expenditures for children and nonelderly adults were about one-fifth those for the elderly or disabled (Table 6.8). Although this gap was similar in other states in our comparison group, California includes more children and nonelderly recipients than other states. Medi-Cal’s low spending per recipient can also be explained by higher managed care penetration in this state. In 2002, 85 percent of Table 6.8 Medicaid Expenditures per Recipient, by Eligibility Group, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 3,113 4,878 3,726 5,393 8,101 3,774 Children 1,715 2,447 1,189 1,503 5,771 1,702 Nonelderly Adults 1,549 2,505 1,988 2,769 3,951 2,694 Elderly or Disabled 9,530 11,592 10,103 15,161 18,669 11,403 92 Medi-Cal recipients were enrolled in some type of managed care plan that provided them with either a comprehensive set of services or a limited set of benefits such as mental health, dental, behavioral health, prenatal, or long-term care. The share of Medicaid recipients enrolled in managed care was 45 percent in the rest of the country and similar in our other comparison states (Table 6.9). In addition, California paid managed care organizations less per enrollee than other states. At $780 per enrollee annually, managed care reimbursements were 41 percent lower than in the rest of the county, 47 percent lower than in New York, 32 percent lower than in Illinois, 29 percent lower than in Texas, and 24 percent lower than in Florida. For recipients not enrolled in managed care plans or those enrolled in plans with limited benefits, Medicaid services are directly purchased from providers on a FFS basis. As the last column of Table 6.9 implies, the bulk of Medicaid expenditures occur under the FFS system. Overall, California reimburses these providers at a rate that is 91 percent of the national average, although there are some differences by type of service (Table 6.10). Variation in Medicaid reimbursement rates may reflect the costs of providing medical services in different states. However, California is a high-cost state in which to practice medicine. According to a 2001 study, it costs 7 percent more to provide physician services in California than the national average. Among the four comparison states, only New York is more expensive than California (Lewin Group, 2001). Table 6.9 Components of Medicaid Expenditures per Recipient, 2002 California U.S.–California Florida Illinois New York Texas Total Managed Care Managed Care Expenditure Expenditures Payment per Enrollment per Share of per Participant Participant Participant Managed Care 3,113 780 0.85 0.21 4,878 1,332 0.45 0.12 3,726 1,023 0.41 0.11 5,393 1,028 0.09 0.02 8,101 1,459 0.42 0.08 3,774 1,099 0.35 0.10 93 California United States Florida Illinois New York Texas Table 6.10 Medicaid Fee Index, 2003 Overall 91 100 95 92 70 99 Primary Care 87 100 96 89 71 96 Obstetric Care 83 100 104 103 88 93 Other Services 109 100 83 93 46 109 One way to adjust for geographic variation in the cost of providing medical services is to compare Medicaid reimbursement rates with the corresponding Medicare fee schedules for similar services in each state. Because Medicare fee schedules are supposed to reflect differences in the price of inputs used in furnishing medical services in each state and not case mix, the ratio of Medicaid to Medicare fees would be the same in California and the other states if the differences in Medicaid reimbursement rates represented solely the difference in cost of providing these services in California and the other states. Although Medicaid fees lagged well behind Medicare fees in all the states, the gap was larger in California. Overall Medi-Cal fees were 59 percent of Medicare fees, less than the national average of 69 percent (Table 6.11). This suggests that California reimbursed Medicaid providers at a lower rate than the national average even after adjusting for Table 6.11 Medicaid-to-Medicare Fee Indexes, 2003 California United States Florida Illinois New York Texas Overall 59 69 65 63 45 69 Primary Care 51 62 60 54 40 62 Obstetric Care 65 84 82 84 65 82 Other Services 74 73 58 68 31 82 94 geographic differences in the cost of medical services. Specifically, California paid significantly less for primary care and obstetrical care than the national average, while it reimbursed hospital services slightly more than the national average. In sum, Medi-Cal is distinct from other Medicaid programs because of its large enrollment and low expenditures per recipient. California has generous eligibility criteria for all eligible groups and high take-up rates among eligible individuals compared to the other large states we have considered. A younger low-income population in California as well as a generous enrollment policy has also helped suppress Medi-Cal expenditures per recipient because children and nonelderly adults have relatively low average medical costs. Extensive use of managed care and low rates of reimbursements for medical services provided on a fee-forservices basis brought expenditures per recipient further down in California than in other states. Own Public Hospitals Public hospitals administered by state and local governments serve a large proportion of Medicaid and uninsured patients and play an important role in the health care safety net. Expenditures per capita by these hospitals have been a bit higher in California than in the rest of the country since 1977 (Figure 6.2). A relative increase in California during the early 1990s reflects higher federal DSH payments to public hospitals in these years. Public hospitals receive revenues from patient charges (including those paid by private and public insurance programs), auxiliary services such as hospital cafeterias and gift shops, and government grants. To determine how much these hospitals rely on government subsidies for daily operations, we subtract these revenues from total hospital expenditures. The net amount measures subsidies that hospitals receive from state, local, and federal governments. Compared to the average for the rest of the country, California subsidized publicly administered hospitals at a higher rate in 2002. Out of public hospital expenditures of $319 per capita, 71 percent was financed with hospital charges and receipts and 29 percent reflected government subsidies (Table 6.12). In the rest of the country, the 95 Expenditures (2002 $) 400 350 300 250 200 150 100 California U.S.–California 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.2—Own Public Hospital Expenditures per Capita, 1977–2002 Table 6.12 Components of Own Public Hospital Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 319 298 257 169 454 341 Charges 227 228 218 72 268 205 Subsidy 92 70 39 97 186 136 average subsidy to public hospitals amounted to $70 per capita, or 23 percent of total spending. Among our comparison states, Florida spent 19 percent less per capita than California on public hospitals and subsidized 15 percent of the total hospital spending. New York and Texas spent more per capita than California and provided higher subsidies (40% of total expenditures). Illinois spent about half as much per capita as California on public hospitals, but 58 percent of this spending came from government subsidies. 96 The ratio of government subsidies to total hospital expenditures does not necessarily reflect the self-sufficiency of state hospitals. Several factors can affect the amount of subsidies hospitals receive. For instance, county hospitals may shoulder more responsibility in providing care to the medically indigent population and as a result may obtain more government funding. It is also important to note that government subsidies include subsidies to general hospitals and institutions for the developmentally disabled. Since the latter institutions are likely to receive more government funding, states with a larger share of such institutions may exhibit higher government subsidies. A rigorous study of the self-sufficiency of public hospitals is beyond the scope of this report. Other Health Services California has always spent more than the average for all other states on the category known as other health services, which covers a broad array of activities including mental health and developmental services, public health, and some environmental protection (Figure 6.3). In 1977, California spent 28 percent more per capita than the average of the rest of the nation on this function. Since then, expenditures per 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 6.3—Other Health Services Expenditures per Capita, 1977–2002 97 Expenditures (2002 $) capita on other health services have grown consistently relative to expenditures in the rest of the country. The early 1990s in particular witnessed a sharp rise in other health expenditures relative to the rest of the nation. This sharp increase coincided with a realignment of health and social services programs in California that transferred responsibilities for certain health care and mental health programs from the state to counties. In exchange, counties received additional revenues from a state-levied halfcent sales tax and increased vehicle license fees. In 1992, these revenues amounted to $1.5 billion. By 2002, they grew to $2.6 billion. The gap between expenditures on other health services between California and the rest of the nation further widened after 1996, when the state increased funding for developmental services related to health. It would be instructive to identify the specific health activities that drive California’s higher spending in this area. However, in California as in most other states, the bulk of activities included in this function are administered at the county level and comparable information across states on these local functions does not exist. Outcomes It is notoriously difficult to link public spending to health outcomes. However, Californians fare well in terms of two commonly accepted measures of quality of care—preventable hospitalizations and birth outcomes. They fare less well on a third measure—access to care. The preventable hospitalization rate reflects whether patients have access to timely and effective ambulatory care. Such care can help prevent hospitalizations for chronic conditions such as diabetes, asthma, and congestive heart failure, or for the deterioration of acute conditions such as infections and cellulitis or preventable diseases such as tetanus and rheumatic fever. A high rate of preventable hospitalizations therefore suggests access problems or poor performance of the health system. According to a study by the Agency for Healthcare Research and Quality (AHRQ), California had lower preventable hospitalizations for all age groups in 1999, the latest year for which complete data are available. Its rate of preventable hospitalizations was particularly low for 98 the nonelderly population compared to the nation overall and the three comparison states for which data are available (Table 6.13). We should be careful in attributing the low preventable hospitalization rate in California to government expenditures on health services. Other factors such as physician practice style and the local health care delivery system in private health care markets can also affect hospitalizations. For example, physicians have discretion over whether to admit a patient to a hospital or to manage the case on an ambulatory basis. The West Coast in general exhibits lower use of hospital services in studies of regional differences in physician practice style (Billings and Weinick, 2003). Similarly, the prevalence of HMO-type managed care plans in the private health insurance market in California could reduce hospital use among those who are insured, which in turn contributes to a low rate of preventable hospitalizations (Glied, 2000). As noted above, birth outcomes provide another measure of quality of care. Common measures include deliveries to mothers who received no or late prenatal care (in the third trimester), low-weight newborns (weighing less than 2,500 grams after a gestation period of 37 weeks or longer), and preterm births (deliveries before 37 weeks gestation). Compared to the national average and the other three large states for which data are available, California reported the lowest rates for all three indicators in 1999 (Table 6.14). On another measure, self-reported access to care, California did not fare as well. In major U.S. metropolitan areas, two of every ten persons with income up to 200 percent of FPL reported that they had no usual Table 6.13 Average Preventable Hospitalizations per 1,000 Persons, 1999 California United States Florida Illinois New York Ages 0–17 8 10 12 11 13 Ages 18–39 5 7 8 8 8 Ages 40–64 16 19 20 22 21 99 Table 6.14 Birth Outcome Indicators per 100 Births, 1999 California United States Florida Illinois New York Late or No Prenatal Care 3 4 3 4 5 Low Birth Weight 2 3 3 3 3 Preterm Births 10 12 13 12 11 place of care if they were sick or in need of health advice. This ratio was higher in certain regions of California (Table 6.15). Four out of ten individuals with incomes up to 200 percent of FPL reported having no usual source of care in the Ventura area, and about three out of ten Table 6.15 Access to Care Measures: Percentage of Population Below 200 Percent FPL, 1999–2000 California Los Angeles–Long Beach PMSA Orange County PMSA Riverside–San Bernardino PMSA San Diego MSA Ventura PMSA Oakland PMSA Sacramento PMSA San Francisco PMSA No Usual Source of Care No Doctor Visit in Past Year 25 26 30 33 28 27 27 21 39 35 12 16 18 23 19 23 Other Areas U.S. national average Miami PMSA Chicago PMSA New York PMSA 21 21 31 27 25 20 15 13 NOTES: MSAs are Metropolitan Statistical Areas as defined by United States Office of Management and Budget. PMSAs are Primary MSAs, or components of larger Consolidated MSAs (CMSAs). 100 responded similarly in the Orange County, Riverside–San Bernardino, San Diego, and Los Angeles–Long Beach areas. By comparison, Miami and Chicago also reported higher rates of no usual source of care than the national average, and New York reported a lower rate. Low-income individuals from areas with high rates of no usual source of care were also more likely to report not having visited a doctor within the last year. Various factors affect the self-reported measure of access to care. For example, the uninsured population is more likely than those who are insured to report problems with access to health care, and the uninsured rate is higher in California than the national average. Some studies suggest that higher Medicaid physician fees increase physicians’ participation in Medicaid (Zuckerman et al., 2004), thereby improving access to care by Medicaid patients. Although this link is not definite, low rates of reimbursement for primary care services of Medi-Cal could contribute to the relatively poor self-reported access to care by lowincome individuals in California. Conclusions California has the largest state and local health budget in the nation, with Medicaid program expenditures accounting for over half of this spending. Although the state’s total health expenditures per capita in 2002 were in line with those in the rest of the nation, per capita Medicaid expenditures were below this average. Medicaid expenditures per recipient were significantly lower in California than in the rest of the country, whereas Medicaid enrollment rates were higher. To a large extent, this pattern is explained by generous eligibility rules for children and adults under Medi-Cal and a younger low-income population in California. Other factors affecting expenditures per recipient in California are high managed care enrollments, low payments per managed care enrollee, and low reimbursements under the fee-forservice system. A quarter of California’s health budget was spent on hospitals, with governments subsidizing publicly administered hospitals at a rate higher than that in the rest of the nation although at a comparable rate to other large states. The remaining quarter of the state and local health budget 101 was spent on miscellaneous health services, most of which were administered at the local level, and California spent more per capita on these services than other states did. 102 7. Social Services In 2002, California state and local governments spent $18.7 billion on support for low-income households, including both cash payments and subsidized services. Expenditures in this area represented 6 percent of total state and local spending, compared to an average of 4 percent in the rest of the country. This chapter analyzes differences in social services expenditures between California and other states, focusing on major assistance programs such as Temporary Assistance for Needy Families (TANF), community services (e.g., child care, foster care, adoption assistance, and supportive services for the elderly and disabled), and the State Supplementary Program to Supplemental Security Income (SSI/SSP). California spends more per capita than the rest of the country on most types of assistance for low-income households. However, there are differences by program. California has higher maximum allowable benefits and serves more recipients per capita under its TANF program than does the rest of the country, but it spends less on average per recipient. By contrast, expenditures per recipient for community services programs and aid to the low-income aged, blind, and disabled are higher in California than in comparable states. Total Expenditures State and local governments in California provide services to the state’s low-income population through several programs. Perhaps the best known program is California Work Opportunity and Responsibility to Kids (CalWORKs), which provides cash aid, child care, and job training assistance to low-income families. CalWORKs implements the federal TANF program, a block grant program created in 1996 to replace the entitlement program AFDC. Under TANF, states receive fixed block grants to operate programs of their own design within federal limits on maximum lifetime participation and work requirements. In 103 exchange for these grants, states must provide matching funds, often referred to as “Maintenance of Effort” (MOE). In addition, states may choose to supplement these expenditures with additional resources. Throughout this chapter, we refer to state expenditures of federal funds and state matching funds as “federally mandated” expenditures and to the state supplements as “additional state spending.” The 1996 federal welfare reform effort also expanded funds for child care services, which are available to TANF recipients under different rules than those applicable to child care for other low-income working families. States can fund child care services for TANF recipients or other low-income Californians with grants from the federal Child Care and Development Fund (CCDF) and the Social Services Block Grant (SSBG) in addition to their own resources. As in welfare, some child care expenditures are mandated under the terms of participating in federal programs. In addition to welfare and child care programs, state and local governments operate a variety of community-based assistance programs. These include child welfare services such as the protection of abused or neglected children, foster care, and adoption assistance. States and localities may also provide supportive services for the elderly and disabled, including the developmentally disabled. States and localities have substantial flexibility in establishing these programs, often with support from federal SSBG funds. However, there are no federally mandated expenditures in this area, and states may supplement federal funds with their own resources. In the case of California, this state supplement is substantial. The final major assistance program is the State Supplementary Program to the federal Supplemental Security Income program, which provides cash transfers to low-income elderly, blind, and disabled individuals. Together, these programs are known as SSI/SSP. Under SSI, the federal government establishes eligibility rules and mandates minimum benefit levels to be funded by both the federal government and the states. States also have the option of providing additional aid. California is one state that chooses to supplement federally mandated aid. It also contracts with the federal Social Security Administration to administer SSI/SSP. 104 The U.S. Census Bureau classifies social services expenditures not by program but by type of aid: federal categorical assistance, other cash assistance, other welfare, vendor payments, welfare institutions, veterans’ services, and social insurance administration.1 The cash portions of CalWORKs and SSI/SSP are included in categorical and other cash assistance, hereafter referred to as cash assistance.2 Other welfare includes in-kind aid such as job preparedness assistance and training provided through CalWORKs, child care for welfare recipients and the working poor, foster care, adoption assistance, and community services. Vendor payments mostly include payments to energy providers under the Low-Income Home Energy Assistance Program (LIHEAP). This program provides heating and weatherization support to low-income individuals through cash transfers or subsidies of energy services. The remaining three Census categories consist of administrative functions of nursing facilities, veterans’ financial benefits programs, and unemployment and disability insurance. In 2002, California spent 70 percent more per capita than the rest of the country on social services, including cash assistance and in-kind services as well as administrative support (Table 7.1). Only New York spent more than California, although it provided relatively less aid in cash and more through services. Vendor payments were notably low in mild-climate states such as California and higher in states facing harsher weather conditions such as Illinois and New York. In 2002, aid to individuals and families as either cash or in-kind assistance (other welfare) accounted for 97 percent of social services spending in California. This share ranged from 88 percent in Illinois to 92 percent in Florida, and was 90 percent in the rest of the nation. Vendor payments and administration expenditures represented a very small fraction of total social services spending in California and the rest _____________ 1The Census Bureau includes an eighth category—veterans’ bonuses—which does not apply to California. Since this category constitutes a small share of social services expenditures (less than one-hundredth of a percentage point), it is dropped from the analysis. 2Categorical cash assistance also incorporates some state intergovernmental aid to local governments through the Medicaid program. These transfers are excluded from the analysis that follows (see Chapter 6). 105 Table 7.1 Social Services Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 533 311 164 363 563 157 Cash Assistance 240 62 21 45 132 38 Other Welfare 276 217 130 273 381 104 Vendor Payments Administration 2 16 8 25 49 18 27 23 27 1 15 of the country. Thus, the remainder of this chapter focuses on the programs included in cash assistance and other welfare. California has dedicated more resources per capita to social services than the rest of the country throughout recent history (Figure 7.1). Only New York has provided more total assistance per capita since 1977. In terms of specific types of assistance, California’s cash payments per capita have also exceeded those in the rest of the country, whereas inkind aid has remained close to the average for the rest of the nation until recently. More recently, California has relied less on cash transfers and more on in-kind assistance and services. The rest of the country 600 550 500 450 400 350 300 250 California 200 U.S.–California 150 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 7.1—Social Services Expenditures per Capita, 1977–2002 Expenditures (2002 $) 106 experienced a similar switch in the importance of these two types of aid as a consequence of the implementation of the 1996 federal welfare reform. California differs from other states in our comparison group in that its welfare programs are highly decentralized. This pattern is more pronounced for cash transfers than for other welfare services. Local governments in California spend 71 percent of total social services expenditures, whereas the average local government share of total social services spending is 40 percent in the rest of the country and as low as 9 percent in Illinois. Only New York exhibits a higher level of decentralization than California, with local governments spending 91 percent of total social assistance funds. The Census Bureau’s categorization of social services expenditures is not ideal for understanding individual programs. For example, TANF funds may be classified under categorical cash assistance or other welfare depending on state practices. Further difficulties arise with in-kind assistance and subsidized services. For instance, vouchers for child care are technically noncash social services expenditures (other welfare) from state and federal funds. However, some of these programs are administered by the Department of Education and the California Community Colleges. The Census Bureau thus classifies these expenditures under its education categories. For these reasons, the remainder of this chapter uses data from alternative sources that allow for more meaningful comparisons of programs across states. The analysis of welfare and community services uses reports to the Department of Health and Human Services’ Administration for Children and Families (ACF) on the use of TANF, CCDF, and SSBG funds, as well as data from the U.S. House of Representatives (2004), commonly known as the Green Book. The analysis of SSI/SSP relies on the Social Security Administration’s annual reports. These alternative data sources have some drawbacks. First, they restrict the analysis to expenditures on three major sets of programs: welfare services, community services, and SSI/SSP. Second, they provide only aggregate information about welfare caseload. This hinders an analysis by type of recipient, an important distinction in the 107 implementation of these assistance programs. Third, there are insufficient data on child care programs across states to permit comparisons of child care spending per recipient.3 Fourth, SSA reports on SSI/SSP include federal direct spending as well as state and local expenditures of federal funds. As a result, the sum of social services expenditures from these alternative sources will necessarily exceed Census Bureau state and local government expenditure totals. Despite these caveats, the data sources described above allow a better analysis of social services spending across states beyond what is permitted by the Census Bureau classification scheme. Throughout this chapter, the low-income population is defined as individuals with income below 200 percent of the FPL, $30,040 for a family of three in 2002. Although this benchmark does not necessarily coincide with eligibility criteria in each state for programs analyzed in this chapter, it encompasses the most generous eligibility criterion. Thus, this threshold accounts for all potentially eligible individuals and establishes a common reference for all states. Welfare (TANF/CalWORKs) In California, 32 percent of social assistance expenditures in 2002 were through CalWORKs, the state’s TANF program targeting lowincome families with children. This program provides cash aid, child care subsidies, employment training, and counseling services to reintegrate parents into the labor force. These activities are predominantly financed with federal expenditures and state maintenance of effort funds. Federal funds are based on state expenditures before the federal welfare reform of 1996. This decision locks in California’s TANF federal grant and mandated state expenditures at the highest level in the nation. In addition, California supplements these mandated funds more generously than other states. As a result, California’s expenditures on welfare programs were nearly twice the average for the rest of the nation _____________ 3For a detailed overview of child care expenditures and policies in California, see O’Brien-Strain, Moyé, and Sonenstein (2003), and Marrufo, O’Brien-Strain, and Oliver (2003). 108 in 2002 (Table 7.2). Compared to other large states, only New York had higher welfare expenditures per capita than California, in terms of both total spending and voluntary state spending. To understand the determinants of spending in this area, it is useful to separate expenditures per capita into caseload, or recipients per capita, and expenditures per recipient. We define caseload as the number of persons (adults plus children) receiving assistance under TANF. This definition differs from the commonly used metric of the number of families receiving assistance. Under the latter definition, a family of three (e.g., one parent and two children) receiving cash or services counts as one case, as does a child living with nonparent relatives, even if those relatives do not receive any assistance. Under our definition of caseload, the first family would include three recipients and the second would include only one. This definition is partly due to data limitations. At the same time, it ensures comparability across types of programs and across other chapters in this report. Following the decomposition method outlined in Chapter 2: TANF Expenditures = TANF Recipients  Expenditures Population Population Recipient (7.1) By this metric, California’s average annual expenditure of $4,700 per recipient was below the national average (Table 7.3). California’s yearly expenditures per case exceeded those of Texas by $2,500 but were the Table 7.2 Welfare Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita 159 81 61 79 205 35 State Separate Expenditures per Capita 14 5 1 0 36 1 109 Table 7.3 Components of Welfare Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 159 71 61 79 205 35 Caseloads per Capita 0.034 0.016 0.008 0.011 0.022 0.016 Expenditures per Case ($) 4,718 5,105 8,053 7,264 9,336 2,236 State Separate Expenditures per Case ($) 409 307 166 25 1,637 66 second lowest in our comparison group. New York spent nearly twice as much per case as California in 2002. By contrast, California’s additional expenditures beyond mandated spending were 33 percent higher than the average for the rest of the nation, although just one-quarter of New York’s. The finding of lower expenditures per case may seem at odds with California’s generous maximum allowable benefits per family. For instance, in 2002, California’s maximum cash grant per month for a family of three (one adult and two children, with no income) was $679, the third highest in the nation behind Alaska ($923) and Hawaii ($712). The gap between the maximum benefit per family and the average expenditures per case can arise from a number of factors, including shorter average spells on welfare in California, larger family sizes, or a different allocation of cash assistance versus in-kind benefits. California’s lower average expenditure per case is offset by the state’s higher caseload. In 2002, the number of welfare recipients per capita was 54 percent higher in California than in New York, and more than twice as high as in Texas or the average for the rest of the country (Table 7.3). California’s high caseload in CalWORKs may be a consequence of factors over which the state has little direct control, such as the poverty rate. Alternatively, differences across states may arise from policy choices, including eligibility rules, benefit levels, lifetime participation allowances within the five-year federal limit, and workforce participation requirements. 