Donate
Independent, objective, nonpartisan research

R 809SLR

Authors

R 809SLR

Tagged with:

Publication PDFs

Database

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

View live version

object(Timber\Post)#3711 (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_809SLR.pdf" ["wpmf_size"]=> string(6) "374699" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(77644) "Special Education Financing in California A Decade After Reform Stephen Lipscomb with contributions from Karina Jaquet Supported with funding from The William and Flora Hewlett Foundation August 2009 The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research on major economic, social, and political issues. The institute’s goal is to raise public awareness and to give elected representatives and other decisionmakers a more informed basis for developing policies and programs. The institute’s research focuses on the underlying forces shaping California's future, cutting across a wide range of public policy concerns, including economic development, education, environment and resources, governance, population, public finance, and social and health policy. PPIC is a private operating foundation. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. PPIC was established in 1994 with an endowment from William R. Hewlett. Mark Baldassare is President and Chief Executive Officer of PPIC. Walter B. Hewlett is Chair of the Board of Directors. Copyright © 2009 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. Research publications reflect the views of t he authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. . Contents Acronyms 4 Summary 5 Acknowledgments 7 INTRODUCTION 8 FINANCING SPECIAL EDUCATION IN CALIFORNI A 10 The Funding Process 10 Funding Levels 11 DISABILITY RATES AND INCOME 14 Findings from the California Health Interview Survey 15 Findings for Children in Special Education Programs 16 SPENDING ON CHILDREN WITH DISABILITIES IN CALIFORNIA 18 Conceptual Framework 18 Expenditure Totals for 2006–07 20 Additional Spending on Disabled Children and Local Support 22 Spending and Regional Non-Teacher Wages 24 SUGGESTIONS FOR IMPROVING SPECIAL EDUCAT ION FINANCE 27 Refine the Allocation Model 27 Clarify the State’s Objective for Special Education Funds 29 References 31 About the Author 33 4 Acronyms AB 602 Assembly Bill 602 (1997) ADA Average Daily Attendance ADD/ADHD Attention Deficit Disorder / Attention Deficit Hyperactivity Disorder AIR American Institutes for Research ARRA American Recovery and Reinvestment Act (2009) CASEMIS California Special Education Management Information System CBEDS California Basic Education Data System CBSA Core Based Statistical Area CDE California Department of Education CHIS California Health Interview Survey COE County Office of Education CWI Comparable Wage Index ESEA Elementary and Secondary Education Act (1965) IDEA Individuals with Disabilities Education Act JPA Joint Powers Agreement LEA Local Education Agency MSA Metropolitan Statistical Area NCES National Center for Education Statistics SACS Standardized Account Code Structure SDA Special Disabilities Adjustment SEEP Special Education Expenditure Project SELPA Special Education Local Plan Area USD Unified School District 5 Summary This report assesses California’s special education finance policy and suggests improvements, about a decade after California overhauled special education financing to address concerns about the efficacy of the previous system. That overhaul , Assembly Bill 602 (1997), sought t o ensure greater funding equity, eliminate inappropriate placement incentives, and streamline the funding model, among other objectives . We find that finance reform has led to positive changes but that the state can do more to implement its desir ed reform goals. Special education programs help in educating California’s children with disabilities, who represent one in ten public school students statewide. Program expenditures amounted to $9.3 billion in 2006 –07, or more than 16 percent of K –12 general fund spending (Lipscomb, 2009b). Special education differs from most educational programs because children with disabilities in the United States are legally entitled to free, appropriate services based on their individual needs. T his service entitlement, along with the magnitude of expenditures and earlier finance reform , makes special education an important part of California’s education finance policy. Special education aid from federal, state, and local source s, at $4.7 billion in 2006 –07, is California’s largest pool of funding specifically for one education program. California allocate s most of this funding on a per -student basis, regardless of disability status; t he underlying assumption is that disabilities vary evenly across the population. Since enacting this funding model, California has reduced but not eliminated historical inequities in the per -student funding rate across the state. Inequities in funding rates reflect historical circumstances and are no t consistent with the type of special education finance system that California chose to adopt in 1997 . By design, California’s special education finance model does not reimburse school districts for expenditures related to greater reported special education needs. This policy encourages districts to serve students c ost-effectively , but it also means that the local share of spending on special education mandates can diff er substantially across the state. This report suggests that California consider funding rate equalization, with adjustments based on factors outside of district control —factors that arguably describe true cost variation. Such adjustments could partly account for differences in costs but without giving districts an incentive to over- represent special education needs. P recedent for this kind of funding adjustment comes from the federal government, which allocates part of its special education funding based on poverty. Within California, severe disability rates tend to be higher among children from lower -income families. California could adapt the federal model, and could add other factors too , such as v arying labor market conditions . Per -pupil spending, which is primarily for salaries, is sensitive to regional variation in labor market conditions for non-teachers with similar qualifications, even holding disability rates constant. These potential adjustment factors would describe different types of costs to meet special education mandates, with the first arguably related to incidence and the second related to the expected price of employee compensation . 6 Following this assessment, t his report then simulates how California could use the principles of the cu rrent system to allocate funds at an equal rate per student, adjusted by these two factors. The simulation adjusts for low-income stu dents and regional labor market conditions , but California could substitute other factors, such as an updated version of th e formula’s existing cost proxy . T hese refinements would redistribut e how the state allocates existing funds, but would also lead to fuller implementation of current funding objectives . All technical appendices to this paper are available on the PPIC website: http://www.ppic.org/content/pubs/other/809SLR_appendix.pdf 7 Acknowledgments I would like to thank several people who helped me learn about California’s special educ ation program, including P aul Goldfinger, Sarge Kennedy, Jack Lucas, Marcia McClish, Rick Miller, Sally Spaeth, Ray Reinhard, and Paul Warren. I also deeply appreciate the insightful comme nts that Kim Connor, Jaqui Guzmán, and Tom Parrish provided. This wo rk would not be possible without their careful assistance. At PPIC, Richard Greene, Hans Johnson, Eric Larsen, Heather Rose, Jon Sonstelie, and Lynette Ubois provided thoughtful suggestions that improved the study and Karina Jaquet provided outstanding research support. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or b oard of directors of the Public Policy Institute of California. 8 Introduction Special education is a mandated but under-evaluated part of state commitments to public education. U nder the federal Indi viduals with Disabilities Education Act (IDEA), children with disabilities have a legal entitlement to a “free a nd appropriate public education.” 1 The service entitlement alone makes special education finance an important policy issue for California and other states. But special education is also a multibillion -dollar program that serves more than 10 percent of California’s public school enrollm ent. In fact, the $4.7 billion in federal, state, and local special education aid that California school districts received in 2006–07 was their largest source of funding specifically for one program. Nevertheless, a growing concern in California is t hat expenses related to meeting special education mandates tend to “encroach” on other education funds. T his service entitlement makes special education different from most educational programs because it supersedes the availabili ty of funding. In effect, school districts must meet their mandate to serve disabled children before supporting other budgeted programs. 2 The state legislature overhauled the special education system in 199 7. Policymakers cited several undesirable properties of the prior system, such as funding inequity, complexity, and inappropriate placement incentives as reasons for pursuing reform. The reform bill, AB 602, enacted a new allocation model that addressed these issues by assuming that disabilities vary evenly across the state. The current formula distributes the predominant share of special education funds based on total student population size rather than the size of disabled populations or the needs of disabled students. T he requirement that school districts meet all special education needs appropriately, whether needs are rare or frequent, inexpensive or costly , still applies . The perception of budgetary tension between special education and other programs is especially clear in California , where school districts have few options for raising additional revenue to meet new special education demands. In recent years, special education spending has grown faster than spending on other programs while special education fund ing has grown more slowly than general -purpose funds. Spending on special education services in California totaled $9.3 billion in 2006 –07, over 16 percent of all K–12 spending (Lipscomb, 2009b). This report examines special education financing in California in 2006 –07, about a decade after AB 602 passed. It explores the funding process, patterns of disability, and patterns of spending on disabled children . It serves as the basis for evaluating California’s finance policy today and consider s whether the state should pursue further refinements. The first sectio n, on funding , examines the extent to which switching to a per -student funding system has led to greater funding equity. The following section, on disability rates , describes disability patterns with respect to income and factors like demography because special education need s may vary across the state despite the assumption that they do not. Lastly, we document expenditures , including the local share, and describe spending patterns among regions of California where personnel costs are arguably higher. 1 IDEA first passed in 1975 as the Education of all Handicapped Children Act. Congress last reauthorized it in 2004. 2 Regular education budgets funded 28 percent of special education spending in California in 2004 (Asimov, 2006). Harr, Parrish , and Chambe rs (2008) refer to encroachment as a growing policy concern in reviewing special education research. 9 We view special education financing and spending through the lens of IDEA requirements that special education aid help defray a school district’s additional spending on children with disabilities —above its average spending on all children. Concerns about e n croachment —the shortfall between additional spending on disabled children and special education aid —suggest that special education is raiding funding for other programs , although practically every state funds special education through a combination of fede ral, state, and local revenue . As Harr, Parrish , and Chambers (2008) explain , what is considered encroachment in one state may be considered the local share in another. This report adopts a similar view and when ever possible refers to encroachment as local support for additional spending on children with disabilities. Th is list summarizes key terminology used in the report: • Spending on children with disabilities —Combined special education and non- special education spending to educat e children with disabilities • Additional s pending on children with disabilities —Spending on children with disabilities above the average for all children • Special education funds —Federal, state, and local revenue reserved for educating children with disabilities • Local support for additional spending on children with disabilities —Additional spending that is not covered by special education funds (i.e., encroachment) T he report concludes that AB 602 improved special education finance from the previous system, b ut that California can take additional steps to implement the desired changes that reform intended to achieve, like greater funding equity . The conclusion suggests improvements , such as refinements to the allocation model itself, and describes how th ose mi ght work. 10 Financing Special Education in California The Funding Process Special education in California receive d $4.7 billion in funding from federal, state, and local sources in 2006 –07. 3 The California Department of Education (CDE) allocates these funds to 120 regional groups of school districts known as Special Education Local Plan Areas (SELPAs) that coordinate special education activities to offer a wider range of services more efficien tly. 4 Under AB 602 (1997), California has distributed most special education funds based on the average daily attendance (ADA) of each SELPA’s entire student population since 1998– 99. Disability counts and special education expenditures are not part of the funding equation. CDE effectively controls the total size of special education grants from all sources because it deduc ts federal and local funding from state aid when determining SELPA allocations. 5 California’s funding process is a census-based, or capitation, model. While not the predominant choice across the states, the use of census -based models has grown since 1991, and the federal government and nine other states now use them. 6 AB 602 addressed several key concerns with the finance model that had been in place. The old system, which allocated funds based on the number of classes in different instructional settings that districts reported , was widely seen as inequitable and overly complex ; d istricts received differ ent amounts of money for serving children in equivalent settings , even in the same SELPA . T ransitioning to a formula based on a flat grant per student incr eased funding equity and transparency. Streamlining the allocation formula also helped the legislature to increase district flexibility in using funds . In addition, s tate policymakers wanted to avoid incentivizing inappropriate special education placements , which can happen when school districts receive funding based on their self -reported needs. Under a census-based model, school districts that classify more children as disabled incur additional costs but receive no additional funding. The pure fiscal ince ntive actually is to identify fewer disabilities and provide less costly services. Most states continue to use more conventional formulas based on the population and reported needs of disabled students. The sensitivity of identification rates to funding incentives is well documented by researchers, whose findings suggest that census-based models are associated with lower disability rates. 7 3 Lipscomb (2009b) describes the components of California’s special education funding process. For example, Dhuey and Lipscomb (2009) estimate a relationship between states adopting census-based model s and an 8 –10 percent average reduction in their disability rate, mostly in the categories of learning disabilities and mental retardation ; n o category experienced a statistically significant increase in identification. Adopting these models w as also associated with greater u se of outside school placements for severely disabled children . These 4 SELPA membership in 2006 -07 ranged from a single district to 47 districts. Tulare is the SELPA with the most member districts. There were 36 single district SELPAs. California has over 1,000 school districts and county offices of education. 5 The exception is a small number of extraordinarily high cost placements . 6 The states are Alabama, Alaska, Connecticut, Idaho, Massachusetts, Montana, New Jersey, North Dakota, and Pennsylvania. Missouri, South Dakota, and Vermont use a partial census model. 7 For example, see Dhuey and Lipscomb (2009), Greene and Forster ( 2002), Kwak (2008), Lipscomb (2009a), and Mahitivanichcha and Parrish (2005). Cullen (2003) does not examine a census -based model but reaches similar conclusions about funding incentives in special education. 11 are the most expensive placements , bu t the increase may be because all census-based formulas provide some reimbursement for extraordinarily high -cost placements. Some evidence suggests a link between census-based reform and a higher rate of requests for dispute resolution in special education matters (Lipscomb 2009a). 8 The general findings from the research literature s upport conclusions about capitation payments in health care applications . Designing educational programs for disabled children may be come slightly more contentious between parents and schools when state s stop providing funds based on reported needs. 9 Funding Levels M anaged care systems typically reimburse providers based on the number of patients they see per month, rather than patient severity or the cost of treatment. The systems are cost-co ntainment strategies, but introduce a direct incentive for providers to seek healthier patients and provide fewer services. Census-based policies in special education are similar because f unding helps school districts meet individual needs but need plays little role in determining funding. In transition ing to a census-based model, California began equalizing the funding amount that students (regardless of disability status) generate for their SELPA. This amount , called the base rate, underpins most special education allocations in California. Figure 1 shows that California has reduced but not eliminated the disparities that existed in 1998. In fact, t he SELPAs that had high base rates in 1998 continue to ha ve high base rates today. The statewide average base rate has remained mostly constant over the last five years in 2006 dollars. Figure 1 SELPA Base Rates in 2006 Dollars NOTE: High and low base rates correspond to the 9 5th and the 5th percentile rates in California. The 5th percentile is higher than the rate that applies to 5 percent of California students. Data come from CDE. Base rate differences translate into differences in funding per student. The first chart in Figure 2 compares SELPA base rates in 2006 –07 with AB 602 funds , which represent 89 percent 8 The evidence comes from an “enrollment weight ed” specification, suggesting nationwide growth in the rate of dispute resolution requests per special education student. Dhuey and Lipscomb (2009) treat states equally regardless of size. They find a statis tically insignificant average response across ref orm states. 