110 For these reasons, it is useful to separate caseload into the potentially eligible population and the participation rate within this group: TANF Recipients = Potential Eligibles  Recipients Population Population Potential Eligibles (7.2) The first component of this equation captures a state’s demographic and economic characteristics. The second component reflects both state eligibility rules or benefit levels and individual participation decisions (i.e., the take-up rate). For the purpose of this analysis, we broadly define potential eligibles as the number of individuals in a low-income family that includes at least one child. In 2002, the fraction of individuals in low-income families with children was 27 percent higher in California than in the rest of the United States. Among these individuals potentially eligible for welfare benefits, the recipiency rate was also higher, by almost 70 percent (Table 7.4). Of our comparison states, only Texas had more low-income individuals with children per capita than California. However, Texas also had a very low number of recipients within the low-income population with children (58 recipients per 1,000 potentially eligible state residents). Only New York’s recipiency rate among potential eligibles was, like California’s, higher than the rest of the country’s. However, New York had 28 fewer recipients per 1,000 low-income residents than California. Table 7.4 Components of Welfare Caseload, 2002 California U.S.–California Florida Illinois New York Texas Recipients per Capita 0.034 0.016 0.008 0.011 0.022 0.016 Low-Income Individuals with Children per Capita 0.243 0.192 0.196 0.169 0.200 0.273 Recipients per Low-Income Individuals with Children 0.138 0.082 0.038 0.064 0.110 0.058 111 Welfare eligibility policies span a number of dimensions. California’s rules are usually regarded as generous and are certainly part of the explanation for the high number of recipients in the state as a share of the potentially eligible population (MaCurdy, Mancuso, and O’Brien-Strain, 2002). Unfortunately, data limitations prevent us from separating the effect of these rules from individual participation decisions. The following thought experiments are useful to see what would happen if California’s eligibility policies and individual participation decisions mimicked those of other states. If California had the same economic and demographic characteristics as it does now with Florida’s recipiency rate, its caseload would drop from around 1.2 million to fewer than 319,000 recipients and spending would drop to only 28 percent of federally mandated funds. Assuming that total spending remained the same, this drop in recipients would allow California to increase annual expenditures per recipient by $12,000. Similarly, if California’s recipiency rate were like New York’s, it would experience a more modest reduction in its caseload, to just over 920,000 recipients, and potentially would spend only 80 percent of its mandatory funds. It would then hypothetically be able to increase yearly benefits by $1,200 per case or 26 percent. Of course, these changes are merely for illustrative purposes. They do not take into consideration adjustments to income eligibility thresholds and the corresponding take-up rate derived from a change in benefit levels.4 Nonetheless, they give a sense of the size and composition of California’s welfare program. One component of welfare expenditures that draws special attention is child care. California devoted $24 dollars per capita (15% of total welfare expenditures) to child care for welfare families in 2002 (Table 7.5). Of this total, $1 per capita was in addition to federally mandated expenditures. By contrast, the rest of the country devoted $11 per capita _____________ 4Specifically, higher benefits increase the income eligibility threshold. Such a policy would have two effects. First, a mechanical increase in the pool of potential eligible people would increase the number of recipients for a constant number of applicants. The second effect is a behavioral one, whereby people are more likely to apply for and receive assistance. The combination of both effects eventually increases caseload and erodes, at least partially, feasible increases in benefits per case. 112 Table 7.5 Welfare Child Care Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Welfare Child Care Expenditures per Capita 24 11 17 29 5 1 State Additional Welfare Child Care Expenditures per Capita <1 <1 0 0 0 0 to child care, with less than $1 per capita from additional state resources. Within our group of comparison states, Illinois spent the highest amount on child care at $29 dollars per capita. However, none of these resources were in addition to federally mandated spending. New York’s expenditures in this area were one-fifth of California’s, and Texas’s expenditures for welfare child care were 20 times smaller than California’s and the lowest in the nation. California’s welfare expenditures declined after the state’s full implementation of welfare reform in 1998. The most important factor behind this reduction—in California and in the rest of our comparison states—was caseload reduction (Figure 7.2). California’s number of TANF recipients per capita decreased 16 percent between 1998 and 2002, compared to 14 percent in the rest of the nation. This decline was accompanied by a 12 percent increase in benefits per recipient. The growth in benefits per capita in California exceeded the average rate of inflation over this period but remained below the 19 percent increase in benefits per recipient in the rest of the country. California’s reduction in caseload during this period lagged those in Florida (23%), Illinois (34%), and New York (22%) and thus its increase in benefits per case was also lower. 113 Caseload 0.09 0.08 California 0.07 U.S.–California 0.06 0.05 0.04 0.03 0.02 0.01 0 1998 1999 2000 2001 2002 Figure 7.2—TANF Caseload per Capita, 1998–2002 Community Services In 2002, 53 percent of California’s expenditures on social services were devoted to community-based programs supported at least in part with SSBG funds. Several of these programs are aimed exclusively at children: day care, foster care, and adoption assistance. Other programs provide day care for adults, home-based services (including meals for the elderly and supportive services for the disabled), and protective services for adults. In addition to SSBG, these programs are financed with supplemental state resources and federal funds, including CCDF and other matching federal grants.5 In recent years, California has consistently concentrated these funds on child day care, foster care, adoption assistance, day care for adults, home-based services including meals for the elderly, and services for the disabled. Other states in our comparison group provide additional services with these funds such as family planning and counseling, employment services, assistance for transitional and independent living, _____________ 5The latter are particularly important for foster care and adoption assistance (“Title IV” grants). 114 substance abuse, and a variety of other services including health, legal, and transportation assistance. As a whole, California’s per capita expenditures in 2002 on community-based services were 68 percent higher than in the rest of the nation but were 45 percent lower than those of Illinois (Table 7.6). In contrast to its welfare spending, Illinois spent the highest amount per capita on community services in our comparison group. Overall, these services are heavily funded by state-generated funds. California’s ownsource expenditures per capita were almost twice those in the rest of the country in 2002, and these funds represented a greater share of spending (70%) than in any of our comparison states except Illinois (73%). As in the case of welfare, the number of recipients of these services depends in large part on the demographic and economic characteristics of the state. However, the variety of programs in this area precludes using any one definition of the potentially eligible population in each state. For some services, individuals can be eligible even in the absence of children. For others, such as foster care and adoption services, there is no low-income requirement. Because of these distinctions among community services programs, we analyze three major types of services separately: child care, child welfare, and services targeted primarily toward adults and the disabled. In the case of child care, data limitations related to sources of funds and administration of programs prevent a meaningful comparison of Table 7.6 Total Community Services Expenditures per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 237 141 78 344 140 94 State-Fund Expenditures 167 87 35 250 64 55 115 caseloads and expenditures per child across states. It is nevertheless useful to have a reference point. In 2002, California’s expenditures on child care for both welfare and nonwelfare recipients totaled $3.5 billion or $100 per capita. Of this total, almost $1.9 billion in spending was federally mandated and the remainder came from state funds. Child care programs served approximately 611,500 children in that year, averaging expenditures of $5,660 per case (Legislative Analyst’s Office, 2002, 2003). Our second group of community-based services is child welfare. The largest programs in this group are foster care and adoption assistance. The goal of these programs is to improve the living conditions of children and their families or to provide substitutes for parents who have difficulty performing their obligations. California’s per capita expenditures on foster care were twice as much as the average in the rest of the country in 2002, although only half those of Illinois (Table 7.7). Nearly half of these expenditures were covered by state and local funds, closely following the average matching rate required to gain federal funds. In contrast, Illinois funded almost 60 percent of these services through state and local funds, well above the federal matching rate. California’s caseload per capita for foster care was practically the same as in the rest of the country and in other large states in 2002. Higher expenditures on foster care therefore reflect higher expenditures per child (26% higher than in the rest of the nation and about twice those of Florida or Texas). On the other hand, California’s foster Table 7.7 Components of Foster Care Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita ($) 65 33 21 138 51 13 Cases per Capita 0.003 0.002 0.002 0.002 0.002 0.001 Total Expenditures per Case ($) 15,388 12,149 7,059 52,633 16,233 8,761 116 care expenditures per child were within 5 percent of those in New York and much lower (by one-third) of those in Illinois. Expenditures for adoptions assistance show a similar pattern in 2002. In that year, California spent 18 percent more per capita than the rest of the country but 70 percent more per case (Table 7.8). Expenditures per case were four times higher in California than in Texas (the state with the lowest expenditures per case) but were 20 percent lower than in Illinois. The level of expenditures per case in California was high despite having the lowest caseload per capita in our comparison group: 40 percent of New York’s caseload and just over 50 percent of the average for all other states. For both foster care and adoption assistance, California’s higher expenditures per case may result from a number of factors. In foster care, assistance payments to foster families constituted 35 percent of total foster care expenditures in California, compared to 45 percent in the rest of the nation in 2002. Among the states in our comparison group, California devoted the lowest share of foster care expenditures to assistance after Florida (20%) and considerably less than this share in Illinois (83%). Instead, a higher share of expenditures in California went toward administration, case management, and placement services. Expenditures for adoption assistance displayed a similar pattern. Direct adoption payments in California accounted for 79 percent of total spending, compared to 76 percent in the rest of the nation, 62 percent in Florida, and 98 percent in Illinois. Table 7.8 Components of Adoptions Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Expenditures per Capita ($) 13 11 7 31 21 4 Adoption Cases per Capita 0.00044 0.00066 0.00064 0.00084 0.00099 0.00051 Total Expenditures per Adoption Case ($) 28,560 16,725 10,524 36,204 20,871 6,835 117 The third group in our community services analysis encompasses all other expenditures of SSBG funds. In California, this group is primarily made up of protective services for adults, home-based care and meals for the elderly, and developmental services for the disabled. In other states, it may include counseling services, substance abuse treatment, transportation, promotion of independent living for adults, and family planning, among other services. Taken together, California’s per capita expenditures on other community services are almost three times those of the rest of the country and the highest among our comparison states (Table 7.9). Similar to Illinois, state and local own funds are the main source of funding for these services in California: The shares of expenditures financed with these sources were 95 percent in California and 92 percent in Illinois in 2002. As in the rest of this chapter, we analyze the roles of caseload and expenditures per case in determining total expenditures through Eq. (7.1). This decomposition shows that the number of community services beneficiaries relative to the population in California was close to New York’s, exceeding it by only 8 recipients per 1,000 residents in 2002. Caseload per capita in California was 44 percent lower than in the rest of the country, and 48 percent of Florida’s (Table 7.10). It was six times lower than the caseload in Illinois. Average expenditures per recipient in California were almost $6,000 higher than in the rest of the country and twice those in New York, Table 7.9 Expenditures on Other Community Services per Capita, by Source of Funding, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 127 46 22 99 35 30 State Separate Expenditures 121 41 15 92 22 23 118 Table 7.10 Components of Expenditures on Other Community Services per Capita, 2002 California U.S.–California Florida Illinois New York Texas Total Other Expenditures per Capita 127 46 22 99 35 30 Total Other Beneficiaries per Capita 0.018 0.041 0.037 0.110 0.010 0.075 Total Other Total Other State Additional Expenditures Expenditures per Beneficiary per Beneficiary 7,011 6,690 1,117 986 611 409 902 834 3,374 2,144 397 312 despite similar numbers of beneficiaries per capita in these two states. In contrast, Illinois provided less than one-seventh the benefits per recipient as California but had six times as many recipients. These differences are consistent with the fact that California spends nearly half of its community services expenditures other than child welfare on the disabled population. In fact, California’s community services other than child care and child welfare have consistently focused on developmental services for the disabled and home-based care in recent years. By comparison, this group accounts for 23 percent of expenditures from these funds in Florida and 7 percent in Illinois. How much would California spend on each of its beneficiaries if it adopted other states’ policies and had similar take-up rates? The relevant exercise is to consider other states’ choices of own-source expenditures. The results are quite dramatic. If California’s expenditures per recipient from its own funds were the same as those in Texas, it would reduce total expenditures to $12 per capita. This would imply a reduction from $6,700 to $662 in total expenditures per recipient. Adopting New York’s state supplementation per case would shrink California’s expenditures to $45 per capita, or $2,500 per case. There has been a wide disparity in the evolution of community services expenditures across states, mostly driven by the change in 119 additional state resources per case (Table 7.11). This is clearly the case for California, where the growth in expenditures from the state’s own sources has outpaced the growth in its caseload. As a consequence, expenditures per beneficiary increased threefold from 1998 to 2002. During this period, all our comparison states also expanded their coverage in terms of caseload per capita. However, all of them have spent less from their own resources on those additional cases almost every year. Florida and New York stand out in this respect, respectively reducing own-source spending by approximately 63 and 60 percent. By contrast, Illinois doubled its expenditures per case in the same period. Table 7.11 Other Community Services Expenditures per Beneficiary from State-Only Sources, 1998–2002 (dollars) California Florida Illinois New York Texas 1998 2,126 1,117 413 5,425 468 1999 2,392 555 127 1,482 322 2000 2,840 1,999 777 8,352 376 2001 4,265 526 136 1,032 683 2002 6,690 409 834 2,144 312 State Supplementary Program for Supplemental Security Income The SSI program provides cash transfers to low-income elderly, blind, and disabled people. Its goal is to provide a means of subsistence for members of these groups who are not eligible for other welfare programs. The federal government establishes eligibility rules for this program. Benefits consist of three components: a minimum level provided by the federal government, a mandated amount to be provided by states, and an optional state supplement. We cannot distinguish between the two types of state spending in our data. Even though states cannot affect the group of eligible individuals by establishing eligibility rules directly, they affect their numbers through their voluntary supplemental assistance. This happens because the maximum combined SSI/SSP benefit (i.e., federal aid plus state supplement) determines the income eligibility threshold. California’s 120 total supplementation to federal aid represented 38 percent of its total SSI/SSP spending in 2002. The state also contracts with the federal government for the administration and distribution of benefits under SSI. In 2002, California’s SSI/SSP expenditures per capita were almost twice those in the rest of the country, Florida, or Illinois and much higher than those in Texas (Table 7.12). Its spending on this program was similar to that of New York. Both states had higher-than-average state supplementation (including the mandated portion) in 2002. However, California’s supplement was more than twice that of New York’s and ten times beyond the average for the rest of the country. In California and other states, the largest share of expenditures was targeted at the low-income blind and disabled populations. California and New York stand out as states with large caseloads and higher expenditures per case (Table 7.13). California’s high expenditures per case result from its choice of a more generous state supplement. State aid per case in 2002 was highest in California at roughly $2,400 per recipient (38% of total assistance). Without state supplementation, California’s benefits per case ($3,927) would be the lowest within our comparison group, although only 15 percent below New York’s and $100 less per recipient than in the rest of the country. It is also important to notice that California is the only state that has converted food stamps into cash and that it includes these payments as part of the state supplement. This conversion further increases California’s Table 7.12 SSI Expenditures per Capita: Use and State Supplementation, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures Total Expenditures on the Blind Total State Expenditures on the Elderly and Disabled Supplementation 207 55 152 79 108 11 93 8 109 20 89 1 101 9 90 2 178 31 147 29 83 14 69 <1 121 Table 7.13 Components of SSI Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 207 108 109 101 178 83 Cases per Capita 0.032 0.022 0.024 0.020 0.033 0.020 Expenditures per Case ($) 6,366 4,406 4,425 4,474 5,450 4,080 expenditures beyond federal requirements and distinguishes it from other states. To examine the components of caseloads, we can use an analog to Eq. (7.3), with the low-income population restricted to the elderly, blind, and disabled. This analysis shows that even though there are fewer low-income elderly, blind, and disabled individuals per capita in California than in the rest of the country, more people within this pool of potential eligibles receive SSI/SSP assistance (Table 7.14). In particular, even though there were fewer potential eligibles in California in 2002 than in the rest of the country (by 14 per 1,000 residents), California’s recipiency rate was almost twice as high as in the rest of the country. New York had a similar situation, with an almost Table 7.14 Components of SSI Caseload, 2002 California U.S.–California Florida Illinois New York Texas Cases per Capita 0.032 0.022 0.024 0.020 0.033 0.020 Low-Income Elderly and Disabled per Capita 0.072 0.086 0.108 0.071 0.090 0.078 Cases per Low-Income Elderly and Disabled Population 0.453 0.258 0.220 0.281 0.361 0.257 122 identical pool of potential eligibles per capita as in the rest of the country but with a recipiency rate 40 percent higher. Both California and New York report high state supplementation per case. However, it is important to recall that the data do not allow separating state supplemental expenditures into their mandated and discretionary components. Despite the link between eligibility and total benefits, it is illustrative to see the effect of California’s higher recipiency rate on expenditures based on the following thought experiment. If California had Florida’s caseload but kept the same benefit levels, it would spend only $100 per capita on SSI/SSP, a 50 percent reduction from its 2002 levels. Alternatively, if it kept its current recipiency rate but provided Texas’s benefit levels, expenditures per capita would only drop to $132 per capita. Thus, California’s higher expenditures on supplementation to SSI are due to a combination of extending additional aid to more people through higher income thresholds and larger supplementation to those cases. Conclusions Social services represent a modest share of total state and local expenditures in California. However, these expenditures have been high relative to those in comparable states in recent years. The state’s welfare programs are a combination of direct cash payments to individuals and families and services that are either subsidized or provided by government agencies. An intricate combination of state policy choices and federal programs influences expenditures in all states. Overall, California’s high per capita expenditures on welfare indicate that the state has relatively generous assistance provisions for these programs. This generosity is reflected in the state’s broad coverage. From the potentially eligible population, more children and adults receive welfare assistance in California than in other states. As a consequence, average benefits per welfare recipient are spread more thinly. In the case of assistance to the low-income blind, elderly, and disabled population, both caseload and expenditures per case reflect California’s generous policies. 123 California has shifted the focus of community services in recent years. Its coverage of at-risk and disabled populations is on par with the rest of the nation although below that in other states. On the other hand, the benefits provided are high. For child welfare services such as foster care and adoptions, expenditures are heavily allocated to administration and management and less to direct assistance. 124 8. Transportation In 2002, transportation expenditures constituted 7 percent of state and local expenditures in California and 9 percent of expenditures in the rest of the country. This expenditure category includes highways, mass transit, airports, water transport, and parking facilities. Because the last three areas are largely enterprise activities and are relatively small, this chapter focuses on highways and mass transit. This focus leads naturally to the issues of transportation within urban areas, highway congestion, and mass transit as an alternative to the automobile. These issues are particularly relevant in California. Since 1977, the state has spent less per capita on highway construction than the rest of the country. As a result, the state’s highway capacity has grown more slowly than its population. On the other hand, California has invested relatively heavily in alternatives to the automobile. Since 1990, it has spent more per capita on mass transit than the rest of the country. Nevertheless, rates of highway travel in California rival those of Texas and Florida, states whose residents rely heavily on the automobile. High demands on California highways, coupled with their relatively low capacity, cause substantial traffic congestion in the state’s urban areas. According to estimates from the Texas Transportation Institute, the freeways of California’s cities are among the most congested in the nation. Total Expenditures Total transportation expenditures, as noted above, include expenditures on highways, mass transit, airports, water transport, and parking. Highway expenditures consist of the maintenance and construction of public roads, streets, highways, and freeways. Within this broad area, the U.S. Census Bureau also includes expenditures on street lighting, traffic signals, and highway storm drains. The main 125 exclusions are expenditures for the patrolling and policing of streets and highways, which are included in the category of police protection. The Census reports expenditures separately for toll and nontoll highways as part of total transportation expenditures. For California and its four comparison states, expenditures on toll highways are generally 10 percent or less of the expenditures of nontoll highways. The sole exception is New York, where this ratio is 14 percent. Because expenditures on toll highways are relatively small, we have combined them with expenditures on nontoll highways in the analysis that follows. California’s roadway is vast. In 2002, the state had 373,000 lane miles of public roadway.1 Of those lane miles, 57,000 were freeways or other principal arterials, 70 percent of which ran through the state’s urban areas. Mass transit expenditures include the expenditures of public transit agencies and also public subsidies for private transit companies. These subsidies are mainly for privately owned railroad companies providing commuter services. The main exclusions from mass transit are the bus systems transporting public school students. These expenditures are included under the category of elementary and secondary education. Although California has 77 transit agencies, most passenger trips are provided by a few agencies. In 2002, public transit agencies in California provided 1.4 billion passenger trips. The largest agency was the Los Angeles County Metropolitan Transportation Agency (LACMTA), which supplied 32 percent of those trips. The three next largest agencies all served the San Francisco Bay Area—San Francisco Municipal Railway (MUNI), the Bay Area Rapid Transit District (BART), and the Alameda–Contra Costa Transit District (AC Transit). These three agencies provided 28 percent of passenger trips in the state. Nearly 90 percent of passenger trips occurred in either the Los Angeles metropolitan area or the San Francisco Bay Area, with roughly the same number of trips in each. Airport expenditures consist of the operation and construction of public airport facilities. This total does not include the expenditures of _____________ 1Lane miles is the product of road length and the number of lanes. A ten-mile stretch of a four-lane highway has 40 lane miles. 126 airline companies leasing space at public airports. In 2002, California had 28 commercial airports offering scheduled airline service, as well as 250 airports serving general aviation. Two-thirds of airport operating expenditures were generated by two airports: Los Angeles and San Francisco. Expenditures on water transport are due mainly to commercial seaports and municipal marinas. As with airports, the Census Bureau excludes the expenditures of lessees such as shipping companies. In 2002, the Port of Los Angeles accounted for 25 percent of California’s water transport operating expenditures. Six ports—Los Angeles, San Diego, San Francisco, Oakland, Stockton, and Long Beach—accounted for 72 percent of these expenditures. The final component of transportation expenditures is parking. This area includes the construction and operation of public parking lots and garages and the purchase and maintenance of on-street parking meters. It excludes parking facilities connected to other public facilities such as sports stadiums and airports. Transportation expenditures differ dramatically across states. In 2002, California spent about the same amount per capita on transportation as the rest of the country (Table 8.1). However, New York spent two-thirds more than California and more than twice as much as Texas. These differences are largely due to mass transit expenditures. New York spent $465 more per capita on mass transit than Texas, accounting for nearly 90 percent of the difference between these two states in total transportation expenditures. Table 8.1 Expenditures on Transportation per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 611 599 586 706 1,004 482 Mass Transit 187 119 62 202 525 60 Water Highways Airports Transport 328 59 32 411 56 10 403 96 20 451 48 1 370 97 10 347 65 9 Parking 5 4 5 4 3 1 127 California stands out in these comparisons in the way it allocates total transportation expenditures between highways and mass transit. In 2002, highway spending per capita was lower in California than in the rest of the country. It was also lower in California than in each of the four comparison states. At the same time, spending per capita on mass transit was higher in California than in the rest of the country. It was also higher than in Texas and Florida, although less than in Illinois and New York, two states dominated by large cities in which fixed-rail transit systems were established before the automobile emerged as a major mode of commuting. Many transportation activities collect revenue from their users. Some highways have tolls for vehicles, buses and subways collect fares from their passengers, airports and seaports impose rents and fees on private companies using their facilities, and parking facilities charge parking rates. These charges defray some public expenditures on these activities. To better measure the effect of transportation activities on public funds, Table 8.2 presents net expenditures, defined as total expenditures minus user fees, charges, and tolls. The user fees, charges, and tolls netted out against highway expenditures in Table 8.2 do not include a variety of taxes and license fees closely related to highway use. These taxes and fees include state gasoline taxes, vehicle license fees, mileage and weight fees on motor carriers, and driver’s license fees. In total, these taxes and fees amounted to just over $153 per capita in California in 2002. Sixty-three Table 8.2 Net Expenditures on Transportation per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 478 484 430 551 615 405 Mass Transit 151 88 51 143 350 54 Highways 309 381 359 417 259 322 Airports 11 14 16 –8 8 28 Water Transport 8 2 5 0 3 1 Parking –2 –1 –1 –1 –5 0 128 percent of this total was the revenue from the state gasoline tax. In other states, highway-related taxes and fees amounted to $186 per capita. When expenditures are netted out against revenues, the smallest three expenditure areas—airports, water transport, and parking— become an even less significant part of transportation spending. For all states except Texas, airport revenues cover at least 75 percent of total expenditures. Ports and marinas also cover most of their expenditures. Parking activities consistently take in more revenue than their expenditures. Because of their small net effect on public funds, these areas are not analyzed further in this chapter. In 2002, net transportation expenditures in California were approximately equal in per capita terms to expenditures in the rest of the country. As Figure 8.1 shows, expenditures in 2002 were somewhat anomalous, although perhaps also the culmination of a long-term trend. From 1977 through 2000, net expenditures per capita were lower in California than in the rest of the country. In the 1970s and 1980s, expenditures per capita were consistently 20 to 30 percent lower in California than in the rest of the country. That relationship began to change in the 1990s as real expenditures per capita grew in California 600 500 400 300 200 California 100 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.1—Net Expenditures on Transportation per Capita, 1977–2002 129 Expenditures (2002 $) relative to the rest of the country, reaching the level of the rest of the country in 2002. Mass Transit Mass transit is an important reason why transportation expenditures per capita differ significantly across states. The five largest states fall into three broad classes (Table 8.3). Florida and Texas spend relatively little on mass transit. At the other extreme, New York spends seven times as much per capita as either Texas or Florida. California and Illinois lie between those two extremes with net expenditures per capita three times higher than in Florida and Texas but half of those in New York. Subsidies to private transit companies are small in all states. Public transit is most practical in urban areas, and thus differences in expenditures per capita across states are at least partly due to differences in the extent and density of a state’s urban areas. California is one of the nation’s most urbanized states. Eighty-eight percent of California residents live in urbanized areas as defined by the Census Bureau (areas with populations of at least 50,000 and densities of at least 1,000 persons per square mile), a higher percentage than in any of the four comparison states (Table 8.4). California’s urbanized areas are also more densely populated than those of all our comparison states except New York. In 2002, capital expenditures were a significant fraction of the total expenditures of public transit agencies (Table 8.5). In New York, capital Table 8.3 Net Expenditures on Mass Transit per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Net Expenditures 151 88 51 146 350 54 Total Expenditures on Public Transit 181 117 62 198 514 60 Fare Revenue from Public Transit 36 30 10 56 175 7 Subsidies to Private Transit 6 2 0 4 11 0 130 Table 8.4 Expenditures on Mass Transit and Urban Density, 2002 California U.S.–California Florida Illinois New York Texas 2002 Expenditures per Capita on Public Transit ($) 181 117 62 198 514 60 Percentage of Population in Urbanized Area 88 66 84 78 82 71 Residents per Square Mile in Urbanized Area 4,571 2,480 2,451 3,390 4,732 2,697 Table 8.5 Expenditures on Public Mass Transit per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 181 117 62 198 514 60 Capital Expenditures 38 40 22 50 172 22 Current Expenditures 142 77 39 148 342 38 expenditures represented half of current expenditures; in Illinois, they accounted for about one-third. In comparison, California’s capital expenditures were modest—26 percent of current expenditures—and slightly less than the average in the rest of the country. However, because capital expenditures vary considerably from year to year, it is more informative to focus on long-run trends. From a long-run perspective, it becomes clear that California has been steadily building its transit capacity (Figure 8.2). In the 1970s and 1980s, expenditures per capita were lower in California than in the rest of the country. That changed in the 1990s, however, with capital expenditures per capita substantially higher in California than in the rest of the country. 131 Expenditures (2002 $) 100 California 75 U.S.–California 50 25 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.2—Capital Expenditures on Mass Transit per Capita, 1977–2002 Current expenditures have followed the same general trend (Figure 8.3). Throughout the 1970s and 1980s, current expenditures per capita on mass transit were slightly higher in California than in the rest of the nation. Throughout this period, real expenditures per capita were virtually unchanged from year to year in both California and the rest of 200 California 150 U.S.–California 100 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.3—Current Expenditures on Mass Transit per Capita, 1977–2002 132 Expenditures (2002 $) the nation. In the 1990s, however, real current expenditures per capita rose sharply in California while remaining flat in the rest of the country. By 2002, current expenditures per capita were twice as high in California as in the rest of the country. To further analyze current expenditures, we turn to the National Transit Database (NTD) compiled by the Federal Transit Administration. The database includes information about ridership and expenditures for all transit agencies receiving federal support under the Urbanized Area Formula Program. Although the NTD uses slightly different definitions of transit expenditures and does not include all transit agencies, in the aggregate its data match the Census of Governments figures very well. The operating expenditures of transit agencies included in the NTD in 2002 were 93 percent of the current expenditures reported for transit agencies in the Census of Governments for that year. For California, however, that figure was only 75 percent. Three-quarters of the gap was due to differences in current expenditures for the Los Angeles Metropolitan Transit Authority (LAMTA). NTD includes only those expenses directed to the operation of the authority’s public transit system. The Census also includes expenses associated with the authority’s role as a regional transportation planning and coordinating agency. We thus believe that the NTD is a reliable source of information about transit operating costs for both California and the rest of the country. In terms of both costs and service, rail transit differs considerably from bus transit. Rail transit primarily services commuters traveling from locations throughout an urban area to its central city. Although bus transit may play this role too, it also provides transit between noncentral-city locations. As a consequence, bus trips are shorter on average (Table 8.6). States also differ considerably in the extent to which they depend on rail transit versus other forms of transit. In California, 21 percent of mass transit trips in 2002 were by rail (Table 8.6). On the other hand, in New York, 60 percent of mass transit trips were by rail, mostly as a result of the extensive commuter rail and subway system serving New York City. At the other extreme are Texas and Florida, with rail accounting for less than 7 percent of transit trips. 133 Table 8.6 Passenger Miles per Trip, by Transit Mode, 2002 California U.S.–California Florida Illinois New York Texas Miles per Trip Rail Nonrail Total 7.5 3.6 6.9 3.8 7.8 4.8 10.1 3.1 6.5 2.5 6.5 5.0 4.4 5.2 5.0 6.0 4.9 5.1 Percentage of Trips by Rail 21.1 43.0 6.9 41.3 62.0 5.5 Rail and bus transit involve the classic tradeoff between fixed cost and variable cost. Rail transit requires high capital costs, particularly for underground systems. Once those construction costs have been incurred, however, the operating cost of transporting a passenger for a given distance is generally cheaper by rail than by bus. In Table 8.7, operating cost per passenger mile is the total miles traveled by all passengers during a year divided by operating costs during that year. For California, Illinois, and New York, operating cost per mile for rail transit is 45 to 60 percent of the cost of nonrail transit. Within commuting modes, operating costs per mile do not differ significantly among states. For rail transit, cost per mile in California is slightly higher than the average in the rest of the nation. For nonrail transit, the cost per mile in California is slightly lower than the average for all other states. Table 8.7 Operating Cost per Passenger Mile, by Mode, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Rail Nonrail Total 0.41 0.67 0.58 0.36 0.72 0.51 0.57 0.67 0.66 0.32 0.71 0.44 0.37 0.79 0.45 0.69 0.64 0.64 134 In the case of both rail and nonrail transit, the largest component of operating cost is employee compensation (Table 8.8). In California, employee compensation represents 70 percent of rail operating costs and 72 percent of nonrail operating costs. These percentages are similar in other states. Moreover, within a transit mode, employee compensation per passenger mile does not differ much across states. Employee compensation per passenger mile can be decomposed into average hourly employee compensation and employee hours per passenger mile. This decomposition follows the general form of Eq. (2.3) in Chapter 2, with passenger miles as the unit of service, employee hours as the unit of input, and compensation per hour as the price of a unit of that input. In the present case, this equation is: Employee Compensation = Employee Hours Passenger Mile Passenger Mile  Employee Compensation Employee Hour (8.1) Table 8.8 Components of Operating Cost per Passenger Mile, by Mode, 2002 (dollars) California U.S.–California Florida Illinois New York Texas California U.S.–California Florida Illinois New York Texas Employee Fuel and Compensation Utilities Rail 0.29 0.05 0.27 0.05 0.30 0.12 0.22 0.05 0.30 0.05 0.31 0.09 Nonrail 0.48 0.07 0.51 0.09 0.43 0.08 0.50 0.08 0.62 0.09 0.46 0.08 Other 0.06 0.04 0.15 0.04 0.03 0.29 0.12 0.12 0.16 0.13 0.09 0.10 Total 0.41 0.36 0.57 0.32 0.37 0.69 0.67 0.72 0.67 0.71 0.79 0.64 135 Table 8.9 presents the components of this equation. For both rail and nonrail modes, average compensation per hour is slightly higher in California than in the rest of the nation. However, for nonrail transit, employee hours per passenger mile are lower in California than the average for all other states, resulting in a lower compensation per passenger mile than in other states. California’s higher average compensation for transit workers is consistent with general salary differences between California and other states (Table 8.10). In California and in other states, salary ranged from 63 to 66 percent of average compensation. Following the procedures described in Chapter 2, average salary per hour in California and other states was adjusted for location differences using the average salary of Zone 3 workers. After that adjustment, the average salary of transit workers in California is slightly lower than that of transit workers in other states. In terms of operating costs, California’s public transit system appears to be relatively efficient. Despite higher employee compensation Table 8.9 Components of Employee Compensation per Passenger Mile, 2002 California U.S.–California Florida Illinois New York Texas California U.S.–California Florida Illinois New York Texas Employee Average Compensation Employee per Passenger Compensation Mile ($) ($/hr) Rail 0.29 45 0.27 44 0.30 34 0.22 39 0.30 50 0.31 29 Nonrail 0.48 34 0.51 30 0.43 26 0.50 31 0.62 36 0.46 27 Employee Hours per Passenger Mile ($) 0.007 0.006 0.009 0.006 0.006 0.011 0.014 0.017 0.017 0.016 0.017 0.017 136 Table 8.10 Mass Transit Salary Adjusted for Location, 2002 (dollars/hour) California U.S.–California Florida Illinois New York Texas Rail Unadjusted Average Salary Average Salary Adjusted for Location 29 25 28 28 23 25 24 23 31 27 19 20 Nonrail Average Salary Unadjusted Adjusted for Average Salary Location 21 19 20 20 18 19 20 19 23 20 18 19 necessitated by California’s labor market, operating costs per passenger mile are comparable to costs per mile in other states. For nonrail transit in particular, these relatively low costs are due to relatively low employee hours per passenger mile, which suggests that California’s bus systems are operating close to full capacity. Highways The dominant mode of commuting is the automobile. The public sector facilitates this essentially private mode of transportation by constructing and maintaining the highways that make automobile travel possible. Revenue from users is not as significant for highways as for public mass transit (Table 8.11). User revenue includes revenue from toll highways and various reimbursements that cities, counties, and states may receive for constructing and maintaining streets. These reimbursements also include maintenance assessments for street lighting and snow plowing. User revenue for highways does not include fuel taxes although these funds may be earmarked for highway construction. In fact, California governments received more user revenue from nontoll highways ($11 per capita) than from toll highways ($8 per capita) in 2002. In every other state, toll revenues were much higher than reimbursements. 137 Table 8.11 Expenditures on Highways per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Net Expenditures 309 381 359 417 259 322 Total Expenditures 328 411 403 451 370 347 User Revenue 19 30 44 34 111 24 California stood out in a more important regard. Although its current expenditures per capita were roughly equivalent to those of other states, its capital expenditures per capita were much lower (Table 8.12). As Figure 8.4 shows, 2002 was not an exception. Since 1977, capital expenditures per capita in California have been between 40 and 60 percent of expenditures per capita in the rest of the country. In fact, as a percentage of expenditures in the rest of the country, 2002 was nearly a high point, with spending per capita in California at 64 percent of the spending in the rest of the country. Only in 1994 did California reach a higher percentage. In that year, the combination of a dip in real spending per capita in the rest of the country and a small increase in real spending per capita in California produced a ratio of 65 percent. The cumulative effect of California’s relatively low capital spending has been a steady decline of highway capacity relative to population (Figure 8.5). In 1980, California had approximately 2,000 lane miles Table 8.12 Current and Capital Expenditures on Highways per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total Expenditures 328 411 403 451 370 347 Current Expenditures 174 171 129 180 154 128 Capital Expenditures 154 240 273 271 215 218 138 Expenditures (2002 $) 250 200 150 100 50 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 8.4—Capital Expenditures on Highways per Capita, 1977–2002 3,000 Lane miles 2,500 2,000 California U.S.–California 1,500 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Figure 8.5—Freeway and Other Principal Arterial Lane Miles per Million Residents, 1980–2002 of freeway and principal arterials for every million residents. That ratio has fallen ever since, resulting in 1,600 lane miles per million residents in 2002. In contrast, the rest of the nation had 2,500 lane miles per million 139 people in 1980 and maintained that ratio through the 1980s and 1990s. By 2002, lane miles per capita in California were only 65 percent of the ratio in the rest of the country. This decline does not square with California’s reputation as a pioneer of the modern freeway. In fact, in the 1950s and 1960s, the state was spending more per capita on highway construction than the rest of the country (Table 8.13). Real spending per capita was also rising each year. This pattern changed in 1972, when real spending per capita fell to the level of 1957. It was also less than in all of our comparison states except New York. California’s real spending per capita declined even further in 1977 to a level less than half of 1957. As Figure 8.4 shows, real spending stayed at about this level throughout the 1980s. In highway construction as in all public services, the need for capital spending is more closely related to the growth in population than to the level of population. In states with low population growth, a low level of spending per capita can maintain the existing ratio of highway capacity to population. New York is a good example. From 1980 to 2002, New York’s population increased by 9 percent, its freeway lane miles increased by 7 percent, and it maintained its ratio of lane miles to population (Table 8.14). New York’s low level of capital spending per person was consistent with its low population growth rate. From 1980 to 2002, California’s lane miles of freeways and arterials increased by 19 percent, an increase larger than New York’s but smaller than that in the rest of the nation or in any of our comparison states. In contrast, its population increased by 48 percent during this period, an increase much larger than in New York, Illinois, and all states other than Table 8.13 Capital Expenditures on Highways per Capita, 1957–1977 (2002 $) California U.S.–California Florida Illinois New York Texas 1957 221 192 220 205 149 196 1962 248 221 184 210 206 242 1967 275 256 229 187 203 305 1972 224 257 249 279 203 286 1977 89 179 153 229 94 151 140 Table 8.14 Freeways and Other Principal Arterial Growth in Lane Miles and Population, 1980–2002 California U.S.–California Florida Illinois New York Texas Lane Miles per Million People 1980 2002 2,038 2,484 2,151 1,287 1,260 3,496 1,639 2,528 1,859 1,955 1,236 2,759 Increase, 1980–2002 (%) Population Lane Miles 48 19 25 27 71 48 10 67 97 53 20 California and almost as large as in Florida and Texas. This combination of a relatively modest increase in lane miles and a rapid increase in population decreased the state’s lane miles per capita by 20 percent. Its ratio of lane miles to population is now the lowest of any of our comparison states except New York. Illinois has followed a different course. In 1980, its ratio of lane miles to population was as low as New York’s. Since then, it has increased its lane miles by 67 percent, but its population has grown by only 10 percent. It now has more lane miles per capita than New York, California, or Florida. The ratio of lane miles to population is a proxy for highway capacity relative to the demand for highway travel. However, population is an imperfect measure of the demand for highway travel. Factors such as the locations of jobs and residences can also have a significant effect on demand. In turn, highway routes can affect settlement patterns, creating a feedback between highway supply and demand. This feedback makes highway demand a complex issue, but there is no doubt that demands differ among states. For example, in 2002, residents of New York averaged a little over 3,000 vehicle miles per person, a little more than half of the average vehicle miles traveled by Californians (Table 8.15). Thus, although residents of California had 30 percent more lane miles per capita than residents of New York, freeways and arterials were used 141 Table 8.15 Traffic on Freeways and Other Principal Arterials, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles per Million Persons 1,639 2,528 1,859 1,955 1,236 2,759 Annual Vehicle Annual Vehicle Miles per Miles per Capita Lane Mile 5,938 3,622,515 5,283 2,090,087 5,191 2,791,882 4,532 2,318,457 3,308 2,676,524 6,010 2,178,663 more intensively in California. The ratio of vehicle miles traveled to lane miles was more than 30 percent higher in California than in New York. This high demand for vehicle travel is California’s main transportation challenge. The state has invested in mass transit, and the use of mass transit is higher in California than in many other states. Nevertheless, the ratio of vehicle miles to population is nearly as high in California as in Texas, a state that has invested relatively little in mass transit. Yet, despite this high demand, California has relatively few lane miles of freeways and arterials. The result is very high levels of travel per freeway mile, almost 30 percent higher than in any of California’s four comparison states. Urban highways are generally used more intensively than rural highways. Thus, combining miles traveled and lane miles in urban and rural areas may be misleading, particularly when comparing states with different percentages of urban and rural highways. For example, in Texas, 72 percent of miles traveled are in urban areas, but only 51 percent of freeway lane miles are in those areas (Table 8.16). In contrast, 82 percent of California’s vehicle miles occur in urban areas, and 71 percent of its lane miles are in urban areas, a closer match between capacity and demand. That closer match notwithstanding, freeways and arterials in both urban and rural areas are used more intensively in California than in any other comparison state. In urban areas, travel per lane mile is at least 18 142 Table 8.16 Traffic on Freeways and Other Principal Arterials in Urban and Rural Areas, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles in Urban Areas (%) 71 41 54 52 64 51 Vehicle Miles in Urban Areas (%) 82 63 69 73 80 72 Annual Vehicle Miles per Lane Mile Urban Areas Rural Areas 4,209,465 3,148,986 3,560,831 3,282,720 3,338,413 3,088,842 2,192,907 1,339,106 1,894,649 1,283,415 1,497,768 1,242,009 percent higher in California than in any of our comparison states. In rural areas, travel per lane mile is at least 15 percent higher. The heavy usage of California highways surely affects the cost of maintaining them. Some roads are owned and maintained by state governments (Caltrans in the case of California), and the remainder are owned and maintained by local governments. In either case, the cost of maintenance is reflected in current highway expenditures. Table 8.17 shows the division of current expenditures between state highway departments and local governments in 2002. With the exception of Texas, state highway departments spent more than local governments on highways. However, most roads are owned and maintained by local governments. In California, for example, the state maintained 50,000 lane miles of roadway and localities maintained 320,000 lane miles. Table 8.17 Current Highway Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas All Roads and Highways 174 171 129 180 154 128 Local Roads and Highways 77 76 61 63 42 65 State Roads and Highways 97 95 69 118 112 64 143 Current highway expenditures per capita can be decomposed into lane miles per capita and expenditures per lane mile. This decomposition is an application of Eq. (2.1) of Chapter 2, although it has a slightly different interpretation from that equation because lane miles play the role of caseload. That unusual interpretation notwithstanding, the equation is: Expenditures = Lane Miles  Expenditures Population Population Lane Mile (8.2) The components of this equation are listed in Table 8.18 for local roads and highways. In 2002, California had less than half of the lane miles per capita of the rest of the country. It also spent more than twice as much per lane mile as the rest of the country. The same general relationship holds between California and each of our four comparison states: fewer lane miles per capita and higher current expenditures per lane mile. A similar pattern holds for current expenditures on state-owned highways (Table 8.19). California had the fewest lane miles per capita of any of our comparison states. It also spent the most per lane mile. In all states, current expenditures per lane mile were substantially higher for state roads and highways than for local roads and highways, a difference reflecting the heavier use of state highways. The heavy use of California highways may partly explain its high current expenditures per lane mile, but another factor is California’s Table 8.18 Components of Current Highway Expenditures per Capita on Local Roads and Highways, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 77 76 61 63 42 65 Lane Miles per 1,000 Residents 9.2 24.3 13.2 19.7 10.6 20.8 Expenditures per Lane Mile ($) 8,376 3,133 4,619 3,200 3,955 3,107 144 Table 8.19 Components of Current Highway Expenditures per Capita on State Roads and Highways, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 97 95 69 118 112 64 Lane Miles per 1,000 Residents 1.4 7.0 2.4 3.3 2.0 8.7 Expenditures per Lane Mile ($) 67,023 13,514 28,173 35,395 56,348 7,311 relatively high payroll expenditures per highway employee. In 2002, California highway employees were paid 52 percent more than equivalent employees in other states (Table 8.20). They were also paid more than equivalent employees in any of the four comparison states. Following the procedure described in Chapter 2, payroll for employee is adjusted for location differences using the average salary of Zone 3 workers. Even after that adjustment, highway employees in California were paid considerably more than similar employees in other states. The Census Bureau does not distinguish between payroll expenditures for capital projects and payroll expenditures for maintenance and other current activities. It seems reasonable to assume, however, that payroll expenditures are a higher proportion of current Table 8.20 Annual Highway Payroll Adjusted for Location, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Payroll per Employee 56,163 36,812 34,831 43,832 45,406 31,856 Payroll per Employee Adjusted for Location 49,280 37,486 38,054 42,474 39,642 34,493 145 expenditures than they are of capital expenditures. With that assumption, the percentage of current highway expenditures allocated to payroll must be larger than total payroll expenditures as a percentage of total expenditures (capital plus current). On the other hand, the percentage of current expenditures allocated to payroll cannot be larger than total payroll expenditures as a percentage of current expenditures. For California, that lower limit (total payroll expenditures as a percentage of total expenditures) is 22 percent, and the upper limit (total payroll expenditures as a percentage of current expenditures) is 41 percent. With current payroll expenditures of that magnitude, California’s substantially higher payroll expenditures per employee is another factor in addition to heavy use explaining its higher current expenditures per lane mile. Outcomes Highway use and capacity is a delicate balance. It is a waste of public funds to add lanes to an uncongested freeway. However, as traffic on a highway grows, at some point travel speeds begin to slow, and the commute from home to work becomes more time-consuming. California freeways and arterials are used intensively, but does the level of use significantly extend the commuting time of the average California resident? Traffic delay is certainly significant on the freeways surrounding California cities. On 40 percent of the state’s urban freeways, traffic moved at 35 miles per hour or less during peak periods of a typical workday in 1998 (Legislative Analyst’s Office, 2000). Traffic delay amounted to over 400,000 hours per day. According to estimates from the Texas Transportation Institute (Schrank and Lomax, 2005), traffic delay in California also tends to be higher than delay in other states. It is important to note that the institute does not directly measure the speed at which cars actually travel on urban highways. Instead, it estimates travel delay with a numerical algorithm based on the ratio of traffic volume to the supply of roadway. Using observations of roadways in many situations, the institute has estimated a relationship between that ratio and the average speed that cars travel. That relationship is flat until the ratio reaches a level at 146 which congestion starts to occur. As the ratio increases beyond that point, travel speed falls steeply. The institute is able to use this relationship and data on traffic volume to estimate average speeds on roadway segments within urban areas. It then compares these estimates to speeds if there were no congestion. The difference is then aggregated to the level of the urban area for 85 urban areas throughout the country. The institute makes other assumptions in its algorithm. For example, it assumes a relationship between traffic volume and the percentage of travel that is congested.2 All of these assumptions can be criticized and surely apply better in some areas than in others. On the other hand, the institute is the only source of information about traffic delay that applies a consistent set of measurements and estimates to a large number of American cities. Its methods are transparent and its data readily available. Appendix Table A.1 lists the urban areas included in California and our comparison states that are included in the institute’s study. Nine of the institute’s 85 urban areas are in California. Residents of these areas tend to use mass transit more than residents of Texas and Florida cities but not as much as residents of Illinois and New York cities (Table 8.21). Despite their more frequent mass transit use, residents of California cities travel as many vehicle miles on average as residents of Florida and Texas cities. Californians have fewer lane miles per resident, however, so the ratio of vehicle miles to lane miles is higher in California cities than in Florida and Texas cities. It is also higher than in Illinois and New York cities. Cities in those two states have fewer lane miles per capita than California cities, but their residents travel less by automobile, presumably because of the extensive mass transit systems in Chicago and New York. According to the institute’s estimates, California’s high ratio of highway usage to capacity leads to substantially slower speeds and extensive travel delay. For the nine California cities, annual delay per _____________ 2For example, the volume of traffic per lane mile in Los Angeles is 52 percent higher than in Jacksonville, Florida. The institute’s algorithm assumes that this difference in volume increases the percentage of traffic incurring congestion from 42 percent to 50 percent. 