9 Newhouse (1996) reviews this literature. 400 450 500 550 600 650 700 1998 19992000200120022003200420052006Dollars per Average Daily Attendance HighStatewide AverageLow 12 of federal, state, and local special education revenue. 10 There is a visible upward relationship. The relationship is not perfect because the allocation formula has several adjustment factors, such as for regionalized services in small SELPAs (less than 15,000 ADA) and for SELPAs eligible for a Special Disabilities Adjustment (SDA). (SDA funds are for SELPAs that the legislature found to have a greater incidence of hi gh cost disabilities in 1998 but relatively lower base rates. The SDA is the only cost proxy in place. The next chapter discusses the SDA in further detail.) Figure 2 Relationships Between Base Rate s and Special Education Funds, 2006–07 SOURCE: AB 602 Funding Exhibits The second chart in Figure 2 removes these adjustment factors. The remaining funds still represent 81 percent of federal, state, and lo cal special education revenue. The link between base rates and funding becomes nearly one to one, suggesting that historical inequities explain much of the difference in funding across SELPAs today. Further descriptive analysis in the Appendix corroborates this conclusion. The anal ysis adjusts the overall amount of special 10 AB 602 does not allocate funds for several types of services, the largest of which are for special education transportation and infants with disabilities. 500 600 700 800 900 1,000 1,100 1,200 500 600 700 800 900 1,000 1,100AB 602 Funds per Student SELPA Base Rate 500 600 700 800 900 1,000 1,100 1,200 500 600 700 800 900 1,000 1,100Base Allocation per Student SELPA Base Rate 13 education funds per student for differences in the base rate, whether SELPAs serve fewer than 15,000 students, and whether they are eligible for SDA funding. These three factors alone explain nearly 80 percent of the variation in funding , primarily because of the association between base rates and funding per student. 11 T he lack of full base rate equalization has little justification more than a decade following AB 602’s enactment because funding equity is one of the main rationales for adopting a census-based model . Figure 2 indicates that factors like the SDA help somewhat by providing more revenue to some SELPAs with lower base rates. But California could further AB 602’s goals of equity and transparency by starting with a level playing field for everyone. The state could then consider funding adjustments based on factors outside of SELPA control that are good proxies for true cost variation. A criticism of pure census models is that they do not account for variation in student need (Parrish, et al., 2003). The prob lem is that identifying good proxies is difficult. Many of the adjustment factors we think of first, like disability rates and spending, are to some extent within a district’s control. Adjusting funding based on these measures reintroduces the same inappropriate incentives issue census-based model s were supposed to avoid. When faced with this same problem in allocating IDEA funds , the federal government decided to adjust apportionments for child poverty rates. In general, health outcomes tend to improve with socioeconomic status, so the poverty adjustment arguably helps account for part of the variation in special education need . The federal formula distributes 85 percent of funds based on population and 15 percent based on poverty. T his report examines patterns of disability and spending in California to identify potential factors that could arguably serve as proxies for true cost variation . It then illustrates how California could incorporate these factors into an allocation formula . 11 Appendix Table A.1 provides summary statistics for the variables. Table A.2 contains the results. 14 Disability Rates and Income All census models assume that disabilities are spread evenly across the population. California justified AB 602’s goal of equalizing funding per student based on the premise that “handicapping conditions of similar severity” occur with “roughly equal frequency.” 12 AB 602 directs special education funds to SELPAs in part to allay fears that the equal frequency assumption may not hold for smaller populations like school districts. 13 A 1998 report by the American Institutes for Research ( AIR) concluded that severe/high cost disabilities did vary across SELPAs more than could be expected randomly. 14 Based on that , the legislature added the SDA program to the allocation formula as a cost proxy. The SDA provides a severity supplement to some lower -funded SELPAs based on the services that their high cost special edu cation students received in 1997. In 2006 –07, SDA funds provided $81 million to 34 SELPAs or about $34 per student in eligible SELPAs. The legislature has never updated the incidence multipliers used to determine eligibility. 15 Although the SDA derives from historical data, it identifies SELPAs with current higher rates of severe disabilities. Table 1 indicates that in 2006–07, the rate of severe disabilities was 15 percent higher in S DA-eligible SELPAs . This study defines disability severity at the category of disability level, following the delineation that California uses in its financial data. 16 Severity clearly varies within categories too, meaning that the delineation is imperfect. Yet the categories in the severe group tend to be more costly to service, suggesting that grouping disabilities by category is reasonable across the population. 17 Table 1 Severe Disability Rates and Income by SDA Funding Status, 2006 –07 Severe Disabilities (% of Students) Free or Reduced-price Meals (% of Students) SELPAs Receiving SDA Funding 2.59 57.64 SELPAs Not Receiving SDA Funding 2.27 46.40 NOTE: Sample based on 1 19 SELPAs. The proportions in each column are statistically different at the 5 percent level. The z -statistics are 26 and 274, respectively. The AIR (1998) study recommended a cost adjustment based on services received by high -cost students partly because a measure like poverty did not reliably explain the observed differences in s everity across California a decade ago . This study find s some evidence in recent 12 AB 602 Bill Analysis (1997) 13 This was a recommendation from a 1995 report published by the Legislative Analyst’s Office, the Department of Education, and the Department of Finance. 14 Parrish, Kaleba, Gerber, and McLaughlin (1998) 15 The SDA incidence multipliers also help determine cost -of -living adjustments and growth funding. 16 Based on the California School Accounting Manual’s (2008), severe disabilities include autism, deafness, d eaf-blindness, emotional disturbance, mental retardation, multiple disabilities, orthopedic impairments, traumatic brain injury, and visual impairments (including blindness). Non -severe disabilities are learning disabilities, speech or language impairments , and other health impairments. 17 Parrish, Harr, Kidron, Brock, Anand (2004) estimate disability costs for California in 2002– 03. 15 data of a negative relationship between disability status and income. SDA eligibility also appears to correlate with income. For instance, Table 1 indicates that SELPAs receiving SDA funds have a higher rate of participation in free or reduced -price meals , a program with an income eligibility cap at 185 percent of the federal poverty level. Findings from the California Health Interview Survey Figur e 3 uses the California Health Interview Survey (CHIS), the largest state- representative health survey in the United States, to show that severe conditions tend to be less common among children from higher income families. CHIS surveyed parents about their children ’s disabilit ies, and the sample includes 6,515 children ages 5 to 11 in 2005. The findings from the analysis are representative of California children in that age range. Figure 3 Income and Child Disability Conditions, California Health Interview Survey, 2005 NOTE: Each income category contains 20 percent of the sample. Table A.3 provides summary statistics on the CHIS variables. Adjusted disability rates come from regression estimates in Appendix Table A.4. The horizontal axis divides the samp le into five equally sized groups based on income. The first and third columns on the horizontal axis show the average rate of severe and non - severe conditions in each group . Both show a negative relationship with income, although the relationship is smoot her for severe disabilities. The second and fourth columns show the rate of each type of condition after adjusting for differences in children’s gender, language spoken at home, rural setting, race /ethnicity, birth weight, and age. Controlling for these factors does not substantively change the relationship between income and disability status. The findings in Figure 3 support general research conclusions about correlations between health outcomes and socioeconomic status. 0 1 2 3 4 5 6 7 0 to 1.4 Times Federal Poverty 1.4 to 2.7 Times Federal Poverty 2.7 to 4.3 Times Federal Poverty 4.3 to 6.5 Times Federal Poverty At Least 6.5 Times Federal Poverty Percentage of Children Income Categories SevereSevere (Adjusted)Non-severeNon-severe (Adjusted) 16 The association between disability and income exists for behavioral and mental conditions but not for physical conditions. Appendix Table A.4 reorganizes the severe/non- severe delineation based on whether a disability is behavioral/mental or physical. The estimates show a strong relationship between income and the former, but a small and weak relationship between income and the latter. The findings suggest that the probability of a behavioral disability is 10 percent lower for a child at 200 percent of the poverty line than it is for a child at th e poverty line, adjusting for other factors . Findings for Children in Special Education Programs Recent CDE data on actual special education enrollment among children with severe disabilities suggests a similar negative relationship with income. Th e analysis in Appendix Table A.5 constructs rates of severe and non-severe disabilities for each SELPA for 2006–07. It then uses a regression to adjust the rates for a similar set of characteristics as in Figure 3. 18 The income measure is the percentage of students in a SELPA enrolled in the free or reduced -price meals program. Unlike the CHIS analysis, which compares family income and disability conditions at the individual level, the analysis using CDE data compares a SELPA’s disability rate to its percentage of students in free or reduced -price meals. In other words, it only reports associations between aggregated data. While this is an important difference, the SELPA is also the level to which AB 602’s assumption of even disability rates applies. Holdin g constant other observable characteristics, a lower rate of free or reduced -price meals (i.e. higher income) is related to a lower severe disability rate. The findings suggest that the severe disability rate is 12 percent higher in a SELPA with 60 percent of students in free or reduced -price meals than it is in a SELPA where 30 percent of students are in free or reduced - price meals. The average rate of free or reduced -price meals across SELPAs is about 46 percent. 19 The relationship between free or reduced-price meals and rates of non -severe disabilities is much weaker. Part of the explanation is that grouping disabilities under severe and non - severe headings masks differences with respect to income at the category of disability level. To illustrate this, th e remaining columns in Table A.5 show the estimated relationship between free or reduced -price meals and the percentage of children classified in each of the six largest categories of disability. 20 Similar opposing relationships exist for severe disabilities too. M ental retardation is more common in lower -income areas while autism is more common in higher -income areas. These categories account for over 90 percent of disabilities in California. The data suggest that there are opposing relationships in some cases. Specifically, learning disabilities are more common in lower -income areas while other health impairments (e.g. ADD and ADHD) are more common in higher -income areas. These opposing relationships contribute to a weak association between income and non -severe disabilities overall. 18 The model includes the following characteristics: percent free or reduced- price meals, an index of regional non-teacher wage levels, percent English learners, town or rural location, race/ethnicity, SELPA enrollment, AB 602 base rate per ADA, the average SELPA revenue limit per pupil, a single- district SELPA indicator, and a constant. 19 Table A.1 provides summary statistics and Table A.5 contains the regression results. 20 The categories are learning disabilities, speech or language impairments, other health impairments, mental retardation, autism, and emotional disturbance. Lipscomb (2009b) provides disability definitions. 17 Overall, however, rates of severe disabilities tend to be higher in SELPAs with higher proportions of low -income students. The findings also suggest several significant relationships with severe disabilities besides that with income. Holding constant other factors, the rate of severe disabilities is lower in SELPAs with higher concentrations of Hispanic and Asian students, lower concentrations of African -American students, towns and rural areas , larger SELPAs, and single -district SELPAs . Hispanic and Asian children have lower rates of special education classification in California while African -Americans have higher rates relative to non-Hispanic white children (Lipscomb, 2009b). Differences in classification rates by race and eth nicity are most pronounced in the categor ies of emotional disturbance , learning disability , and other health impairment. The finding about small towns and rural areas suggests that urban settings may offer a wider range of local care options : the greater availability of therapy and medical services overall may attract families with severely disabled children. The fact that t hat severe disability rates are also higher in larger and single -district SELPAs appears to support this theory. Single -district SELPA s tend to have both above -average district enrollments and to be located in urban locations. The AIR study by Parrish et al (1998) found that single-district SELPAs spend more per student than do others . In sum, both the CHIS and actual special education enrollment data suggest a relationship between income and disability status, particularly for severe disabilities. A factor related to income, such as free or reduced -price meal eligibility, may be an appropriate proxy to identify SELPAs that face a higher rate of severe special education needs. An income-based adjustment would have both potential advantages and disadvantages in relation to the existing cost proxy, the SDA. The advantages are that it is entirely out of SELPA control , that it could be update d regularly , and that it would align closely with the federal formula. But an income -based adjustment has potential disadvantages too. A 2004 follow -up report by AIR recommended that California instead update the SDA , partly because of the differing patterns of mental retardation and autism with respect to poverty. AIR also cite d the possible social stigma attached to enrolling in free or reduced -price meals at the high school level. California could address this issue by coll ecting data on income, rather than enrollment, to determine meal program eligibility . Further, relationships between poverty and disability rates do not inform the question of how much additional funding SELPAs need because of their higher poverty rate. Wh en the federal government experienced this issue, it adopted an 85/15 compromise between the population and poverty -based portions of its allocation formula. Because of the valid concerns about an income -based adjustment, maintaining the existing cost prox y, the SDA, would be a sensible way to go , too. In this case, California should update the multipliers that determine funding to maximize the SDA’s effectiveness as a proxy for costs today. 18 Spending on Children with Disabilities in California Both the prevalence of disabilities and the amount of special education funds affect s chool spending on disabled children . The link to funding comes from IDEA , which requires that districts use federal assistance for disabled children to help pay the “excess costs” of educating them (Federal Register, 2006). 21 Conceptual Framework School districts incur excess costs when they spend more educating disabled children than they spend on average on all children. T his chapter documents spending levels and patterns across SELPAs. As in the previous chapter , the emphasis is partly on identifying a factor outside of SELPA control that arguably serves as a proxy for cost variation. The focus here is to account for differences in regional labor market conditions for educators , a different type of cost from the severity of student needs. Despite the name, excess cost is actually a measure of spending. Costs are defined as the minimum expenditure for the services a student needs. Expenditures exceed costs when needs are not identified correctly or when districts are not providing services efficiently. Patterns of spending may resemble, but are not necessarily the same as, patterns of cost. To underscore the distinction, t his report refers to excess costs as ad ditional spending on children with disabilities. Special education and regular education share spending on children with disabilities. For example , children with speech impairments may receive speech therapy instruction on a regular basis but are otherwise in the regular classroom, while children with severe mental retardation may spend most of the school da y outside the regular classroom. Figure 4 illustrates how this sharing works. The dashed line represents a verage spending on all children. Additional spending on children with disabilities is above the dashed line. Special education funds help to defray these amounts, and local funds pay the rest. Figure 4 Illustrating School Spending on Nondisabled and Disabled Children NOTE: Figure 4 is strictly illustrative, including the dollars per student shown on the vertical axis. 21 This requirement is also listed in California Education Code Section 56841(a). 