147 Table 8.21 Road Congestion and Delay in 85 Urban Areas, 2002 California U.S.–California Florida Illinois New York Texas Lane Miles of Freeways and Daily Annual Annual Mass Daily Vehicle Arterials Vehicle Hours of Transit Trips Miles Traveled per Million Miles per Delay per Capita per Capita Residents Lane Mile per Capita 51 16 1,388 11,803 39 59 14 1,327 10,293 22 18 15 1,584 9,355 25 73 12 1,033 11,362 29 169 10 923 10,544 27 24 17 1,576 10,582 30 capita is estimated to be 40 hours, one-third higher than in Texas and Florida cities. Annual delay in California cities is also significantly higher than in New York and Illinois cities, although those comparisons may be misleading. The institute estimates only highway delays. It does not estimate the lengthy travel time that sometimes accompanies travel by mass transit. Traffic delay can vary considerably across cities within the same state. For urban areas with more than one million residents, Figure 8.6 plots hours of delay per capita against the ratio of miles traveled to lane miles. Los Angeles and San Jose experienced the most delay. San Francisco ranked with Dallas and Houston. New York and Chicago have less delay, but, to repeat, the estimates do not include travel time for mass transit commuters. Conclusions The current congestion on California’s freeways can be traced to policy decisions made during the 1970s. During that period, California began to invest less than other states on highways. This trend has continued through the 1980s and 1990s. California’s lower investment in highways and its rapid population growth have led to a decline in highway capacity relative to population. 148 60 • San Jose •50 Los Angeles Hours 40 •Dallas Riverside– • San Francisco • Houston • •30 San Bernardino Orlando •Tampa Chicago • San Diego • •New York • •20 San Antonio Sacramento 10 0 6,000 • Buffalo 8,000 • Miami • Jacksonville 10,000 12,000 14,000 16,000 Daily vehicle miles traveled per lane mile Figure 8.6—Annual Hours of Delay per Capita in the Largest Urban Areas, 2002 Over the same period of time, California has steadily increased real spending per capita on mass transit. Perhaps as a result, mass transit use is somewhat higher in California than in other states. However, despite this mass transit use, automobile travel is high in California. The average Californian travels as many automobile miles as the average Texan or Floridian. This high demand for automobile travel has run up against lower and lower highway capacity per resident. The result is relatively high levels of congestion and traffic delay. The heavy use of California highways may also partly explain its relatively high current expenditures per lane mile. 149 9. Public Safety California spent $27.2 billion on public safety in 2002. This represented 9 percent of total state and local government expenditures, compared to 7 percent in the rest of the country. This chapter analyzes expenditure differences between California and other states in this area. It finds that California’s total expenditures per capita on public safety are higher than in the rest of the country. Underlying conditions are partially responsible for the state’s higher expenditures, but service provision costs—in particular labor costs—also contribute importantly to California’s spending levels. Staffing levels in police and fire are also lower in California than in comparable states. Expenditures on inspection and regulation activities are considerably higher in California than in the rest of the country. Total Expenditures The U.S. Census Bureau classifies public safety expenditures under four functions: corrections, fire protection, police protection, and protective inspection and regulation. Spending on corrections includes the operation and administration of state prisons, local jails, youth corrections, and probation and parole programs. Fire protection includes only the activities of local governments (cities, counties, and special fire districts). The Census Bureau classifies state fire operations, which focus on forest fire protection, under natural resources (see Chapter 10). Police protection covers all police departments and law enforcement agencies with the exception of special forces, park rangers, fish and game wardens, and enforcement or police who serve specific purposes such as judicial services or correctional functions. It includes activities by state governments as well as a variety of local jurisdictions. The protective inspection and regulation category includes expenditures related to a variety of regulatory activities of state and local governments. Some of these activities include the regulation of utilities, 151 vehicle registration, and consumer protection at the state level, and construction regulation and building inspection at the local level. The components of this category are likely to differ from state to state, depending on the scope of their regulatory needs and activities. Thus, we keep comparisons between California and other states to a minimum. As in other areas, California’s per capita expenditures on public safety were higher than in the rest of the country in 2002 (Table 9.1). Differences were pronounced for corrections and protective inspection and regulation. Fire and police protection expenditures in California were lower than in New York and did not differ much from those in Florida or Illinois. With the exception of fire protection, state and local authorities share public safety responsibilities. Local governments spent roughly 66 percent of the total for this category in California in 2002, the same as in the rest of the nation but less than the local share of public safety expenditures in New York (77%) or Florida (71%). Compared to the rest of the country, California’s public safety expenditures have been high through most of recent history (Figure 9.1). Only New York spent as much per capita during the years 1977–2002. Nationwide, public safety expenditures have been increasing, but California experienced a relative acceleration in the mid-1980s. The fall and rise in total expenditures in the late 1980s resulted mainly from fluctuations in corrections capital expenditures and fluctuations in fire protection spending at the fire district level. The increase in public safety Table 9.1 Public Safety Expenditures per Capita, by Type of Service, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 778 512 610 538 761 464 Corrections 257 181 202 149 244 202 Fire Protection 119 86 111 108 128 69 Police Protection 290 215 263 256 351 191 Protective Inspection and Regulation 111 31 34 24 38 21 152 Expenditures (2002 $) 900 800 700 600 500 400 California 300 U.S.–California 200 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.1—Public Safety Expenditures per Capita, 1977–2002 expenditures in 2002 was also a consequence of several factors: substantially higher inspection and regulation expenditures, growth in nonpayroll costs in corrections and police protection, and increases in payroll costs for fire protection. Relative to total public safety expenditures, corrections have increased in importance both in California and in the rest of the nation. In 1977, 23 percent of public safety expenditures were devoted to corrections in California, compared to 20 percent nationwide. By 2002, that share was 33 percent in California and 35 percent nationwide. Over the same period, California’s police protection expenditures fell from 46 to 42 percent of total public safety spending, whereas the share of public safety expenditures allocated to fire protection fell from 21 to 15 percent. Similarly, in the rest of the nation, shares allocated to police and fire dropped from 21 to 17 percent and 51 to 42 percent, respectively. Corrections State and local expenditures per capita on corrections were higher in California than in the rest of the nation in 2002 (Table 9.2). In California as in other states, state and local authorities share the operational and financial responsibilities of running correctional 153 Table 9.2 Corrections Expenditures per Capita, by Type, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 257 181 202 149 244 202 Current 247 169 193 138 225 190 Capital 10 12 8 12 19 12 institutions. Local jurisdictions run jails and detention centers, and the state administers adult prisons. Both levels of government run juvenile centers and parole and probation programs. In California, the state government spent 59 percent of total corrections expenditures in 2002. The remaining expenditures were fairly evenly distributed across cities and counties. The growth in per capita expenditures on corrections in California outpaced that of the rest of the nation in the mid-1980s (Figure 9.2). Some of our comparison states, notably New York and Florida, had similar experiences. Texas’s expenditures rose above the national average after a particularly strong increase in the 1990s. Fluctuations in California’s spending on corrections between 1977 and 2002 stemmed mainly from capital expenditures. However, their average share of total corrections spending over this period was only 10 percent. Thus, current expenditures explain most of the long-run growth in total expenditures. A basic way to analyze the evolution of total expenditures is through the following decomposition: Expenditures = Inmates  Expenditures Population Population Inmate (9.1) A natural question to ask is whether more sentencing and longer confinement periods explain the growth in per capita corrections spending in California (Mackenzie, 2001; Legislative Analyst’s Office, 2004). The growth of the inmate population—defined as people in prison, youth corrections, probation, and parole—throughout the 154 Expenditures (2002 $) 300 250 200 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.2—Corrections Expenditures per Capita, 1977–2002 country is consistent with this conjecture.1 Nationwide, the number of inmates increased from 2.4 to 4.2 million over the 1983–1990 period, a 75 percent increase. From 1991 to 2002, it increased from 4.4 million to 6.6 million, or by 50 percent. Although California experienced a similar trend, its caseload, or number of inmates per capita, does not differ substantially from that of most other large states (Table 9.3). Texas holds the largest inmate population in our comparison group, but it also has the lowest expenditures per inmate. California’s inmate-to-population ratio was at or above the average in the rest of the nation until the early 1990s, when it experienced a decline. Even though the state’s inmate population increased slightly after the approval of Proposition 184 in 1994, it has remained lower than in the rest of the country in recent years (Figure 9.3).2 _____________ 1We use this definition of inmate population because of limited data on jail inmates for all states during our study period. In California, jail inmates made up only 13 percent of the total correctional population, similar to the nation as a whole at 11 percent, as of 2001. 2Proposition 184, also known as the “three strikes and you’re out” initiative, increased penalties for felons convicted of a third crime. In particular, if a criminal has had one previous serious or violent felony conviction, the mandatory sentence for a 155 Table 9.3 Components of Corrections Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Capita ($) 257 181 202 149 244 202 Inmates per Capita 0.018 0.021 0.023 0.017 0.017 0.032 Expenditures per Inmate ($) 14,222 8,655 8,619 8,555 14,521 6,291 0.025 0.020 Inmates per capita 0.015 0.010 0.005 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.3—Inmates Relative to Population, 1977–2002 Consequently, California’s per capita expenditures are high because of above-average spending per inmate. Since expenditures per inmate drive most of California’s spending on corrections, we will analyze their components in the following way: ______________________________________________________________ second such conviction is doubled. After two violent or serious felony convictions, any further felony, nonviolent or not, will trigger a third strike. The mandatory sentence will then be the greater of: three times the term ordinarily required, 25 years, or a term determined by the court. 156 Expenditures Inmate =  Employees   Inmate  Payroll Employee    (9.2) + Nonpayroll Expenditures Inmate The first component in this equation separates labor inputs and their average costs. Payroll expenditures account for nearly half of all corrections expenditures, except in Illinois and New York, where they constitute 61 and 73 percent of total spending, respectively. The second component—nonpayroll expenditures per inmate—consists of total expenses for the provision of inmate services such as dining, supplies, health care, and job training. Governments might incur these costs directly through their own provision or indirectly through contracts with private enterprises. Nonpayroll costs also include some employee benefits but exclude pension plans.3 California’s corrections staffing ratio is very close to the national average, but its expenditures per employee are significantly higher than in comparable states (Table 9.4). By contrast, in New York, payroll costs are higher than in California, and the staffing ratio is much higher. On the other hand, New York’s nonpayroll costs are much lower. The gap between wages in California and in other states stems in part from the need for all employers in California to compensate workers more in this state. For this group of workers (Zone 3 as described in Chapter 2), employers must pay 14 percent more than their counterparts in other states. However, California governments pay their corrections employees nearly 40 percent more than the average for all other states (Table 9.5). Thus, more than half the difference in corrections payroll between California and other states remains unexplained by general labor market conditions. _____________ 3The Census Bureau tracks these expenditures in a separate retirement category in the insurance trusts sector. It does not distinguish among government functions (such as corrections) in these data. 157 Table 9.4 Components of Corrections Expenditures per Inmate, 2002 California U.S.–California Florida Illinois New York Texas Expenditures per Inmate ($) 14,222 8,655 8,619 8,555 14,521 6,291 Payroll Expenditures Employees per Employee per Inmate ($) 0.12 52,279 0.12 37,707 0.11 34,679 0.12 42,302 0.19 55,228 0.10 30,281 Nonpayroll Expenditures per Inmate ($) 7,777 4,151 4,929 3,379 3,893 3,224 Table 9.5 Average Yearly Payroll in the Correctional System, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 52,279 37,707 34,679 42,302 55,228 30,281 Location Adjusted 46,267 38,397 37,889 40,991 48,217 32,788 As mentioned above, nonpayroll costs include employee benefits as well as services to inmates and other operational expenditures. Examples of inmate services are rehabilitation and vocational training programs and health care services (e.g., drug treatment). Other operational expenditures include utilities, leases, and payments to outside contractors. Although our data do not allow us to separate these individual components, nonpayroll costs as a group are considerably higher in California than in the rest of the country. They are even higher than in New York, typically considered a high-cost state. Differences across states in nonpayroll costs may be due to different choices of which supplemental services to provide and how to provide them (e.g., different practices of contracting out). 158 The evidence presented so far suggests that California’s higher expenditures on corrections result mainly from higher operating costs. However, it is undeniable that the correctional system has also faced an increasing workload over the past 30 years (Figure 9.3). This trend raises the question of whether California’s correctional system has sufficient infrastructure to support a very large number of prisoners (e.g., Warren and Reiterman, 2005). The answer to this question depends on which measure of crowding is used. The “rated capacity,” or the number of beds judged appropriate for prison facilities, does not show that California’s large inmate population has induced an overcrowding of facilities. By this measure, California state prisons were slightly under full capacity (98%) in 2002 (Table 9.6). The ratio of inmates to beds officially assigned at each facility was also below that of Illinois or New York. At the local level, California’s jails were more crowded than those in the rest of the nation and all of our comparison states, but they were only slightly above full capacity. These data suggest that anecdotal evidence of overcrowding might be due not only to overall capacity constraints but also to the alternative use of spaces originally conceived for rehabilitation or recreation to house inmates. In addition, there may be limits on availability in certain types of facilities (e.g., maximum security). The ratio of inmates to design capacity, or the original number of beds planned for prison facilities, presents a different picture for state prisons. Here, California state prisons are more crowded than those in other large states (Table 9.6). (Data on design capacity are not available for local facilities.) (See Bailey and Hayes, 2006.) The perception of overcrowding could also reflect an ongoing mismatch between the growth in inmate population and the allocation of resources for capital improvements. A way to assess this conjecture is by comparing cumulative capital outlays on corrections with the growth of the inmate population (Table 9.7). Between 1992 and 2002, California’s investment per additional inmate was not very different from that in the rest of the country. It was far ahead of investment per additional inmate in Texas but below expenditures in Illinois or New York. 159 Table 9.6 Average Occupancy of Correctional Facilities, 2002 California U.S.–California Florida Illinois New York Texas Percentage of Rated Capacity State Prisons Local Jails 98 103 — 102 95 95 143 98 110 75 91 94 Percentage of Design Capacity State Prisons 201 113 126 164 124 91 Table 9.7 Real Capital Outlay per Additional Inmate in State Prisons, 1992–2002 California U.S.–California Florida Illinois New York Texas Increase in State Prison Inmate Population 52,821 422,065 26,908 11,053 5,329 101,536 Real Capital Outlay ($) 4,612,590,963 38,490,431,059 2,532,558,435 1,801,186,458 4,797,594,035 4,270,701,794 Real Capital Outlay per Additional Inmate ($) 87,325 91,196 94,119 162,959 900,280 42,061 Police Protection California spent more than the average of all other states on police protection in 2002 ($290 per capita compared to $215 per capita). However, its spending was on par with other large states—including Florida ($263 per capita) and Illinois ($256 per capita)—and less than New York’s ($351 per capita) (Table 9.8). These expenditures occurred mostly at the local level. In the last 35 years, California has spent more per capita on police protection than the average in the rest of the nation and any other state in our comparison group, with the exception of New York (Figure 9.4). 160 Table 9.8 Police Protection Expenditures per Capita, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 290 215 263 256 351 171 State Local 32 258 33 182 25 238 29 228 30 321 18 154 300 250 Expenditures (2002 $) 200 150 California U.S.–California 100 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.4—Police Protection Expenditures per Capita, 1977–2002 As in the case of corrections, capital expenditures are a modest share of total police protection spending. Therefore, this section focuses on operating costs. To gain some insight into police services, it is useful to look at a simple decomposition of expenditures per capita: 161 Expenditures Population =  Crimes   Population  Employees Crime  Payroll Employee    + Nonpayroll Population (9.3) California’s crime rate and staffing level per crime were close to the average in the rest of the nation in 2002 (Table 9.9). California’s ratio of police staff to the number of crimes contrasts with New York’s, the highest in our comparison group. In California and most comparison states, payroll expenditures represent roughly two-thirds of expenditures per employee. The exception is New York, where payroll takes up 95 percent of total expenditures per employee. Operating costs other than wages, which include employee benefits, have also been high in California. It is unclear whether this is a consequence of state-specific factors. For example, New York’s nonpayroll expenditures per capita have historically been quite low whereas Florida’s are large and closer to those in California. As shown in Table 9.9, wages for police officers and other police protection staff members are higher in California than in the rest of the nation and in all of our comparison states except New York. As in the other professions analyzed in this chapter, this is partially due to a location premium. However, even after taking this premium into account, wages remain higher in New York and California (Table 9.10). In recent years, California’s per capita expenditures for police protection have not grown as fast as in the rest of the nation (1.6% compared to 2.8%), outpacing only Florida’s (0.9%) and Texas’s (1%). This trend has mirrored the decrease in the number of crimes per capita in California, especially in recent years (Figure 9.5). Taken together, Table 9.9 and Figure 9.5 raise questions about California’s resource allocation in police protection. As a consequence of the reduction in the crime rate and a moderate growth in police staffing per capita, the number of police staff per crime has increased. To be sure, staffing levels are not determined in isolation. For most of the past 162 Table 9.9 Components of Police Protection Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Total per Crimes per Employees Employee Capita ($) Capita per Crime ($) 268 0.040 0.072 64,456 203 0.041 0.075 47,455 244 0.054 0.065 45,698 257 0.040 0.089 49,194 315 0.028 0.165 71,942 161 0.052 0.054 40,806 Nonpayroll per Capita ($) 105 67 101 81 17 58 Table 9.10 Average Yearly Payroll for Police Protection, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 64,456 47,455 45,698 49,194 71,942 40,806 Location Adjusted 56,557 48,324 49,927 47,670 62,809 44,184 35 years, California’s crime rate exceeded that in the rest of the country. Like many states, California has experienced fewer reported crimes per capita since the mid-1990s. This trend could be due to any number of factors including a better trained police force as well as changing demographics, longer prison sentences, or a strong economy. In any event, this may be an opportune time to reexamine resources dedicated to staffing versus other priorities within police protection and across state and local government budgets. Fire Protection California’s fire protection expenditures were high relative to those in the rest of the nation in 2002 (Table 9.11). With the exception of 163 Reported offenses 0.070 0.065 0.060 California U.S.–California 0.055 0.050 0.045 0.040 0.035 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Figure 9.5—Reported Offenses per Capita, 1990–2002 Table 9.11 Fire Protection Expenditures per Capita, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures per Capita 119 86 111 108 128 69 Counties 28 11 47 <1 2 1 Cities 73 64 55 79 102 65 Special Districts 18 10 9 29 24 2 Texas, other comparison states also displayed high spending levels in this area. Although all types of local governments (counties, cities, and special districts) provide these services, responsibilities for fire protection are concentrated at the city level. However, the role of cities is less dominant in California than in other states. 164 Expenditures (2002 $) 130 120 110 100 90 80 70 60 California 50 U.S.–California 40 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.6—Fire Protection Expenditures per Capita, 1977–2002 California’s expenditures on fire protection have consistently exceeded spending in the rest of the nation (Figure 9.6).4 Operating costs drive most fire protection expenditures. As a result, we can gain a better understanding of expenditures per capita through their decomposition into staffing levels, payroll, and nonpayroll expenditures (Table 9.12). California has the lowest fire protection staffing ratio within our comparison group, even below that of Texas, where per capita spending is much lower. Higher expenditures per capita in California mainly result from higher expenditures per employee, which mainly consist of payroll. This contrasts with Florida, where expenditures per employee are moderate and staffing ratios are high. As a result, its per capita expenditures are high not only because of the costs of labor but also because Florida maintains a higher level of labor inputs. As in the case of corrections and police protection, the differences in payroll costs are not fully explained by location factors (Table 9.13). _____________ 4The unusually large variation in fire protection expenditures for California between 1988 and 1991 is due to fluctuations in special districts’ expenditures. Most of this variability is attributable to reduced precision in the Census Bureau’s estimates stemming from the nature and size of the sample in those years. 165 Table 9.12 Components of Fire Protection Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Employees Expenditures per 1,000 per Capita ($) Residents 119 0.9 86 1.1 111 1.4 108 1.3 128 1.2 69 1.0 Payroll Expenditures per Employee ($) 80,372 50,324 50,443 55,019 74,590 47,978 Nonpayroll Expenditures per Capita ($) 46 32 41 38 36 22 Table 9.13 Average Yearly Payroll for Fire Protection, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted 80,372 50,324 50,443 55,019 74,590 47,978 Location Adjusted 70,522 51,245 55,112 53,314 65,121 51,950 The low-staff/high-wage situation in California is not new. The state’s per capita expenditures exhibited only moderate real growth from 1993 to 2002 (6%).5 This was the slowest pace within our group of states and well below the growth in the rest of the nation. California is also the only state with a decline in the number of employees per capita from 1993 to 2002 (Figure 9.7). On the other hand, expenditures per employee have been higher than in the rest of the nation, primarily driven by increases in payroll. However, their growth has been modest in recent years, averaging 1 percent per year in California. By contrast, in New York, payroll _____________ 5The break in Figure 9.7 occurs because the Census Bureau did not report employment data for fire protection in 1996. 166 Fire protection employees 0.12 California U.S.–California 0.11 0.10 0.09 0.08 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Figure 9.7—Fire Protection Employees per 10,000 Residents, 1993–2002 expenditures have grown by an average of almost 2 percent per year. This trend is similar to the national average excluding California, although above the growth rate in some of our comparison states. Protective Inspection and Regulation This category encompasses several regulatory, oversight, and enforcement activities that do not entail a direct provision of protective services. In California, the largest expenditures in this category are related to regulation of public utilities, licensing and expedition of identifications, and consumer protection. Smaller functions include fraud control, regulation and inspection of medical boards and insurance and financial institutions, and supervision of industrial relations. It is important to realize that because of the residual nature of this Census category, the range of activities it includes for other states is likely to include similar functions as in California as well as many others that differ from one state to another. California’s state government incurred 74 percent of total protection and inspection expenditures in 2002, leaving the rest to cities and counties (Table 9.14). The historical averages for those shares are 61 and 39 percent, respectively. The higher share of state expenditures in 2002 167 Table 9.14 Protective Inspection and Regulation Expenditures, by Type of Government, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures per Capita 111 31 34 24 38 21 State Expenditures per Capita 82 20 14 16 25 14 Local Expenditures per Capita 29 11 20 9 13 7 may stem from one-time spikes in the costs of regulating utilities after the energy crisis in 2000–2001. Despite the small share of local expenditures in California in 2002 (26%), the state has had a historically high degree of decentralization of inspection and regulation services relative to the rest of the country. In California, state expenditures in this category consisted of 49 programs distributed among 18 agencies in 2002. Within state activities, the Public Utilities Commission, the Department of Motor Vehicles, and some programs in the Department of Consumer Affairs spent the largest share of expenditures (82%). Smaller state oversight activities were aimed at labor relations, insurance and financial institutions, and alcohol, games, and gambling. Regulatory activities by California’s local governments were mostly under control of city governments. Construction regulation and enforcement constituted 83 percent of total local expenditures. The rest was aimed at building inspection and agricultural regulation and was under the responsibility of county governments. California’s spending in such activities has always been considerably higher than in the rest of the country (Figure 9.8). Expenditures in 2002 were exceptionally high, perhaps as a consequence of additional regulatory burdens stemming from the 2000–2001 energy crisis. In 2000, protective inspection and regulation expenditures amounted to 168 Expenditures (2002 $) 120 California 100 U.S.–California 80 60 40 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 9.8—Protective Inspection and Regulation per Capita, 1977–2002 $73 per capita, an amount much closer to the state’s historical average expenditures on these activities. It is difficult to determine what has caused these high expenditures in California. One reason could be the higher costs of enforcement, regulation, or oversight. Another possibility is a large number of programs that need to be monitored. If the former reason dominates, California’s spending level could indicate high compliance burdens. Alternatively, the state’s size and economic diversity make a case for a need to oversee a wider set of activities in California. Both explanations are likely to carry some weight. Conclusions California devotes more resources to public safety than many comparable states do. The size of the state’s population and the underlying need for safety services explain some of these expenditures. A closer examination reveals that provision costs are also higher in California than in most of our comparison states. Labor costs are the most important component of operating expenses. The observed gap between wages in California and in other states cannot be completely 169 explained by the need to compensate employees for living in a high-cost state. California’s spending on corrections is high relative to that in the rest of the nation and to our comparison group, even though the number of inmates per capita is among the lowest. California’s high correctional payroll is consistent with high overall operations expenditures. California does not exhibit overcrowding of state prisons in terms of the number of inmates to beds officially assigned at each facility. However, local jails are operating above capacity according to this measure. California’s state prison occupancy also looks much worse compared to other states as a percentage of design capacity. In addition, California’s capital outlay per additional inmate has been slightly below the average for the rest of the nation in recent years. It has also been lower than the infrastructure investment of other states with large confined populations, such as Illinois and New York. Fire and police protection are also labor-intensive activities where California’s labor costs remain higher than our comparison group, except for police officer wages in New York. Payroll expenditures have been the primary source of recent growth in per capita expenditures on these functions. Finally, California’s expenditures in government activities aimed at inspection and regulation are substantially higher than those in the rest of the country. Possible explanations for these high levels of spending include a wider set of activities that need regulation because of the size and diversity of California’s economy, higher regulatory costs, and an inherently stronger emphasis on regulation than in other states. 170 10. Environment and Housing In 2002, California state and local governments allocated 7 percent of total expenditures to environment and housing, compared to 6 percent in the rest of the nation. Environment and housing is a broad category as defined by the U.S. Census Bureau, combining two areas not typically considered together. The environment component includes public spending on agriculture, fish and game, forestry, and other natural resources as well as on parks and recreational facilities. It also includes spending on public services that can affect environmental quality, such as sewerage and solid waste management.1 For similar reasons, we consider these enterprises together with publicly owned water, gas, and electric utilities at the end of the chapter. In 2002, California’s state and local governments spent one-third more than the average for all other states on environment and housing. They spent 50 percent more on housing and community development, even though fewer households received direct housing assistance than in the rest of the nation. California’s higher spending on this function may be due to the higher costs of providing housing assistance in this state or to a greater emphasis on redevelopment rather than housing assistance compared to other states. California spent 80 percent more than the rest of the country on natural resources in 2002. Its total expenditures were comparable to those of Florida and other western states. Natural resources spending net of revenues (including fees from local water agencies in California) was only 30 percent greater than the average for all other states, and it was 40 percent less than in Florida. Per capita expenditures on parks and recreation and enterprise activities including public water, gas, and _____________ 1This category does not include state and local spending on environmental protection, such as monitoring and regulation of air and water quality. These expenditures are captured in the “other health” category discussed in Chapter 6. 171 electric utilities did not differ markedly from those of other states once user fees were subtracted from the total. Total Expenditures Environment and housing is a diverse category as defined by the Census Bureau. The housing and community development function itself comprises several types of programs, including those that build and maintain affordable housing, provide assistance to low-income renters and homebuyers, and revitalize urban and rural areas. Expenditures in this category are predominantly local, although they often include passthroughs of state and federal funds. In particular, this category includes the expenditures of local public housing authorities (PHAs), locally chartered and governed agencies responsible for administering federal housing programs such as the Section 8 rental subsidy program and public housing projects overseen by the U.S. Department of Housing and Urban Development (HUD). It also includes local spending from federal Community Development Block Grants (CDBGs) for neighborhood revitalization. It is worth noting that the largest active form of housing assistance in the United States, the low-income housing tax credit (LIHTC), is not reflected in these data because it represents forgone revenue rather than a direct expenditure.2 Similarly, Census Bureau data do not reflect the proceeds of tax-exempt private activity bonds (PABs), including bonds for the construction of multifamily housing as well as bonds that finance below-market interest rate mortgages or reduced federal tax liability for eligible homeowners (i.e., mortgage revenue bonds and mortgage credit certificates).3 _____________ 2In 2002, the federal government authorized $60 million in tax credits for California. California also runs its own tax credit program, with $70 million authorized in 2002 (California Tax Credit Allocation Committee, 2006). 3In 1999, the federal government authorized $1.6 billion in tax-exempt bonds for private purposes in California, of which the state allocated $1.4 billion to housing. California accounted for nearly 30 percent of the value of mortgage credit certificates issued in 2000 and more than half of the value of these bonds issued from 1992 to 2000 (Greulich and Quigley, 2003). 172 The natural resources component of environment and housing describes government conservation, improvement, and promotion efforts in agriculture, fish and game, forestry, and resources “not elsewhere classified” (NEC). In California, the NEC category corresponds mainly to water resources. Roughly half of this spending occurs at the state level through the Department of Water Resources and several state conservancies (e.g., the State Coastal Conservancy, Santa Monica Mountains Conservancy, Tahoe Conservancy). The remainder is spent by counties and special districts on flood control, drainage, land reclamation, and levee maintenance. In other states, this subcategory may correspond to the protection of water, soil, energy, or mineral resources. Parks and recreation is the umbrella term for recreational, cultural, or scientific facilities. The bulk of spending in this area comes from local parks, golf courses, public beaches, museums, zoos, and convention centers. At the state level, this category includes expenditures of the state Department of Parks and Recreation as well as the California Science Center and Arts Council. Sewerage refers to government financing of sanitary sewers and storm water facilities. Sanitary sewers dispose of wastewater from buildings, whereas storm sewer systems channel runoff from storm drains. Solid waste management includes garbage collection and disposal as well as recycling, toxic cleanup, and street cleaning. The Census Bureau records expenditures on these activities whether they are performed by the government itself or through a private contractor. Finally, the water, gas, and electric utility functions comprise all government activities relating to the production, acquisition, and distribution of these resources to the general public or to other public or private utilities. In 2002, California state and local governments spent $20.6 billion on environment and housing combined, translating into $590 per capita (Table 10.1). The bulk of these expenditures (97%) occurred at the local level, although they often involved state and federal funds. California spent one-third more than the rest of the nation on this category but a similar amount to other large states including Florida ($539), Illinois 173 Table 10.1 Environment and Housing Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 590 448 539 548 566 336 Housing and Community Natural Parks and Solid Development Resources Recreation Sewerage Waste 154 126 119 113 77 104 69 102 108 65 71 144 110 103 111 142 51 201 118 37 197 24 105 117 124 72 46 94 82 42 ($548), and New York ($566). By contrast, California spent 75 percent more than Texas ($336). The pattern has persisted in California since 1977. Expenditures increased in relative terms during the late 1980s and mid-1990s, mainly because of increases in housing and community development expenditures. Among other large states, California’s spending over time has been similar to that of New York, whereas Florida, Texas, and Illinois more closely resemble the rest of the nation. The diversity of programs included in this category limits the usefulness of further comparisons of total expenditures across states. We therefore move to a discussion of the major components of this category, starting with housing and community development and then turning to natural resources, parks and recreation, and enterprise activities including sewerage, solid waste, and other utilities. Housing and Community Development In 2002, California state and local governments allocated $5.4 billion to housing and community development, equivalent to $154 per capita or about 50 percent more than the average for all other states (Table 10.2). Compared to other large states, California spent roughly the same amount per capita as Illinois, about 20 percent less than New York, and more than twice as much as Florida or Texas. Capital spending represented one-third of total spending on housing and community development, a similar proportion as in other states except 174 Table 10.2 Housing and Community Development Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 154 104 71 142 197 72 Current 103 83 64 92 153 50 Capital 51 20 7 50 44 22 Florida. However, these amounts fluctuate over time, as discussed below. California has not always been a “high-spending state” on housing and community development. It was at or below the average for all other states during the 1970s and early 1980s (Figure 10.1). Spending doubled during the 1980s and grew another 30 percent in the early 1990s before dropping sharply in 1997. Among our comparison states, New York exhibited a similar pattern, although accelerated by a few years, with expenditures peaking in the early 1990s. 200 180 160 140 120 100 80 60 California 40 U.S.–California 20 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.1—Housing and Community Development Expenditures per Capita, 1977–2002 Expenditures (2002 $) 175 As with all functional categories defined by the Census Bureau, it is not possible to further disaggregate spending on housing and community development by program. It is clear from Table 10.3, however, that total spending in this area is driven largely by federal grants. In 2002, for example, federal intergovernmental revenue accounted for over half of all California spending on housing and community development and an even greater share of spending in all of our comparison states, reaching a high of 95 percent in New York. To better understand the drivers of housing and community development expenditures in California and other states, we can apply our standard decomposition framework from Chapter 2: Expenditures = Caseload  Expenditures Population Population Case (10.1) and Caseload = Participants  Eligibles Population Eligibles Population (10.2) Identifying the caseload for spending on housing and community development is complicated by the variety of programs in this area. Cases may include families receiving rent subsidies under federal housing programs as well as those benefiting from local planning or redevelopment efforts. Table 10.3 Housing and Community Development Federal Intergovernmental Revenue per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 154 104 71 142 197 72 Federal Intergovernmental Revenue 84 78 51 109 188 45 176 For major federal housing assistance programs, eligibility is restricted to “low-income” households, defined as those with earnings at 80 percent of the local median income. There are also adjustments for family size and incentives for local authorities to target “very” and “extremely” low-income families at 50 and 30 percent of the area median income, respectively. For local redevelopment efforts supported by federal funds, the Community Development Block Grant program requires that funded activities benefit areas with a concentration of lowincome individuals.4 Table 10.4 shows that there is little variation by state in the proportion of households meeting these eligibility criteria. Where states do differ is in the proportion that actually receives federal, state, or local rental subsidies or resides in federally supported public housing. Although participation rates are generally quite low, they are particularly low in California. In 2002, 8.3 percent of all eligible households in California received rent subsidies or lived in public housing, compared to 11.5 percent in the rest of the nation. Moreover, California’s participation rate was lower than that of all other large states in our comparison group except Texas. Although figures differ somewhat from year to year, California was below the average for all other states in the Table 10.4 Housing Assistance Caseloads and Components, 2002 Total Households Recipients of Low-Income Housing Families as Assistance as Percentage of Percentage of Total Total Participation in Housing Assistance as Percentage of Eligible California 11,935,960 41.8 3.5 8.3 U.S.–California 97,452,805 41.4 4.8 11.5 Florida 6,683,618 39.6 3.8 9.5 Illinois 4,836,880 42.0 4.7 11.2 New York 7,294,127 43.4 7.9 18.3 Texas 7,493,242 43.5 2.6 6.0 _____________ 4Similarly, in California, Redevelopment Agencies (RDAs) must spend 20 percent of their tax increment financing (i.e., increases in property tax revenues above inflation) on low- to moderate-income housing. 177 share of all households receiving housing assistance from 1977 to 2002 (Figure 10.2). Unlike other forms of public support, housing assistance is not an entitlement. Rather, households must apply for aid and there are generally many more applicants than subsidies available. Local PHAs determine who receives federal assistance within HUD guidelines. Table 10.5 suggests that it is unlikely that California’s lower participation rate in rent subsidy and public housing programs stems from a lack of demand. In fact, California includes more households with “moderate” to “severe” rent burdens (defined by HUD as households paying rents in excess of 30 and 50 percent of their income) than in the rest of the nation and in all other large states except New York. Moreover, according to the most recent data available, California residents who are eventually accepted into federal programs tend to spend a longer time on waiting lists than the average for the rest of the nation and in many of our comparison states (Table 10.6). California’s higher rent burdens also suggest that the cost of subsidizing each household is higher. This is evident in the Fair Market 6.0 5.0 4.0 3.0 2.0 California 1.0 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Figure 10.2—Percentage of All Households Receiving Rent Subsidies or Public Housing, 1977–2002 178 Percentage Table 10.5 Rent Burdens as a Percentage of Household Income, 2000 California U.S.–California Florida Illinois New York Texas Renters Paying 30 to 50 Percent of Income 22.1 18.7 21.3 18.0 18.9 18.6 Renters Paying More Than 50 Percent of Income 20.2 17.2 19.6 17.3 21.6 15.7 Table 10.6 Average Months on Waiting List for Housing Program Participants, 1998 California U.S.–California Florida Illinois New York Texas All Programs 33 22 28 40 29 22 Section 8 Certificates and Vouchers 37 28 36 63 34 31 Public Housing 18 11 16 9 10 11 Rents (FMRs) used by HUD in calculating rent subsidies under the Section 8 program.5 Table 10.7 confirms that FMRs are higher for nearly all apartment sizes in California than in other large states. The average rent for a two-bedroom unit in 2002 was $934 in California, compared to $673 in the rest of the nation and $903 in the next highest rental market, New York. It is worth noting that these higher rents may reflect underlying differences in the quality of the housing stock apart from apartment size. _____________ 5FMRs are defined as the 40th percentile of rents in a metropolitan area or nonmetropolitan county (or the 50th percentile in a few metropolitan areas). 179 Table 10.7 Statewide Average Fair Market Rents Weighted by Population in Each Market Area, 2002 (dollars) California U.S.–California Florida Illinois New York Texas 01 2 3 4 Bedrooms Bedroom Bedrooms Bedrooms Bedrooms 639 746 934 1,283 1,484 459 547 673 886 1,021 478 560 686 916 1,068 535 639 774 978 1,099 661 768 903 1,161 1,296 434 499 635 874 1,023 In 2002, capital outlay on housing and community development amounted to $51 per capita in California and $20 per capita in the rest of the nation. However, because capital expenditures differ significantly from year to year and contribute to a preexisting stock, a more appropriate measure would consider spending over a longer period as a function of growth in demand, measured here as population growth. Capital spending also depends on the age of the housing stock. Presumably, states with an older stock will need to spend more because of depreciation regardless of the population growth rate. From 1992 to 2002, California spent $14.5 billion on capital outlay for housing and community development.6 This translated to $3,610 per additional state resident, more than twice that in the rest of the nation but less than Illinois ($4,460 per additional person) and New York ($8,049 per additional person) (Table 10.8). In terms of the age of the housing stock, California is at the midpoint of our comparison group with the median housing unit being built in 1970, compared to 1962 in Illinois, 1954 in New York, 1977 in Texas, and 1980 in Florida. In sum, California spent 50 percent more than the average for all other states on housing and community development, although a similar amount as other large states including Illinois and New York. California served fewer households than other states, however, and this may be due to higher costs per subsidized household. However, this spending _____________ 6These expenditures do not reflect the proceeds of a $2.1 general obligation bond, authorized with the passage of Proposition 46 in November 2002. 180 Table 10.8 Housing and Community Development Real Capital Outlay, 1992–2002 California U.S.–California Florida Illinois New York Texas Cumulative Real Capital Outlay (2002 $) 14,485,723,956 44,854,701,852 1,124,716,237 3,973,465,154 7,276,633,908 2,532,155,424 Change in Population 4,013,000 27,411,000 3,030,000 891,000 904,000 3,963,000 Real Capital Outlay per Additional Person (2002 $) 3,610 1,636 371 4,460 8,049 639 category also encompasses local redevelopment and planning efforts, and there may be a greater emphasis on these efforts in California. Unfortunately, there are not sufficient data on redevelopment spending to distinguish between these two explanations. Natural Resources In 2002, California spent $4.4 billion overseeing its natural resources, including agriculture, fish and game, forestry, and “other” resources not elsewhere classified, which in California corresponds mainly to water resources. At $126 per capita, these expenditures were 80 percent higher than the average for the rest of the nation and significantly greater than expenditures in all of our comparison states except Florida, which spent 15 percent more than California in 2002 (Table 10.9). Spending in this category clearly depends on a state’s endowment of natural resources. For example, it is not surprising that California spent more in 2002 than did any of our comparison states on forestry, which includes spending on forest fire prevention, control, and suppression efforts. Sparsely populated states with rich oil and mineral resources as well as higher fish and game, forestry, and agriculture expenditures such as Alaska, Wyoming, Montana, and North Dakota tend to dominate 181 Table 10.9 Natural Resources Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 126 69 144 51 24 46 Agriculture 27 21 30 16 6 13 Fish and Game 10 10 9 4 2 7 Forestry 14 4 4 1 1 1 Other Resources 75 35 100 30 14 25 expenditures per capita in this category. California and Florida occupy a second tier along with other western states, including Arizona ($76 per capita in 2002), Nevada ($99), Utah ($93), and Washington ($118). The variety of programs included in the “other” category makes further comparisons difficult. For example, in California, roughly onethird of all “other” natural resources spending corresponds to the State Water Project (SWP), or the system of reservoirs, aqueducts, and pumps that transport water from the northern to the southern regions of the state. The Census Bureau categorizes the SWP under water resources because of its role in irrigation, drainage, flood control, soil conservation, and reclamation, even though the SWP also acts as a wholesaler to local water agencies, whose activities are recorded in the “water supply” category discussed below.7 In other western states, the “other” category probably also corresponds to water resources management activities, but we have no way to know for certain. In Florida, this category encompasses spending on an ambitious state and federal Everglades restoration program begun in the mid-1990s. To properly account for the public subsidy to natural resources, we must calculate spending net of user charges, such as fees paid to the State Water Project or local flood control districts. After subtracting these _____________ 7Apart from the SWP, two major federal water conveyance and storage systems— the Central Valley Project (CVP) and Colorado River Project—and several smaller Army Corps of Engineers projects provide water to local utilities in California. Expenditures of these projects are not reflected in Census of Governments data. However, the presence of federal projects may have the effect of lowering state and local government outlays for water in California and other western states. 182 revenues, California’s net spending on natural resources was only 30 percent higher than the average for all other states in 2002. California spent 2.5 times as much as New York, nearly twice as much as Texas, and two-thirds more than Illinois. However, California spent 40 percent less than Florida in 2002 (Table 10.10). The state spent one-third more than Arizona and about 90 percent more than Colorado on a net expenditure basis, although a comparable amount to Nevada, Utah, and Washington in 2002. Natural resources expenditures may differ considerably from year to year (Figure 10.3). For instance, both gross and net expenditures dropped by 40 percent during the early 1990s, perhaps as a result of declining water resource management responsibilities following the drought of 1987 to 1992. Nevertheless, California’s net spending on natural resources has consistently exceeded the average for all other states and that of all of our comparison states except Florida since 1977. It has been within the range of other western states over this period. Compared to its own historical record, California used to spend more on natural resources. Notably, in 1960, voters approved a $1.75 billion general obligation bond ($8.2 billion in 2003 dollars) to finance construction of the SWP (de Alth and Rueben, 2005). More recently, state funds for water resources management have increased with the passage of Propositions 40 and 50, which provide funds for ecological restoration and water supply management among other priorities. Table 10.10 Natural Resources Net Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Expenditures 126 69 144 51 24 46 Fees and Charges 44 6 2 1 0 3 Net Expenditures 83 64 142 50 23 44 183 Expenditures (2002 $) 100 90 80 70 60 50 40 30 20 California U.S.–California 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.3—Natural Resources Net Expenditures per Capita, 1977–2002 Parks and Recreation California spent $4.2 billion on parks and recreation in 2002. Nearly two-thirds of this spending occurred at the city level. The remainder was equally distributed between other types of local governments (counties and special districts) and the state (including local assistance). At nearly $120 per capita, California’s expenditures on parks and recreation were 16 percent higher than the average for all other states but similar to other large states, including Florida (whose expenditures were 8% lower than California’s), New York (12% lower), and Texas (26% lower) (Table 10.11). In contrast, Illinois spent nearly 70 percent more than California on parks and recreation in 2002. Capital expenditures represented approximately one-quarter of this total in California as in the rest of the nation and our other comparison states except Texas. In addition to spending more than the average for all other states, California also brings in more revenue from its parks and recreational facilities. In 2002, California state and local governments recouped nearly 30 percent of total expenditures from park entrance fees, parking 184 Table 10.11 Parks and Recreation Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 119 102 110 201 105 94 Current 91 70 80 132 80 49 Capital 28 32 30 69 25 46 fees, campground fees, and other user charges.8 Illinois captured a similar share of expenditures in 2002, whereas in other large states, revenues offset expenditures by around 15 to 20 percent (Table 10.12). Subtracting these fees and charges, California’s net expenditures on parks and recreation were $86 per capita in 2002, only 10 percent higher than in the rest of the nation. California used to spend as much as 80 percent more per capita than the average for all other states, but this gap has been generally narrowing over time (Figure 10.4). California’s higher net spending on parks and recreation after 2001 reflects an influx of new bond funds. For example, the passage of Proposition 12 in Table 10.12 Parks and Recreation Net Expenditures per Capita, 2002 (dollars) Fees and Net Expenditures Charges Expenditures California U.S.