0 4,000 8,000 12,000 16,000 Children without Disabilities Children with Disabilities Dollars per Student Non-Special Education SpendingSpecial Education Spending Special Ed Funds Average Spending on All Children Local Support (Encroachment)AdditionalSpending (Excess Cost) 19 In California, s tate aid for special education effectively shares the same purpose federal funds do in helping d efray additional spending on children with disabilities. The effective intent of state aid is the same b ecause CDE deducts federal and local special education funds from state aid when calculating SELPA allocations . The only way for California to show that districts are using IDEA funds as intended is to treat state allocations in the same manner. That said, the language in California’s education code suggests a slightly different purpose . The cited purpose is to help local districts “provide special education and related services to individuals with exceptional needs.” 22 The language of the education code contribute s to confusion about the term “encroachment ,” which refers just to the local support amount above the dashed line. “Encroachment” conventionally describes any situation where local funds close a gap between special education spending and funding. In other words, state aid is for the entire special education area in Figure 4 according to the education code , although it is actually just for the part above the dashed line because of the way California deducts funding from non-state sources when calculating SELPA allocations . 23 This usage may be common and appears to derive straight from the education code, but it needs to be better defined because special education funds are not meant to pay for all progra m spending. Children with disabilities generate other education funds for schools too, which schools can use for special education spending. As mentioned in the Introduction, partial local support for special education finance is the norm nationwide. Only Wyoming and Hawaii reimburse 100 percent of special education spending (Parrish, et al., 2003). 24 Districts draw from their other education funds to pay the local support portion shown in Figure 4. Although this money relates to a school district’s ex cess fiscal needs, it is an imperfect measu re because it describes expenditures and not costs. (It is important to remember that patterns of spending are not necessarily the result of differences in cost. ) Variations among SELPAs in local support could result from other reason s too, such as parents who have g reater demands for special education services in one SELPA than in another. In California, state special education aid help s pay just the part of spending above the dashed line in Figure 4. Clarifying the intent of state special education aid is a suggestion for California to consider. That said, special education expenditures are likely to stand as a better proxy for costs than other types of school spending because legal contracts between parents and schools under IDEA delineate the special education services that schools must provide (Harr, et al., 2006). School districts also have a clear incentive to operate special education programs efficiently because each additional dollar of local support diminishes available funds for other programs. Maximizing efficiency brings expenditure s and costs in closer alignment. School administrators tend to view local support for additional spending on disabled children as evidence of insufficient government fundin g for meeting special education 22 California Education Code Section 56836.04(b) 23 For instance, the first page returned from a Google search for “special education encroachment California” on December 8, 2008 was a financial report from the Fullerton School District (2007). The report budgets $15.1 million for special education serv ices, in cluding $6.7 million in encroachment costs. 24 Hawaii reimburses at a 100 percent rate because it operates a single school district for the entire state. 20 mandates. Additional funding is certainly a possible solution. So is looking for further cost - saving efficiencies and improving the way the state allocates existing funds. Federal stimulus money (The American Recovery and Re investment Act [ARRA] of 2009) may provide a rare opportunity for schools to reduce local spending on children with disabilities , through an additional $1.3 billion in one-time IDEA funds to California school districts. D istricts may be able to use half of any increase in federal assistance to reduce local and state spending. 25 Federal stimulus funds may also lead to greater efficiencies over a longer period if schools invest them with that potential in mind. For instance, schools could invest in professional development strategies aimed at serving children with disabilities in the regular classroom as effectively as possible, or in new diagnostic equipment to better identify student needs. 26 Expenditure Totals for 2006 –07 California public schools spent $53.1 billion , or about $8,447 per pupil, educating students in 2006–07 (Table 2). More than $9.3 billion went to providing special education and related services. This is part of total spending on children with disabilities. The other part is non -special education spending on disabled children . 27 This latter amount must be estimated because the state’s education finance data do not separate regular education spending for disabled and non disabled children. 28 We find that the regular education port ion was about $2.7 billion for 2006 –07. In other words, total spending on children with disabilities was $12.0 billion, or about $17,633 per disabled child in the state. 29 The average amount of spending per disabled child in California is roughly 19 percent higher than previous estimates for the nation. 30 Lipscomb (2009b) finds a similar difference between California and past national estimates for the special education component of spending alone. 25 Ordinarily, school districts must spend more on special education from local or combined state and local funds in a year than in the prior year (in total or in per pupil terms) to meet funding “Maintenance of Effort” requirements. 26 The ARRA example considers an increase in funding for children with disabilities. In theory, an increase in general- purpose funding can also reduce local support for additional spending by raising the dashed line in Figure 4. 27 Non -special education spending is not the same as spending in the regular classroom environment because children can receive special education serv ices in the regular classroom. 28 The method in this report is to prorate non- special education spending based on the percentage of the school day that children with the same disability in California spend inside the regular classroom. For instance, each nondisabled student gets one share of non- special education spending. Disabled students get a fractional share based on their disability. See Appendix B for full details. 29 The financial data includes all special education expenditures on behalf of children up to age 22. As a result, average spending per disabled child is found by dividing $12.0 billion into total special education enrollment among children up to age 22. 30 The Center for Special Education Finance (2003) reports $13,054 in average spending per school -aged special education student in 2001– 02. This converts to about $14,819 in 2006 dollars. The inflation adjustment is the Consumer Price Index for the west region of the United States. 21 Table 2 Education Spending per Student by Disability Status, 2006 –07 Dollars (billions) Dollars per Total Enrollment Dollars per Disabled Child Dollars per Nondisabled Child Total Expenditures 53.1 8,447 Special Education Expenditures 9.3 1,474 13,642 Non-Special Ed Spending on Children with Disabilities 2.7 431 3,991 Total Spending on Children with Disabilities 12.0 1,905 17,633 Total Spending on Children without Disabilities 41.1 6,542 7,334 NOTE: Appendix B describes the methodology . All amounts are net of capital outlay and debt service expenditures. The sample includes data from all school district s, county office s of education, and transportation joint powers agencies . Total enrollment is 6,282,036 K–12, ungraded, and adult educatio n students. S pecial education enrollment is 678,699 children with disabilities ages 0 –22. By comparison, California schools spent ab out $41.1 billion educating non disabled children in 2006 –07, about $7,334 per child . Spending on disabled children was 2.4 times higher than spending on nondisabled children. The spending ratio in California is somewhat higher than spending ratios estimated for the nation. Harr, Parrish, and Chambers (2008) summarize the results from four studies analyzing data between 1968 –69 and 1999 –2000. The spending ratios for the nation in the se studies range from 1.90 to 2.29. 31 California ’s higher spending ratio may reflect several factors. For one, the average severity level of special education students in California may be higher than in other states. California has the nation’s lowest rate of special education identification, well below the national average of about 14 percent (Lipscomb, 2009b). A lower rate of classification among children with the least severe disabilities may acco unt for much of this difference. If so, the population of special education students in California has a bigger share of severely disabled children, helping to explain its relatively higher spending ratio. The interaction between the service entitlement, California’s higher personnel costs, and its lower overall rate of spending per student may also contribute to a higher spending ratio. Personnel costs represent 80 percent of district expenditures in California (Rose and Sengupta, 2007) . According to the U .S. Department of Education’s National Center for Education Statistics (NCES), college graduates in California earn higher salaries than in practically every other state. 32 The share of school spending devoted to children with disabilities resembles what is found in other states, but California classifies fewer students that way. Disabled children account for 10 percent of enrollment in California but 22.5 percent of spending. Nationwide estimates for 1999 –2000 found that 12.1 percent of students were disabled, and that 21.4 percent California s chool districts need to pay these higher salaries to attract quality educators and, like other states, they need to meet special education mandates . To the extent that school districts in California meet special education mandates at higher costs than other states but spend less overall, they have less revenue available to support other programs. 31 It is possible that the national spending ratio has grown since 1999–2000 because of growth in high cost disabilities like autism. 32 See the Comparable Wage Index data at www.nces.ed.gov. 22 of spending was devoted to them (Harr, et al., 2006). California may spend more, but it does not spend the most. The Special Education Expenditure Project (SEEP) analyzed spending in 11 states between 1999 and 2001 . Adjusted for inflation, the findings suggest that five of the states exceeded California’s average spending on disabled children in 2006–07. 33 Additional Spending on Disabled Children and Local Support Figure 5 illustrates California’s $12 billion in spending on children with disabilities in the same way as Figure 4. The first portion , at the bottom of the column, is for non -special education spending. The next part is special education spending below the dashed line that represents average spending on all children. Together, these components total ed $4.4 billion in 2006– 07. Additional spending accounts for the remaining $7.6 billion. Special education funding offsets about $4.7 billion in additional spending, leaving $2.9 billion in local support. Altogether, special education funds offset 62 percent of additional spending on disabled children. The local support portion is the remaining 38 percent , or 24 percent of total spending on children with disabilities. Figure 5 California School Spending on Children with Disabilities, 2006 –07 NOTE: Appendix B describes the method used for each calculation. Data come from the Standardized Account Code Structure a nd IDEA Educational Environment records for 2006. We measure additional spending and local support at the SELPA level based on each SELPA’s entire student population. Spending and funding at the district level has more to do with local plan a greements than how the funding formula works. The methodology for calculating spending in this report follows federal regulations closely (see A ppendix B) except for two main differences. First, the federal regulations describe a district-level calculation. 33 SEEP examined spending in Alabama, Delaware, Indiana, Kansas, Missouri, New Jersey, Ne w York, Ohio, Rhode Island, Maryland, and Wyoming (Center for Special Education Finance, 2003). 2.7 1.7 4.2 0.5 2.9 0 1 2 3 4 5 6 7 8 9 10 11 12 Spending ($ billions) Local Support for Additional Spending Special Education Funds (Non-AB 602) Special Education Funds (AB 602) Special Education Spending to Reach Excess Cost Threshold Non-Special Education Spending Additional Spending ($7.6 billion) Special Education Spending ($9.3 billion) 23 Second, they specify that districts should make separate calculations for elementary and secondary students. Neither of these differences is likely to change the finding that there is a wide range of additional spending and local support per student across California. Fo r instance, Table 3 shows that local support represented $1,004 of $1,830 in additional spending per student at Mt. Diablo Un ified School District (USD) SELPA. This is 120 percent and 52 percent higher , respectively, than the statewide average of $455 and $1,201 per student. Special education funds appear to be more than sufficient to offset all additional spending on children with disabilities in a few SELPAs. For example, Sierra County Office of Education ( COE) SELPA ha d about $930 per student in additional spending but receive d $1,188 per student in special education grants. Table 3 SELPAs with the 10 Highest and Lowest Levels of Local Support per Student , 2006–07 SELPAs with the Most Local Support per Student SELPAs with the Least Local Support per Student SELPA Name Additional Spending Special Education Revenue Local Support SELPA Name Additional Spending Special Education Revenue Local Support Mt. Diablo USD 1,830 826 1,004 Sierra COE 930 1,188 -258 Los Angeles USD 1,794 851 943 Trinity COE 1,016 1,132 -116 San Diego CUSD 1,831 900 932 Lassen COE 972 1,040 -68 Newport -Mesa USD 1,636 728 908 Colusa COE 747 773 -27 Santa Clara I 1,557 672 885 Santa Clara III 1,020 983 37 North Orange 1,594 764 830 Siskiyou COE 955 902 53 Tri-City (Culver City USD) 1,489 780 709 Tehama COE 812 740 71 Santa Clara II 1,417 710 707 Imperial COE 728 649 79 North Region (Albany) 1,419 729 691 Humboldt/Del Norte 804 725 79 San Mateo COE 1,464 779 685 Modoc COE 1,403 1,320 83 NOTES: See Appendix Table B .2 for the complete list of SELPAs. The statewide per pupil average values are $1,201 of additional spending , $746 of special education revenue, and $455 of local support. SELPA funding clearly differs across California , but not because of differences in spending ; SELPAs with higher additional spending per student tend to have higher amounts of local support per student (Figure 6) . The solid line in the figure would be flatter if SELPAs with higher additional spending instead tended to receive more special education aid per student. Figure 6 describes an empirical relationship in the data, but not necessarily a policy concern because census-based models are not supposed to track spending. To the extent that spending and costs align closely, however, Figure 6 suggests that adjusting allocations based on cost proxies may help equalize the amount of local support per student across the state. 24 Figure 6 Additional Spending and Local Support for California SELPAs , 2006–07 SOURCE: Author’s calculation based on A ppendix Table B.2 Spending and Regional Non- Teacher Wages Patterns of spending on children with disabilities relate to regional labor market conditions in California . As mentioned earlier, the average salary for college -educated workers in California is among t he highest across the states, and employee compensation is the predominant expenditure for schools. T eacher compensation varies considerably across California (Rose and Sengupta, 2007) ; for 2003–04, for a mid -career teacher, it ranged from less than $55,000 in Yolo and the North Coast counties to more than $70,000 in Santa Clara and Orange Count ies. 34 When the price of resources is high er, school districts have less purchasing power under a fixed budget , meaning that they need to look for efficiencies in their program offerings. D istricts have less flexibility in special education offerings because meeting special education mandates supersedes budgetary concerns. For example, if a student needs a special education aide in the regular classroom, schools must provide one whether they are in Yolo or Santa Clara Counties . Rose and Sengupta (2007) f ound that the wages of non -teachers with similar educational attainment as teachers provide a good benchmark for contextualizing differences in teacher compensation across labor markets in California . They developed a comparable wage index (CWI) that compares non -teacher wages in a regional labor market to the statewide average. Figure 7 shows a positive relationship between non-teacher wages and local support for additional spending on disabled children. The value of the index is higher for SELPAs located where non -teachers earn higher average wages. 34 These salaries are for teachers with 10 years of experience and 60 credits beyond a bachelor’s degree. -300 -100100 300 500 700 900 1,100 0 500 1,000 1,5002,000 Local Support per Pupil Additional Spending per Pupil 25 Figure 7 Local Support for Additional Spending and Regional Non-Teacher Wages, 2006–07 NOTE: The statewide average value of the CWI is 1. The CWI serves as a proxy for variation in the personnel costs of hiring and retaining educators because it corresponds to the wage that teachers can expect outside of teaching. 35 It is helpful in studying patterns of spending because it is outside the control of school districts and teachers’ unions. External labor market conditions are a different type of cost for districts than the amount and severity of special education needs. 36 Appendix Table B.3 adjusts the relationship in Figure 7 for differences in a number of observable SELPA characteristics. These includ e the percent enrolled in free or reduced-price meals, percent English learners, urban or rura l location, race-ethnicity, total enrollment, measures of special education and general purpos e funding, and an indicator for single-district SELPAs. The findings indicate that both addi tional spending and local support per student relate positively to the comparable wage index, holding these factors constant. 37 The analysis supports the same conclusion by further controllin g for the rate of both severe and non-severe disabilities. Patterns of spending relate to factors beyond types of disability. 