–California Florida Illinois New York Texas 119 102 110 201 105 94 33 86 23 79 23 87 61 139 23 81 13 81 _____________ 8This figure does not include revenues from hunting and fishing licenses, which contributed $2 per capita in 2002 in California and similar amounts in other states. 185 Expenditures (2002 $) 100 90 80 70 60 50 40 30 20 California U.S.–California 10 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 10.4—Parks and Recreation Net Expenditures per Capita, 1977–2002 March 2000 provided funds for land acquisition and local assistance grants. California’s roughly average expenditures per capita in this category are perhaps surprising in light of the state’s reputation as home to several landmark parks and recreation areas. One way to evaluate these expenditures is as a function of acres provided by state and local governments, similar to the decomposition of expenditures per capita into caseload and expenditures per case in Chapter 2: Expenditures = Acres  Expenditures Population Population Acre (10.3) California leads our comparison group in absolute state and local park acreage, although New York’s parks represent a larger share of nonfederal land (Table 10.13). After adjusting expenditures for land area in state and local parks, California spent one-third more than New York, a comparable amount to Florida and Texas, and about half as much as Illinois (Table 10.14). Illinois’ higher expenditures per acre may be due to the inclusion of museums in this category or to a greater presence of high-amenity parks such as golf courses. 186 Table 10.13 Federal, State, and Local Parks and Natural Resource Lands, 2002 (thousands of acres) California Florida Illinois New York Texas Local Parks 600 324 279 186 340 State Parks and Federal Parks, Natural Resource Forests, and Lands Wildlife Refuges 1,457 31,691 592 4,982 306 3,764 1,159 70 668 970 Total Land Area in State 99,814 34,513 35,573 30,217 167,550 In addition to state parks, California is home to over 31 million acres of federal parks, forests, and wildlife refuges, more by far than any of our comparison states. In 2002, California received more visitors per capita to its national parks than any other state. California also includes private alternatives to public recreation areas, such as private land trusts or land protected by conservation easements. According to the Land Trust Alliance (2004), California led the nation with 1.5 million acres covered through these arrangements in 2003. Within our comparison group, New York had 571,000 privately protected acres, Texas had 194,000, Florida 79,000, and Illinois 52,000 in 2003. Perhaps as a result of these alternatives, visits per capita to California state parks (including both state residents and nonresidents) lag those of Illinois and New York (Table 10.15). Table 10.14 Parks and Recreation Expenditures per Acre, 2002 California Florida Illinois New York Texas Total Expenditures ($ 1,000s) 4,172,705 1,836,837 2,523,951 2,006,117 2,050,467 State and Local Parks (1,000 acres) 2,057 916 585 1,345 1,008 Expenditures per Acre ($) 2,029 2,005 4,312 1,491 2,034 187 Table 10.15 State and National Park Visits per Capita, 2002 California U.S.–California Florida Illinois New York Texas State Parks 2.53 2.72 0.93 3.51 2.73 1.07 National Parks 0.98 0.85 0.29 0.03 0.25 0.32 Sewerage, Solid Waste, and Utilities California’s per capita expenditures on sewerage and solid waste public enterprises were comparable to the rest of the nation and many other large states in 2002 (Table 10.16). Its public water, gas, and electric utilities spent nearly three times more than the rest of the nation as a group. California’s public electric utility expenditures were particularly high in 2002 because of the state energy crisis.9 Table 10.16 Utility Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Sewerage 113 108 103 118 117 82 Solid Waste 77 65 111 37 124 42 Water 231 128 168 140 98 173 Electric 401 166 188 36 251 171 Gas 3 22 13 5 1 14 _____________ 9In January 2001, the state Department of Water Resources (DWR) purchased roughly $10 billion of electricity on the wholesale market and later resold it to investorowned utilities for approximately $3 billion. These purchases were underwritten by a loan of approximately $6 billion from the state General Fund and a $4 billion interim loan. Although the DWR eventually repaid this borrowing with proceeds from a bond secured by future rate payments, it appeared in the fiscal year 2002–2003 budget as a $6.9 billion expenditure from “unclassified state funds.” 188 Excluding electric utilities, expenditures net of user charges for these activities are quite low. For example, California state and local governments spent only $7 per capita on sewerage beyond user fees in 2002, compared to $13.5 nationally not including California (Table 10.17). Table 10.17 Net Enterprise and Utility Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Sewerage 7 14 6 52 40 –9 Solid Waste 25 28 21 27 101 9 Water 31 23 38 24 19 31 Electric 38 1 2 –6 12 0 Gas 0 0 –1 0 0 1 Conclusions California’s higher than average spending on environment and housing is evident in each of its major subcategories. California resembles Illinois and New York in its housing and community development expenditures per capita and is close to but lower than Florida and other western states in its natural resources spending. Expenditures on parks and recreation were comparable to those of other states once user fees were subtracted from the total. Similarly, net spending on sewerage and solid waste enterprise activities were relatively small and similar to other states. 189 11. Governmental Administration The previous seven chapters have analyzed specific government services, such as police protection, transportation, and housing. In the typical case, these services are the responsibility of a department or agency of a general government such as a city, county, or state. This government then has the responsibility of overseeing and supporting these departments and agencies. In the case of the city, for example, a city manager may supervise all the city’s departments, and these departments may receive support services from the city’s personnel and finance departments. The U.S Census Bureau classifies these supervising and supporting activities as governmental administration. It also includes the legal and judicial activities of government in this category. In 2002, governmental administration accounted for 5 percent of state and local expenditures in California and 4 percent in the rest of the country. California’s relatively high spending per capita in this area was due to both high nonpayroll expenditures and high payroll expenditures. California’s high payroll expenditures were not due to significantly higher employees per capita but rather to high average salaries, even after adjusting for the generally higher level of salaries and wages in the California labor market. Total Expenditures Governmental administration includes five Census Bureau functions: financial administration, central staff services, judicial and legal, general public buildings, and legislative. Because financial administration and central staff services are similar in nature, we combined them into an area labeled “finance and central staff.” At the state level in California, the five departments with the largest expenditures in this area are the Franchise Tax Board, the Department of 191 Motor Vehicles (DMV), the Department of General Services, the Public Employees’ Retirement System, and the State Board of Equalization.1 At the local level, the finance and central staff function encompasses the offices of county administrators, auditors, assessors, recorders, supervisors, and treasurers. At the city level in particular, this area includes the offices of city mayors, managers, councils, and clerks. It also includes offices of planning and zoning. However, administrative activities particular to a specific government service are excluded from this area. For example, the administration of a city’s police department would be included in public safety, not finance and central staff. For similar reasons, this area excludes administrative functions of singlepurpose special districts, such as water, sewer, fire, and mosquito abatement districts. The “judicial and legal” area comprises the courts, public prosecutors, public defenders, and legal counselors to city, county, and state government. In California, this area encompasses the 58 state trial courts located in each county, the six Courts of Appeal, the Supreme Court, the Attorney General’s Office, and the state Department of Justice. At the city and county level, it includes district attorneys, public defenders, and grand juries. The “general public buildings” category includes the construction and operation of buildings not related to specific functions and agencies. Examples are courthouses, city halls, and city and county administrative buildings. Excluded from this category are schools, police stations, firehouses, libraries, jails, and hospitals—buildings associated with specific government services. The legislative function encompasses state legislatures and related research and investigative agencies. In California, it includes the Assembly and Senate as well as the Office of Legislative Counsel. The category excludes similar activities at the city and county level, such as city councils and county boards of supervisors. These activities are _____________ 1Drivers’ licensing and other public safety aspects of the DMV are included in Chapter 8, whereas general government support such as Vehicle/Vessel ID and Compliance is included here. 192 included in central staff services, which we have combined with financial administration as previously discussed. In 2002, per capita spending on governmental administration was significantly higher in California than in other states (Table 11.1). Spending per capita in California was 47 percent higher than in the rest of the country. It was more than twice as high as in Texas, and even 15 percent higher than in New York, the next highest spending state in this area. Administrative spending per capita in California has exceeded spending per capita in other states for every year from 1977 to 2002 (Figure 11.1). Even at its lowest level in relative terms, in 1983, California’s spending per capita was 25 percent higher than the average for all other states. In California and its comparison states, finance and central staff represent the largest area of administrative spending. Legal and judicial spending is also significant in every state. On the other hand, legislative expenditures and expenditures on general public buildings are relatively small in every state—the two areas together constitute less than 20 percent of administrative spending. As a consequence, these two areas cannot be an important explanation of why administrative expenditures differ so widely across states, and we do not consider them further in this chapter. Table 11.1 Governmental Administration Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 447 305 293 330 390 208 Finance and Central Judicial Staff and Legal 216 201 161 96 160 99 198 88 178 143 97 75 General Public Buildings 21 40 26 36 59 30 Legislative 8 8 9 7 11 5 193 Expenditures (2002 $) 500 450 400 350 300 250 200 150 California 100 U.S.–California 50 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.1—Governmental Administration Expenditures per Capita, 1977–2002 Finance and Central Staff Spending per capita on finance and central staff has been higher in California than in the rest of the country for every year from 1977 to 2002 (Figure 11.2). The lowest margin occurred in 1998, when spending per capita in California was 17 percent higher than in other states. The highest, 48 percent, occurred in 1978. In 2002, expenditures per capita on finance and central staff were 35 percent higher in California than in other states. Expenditures in this area are primarily current rather than capital expenditures. In 2002, 95 percent of California’s total finance and central staff expenditures were current expenditures, a similar proportion as in other states. Although the mix between local and state expenditures differs widely among states, in most states at least half of finance and central staff expenditures occurred at the local level. In California, this percentage was 54 percent. Among our comparison states, Texas had the highest percentage at 71 percent, and New York had the lowest at 44 percent. 194 250 200 Expenditures (2002 $) 150 100 California 50 U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.2—Finance and Central Staff Expenditures per Capita, 1977–2002 The activities in the area of finance and central staff are laborintensive, which implies that payroll expenditures are a large portion of total expenditures (Table 11.2). In California, payroll expenditures were 52 percent of the total. This percentage was even higher in Florida, New York, and Texas. In contrast, payroll expenditures were only 44 percent of total expenditures in Illinois. In general, payroll expenditures understate personnel costs because they do not include the cost of employee benefits, which are included in nonpayroll expenditures. California’s relatively high spending in the area of finance and central staff is due to both high payroll expenditures per capita and high Table 11.2 Components of Total Expenditures per Capita on Finance and Central Staff, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 216 161 160 198 178 97 Payroll 113 90 88 85 111 67 Nonpayroll 103 71 72 113 67 30 195 nonpayroll expenditures. Payroll expenditures per capita in California were 25 percent higher than in other states. They were also higher than in the four comparison states. Nonpayroll expenditures per capita were 45 percent higher in California than in other states and higher than in all of our comparison states except Illinois. Following the general approach outlined in Chapter 2, payroll expenditures per capita can be decomposed into two components: payroll per employee and employees per capita. Specifically, applying Eq. (2.3): Payroll = Payroll  Employees Population Employee Population (11.1) California had slightly fewer employees per capita than the rest of the country, but it paid its employees much more generously than did other states (Table 11.3). Equation (2.4) in Chapter 2 demonstrates how the difference in payroll expenditures per capita can be attributed to each of these two components, where  represents the percentage difference in a variable between two states or between California and all other states:  Payroll   Payroll +  Employees Population Employee Population (11.2) Table 11.3 Components of Finance and Central Staff Payroll Expenditures, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Capita ($) 113 90 88 85 111 67 Employees per Payroll per 10,000 Employee ($) Residents 51,450 22 39,451 23 40,528 22 41,853 20 45,392 24 37,332 18 196 Applying this equation, in 2002, payroll expenditures per employee were 30 percent higher in California than in other states (the second term in Eq. (11.2)). These higher expenditures per employee were partially offset by a lower ratio of employees to population. Employees per capita (the third term in Eq. (11.2)) were 9 percent lower in California than in other states. As a consequence, payroll expenditures per capita (the first term) were 25 percent higher in California than in other states. California’s high average payroll for employees in finance and central staff is partly due to the generally high level of salaries and wages in the California labor market. Even after adjusting for these labor market conditions, however, California’s average payroll in this category is higher than the average in other states. The first column of Table 11.4 lists the average payroll for employees in finance and central staff. In the second column, average salaries are adjusted for labor market conditions using the method described in Chapter 2. In this particular case, these adjustments were made using the average salaries of Zone 3 workers, who have the education and skills typical of administrative employees in business and government. Before adjusting for labor market conditions, average payroll in California was 30 percent higher than in the rest of the country. After those adjustments, it is 12 percent higher. Thus, labor market conditions account for a little more than half of California’s relatively high average payroll for finance and central staff employees. Table 11.4 Average Payroll for Finance and Central Staff Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Payroll 51,450 39,451 40,528 41,853 45,392 37,332 Average Payroll Adjusted for Location 45,145 40,173 44,279 40,556 39,629 40,423 197 We do not have specific measures of outcomes or performance in the area of finance and central staff. In large part, this is due to the indirect nature of many of the activities in this area. An efficient human resources department is a key part of a well-run city, but it is inherently difficult to measure the contribution of that department to the effectiveness of any particular government service. Judicial and Legal In 2002, judicial and legal expenditures per capita were twice as high in California as in other states. As Figure 11.3 demonstrates, they have also been higher in California in every year since 1977. In that year, judicial and legal expenditures per capita were $12 per capita higher in California than in the rest of the nation. That margin remained roughly constant through 1981, then increased steadily until 1991, when California spent $112 per capita more than the rest of the country. From 1991 to 1997, this margin increased only slightly. From 1997 on, however, it increased dramatically. In 2002, California spent $181 more per capita than the rest of the country. As in the case of finance and central staff expenditures, judicial and legal expenditures are overwhelmingly current expenditures. In 250 200 150 100 50 California U.S.–California 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2002 Figure 11.3—Total Judicial and Legal Expenditures per Capita, 1977–2002 198 Expenditures (2002 $) California in 2002, 97 percent of total expenditures in this area were current expenditures. Other states had similar percentages. Our analysis of judicial and legal expenditures follows the same approach as our analysis of finance and central staff expenditures. Total expenditures are first broken down into payroll and nonpayroll expenditures. Payroll expenditures are then decomposed into payroll per employee and employees per capita. The final step is to adjust average payroll expenses for labor market conditions. Both payroll and nonpayroll expenditures per capita are higher in California than in the average in the rest of the country (Table 11.5). California’s nonpayroll expenditures are particularly remarkable. In 2002, nonpayroll expenditures per capita in California were triple the average in all other states. Nonpayroll expenditures include some personnel-related costs, such as health benefits. These expenditures also include payment for court-appointed lawyers, witness and jury fees, and private attorneys on retainer. We do not know, however, what specific factors account for California’s extraordinarily high nonpayroll expenditures. Our lack of this knowledge is particularly frustrating because nonpayroll expenditures account for much of the widening in the margin between expenditures per capita in California and those in the rest of the nation. In 1982, nonpayroll expenditures per capita were 13 percent higher in California than in the rest of the nation. By 2002, they were three times higher. In contrast, in 1982 payroll expenditures per capita were 33 percent higher in California than in the rest of the nation. By Table 11.5 Components of Total Judicial and Legal Expenditures per Capita, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Total 201 96 99 88 143 75 Payroll 93 61 75 65 102 45 Nonpayroll 108 35 24 23 31 30 199 2002, that margin had grown, but only modestly—it was 52 percent higher. Although less dramatic, payroll expenditures per capita are also higher in California than in the rest of the nation. In 2002, California exceeded other states in both components of judicial and legal payroll per capita (Table 11.6). It had higher average payroll per employee (34% higher) and more employees per resident (14% more) than all other states. The result was payroll expenditures per capita that were approximately 50 percent higher than in other states. New York followed a similar pattern—higher average payroll and more employees. Its payroll expenditures per capita were even higher than California’s. Labor market conditions in California are partly responsible for the high average payroll of California’s judicial and legal employees. In 2002, average payroll in other states was 26 percent lower than in California (Table 11.7). After adjustments for labor market conditions, this difference was 14 percent. Because many employees in the judicial and legal area are highly educated, we used the average salary of Zone 4 workers in making these adjustments. For the most part, Zone 4 workers have college degrees. As in the case of finance and central staff, we do not have performance or outcome measures in the judicial and legal area. This area also is a difficult one for defining and measuring output. Table 11.6 Components of Judicial and Legal Payroll Expenditures per Capita, 2002 California U.S.–California Florida Illinois New York Texas Payroll per Capita ($) 93 61 75 65 102 45 Employees per Payroll per 10,000 Employee ($) Residents 59,104 16 43,993 14 40,422 18 45,240 14 61,679 17 40,259 11 200 Table 11.7 Average Payroll for Judicial and Legal Employees, 2002 (dollars) California U.S.–California Florida Illinois New York Texas Unadjusted Average Payroll 59,104 43,993 40,422 45,240 61,679 40,259 Average Payroll Adjusted for Location 52,105 44,796 43,010 45,453 50,450 42,763 Conclusions The activities in governmental administration are among the most important in the public sector. No other area of government has the power of the courts to affect the welfare of individual citizens. The laws and regulations enacted by city councils and state legislators influence all government services. Partly because of their wide reach and indirect consequences, it is inherently difficult to measure performance in this category of governmental administration. We are left, therefore, with some observations about the costs of these activities. Per capita expenditures on governmental administration were significantly higher in California than in other states, and this condition has persisted for many years. Part of the explanation for California’s high expenditures is generous compensation for employees—more generous than dictated by general labor market conditions. On the other hand, California does not employ significantly more workers in this category than do other states. Nonpayroll expenditures are also part of the explanation. These expenditures are significantly higher in California than in other states. 201 12. California’s Fiscal Tradeoffs Previous chapters have examined California’s revenue system and spending patterns in each major functional area, such as education, health, and public safety. This chapter draws this information together in one place to ask: What tradeoffs has California made, either explicitly or implicitly, across budget categories? What might other tradeoffs look like? The chapter begins by reviewing what we know about Californians’ fiscal preferences based on public opinion data. Where expressed preferences differ from actual choices, it then poses some alternative scenarios. These scenarios are not policy prescriptions but illustrations of past fiscal decisions and the kinds of changes that would be necessary to undo these tradeoffs. Californians’ Fiscal Preferences Public opinion data provide a window to Californians’ thoughts about where the state should be in its revenues and public spending. The data suggest that Californians have a preference for a larger public sector. In January 2006, 61 percent of state residents said that they would rather pay higher taxes and have a state government that provided more in services, compared to 31 percent who preferred lower taxes and a smaller public sector (Baldassare, 2006a). In previous years, however, Californians were more divided on this question, at 49 to 44 percent in 2005 and 51 to 44 percent in 1998 (Baldassare, 2005a; 1998).1 Moreover, statewide average responses often mask divisions by political party and race or ethnicity. Democrats are much more likely _____________ 1In 1998, the question was worded differently: “If the government had a choice between reducing taxes or spending more on social programs like health care, social security, and unemployment benefits, which do you think it should do? (a) reduce taxes, even if this means spending less on social programs. (b) spend more on social programs, even if this means higher taxes.” 203 than Republicans to be in favor of a larger state government (76% compared to 53% in 2006). Independents, or those declining to state a party affiliation, fall between (61 percent). Also, white respondents are less likely than Latinos (52% versus 79%) to select the option of higher taxes and more services (Baldassare, 2006a). In terms of specific spending priorities, Californians nearly always single out K–12 education as their top concern. In January 2006, 71 percent of state residents said that they thought the state should spend more on public schools (Table 12.1). Similarly, when asked to name the most important problem facing the governor and state legislature, more respondents (25%) chose education and schools than any other issue (Baldassare, 2006a). Although top concerns often fluctuate with issues of the day, K–12 education is usually mentioned along with jobs and the economy, the state budget, and immigration. Health and human services is typically the second area where Californians would most like to increase state funds. In 2006, 61 percent of state residents indicated that they would like the state to spend more on this area (Table 12.1). Health and human services is also the area after K–12 education that Californians would most like to protect from spending cuts (Baldassare, 2003). Attitudes differ significantly Table 12.1 California Preferences for Government Spending, January 2006 (percent) K–12 public education Health and human services Roads and other infrastructure projects Colleges and universities Corrections, including prisons “For each area, please tell me if you think that the state government should spend more money than it does now, the same amount as now, or less money than now.” More Same Don’t Less Know 71 19 82 61 25 11 3 58 31 83 56 30 95 24 37 32 7 204 across party lines, however, with majorities of Democrats and Independents favoring additional spending on health and human services and Republicans as likely to want to maintain current spending levels as to increase funds in this area (Baldassare, 2006a). Usually vying for third place among Californians’ spending priorities are higher education and roads or infrastructure. In January 2006, 58 percent of California residents and 59 percent of likely voters voiced support for increased funding for roads and infrastructure. At the same time, 56 percent of state residents and 51 percent of likely voters said that they wanted the state to spend more on public colleges and universities. Support for higher education depends somewhat on party affiliation, with a plurality of Republicans (46%) favoring current funding levels. Similarly, support for roads and infrastructure varies by region, with residents of Los Angeles and the rest of Southern California more likely than other Californians to favor additional state funds in this area (Baldassare, 2006a). The only area where many Californians would like to decrease state spending is the correctional system. In January 2006, 32 percent of state residents said that they would like to spend less on corrections, including prisons (Table 12.1). The state corrections system is also the only area where majorities from both political parties support the same or less funding (Baldassare, 2006a). In prior years, even more Californians (42% in May 2005) expressed support for reducing state spending in this area (Baldassare, 2005a). Many Californians (e.g., 57% in February 2003) say that they would be willing to pay higher taxes to fund their budget priorities. Support for tax increases is higher if revenues are to be earmarked for specific purposes or if tax increases are to be limited in size (e.g., Baldassare, 2003). However, when asked about the level of taxation apart from spending priorities, Californians are often evenly split on whether they pay “about the right amount” or “somewhat” to “much more” in taxes than they should (e.g., Baldassare, 2004a). In addition, a significant proportion of Californians (nearly two-thirds in January 2005) believes that there is sufficient waste in state government to spend less without cutting services. Among these respondents, 70 percent believe the state wastes 10 percent or more of total expenditures (Baldassare, 2005a). 205 The latter responses suggest an underlying distrust of government. When asked in 2005 how often they could trust statewide elected officials “to do what’s right,” a large majority (73%) suggested “only some of the time” (Baldassare, 2005b). Californians are only slightly more confident in their local government officials (e.g., Baldassare, 2005c). In contrast, many Californians (40%) have a “fair amount” of confidence in their own ability to decide fiscal and other issues at the ballot box (Baldassare, 2005b). Although regional differences in values or culture are often invoked as explanations for divergent fiscal choices across the nation, Californians do not differ from other Americans in their stated fiscal preferences. For example, in a national Gallup poll, nearly 80 percent of Americans were unwilling to cut spending on K–12 education or health and human services to balance their state budgets (Table 12.2). On the other hand, 56 percent of Americans were willing to reduce funding for their state correctional systems (The Gallup Organization, 2003). Similarly, survey respondents in other large states including Florida, Illinois, and Texas also typically rank K–12 education, health, social services, and higher education as their top budget priorities (Table 12.2). Residents of these states also voice doubts about their elected officials. For example, majorities of New Yorkers have disapproved of their state legislature’s job performance since 2004 (Quinnipiac University, 2006). In Florida, most residents (e.g., 57% in 2004) trust their state government to do what is right only “some of the time” (McDonald, 2004). Californians’ desires for enhanced public spending, mixed sentiments toward taxes, and distrust of elected officials are far from unique but present challenges to state and local decisionmakers. Alternative Tradeoffs Results presented in foregoing chapters suggest that there are gaps between Californians’ fiscal preferences and budget realities. These gaps may be due to difficulties that state and local governments face in reconciling higher costs, larger service populations, and federal program requirements, among other pressures. Alternatively, they may reflect unrealistic preferences among California voters. In any event, it is 206 Table 12.2 Budget Priorities in Other States, 2002 and 2003 (percent) Texas, Florida, 2002 Illinois, 2003 2003 New York, 2003 United States, 2003 “What areas “Would you “Bad idea to of state be willing to “Should balance the government “Should state pay $25 to state state budget by “Would you be would be “Please tell me . . . whether you spending in maintain spending in cutting willing to cut most think the amount now being this area be current service this area be spending in state funding appropriate spent should be . . .” increased?” levels?” cut . . . ?” this area?” for this area?” to cut?” Yes, No, Not Increased Maintained Decreased Yes Yes Not at All Yes Willing Willing 207 K–12 education 85 13 2 77 78 Health services 74 23 4 61 75 Social services 62 32 7 61 63 Higher education 58 38 4 54 58 Protecting the environment 57 36 7 48 State highways and road systems 41 50 8 37 State prisons and correctional facilities 34 48 18 23 77 69 64 59 34 25 84 22 77 8 85 21 78 16 56 useful to explore these gaps further by illustrating the kinds of changes that would be required to undo California’s current fiscal tradeoffs. These scenarios are not meant to be exhaustive. There are many possible alternatives, depending on choices of policy levers, comparison states, and base year. Readers may also look across scenarios and juxtapose different options. More important, these scenarios are not prescriptive. We do not explore how to implement these changes and, indeed, there may be substantial barriers to implementation, including federal requirements or state constitutional restrictions on government tax and spending authority. On the other hand, institutional barriers also represent policy choices. For example, the state could apply for waivers from rules to participate in federal programs, and voters could amend constitutional limits on state and local revenues and spending. Raising Staffing Levels in K–12 Schools As we have seen, more than 70 percent of Californians would like to spend more on K–12 education. In particular, they would like to put more teachers in the classroom. Class size tops the list when Californians are asked what feature of California’s public schools most needs improvement (Baldassare, 2005d). Achieving this goal would be expensive. This is partly because the state is already at a disadvantage compared to the rest of the nation when it comes to staff resources per student. In 2002, California’s public schools employed 92 people per 1,000 students, compared to 129 per 1,000 in the rest of the nation and even higher ratios in Texas and New York (140 and 147 employees per 1,000 students, respectively). In addition, average compensation per school employee is high in California at $65,710 in 2002, compared to $63,456 in New York, the next highest-paying state in our comparison group, and $38,269 in Texas, the lowest-paying state in our group. If California were to increase school staffing ratios to the average for all other states, assuming that new teachers would command the same average salary as current teachers, total state and local government 208 expenditures would increase by $15 billion, or $430 per capita.2 This change is equivalent to a 30 percent increase in K–12 education expenditures or a 5 percent increase in total state and local spending as of 2002 (Table 12.3). Where might these additional resources come from? As we have seen, many Californians would like to reduce spending on corrections (Table 12.1). In 2002, California’s state and local governments spent more per capita in this area than the average for all other states ($257 versus $181). They also spent more than any of our comparison states, although the difference with respect to New York was not large. This higher spending came about despite California’s having fewer inmates per capita than the rest of the nation (0.018 versus 0.021) as well as fewer inmates than in some of our comparison states (Florida and Texas). As a result, expenditures per inmate were two-thirds higher in 2002 for California than the average for the rest of the nation ($14,222 compared to $8,655 per inmate). Table 12.3 Alternative Scenario #1: Raising Staffing Levels in K–12 Schools Expected Change in State and Local Expenditures Raise number of K–12 education staff per student to U.S. average excluding California Reduce corrections payroll per employee to U.S. average excluding California Reduce corrections nonpayroll costs per inmate to U.S. average excluding California Reduce K–12 education payroll per employee to U.S. excluding California Increase property taxes to U.S. average excluding California Increase revenues per higher education student to U.S. average excluding California $ per Capita 430 –24 –70 –150 –192 –123 % Total 5 –0.3 –0.8 –2 –2 –1 _____________ 2This assumption may be violated if new teachers draw lower salaries (e.g., Reed, Rueben, and Barbour, 2006). 209 Roughly half of all corrections expenditures per inmate are payroll costs, including salaries, wages, overtime, and bonuses but excluding employer contributions to retirement saving plans or health insurance. As explained in Chapter 9, some of these payroll expenditures certainly reflect a location premium, or the higher labor costs for all California employers seeking to attract workers with a given background or level of experience. However, the unexplained component of payroll costs accounts for about half the difference between California and other states. If California were to reduce corrections payroll expenses to the national average while continuing to pay workers the same location premium, payroll costs would drop from about $52,000 to $44,000 per employee. However, because California also employs fewer corrections personnel per capita than other states, total savings from this change would be only $800 million, or $24 per capita (Table 12.3). Reducing nonpayroll costs per inmate to the average for all other states ($4,151 instead of $7,777 per inmate) would provide an additional $2.4 billion, or $70 per capita (Table 12.3). Thus, total savings in corrections would be $94 per capita, less than one-quarter of the savings necessary to reach national staffing levels in K–12 education. Note that California could enact the same payroll adjustment for K–12 education as discussed above for corrections. That is, it could reduce salaries to the average for all other states while maintaining the location premium that all employers in California must pay for collegeeducated workers. This change would be analogous to the class size reduction (CSR) policy enacted in 1996, which earmarked funds for smaller class sizes rather than higher salaries.3 If this change were implemented, compensation per public school employee would fall from an average of about $66,000 to $57,000 per year, providing an annual $5.2 billion or $150 per capita in savings (Table 12.3). The substantial savings from this change in K–12 education as opposed to corrections result from the higher number of employees in elementary and secondary education (669,238 full-time_____________ 3However, demand for more teachers may also drive up salaries in the longer term (Jepsen and Rivkin, 2002). 210 equivalent employees, compared to 77,455 in corrections as of March 2002). California’s school finance system was transformed with the passage of Proposition 13 in 1978. In addition to capping property tax rates and limiting the growth in assessed values, Proposition 13 made the state responsible for allocating property tax revenues among schools and local governments (e.g., Sonstelie, Brunner, and Ardon, 2000). The effects of lifting Proposition 13 restrictions on local property tax rates and assessment practices are unclear. Sources of uncertainty include court mandates to equalize school funding across districts and feedback effects on real estate markets. Nevertheless, the experiences of other states provide a basis of comparison. As a result of Proposition 13, California state and local governments collect less in property taxes than the average state—$26 per $1,000 of personal income in 2002, compared to $32 per $1,000 in the rest of the nation—and even less than in other large states such as New York, Illinois, and Texas. If California were to raise the same amount of property tax revenues per $1,000 of personal income as in the rest of the nation, it would receive an additional $192 per capita in revenues (Table 12.3).4 Finally, California’s K–12 and higher education systems are clearly interdependent. The state’s public schools depend on its colleges and universities to provide skilled teachers. The higher education system, in turn, relies on K–12 schools for adequate student preparation. Institutions of higher learning also have potential social benefits as centers of innovation, public service, and regional economic growth. On the other hand, it is inescapable that average net support per fulltime-equivalent student in California’s colleges and universities is 50 percent higher than in its K–12 public schools ($11,300 per student, compared to $7,400 per student). This result stems in part from higher labor costs and in part from a tradition of low tuition and fees in _____________ 4Another proposal is to move to a “split roll,” under which commercial and industrial property would be assessed at market value and residential property would continue to be assessed at acquisition value. Revenue estimates for these proposals range from $1.9 to $3.3 billion, or roughly $54 to $94 per capita (Sexton and Sheffrin, 2002; Auerbach, 2003; California State Board of Equalization, 2002). 211 California. For example, enrollment in California’s community colleges was free as recently as 1985. Similarly, in the latter half of the 1990s, higher education fees were held constant in nominal terms, resulting in a real decline in revenues per student over that period. Despite more recent policy shifts, average fees for the UC and CSU systems remain below those of comparable institutions and CCC fees are the lowest in the nation (Legislative Analyst’s Office, 2006a). In 2002, higher education revenues (including both tuition and auxiliary revenues) were $5,404 per full-time-equivalent student in California, compared to $7,027 per FTES in the rest of the nation. If California were to raise revenues per FTES to the average for all other states, it would gain $123 per capita (Table 12.3). This change would also result in a significant reduction (more than $3,000) in net support per FTES, part of which could be offset by an increase in student assistance or a greater reliance on federal sources. (California is currently slightly below average in direct student aid, at $1,351 versus $1,537 per student.) It is important to note that this scenario assumes no change in overall higher education participation rates. In sum, achieving one of Californians’ most desired policy goals would be expensive. Potential tradeoffs with other budget priorities often involve difficult choices and may reap less in savings than would be expected. On the other hand, Californians may wish to implement some variant of these changes, particularly if low school staffing ratios harm the least advantaged in our society. For example, in April 2006, more than 70 percent of Californians indicated that they would favor adding support staff (counselors, mentors, and social workers) in low-income schools to reduce dropout rates, even if this change cost the state more money (Baldassare, 2006b). Expanding Coverage in Health and Human Services As noted above, a majority of Californians (52%) would like to dedicate more resources to health and human services (Table 12.1). This category encompasses several types of programs, ranging from cash assistance for low-income families to subsidized medical care for children and their parents to meals and transportation for elderly and disabled individuals, and it is difficult to know which programs in particular are 212 drawing popular support. Moreover, support for expanding health and human services is not necessarily shared across political parties, regions, or other breakdowns of survey respondents. Nevertheless, to understand California’s current fiscal tradeoffs, it is useful to consider an alternative scenario whereby California expands its two primary health and human services programs, Medi-Cal and CalWORKs. Both programs are already characterized by broader coverage than in other states. In 2002, Medi-Cal enrolled 21 percent of the state population and 46 percent of all Californians with incomes below 300 percent of the FPL, compared to shares of 15 and 32 percent, respectively, in the rest of the nation. Similarly, CalWORKs recipients represented 3.4 percent of the overall population and 13.8 percent of families with children and incomes below 200 percent of the FPL, compared to 1.6 and 8.2 percent, respectively, in the rest of the nation. California’s higher enrollment rates may be driven by broader rules about who is eligible for these programs or by individual decisions about whether or not to participate. Although we cannot always separate these two influences, under both CalWORKs and Medi-Cal the state exceeds federally mandated expenditures. For example, in 2002, the maximum allowable CalWORKs grant for a family of three (one adult and two children) and no earnings was $679 per month, the third highest in the nation after Alaska ($923) and Hawaii ($712).5 Higher maximum cash grants increase the pool of potential TANF recipients both because they raise the level of income at which individuals can start receiving benefits and because more individuals may apply for and receive benefits. Similarly, Medi-Cal covers individuals at higher income and resource levels than Medicaid programs in several other large states and provides more benefits that are “optional” under federal program rules. In addition, California has extended coverage to groups who are not traditionally covered under Medicaid, such as working parents with family incomes up to 107 percent of the FPL who were not otherwise eligible. _____________ 5These grants reflect in part higher spending in California before the 1996 federal welfare reform, which “locked in” a certain floor of public spending in the states. 213 For both Medi-Cal and CalWORKs, the per capita costs of expanding coverage even further would be modest relative to some of the changes considered in the previous scenario, for example. The reason is that, despite higher enrollments in each of these programs, California has a lower cost per beneficiary relative to other states. For CalWORKs in 2002, for example, the average cost per recipient was roughly $4,700, compared to $5,100 in the rest of the nation and higher amounts in all of our comparison states except Texas.6 The differences between average Medi-Cal expenditures in California and the rest of the country were even more dramatic, at about $3,100 versus $4,900, respectively. Thus, if California were to adopt a policy like that of New York, offering Medicaid to all working parents with family incomes up to 150 percent of the FPL and adults without children below the poverty line, the state would gain roughly 800,000 enrollees, a 40 percent increase for this group.7 However, because of California’s lower cost per recipient and particularly low relative costs for nonelderly adults ($1,549 in 2002, compared to $3,951 in New York), the cost would be only $1.2 billion, or $35 per capita. Similarly, if California were to expand CalWORKs coverage by 10 percent (it already has the maximum coverage among states in our comparison group), the additional cost would be $660 million, or $19 per capita (Table 12.4). Where might these resources come from? States typically expand Medicaid eligibility to nontraditional groups through Section 1115 waivers from federal program requirements. However, the federal government requires that these demonstration projects be “budget neutral.” States thus often use savings from other program areas such as enhanced managed care enrollment, limited benefits for some groups, or requiring greater cost-sharing for recipients. Given Medi-Cal’s already _____________ 6Recipients are defined here as individuals rather than families. Thus, a family of three with one parent and two children all receiving cash assistance would include three recipients, whereas one including a child and two nonparent relatives who do not receive assistance would include only one recipient. 7If only a fraction of potentially eligible adults applied for and received benefits, the enrollment gain would obviously be smaller. As noted in Chapter 6, the overall take-up rate in California for both Medi-Cal and SCHIP is 78 percent, higher than the average rate for all other states but somewhere between that for Illinois and New York. 214 Table 12.4 Alternative Scenario #2: Expanding Medi-Cal and CalWORKs Raise Medi-Cal maximum allowable incomes for working parents, extend coverage to all adults without children below FPL Increase CalWORKs caseloads per capita by 10 percent Increase selective sales taxes to U.S. average excluding California Reduce central finance and staff to Illinois level Reduce judicial and legal to New York level Expected Change in State and Local Expenditures $ per Capita % Total 35 0.4 19 0.2 –94 –1.1 –18 –0.2 –59 –0.7 high managed care enrollments and cost increases on par with those of traditional fee-for-service (FFS), these savings may be limited for California. However, it is also possible that California would fund such an expansion using new resources. A commonly suggested way to finance health program expansions is by increases in “sin” taxes such as levies on alcoholic beverages and tobacco sales. Although California voters rejected a proposal on the November 2006 ballot to increase cigarette taxes by $2.60 per pack, survey respondents generally favor these taxes (e.g., Baldassare, 2006a, 2005a, 2004a). Chapter 3 shows that, as a group, selective sales tax revenues are lower in California than in other states ($9 versus $12 per $1,000 of personal income). If selective sales tax collections were the same in California as in other states, revenues per capita would climb by $3.3 billion, or $94 per capita (Table 12.4). It is worth noting that California’s taxes on tobacco products in particular are already relatively high ($0.87 per pack, compared to $1.50 in New York, $0.98 in Illinois, $0.41 in Texas, and $0.34 in Florida as of 2002). Moreover, cigarette taxes can discourage consumption and have negative distributional effects. More generally, economists are already critical of California’s current sales and use tax because of its high rate (7.25% statewide, plus local levies) and narrow base (e.g., exemptions for most services as well as utilities and food products). 215 An unpopular area of public spending and one where California differs markedly from other states is government administration. In 2002, California’s per capita spending in this area was $447. California’s expenditures on administration exceeded the average for all other states by nearly 50 percent and were higher than in any of our comparison states including New York (by 15%). Administrative services are necessary to keep government functioning. They are also labor-intensive, suggesting that payroll costs could explain much of the variation between California and other states. However, nonpayroll costs are also high in this area, and differences in salaries are not explained by general labor market conditions or by the location premiums that all employers in California must pay to attract qualified workers. The largest components of administrative expenditures are finance and central staff (e.g., general government support, revenue collection, and the administration of public employee retirement funds) and judicial and legal functions. In 2002, state and local governments in California spent $216 and $201 per capita on these areas, respectively. If California were to reduce per capita expenditures in each category to the level of the next highest state in our comparison group (Illinois for central finance and staff and New York for judicial and legal), it would save $632 million (or $18 per capita) on central finance and staff and $2 billion ($59 per capita) on judicial and legal costs. The breadth of activities included in each of these categories makes it difficult to understand where these cuts might come from and how they might affect government outputs. In sum, health and social services is a complicated area in which to consider alternative tradeoffs because of complex and often interlocking federal program requirements. However, the above scenario reinforces a key finding of this report, namely, that California enrolls larger shares of its population in Medicaid and TANF than other states do and it incurs lower expenditures per case. As a result, hypothetical program expansions would not be very costly in per capita terms relative to some of the other scenarios presented here. These calculations do not take into account differences between new enrollees and current beneficiaries of these programs. Differences in 216 health care use patterns among these groups, for example, could alter the conclusions presented here (e.g., MaCurdy et al., 2005). In addition, we have not addressed interactions between public and private health care markets, including potential cross-subsidies, which might be affected by changes in public programs. Investing in Roads and Infrastructure Twenty-seven percent of Californians would like the state to spend more on roads and infrastructure (Table 12.1). When asked what types of infrastructure should receive funding priority, survey respondents typically choose school facilities (e.g., 40% in August 2004), surface transportation (24%), and water systems (21%) (Baldassare, 2004b). These three sectors represent the bulk (85%) of $220 billion in proposed spending under the governor’s ten-year strategic growth plan (Office of the Governor, 2006) and more than 90 percent of borrowing approved by California voters in the November 2006 election. We find that California’s capital outlays in these areas have lagged those of other states in recent years. For example, in 2002, California spent roughly half as much as the average for all other states on higher education facilities ($35, compared to $65 per capita). Although it spent more than other states on K–12 educational facilities ($201 versus $182 per capita), this situation was without precedent since the 1970s. Similarly, capital outlays on mass transit were similar to the national average excluding California but those on highways lagged the average for other states ($154, compared to $240 per capita) and expenditures in every state within our comparison group. Capital spending on water supply and quality was higher than average ($77, compared to $53 per capita) but lower than spending in Florida ($97 per capita) and other western states. More important, California’s capital investments have often failed to keep pace with growing demand. From 1992 to 2002, California experienced increases in K–12 and higher education enrollment comparable to those of Texas and Florida. However, its capital investment per additional student was lower than in either of these states 217 or the national average excluding California.8 Similarly, in transportation, California’s recent population growth and trends in vehicle miles traveled mirror those of Florida and Texas, yet its addition of freeway lane miles has been modest, more akin to that of New York. As a result, California’s freeways have been ranked among the most congested in the nation (Schrank and Lomax, 2005). The experiences of other states provide a range of alternative scenarios. If California were to fund highway construction at the level of another fast-growing state with high automobile use, such as Florida, state and local annual expenditures would rise by $4.2 billion, or $120 per capita (Table 12.5). If it were to invest in higher education facilities at a level comparable to another state with a large public college and university system and rising enrollments, such as Texas, spending would be $42 per capita higher. If California’s capital investments in water supply and other natural resources were equivalent to those of Florida in 2002, its annual expenditures would increase by $19 per capita. How big are these changes? California state and local governments might offset increases in highway capital spending with reductions in mass transit. If California reduced these expenditures to the level of another state with high annual vehicle miles traveled, Texas, it would produce savings of $4.4 billion, or $127 per capita (Table 12.5). California could also charge user fees for mass transit comparable to those in Illinois ($59 per capita) rather than the current $36 per capita in fares. If it followed this course, California would raise an additional $805 million ($23 per capita) (Table 12.5). Yet another alternative would be to increase highway-related taxes and fees, including state _____________ 8The funding picture for educational facilities is changing because of several recent ballot propositions. In 1998, voters authorized $9 billion in general obligation bonds for the construction and modernization of buildings in public schools, colleges, and universities (Proposition 1A). In 2000, they passed Proposition 39, lowering the vote requirement for local bonds for these purposes from two-thirds to 55 percent. In 2002, voters passed Proposition 47, which authorized the state to sell bonds worth $13 billion for capital improvements to public schools and public colleges and universities. In 2004, they passed Proposition 55, which authorized $21 billion for the same purposes, and in 2006, voters approved Proposition 1D, authorizing $10.4 billion in bonds for K–12 and higher education facilities. 218 Table 12.5 Alternative Scenario #3: Investing in Roads and Other Infrastructure Increase highway capital expenditures to level in Florida Increase higher education capital spending to level in Texas Invest in natural resources and water supply at level of Florida Reduce expenditures per capita on mass transit to level in Texas Increase transportation user fees to level in Illinois Increase fuel tax to average for U.S.–California Expected Change in State and Local Expenditures $ per Capita % Total 120 1.4 42 0.5 19 0.2 –127 –23 –33 –1.5 –0.3 –0.4 gasoline taxes, vehicle license fees, mileage and weight fees on motor carriers, and driver’s license fees. In 2002, these taxes and fees amounted to roughly $153 per capita in California, compared to $186 per capita on average in all other states. If California had the same such tax burden as in the rest of the nation, it would collect an additional $33 per capita, or $1.2 billion (Table 12.5). In sum, raising annual capital expenditures on California’s infrastructure would not require a large infusion of spending, although eradicating years of lower-than-average spending may require additional resources. California’s capital spending, although uneven across categories, is not widely off the mark of the average for other states. Conclusions These tradeoffs have served to reinforce certain themes of our report. In many areas, the state faces higher-than-average caseloads. This may be due to demographics (as in the case of K–12 education, for example), policy decisions (as in Medi-Cal), or both (CalWORKs and higher education, for example). Like all employers in California, state and local governments must pay higher wages to attract employees who can afford to live in the state. However, governments often pay more than the 219 premium required at comparable skill and experience levels. Higher payroll expenditures translate into lower staffing ratios in K–12 education, police, and fire protection. The state also has relatively low user fees in certain areas (higher education and mass transit) and higher administration costs overall. This is an opportune time for state residents to reexamine these choices and to consider alternative tradeoffs. 