38 Figure 8 illustrates the expected rate of spending if the CWI were 10 percent above average, adjusting for observable SELPA characte ristics and rates of disability. The findings suggest that additional spending on disabled children pe r pupil would be about 4.3 percent higher than average. 35 See Rose and Sengupta (2007) for illustrations of the relation ship between teacher and non-teacher wage levels across counties in California. 36 Comparable plots to Figure 7, available up on request, show little evidence of a relationship between the comparable wage index and either the overall rate of special educat ion or the rate of severe disability. 37 The findings are numerically identical because specia l education funding per student is held constant. 38 Several other variables are significant as well. Higher addition al spending relates to lower enrollment in free or reduced-price meals, a higher concentration of English learners, larger SELP As, more special education funds per student, single-district SEL PAs, and higher severe disability rates. The lowe r rate of additional spending in SELPAs with larger rates of free or reduced-price meals appears to reflect a higher rate of spending ov erall (i.e. a higher dashed line in Figure 4). -400 -2000 200 400 600 800 1,000 1,200 0.7 0.8 0.9 1 1.1 1.2 Local Support for Additional Spending per Pupil Comparable Wage Index 26 Figure 8 Spending in a SELPA with a CWI 10 Percent above Average, 2006 –07 NOTE: Asterisk indicates statistical significance at the 10 percent level. The findings come from columns 3 through 5 in A ppendix Table B.3. Per-pupil spending includes both disabled and nondisabled students in the denominator. For comparison, Figure 8 illustrates similar relationships for the overall rate of spending on disabled and nondisabled children. T he association between the CWI and spending per pupil on disabled children is the larger of the two, holding constant other factors. Spending on disabled children is expected to be about 3.7 percent higher than average while spending on the nondisabled is expected to be 2.2 percent higher. The association is more precise for disabled children as well. In fact, the CWI in Ap pendix Table B.3 is not statistically significant in describing patterns of spending per pupil on nondisabled children. In most states, school districts have the option to try raising additional income if resources are expensive. In California, however, the state largely sets education revenue and leaves few options for local school districts to raise funds. Rose and Sengupta (2007) propose using the CWI to help equalize the purchasing power of school distric t budgets. At least two other recent policy reports on California education finance share this recommendation (Sonstelie, 2007; Bersin, Kirst, and Liu, 2007) . Sonstelie (2007) reaches this conclusion after applying a theoretical framework of economic decisionmaking under a fixed budget and a set of resource costs to school finance and survey data. Using a theoretical model to guide the analysis helps to moderate concerns about using expenditure data to describe patterns arguably related to cost. The empirical strategy in this study resembles the one suggested by the theoretical model in Sonstelie (2007). The Rose and Sengupta (2007) and Sonstelie (2007) studies were written as part of the Getting Down to Facts research project on California school finance and governance that was organized by Stanford University. The Bersin, Kirst, and Liu (2007) proposal grew out of the findings. This latter study proposes a funding system that consists of a base grant per student, an equalized special education grant per student, targeted funding for low-income students, and a regional cost adjustment. The findings in this section support a similar conclusion about the CWI and special education funds. 0% 1% 2% 3% 4% 5% Additional Spending per Pupil * Spending on Disabled Children per Pupil * Spending on Nondisabled Children per Pupil Rate of Spending above the Average Across SELPAs 27 Suggestions for Improving Special Education Finance California’s switch to per -student funding via AB 602 (1997) improved the state allocation formula by placing a greater e mphasis on funding equity, transparency , and flexibility while minimizing incentives to classify students inappropriately. But more than a decade later, California’s census -based funding system shows signs that still more could be done. Children still gene rate different amounts of special education funding depending on the SELPA in which they live. Moreover, the variation s in funding rates today reflect the historical disparities that existed under the previous funding regime. California could take the following two steps to ward implementing more fully the type of special education finance system it chose to adopt in 1997. • Refine the allocation model. Equalize the base rates and adjust funding for a small number of factors outside of SELPA control. • Clarify the state’s objective for special education funds. Emphasize providing appropriate services for educating disabled children rather than providing special education services alone. Refine the Allocation Model California should preserve its census-based approach for special education finance, but it can improve upon the existing design. The first step is to complete the equalization of base rate s per student across the state. The second step is to adjust funding for a small number of factors that are outside of SELPA control and could serve as a proxy for true cost variation . The resulting model would aim to be sensible and simple, furthering the existing funding goals. As a census -based model, schools would be unable to influence funding levels based on the way they classify and serve students. F urther, it could offer SELPAs greater flexibility in using funding , should policymakers decide to consolidate several sources of special education revenue into one allocation. According to a recent report by the Legislative Analyst’s Office (2008), SELPAs receive state special education funds through 15 separate programs. The report argues for merging many of these funds because the current allocation method makes it hard to see how much funding the state provides and how it ultimately distributes funds. This study suggest s a possible form for a consolidated formula that involves equal base rates, with adjustments for eligibility in free or reduced -price meals, and regional non -teacher wage levels. The formula could resemble the federal IDEA funding process, distribut ing 85 percent of funds based on enrollment and 15 percent based on poverty. Such a formula would also re semble the Bersin, Kirst, and Liu (2008) prop osal for California’s K–12 education finance system that includes both funding for low -income students and a regional wage adjustment. Ap pendix C provides technical information. 28 Each SELPA’s funding allocation, F , w ould take the following form: The formula has two components, a base and a regional wage adjustment. Most of the base (a 0.85 weight in this example) would come fr om multiplying the statewide base rate by total SELPA enrollment. The rest is an a djustment for low-income students. The adjustment is the statewide base rate multiplied by the number of students eligible for free or reduced -price meals. The formula then adjusts a SELPA’s base funding for regional wage levels. The average value of the regional wage adjustment is one . The statewide base rate depends on how much money California consolidates into the allocation. Table 4 compares the simulated allocation to the current allocation using funds from the AB 602 base entitlement (second chart in Figure 2). Appendix Table C.1 provides a similar analysis using all AB 602 funds (first chart in Figure 2). 39 Table 4 Base Allocation Funds U nder Simulated and Current Models, 2006 –07 The respective statewide base rates are $673 or $740 per student. California could apply future cost of living adjustments or other funding supplements directly to the base rates. A. Statewide Average, Low, and High Values Average Lowest Highest Simulated Allocation 605 491 692 Current Base Allocation 605 547 995 B. Model Comparison Comparable Wage Index Simulated Allocation Current Base Allocation Low Medium High Low Medium High .74–.92 .93–.98 .99–1.15 .74–.92 .93–.98 .99–1.15 Percent Free or Reduced Price Meals Low 0–38 533 581 620 632 604 625 Medium 39–56 542 596 647 616 582 594 High 57–100 560 618 659 652 599 646 NOTES: Summary statistics are weighted by total SELPA enrollment. The low, medium, and high categories in the model comparison each include about one -third of SELP As. Table C.1 provides a similar model comparison using all AB 602 funds. Section A shows that the simulated allocation maintains the actual level of base funding that was available to SELPAs in 2006 –07. The difference is in how funding is distributed, wi th 39 The simulation in Appendix Table C.1 uses all AB 602 funds for simplicity. Policymakers should leave some AB 602 programs, such as funding for out- of-home care, unchanged. 29 the simulated allocation reducing funding variation across the state. By design, funding per student under the simulated model (see Section B) is highest for SELPAs with large proportions of low er-income students in regions with higher expected personnel expenses. Funding per student is smallest for SELPAs that have the opposite student and personnel characteristics. 40 The simulated allocation in Table 4 is just one possible way for policymakers to refine the funding formula and stay consistent with the finance reform goals of 1997. In practice, the state legislature c ould adjust the 85/15 weighting between the enrollment portion and the low- income adjustment. It could also identify and use other adjustment factors. For instance, California could maintain the SDA, the existing AB 602 cost proxy. In this case, the state should consider whether updating the SDA’s eligibility criteria after a decade would provide an even better cost proxy . T he existing allocation shows less consistent patterns with respect to regional wage levels and proportions of low -income students. For instance, SELPAs in the low/low categories currently have among the higher average rates of base funding. R efining the funding formula along the lines of Table 4 would help California move closer to fully implementing its census -based finance reform goals. Recent policy research on education finance in California supports this type of model . It could even lead to greater efficiencies through increased flexibility . Implementation would require additional state funds only if California chose initially to hold SELPAs harmless (i.e. , prevent them from losing funds under the new formula) and then phase out the hold -harmless provision over several years. Clarify the State’s Objective for Special Education Funds California can have more constructive special education debates at state and local levels by focusing on total spending to educate children with disabilities rather than on special education expenditures alone. This broader frame of reference provides a more complete picture of how schools use both the regular education and special education environments to meet the needs of children with disabilities. By examining total expenditures on disabled children in rela tion to total expenditures on nondisabled children, school districts can more meaningfully account for what is spent and why. The state can take an important step by aligning the intent of state special education funds with IDEA. Th is would provide a clearer rationale for funding because both state and federal funds share the same purpose—helping to defray the additional spending to educate children with disabilities . A secondary benefit would be to help clarify the actual meaning of the term encroachment ( the local share of additional spending on disabled children ). Doing so would underscore how the debate should focus on whether the size of the local share is fair for providing a free a nd appropriate public education, not on whether the local share should exist at all. In addition, better data would improve special education discussions and ensure greater program accountability. Currently, state data systems are not equipped to calculate school spending by a child’s disability status. The missing ingredient is the capacity to 40 The low/low cell includes San Juan, Lassen, Tuolumne, San Luis Obispo, Amador, Calaveras, El Dorado, Nevada, Clovis, Sierra, and Placer. The high/high cell includes Garden Grove, West Contra Costa, San Fr ancisco, Santa Ana, Anaheim, and Oakland. 30 account for non -special education spending on children with disabilities. This report is able to provide an estimate, but calculations that are more exact would be possible if the current data collection could track spending by disability status. 41 California’s current funding formula has several desirable properties, but more can be done. By building upon the state’s existing accomplishments in reforming special education finance, it can develop a system to serve as a role model for education finance reforms in the future. 41 Appendix B describes the methodology used in the report. 31 References AB 602 and Bill Analysis, Senate Rules Committee, Office of Senate Floor Analyses, September 9, 1997. Asimov, Nanette, “Extra -Special Education at Public Expense,” San Francisco Chro nicle, February 19, 2006. Bersin, Alan, Michael W. Kirst, and Goodwin Liu, “Getting Beyond the Facts: Reforming California School Finance,“ Chief Justice Earl Warren Institute on Race, Ethnicity, and Diversity, University of California, Berkeley, Californ ia, 2007. California School Accounting Manual, California Department of Education, Sacramento, California, 2008. Center for Special Education Finance, “Comparison of Special Education Expenditures across SEEP States and the Nation,” American Institutes f or Research, 2003. Cullen, Julie Berry, “The Impact of Fiscal Incentives on Student Disability Rates,” Journal of Public Economics , Vol. 87, 2003, pp. 1557 –1589. Dhuey, Elizabeth , and Stephen Lipscomb, “The Effects of Fiscal Incentives in Special Education: Evidence from Capitation Finance Re forms,” PPIC Working Paper, 2009. Federal Register, “34 CFR Parts 300 and 301: Assistance to States for the Education of Children With Disabil ities and Preschool Grants for Children With Disabilities; Final Rule,” Federal Register, Vol. 71, No. 156, 2006. Fullerton School District, “2007/08 First Interim Financial Report,” 2007, available at www.fsd.k12.ca.us/menus/bussvcs/fiscal/InterimReports/0708FirstInterim Narrative.pdf . Greene, Jay P. , and Greg Forster, “Effects of Funding Incentives on Special Education Enrollme nt,” Civic Report , Vol. 32, 2002, pp. 1 –13. Harr, Jenifer J., Tom Parrish, and Jay Chambers, “Special Education,” in Handbook of Research in Education Finance and Policy , Eds. Helen F. Ladd and Edward B. Fiske, Routledge, New York, New York, 2008, pp. 575 –590. Harr, Jenifer J., Tom Parrish, Jay Chambers, Jesse Levin, and Maria Segarra, “Considering Special Education Adequacy in California,” American Institutes for Research, 2006. Kwak, Sally, “The Impact of Intergovernmental Incentives on Special Education Spending,” Working Paper, 2008. Legislative Analyst’s Office, “LAO Recommended Legislation,” www.lao.ca.gov, 2008. Lipscomb, Stephen, “Resolving Special Education Disputes in California,” Public Policy Institute of California, San Francisco, CA, 2009. Lipscomb, Stephen, “Students with Disabilities and California’s Special Education Program,” Public Policy Institute of California, San Francisco, CA, 2009. 32 Mahitivanichcha, Kanya , and Thomas Parrish, “Do Non-Census Fund ing Systems Encourage Special Education Identification? Reconsidering Greene and Forster ,” Journal of Special Education Leadership , Vol. 18, No. 1, 2005, pp. 38 –46. McLaughlin, Margaret J. , and Maria F. Owings, “Relationships Among States’ Fiscal and Demo graphic Data and the Implementation of PL 94 –142,” Exceptional Children, Vol. 59, 1993, pp. 247 –261. “New Funding Model for Special Education: Final Report,” California Legislative Analyst’s Office, Department of Education, and Department of Finance, 1995. Newhouse, Joseph P., “Reimbursing Health Plans and Health Providers: Efficiency in Production versus Selection,” Journal of Economic Literature , Vol. 34, pp. 1236–1263. Parrish, Thomas, Jenifer Harr, Jennifer Anthony, Amy Merickel, and Phil Esra, “State S pecial Education Finance Systems, 1999 –2000: Part I,” American Institutes for Research, Center for Special Education Finance, 2003. Parrish, Tom, Jenifer Harr, Yael Kidron, Leslie Brock, and Priyanka Anand, “Study of the Incidence Adjustment in the Special Education Funding Model,” American Institutes for Research, 2004. Parrish, Thomas B., Daniel Kaleba, Michael Gerber, and Margaret McLaughlin, “ Special Education: Study of Incidence of Disabilities Final Report,” American Institutes for Research, Center fo r Special Education Finance, 1998. Rose, Heather, and Ria Sengupta, “Teacher Compensation and Local Labor Market Conditions in California: Implications for School Funding,” Public Policy Institute of California, Occasional Paper, 2007. Sonstelie, Jon, “Ali gning School Finance With Academic Standards: A Weighted -Student Formula Based on a Survey of Practitioners,” Public Policy Institute of California, 2007. About the Author Stephen Lipscomb is a research fellow at the Public Policy Institute of California, where he studies education policy issues. His current work focuses on special education, school accountability, and patterns of early grade retention. He holds a Ph.D. in economics from the University of California, Santa Barbara. PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Walter B. Hewlett, Chair Director Center for Computer Assisted Research in the Humanities Mark Baldassare President and Chief Executive Officer Public Policy Institute of California Ruben Barrales President and Chief Executive Officer San Diego Regional Chamber of Commerce John E. Bryson Retired Chairman and CEO Edison International Gary K. Hart Former State Senator and Secretary of Education State of Californ ia Donna Lucas Chief Executive Officer Lucas Public Affairs Ki Suh Park Design and Managing Partner Gruen Associates Constance L. Rice Co -Director The Advancement Project Thomas C. Sutton Retired Chairman and Chief Executive Officer Pacific Life Insurance Company Raymond L. Watson Vice Chairman of the Board Emeritus The Irvine Company Carol Whiteside President Emeritus Great Valley Center PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 p hone: 415. 291.4400 f ax: 415. 291.4401 PPIC SACRAMENTO CENTER Senator Office Building 1121 L Street, Suite 801 Sacramento, California 95814 p hone: 916.440.1120 f ax: 916.440.1121 www.ppic.org" } ["___content":protected]=> string(102) "

R 809SLR

" ["_permalink":protected]=> string(104) "https://www.ppic.org/publication/special-education-finance-in-california-a-decade-after-reform/r_809slr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8631) ["ID"]=> int(8631) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:39:20" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3886) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 809SLR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_809slr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_809SLR.pdf" ["wpmf_size"]=> string(6) "374699" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(77644) "Special Education Financing in California A Decade After Reform Stephen Lipscomb with contributions from Karina Jaquet Supported with funding from The William and Flora Hewlett Foundation August 2009 The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research on major economic, social, and political issues. The institute’s goal is to raise public awareness and to give elected representatives and other decisionmakers a more informed basis for developing policies and programs. The institute’s research focuses on the underlying forces shaping California's future, cutting across a wide range of public policy concerns, including economic development, education, environment and resources, governance, population, public finance, and social and health policy. PPIC is a private operating foundation. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. PPIC was established in 1994 with an endowment from William R. Hewlett. Mark Baldassare is President and Chief Executive Officer of PPIC. Walter B. Hewlett is Chair of the Board of Directors. Copyright © 2009 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. Research publications reflect the views of t he authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. . Contents Acronyms 4 Summary 5 Acknowledgments 7 INTRODUCTION 8 FINANCING SPECIAL EDUCATION IN CALIFORNI A 10 The Funding Process 10 Funding Levels 11 DISABILITY RATES AND INCOME 14 Findings from the California Health Interview Survey 15 Findings for Children in Special Education Programs 16 SPENDING ON CHILDREN WITH DISABILITIES IN CALIFORNIA 18 Conceptual Framework 18 Expenditure Totals for 2006–07 20 Additional Spending on Disabled Children and Local Support 22 Spending and Regional Non-Teacher Wages 24 SUGGESTIONS FOR IMPROVING SPECIAL EDUCAT ION FINANCE 27 Refine the Allocation Model 27 Clarify the State’s Objective for Special Education Funds 29 References 31 About the Author 33 4 Acronyms AB 602 Assembly Bill 602 (1997) ADA Average Daily Attendance ADD/ADHD Attention Deficit Disorder / Attention Deficit Hyperactivity Disorder AIR American Institutes for Research ARRA American Recovery and Reinvestment Act (2009) CASEMIS California Special Education Management Information System CBEDS California Basic Education Data System CBSA Core Based Statistical Area CDE California Department of Education CHIS California Health Interview Survey COE County Office of Education CWI Comparable Wage Index ESEA Elementary and Secondary Education Act (1965) IDEA Individuals with Disabilities Education Act JPA Joint Powers Agreement LEA Local Education Agency MSA Metropolitan Statistical Area NCES National Center for Education Statistics SACS Standardized Account Code Structure SDA Special Disabilities Adjustment SEEP Special Education Expenditure Project SELPA Special Education Local Plan Area USD Unified School District 5 Summary This report assesses California’s special education finance policy and suggests improvements, about a decade after California overhauled special education financing to address concerns about the efficacy of the previous system. That overhaul , Assembly Bill 602 (1997), sought t o ensure greater funding equity, eliminate inappropriate placement incentives, and streamline the funding model, among other objectives . We find that finance reform has led to positive changes but that the state can do more to implement its desir ed reform goals. Special education programs help in educating California’s children with disabilities, who represent one in ten public school students statewide. Program expenditures amounted to $9.3 billion in 2006 –07, or more than 16 percent of K –12 general fund spending (Lipscomb, 2009b). Special education differs from most educational programs because children with disabilities in the United States are legally entitled to free, appropriate services based on their individual needs. T his service entitlement, along with the magnitude of expenditures and earlier finance reform , makes special education an important part of California’s education finance policy. Special education aid from federal, state, and local source s, at $4.7 billion in 2006 –07, is California’s largest pool of funding specifically for one education program. California allocate s most of this funding on a per -student basis, regardless of disability status; t he underlying assumption is that disabilities vary evenly across the population. Since enacting this funding model, California has reduced but not eliminated historical inequities in the per -student funding rate across the state. Inequities in funding rates reflect historical circumstances and are no t consistent with the type of special education finance system that California chose to adopt in 1997 . By design, California’s special education finance model does not reimburse school districts for expenditures related to greater reported special education needs. This policy encourages districts to serve students c ost-effectively , but it also means that the local share of spending on special education mandates can diff er substantially across the state. This report suggests that California consider funding rate equalization, with adjustments based on factors outside of district control —factors that arguably describe true cost variation. Such adjustments could partly account for differences in costs but without giving districts an incentive to over- represent special education needs. P recedent for this kind of funding adjustment comes from the federal government, which allocates part of its special education funding based on poverty. Within California, severe disability rates tend to be higher among children from lower -income families. California could adapt the federal model, and could add other factors too , such as v arying labor market conditions . Per -pupil spending, which is primarily for salaries, is sensitive to regional variation in labor market conditions for non-teachers with similar qualifications, even holding disability rates constant. These potential adjustment factors would describe different types of costs to meet special education mandates, with the first arguably related to incidence and the second related to the expected price of employee compensation . 6 Following this assessment, t his report then simulates how California could use the principles of the cu rrent system to allocate funds at an equal rate per student, adjusted by these two factors. The simulation adjusts for low-income stu dents and regional labor market conditions , but California could substitute other factors, such as an updated version of th e formula’s existing cost proxy . T hese refinements would redistribut e how the state allocates existing funds, but would also lead to fuller implementation of current funding objectives . All technical appendices to this paper are available on the PPIC website: http://www.ppic.org/content/pubs/other/809SLR_appendix.pdf 7 Acknowledgments I would like to thank several people who helped me learn about California’s special educ ation program, including P aul Goldfinger, Sarge Kennedy, Jack Lucas, Marcia McClish, Rick Miller, Sally Spaeth, Ray Reinhard, and Paul Warren. I also deeply appreciate the insightful comme nts that Kim Connor, Jaqui Guzmán, and Tom Parrish provided. This wo rk would not be possible without their careful assistance. At PPIC, Richard Greene, Hans Johnson, Eric Larsen, Heather Rose, Jon Sonstelie, and Lynette Ubois provided thoughtful suggestions that improved the study and Karina Jaquet provided outstanding research support. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or b oard of directors of the Public Policy Institute of California. 8 Introduction Special education is a mandated but under-evaluated part of state commitments to public education. U nder the federal Indi viduals with Disabilities Education Act (IDEA), children with disabilities have a legal entitlement to a “free a nd appropriate public education.” 1 The service entitlement alone makes special education finance an important policy issue for California and other states. But special education is also a multibillion -dollar program that serves more than 10 percent of California’s public school enrollm ent. In fact, the $4.7 billion in federal, state, and local special education aid that California school districts received in 2006–07 was their largest source of funding specifically for one program. Nevertheless, a growing concern in California is t hat expenses related to meeting special education mandates tend to “encroach” on other education funds. T his service entitlement makes special education different from most educational programs because it supersedes the availabili ty of funding. In effect, school districts must meet their mandate to serve disabled children before supporting other budgeted programs. 2 The state legislature overhauled the special education system in 199 7. Policymakers cited several undesirable properties of the prior system, such as funding inequity, complexity, and inappropriate placement incentives as reasons for pursuing reform. The reform bill, AB 602, enacted a new allocation model that addressed these issues by assuming that disabilities vary evenly across the state. The current formula distributes the predominant share of special education funds based on total student population size rather than the size of disabled populations or the needs of disabled students. T he requirement that school districts meet all special education needs appropriately, whether needs are rare or frequent, inexpensive or costly , still applies . The perception of budgetary tension between special education and other programs is especially clear in California , where school districts have few options for raising additional revenue to meet new special education demands. In recent years, special education spending has grown faster than spending on other programs while special education fund ing has grown more slowly than general -purpose funds. Spending on special education services in California totaled $9.3 billion in 2006 –07, over 16 percent of all K–12 spending (Lipscomb, 2009b). This report examines special education financing in California in 2006 –07, about a decade after AB 602 passed. It explores the funding process, patterns of disability, and patterns of spending on disabled children . It serves as the basis for evaluating California’s finance policy today and consider s whether the state should pursue further refinements. The first sectio n, on funding , examines the extent to which switching to a per -student funding system has led to greater funding equity. The following section, on disability rates , describes disability patterns with respect to income and factors like demography because special education need s may vary across the state despite the assumption that they do not. Lastly, we document expenditures , including the local share, and describe spending patterns among regions of California where personnel costs are arguably higher. 1 IDEA first passed in 1975 as the Education of all Handicapped Children Act. Congress last reauthorized it in 2004. 2 Regular education budgets funded 28 percent of special education spending in California in 2004 (Asimov, 2006). Harr, Parrish , and Chambe rs (2008) refer to encroachment as a growing policy concern in reviewing special education research. 9 We view special education financing and spending through the lens of IDEA requirements that special education aid help defray a school district’s additional spending on children with disabilities —above its average spending on all children. Concerns about e n croachment —the shortfall between additional spending on disabled children and special education aid —suggest that special education is raiding funding for other programs , although practically every state funds special education through a combination of fede ral, state, and local revenue . As Harr, Parrish , and Chambers (2008) explain , what is considered encroachment in one state may be considered the local share in another. This report adopts a similar view and when ever possible refers to encroachment as local support for additional spending on children with disabilities. Th is list summarizes key terminology used in the report: • Spending on children with disabilities —Combined special education and non- special education spending to educat e children with disabilities • Additional s pending on children with disabilities —Spending on children with disabilities above the average for all children • Special education funds —Federal, state, and local revenue reserved for educating children with disabilities • Local support for additional spending on children with disabilities —Additional spending that is not covered by special education funds (i.e., encroachment) T he report concludes that AB 602 improved special education finance from the previous system, b ut that California can take additional steps to implement the desired changes that reform intended to achieve, like greater funding equity . The conclusion suggests improvements , such as refinements to the allocation model itself, and describes how th ose mi ght work. 10 Financing Special Education in California The Funding Process Special education in California receive d $4.7 billion in funding from federal, state, and local sources in 2006 –07. 3 The California Department of Education (CDE) allocates these funds to 120 regional groups of school districts known as Special Education Local Plan Areas (SELPAs) that coordinate special education activities to offer a wider range of services more efficien tly. 4 Under AB 602 (1997), California has distributed most special education funds based on the average daily attendance (ADA) of each SELPA’s entire student population since 1998– 99. Disability counts and special education expenditures are not part of the funding equation. CDE effectively controls the total size of special education grants from all sources because it deduc ts federal and local funding from state aid when determining SELPA allocations. 5 California’s funding process is a census-based, or capitation, model. While not the predominant choice across the states, the use of census -based models has grown since 1991, and the federal government and nine other states now use them. 6 AB 602 addressed several key concerns with the finance model that had been in place. The old system, which allocated funds based on the number of classes in different instructional settings that districts reported , was widely seen as inequitable and overly complex ; d istricts received differ ent amounts of money for serving children in equivalent settings , even in the same SELPA . T ransitioning to a formula based on a flat grant per student incr eased funding equity and transparency. Streamlining the allocation formula also helped the legislature to increase district flexibility in using funds . In addition, s tate policymakers wanted to avoid incentivizing inappropriate special education placements , which can happen when school districts receive funding based on their self -reported needs. Under a census-based model, school districts that classify more children as disabled incur additional costs but receive no additional funding. The pure fiscal ince ntive actually is to identify fewer disabilities and provide less costly services. Most states continue to use more conventional formulas based on the population and reported needs of disabled students. The sensitivity of identification rates to funding incentives is well documented by researchers, whose findings suggest that census-based models are associated with lower disability rates. 7 3 Lipscomb (2009b) describes the components of California’s special education funding process. For example, Dhuey and Lipscomb (2009) estimate a relationship between states adopting census-based model s and an 8 –10 percent average reduction in their disability rate, mostly in the categories of learning disabilities and mental retardation ; n o category experienced a statistically significant increase in identification. Adopting these models w as also associated with greater u se of outside school placements for severely disabled children . These 4 SELPA membership in 2006 -07 ranged from a single district to 47 districts. Tulare is the SELPA with the most member districts. There were 36 single district SELPAs. California has over 1,000 school districts and county offices of education. 5 The exception is a small number of extraordinarily high cost placements . 6 The states are Alabama, Alaska, Connecticut, Idaho, Massachusetts, Montana, New Jersey, North Dakota, and Pennsylvania. Missouri, South Dakota, and Vermont use a partial census model. 7 For example, see Dhuey and Lipscomb (2009), Greene and Forster ( 2002), Kwak (2008), Lipscomb (2009a), and Mahitivanichcha and Parrish (2005). Cullen (2003) does not examine a census -based model but reaches similar conclusions about funding incentives in special education. 11 are the most expensive placements , bu t the increase may be because all census-based formulas provide some reimbursement for extraordinarily high -cost placements. Some evidence suggests a link between census-based reform and a higher rate of requests for dispute resolution in special education matters (Lipscomb 2009a). 8 The general findings from the research literature s upport conclusions about capitation payments in health care applications . Designing educational programs for disabled children may be come slightly more contentious between parents and schools when state s stop providing funds based on reported needs. 9 Funding Levels M anaged care systems typically reimburse providers based on the number of patients they see per month, rather than patient severity or the cost of treatment. The systems are cost-co ntainment strategies, but introduce a direct incentive for providers to seek healthier patients and provide fewer services. Census-based policies in special education are similar because f unding helps school districts meet individual needs but need plays little role in determining funding. In transition ing to a census-based model, California began equalizing the funding amount that students (regardless of disability status) generate for their SELPA. This amount , called the base rate, underpins most special education allocations in California. Figure 1 shows that California has reduced but not eliminated the disparities that existed in 1998. In fact, t he SELPAs that had high base rates in 1998 continue to ha ve high base rates today. The statewide average base rate has remained mostly constant over the last five years in 2006 dollars. Figure 1 SELPA Base Rates in 2006 Dollars NOTE: High and low base rates correspond to the 9 5th and the 5th percentile rates in California. The 5th percentile is higher than the rate that applies to 5 percent of California students. Data come from CDE. Base rate differences translate into differences in funding per student. The first chart in Figure 2 compares SELPA base rates in 2006 –07 with AB 602 funds , which represent 89 percent 8 The evidence comes from an “enrollment weight ed” specification, suggesting nationwide growth in the rate of dispute resolution requests per special education student. Dhuey and Lipscomb (2009) treat states equally regardless of size. They find a statis tically insignificant average response across ref orm states. 