220 Appendix Data Sources Chapter 2 State Rankings State geography, population, and economy data and rankings reported in Table 2.2 come from the U.S. Census Bureau, Statistical Abstract of the United States: 2006, available at http://www. census. gov/prod/www/statistical-abstract.html. Wage Data Wage data come from the “Occupation Employment Statistics” (OES) survey, available at www.bls.gov/OES/. The United States Bureau of Labor Statistics administers this survey semi-annually to every employer in the United States with more than 250 employees and to a random sample of smaller firms. Wage estimates for 2002 are based on responses from 1999 through 2002 for a total sample size of approximately 1.4 million establishments. The nationwide response rate in 2002 was 78 percent, covering 71 percent of employment. Job Zones Job Zones (Tables 2.3–2.5) come from the O*NET or Occupational Information Network database, version 7.0, available at http://www. onetcenter.org/database.html. This database combines information on worker and occupation characteristics from occupational analysts and surveys of employees, their supervisors, and co-workers. We weighted average salaries in each zone by the national employment shares in each occupation. We use national rather than state employment shares as weights so that our location premium reflects only price differences and not the composition of labor markets in each state. 221 Payroll Public employment and payroll expenditures used throughout the report come from the Census of Governments Employment Phase in 2002, available at www.census.gov/govs/www/apcs.html. The Census Bureau canvasses all U.S. state and local governments in Census years (ending in 2 or 7) and all states and a sample of local governments annually. Statistics for part- and full-time employees are reported for the same functional categories used in the Finance Phase (de Alth and Haskel, 2007). Gross payrolls include all salaries, wages, fees, overtime, bonuses, and awards paid to employees as of the March 12 pay period. To convert these monthly payrolls to yearly payrolls, the Census Bureau payroll total was multiplied by 11.77 (365 days/31 days). Fringe benefits including health and life insurance, Social Security, and retirement contributions are not included. Chapter 3 Personal Income Personal income data (Tables 3.3, 3.4, and 3.7) come from the Bureau of Economic Analysis, SA04 current series, available at http:// www.bea.gov/bea/regional/spi/default.cfm?satable=SA04&series=ancillary. As noted in the chapter, this series excludes some types of income that are part of the tax base, such as capital gains, pension benefit payments, and personal contributions to social insurance programs. It includes other sources of nontaxable income, such as the nontaxable portion of Social Security benefits and certain investment income. It is not clear how adjusting for these factors would affect California’s relative position. Tax Rates and Incidence State tax rates and provisions (Tables 3.5 and 3.6) are from the Federation of Tax Administrators (FTA), available at http://www. taxadmin.org/fta/rate/tax_stru.html, and the Urban-Brookings Tax Policy Center, available at http://www.taxpolicycenter.org/TaxFacts/ tfdb/TFTemplate.cfm? topic2id=90. In the case of sales tax rates, the highest local rate is that known to be actually levied by at least one 222 jurisdiction. Average tax rates (e.g., in Table 3.6) are not weighted by population or sales. Income thresholds for state income taxes are from Johnson, Llobrera, and Zahradnik (2003), Table 3, also available at http://www. taxpolicycenter.org/TaxFacts/tfdb/TFTemplate.cfm?topic2id=90. Information on income tax progressivity is based on reports to the California Franchise Tax Board and comparable bodies in New York and Illinois. Sales taxes on services are compiled from a survey undertaken in 2004 by the FTA, available at http://www.taxadmin.org/fta/pub/services/ services04.html#summary. Chapter 4 Children Eligible for Public School Children eligible for public school (Table 4.4) is from Census 2000, Table QT-P19, School Enrollment: 2000, available at http://factfinder. census.gov/servlet/QTTable. It is the sum of public and private school enrollment in kindergarten, elementary school, and high school plus the population ages 16 to 19 who are not enrolled in school and are not high school graduates. Employees Employees in elementary and secondary education (Tables 4.6 and 4.7) are from the National Center for Education Statistics, State Nonfiscal Survey of Public Elementary and Secondary Education, 2001–02, Common Core of Data, available at http://nces.ed.gov/ccd/ccddata.asp. Staff are measured in full-time-equivalents. Support staff in Table 4.7 consists of instructional aides, instructional coordinators, guidance counselors, librarians, library support staff, and student support staff. Administrative staff in Table 4.7 consists of local education agency (LEA) administrators, LEA administrative support staff, school administrators, and school administrative support staff. Other support staff in Table 4.7 includes the category “all other staff” in the NCES survey. 223 Salaries and Benefits of Public School Employees Compensation for public school employees (Tables 4.5, 4.6, and 4.8) is from the National Center for Education Statistics, National Public Education Financial Survey, School Year 2001–02, Common Core of Data, available at http://nces.ed.gov/ccd/ccddata.asp. These salaries and benefits are the sum of salaries and benefits for the following four areas: instructional expenditures, support expenditures, noninstructional food service expenditures, and noninstructional enterprise expenditures. School Enrollments Public school enrollments (Tables 4.3–4.7 and 4.10 and Figures 4.2–4.4) are from the National Center for Education Statistics, State Nonfiscal Survey of Public Elementary and Secondary Education, 2001– 2002, Common Core of Data, available at http://nces.ed.gov/ccd/ ccddata.asp. In Table 4.4, public and private school enrollment is from Census 2000, Table QT-P19, School Enrollment: 2000, available at http://factfinder.census.gov/servlet/QTTable. Stanford Achievement Test, Ninth Edition SAT9 scores are available at the web site of the California Department of Education for its Standardized Testing and Reporting Program (http://star.cde.ca.gov/). Average Scores on NAEP Average NAEP scores for different states and groups of students (Tables 4.11 and 4.12) are from the NAEP website, available at http://nces.ed.gov/nationsreportcard/naepdata/search.asp. Chapter 5 Higher Education Enrollment Data The National Center for Education Statistics provided enrollment figures. These data were extracted from the Integrated Postsecondary Education Data System (IPEDS) using the Dataset Cutting Tool accessed via the IPEDS website, available at http://nces.ed.gov/ipedspas. 224 For each state, enrollment figures were collected for public two- and four-year, degree-granting institutions for the fall 2002 period. Full-Time-Equivalent Students The IPEDS dataset derives full-time-equivalent student numbers (Tables 5.4, 5.6, 5.9, 5.11, and 5.12; Figures 5.3, 5.5, 5.6, and 5.7) from reported full- and part-time enrollment for the fall period. A formula is used to convert an institution’s part-time enrollment into full-timeequivalents and then that number is added to the reported full-time enrollment of the institution. The full-time-equivalent of part-time enrollment is estimated by multiplying reported part-time enrollment by factors that vary by control, level of institution, and level of student. The formula used is published annually in the Digest of Education Statistics. In the case of California, this formula produces a smaller number of FTES than those reported by the three state systems. Student Migration Figures for enrollment in home state versus out-of-state schools (Table 5.5) were derived from Table 203, “Residence and Migration of All Freshmen Students in Degree-Granting Institutions, by State or Jurisdiction, Fall, 2002” in the Digest of Education Statistics, 2004. Degrees Awarded IPEDS provides data on the number of degrees completed in a given academic year (in this case, July 1, 2002, and June 30, 2003). Data were collected for associate’s, bachelor’s, master’s, doctorates, and professional degrees (Tables 5.10, 5.11, and 5.12). Awards for programs shorter than two years were not included in the total. Chapter 6 Medicaid Expenditures, Recipients, Enrollment, and Service Use Medicaid expenditures, recipients, and enrollment data (Tables 6.2 and 6.4–6.9) are from the Medicaid Statistical Information System (MSIS) State Summary Datamart (SSD), an exclusive web-based 225 Medicaid data analysis tool from the Centers for Medicare & Medicaid Services, available at http://www.cms.hhs.gov/MSIS/. Medicaid participants are recipients of Medicaid services during the fiscal year. Medicaid enrollment is the total count of recipients of Medicaid during the fiscal year, excluding those enrolled in Medicaid expansion SCHIP programs and those enrolled in Family Planning Services programs under a section 1115 waiver. Medicaid expenditures measure payments for claims adjudicated during the fiscal year. Population with Income Below 300 Percent of Federal Poverty Level Estimates of the state population composition in Table 6.3, and estimates of state population groups with incomes below 200 and 300 percent of FPL (Tables 6.4–6.6) are the authors’ calculations using the Census Bureau’s American Community Survey 2002, available at www.census.gov/acs/www. Children ever enrolled in SCHIP in Table 6.6 (both Medicaid expansion SCHIP and Separate SCHIP) are obtained from the Centers for Medicare & Medicaid Services, SCHIP Enrollment Report FY2002, available at http://www.cms.hhs.gov/schip/enrollment/schip02.pdf. Take-Up Rate and Eligibility Rate Monthly take-up rate and eligibility rates (Table 6.7) are from the Office of Health Policy, U.S. Department of Health and Human Services, TRIM3’s 2001 Baseline Simulation of Medicaid and SCHIP Eligibility and Enrollment: Methods and Results, Table 4, available at http://aspe.hhs.gov/health/reports/05/medicaid-schip-simulation/. Managed Care Enrollment and Payment Managed care enrollment and capitated payments (Table 6.9) are from the Centers for Medicare and Medicaid Services, National MSIS Tables by States—2002, Tables 10 and 17, available at http://www.cms. hhs.gov/medicaid/msis/tables2002.asp. 226 Medicaid Fee Indexes and Medicaid-Medicare Fee Indexes Table 6.10 and Table 6.11 are extracted from “Exhibit 2: Medicaid Fee Indexes and Medicaid-to Medicare Fee Indexes 2003,” in Zuckerman et al. (2004). The computation of Medicare obstetrics fees uses the formulas (conversion factors and geographic adjusters) for Medicare fees based on relative value units (RVUs) in the Medicare Physician Fee Schedule. Although not many Medicare patients get obstetrics care, the RVUs are created to be service-specific, and do not take into account patient characteristics. Charges of Own Public Hospitals Charges of publicly administered hospitals (Table 6.12) include charges from patients, private insurance companies, public insurance programs (such as Medicare) of public hospitals and of institutions for care and treatment of the handicapped, and receipts from hospital canteens, cafeterias, gift shops, etc. Outcome Measures Performance measures of preventable hospitalizations, birth outcomes, and access to care (Tables 6.13–6.15) are extracted from Billings and Weinick (2003), available at http://www.ahrq.gov/data/ safetynet/netfact.htm. Chapter 7 Low-Income Population All low-income population figures (Tables 7.4 and 7.14) were estimated using the U.S. Bureau of the Census March 2002 Current Population Survey (CPS). Low income was defined at 200 percent of the FPL. SSI Caseload and Aid Payment and recipient data (Tables 7.12 and 7.14) come from the Social Security Administration’s Supplemental Security Income Annual Statistical Supplement and Annual Statistical Supplement, available at http://www.ssa.gov/policy/docs/statcomps/supplement/2003/index.html. 227 TANF, CCDF, and SSBG Caseload and Financial Information The analysis of welfare, child care, and community-based services (Tables 7.2–7.5, 7.6, 7.9–7.11) draws from several publications of the U.S. Department of Health and Human Services’ Administration for Children and Families (ACF): TANF Annual Reports to Congress, CCDF data tables, SSBG Annual Reports on Expenditures and Recipients, and SSBG Focus Reports. All these publications are available at http://www. acf.hhs.gov. Recipiency rates in Table 7.3 were calculated by adding the number of recipients from services funded with TANF and CCDF. This potentially induces some double-counting and underestimates benefits per recipient. Exclusive use of TANF expenditure and caseload data increases the variability in recipiency rates and benefits per case but leaves the qualitative conclusions unchanged. It is important to keep in mind that in Table 7.3, the number of recipients for state supplementation is assumed to be exactly the same as the total number of recipients. This is not necessarily true, but it is not possible to have a detailed breakdown of the respective recipient populations with the available information. Adoptions and Foster Care The Green Book reports only federal funds expended through the matching grants authorized under Title IV, Sections B and E, of the Social Security Act. However, it was possible to impute state expenditures from the matching requirements associated with each type of service and, with them, total spending. Caseload figures for these programs (Tables 7.7 and 7.8) were extracted from ACF’s Child Welfare Outcomes Annual Report. SSI/SSP The Social Security Administration’s data (Tables 7.12–7.14) include payments made by the federal government, which the Census Bureau does not count toward each state’s expenditures. As a consequence, the sum of expenditures in the three programs presented in 228 this chapter exceeds, by construction, the total of cash assistance, categorical assistance, and other welfare from the Census categories. Chapter 8 Lane Miles of Freeways and Other Principal Arterials Lane miles of freeways and other principal arterials (Tables 8.13–8.15 and Figure 8.5) are from Table HM-60, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm. These are the sum of lane miles for rural interstates, other rural principal arterials, urban interstates, other urban freeways and expressways, and other urban principal arterials. Lane Miles of Local and State Highways State highway lane miles (Table 8.18) are from Table HM-81, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/ hss/hsspubs.htm. Local highway lane miles (Table 8.17) are total lane miles from Table HM-60 less state highway lane miles from Table HM81. Operating Costs and Employee Compensation for Public Transit Operating costs and employee compensation for public transit (Tables 8.6–8.9) are from the 2002 National Transit Database of the Federal Transit Administration, available at http://www.ntdprogram. com. Passenger Trips and Passenger Miles Passenger trips and passenger miles (Tables 8.5–8.8) are from the 2002 National Transit Database of the Federal Transit Administration, available at http://www.ntdprogram.com. 229 Population Density of Urbanized Areas The percentage of a state’s population in urbanized areas and the density of urbanized areas are from the Census 2000 Summary File 1, Table GCT-PH1, Population, Housing Units, Area, and Density: 2000. Road Congestion and Delay Data on road congestion and delay in urban areas (Table 8.12 and Figure 8.6) are from The 2005 Urban Mobility Report, Texas Transportation Institute, May 2005, available at http://mobility.tamu. edu. Vehicle Miles of Travel Vehicle miles of travel (Tables 8.14 and 8.15) are from Table VM-2, Highway Statistics, Federal Highway Administration, U.S. Department of Transportation, available at http://www.fhwa.dot.gov/policy/ohpi/hss/ hsspubs.htm. Chapter 9 Employment and Payroll Throughout the chapter, payroll costs per employee (Tables 9.4, 9.5, 9.9, 9.10, 9.12, and 9.13) are calculated by dividing total payroll by the number of full-time-equivalent employees in March of each year. The sources are described above. Confined Population and People Under Probation or Parole The U.S. Department of Justice’s Bureau of Justice Statistics provided a variety of documents for complementary data. Counts on people in jails, in prisons, and under probation or parole (Figure 9.2 and Tables 9.3 and 9.4) come from Bureau of Justice Statistics available at http://www.ojp.usdoj.gov/bjs/dtdata.htm#ncrp, for data before 1999, as well as from the Annual Survey of Jails and the Sourcebook of Criminal Justice Statistics for 1999–2002. 230 Correctional Facilities Information on the number of state prisons and local jails (Table 9.6) comes directly from each state’s Department of Corrections. The rated and design capacity of both state prisons and local jails was extracted from the Department of Justice, Prisoners in 2002, and Sourcebook of Criminal Justice Statistics, respectively. “Design capacity” refers to how many inmates the original design of a facility was intended in to house. “Rated capacity” is a more subjective measure determined by officials within each jurisdiction. Crimes The Federal Bureau of Investigation’s Uniform Crime Reports was the source for the number of crimes (Table 9.9). We used the number of criminal offenses for the analysis in this chapter. Criminal offenses encompass more incidents than those included in the FBI’s commonly used crime index. The latter uses information of offenses cleared by arrest or by exceptional means and which generally entail higher levels of seriousness. Examples include criminal homicide, aggravated assault, forcible rape, and arson, among others. In addition to the number of incidents the crime index comprises, the number of criminal offenses includes nonviolent crimes, such as assaults without weapons, fraud, vandalism, drug abuse violations, driving under the influence, etc. Chapter 10 HUD Income Limits, Rent Burdens, Wait Lists, and Fair Market Rents Income limits for public housing and other housing assistance (Table 10.4) come from the U.S. Department of Housing and Urban Development, FY2002 HUD Income Limits Briefing Material, available at http://www.huduser.org/DATASETS/il/fmr02/index.html. These data were merged with income data by family characteristics from the 2002 March CPS. Data on the receipt of housing assistance also come from the CPS’s HPUBLIC and HRENT questions. Data on rent burdens (Table 10.5) come from the 2000 Census of Population and Housing. HUD Fair Market Rents are taken from http://www. 231 huduser.org/datasets/fmr.html. Data on average waiting times for housing programs (Table 10.6) are taken from “Picture of Susidized Households, 1998,” available at http://www.huduser.org/datasets/ assthsg/statedata98/. Fair market rents in 2002 (Table 10.7) come from the HUDuser website, available at http://www.huduser.org/datasets/ fmr/fmr2002/index.html. Park Acreage and Visits For Tables 10.13 and 10.14, the source for local acreage was State Comprehensive Outdoor Recreation Plans (SCORPs) reported by states to the National Park Service. For state park acreage, we drew on the 2003 Annual Information Exchange by the National Association of State Park Directors, available at http://www.naspd.org. National park data came from U.S. Fish and Wildlife Service, Land Areas Report, 2004, and data from the Land Resource Division of the U.S. National Park Service, available at http://ca.rand.org/stats/community/parkvisitsCA.html. Data on visits for state and national parks came from the 2005 Statistical Abstract, Table 1250. Chapter 11 Expenditures Expenditures for finance and central staff, judicial and legal, general public buildings, and legislative (Tables 11.1–11.3, and 11.5 and Figures 11.1–11.3) are from the Census of Governments. Expenditures on finance and central staff combine the Census categories of financial administration (the Census Bureau series E23, F23, and G23) and central staff services (E29, F29, and G29). Legal and judicial expenditures are the sum of Census categories E25, F25, and G25. Legislative expenditures are the sum of E26, F26, and G26. Expenditures on general public buildings are the sum of E31, F31, and G31. Employees and Payroll Employees and payroll expenditures (Tables 11.2–11.7) come from the employment portion of the 2002 Census of Governments, available 232 at http://www.census.gov/govs/www/apes.html. The Census Bureau measures the number of employees in full-time-equivalents and the gross payroll for the month of March. Population When expenditures are expressed in per capita terms (Tables 11.1–11.3, 11.5, and 11.6 and Figures 11.1–11.3), the denominator is the population estimate by states released by the U.S. Census Bureau on December 22, 2004. 233 Table A.1 Urban Areas in Texas Transportation Institute (TTI) Study, 2002 Population (1,000s) % of State Population California Bakersfield Oxnard-Ventura Fresno Sacramento Riverside–San Bernardino San Jose San Diego San Francisco–Oakland Los Angeles–Long Beach–Santa Ana All urban areas in TTI study Florida Pensacola Cape Coral Sarasota-Bradenton Jacksonville Orlando Tampa–St. Petersburg Miami All urban areas in TTI study Illinois Chicago New York Albany-Schenectady Rochester Buffalo New York–Long Islanda All urban areas in TTI study Texas Beaumont Brownsville Laredo Corpus Christi El Paso Austin San Antonio Houston Dallas–Fort Worth–Arlington All urban areas in TTI study 425 560 585 1,560 1,610 1,675 2,825 12,400 12,400 25,760 310 315 560 905 1,255 2,025 5,000 10,370 8,120 525 660 1,115 12,736 15,036 145 160 190 320 665 835 1,280 3,720 4,200 11,515 1.3 1.7 1.7 4.6 4.8 4.8 8.3 36.6 76.1 76.1 1.9 2.0 3.5 5.7 7.9 12.7 31.3 64.9 65.4 2.8 3.5 5.9 67.1 9.2 0.7 0.8 0.9 1.5 3.2 4.0 6.1 17.8 20.1 55.2 aStatistics for New York CMSA prorated to portion in New York State. 234 References Auerbach, Richard, “Estimates of Los Angeles County’s Commercial Property Values Too High in Proposition 13–Related Study,” press release, Los Angeles County Assessor, Los Angeles, California, February 25, 2003. Bailey, Amanda, and Joseph P. Hayes, “Who’s in Prison? The Changing Demographics of Incarceration,” California Counts, Vol. 8, No. 1, Public Policy Institute of California, San Francisco, California, August 2006. Baldassare, Mark, PPIC Statewide Survey: May 1998, Public Policy Institute of California, San Francisco, California, May 1998. Baldassare, Mark, PPIC Statewide Survey: Californians and Their Government, Public Policy Institute of California, San Francisco, California, February 2003. Baldassare, Mark, PPIC Statewide Survey: Special Survey on California’s Fiscal System, Public Policy Institute of California, San Francisco, California, January 2004a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on California’s Fiscal System, Public Policy Institute of California, San Francisco, California, August 2004b. Baldassare, Mark, PPIC Statewide Survey: Special Survey on the California State Budget, Public Policy Institute of California, San Francisco, California, May 2005a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Californians and the Initiative Process, Public Policy Institute of California, San Francisco, California, November 2005b. Baldassare, Mark, PPIC Statewide Survey: Special Survey of Los Angeles, Public Policy Institute of California, San Francisco, California, March 2005c. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Education, Public Policy Institute of California, San Francisco, California, April 2005d. 235 Baldassare, Mark, PPIC Statewide Survey: Special Survey on the California State Budget, Public Policy Institute of California, San Francisco, California, January 2006a. Baldassare, Mark, PPIC Statewide Survey: Special Survey on Education, Public Policy Institute of California, San Francisco, California, April 2006b. Billings, John, and Robin M. Weinick, eds., Monitoring the Health Care Safety Net—Book I: A Data Book for Metropolitan Area, Agency for Healthcare Research and Quality, AHRQ Publication No, 030025, 2003, available at http://www.ahrq.gov/data/safetynet/ netfact.htm. California Postsecondary Education Commission, Fiscal Profiles, 2004, Commission Report 04-20, Sacramento, California, December 2004. California State Board of Equalization, “Staff Legislative Bill Analysis— SB 1662 (Peace),” Sacramento, California, 2002. California Tax Credit Allocation Committee (CTAC), “A Description of California Tax Credit Allocation Committee Programs,” California Office of the Treasurer, Sacramento, California, 2006, available at http://www.treasurer.ca.gov/ctcac/program.pdf. Carroll, Stephen J., Cathy S. Krop, Jeremy Arkes, Peter A. Morrison, and Ann Flanagan, California’s K–12 Schools: How Are They Doing? RAND Corporation, Santa Monica, California, 2005. Centers for Medicare & Medicaid Services, Medicaid At-a-Glance 2003, available at http://www.cms.hhs.gov/states/maag2003.pdf. de Alth, Shelley, and David Haskel, Data Appendix, Public Policy Institute of California, San Francisco, California, 2007. Gallup Organization, “Gallup Poll: April Wave 3” and “May Wave 2,” 2003, available at http://brain.gallup.com/. General Accountability Office, Medicaid: Spending Pressures Drive States toward Program Reinvention, GAO/HEHS-95-122, 1995, available at http://www.gao.gov/archive/1995/he95122.pdf. General Accountability Office, Sales Taxes: Electronic Commerce Growth Presents Challenges; Revenue Losses Are Uncertain, GGD/OCE-00165, 2000, available at http://www.gao.gov/archive/2000/g600165. pdf. Glied, Sherry, “Managed Care,” Chapter 13, Volume 1A, in Anthony J. Culyer and Joseph P. Newhouse, eds., The Handbook of Health Economics, North-Holland, Amsterdam, 2000. 236 Greulich, Erica, and John M. Quigley, “Housing Subsidies and the Tax Code: The Case of Mortgage Credit Certificates,” unpublished manuscript, April 2003. Hanak, Ellen, and Elisa Barbour, “Sizing Up the Challenge: California’s Infrastructure Needs and Tradeoffs,” in Ellen Hanak and Mark Baldassare, eds., California 2025: Taking on the Future, Public Policy Institute of California, San Francisco, California, 2005. Hanushek, Eric A., “Assessing the Effects of School Resources on Student Performance: An Update,” Education Evaluation and Policy Analysis, Vol. 19, No. 2, 1997, pp. 141–164. Jepsen, Christopher, and Steven Rivkin, Class Size Reduction, Teacher Quality, and Academic Achievement in California Public Elementary Schools, Public Policy Institute of California, San Francisco, California, 2002. Johnson, Nicholas, Joseph Llobrera, and Bob Zahradnik, A Hand Up, How State Earned Income Tax Credits Help Working Families Escape Poverty in 2003, Center on Budget and Policy Priorities, Washington, D.C., March 3, 2003. Jones, Martha, The Streamlined Sales and Use Tax Agreement: A California Perspective, California Research Bureau, Sacramento, California, February 2005. The Kaiser Commission on Medicaid and the Uninsured, Medicaid Benefit, available at http://www.kff.org/medicaidbenefits/index.cfm. The Kaiser Commission on Medicaid and the Uninsured, The Medicaid Resource Book, July 2002, available at http://www.kff.org/medicaid/ 2236-index.cfm. Krueger, Alan B., “Economic Considerations and Class Size,” Economic Journal, Vol. 113, 2003, pp. 34–63. Land Trust Alliance, “National Land Trust Census,” available at http://www.lta.org/aboutlt/census.shtml. Legislative Analyst’s Office, California Travels: Financing Our Transportation, Sacramento, California, May 2000. Legislative Analyst’s Office, Analysis of the 2001–02 Budget Bill, Sacramento, California, 2002. Legislative Analyst’s Office, Analysis of the 2002–03 Budget Bill, Sacramento, California, 2003. Legislative Analyst’s Office, Cal Facts: California’s Economy and Budget in Perspective, Sacramento, California, December 2004. 237 Legislative Analyst’s Office, Revenue Volatility in California, Sacramento, California, January 2005. Legislative Analyst’s Office, A Primer: Funding Higher Education, Sacramento, California, May 15, 2006a. Legislative Analyst’s Office, Overview of the 2006–2007 May Revision, Sacramento, California, May 15, 2006b. Lewin Group, Comparing Physician and Dentist Fees among Medicaid Programs, Fairfax, Virginia, 2001. Mackenzie, Doris Layton, Sentencing and Corrections in the 21st Century: Setting the Stage for the Future, U.S. Department of Justice Report 189089, July 2001. MaCurdy, Thomas, David C. Mancuso, and Margaret O’Brien-Strain, Does California’s Welfare Policy Explain the Slower Decline of Its Caseload? Public Policy Institute of California, San Francisco, California, January 2002. MaCurdy, Thomas, Raymond Chan, Rodney Chun, Jans Johnson, and Margaret O’Brien-Strain, Medi-Cal Expenditures: Historical Growth and Long Term Forecasts, Public Policy Institute of California, San Francisco, California, 2005. Marrufo, Grecia, Margaret O’Brien-Strain, and Helen Oliver, Child Care Price Dynamics in California, Public Policy Institute of California, San Francisco, California, December 2003. Mazerov, Michael, Developing the Capacity to Analyze the Distributional Impact of State and Local Taxes: Issues and Options for States, Center for Budget and Policy Priorities, Washington, D.C., January 2002. McDonald, Steve, 2003–2004 Florida Annual Policy Survey Report, Survey Research Laboratory, Talahassee, Florida, 2004, available at http://www.fsu.edu/~survey/FAPS/#Reports. National Association of Realtors, “Median Sales Price of E