9 Newhouse (1996) reviews this literature. 400 450 500 550 600 650 700 1998 19992000200120022003200420052006Dollars per Average Daily Attendance HighStatewide AverageLow 12 of federal, state, and local special education revenue. 10 There is a visible upward relationship. The relationship is not perfect because the allocation formula has several adjustment factors, such as for regionalized services in small SELPAs (less than 15,000 ADA) and for SELPAs eligible for a Special Disabilities Adjustment (SDA). (SDA funds are for SELPAs that the legislature found to have a greater incidence of hi gh cost disabilities in 1998 but relatively lower base rates. The SDA is the only cost proxy in place. The next chapter discusses the SDA in further detail.) Figure 2 Relationships Between Base Rate s and Special Education Funds, 2006–07 SOURCE: AB 602 Funding Exhibits The second chart in Figure 2 removes these adjustment factors. The remaining funds still represent 81 percent of federal, state, and lo cal special education revenue. The link between base rates and funding becomes nearly one to one, suggesting that historical inequities explain much of the difference in funding across SELPAs today. Further descriptive analysis in the Appendix corroborates this conclusion. The anal ysis adjusts the overall amount of special 10 AB 602 does not allocate funds for several types of services, the largest of which are for special education transportation and infants with disabilities. 500 600 700 800 900 1,000 1,100 1,200 500 600 700 800 900 1,000 1,100AB 602 Funds per Student SELPA Base Rate 500 600 700 800 900 1,000 1,100 1,200 500 600 700 800 900 1,000 1,100Base Allocation per Student SELPA Base Rate 13 education funds per student for differences in the base rate, whether SELPAs serve fewer than 15,000 students, and whether they are eligible for SDA funding. These three factors alone explain nearly 80 percent of the variation in funding , primarily because of the association between base rates and funding per student. 11 T he lack of full base rate equalization has little justification more than a decade following AB 602’s enactment because funding equity is one of the main rationales for adopting a census-based model . Figure 2 indicates that factors like the SDA help somewhat by providing more revenue to some SELPAs with lower base rates. But California could further AB 602’s goals of equity and transparency by starting with a level playing field for everyone. The state could then consider funding adjustments based on factors outside of SELPA control that are good proxies for true cost variation. A criticism of pure census models is that they do not account for variation in student need (Parrish, et al., 2003). The prob lem is that identifying good proxies is difficult. Many of the adjustment factors we think of first, like disability rates and spending, are to some extent within a district’s control. Adjusting funding based on these measures reintroduces the same inappropriate incentives issue census-based model s were supposed to avoid. When faced with this same problem in allocating IDEA funds , the federal government decided to adjust apportionments for child poverty rates. In general, health outcomes tend to improve with socioeconomic status, so the poverty adjustment arguably helps account for part of the variation in special education need . The federal formula distributes 85 percent of funds based on population and 15 percent based on poverty. T his report examines patterns of disability and spending in California to identify potential factors that could arguably serve as proxies for true cost variation . It then illustrates how California could incorporate these factors into an allocation formula . 11 Appendix Table A.1 provides summary statistics for the variables. Table A.2 contains the results. 14 Disability Rates and Income All census models assume that disabilities are spread evenly across the population. California justified AB 602’s goal of equalizing funding per student based on the premise that “handicapping conditions of similar severity” occur with “roughly equal frequency.” 12 AB 602 directs special education funds to SELPAs in part to allay fears that the equal frequency assumption may not hold for smaller populations like school districts. 13 A 1998 report by the American Institutes for Research ( AIR) concluded that severe/high cost disabilities did vary across SELPAs more than could be expected randomly. 14 Based on that , the legislature added the SDA program to the allocation formula as a cost proxy. The SDA provides a severity supplement to some lower -funded SELPAs based on the services that their high cost special edu cation students received in 1997. In 2006 –07, SDA funds provided $81 million to 34 SELPAs or about $34 per student in eligible SELPAs. The legislature has never updated the incidence multipliers used to determine eligibility. 15 Although the SDA derives from historical data, it identifies SELPAs with current higher rates of severe disabilities. Table 1 indicates that in 2006–07, the rate of severe disabilities was 15 percent higher in S DA-eligible SELPAs . This study defines disability severity at the category of disability level, following the delineation that California uses in its financial data. 16 Severity clearly varies within categories too, meaning that the delineation is imperfect. Yet the categories in the severe group tend to be more costly to service, suggesting that grouping disabilities by category is reasonable across the population. 17 Table 1 Severe Disability Rates and Income by SDA Funding Status, 2006 –07 Severe Disabilities (% of Students) Free or Reduced-price Meals (% of Students) SELPAs Receiving SDA Funding 2.59 57.64 SELPAs Not Receiving SDA Funding 2.27 46.40 NOTE: Sample based on 1 19 SELPAs. The proportions in each column are statistically different at the 5 percent level. The z -statistics are 26 and 274, respectively. The AIR (1998) study recommended a cost adjustment based on services received by high -cost students partly because a measure like poverty did not reliably explain the observed differences in s everity across California a decade ago . This study find s some evidence in recent 12 AB 602 Bill Analysis (1997) 13 This was a recommendation from a 1995 report published by the Legislative Analyst’s Office, the Department of Education, and the Department of Finance. 14 Parrish, Kaleba, Gerber, and McLaughlin (1998) 15 The SDA incidence multipliers also help determine cost -of -living adjustments and growth funding. 16 Based on the California School Accounting Manual’s (2008), severe disabilities include autism, deafness, d eaf-blindness, emotional disturbance, mental retardation, multiple disabilities, orthopedic impairments, traumatic brain injury, and visual impairments (including blindness). Non -severe disabilities are learning disabilities, speech or language impairments , and other health impairments. 17 Parrish, Harr, Kidron, Brock, Anand (2004) estimate disability costs for California in 2002– 03. 15 data of a negative relationship between disability status and income. SDA eligibility also appears to correlate with income. For instance, Table 1 indicates that SELPAs receiving SDA funds have a higher rate of participation in free or reduced -price meals , a program with an income eligibility cap at 185 percent of the federal poverty level. Findings from the California Health Interview Survey Figur e 3 uses the California Health Interview Survey (CHIS), the largest state- representative health survey in the United States, to show that severe conditions tend to be less common among children from higher income families. CHIS surveyed parents about their children ’s disabilit ies, and the sample includes 6,515 children ages 5 to 11 in 2005. The findings from the analysis are representative of California children in that age range. Figure 3 Income and Child Disability Conditions, California Health Interview Survey, 2005 NOTE: Each income category contains 20 percent of the sample. Table A.3 provides summary statistics on the CHIS variables. Adjusted disability rates come from regression estimates in Appendix Table A.4. The horizontal axis divides the samp le into five equally sized groups based on income. The first and third columns on the horizontal axis show the average rate of severe and non - severe conditions in each group . Both show a negative relationship with income, although the relationship is smoot her for severe disabilities. The second and fourth columns show the rate of each type of condition after adjusting for differences in children’s gender, language spoken at home, rural setting, race /ethnicity, birth weight, and age. Controlling for these factors does not substantively change the relationship between income and disability status. The findings in Figure 3 support general research conclusions about correlations between health outcomes and socioeconomic status. 0 1 2 3 4 5 6 7 0 to 1.4 Times Federal Poverty 1.4 to 2.7 Times Federal Poverty 2.7 to 4.3 Times Federal Poverty 4.3 to 6.5 Times Federal Poverty At Least 6.5 Times Federal Poverty Percentage of Children Income Categories SevereSevere (Adjusted)Non-severeNon-severe (Adjusted) 16 The association between disability and income exists for behavioral and mental conditions but not for physical conditions. Appendix Table A.4 reorganizes the severe/non- severe delineation based on whether a disability is behavioral/mental or physical. The estimates show a strong relationship between income and the former, but a small and weak relationship between income and the latter. The findings suggest that the probability of a behavioral disability is 10 percent lower for a child at 200 percent of the poverty line than it is for a child at th e poverty line, adjusting for other factors . Findings for Children in Special Education Programs Recent CDE data on actual special education enrollment among children with severe disabilities suggests a similar negative relationship with income. Th e analysis in Appendix Table A.5 constructs rates of severe and non-severe disabilities for each SELPA for 2006–07. It then uses a regression to adjust the rates for a similar set of characteristics as in Figure 3. 18 The income measure is the percentage of students in a SELPA enrolled in the free or reduced -price meals program. Unlike the CHIS analysis, which compares family income and disability conditions at the individual level, the analysis using CDE data compares a SELPA’s disability rate to its percentage of students in free or reduced -price meals. In other words, it only reports associations between aggregated data. While this is an important difference, the SELPA is also the level to which AB 602’s assumption of even disability rates applies. Holdin g constant other observable characteristics, a lower rate of free or reduced -price meals (i.e. higher income) is related to a lower severe disability rate. The findings suggest that the severe disability rate is 12 percent higher in a SELPA with 60 percent of students in free or reduced -price meals than it is in a SELPA where 30 percent of students are in free or reduced - price meals. The average rate of free or reduced -price meals across SELPAs is about 46 percent. 19 The relationship between free or reduced-price meals and rates of non -severe disabilities is much weaker. Part of the explanation is that grouping disabilities under severe and non - severe headings masks differences with respect to income at the category of disability level. To illustrate this, th e remaining columns in Table A.5 show the estimated relationship between free or reduced -price meals and the percentage of children classified in each of the six largest categories of disability. 20 Similar opposing relationships exist for severe disabilities too. M ental retardation is more common in lower -income areas while autism is more common in higher -income areas. These categories account for over 90 percent of disabilities in California. The data suggest that there are opposing relationships in some cases. Specifically, learning disabilities are more common in lower -income areas while other health impairments (e.g. ADD and ADHD) are more common in higher -income areas. These opposing relationships contribute to a weak association between income and non -severe disabilities overall. 18 The model includes the following characteristics: percent free or reduced- price meals, an index of regional non-teacher wage levels, percent English learners, town or rural location, race/ethnicity, SELPA enrollment, AB 602 base rate per ADA, the average SELPA revenue limit per pupil, a single- district SELPA indicator, and a constant. 19 Table A.1 provides summary statistics and Table A.5 contains the regression results. 20 The categories are learning disabilities, speech or language impairments, other health impairments, mental retardation, autism, and emotional disturbance. Lipscomb (2009b) provides disability definitions. 17 Overall, however, rates of severe disabilities tend to be higher in SELPAs with higher proportions of low -income students. The findings also suggest several significant relationships with severe disabilities besides that with income. Holding constant other factors, the rate of severe disabilities is lower in SELPAs with higher concentrations of Hispanic and Asian students, lower concentrations of African -American students, towns and rural areas , larger SELPAs, and single -district SELPAs . Hispanic and Asian children have lower rates of special education classification in California while African -Americans have higher rates relative to non-Hispanic white children (Lipscomb, 2009b). Differences in classification rates by race and eth nicity are most pronounced in the categor ies of emotional disturbance , learning disability , and other health impairment. The finding about small towns and rural areas suggests that urban settings may offer a wider range of local care options : the greater availability of therapy and medical services overall may attract families with severely disabled children. The fact that t hat severe disability rates are also higher in larger and single -district SELPAs appears to support this theory. Single -district SELPA s tend to have both above -average district enrollments and to be located in urban locations. The AIR study by Parrish et al (1998) found that single-district SELPAs spend more per student than do others . In sum, both the CHIS and actual special education enrollment data suggest a relationship between income and disability status, particularly for severe disabilities. A factor related to income, such as free or reduced -price meal eligibility, may be an appropriate proxy to identify SELPAs that face a higher rate of severe special education needs. An income-based adjustment would have both potential advantages and disadvantages in relation to the existing cost proxy, the SDA. The advantages are that it is entirely out of SELPA control , that it could be update d regularly , and that it would align closely with the federal formula. But an income -based adjustment has potential disadvantages too. A 2004 follow -up report by AIR recommended that California instead update the SDA , partly because of the differing patterns of mental retardation and autism with respect to poverty. AIR also cite d the possible social stigma attached to enrolling in free or reduced -price meals at the high school level. California could address this issue by coll ecting data on income, rather than enrollment, to determine meal program eligibility . Further, relationships between poverty and disability rates do not inform the question of how much additional funding SELPAs need because of their higher poverty rate. Wh en the federal government experienced this issue, it adopted an 85/15 compromise between the population and poverty -based portions of its allocation formula. Because of the valid concerns about an income -based adjustment, maintaining the existing cost prox y, the SDA, would be a sensible way to go , too. In this case, California should update the multipliers that determine funding to maximize the SDA’s effectiveness as a proxy for costs today. 18 Spending on Children with Disabilities in California Both the prevalence of disabilities and the amount of special education funds affect s chool spending on disabled children . The link to funding comes from IDEA , which requires that districts use federal assistance for disabled children to help pay the “excess costs” of educating them (Federal Register, 2006). 21 Conceptual Framework School districts incur excess costs when they spend more educating disabled children than they spend on average on all children. T his chapter documents spending levels and patterns across SELPAs. As in the previous chapter , the emphasis is partly on identifying a factor outside of SELPA control that arguably serves as a proxy for cost variation. The focus here is to account for differences in regional labor market conditions for educators , a different type of cost from the severity of student needs. Despite the name, excess cost is actually a measure of spending. Costs are defined as the minimum expenditure for the services a student needs. Expenditures exceed costs when needs are not identified correctly or when districts are not providing services efficiently. Patterns of spending may resemble, but are not necessarily the same as, patterns of cost. To underscore the distinction, t his report refers to excess costs as ad ditional spending on children with disabilities. Special education and regular education share spending on children with disabilities. For example , children with speech impairments may receive speech therapy instruction on a regular basis but are otherwise in the regular classroom, while children with severe mental retardation may spend most of the school da y outside the regular classroom. Figure 4 illustrates how this sharing works. The dashed line represents a verage spending on all children. Additional spending on children with disabilities is above the dashed line. Special education funds help to defray these amounts, and local funds pay the rest. Figure 4 Illustrating School Spending on Nondisabled and Disabled Children NOTE: Figure 4 is strictly illustrative, including the dollars per student shown on the vertical axis. 21 This requirement is also listed in California Education Code Section 56841(a). 0 4,000 8,000 12,000 16,000 Children without Disabilities Children with Disabilities Dollars per Student Non-Special Education SpendingSpecial Education Spending Special Ed Funds Average Spending on All Children Local Support (Encroachment)AdditionalSpending (Excess Cost) 19 In California, s tate aid for special education effectively shares the same purpose federal funds do in helping d efray additional spending on children with disabilities. The effective intent of state aid is the same b ecause CDE deducts federal and local special education funds from state aid when calculating SELPA allocations . The only way for California to show that districts are using IDEA funds as intended is to treat state allocations in the same manner. That said, the language in California’s education code suggests a slightly different purpose . The cited purpose is to help local districts “provide special education and related services to individuals with exceptional needs.” 22 The language of the education code contribute s to confusion about the term “encroachment ,” which refers just to the local support amount above the dashed line. “Encroachment” conventionally describes any situation where local funds close a gap between special education spending and funding. In other words, state aid is for the entire special education area in Figure 4 according to the education code , although it is actually just for the part above the dashed line because of the way California deducts funding from non-state sources when calculating SELPA allocations . 23 This usage may be common and appears to derive straight from the education code, but it needs to be better defined because special education funds are not meant to pay for all progra m spending. Children with disabilities generate other education funds for schools too, which schools can use for special education spending. As mentioned in the Introduction, partial local support for special education finance is the norm nationwide. Only Wyoming and Hawaii reimburse 100 percent of special education spending (Parrish, et al., 2003). 24 Districts draw from their other education funds to pay the local support portion shown in Figure 4. Although this money relates to a school district’s ex cess fiscal needs, it is an imperfect measu re because it describes expenditures and not costs. (It is important to remember that patterns of spending are not necessarily the result of differences in cost. ) Variations among SELPAs in local support could result from other reason s too, such as parents who have g reater demands for special education services in one SELPA than in another. In California, state special education aid help s pay just the part of spending above the dashed line in Figure 4. Clarifying the intent of state special education aid is a suggestion for California to consider. That said, special education expenditures are likely to stand as a better proxy for costs than other types of school spending because legal contracts between parents and schools under IDEA delineate the special education services that schools must provide (Harr, et al., 2006). School districts also have a clear incentive to operate special education programs efficiently because each additional dollar of local support diminishes available funds for other programs. Maximizing efficiency brings expenditure s and costs in closer alignment. School administrators tend to view local support for additional spending on disabled children as evidence of insufficient government fundin g for meeting special education 22 California Education Code Section 56836.04(b) 23 For instance, the first page returned from a Google search for “special education encroachment California” on December 8, 2008 was a financial report from the Fullerton School District (2007). The report budgets $15.1 million for special education serv ices, in cluding $6.7 million in encroachment costs. 24 Hawaii reimburses at a 100 percent rate because it operates a single school district for the entire state. 20 mandates. Additional funding is certainly a possible solution. So is looking for further cost - saving efficiencies and improving the way the state allocates existing funds. Federal stimulus money (The American Recovery and Re investment Act [ARRA] of 2009) may provide a rare opportunity for schools to reduce local spending on children with disabilities , through an additional $1.3 billion in one-time IDEA funds to California school districts. D istricts may be able to use half of any increase in federal assistance to reduce local and state spending. 25 Federal stimulus funds may also lead to greater efficiencies over a longer period if schools invest them with that potential in mind. For instance, schools could invest in professional development strategies aimed at serving children with disabilities in the regular classroom as effectively as possible, or in new diagnostic equipment to better identify student needs. 26 Expenditure Totals for 2006 –07 California public schools spent $53.1 billion , or about $8,447 per pupil, educating students in 2006–07 (Table 2). More than $9.3 billion went to providing special education and related services. This is part of total spending on children with disabilities. The other part is non -special education spending on disabled children . 27 This latter amount must be estimated because the state’s education finance data do not separate regular education spending for disabled and non disabled children. 28 We find that the regular education port ion was about $2.7 billion for 2006 –07. In other words, total spending on children with disabilities was $12.0 billion, or about $17,633 per disabled child in the state. 29 The average amount of spending per disabled child in California is roughly 19 percent higher than previous estimates for the nation. 30 Lipscomb (2009b) finds a similar difference between California and past national estimates for the special education component of spending alone. 25 Ordinarily, school districts must spend more on special education from local or combined state and local funds in a year than in the prior year (in total or in per pupil terms) to meet funding “Maintenance of Effort” requirements. 26 The ARRA example considers an increase in funding for children with disabilities. In theory, an increase in general- purpose funding can also reduce local support for additional spending by raising the dashed line in Figure 4. 27 Non -special education spending is not the same as spending in the regular classroom environment because children can receive special education serv ices in the regular classroom. 28 The method in this report is to prorate non- special education spending based on the percentage of the school day that children with the same disability in California spend inside the regular classroom. For instance, each nondisabled student gets one share of non- special education spending. Disabled students get a fractional share based on their disability. See Appendix B for full details. 29 The financial data includes all special education expenditures on behalf of children up to age 22. As a result, average spending per disabled child is found by dividing $12.0 billion into total special education enrollment among children up to age 22. 30 The Center for Special Education Finance (2003) reports $13,054 in average spending per school -aged special education student in 2001– 02. This converts to about $14,819 in 2006 dollars. The inflation adjustment is the Consumer Price Index for the west region of the United States. 21 Table 2 Education Spending per Student by Disability Status, 2006 –07 Dollars (billions) Dollars per Total Enrollment Dollars per Disabled Child Dollars per Nondisabled Child Total Expenditures 53.1 8,447 Special Education Expenditures 9.3 1,474 13,642 Non-Special Ed Spending on Children with Disabilities 2.7 431 3,991 Total Spending on Children with Disabilities 12.0 1,905 17,633 Total Spending on Children without Disabilities 41.1 6,542 7,334 NOTE: Appendix B describes the methodology . All amounts are net of capital outlay and debt service expenditures. The sample includes data from all school district s, county office s of education, and transportation joint powers agencies . Total enrollment is 6,282,036 K–12, ungraded, and adult educatio n students. S pecial education enrollment is 678,699 children with disabilities ages 0 –22. By comparison, California schools spent ab out $41.1 billion educating non disabled children in 2006 –07, about $7,334 per child . Spending on disabled children was 2.4 times higher than spending on nondisabled children. The spending ratio in California is somewhat higher than spending ratios estimated for the nation. Harr, Parrish, and Chambers (2008) summarize the results from four studies analyzing data between 1968 –69 and 1999 –2000. The spending ratios for the nation in the se studies range from 1.90 to 2.29. 31 California ’s higher spending ratio may reflect several factors. For one, the average severity level of special education students in California may be higher than in other states. California has the nation’s lowest rate of special education identification, well below the national average of about 14 percent (Lipscomb, 2009b). A lower rate of classification among children with the least severe disabilities may acco unt for much of this difference. If so, the population of special education students in California has a bigger share of severely disabled children, helping to explain its relatively higher spending ratio. The interaction between the service entitlement, California’s higher personnel costs, and its lower overall rate of spending per student may also contribute to a higher spending ratio. Personnel costs represent 80 percent of district expenditures in California (Rose and Sengupta, 2007) . According to the U .S. Department of Education’s National Center for Education Statistics (NCES), college graduates in California earn higher salaries than in practically every other state. 32 The share of school spending devoted to children with disabilities resembles what is found in other states, but California classifies fewer students that way. Disabled children account for 10 percent of enrollment in California but 22.5 percent of spending. Nationwide estimates for 1999 –2000 found that 12.1 percent of students were disabled, and that 21.4 percent California s chool districts need to pay these higher salaries to attract quality educators and, like other states, they need to meet special education mandates . To the extent that school districts in California meet special education mandates at higher costs than other states but spend less overall, they have less revenue available to support other programs. 31 It is possible that the national spending ratio has grown since 1999–2000 because of growth in high cost disabilities like autism. 32 See the Comparable Wage Index data at www.nces.ed.gov. 22 of spending was devoted to them (Harr, et al., 2006). California may spend more, but it does not spend the most. The Special Education Expenditure Project (SEEP) analyzed spending in 11 states between 1999 and 2001 . Adjusted for inflation, the findings suggest that five of the states exceeded California’s average spending on disabled children in 2006–07. 33 Additional Spending on Disabled Children and Local Support Figure 5 illustrates California’s $12 billion in spending on children with disabilities in the same way as Figure 4. The first portion , at the bottom of the column, is for non -special education spending. The next part is special education spending below the dashed line that represents average spending on all children. Together, these components total ed $4.4 billion in 2006– 07. Additional spending accounts for the remaining $7.6 billion. Special education funding offsets about $4.7 billion in additional spending, leaving $2.9 billion in local support. Altogether, special education funds offset 62 percent of additional spending on disabled children. The local support portion is the remaining 38 percent , or 24 percent of total spending on children with disabilities. Figure 5 California School Spending on Children with Disabilities, 2006 –07 NOTE: Appendix B describes the method used for each calculation. Data come from the Standardized Account Code Structure a nd IDEA Educational Environment records for 2006. We measure additional spending and local support at the SELPA level based on each SELPA’s entire student population. Spending and funding at the district level has more to do with local plan a greements than how the funding formula works. The methodology for calculating spending in this report follows federal regulations closely (see A ppendix B) except for two main differences. First, the federal regulations describe a district-level calculation. 33 SEEP examined spending in Alabama, Delaware, Indiana, Kansas, Missouri, New Jersey, Ne w York, Ohio, Rhode Island, Maryland, and Wyoming (Center for Special Education Finance, 2003). 2.7 1.7 4.2 0.5 2.9 0 1 2 3 4 5 6 7 8 9 10 11 12 Spending ($ billions) Local Support for Additional Spending Special Education Funds (Non-AB 602) Special Education Funds (AB 602) Special Education Spending to Reach Excess Cost Threshold Non-Special Education Spending Additional Spending ($7.6 billion) Special Education Spending ($9.3 billion) 23 Second, they specify that districts should make separate calculations for elementary and secondary students. Neither of these differences is likely to change the finding that there is a wide range of additional spending and local support per student across California. Fo r instance, Table 3 shows that local support represented $1,004 of $1,830 in additional spending per student at Mt. Diablo Un ified School District (USD) SELPA. This is 120 percent and 52 percent higher , respectively, than the statewide average of $455 and $1,201 per student. Special education funds appear to be more than sufficient to offset all additional spending on children with disabilities in a few SELPAs. For example, Sierra County Office of Education ( COE) SELPA ha d about $930 per student in additional spending but receive d $1,188 per student in special education grants. Table 3 SELPAs with the 10 Highest and Lowest Levels of Local Support per Student , 2006–07 SELPAs with the Most Local Support per Student SELPAs with the Least Local Support per Student SELPA Name Additional Spending Special Education Revenue Local Support SELPA Name Additional Spending Special Education Revenue Local Support Mt. Diablo USD 1,830 826 1,004 Sierra COE 930 1,188 -258 Los Angeles USD 1,794 851 943 Trinity COE 1,016 1,132 -116 San Diego CUSD 1,831 900 932 Lassen COE 972 1,040 -68 Newport -Mesa USD 1,636 728 908 Colusa COE 747 773 -27 Santa Clara I 1,557 672 885 Santa Clara III 1,020 983 37 North Orange 1,594 764 830 Siskiyou COE 955 902 53 Tri-City (Culver City USD) 1,489 780 709 Tehama COE 812 740 71 Santa Clara II 1,417 710 707 Imperial COE 728 649 79 North Region (Albany) 1,419 729 691 Humboldt/Del Norte 804 725 79 San Mateo COE 1,464 779 685 Modoc COE 1,403 1,320 83 NOTES: See Appendix Table B .2 for the complete list of SELPAs. The statewide per pupil average values are $1,201 of additional spending , $746 of special education revenue, and $455 of local support. SELPA funding clearly differs across California , but not because of differences in spending ; SELPAs with higher additional spending per student tend to have higher amounts of local support per student (Figure 6) . The solid line in the figure would be flatter if SELPAs with higher additional spending instead tended to receive more special education aid per student. Figure 6 describes an empirical relationship in the data, but not necessarily a policy concern because census-based models are not supposed to track spending. To the extent that spending and costs align closely, however, Figure 6 suggests that adjusting allocations based on cost proxies may help equalize the amount of local support per student across the state. 24 Figure 6 Additional Spending and Local Support for California SELPAs , 2006–07 SOURCE: Author’s calculation based on A ppendix Table B.2 Spending and Regional Non- Teacher Wages Patterns of spending on children with disabilities relate to regional labor market conditions in California . As mentioned earlier, the average salary for college -educated workers in California is among t he highest across the states, and employee compensation is the predominant expenditure for schools. T eacher compensation varies considerably across California (Rose and Sengupta, 2007) ; for 2003–04, for a mid -career teacher, it ranged from less than $55,000 in Yolo and the North Coast counties to more than $70,000 in Santa Clara and Orange Count ies. 34 When the price of resources is high er, school districts have less purchasing power under a fixed budget , meaning that they need to look for efficiencies in their program offerings. D istricts have less flexibility in special education offerings because meeting special education mandates supersedes budgetary concerns. For example, if a student needs a special education aide in the regular classroom, schools must provide one whether they are in Yolo or Santa Clara Counties . Rose and Sengupta (2007) f ound that the wages of non -teachers with similar educational attainment as teachers provide a good benchmark for contextualizing differences in teacher compensation across labor markets in California . They developed a comparable wage index (CWI) that compares non -teacher wages in a regional labor market to the statewide average. Figure 7 shows a positive relationship between non-teacher wages and local support for additional spending on disabled children. The value of the index is higher for SELPAs located where non -teachers earn higher average wages. 34 These salaries are for teachers with 10 years of experience and 60 credits beyond a bachelor’s degree. -300 -100100 300 500 700 900 1,100 0 500 1,000 1,5002,000 Local Support per Pupil Additional Spending per Pupil 25 Figure 7 Local Support for Additional Spending and Regional Non-Teacher Wages, 2006–07 NOTE: The statewide average value of the CWI is 1. The CWI serves as a proxy for variation in the personnel costs of hiring and retaining educators because it corresponds to the wage that teachers can expect outside of teaching. 35 It is helpful in studying patterns of spending because it is outside the control of school districts and teachers’ unions. External labor market conditions are a different type of cost for districts than the amount and severity of special education needs. 36 Appendix Table B.3 adjusts the relationship in Figure 7 for differences in a number of observable SELPA characteristics. These includ e the percent enrolled in free or reduced-price meals, percent English learners, urban or rura l location, race-ethnicity, total enrollment, measures of special education and general purpos e funding, and an indicator for single-district SELPAs. The findings indicate that both addi tional spending and local support per student relate positively to the comparable wage index, holding these factors constant. 37 The analysis supports the same conclusion by further controllin g for the rate of both severe and non-severe disabilities. Patterns of spending relate to factors beyond types of disability. 38 Figure 8 illustrates the expected rate of spending if the CWI were 10 percent above average, adjusting for observable SELPA characte ristics and rates of disability. The findings suggest that additional spending on disabled children pe r pupil would be about 4.3 percent higher than average. 35 See Rose and Sengupta (2007) for illustrations of the relation ship between teacher and non-teacher wage levels across counties in California. 36 Comparable plots to Figure 7, available up on request, show little evidence of a relationship between the comparable wage index and either the overall rate of special educat ion or the rate of severe disability. 37 The findings are numerically identical because specia l education funding per student is held constant. 38 Several other variables are significant as well. Higher addition al spending relates to lower enrollment in free or reduced-price meals, a higher concentration of English learners, larger SELP As, more special education funds per student, single-district SEL PAs, and higher severe disability rates. The lowe r rate of additional spending in SELPAs with larger rates of free or reduced-price meals appears to reflect a higher rate of spending ov erall (i.e. a higher dashed line in Figure 4). -400 -2000 200 400 600 800 1,000 1,200 0.7 0.8 0.9 1 1.1 1.2 Local Support for Additional Spending per Pupil Comparable Wage Index 26 Figure 8 Spending in a SELPA with a CWI 10 Percent above Average, 2006 –07 NOTE: Asterisk indicates statistical significance at the 10 percent level. The findings come from columns 3 through 5 in A ppendix Table B.3. Per-pupil spending includes both disabled and nondisabled students in the denominator. For comparison, Figure 8 illustrates similar relationships for the overall rate of spending on disabled and nondisabled children. T he association between the CWI and spending per pupil on disabled children is the larger of the two, holding constant other factors. Spending on disabled children is expected to be about 3.7 percent higher than average while spending on the nondisabled is expected to be 2.2 percent higher. The association is more precise for disabled children as well. In fact, the CWI in Ap pendix Table B.3 is not statistically significant in describing patterns of spending per pupil on nondisabled children. In most states, school districts have the option to try raising additional income if resources are expensive. In California, however, the state largely sets education revenue and leaves few options for local school districts to raise funds. Rose and Sengupta (2007) propose using the CWI to help equalize the purchasing power of school distric t budgets. At least two other recent policy reports on California education finance share this recommendation (Sonstelie, 2007; Bersin, Kirst, and Liu, 2007) . Sonstelie (2007) reaches this conclusion after applying a theoretical framework of economic decisionmaking under a fixed budget and a set of resource costs to school finance and survey data. Using a theoretical model to guide the analysis helps to moderate concerns about using expenditure data to describe patterns arguably related to cost. The empirical strategy in this study resembles the one suggested by the theoretical model in Sonstelie (2007). The Rose and Sengupta (2007) and Sonstelie (2007) studies were written as part of the Getting Down to Facts research project on California school finance and governance that was organized by Stanford University. The Bersin, Kirst, and Liu (2007) proposal grew out of the findings. This latter study proposes a funding system that consists of a base grant per student, an equalized special education grant per student, targeted funding for low-income students, and a regional cost adjustment. The findings in this section support a similar conclusion about the CWI and special education funds. 0% 1% 2% 3% 4% 5% Additional Spending per Pupil * Spending on Disabled Children per Pupil * Spending on Nondisabled Children per Pupil Rate of Spending above the Average Across SELPAs 27 Suggestions for Improving Special Education Finance California’s switch to per -student funding via AB 602 (1997) improved the state allocation formula by placing a greater e mphasis on funding equity, transparency , and flexibility while minimizing incentives to classify students inappropriately. But more than a decade later, California’s census -based funding system shows signs that still more could be done. Children still gene rate different amounts of special education funding depending on the SELPA in which they live. Moreover, the variation s in funding rates today reflect the historical disparities that existed under the previous funding regime. California could take the following two steps to ward implementing more fully the type of special education finance system it chose to adopt in 1997. • Refine the allocation model. Equalize the base rates and adjust funding for a small number of factors outside of SELPA control. • Clarify the state’s objective for special education funds. Emphasize providing appropriate services for educating disabled children rather than providing special education services alone. Refine the Allocation Model California should preserve its census-based approach for special education finance, but it can improve upon the existing design. The first step is to complete the equalization of base rate s per student across the state. The second step is to adjust funding for a small number of factors that are outside of SELPA control and could serve as a proxy for true cost variation . The resulting model would aim to be sensible and simple, furthering the existing funding goals. As a census -based model, schools would be unable to influence funding levels based on the way they classify and serve students. F urther, it could offer SELPAs greater flexibility in using funding , should policymakers decide to consolidate several sources of special education revenue into one allocation. According to a recent report by the Legislative Analyst’s Office (2008), SELPAs receive state special education funds through 15 separate programs. The report argues for merging many of these funds because the current allocation method makes it hard to see how much funding the state provides and how it ultimately distributes funds. This study suggest s a possible form for a consolidated formula that involves equal base rates, with adjustments for eligibility in free or reduced -price meals, and regional non -teacher wage levels. The formula could resemble the federal IDEA funding process, distribut ing 85 percent of funds based on enrollment and 15 percent based on poverty. Such a formula would also re semble the Bersin, Kirst, and Liu (2008) prop osal for California’s K–12 education finance system that includes both funding for low -income students and a regional wage adjustment. Ap pendix C provides technical information. 28 Each SELPA’s funding allocation, F , w ould take the following form: The formula has two components, a base and a regional wage adjustment. Most of the base (a 0.85 weight in this example) would come fr om multiplying the statewide base rate by total SELPA enrollment. The rest is an a djustment for low-income students. The adjustment is the statewide base rate multiplied by the number of students eligible for free or reduced -price meals. The formula then adjusts a SELPA’s base funding for regional wage levels. The average value of the regional wage adjustment is one . The statewide base rate depends on how much money California consolidates into the allocation. Table 4 compares the simulated allocation to the current allocation using funds from the AB 602 base entitlement (second chart in Figure 2). Appendix Table C.1 provides a similar analysis using all AB 602 funds (first chart in Figure 2). 39 Table 4 Base Allocation Funds U nder Simulated and Current Models, 2006 –07 The respective statewide base rates are $673 or $740 per student. California could apply future cost of living adjustments or other funding supplements directly to the base rates. A. Statewide Average, Low, and High Values Average Lowest Highest Simulated Allocation 605 491 692 Current Base Allocation 605 547 995 B. Model Comparison Comparable Wage Index Simulated Allocation Current Base Allocation Low Medium High Low Medium High .74–.92 .93–.98 .99–1.15 .74–.92 .93–.98 .99–1.15 Percent Free or Reduced Price Meals Low 0–38 533 581 620 632 604 625 Medium 39–56 542 596 647 616 582 594 High 57–100 560 618 659 652 599 646 NOTES: Summary statistics are weighted by total SELPA enrollment. The low, medium, and high categories in the model comparison each include about one -third of SELP As. Table C.1 provides a similar model comparison using all AB 602 funds. Section A shows that the simulated allocation maintains the actual level of base funding that was available to SELPAs in 2006 –07. The difference is in how funding is distributed, wi th 39 The simulation in Appendix Table C.1 uses all AB 602 funds for simplicity. Policymakers should leave some AB 602 programs, such as funding for out- of-home care, unchanged. 29 the simulated allocation reducing funding variation across the state. By design, funding per student under the simulated model (see Section B) is highest for SELPAs with large proportions of low er-income students in regions with higher expected personnel expenses. Funding per student is smallest for SELPAs that have the opposite student and personnel characteristics. 40 The simulated allocation in Table 4 is just one possible way for policymakers to refine the funding formula and stay consistent with the finance reform goals of 1997. In practice, the state legislature c ould adjust the 85/15 weighting between the enrollment portion and the low- income adjustment. It could also identify and use other adjustment factors. For instance, California could maintain the SDA, the existing AB 602 cost proxy. In this case, the state should consider whether updating the SDA’s eligibility criteria after a decade would provide an even better cost proxy . T he existing allocation shows less consistent patterns with respect to regional wage levels and proportions of low -income students. For instance, SELPAs in the low/low categories currently have among the higher average rates of base funding. R efining the funding formula along the lines of Table 4 would help California move closer to fully implementing its census -based finance reform goals. Recent policy research on education finance in California supports this type of model . It could even lead to greater efficiencies through increased flexibility . Implementation would require additional state funds only if California chose initially to hold SELPAs harmless (i.e. , prevent them from losing funds under the new formula) and then phase out the hold -harmless provision over several years. Clarify the State’s Objective for Special Education Funds California can have more constructive special education debates at state and local levels by focusing on total spending to educate children with disabilities rather than on special education expenditures alone. This broader frame of reference provides a more complete picture of how schools use both the regular education and special education environments to meet the needs of children with disabilities. By examining total expenditures on disabled children in rela tion to total expenditures on nondisabled children, school districts can more meaningfully account for what is spent and why. The state can take an important step by aligning the intent of state special education funds with IDEA. Th is would provide a clearer rationale for funding because both state and federal funds share the same purpose—helping to defray the additional spending to educate children with disabilities . A secondary benefit would be to help clarify the actual meaning of the term encroachment ( the local share of additional spending on disabled children ). Doing so would underscore how the debate should focus on whether the size of the local share is fair for providing a free a nd appropriate public education, not on whether the local share should exist at all. In addition, better data would improve special education discussions and ensure greater program accountability. Currently, state data systems are not equipped to calculate school spending by a child’s disability status. The missing ingredient is the capacity to 40 The low/low cell includes San Juan, Lassen, Tuolumne, San Luis Obispo, Amador, Calaveras, El Dorado, Nevada, Clovis, Sierra, and Placer. The high/high cell includes Garden Grove, West Contra Costa, San Fr ancisco, Santa Ana, Anaheim, and Oakland. 30 account for non -special education spending on children with disabilities. This report is able to provide an estimate, but calculations that are more exact would be possible if the current data collection could track spending by disability status. 41 California’s current funding formula has several desirable properties, but more can be done. By building upon the state’s existing accomplishments in reforming special education finance, it can develop a system to serve as a role model for education finance reforms in the future. 41 Appendix B describes the methodology used in the report. 31 References AB 602 and Bill Analysis, Senate Rules Committee, Office of Senate Floor Analyses, September 9, 1997. Asimov, Nanette, “Extra -Special Education at Public Expense,” San Francisco Chro nicle, February 19, 2006. Bersin, Alan, Michael W. Kirst, and Goodwin Liu, “Getting Beyond the Facts: Reforming California School Finance,“ Chief Justice Earl Warren Institute on Race, Ethnicity, and Diversity, University of California, Berkeley, Californ ia, 2007. California School Accounting Manual, California Department of Education, Sacramento, California, 2008. Center for Special Education Finance, “Comparison of Special Education Expenditures across SEEP States and the Nation,” American Institutes f or Research, 2003. Cullen, Julie Berry, “The Impact of Fiscal Incentives on Student Disability Rates,” Journal of Public Economics , Vol. 87, 2003, pp. 1557 –1589. Dhuey, Elizabeth , and Stephen Lipscomb, “The Effects of Fiscal Incentives in Special Education: Evidence from Capitation Finance Re forms,” PPIC Working Paper, 2009. Federal Register, “34 CFR Parts 300 and 301: Assistance to States for the Education of Children With Disabil ities and Preschool Grants for Children With Disabilities; Final Rule,” Federal Register, Vol. 71, No. 156, 2006. Fullerton School District, “2007/08 First Interim Financial Report,” 2007, available at www.fsd.k12.ca.us/menus/bussvcs/fiscal/InterimReports/0708FirstInterim Narrative.pdf . Greene, Jay P. , and Greg Forster, “Effects of Funding Incentives on Special Education Enrollme nt,” Civic Report , Vol. 32, 2002, pp. 1 –13. Harr, Jenifer J., Tom Parrish, and Jay Chambers, “Special Education,” in Handbook of Research in Education Finance and Policy , Eds. Helen F. Ladd and Edward B. Fiske, Routledge, New York, New York, 2008, pp. 575 –590. Harr, Jenifer J., Tom Parrish, Jay Chambers, Jesse Levin, and Maria Segarra, “Considering Special Education Adequacy in California,” American Institutes for Research, 2006. Kwak, Sally, “The Impact of Intergovernmental Incentives on Special Education Spending,” Working Paper, 2008. Legislative Analyst’s Office, “LAO Recommended Legislation,” www.lao.ca.gov, 2008. Lipscomb, Stephen, “Resolving Special Education Disputes in California,” Public Policy Institute of California, San Francisco, CA, 2009. Lipscomb, Stephen, “Students with Disabilities and California’s Special Education Program,” Public Policy Institute of California, San Francisco, CA, 2009. 32 Mahitivanichcha, Kanya , and Thomas Parrish, “Do Non-Census Fund ing Systems Encourage Special Education Identification? Reconsidering Greene and Forster ,” Journal of Special Education Leadership , Vol. 18, No. 1, 2005, pp. 38 –46. McLaughlin, Margaret J. , and Maria F. Owings, “Relationships Among States’ Fiscal and Demo graphic Data and the Implementation of PL 94 –142,” Exceptional Children, Vol. 59, 1993, pp. 247 –261. “New Funding Model for Special Education: Final Report,” California Legislative Analyst’s Office, Department of Education, and Department of Finance, 1995. Newhouse, Joseph P., “Reimbursing Health Plans and Health Providers: Efficiency in Production versus Selection,” Journal of Economic Literature , Vol. 34, pp. 1236–1263. Parrish, Thomas, Jenifer Harr, Jennifer Anthony, Amy Merickel, and Phil Esra, “State S pecial Education Finance Systems, 1999 –2000: Part I,” American Institutes for Research, Center for Special Education Finance, 2003. Parrish, Tom, Jenifer Harr, Yael Kidron, Leslie Brock, and Priyanka Anand, “Study of the Incidence Adjustment in the Special Education Funding Model,” American Institutes for Research, 2004. Parrish, Thomas B., Daniel Kaleba, Michael Gerber, and Margaret McLaughlin, “ Special Education: Study of Incidence of Disabilities Final Report,” American Institutes for Research, Center fo r Special Education Finance, 1998. Rose, Heather, and Ria Sengupta, “Teacher Compensation and Local Labor Market Conditions in California: Implications for School Funding,” Public Policy Institute of California, Occasional Paper, 2007. Sonstelie, Jon, “Ali gning School Finance With Academic Standards: A Weighted -Student Formula Based on a Survey of Practitioners,” Public Policy Institute of California, 2007. About the Author Stephen Lipscomb is a research fellow at the Public Policy Institute of California, where he studies education policy issues. His current work focuses on special education, school accountability, and patterns of early grade retention. He holds a Ph.D. in economics from the University of California, Santa Barbara. PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Walter B. Hewlett, Chair Director Center for Computer Assisted Research in the Humanities Mark Baldassare President and Chief Executive Officer Public Policy Institute of California Ruben Barrales President and Chief Executive Officer San Diego Regional Chamber of Commerce John E. Bryson Retired Chairman and CEO Edison International Gary K. Hart Former State Senator and Secretary of Education State of Californ ia Donna Lucas Chief Executive Officer Lucas Public Affairs Ki Suh Park Design and Managing Partner Gruen Associates Constance L. Rice Co -Director The Advancement Project Thomas C. Sutton Retired Chairman and Chief Executive Officer Pacific Life Insurance Company Raymond L. Watson Vice Chairman of the Board Emeritus The Irvine Company Carol Whiteside President Emeritus Great Valley Center PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 p hone: 415. 291.4400 f ax: 415. 291.4401 PPIC SACRAMENTO CENTER Senator Office Building 1121 L Street, Suite 801 Sacramento, California 95814 p hone: 916.440.1120 f ax: 916.440.1121 www.ppic.org" ["post_date_gmt"]=> string(19) "2017-05-20 09:39:20" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_809slr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:39:20" ["post_modified_gmt"]=> string(19) "2017-05-20 09:39:20" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_809SLR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }