Donate
Independent, objective, nonpartisan research

R 799PLR

Authors

R 799PLR

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_799PLR.pdf" ["wpmf_size"]=> string(7) "1027884" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(257769) "California Cities and the Local Sales Tax ••• Paul G. Lewis Elisa Barbour 1999 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data Lewis, Paul George, 1966– California cities and the local sales tax / Paul G. Lewis, Elisa Barbour. p. cm. Includes bibliographical references. ISBN: 1-58213-010-8 (alk. paper) 1. Sales tax—California. 2. California—Economic policy. I. Barbour, Elisa, 1961— II. Title. HJ5715.U62C25 1999 336.2'713'09794—dc21 99-23683 CIP Copyright © 1999 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 the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword As visitors are quick to learn, California’s sales tax rate is higher than average. Very few residents, let alone visitors, are aware that the rate varies throughout the state, depending on local preferences. Nor are they aware that 1 percent of the overall rate of 7.25 to 8.5 percent is allocated to local governments on a point-of-sale basis. This 1 percent rate is a “diamond in the rough” for cities and counties. It is particularly a “diamond” for cities because it is a major source of their discretionary revenues. Moreover, sales tax revenues have the potential to grow as a city expands and its commercial land uses generate more sales. The “rough” is the restricted arena for raising local revenues after the passage of Proposition 13 in 1978. The authors of this report, Paul Lewis and Elisa Barbour, take a thoughtful look at the consequences of a sales tax regime that is so important to local governments. They address two key questions: What are the effects of the point-of-sale or situs-based sales tax on land-use iii decisions? and How do California cities vary in the benefits they receive from the tax? In conducting this study, the authors first surveyed top administrative officials in 330 cities throughout the state. They then integrated their findings with a thorough analysis of recorded revenue distributions over a 24 year period from 1971 to 1995. Their survey supports the argument, advanced by Dean Misczynski in the mid-1980s, that there has been a “fiscalization" of land-use decisions: Retail land uses are preferred over residential and heavy industrial uses by a substantial margin. City policymakers pursue land uses that generate high levels of sales tax revenues—shopping centers and auto malls. At the same time, the authors demonstrate that despite all the competition for revenue-generating retail businesses, the hierarchy among cities in their sales tax success has changed little over the study period. In other words, the competition is intense and the incentives offered to developers can be substantial, but the pattern of winners and losers remains largely the same. They argue that there is a relatively fixed amount of retail activity that can be supported by the market in a region at any given level of population and that, in the broadest context, business decisions about where to locate are not greatly affected by local incentives. The downside of each city’s giving priority to retail uses, however, is that residential and industrial uses may become more difficult and expensive to site. The authors conclude the study with an examination of various options for reducing distortions in land-use preferences stemming from sales tax policy. They suggest that one promising approach might be to reallocate a greater share of local property taxes to the cities, perhaps in exchange for returning some of the sales tax revenues to the state. Whatever policy correction is chosen, it is clear that cities compete iv actively for retail sales, that the consequences only marginally affect overall ranking in revenues, and that there are real costs to city budgets, even though the returns are negligible. This volume is one of a series that the Public Policy Institute of California is publishing on the status of public finance in California. The policy issues involved in understanding public finance issues today, including the question of equity, the fiscal relationship between state and local governments, and the level and quality of government services are large and important concerns that need to be analyzed with an objective and independent eye. They are exactly the kinds of issues that PPIC was founded to study. We trust that this growing body of research and findings on local government finance will reduce the level of disagreement and set the stage for a more informed reformulation of public policy. David W. Lyon President and CEO Public Policy Institute of California v Summary Although it is somewhat hidden from public view, controversy surrounds the local sales tax in California. A 1 percent local sales tax is collected by the state as part of the larger sales and use tax levied on most items for sale. It generates revenues that are returned to the local government in which the sale occurred (see Table S.1). This return of funds to the point-of-sale jurisdiction is known as the situs rule. The situs rule has tremendous implications, since it gives cities a special incentive to promote the location of retail businesses within their boundaries—an incentive that does not exist for residential or industrial development. For California’s cities, the significance of the local sales tax is much greater than one might suspect, given its relatively small share of their overall revenues. In recent years, the situs-based sales tax has constituted about one-tenth of total city revenues. Nevertheless, the sales tax—along with the property tax—is one of the major sources of discretionary revenues for cities. Revenue from most other major sources, by contrast, vii Table S.1 Components of California’s Overall Sales Tax Rate Rate, % 6.00 5.00 0.50 0.50 1.25 1.00 0.25 Up to 1.25 7.25 to 8.50 Purpose State sales tax, consisting of: State general fund Local Revenue Fund—distributed to counties for health and welfare responsibilities Public Safety Fund—distributed to counties, some cities Bradley-Burns sales tax, consisting of: Local sales tax—directed to general fund of jurisdiction where sale occurred Local transportation tax—directed to county where sale occurred Local special taxes, generally for transportation— optional, require voter approval, used in 24 counties and a few cities (note: in most counties, the maximum rate authorized is 1.50 percent). Total rate SOURCES: Adapted from State Board of Equalization (1998), p. 25; California State Controller (1995–96), p. ix; and the California Revenue and Taxation Code, Section 7251. is earmarked or restricted for certain functions. Moreover, city policymakers perceive that the sales tax has the potential for substantial revenue growth in good economic times—if supportive land-use decisions are made. In the post-Proposition 13 world, these factors— budget flexibility and growth potential—make it a highly prized revenue source for the state’s municipalities. This report focuses on two of the major issues arising from cities’ heavy interest in sales tax revenues. First, what are the effects of the situs-based sales tax on cities’ land-use decisionmaking? Are localities systematically favoring retail development over other types of growth? If they are, this is potentially a negative scenario for economic development in the state as a whole. Second, how do California cities vary in the viii benefits they receive from the local sales tax? Which types of communities are doing better or worse in their quest for sales tax revenues? The Sales Tax and Local Development Decisions Many critics, pointing to local government efforts to recruit retail businesses, have worried about public-sector “giveaways” to retailers or developers. Critics also complain that the “fiscalization of land use”— development decisions favoring tax-generating activities—has retarded housing and other non-retail development. But the evidence of fiscalized land-use decisionmaking has thus far been largely anecdotal. To shed light on this question, PPIC conducted a mail survey of local officials on the topic of “Development Strategies in California Cities” in late 1998. The survey was sent to the city manager, or other top administrative official, in each of the 471 cities then in existence in the state. The response rate was 70 percent, and the sample was quite representative. The survey results provide strong evidence that city governments do systematically favor retail development over other land uses when it comes to new development on vacant land, as well as for redevelopment in designated “blighted” areas. Figure S.1 shows how cities ranked retail and six other land-use categories on a seven-point scale of desirability. (The responses were screened to include only those cities with vacant land available for development, and only those saying they engage in redevelopment.) Retail is the most favored category, although office, mixed-use, and light industrial development are also considered quite desirable. Respondents also indicated that retail projects were the most ix Average desirability score 7 6 New development Redevelopment 5 4 3 2 1 Retail Office Mixed- Light Single- Multi- Heavy use industrial family family industrial development residential residential SOURCE: PPIC City Manager Survey, 1998. Figure S.1—Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers likely to receive a general plan change or financial incentive from their cities. Why the favor toward retail? Our survey also asked respondents to rank the importance of 18 motivations for evaluating development and redevelopment projects, including such factors as job creation, property tax generation, environmental effects, and effects on neighborhoods. Of all these factors, “new sales tax revenues generated” was tied with “city council support for the project” as the top consideration for redevelopment. In the case of new development, sales taxes actually exceeded city council support in importance. Respondents whose cities plan to annex land in the next five years also indicated that gaining new sales tax revenues was the second most important consideration of 12 x factors in deciding which lands to annex, after the desire to control development patterns in surrounding areas. There are regional variations in the degree of importance accorded to sales tax considerations and the favor shown to retail. A set of “pragmatic” factors regarding land use—including sales tax and other fiscal concerns—are considered most important in Central Valley cities and in other mostly rural portions of the state but less important in the San Francisco Bay area. The 36 cities designated as “central cities” by the U.S. Census Bureau tend to have more balanced land-use perspectives than other communities, with retail and sales tax motivations ranked lower relative to other concerns. In general, however, few characteristics can be said to distinguish cities that are more actively pursuing retail development. Instead, the practice is so common that it can be called nearly ubiquitous. Disparities in City Revenues Under the Local Sales Tax City success in the sales tax game varies a great deal. In the 1995–96 fiscal year, the amount of local sales tax revenues per resident received by California cities ranged from $2.25 to $56,891.84. In short, “winners” and “losers” emerge in the local sales tax revenue game. Critics complain that these inequities in revenues do not reflect communities’ social needs for government services, nor do they correspond directly to the amount of effort cities have devoted to economic development. As retail patterns have shifted—given the upswing in suburbanization, shopping malls, and new types of retail-market activity—certain local governments have benefited and others have lost ground. As Figure S.2 shows, the 36 central cities have seen lower levels xi Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Central cities All other cities 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.2—Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 of sales tax revenues per capita over the period since 1971 relative to other municipalities. Also in this period, sales tax revenues per capita in the San Francisco Bay area moved well ahead of those in the greater Los Angeles region. Multivariate statistical analysis indicates that a city’s success in receiving per capita sales tax revenues (as of fiscal years 1990–91 through 1992–93) can be partially predicted based on certain city characteristics. Not surprisingly, the findings reveal that sales tax success is associated with certain market characteristics—cities with higher populations, lower densities, and fewer persons per household. Cities devoting more of their land to redevelopment projects also performed better. But demographic characteristics of cities do not relate to sales tax success in a totally xii straightforward fashion: Cities with very low incomes and higher shares of blacks in their populations had less success, but so did cities with high population growth rates and very high incomes. This last finding emerges in part because many of the “least successful” sales-taxgenerating cities tend to be very wealthy residential suburbs that eschew commercial development. It indicates that although the situs rule does systematically favor some cities, it is not necessarily biased in favor of high-status communities. Sales tax disparities among cities have not been increasing over the past two decades. In fact, the cities that were the most successful in receiving per capita sales tax revenues as of the early 1970s tended to have significantly lower gains over time, relative to other cities. Over the entire two-decade period, most communities experienced a relatively stable position in the sales tax hierarchy, as Figure S.3 illustrates. Fighting for Slices of a Fixed-Size Pie Overall, from the mid-1970s through the mid-1990s, sales taxes, measured in real dollars per capita, were a fairly stagnant source of funds for California cities, allowing for business-cycle swings (see Figure S.4). This is due to the relatively fixed nature of retail spending per capita, the many exemptions that state law provides from sales taxation, and the increasing share of consumer spending going to transactions that are not subject to the sales tax (for example, personal services and catalog sales). Cities are clearly trying to attract retail development. Despite this, the hierarchy among cities in their sales tax success has not changed much. And since per capita sales tax collections are steady or declining overall, it is likely that cities are competing over a relatively fixed amount xiii Log of per capita sales tax revenues, 1995 8 7 6 5 4 3 2 1 0 012345678 Log of per capita sales tax revenues, 1971 (in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.3—Cities’ Sales Tax Revenues per Capita, 1971 and 1995 of per capita revenue. There is only a certain amount of retail activity that can be supported in a region at any given level of population. This makes the growth prospects of retail different from industries in which regions can experience indigenous growth and local gains can lead to broader economic benefits outside the host city. Cities that succeed in recruiting retail businesses within their borders, by contrast, can generally be viewed as simply shifting retail sales geographically within their market region. The issue then becomes whether cities’ efforts to attract retail could systematically affect land-use patterns in the state. First, the actions of cities will not induce “extra” retail development to occur, if the additional retailers cannot hope to make a profit. There is no evidence xiv Revenue (constant 1995 dollars) 140 130 120 110 100 90 80 70 60 1971 1976 1981 1986 1991 1996 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.4—Sales Tax Revenues per Capita for All California Cities that California has substantially more retail space per person than other states. Second, it is also questionable whether all of the favor cities shine on retail leads to major differences in retail location. Retailers generally locate in relation to their customer base, transportation accessibility, suppliers, and competitors. Specific jurisdictions may win and lose in competing for a given retailer, but the winner and loser locales are likely to be fairly similar places. Third, however, the competition for sales taxes may have a broader effect. If cities favor retail development substantially over housing and industrial development, then those land uses will likely be somewhat more difficult, uncertain, and expensive to develop. xv If all localities use similar policy efforts as inducements to retailers, then the advantages that individual cities might gain from these policies will likely be “competed away.” Presumably, the aggregate result would be a shift of resources from local governments to the retail sector (store owners, developers, landowners—and potentially even customers). Considerations for State Policymakers The effects of the sales tax on land-use decisions, along with the vast disparities in sales tax revenues among cities, often lead reformers to urge a change in the situs basis for sales tax distribution. The most significant recent reform was Proposition 11, passed by the voters in November 1998. This constitutional amendment allows any two or more local governments to negotiate sales tax revenue-sharing agreements, with a two-thirds vote of their governing boards. It will enable some experiments in cooperation but is likely to have only a modest overall effect, since local governments that are doing well at attracting retailers will have little incentive to “disarm” in the sales tax race. An alternative to the situs rule often discussed would be to switch toward a more population-based system for distributing sales tax revenues among local governments. Using 1993–94 fiscal data, we calculated that 56 percent of cities would have improved their fiscal situation if sales tax revenues were redistributed statewide on a population basis; 51 percent of cities would have improved their position if the revenues were redistributed by population within each county. The vast majority of county governments would have received more funds under either scenario. A large majority of the California population lives in jurisdictions that would have been better off under each of these population-based approaches to distributing sales tax revenues (56.2 xvi percent and 59.5 percent, respectively). However, the intra-county redistribution of revenues would not advance social equity goals, in the sense of providing more funds to lower-income communities. Distribution of sales tax revenue by population has the advantages of better reflecting fiscal needs, potentially encouraging housing development, and minimizing local “chasing” of retail development. However, the population-based approach has potential disadvantages as well. These include a possible decline in entrepreneurial economic development policy among cities as well as difficult-to-justify windfall gains for wealthy residential enclave communities. The reform also could shortchange some cities with very large retail sectors, which often have greater need for public revenues. Because of the influx of shoppers and workers, their “daytime populations” are far higher than their resident populations, causing burdens for roads and infrastructure, public safety, and other local public services. A more important critique of a population-based distribution of the sales tax is that it fails to address the broader problem of providing fiscal incentives for balanced development. The way around this dilemma is to widen our focus beyond the sales tax—which is merely one component of local revenues. As it stands, retail is looked upon with favor not merely because it is associated with the sales tax but because other forms of development are perceived to simply not pay their own way. Cities often are particularly unenthusiastic about housing because they receive a relatively small slice of the property tax dollar and a fairly large service burden. The solution to this problem would involve adjusting local finances to provide a greater incentive to take on such development— rather than merely adjusting the sales tax to provide less incentive to develop retail. Reallocating a substantially greater share of local property xvii taxes to local governments—perhaps in exchange for a withdrawal of some state aid and a portion of the local sales tax—is one promising approach to this complex problem. xviii Contents Foreword..................................... iii Summary..................................... vii Figures ...................................... xxiii Tables ....................................... xxv Acknowledgments ............................... xxvii 1. INTRODUCTION ........................... Historical Overview of Local Sales Taxes in the United States ................................. Local Sales Taxes in California .................... Contemporary Importance of the Sales Tax to Cities ...... Policy Issue I: Disparities Among Communities in Revenues ............................... Policy Issue II: Accusations of Fiscalized Land-Use Decisions ............................... Recent State Legislative Proposals to Address Sales Tax Concerns ............................... Organization of This Report ...................... 2. TRENDS IN CITY SALES TAX REVENUES ......... Sales Taxes as a Component of Cities’ Revenue Streams .... Revenue Gainers and Losers ...................... Changes in the Sales Tax Hierarchy over Time .......... Regional Trends ............................. 1 3 4 8 10 11 13 14 17 18 22 25 28 xix Trends in Central Cities ........................ The Experience of Newly Incorporated Cities ........... Summary .................................. 3. WHICH TYPES OF CITIES HAVE BENEFITED FROM THE SITUS RULE?........................... Variations Across Cities: The Extremes .............. Variations Across Cities: A Cross-Sectional Model ....... Measuring Sales Tax Revenues: The Dependent Variable ... City Characteristics: Independent Variables ............ Results of Model Estimation ...................... Factors Related to Sales Tax Success: Interpreting the Results ................................ Cities’ Relative Gains over Time: A Longitudinal Model ... Summary .................................. 4. THE SALES TAX AND LOCAL LAND-USE DECISIONS ............................... What Is Meant by the “Fiscalization” of Land Use? ....... Thinking Carefully About the Fiscalization Argument ..... Favoring Retail? ............................ Disfavoring Other Types of Growth? ............... Changing the Landscape? ...................... Implications and Policy Considerations ............... 5. THE PREFERENCE FOR RETAIL DEVELOPMENT ... Survey Methodology ........................... Development Activity in California Cities ............. Introduction to Survey Findings ................... Land-Use Preferences .......................... Factors Influencing City Development Decisions ........ Regional Variation in Attitudes About Development ...... Factors Influencing Annexation Decisions ............. Do City Governments Act on Their Development Preferences? ............................. Which Types of Cities Favor Retail Development the Most? ................................. Interest in Retail/Sales Tax: Dependent Variables ...... City Characteristics: Independent Variables .......... 33 36 39 41 41 45 45 46 51 52 58 63 67 68 71 72 75 75 78 81 82 83 86 87 88 91 95 96 97 98 98 xx Results of Model Estimation .................... 98 Factors Related to Fiscalization: Interpreting the Results .. 104 Characterizing Regional Patterns in Development Priorities .. 105 Summary .................................. 106 6. POLICY CONSIDERATIONS ................... 109 Summarizing the Results ........................ 109 Policy Debates over Sales Tax Revenue Sharing .......... 111 Arguments For and Against Moving Toward a Population- Based Distribution of Sales Tax Revenues .......... 113 Which Local Governments Would Benefit from a Population-Based Distribution? ................ 117 Statewide Redistribution ....................... 118 Within-County Redistribution ................... 120 Broader Remedies: Balancing Fiscal Rewards for Growth in California Communities ..................... 122 Conclusion................................. 125 Appendix The Mail Survey of City Development Strategies ......... 127 Bibliography .................................. 139 About the Authors ............................... 145 xxi Figures S.1. Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers ............................... S.2. Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ............ S.3. Cities’ Sales Tax Revenues per Capita, 1971 and 1995 .. S.4. Sales Tax Revenues per Capita for All California Cities .................................. 2.1. Revenue to California Cities, by Source ............ 2.2. Sales Tax Revenues per Capita for All California Cities .................................. 2.3. Dispersion of Adjusted per Capita Sales Tax Revenues Among California Cities, 1971–1995 ............. 2.4. Cities’ Sales Tax Revenues per Capita, 1971 and 1995 .. 2.5. Relationship Between Cities’ Sales Tax “Success” in the Early 1970s and Subsequent Gains .............. 2.6. Relationship Between Changes in Cities’ Sales Tax “Success” over the 1970s and 1980s .............. 2.7. Sales Tax Revenues per Capita in Major Metropolitan Areas .................................. x xii xiv xv 18 20 25 26 27 29 30 xxiii 2.8. Per Capita Sales Tax Revenues in the Los Angeles Area, 1995 .................................. 31 2.9. Per Capita Sales Tax Revenues in the San Francisco Bay Area, 1995 .............................. 32 2.10. Changes in per Capita Sales Tax Revenues in the Los Angeles Area ............................. 33 2.11. Changes in per Capita Sales Tax Revenues in the San Francisco Bay Area ......................... 34 2.12. Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ............ 35 2.13. Sales Tax Revenues per Capita: Cities Incorporated after Proposition 13 Compared to Other Cities .......... 37 3.1. Relationship Between Average Household Size and City Sales Tax “Success” ......................... 54 3.2. Relationship Between Hispanic Share of Population and City Sales Tax “Success”...................... 56 3.3. Relationship Between Income Level and City Sales Tax “Success” ............................... 57 5.1. Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers ............................... 87 6.1. Relationship Between Retail Concentrations and Crime in California Cities ......................... 115 xxiv Tables S.1. Components of California’s Overall Sales Tax Rate .... viii 1.1. Components of California’s Overall Sales Tax Rate .... 7 1.2. Cities with Highest and Lowest Sales Tax Revenues per Capita, 1996 ............................. 10 2.1. Changes in Real per Capita Sales Tax Revenues Between 1970s and 1990s Among California Cities .......... 22 3.1. Summary Statistics for Cities with Extreme Levels of per Capita Sales Tax Revenues as of 1995 ............. 43 3.2. Regression Model of City Sales Tax “Success” as of the Early 1990s .............................. 53 3.3. Regression Model of Changes in City Sales Tax “Success” Between Early 1970s and Early 1990s ...... 61 5.1. Survey Respondents, by City Population Size ........ 83 5.2. Survey Respondents, by Region ................. 84 5.3. New Development and Redevelopment Activity in California Cities ........................... 84 5.4. Development Activity, by Region ................ 85 5.5. Factors Influencing Development and Redevelopment Decisions ............................... 89 xxv 5.6. Importance Scores for Development Motivations, by Region ................................. 92 5.7. Factors Influencing Annexation Decisions, by Region ... 95 5.8. Dependent Variables in the Regression Model ........ 99 5.9. Independent Variables in the Regression Model ....... 100 5.10. Results of Regression Models: City Characteristics Associated with Preference for Retail Development..... 102 5.11. Results of Regression Models: City Characteristics Associated with Two Patterns of Development Priorities................................ 107 6.1. Number of Jurisdictions That Would Be Better or Worse Off Under a Statewide, Population-Based Distribution of the Local Sales Tax, 1994 ................... 118 6.2. Characteristics of Cities Gaining and Losing Revenues Under a Statewide, Population-Based Distribution of the Local Sales Tax............................ 119 6.3. Number of Jurisdictions That Would Be Better or Worse Off Under an Intracounty, Population-Based Distribution of the Local Sales Tax, 1994 ........... 120 6.4. Characteristics of Cities Gaining and Losing Revenues Under an Intracounty, Population-Based Distribution of the Local Sales Tax ......................... 121 A.1. Survey Respondents, by Title .................. 128 xxvi Acknowledgments This project proved much more complicated than we initially imagined. Through the help of numerous people, we were able to get a better grasp of the facts and issues, complete the research in a more timely fashion, and improve our writing and presentation. Seven people read the draft report in its entirety and provided many helpful comments: Michael Coleman, Michael Dardia, Peter Detwiler, Max Neiman, Joyce Peterson, Peter Schaafsma, and Michael Teitz. The reviews were clearly done with great care. Although we were not able to follow through on all of the suggestions these able reviewers made, their help resulted in significant improvements to the report. At an early stage of the research, a group of city managers, finance directors, and PPIC staff met with us to discuss the project and offered very constructive comments and suggestions. City officials present included Kenneth Blackman, Pamela Gibson, Rod Gould, Brad Grant, Derek Hanway, and Kathleen Millison. We thank Fred Silva for his help in organizing the session and for his continuing advice. Also present at xxvii the meeting was Don Benninghoven, the executive director (now retired) of the League of California Cities. We thank Mr. Benninghoven and the League’s staff for their help in reviewing our mail survey of local officials before its release and encouraging recipients of the survey to respond. Don Blubaugh also made useful suggestions on question wording. The mail survey drew responses from 330 cities and we would like to thank each of the busy professionals who took time to respond. At PPIC, Sun Kim did a great deal of work in organizing the administration of the survey; we could not have made this step work without his help. Sun also provided mapping analysis that proved important to the research and found ways to overcome software limitations along the way. In addition to the reviewers listed above, Mark Baldassare, Kim Rueben, Michael Shires, and other colleagues provided useful research advice. Mark Butala, a summer intern at PPIC in 1997, got the ball rolling by collecting much of the sales tax data and performing some preliminary data analysis. Others at PPIC who provided key help at various stages included Gary Bjork, Meghan Cashman, Arabella Cureton, Shannon Goecke, Kathy Kayhour, Judi Leonor, and Heather McMann. Participants in a colloquium session at California State University at Sacramento’s Graduate Program in Public Policy and Administration provided helpful feedback. In particular, Laurie Watson helped save us from an error. Although all of these persons helped improve the report, any errors, omissions, or other shortcomings are the authors’ responsibility. xxviii 1. Introduction In recent years, observers of California local government have given increasing amounts of attention to the local sales tax. This interest stems from two major concerns relating to the sales tax as an outcome and also as a cause of important economic and geographic trends in the state. First, the differences in levels of sales tax revenue flowing to the various local governments are an outcome of retailers’ location decisions in California. This is because of California’s so-called situs rule, which returns local sales tax revenue—although collected by the state—to the jurisdiction in which the sale occurred. As retail patterns in California and the nation have changed in recent decades—given the upswing in suburbanization, shopping malls, and new types of retailing such as power centers and “big box” stores—certain local governments have benefited and others have lost revenues. In some cases, neighboring cities receive vastly different flows of sales tax funds per capita. Some state and local officials and other observers have complained about these inequities among communities, arguing that such differences do not reflect 1 communities’ social needs or their economic development effort, and that the relative lack of sales tax receipts can complicate local budgeting. The second area of concern regarding the local sales tax is a natural outgrowth of the first issue. Since cities do not want to be “losers” in the quest for local revenues, they increasingly compete for and recruit retail land uses and provide incentives, according to many observers. This behavior may in turn cause land-use and development patterns to be different than what might exist under a more “neutral” system of local finance. That is, local land-use decisions may be biased or distorted in favor of retail—and thus against residential or industrial projects that might otherwise occupy the same properties. Moreover, local governments may engage in a difficult game of enticing retailers to relocate from other places in their market area, thus redistributing the location of retail. These strategies are often referred to as the “fiscalization of land use” and are generally thought to have become more pronounced since the passage of Proposition 13 in 1978, which limited local governments’ access to property tax revenues. This report examines both of these issues—disparities across communities in sales tax revenues and the possibility of sales-tax-driven land-use decisionmaking. We analyze the state’s municipalities— generally all referred to as cities in California—since these governments are by far the biggest beneficiaries of the local sales tax.1 The issues are complex, and some questions, particularly those related to the fiscalization of land use, cannot be answered definitively. Nevertheless, ____________ 1In 1998, when most of our research was completed, there were 471 cities in California. The incorporation of two new cities, Oakley and Laguna Woods, occurred after our study period. Incorporation attempts are currently under way in several other communities. 2 we seek to provide some useful basic analysis of these issues, which have been subject to much controversy but little serious empirical study. Questions to be addressed include: Which types of cities are helped or hindered by the situs rule? How might a different rule for distributing local sales tax revenues affect various types of communities? Which types of cities have gained or lost ground in receiving such revenues in recent decades? Do city officials seek retail development more than other types of land uses? How important is the sales tax as a motivation for city land-use strategies, as compared to other factors? This introductory chapter provides background on the local sales tax and some of the controversies related to it. Historical Overview of Local Sales Taxes in the United States2 Today, Californians pay about 4 percent of their personal income as state and local sales taxes, with the state claiming three-quarters of that amount.3 But the sales tax has not always been part of the public finance system in the state. California, and many other states and localities, began using sales taxes during the Great Depression, when existing revenue sources proved drastically insufficient to meet the extreme needs of their populations. During the Depression, not only did the traditional property-based revenues that local governments relied upon drop as a result of free-falling property values and foreclosures, but also social welfare activities were funded overwhelmingly at the state and local level. ____________ 2This section draws upon Advisory Commission on Intergovernmental Relations (ACIR) (1989); Due and Mikesell (1994); and Vercoutere (1979). 3See the state-by-state comparison table in Governing magazine’s State and Local Sourcebook 1998, p. 34. 3 California began its state sales tax (at a 2.5 percent rate) in 1933 (Benson, 1997). By 1936, 25 states had sales taxes. It is believed that New York City, in 1934, was the first city to make use of a local sales tax; New Orleans followed in 1936. Once the substantial revenue possibilities inherent in local sales taxes became apparent, the innovation spread, and even after the Depression emergency, hundreds of other local governments turned to this source. Policymakers were attracted to the local sales tax because of its relative invisibility, compared to property tax bills, and its responsiveness to inflation. In the 1940s, California and Illinois joined the list of states employing municipal sales taxes. Another set of states turned toward local sales taxes in the 1960s, and the share of local revenues accounted for by sales taxes thereby approximately doubled in the two decades after 1966 (ACIR, 1989). Nationwide, the 1970s saw a trend toward substituting local sales tax revenues for local property tax revenues (Krmenec, 1991). In fiscal year 1991, 16 percent of total tax revenues4 received by U.S. municipalities came from sales taxes, making it the second leading source of tax revenue behind the property tax (Due and Mikesell 1994, p. 277). By 1994, 33 states and the District of Columbia had local sales taxes. Local Sales Taxes in California When California’s economy picked up after the Great Depression, the state lowered its state sales tax from 3 percent to 2.5 percent in 1943. Soon after this reduction, numerous cities around the state followed up ____________ 4That is, revenues excluding enterprise fees, service charges, intergovernmental grants, and so on. 4 by imposing a locally levied sales tax—a propitious time to do so, since consumers probably did not notice the imposition (Vercoutere, 1979, pp. 22–26). Postwar population growth, infrastructure needs, and the declining property values of the Depression had stressed city budgets, which were only beginning to recover. Beginning with San Bernardino in 1944, cities around the state began to levy their own sales taxes (typically at 0.5 percent), with 35 doing so by the end of 1946. “By FY 1954–55, 175 (56%) California cities had a sales tax ordinance. These taxing cities were the residences of more than 87% of the state’s population” (Vercoutere, 1979, p. 32). Cities instituting the tax at this time tended not only to be larger than average in population size but also were typically located near other cities that also levied sales taxes. Ultimately, a number of perceived problems with locally administered sales taxes led the state legislature to pass the Bradley-Burns Uniform Sales and Use Tax Law in 1955 (taking effect January 1, 1956). These difficulties included inefficiencies in local administration, complications for retailers in complying with varied local rates and ordinances, worries about locational distortions for businesses because of varying rates, and variations in the exemptions from sales taxes granted by localities for certain goods. The Bradley-Burns law created a uniform local sales tax rate of 1 percent throughout California among the cities and counties choosing to levy the tax. All have so chosen. The law also required that sales taxes be collected by the state and be distributed on a situs basis, meaning that the jurisdiction hosting the retail facility gets the entire local share of sales taxes for that establishment. Counties receive the situs revenues only for sales in unincorporated areas. The distribution is different for manufacturers selling to businesses and for mail-order sales and auto 5 leases, where sales and use tax revenues are distributed into a shared county pool. Today, many local government observers refer to the onecent local sales tax simply as the “Bradley-Burns tax.” Thus, the state collects a statewide sales tax and also collects the 1 percent local tax that is levied by local governments. Although the state collection and distribution superficially give this the appearance of a revenue-sharing scheme, it is important to keep in mind that BradleyBurns is actually a locally levied tax. Additional sales taxes subsequently enacted and earmarked for specific purposes have complicated the picture further. Table 1.1 illustrates the way sales tax revenues are distributed in California. The state’s overall sales tax rate varies between 7.25 and 8.5 percent, depending on where the sale takes place. Of this amount, only a small to medium share goes to local governments. One cent for each dollar in sales goes to the general fund of the city in which the sale occurred—or, in unincorporated areas, to the county. A one-quarter cent share is dedicated to transportation development funds (mostly targeted to mass transit) in each county, under the provisions of the Transportation Development Act of 1972.5 One-half cent is distributed to counties (and some cities) for public safety programs, under the provisions of Proposition 172, adopted in 1993.6 An additional half cent goes to ____________ 5Technically, the 0.25-cent transportation sales tax is also considered a part of the Bradley-Burns tax. However, since it is distributed to counties and dedicated to a single purpose, we do not consider it in our analysis. 6Counties are allowed to shift other funds away from public safety, so Proposition 172 money is more akin to a general revenue (Legislative Analyst’s Office, 1998, p. 12). Counties are required to direct a small share of Proposition 172 funds (currently about 6 percent) to cities as a small way of making up some of the state’s early-1990s property tax shift. Overall, however, Proposition 172 amounts are not subject to the situs rule within cities and are not dealt with in this report. 6 Table 1.1 Components of California’s Overall Sales Tax Rate Rate, % 6.00 5.00 0.50 0.50 1.25 1.00 0.25 Up to 1.25 7.25 to 8.50 Purpose State sales tax, consisting of: State general fund Local Revenue Fund—distributed to counties for health and welfare responsibilities Public Safety Fund—distributed to counties, some cities Bradley-Burns sales tax, consisting of: Local sales tax—directed to general fund of jurisdiction where sale occurred Local transportation tax—directed to county where sale occurred Local special taxes, generally for transportation— optional, require voter approval, used in 24 counties and a few cities (note: in most counties, the maximum rate authorized is 1.50 percent). Total rate SOURCES: Adapted from State Board of Equalization (1998), p. 25; California State Controller (1995–96), p. ix; and the California Revenue and Taxation Code, Section 7251. counties to help carry out their health and welfare functions. Finally, some local governments levy an additional sales tax of up to 1 percent for special purposes—mainly transportation—subject to the approval of their voters (1.25 percent in the unified City and County of San Francisco). In this study, we focus upon the uniform local sales tax rate of 1 percent. For this principal portion of the Bradley-Burns tax, revenues return to the jurisdiction of sale. In a legal sense, the city sales tax is considered a credit against the countywide 1 percent rate. If a city chose to forgo this credit, the funds collected would remain with the county government. Not surprisingly, all cities in the state take advantage of this option, although some cities receive slightly less than the full 1 percent 7 rate because of tax-sharing agreements with their counties. This 1 percent sales tax is particularly attractive to cities because the use of the resulting revenues is discretionary—not earmarked by the state for specific, required activities. Contemporary Importance of the Sales Tax to Cities The passage of the Proposition 13 property tax limitation in 1978 compelled local governments to take a long look at alternative sources of revenues beyond the property tax. Later, the state’s massive shift of property taxes from cities and counties to school districts in fiscal years 1992–93 and 1993–94 during the state’s own budget crisis only hastened this search. The loss of control over property tax revenues has been particularly difficult for cities because those revenues represented a major source of discretionary, general-fund revenues. Today, city officials lack discretion over the spending of much of their revenue base. The League of California Cities (n.d., p. 2) has estimated that, as of 1994–95, only 36 percent of city revenues were not restricted by law to specific purposes. Using a different methodology, Shires (1999, pp. 32–34) concluded that cities in 1995 had discretion over 44 percent of their revenues (excluding bond proceeds), down from 49 percent in 1978. In either case, the message is the same: Cities lack flexibility over how more than half of their revenues are spent. Cities’ three main options for augmenting revenues in the post-1978 years have been to impose new general taxes, service-related fees and charges (including those on developers), and to encourage sales-taxoriented retail development. Growth in general taxes such as business license taxes, franchise taxes, utility users’ taxes, and transient lodging taxes was indeed significantly greater than growth in property or sales tax 8 revenue in the period after Proposition 13 (Shires, 1999, p. 46). The growth in this revenue stream, however, may have been limited by the passage of Propositions 62 and 218, which require majority voter approval for new or increased general taxes. The second option, fees and charges, has also grown in relative importance. However, fees and charges typically are limited by law to funding the cost of providing the service in question or mitigating the costs of development—although with some creativity they can sometimes be used to subsidize a city’s general-fund operations. The limitations on these two revenue options have left the local sales tax as one of the few general-purpose revenue sources with much growth potential to remain under city officials’ control.7 “Control” here, however, is indirect, operating through a city’s land-use policies. Sales taxes in California amounted to 9.6 percent of city revenues in fiscal year 1995–96, although this figure differs substantially from one jurisdiction to another. The vast majority of counties received far less than the average city in per capita sales tax revenues. Recall that counties are allocated only the sales taxes generated by sales in unincorporated areas, which typically are less urbanized than cities. Therefore, counties receive only 1.4 percent of their overall revenues from the sales tax, although such revenues are in many cases important to counties for the provision of “municipal”-type services in the unincorporated areas— services that many observers argue are underfunded. Given that cities rather than counties are most heavily invested in the competition for sales tax revenues, we have chosen in this report to focus on the former. In doing so, we concentrate on two main policy issues: the disparities in ____________ 7Cities may also turn to redevelopment to maximize the amount of property tax increment flowing to the city, but this revenue flow is earmarked. 9 sales tax revenues across cities and the effects of local sales tax competition on land-use decisionmaking. Policy Issue I: Disparities Among Communities in Revenues Sales taxes are important sources of revenues for city governments. But some cities have done better than others at capturing taxable sales. Under the situs rule, there are benefits to jurisdictions that manage to capture large shares of the taxable sales in their market area. The variation among cities in sales tax revenues is enormous, ranging from $2.25 per capita in Rolling Hills to $56,891.84 per capita in Vernon in fiscal year 1995–96. Table 1.2 lists the top and bottom ten cities in this sales tax “hierarchy.” Cities that serve as retailing centers in their region frequently enjoy far higher per capita collections of local sales taxes than ordinary jurisdictions do. Historically, central cities, with their downtown Table 1.2 Cities with Highest and Lowest Sales Tax Revenues per Capita, 1996 Ten Cities with Highest Revenues Vernon Industry Sand City Colma Irwindale Santa Fe Springs Signal Hill Commerce Emeryville Brisbane Per Capita Sales Tax Revenues, $ 56,891.84 30,130.96 6,304.87 4,400.14 2,002.12 1,172.58 841.73 824.05 799.93 515.94 Ten Cities with Lowest Revenues Rolling Hills Bradbury Tehama Hidden Hills Monte Sereno Hillsborough Canyon Lake Atherton La Habra Heights Avenal Per Capita Sales Tax Revenues, $ 2.25 2.57 4.19 4.54 5.14 5.31 5.71 5.92 6.68 9.71 SOURCES: Calculated from sales tax revenue data in California State Controller (1995–96), and annual city population estimates of the California Department of Finance, Demographic Research Unit. 10 shopping districts, served as the commercial centerpieces of metropolitan areas. Today, however, it is frequently suburban cities with major shopping malls, such as Irvine or Pleasanton, that are the focus of the retail market. In addition, some relatively small cities that specialize in retail activities—or host manufacturers selling equipment to other businesses—collect vast sums of sales taxes per capita, such as Irwindale, Vernon, Colma, and Sand City.8 These retail-centered economies can, in turn, lead to further concentrations of retail in such places. In short, revenue “winners” and “losers” have emerged in the local sales tax game.9 A major task of this report, therefore, will be to examine the disparities among cities in per capita sales tax revenues. What factors influence these disparities? What types of cities are advantaged and disadvantaged by the situs rule? Policy Issue II: Accusations of Fiscalized Land-Use Decisions In the wake of voter initiatives such as Propositions 13 and 218, which put serious constraints on the local property tax, the sales tax has been one of the few major revenue sources subject to at least partial control of local officials. By influencing land-use decisions, cities and counties capture greater or lesser shares of regional retail trade—and thus, greater or lesser shares of taxable sales transactions. Local ____________ 8Miller (1981, Chap. 3) shows how some of these types of cities, such as Industry and Commerce, were founded in the 1950s and 1960s as property tax havens for business landowners. Out-of-town consumers paying sales taxes provided disproportionate support for the local treasury. 9As we will discuss in Chapter 6, however, retail concentrations typically increase the amount of public services that local government must offer, because of the added “daytime population” and activity associated with retailing. Although we cannot measure these additional costs in this report, it is important to keep them in mind when evaluating the distribution of sales tax revenues across cities. 11 governments may favor commercial over other types of development and may extend services and boundaries considerably to encompass actual or potential retail growth. In short, overall city development strategies and plans may be heavily affected by the search for land uses that are lucrative in the amounts of revenue they contribute. These considerations suggest an emphasis on the “supply side” of the local land market, as opposed to the demand-side responses of customers and retailers. Local governments can use their land-use powers of zoning and subdivision regulation, as well as their ability to build infrastructure, to reward some types of growth and penalize others. In this fashion, land use is determined politically as well as economically (Lewis, 1996). The presence of local sales taxes may encourage competition among local governments for retail developments, may lead to the costs of retail developments being externalized onto local governments, and may lead cities to favor retail over other types of development that would not generate sales taxes (LAO, 1994, p. 125). A former legislative representative of the California State Association of Counties has argued, “It is . . . clear that the situs allocation of sales tax greatly encourages cities to pursue commercial development over both industrial and residential development. This incentive is so strong that cities and counties actually attempt to ‘steal’ sales tax generators from other localities” (Senate Local Government Committee, 1989). Or as a Los Angeles Times reporter put it, cities are “pursuing sales tax dollars at a fever pitch” (Shuit, 1998). Most observers stress that this is not a good long-term economic strategy, as local governments focus their recruitment efforts on businesses like auto malls and big box retail stores that do little to expand the economy. Developers and retailers have been 12 known to play one jurisdiction off against another, seeking the most advantageous “deal” for their projects. This issue is the so-called fiscalization of land-use decisionmaking. It is the second policy focus of this report. Recent State Legislative Proposals to Address Sales Tax Concerns At least three recent proposals from the state legislature have attempted concrete reforms of the sales tax allocation system. Assemblymember Valerie Brown introduced AB 3505 in 1994, which would have altered the distribution of the local sales tax by basing it more on cities’ relative population, as opposed to a strictly situs-based system. Although the bill did not pass, it did generate a dialogue about the possible merits of sales tax sharing. AB 1835, a 1998 bill sponsored by Assemblymember Tom Torlakson, attempted a ban on the sales tax “wars.” The bill would have prohibited the use of public funds by local governments attempting to lure retail businesses from other localities within any given market area. This bill was in part a response to events in Martinez, in Contra Costa County, which reportedly received an ultimatum from a large local discount retailer that the store would move to a nearby city if Martinez did not pay $2 million in costs connected with the planned expansion of the store property. The retailer also reportedly sought $200,000 in rent per year from the city for the privilege of using its parking lot during hours when the store was closed (Newman, 1998, p. 2). This bill passed the Assembly, but died in the state Senate, where the League of California Cities as well as auto dealers lobbied against it (Shuit, 1998). 13 In 1999, Torlakson offered a similar proposal, AB 178, which is pending as of this writing. Proposition 11, which appeared on the November 1998 ballot, was an attempt to help cities stave off ruinous competition. A constitutional amendment authored by Assemblymember George Runner, the proposition allows any two or more cities or counties to negotiate sales tax revenue-sharing agreements, with a two-thirds vote of their governing boards.10 Previously, local governments were permitted to engage in such revenue-sharing only with an affirmative popular vote, a requirement seen as stymieing possible cooperative efforts. The proposition, which passed with 53 percent of the vote, will enable some experiments in cooperation among local governments, but many informed observers say that its effects are likely to be modest. Local governments that are “doing well” in generating sales tax revenues will have little incentive to engage in such agreements. Organization of This Report The remaining chapters of this report provide evidence that should be essential to any further discussions about reforming the local sales tax. Chapter 2 looks at general trends in local sales tax collections and the tax’s place in cities’ overall revenue picture. Chapter 3 presents statistical analyses of sales tax data that allow us to draw conclusions about the types of cities that are doing best and worst under the situs rule. The subsequent two chapters address the issue of fiscalized land-use decisions. Chapter 4 reviews the allegations that have been made against city ____________ 10There are useful analyses of Torlakson’s AB 1835 and Runner’s ACA 10 by the Senate Local Government Committee and other committees. These may be accessed at http://www.leginfo.ca.gov/. 14 development decisions regarding the sales tax and considers the conceptual issues involved. Chapter 5 discusses results from a survey of city officials regarding the degree to which they seek retail development as opposed to other types of land uses, and the degree to which sales tax motivations influence their views toward development, redevelopment, and annexation. Finally, Chapter 6 considers the policy implications of our findings and briefly highlights some alternative policy proposals that have been discussed by those who would reform the situs-based local sales tax. 15 2. Trends in City Sales Tax Revenues How important is the local sales tax to city budgets? Clearly, the answer is different for each city. The sales tax is an almost trivial revenue source to a few, a leading source for others, and a significant but not overwhelming source for most. In this chapter, we use data calculated from the California State Controller’s annual volume, Financial Transactions Concerning Cities of California, to help readers get a more informed sense of the relative importance of the sales tax over the period since 1970. This time period allows us to examine patterns before and after Proposition 13, generally considered a watershed in local public finance. We present data on overall sales tax collections, the disparity among cities, and the patterns in specific subtypes of cities. 17 Sales Taxes as a Component of Cities’ Revenue Streams Aggregating revenue data across all cities, we can get a better sense of the overall role that the local sales tax plays. Figure 2.1 charts the various categories of municipal revenues over the period 1970 to 1995 in percentage terms, looking at the various revenue sources as shares of total revenue in all California cities. Sales taxes are clearly not an overwhelming source of city funds in comparison to other types of revenue, but they are a significant slice of the revenue pie. And despite the increased attention to the sales tax since the Proposition 13 property tax limitation in 1978, the sales tax has been a remarkably consistent share of overall revenues. Public service enterprises Licenses, fees, and assessments Subventions Property taxes Sales taxes 40 30 Percentage 20 10 0 1970 1975 1980 1985 1990 SOURCE: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual). 1995 Figure 2.1—Revenue to California Cities, by Source (as a percentage of total revenue) 18 Note that after the passage of the Proposition 13 property tax limitation in 1978, sales taxes surpassed property taxes as a share of city revenues—a situation that persisted until 1990. In the early 1990s, the sales tax once again approached parity with property taxes. This was due in part to the state’s shift of property taxes away from cities to schools in this period and also in part to a deep recession. The recession held down increases in property taxes, as fewer new structures were built or changed hands and many property owners won downward reassessments. The importance of sales taxes is all the more clear when one considers the issue of spending discretion associated with each revenue source. Many other leading revenue sources have their proceeds restricted to certain types of uses. For example, city enterprise revenues—the rates and service charges paid by users of city-provided water, electricity, or other such utility-like functions—normally must flow back into the operation of the enterprise in question. Similarly, many state subventions and federal grants are earmarked for particular functions or programs. Thus, sales and property taxes (along with Vehicle License Fee revenues) are the major single sources of discretionary income that may be used for general purposes by cities. This makes them particularly attractive and sought after forms of revenue. Since the passage of Proposition 13, however, cities have been very limited in their ability to raise new revenues from the property tax. This is because property tax rates have a ceiling of 1 percent, properties are reassessed only when sold, and the division of the property tax dollar among local governments is set by a state formula—one that often provides a relatively small share to cities.1 Thus, the sales tax has taken ____________ 1In aggregate, about 14 percent of the property taxes paid by California residents goes to cities, with the rest going to schools, special districts, and county governments. 19 on increasing significance since 1978, despite its relatively flat trend in revenues per capita. Figure 2.2 takes a closer look at this trend for selected years since 1970–71. The bottom line on the graph, with sales taxes expressed in nominal dollars per capita (that is, unadjusted for inflation) shows almost continual growth in revenues until the early-1990s recession caused the retail sector to slow. In constant 1995 dollars, however (the top line), we can see that sales tax revenues have been a somewhat stagnant source of funds over this period, generally hovering between $100 and $120 per Revenue (dollars) 140 120 100 80 60 Constant (1995) dollars 40 Nominal dollars 20 1971 1976 1981 1986 1991 1996 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.2—Sales Tax Revenues per Capita for All California Cities 20 capita.2 Although these revenues grew throughout the 1970s, the early1980s recession saw declining sales per capita—and thus declining sales tax revenues. There was only a muted recovery by 1985, and another decline in the period since. These trends illustrate the sales tax’s sensitivity to broader economic conditions and trends. It is well known that “the retail sector is among the most cyclically sensitive segments of the economy” (Kroll and Marrinan, 1985a, p. 1). A factor that may be of even more long-term significance than the business cycle is the changing nature of consumer purchases themselves. Many analysts have pointed out that an increasing share of Americans’ disposable income in recent years has been spent on items not typically subject to local sales taxation—personal services, mailorder sales, and purchases via the Internet. In fact, the California Budget Project has found that taxable sales as a proportion of personal income in the state fell by more than a third between 1950 and 1995. In addition, the state legislature has carved out numerous exemptions to the sales tax for various products over the years, ranging from diaper services to cattle feed to magazine subscriptions (see Benson, 1997). These trends have further constrained California’s local governments’ ability to squeeze any growth from this desirable revenue source. Not surprisingly, the ratio of local sales tax revenues to personal income in cities has declined since the 1970s (Coleman, 1998, p. 24). Thus, the battle among cities for sales tax dollars is highly competitive. Not only are retail sales for a given population relatively fixed in amount, but the overall amount of sales per capita subject to the ____________ 2Throughout this report, dollar figures are adjusted for inflation by using the California Consumer Price Index, which is calculated by the California Department of Labor and Industrial Relations, Division of Labor Statistics Research. 21 sales tax levy appears to be in long-term decline. Thus, cities that “succeed” in recruiting retail businesses within their borders can generally be viewed as simply shifting retail sales geographically within a region. In short, the competition for retail appears to be a zero-sum game (Senate Local Government Committee, 1989). Revenue Gainers and Losers As Figure 2.2 shows, experiences with the local sales tax clearly were quite different in the 1970s from the 1980s, with overall per capita growth in revenues in the earlier decade and stagnation in the latter decade. How many cities gained and lost ground over this period? How much commonality was there in city experiences? Table 2.1 shows the number of cities increasing and decreasing their real per capita sales tax revenues in the period from the early 1970s to the early 1980s, in the decade that followed, and over the entire period. In this case, we have used three-year averages of sales tax revenues and population for each city: fiscal years 1970–71 through 1972–73, Table 2.1 Changes in Real per Capita Sales Tax Revenues Between 1970s and 1990s Among California Cities No. of cities gaining Median amount gained No. of cities losing Median amount lost 1971–73 to 1981–83 315 $19.04 88 $8.32 1981–83 to 1991–93 181 $14.59 245 $17.93 1971–73 to 1991–93 234 $30.83 169 $15.22 NOTE: Revenues measured in constant 1995 dollars. SOURCES: Calculated from California State Controller, Financial Transactions Concerning Cities of California (annual); and annual city population estimates from the California Department of Finance, Demographic Research Unit. 22 1980–81 through 1982–83, and 1990–91 through 1992–93. (For the sake of simplicity, henceforth in this report, we will refer to fiscal year 1970–71 simply as 1971, 1994–95 as 1995, and so on. California’s fiscal years actually run from July 1 to June 30.) We have chosen to take three-year averages because individual cities tend to have a fair amount of variation in sales tax revenues per capita from year to year; a three-year average is a more stable and reliable measure than a single-year measure and is less sensitive to business-cycle swings. Focusing on the time periods immediately after the decennial federal Census (1970, 1980, 1990) will allow us to use Census data to analyze city characteristics. As the table indicates, the vast majority of cities—315 of 403, or 78 percent—were better off in the early 1980s than a decade earlier, by this measure. Moreover, the average amount of increase in per capita sales tax revenues for the revenue gainers was far greater than the median amount lost by the revenue losers. In the following decade, however, most cities reversed course, with 245 of 426 (58 percent) losing ground—and the losing cities were losing more, on average, than the winners were gaining. Over the entire 20-year period, most cities—58 percent—gained some ground.3 It is possible that these numbers mask a growing disparity among cities. That is, it may be the case that as competition for retail businesses heated up, the sales tax “winners” became significantly better off than the sales tax “losers.”4 In attempting to measure this phenomenon, however, ____________ 3To be counted in the table, the city must have been in existence both at the beginning and at the end of the relevant period. 4Although we sometimes use the terms “winners” and “losers” in this report as a convenient shorthand for cities’ levels of per capita sales tax revenues, we do not mean to imply that the former group of cities has employed “better” policies than the latter. Indeed, one can readily conceive of circumstances under which cities make short-term- 23 a key issue is to define the winners and losers. Some traditional measures of variance—highest versus lowest, or standard deviations, for example— are potentially misleading in this context, because they are highly sensitive to outliers or extreme observations. To correct for this problem, we calculated a statistic called the interquartile range (IQR) among cities for each year measured. The IQR simply measures the distance between the 75th percentile of a distribution and the 25th percentile—in this case, the gap between the lowest city in the top quarter of the sales tax success list and the highest city in the bottom quarter. Figure 2.3 displays the IQR over the period 1971 to 1995. It shows a widening dispersion (or “spread”) among cities through the 1970s, with a flatter trend in the period since. Thus, cities were becoming more unequal in their success at attracting sales tax revenues until about the time of Proposition 13 (1978), with no lasting increase in city inequality in the period since.5 Note, too, that the amount of dispersion decreased during the recessions of the early 1980s and 1990s. One possible interpretation for the convergence during these periods is the following: When the economy is expanding, consumers are willing to buy more expensive and luxury goods (automobiles, jewelry, and so on). Sales tax revenues from such purchases may tend to accrue to certain cities specializing in those retail activities. During recessionary periods, ________________________________________________________ oriented, even ill-conceived, land-use decisions that might help them land in the “winner” group. 5As a check of these conclusions, we also examined the percentage of each city’s own-source, general-purpose revenues accounted for by the sales tax in 1973, 1983, and 1993. We then calculated the standard deviation in this share across all cities. The standard deviation was 15 percent in 1973, 13 percent in 1983, and 12 percent in 1993, again indicating that disparities among cities in sales tax reliance were level or slightly decreasing. 24 Interquartile range (in 1995 dollars) 100 90 80 70 60 50 1971 1974 1977 1980 1983 1986 1989 1992 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.3—Dispersion of Adjusted per Capita Sales Tax Revenues Among California Cities, 1971–1995 however, consumers defer making such purchases, and thus the fortunes of these “winner” cities decline relative to other places, making the overall distribution more equal. It is also conceivable that the overall flatness of the IQR measure since 1978 may indicate that the enhanced competition for retailers among cities since Proposition 13 (if the enhanced competition in fact exists) is driving worse-off cities to offer more incentives to retailers and developers. If so, these actions might shift retail sales, at the margin, to places that were falling behind. Changes in the Sales Tax Hierarchy over Time Scatterplot diagrams may also be used to illustrate cities’ performances over time. These have the advantage of showing each 25 Log of per capita sales tax revenues, 1995 individual city’s performance, rather than aggregating them together. In this series of scatterplots, we have used the natural logarithm of sales taxes per capita, rather than the actual amounts of dollars per capita, as a way to improve the graphic presentation of the data.6 Figure 2.4 plots each city’s position in 1971 on the horizontal axis and its 1995 position on the vertical axis. What emerges quite clearly from this picture is the relative stability of the sales tax hierarchy. That is, cities that enjoyed high receipts at the beginning of our study period 8 7 6 5 4 3 2 1 0 012345678 Log of per capita sales tax revenues, 1971 (in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.4—Cities’ Sales Tax Revenues per Capita, 1971 and 1995 ____________ 6Because some cities are extreme outliers, boasting thousands of dollars of sales tax revenues per capita, and most other cities are in the lower ranges, with dozens of dollars per capita, it would be difficult—and not very helpful—to show them all on the same graph. Taking the natural log of these values, a standard technique in the social sciences for transforming skewed data, has the effect of “pulling in” very high values to form a more normal distribution. 26 were still in an advantageous position in the early 1990s, and the lowerranking cities tended to stay in the low ranges. Using this technique to examine the data, then, it would appear that cities can generally expect to maintain their relative positions in retail markets, if past trends hold. This is not to say, however, that there are no factors systematically affecting the capacity of cities to improve or worsen their positions in attracting retail sales. Chapter 3 will illustrate the combination of factors that influence cities’ performance over time. Were cities that started the sales tax “race” in an advantageous position able to gain more ground over time relative to the other cities? No, as we have seen in the discussion of the IQR. Figure 2.5 is a Change in log of per capita sales tax revenues (log of 1991–93 average minus log of 1971–73 average, calculated in 1995 dollars) 3 2 1 0 –1 –2 –3 –4 01 2 3 4 5 6 7 Log of per capita sales tax revenues (average of 1971–73, calculated in 1995 dollars) 8 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.5—Relationship Between Cities’ Sales Tax “Success” in the Early 1970s and Subsequent Gains 27 scatterplot presenting the same basic idea in a different manner. On the horizontal axis, we plot the log of each city’s per capita sales tax revenues in the 1971–73 three-year average period. On the vertical axis, we plot each city’s “success” over time—that is, the log of its early-1990s revenues minus the log of its early-1970s revenues. (Note that some cities have negative values on this measure; they worsened over time in per capita sales tax collections, in real-dollar terms.) The data points do not show an increasing relationship between a high standing in the early 1970s period and subsequent gains. If anything, the scatterplot appears to show a negative relationship between early success and subsequent gains—although there is a great deal of variation. Moreover, cities’ performances through the 1970s do not appear to be related to their performances through the 1980s, as Figure 2.6 reveals. This scatterplot graphs 1970s gains or losses on the horizontal axis and 1980s gains or losses on the vertical axis, again using changes in natural logs for illustrative purposes. The data points appear as a random cloud, indicating that there is no discernible relationship between individual cities’ performances in the two decades. Regional Trends Sales tax patterns among different types or categories of California cities may also be compared. One such comparison is the distinction between municipalities in the two major metropolitan areas of the state. Figure 2.7 displays per capita sales tax collections for cities within the five-county region centered on Los Angeles and the nine-county San Francisco Bay area. Here we can see a clear divergence in the fortunes of these two regions, with the Bay area overtaking the Southland in the early 1980s, and a widening disparity between the two areas in the period 28 1980s changes (log of 1991–93 average per capita sales tax revenues minus log of 1981–83 average per capita sales tax revenues, calculated in 1995 dollars) 2 1.5 1 .5 0 –.5 –1 –1.5 –2 –1 –.5 0 .5 1 1.5 2 1970s changes (log of 1981–83 average per capita sales tax revenues minus log of 1971–73 average per capita sales tax revenues, calculated in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.6—Relationship Between Changes in Cities’ Sales Tax “Success” over the 1970s and 1980s since. The Los Angeles area experienced deeper declines in retail sales per capita during the two major recessions of this period, and the Bay area was also quicker to recover from these troughs in the business cycle. Overall, the improving fortunes of the Bay area relative to Los Angeles during this period can largely be attributed to the former region’s significantly increasing level of per capita income, relative to Los Angeles. The decline of the aerospace industry in Southern California in the 1990s also may have particularly depressed the sales of goods from that region. 29 Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Los Angeles region San Francisco Bay area 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.7—Sales Tax Revenues per Capita in Major Metropolitan Areas Although the graph shows the differing long-term trends in these two regions, it is important to emphasize that the experience of individual cities within each region varies a great deal. Figures 2.8 and 2.9 are maps of the central portions of the Los Angeles region and Bay area, showing each city’s 1995 level of per capita sales tax revenues. Although the maps illustrate that there are proportionally more Bay area cities with high levels of sales tax revenues, both regions exhibit much variation in the performance of individual cities. Another interesting impression one can gain from the maps—a not unexpected finding—is that virtually all of the “top performing” sales tax cities in these two regions are located along major highways, which are displayed in the maps. The accessibility advantages of the freeway can help make an area ripe for modern auto-oriented retail. However, it would be premature to 30 ,,,,,PacificOcean ,,,Lessthan$50 $50 to $99 $100 to $149 ,$150orgreater Figure 2.8—Per Capita Sales Tax Revenues in the Los Angeles Area, 1995 conclude from these maps that freeways “cause” higher levels of retail sales relative to other areas without highway access. After all, it may well be the case that highways were routed in such a way as to connect areas of heavy commerce, so that to some degree highways resulted from retail clusters, as opposed to stores clustering around highways. In Figures 2.10 and 2.11, we present complementary maps that show each city’s change in per capita sales tax revenues between the early 1970s (average of 1971, 1972, and 1973) and the period 20 years later. 31 ,,,PacificOcean San Francisco Bay ,,,Lessthan$50 $50 to $99 $100 to $149 , ,$150orgreater Figure 2.9—Per Capita Sales Tax Revenues in the San Francisco Bay Area, 1995 Again, the Bay area has more cities with gains over this period (darker shaded areas), particularly in the South Bay region, where wealth increased markedly as Silicon Valley industries developed. And once again, there are noticeable clusters of “gainer” cities along some of the major highways. 32 ,PacificOcean ,,,Lessthan–$15 –$15 to 0 $1 to $14 ,,$15to$44 $45 or greater Figure 2.10—Changes in per Capita Sales Tax Revenues in the Los Angeles Area (Average of 1971 Through 1973 Subtracted from Average of 1991 Through 1993) Trends in Central Cities Another comparison of interest involves the relative fortunes of older central cities in relation to other types of jurisdictions. As California and the nation have increasingly suburbanized—both in population 33 ,,,,,,PacificOcean San Francisco Bay ,,,,,Lessthan–$15 –$15 to 0 $1 to $14 $15 to $44 , ,$45orgreater Figure 2.11—Changes in per Capita Sales Tax Revenues in the San Francisco Bay Area (Average of 1971 Through 1973 Subtracted from Average of 1991 Through 1993) and economic functions —we might expect the traditional business centers to experience a relative decline in their retail strength. A difficulty involved in this comparison involves a definitional issue: Which cities qualify as “central” cities? To avoid making our own subjective judgments, we relied upon the designation of central cities by the U.S. Census Bureau. There are 36 such cities in California. Some 34 might object that the list is overly broad, since it includes cities such as Petaluma and Seaside along with more traditional older cities, but the Census Bureau does at least use consistent reasoning in identifying such places. The Bureau has identified one or more traditional central places—historic business and population centers of long standing— within each designated metropolitan statistical area of the state.7 Figure 2.12 compares per capita sales tax revenues for these 36 cities to the other cities of the state. Here again, we see the changing fortunes Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Central cities All other cities 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.12—Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ____________ 7The central cities are Anaheim, Bakersfield, Chico, Fairfield, Fresno, Lompoc, Long Beach, Los Angeles, Merced, Modesto, Monterey, Napa, Oakland, Oxnard, Petaluma, Porterville, Redding, Riverside, Sacramento, Salinas, San Bernardino, San Diego, San Francisco, San Jose, Santa Ana, Santa Barbara, Santa Cruz, Santa Maria, Santa Rosa, Seaside, Stockton, Tulare, Vallejo, Ventura, Visalia, and Yuba City. 35 of different categories of communities. The central cities began the period with slightly higher sales tax receipts per capita than other places but were overtaken in 1973. The disparity has grown wider in the period since. Central cities appear to have been less robust than other cities in their emergence from the two recessions. Overall, the movement of commerce outward from traditional downtowns toward suburban freeway locations is probably at the root of these trends.8 The Experience of Newly Incorporated Cities A number of observers have alleged that many new cities in California have incorporated as a way to capture sales tax revenues in their area, for local use—thus denying these revenues to the broader unincorporated county area. Although some examples of this cityformation-for-profit phenomenon undoubtedly exist (Miller, 1981), overall there has been no noticeable upsurge in incorporations over the past 30 years (Lewis, 1998). In Figure 2.13, we compare the per-capita sales tax revenues for 1981–1995 of cities incorporated since the passage of Proposition 13 in 1978 with those of cities existing before 1978. (The number of post1978 cities, which grows over time, is displayed on the appropriate line on the graph.) With property tax rates limited by that proposition—and disadvantageous property tax distribution formulas typical for new ____________ 8It is worth noting, however, that “central cities” cannot be equated with the fortunes of downtowns. Many California central cities are quite extensive in territory and include a large amount of suburban-style shopping strips and centers. The inclusion of cities like Anaheim or Fairfield that many observers would label suburbs—these cities have only a small traditional “downtown” presence, in comparison with their more modern commercial areas—makes the relative decline of the 36 central cities all the more noteworthy. 36 Revenue (in 1995 constant dollars) 140 120 100 80 15 60 11 40 9 cities 20 1981 1983 21 30 40 48 51 53 53 Post-Proposition 13 cities Other 417 cities 1985 1987 1989 1991 1993 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.13—Sales Tax Revenues per Capita: Cities Incorporated after Proposition 13 Compared to Other Cities cities—we might expect new cities in this period to pay especially close attention to sales tax revenues. But as the graph shows, the new cities are significantly worse off in sales taxes, on average, than the pre-1978 cities. Clearly, creating a new city is no guarantee of fiscal bliss—even though incorporation proponents pay close and careful attention to the new city’s boundaries, typically trying to include as much taxable retail as possible. Several factors militate against high sales tax levels in these new cities. The first fact to consider is the typical position of these cities within their regional context. Typically, inner areas, which would be more likely to have developed a longstanding retail base, are already incorporated, whereas newly incorporated cities tend to be at the outer fringe of 37 metropolitan areas. These fringe areas tend to be heavily residential, with commercial properties few in number and a limited population base in the areas surrounding the new city. Thus, we would expect them to have less retail sales per resident. Another way of expressing this phenomenon is that areas that could experience a windfall gain of tax revenues from incorporation presumably would have already incorporated some time ago; thus, the areas remaining unincorporated in recent years will tend to be more marginal retail locations. Other factors, more related to intergovernmental politics, may also help explain the relatively poor sales tax success of new cities. One is the strategic response of county governments (and to a lesser degree, existing cities) to the proposed formation of a new city. Representatives of counties and existing cities, acting through the Local Agency Formation Commission (LAFCO) in their county, have some veto powers over the formation of new local governments. Incorporations that would seriously weaken the county fiscally (and thus potentially hinder county service provision even in existing incorporated areas) may be vigorously opposed by county delegates on LAFCO boards. Thus, some incorporation proposals that would result in large shifts of sales tax revenues to new cities are likely to be quashed. Moreover, state law dictates that the new city must cement a mutually agreeable tax distribution agreement with its county before a LAFCO can approve an incorporation proposal. As counties have become considerably more fiscally stressed since Proposition 13 and the early-1990s property tax shift, these negotiations are often protracted. The negotiations have 38 resulted in significant concessions by the new cities, whereby a large share of sales taxes in the new city will be passed on to the county.9 Summary Sales taxes are a fairly stagnant source of revenues, overall, for California cities. Nevertheless, some cities have done much better than others. Central cities have tailed off in per capita sales tax revenues relative to other communities, for example, and Bay area cities have continued to gain relative to Los Angeles area cities. There is, however, no evidence of an increase in overall disparities among cities over time and nothing to support the idea that cities that were leading the pack in sales tax collections a decade or two ago have pulled still farther ahead since then. Nor can one make the case that cities newly incorporated after Proposition 13 have an advantageous position with respect to sales taxes. In the next chapter, we will look more systematically at the characteristics of cities that are associated with “success” in the competition for sales tax revenues. ____________ 9Similar agreements are required when a city wishes to annex land. The city and county must work out a tax-sharing arrangement. These negotiations have become increasingly contentious in recent years. 39 3. Which Types of Cities Have Benefited from the Situs Rule? We have no property tax, we have no lighting tax, we have no library tax, we have no utility users tax . . . all because of the sales tax and the economic base we have developed (City manager of Cerritos, quoted in Shuit, 1998). In Chapter 2, we looked at some general patterns relating to cities’ “success” in receiving sales tax revenues. This chapter looks more systematically at the relationship in a recent period between various city characteristics and levels of per capita sales tax collection. We also examine the factors related to cities’ changing fortunes over time, looking at the period from the early 1970s to the early 1990s. Variations Across Cities: The Extremes At the time our data were gathered, the most recent and complete local fiscal data available from the State Controller’s office were from Fiscal Year 1994–95. (Again, for convenience we will refer to this as 1995, and so on for other fiscal years.) In that year, sales tax revenues for 41 California cities ranged from $3.12 per capita in Hillsborough, a residential Bay area enclave, to $50,999.52 per capita in Vernon, a commercial city in Los Angeles county with just 80 residents in that year. The median or “typical” city received $85.52 in sales taxes per capita. Counties, meanwhile, which receive situs-based sales tax revenues only for sales in their unincorporated areas, fared worse than most cities. This finding reflects the relatively weak presence of retail in outlying areas. The median county raised $42.32 per unincorporated area resident, with a range from $19.99 in Modoc County to $174.53 in San Mateo County.1 This tremendous range of “success” in generating sales tax revenues is interesting in and of itself, but it also raises a basic question: What types of cities tended to be the most or least successful? One way to examine this question is to look at the “extremes” of the distribution—those places with very low or very high per capita sales tax revenues. Table 3.1 provides some summary statistics regarding demographic and other characteristics of these communities. From the table, we can conclude that, on average, the cities that have been most successful at getting sales tax revenues tend to be older (in terms of incorporation date) and are more likely to be urbanized than the cities with extremely low sales tax revenues per capita. The “top 20” have lower household sizes, a smaller percentage of children in the population, and perhaps surprisingly, a higher percentage of Hispanics. It also may be surprising that the ____________ 1The City and County of San Francisco is treated as a city for purposes of analysis in this report. All population data are from the official annual estimates made by the California Department of Finance, Demographic Research Unit. We do not rely on the annual population totals listed for each city in the California State Controller’s Financial Transactions volumes, as these proved to be inaccurate and not consistent with the Department of Finance population data. 42 Table 3.1 Summary Statistics for Cities with Extreme Levels of per Capita Sales Tax Revenues as of 1995 Population, 1995 Incorporation date Median year housing built Area (square miles), 1990 Per capita income, 1989 Aggregate income in city and 10- mile buffer around it Household size, 1990 % black, 1990 % Hispanic, 1990 % non-Hispanic white, 1990 % under age 18, 1990 % over age 64, 1990 % central cities % in urbanized areas % in Central Valley % in Bay area % in Los Angeles area % of own-source general revenues from property taxes, 1993 % of own-source general revenues from sales taxes, 1993 Average of Top 20 Cities Bottom 20 Cities 12,020 7,460 1929 1944 1961 1965 5.6 13.4 $17,569 $34,277 $24.2 billion 2.7 3.6 29.6 71.5 22.4 13.4 0 85 5 20 50 $12.0 billion 3.1 4.6 22.7 76.0 27.0 11.6 0 58 35 30 30 9.1 25.6 35.9 5.2 SOURCES: U.S. Census (1990); California State Controller, Financial Transactions Concerning Cities of California (1993, 1995). “bottom 20” cities have much higher income levels. Note, however, that the bottom 20 also tend to be located in market areas with much lower levels of aggregate income—that is, purchasing power. A partial explanation for these findings is that several of the “bottom 20” cities are wealthy bedroom communities. These include cities such as Hillsborough and Atherton in the Bay area, and Rolling Hills and Hidden Hills in the Los Angeles area, all of which had per capita incomes 43 of over $63,000 in the 1990 Census. (By comparison, the statewide per capita income was $16,409.) These are relatively low-density suburbs with very little commercial development. We might surmise that the wealthy residents of these areas have preferred to “zone out” retailing to preserve their community character. This is an important point, since it means that the situs rule does not necessarily work to the advantage of the state’s wealthiest communities. It should be pointed out, however, that the “bottom 20” also includes a number of very poor cities with depressed economies, many of which are in the Central Valley. Examples include Avenal, McFarland, Parlier, and Biggs, each of which had a per capita income of less than $8,600 in 1989—about half the statewide average. Many of the “top 20” cities also fall into certain types. At least seven are commercial or industrial enclaves with small populations. These include the top five cities—Vernon, Industry, Colma, Sand City, and Irwindale—none of which had a population above 1,220 in 1995. Others are the commercial centers of regions of tourism with high daytime populations, notably Carmel, Capitola, and Sonora. It is difficult to generalize about the remaining cities in the top 20, other than to say that most of them have fairly small populations. The only very wealthy city in this group is Beverly Hills. It is dangerous to generalize to all California cities from this set of outliers. Nevertheless, differences between the top and bottom 20 do begin to provide some clues about factors that might be important to the overall group. The next section examines the issue more systematically. 44 Variations Across Cities: A Cross-Sectional Model A more thorough approach is to estimate statistical models that can help account for the variations among cities in per capita sales tax revenues. We do not seek to “explain” sales tax revenue-raising success, per se; rather, we wish to discover community characteristics associated with such success. In devising such a model, it is useful to consider the factors that we might expect to influence retailer location decisions. There are two major ways to analyze retail location. One is from the viewpoint of the practitioner, looking at optimal locations for new stores within market areas. National retail chains tend to have their own store location appraisal divisions, and typically have sets of criteria that any location must meet. A second perspective on retail location is that of the academic urban geographer, concerned with the spatial hierarchies of retailing across the landscape (Berry et al., 1988; Davies, 1976; Guy, 1994; see also Schneider, 1986). We reviewed this literature and, where possible, located data that would help replicate these studies for California (see also Kroll and Marrinan (1985a, 1985b). Except where indicated, the source of our data is the U.S. Census. We describe each variable in the analysis below. Readers who are less interested in the estimation strategy may wish to skip ahead to the summary at the end of this chapter. Measuring Sales Tax Revenues: The Dependent Variable Given the skewness of the data, we use the natural log of per capita sales tax revenue in each city as the dependent variable. We have averaged 45 these data for 1991, 1992, and 1993 to improve reliability, since individual cities tend to vary upward and downward a fair amount from year to year. The early-1990s data also have the advantage of being closer in time to the independent variables, most of which are drawn from 1990 Census data.2 We eliminate from the analysis four cities that are extreme outliers—Vernon, Industry, Colma, and Sand City. Each has a tiny population and tremendously high per capita sales tax revenues. City Characteristics: Independent Variables A number of demographic variables are employed to help illuminate cities’ levels of per capita sales tax revenues. One is the log of the 1991 population. A larger population in a city increases its potential purchasing power but also increases the “denominator” for our measure of sales tax revenues per capita. A city may have many residents, but they may or may not shop locally. Thus, the expected effect of population is unclear. Population change between 1985 and 1991 (in percent) is also included, since we felt that high population growth levels may spur additional retail sales because of the need for housing-related products. Household characteristics of the population are potentially important determinants of retail sales per capita. We include as a variable the percentage of senior citizens in each city, since elderly persons typically consume less than younger adults do. Related is the issue of average household size. Larger household sizes imply cities with more children, who tend to consume less than adults. Moreover, some retail goods—refrigerators and furniture, for example—tend to be bought on a ____________ 2However, performing the same analysis on 1995 per capita sales tax data yields extremely similar results. 46 “per household” basis. The literature also suggests that areas with small household sizes may tend to have a greater share of disposable income and be more oriented to consumption goods. Thus, we expect household size to be negatively related to per capita sales tax revenues. (A variable representing the percentage of children in the population also was considered but not included in the model because of its very high collinearity with household size.) A city’s racial characteristics also may be important. We include percentage black and percentage Hispanic as of 1990. Schneider notes that disproportionately black suburbs, in addition to having lower incomes on average, “are also less prestigious than white suburbs, and firms located in black areas will have difficulty attracting white shoppers, who constitute the bulk of consumers in most suburban areas” (1986, p. 27). We have not found any literature suggesting a relationship between Hispanic population shares and retail sales, but we include this variable for its potential importance in the California context. We also include variables related to the potential purchasing power of the city in question. One is the city’s median household income (as of 1989). As we have discussed, cities with higher incomes have more potential purchasing power—but very high-income people often seek “exclusive” residential environments devoid of retailing and other businesses. This relationship between income and sales tax success thus appears nonlinear—that is, increasing incomes should generally add to sales tax success, but beyond some threshold of wealth, cities may have lower sales tax revenues. For this reason, we specify a quadratic form, using household income and household income squared as variables. Potentially just as important is the log of the aggregate income of the market area around each city—a variable specially constructed for this 47 study. Using a computerized mapping program that allowed us to overlay city boundaries on 1990 Census information, we drew a 10-mile wide “buffer zone” around each city, and calculated the aggregate income of the population within this line—that is, in the city and its surrounding buffer zone. This aggregate income measure, which is reasonably though not perfectly estimated, can be thought of as the potential purchasing power of residents who live within a given travel radius around each city.3 Simply put, we expect retailers to gravitate toward locations where there is sufficient buying power to keep their stores afloat. A final factor related to purchasing power is tourism. We include a county-level estimate of 1992 travel-related expenditures (again in log form because of skewed data among the state’s counties). These data are from the Division of Tourism (1998, p. 9). ____________ 3Data complications led us to substitute a slightly different measure of the aggregate income around each city. Our mapping program was unable to construct geographic “buffers” for a number of cities. We developed an alternative method in which a circle was drawn, centered around each city’s centroid. The radius of each circle was determined by adding 10 miles (the width of the buffer in the earlier method) to the square root of the city’s square mileage divided by π. (In short, we made the simplifying assumption that cities were circular in shape and used the equation area = π x radius2 to calculate an approximate city radius, then added 10 miles to that radius figure.) We then used the mapping program to calculate the aggregate income of the population within this circle. As it turned out, using this method gives results that are extremely closely related (r = .9992) to the preferred “buffer” method, in those cities where both could be computed. There is one further caveat to note. The mapping program apportions populations within each census tract to our circle area based simply on the percentage of the tract’s land area that falls within the circle. This procedure is not generally a problem in most cities (particularly in metropolitan areas) where tracts are quite small. It is, however, a potential problem in a small number of rural cities, where tracts surrounding the city can be huge and sparsely settled (thus leading to an unrealistically low apportionment of the population to the circle). After some diagnostic checks, it was decided to drop all observations where the city’s aggregate income (within its actual boundaries) comprised more than 60 percent of the aggregate income in the wider circle area. We concluded that this deleted the “problem” cases. As a result, 21 cities were dropped from our model estimation. Please contact the authors for any further information on the construction of the aggregate income variable. 48 Another set of variables concerns cities’ locations and their “age”—or stage of urbanization. Using the mapping program, we were able to determine whether each city has an interstate highway within its borders as of 1990, and we include a dummy variable measuring the presence of one of these major freeways. Highway accessibility is often thought to be vital for many high-volume retailers. We also include two dummy variables—urbanization and central city—that relate to whether the Census Bureau defined each city, as of 1990, as being part of an urbanized area, or as being a core city in a metropolitan area. Urbanized areas present more of a “critical mass” of shoppers that can lead to extensive retail competition and support specialty or luxury stores. Chain stores tend to seek agglomeration economies, sometimes avoiding smaller regions which are more difficult to service logistically from centralized warehousing and distribution networks (Kroll and Marrinan, 1985b, pp. 25–28). Research also shows that metropolitan areas experience more internal specialization in their retailing geography than rural areas (Berry et al., 1988, p. 164). Central cities, in particular, are traditional centers of commerce and are often the most accessible location in their respective regions. Thus, one might expect these variables to be positively associated with sales tax revenues per capita. We also include population density, because of the possibility that retailers might be attracted by concentrated populations—or, alternatively, repelled by the congestion and high land costs associated with such concentrations. Cities with larger land areas per person (holding other factors constant) may be those that have “cherry-picked” retail properties on their periphery by annexation. On the other hand, low-density cities may include more outlying rural areas or exurban 49 residential development, which are less likely to include retail facilities. Thus, the effect of density on sales tax success is uncertain. In addition, we include dummy variables representing location in three key, distinctive regions of California: the five-county Los Angeles area, the nine-county San Francisco Bay area, and the 12-county Central Valley region. (The reference category is cities in other parts of the state.) Through use of mapping software, we also were able to identify geographically isolated cities—those with no neighboring cities within five miles of the city boundary. We also include this geographic variable, given our suspicion that cities without nearby “competitors” might be in a better position to attract and retain retail centers. In terms of cities’ degree of development, we include measures of the number of years incorporated—that is, the length of time since the community officially became a city—and the median age of housing structures in the city (among occupied units only). “Older” places, as measured by incorporation dates, are probably more likely to have established business districts. However, older housing may indicate a filtering-down of households and thus perhaps imply lower spending on retail. We therefore expect a positive relationship between sales tax revenues and number of years incorporated but perhaps a negative relationship between sales taxes and age of housing. Finally, we include two variables that measure local policies relevant to sales tax revenues. One is the percentage of sales taxes collected within the city that are shared with the county under a recognized agreement. Many cities in the state have such agreements with their counties to automatically pass through a percentage share of their sales tax revenues 50 to the county.4 Such sharing would have the effect, of course, of reducing a city’s per capita revenues. Second, we turn to the question of city redevelopment policies. Cities frequently employ redevelopment in their quest to attract or retain retail stores and shopping centers. We therefore include a variable measuring the percentage of each city’s land area taken up by public redevelopment projects. This serves as a measure of local effort devoted to redevelopment, which is a widely used policy instrument in California cities (Dardia, 1998). If it is true that redevelopment effort is related to success in attracting retail development, then this variable should be positively associated with per capita sales tax revenues. Results of Model Estimation An ordinary least-squares regression model was estimated to calculate the relationship between these independent variables and per capita sales tax revenues. It is important to note that the regression technique measures the effect of each variable, holding each of the other variables constant. Thus, although cities in the Central Valley tend to have low per capita incomes, the model is able to show the association between Central Valley location and sales tax revenues, independent of each city’s per capita income. We should also note that because of data limitations, the model includes 398 cities as observations, rather than the full 457 ____________ 4We rely on the sales-tax-sharing arrangements listed in the 1992 volume of the California State Controller, Financial Transactions Concerning Counties of California (p. 142). Many other cities have negotiated ad hoc tax-sharing agreements with their counties concerning specific properties that the cities have annexed. There is no source that would allow us to account for such arrangements. In addition, a small number of cities have special agreements to share Bradley-Burns revenues with their redevelopment agencies. We do not expect that these infrequent arrangements would affect our results here in any substantial way. 51 existing and reporting sales tax revenues as of 1991.5 Table 3.2 provides the results of the regression model, which accounts for 42 percent of the variation in per capita sales tax revenues. Of the 22 independent variables, 12 have a statistically significant relationship with per capita sales tax revenues, generally in the expected direction. The variables positively associated with sales tax “success” are population, household income, redevelopment effort, the presence of an interstate highway, and, more surprisingly, Hispanic share of the population. Other city characteristics are negatively associated with sales tax “success.” These include most strongly average household size as well as population density, black share of population, household income squared, location in the Central Valley or (weakly) San Francisco Bay area, and, perhaps surprisingly, population change. None of the other variables are statistically significant. Factors Related to Sales Tax Success: Interpreting the Results What can we make of these findings? Many are not particularly surprising. First, cities with large populations—thus, a larger potential customer base that tends to attract retailers—have more sales per capita. Another factor that helps bring in shoppers—the presence within a city of a major highway—is also clear in the results. Cities with large average household sizes, on the other hand, tend to have less sales tax revenues ____________ 5Including the population change (1985–1991) variable means that only the 441 cities incorporated before 1985 can be included. Missing data relating to the redevelopment variable result in several other cities being dropped. In addition, as discussed above, certain small rural cities were eliminated when calculating aggregate income in the city market area; and four tiny commercial enclaves with huge sales tax revenues per capita were dropped because they skew the dependent variable. 52 Table 3.2 Regression Model of City Sales Tax “Success” as of the Early 1990s Independent Variable Log of 1991 population** % change in population, 1985–91*** % senior citizens Average household size*** % black*** % Hispanic*** Median household income, 1989*** Median household income squared, 1989*** Log of aggregate income in city and adjoining market area Log of travel-related spending (tourism) in county, 1992 Urbanized (dummy variable) Central city (dummy variable) Log of population density** Interstate highway through city (dummy variable)*** Los Angeles region (dummy variable) San Francisco Bay area (dummy variable)* Central Valley (dummy variable)*** No other cities within 5-mile radius (dummy variable) Years since city incorporated Years since median housing unit built % of sales tax within city allocated to county, 1992 Redevelopment areas (1994) as % of 1990 land area*** Adjusted R-squared = 0.42 No. of cases = 398 cities Unstandard- Standardized ized Coef. (beta) Coef. .103 .17 –.467 –.15 –.008 –.06 –1.060 –.66 –.016 –.13 .014 .36 .000 .59 –.000 –.83 .040 .09 –.054 .033 .057 –.125 –.09 .02 .02 –.12 .310 –.106 –.229 –.317 .18 –.06 –.12 –.14 .077 .001 –.003 .03 .03 –.03 .009 .02 .728 .19 T-Value 2.12 –2.95 –1.01 –6.67 –2.80 3.90 2.87 –4.77 0.77 –1.14 0.24 0.38 –1.97 3.75 –0.86 –1.69 –2.80 0.66 0.57 –0.57 0.42 4.43 NOTES: *p < .1, **p < .05, ***p < .01. Dependent variable is the log of per capita sales tax revenues (average of 1991, 1992, and 1993), measured in constant 1995 dollars. All independent variables measured as of 1990, except where noted. 53 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) per capita. Household size is, in fact, the most statistically significant variable in this and every other model that was estimated. Figure 3.1 is a scatterplot diagram illustrating the negative bivariate relationship between household size and sales tax success. Some less obvious outcomes emerge as well. Cities engaging in a great deal of redevelopment effort—as measured by the percentage of land area in redevelopment project areas—tend to have more successful sales tax results. Thus, it may well be that the great amount of effort California cities have devoted to redevelopment has shown some results, at least in the narrow fiscal sense of sales taxes received. (Dardia, 1998, shows the limitations of redevelopment as a broader economic 8 7 6 5 4 3 2 1 0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Average number of persons per household, 1990 SOURCES: U.S. Census (household size); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.1—Relationship Between Average Household Size and City Sales Tax “Success” 54 development tool.) Or, perhaps it is the case that cities with larger retail districts tend to be most predisposed to use redevelopment. Cities with lower population densities (controlling for their urban/rural status) also have more retail sales per capita, perhaps because retailers may seek less-congested highway-oriented locations. Cities located in Central Valley counties—and surprisingly, the Bay area—tend to have less success in generating sales tax revenues, holding constant for socioeconomic and other factors. The Valley effect is considerably stronger.6 Cities’ ethno-racial characteristics also may have a role. Localities with higher shares of blacks in the population have lower sales per capita (again, this is controlling for income and the other variables in the model). One might conclude that retailers and shoppers avoid such locations. However, cities’ Hispanic population shares are positively related to sales tax receipts. It is not obvious why this is so. It is important to note, however, that among California cities, Hispanic population share is highly correlated with average household size (r = .82). This high degree of collinearity means that we should not be too confident in interpreting the sign and coefficient of the Hispanic variable (although the variable never failed to attain high levels of statistical significance in many specifications of the model). This issue illustrates some of the limitations of a regression model—where all other factors are held constant—if one is seeking policy implications. One should keep in mind that, because cities that are heavily Hispanic tend to have large household sizes, such cities actually tend to receive lower per capita sales tax revenues, as shown in Figure 3.2. ____________ 6The Bay area effect is only marginally statistically significant and is not robust to alternative specifications of the model. 55 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) 8 7 6 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Percentage of city population Hispanic, 1990 SOURCES: U.S. Census (percentage Hispanic); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.2—Relationship Between Hispanic Share of Population and City Sales Tax “Success” The relationship between cities’ median household income and per capita sales tax revenues reflects the finding discussed above—that many of the lowest-ranked sales tax cities are very wealthy bedroom communities. The bivariate relationship between per capita income and sales tax success is shown in Figure 3.3. The graph reveals that the relationship between city income and sales taxes is generally (though not dramatically) positive at lower- and middle-income levels, but a negative relationship emerges fairly clearly at upper-income levels. The regression results—income positive, income-squared negative—bear this out. City population change, though expected to increase sales because of housing-related purchases, actually is associated with lower sales taxes per 56 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) 8 7 6 5 4 3 2 1 0 0 50,000 100,000 150,000 200,000 250,000 Median household income, 1989 (dollars) SOURCES: U.S. Census (household income); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.3—Relationship Between Income Level and City Sales Tax “Success” capita. This may be the case because the fastest-growing cities tend to be at the fringes of metropolitan areas—places that are relatively heavy in housing and low in commercial land uses. In short, the retailers may simply have not caught up with population movement, with the effect being that rapidly growing cities are “under-retailed,” with residents traveling to more established business centers to do their shopping. One of the most puzzling results is that the aggregate income of a city’s wider market area does not seem to influence its sales tax success. 57 Although the relationship is in the expected positive direction, it is not statistically significant in this specification.7 Cities’ Relative Gains over Time: A Longitudinal Model We turn next to a related question: Which types of cities have gained and lost ground in sales tax revenues relative to other jurisdictions in recent years? To address this issue, we examine each city’s change in per capita sales tax revenues from the early 1970s to the early 1990s (in inflation-adjusted 1995 dollars). Specifically, we take the average of the per capita sales tax revenues for each city for 1971, 1972, and 1973, and again for 1991, 1992, and 1993. To deal with the extremely high per capita sales taxes some cities exhibit, we again use logarithms, subtracting the log of the average per capita sales tax revenues in the early 1970s from the log of the per capita sales tax revenues in the early 1990s. For this analysis, we are interested in how changes in the various demographic and other characteristics of each city are related to changes in sales tax receipt. Most of the variables are analogous to those in the cross-sectional model above, except that they reflect longitudinal change. We do include the initial (1970 or 1971) levels of a few potentially important variables, such as population size and urbanization status, because we might expect these initial characteristics to influence subsequent changes in cities’ retail market activity as well. Because ____________ 7We also estimated a regression model with county-level dummy variables for all counties containing two or more cities, to control for any unobserved characteristics of each county that might affect sales tax success. Other independent variables were the same as above, except that the tourism, Bay area, Los Angeles area, and Central Valley variables were eliminated as they were subsumed by the county dummies. The results of this model were essentially the same as those reported above. The county dummies lacked (joint) significance, and the explanatory power of the model was unchanged. 58 certain data (specifically data on Hispanic population shares and on land area) were unavailable for cities with populations below 2,500 in the 1970 Census, we include in the model only cities with 1970 populations of 2,500 or more.8 This also has the benefit of eliminating the outlier cities of Colma, Industry, Sand City, and Vernon and substantially improving the fit of the model. Another data limitation is that our measure of the aggregate income, or purchasing power, of each city’s wider market area cannot be constructed from the 1970 and 1980 data. As an (inferior) alternative, we use the change in aggregate income of the entire county in which each city is located.9 Another difficulty involves measuring changes in redevelopment effort, since we do not know the share of city land area accounted for by redevelopment projects in the 1970s. Instead, we use a variable that measures the number of new redevelopment sites designated in each city between 1970 and 1989. In addition to the “change” variables, we include dummy variables for cities considered central cities by the Census Bureau, and those that were considered urbanized as of 1970. We also include a dummy for cities experiencing a change from rural to urbanized status between 1970 and 1990. The model also takes into account the initial population level, ____________ 8The Census Bureau used a somewhat different concept of Hispanic populations in the 1970 Census than in later years. Thus, strictly speaking, cities’ “percent Hispanic” are not fully comparable across decades. However, cities’ shares of Hispanic population relative to one another are comparable in each Census, and thus cities’ relative change in Hispanic share—the variable we use here—should be only minimally affected by the definitional change. In any event, the Hispanic variable turns out to be unimportant for the longitudinal analysis. 9A model using county-level dummy variables, and therefore absorbing the variation of the county aggregate income variable and region variables, produces very similar results to those presented here. 59 average household size, and median family income of each city.10 (Unfortunately, we lack data on the presence of an interstate highway in each city as of 1970, so this variable is not included.) Finally, we have two measures of each city’s “initial position” with respect to fiscal policy. One is its property tax revenue percentage of own-source general revenues in 1971. We might anticipate that low property tax cities sought sales tax revenues more vigorously—although we cannot predict whether they were successful in attracting retail. The second fiscal measure is each city’s sales tax “success” as of 1971–73. This variable is the log of each city’s average per capita sales tax revenues for 1971, 1972, and 1973. We include this variable on the thought that perhaps “the rich got richer”— that is, initially successful cities might have consolidated their advantages as retail centers over time. Table 3.3 displays the results of the model estimated, which accounts for about half of the variation among cities. Nine of the 20 variables show statistically significant relationships with the change-in-sales-tax variable. Three of these were positively associated. Cities that were urbanized in 1970 did better relative to their nonurbanized counterparts. So did those changing their status from nonurban to urbanized, and those with increases in population density.11 Six variables are negatively associated with sales tax “success” over time. Some of these are not surprising, given our earlier results. For example, a growth in average household size is negatively related to changes in per capita sales tax revenues. And once again, cities seeing ____________ 10Household income is not available for 1970, so we use family income instead. 11Cities with greater gains in county aggregate income—that is, more dollars of potential spending money within broad commuting range—did better over time at gaining sales tax revenues, but this result was just below the level of statistical significance (p < .104). 60 Table 3.3 Regression Model of Changes in City Sales Tax “Success” Between Early 1970s and Early 1990s Unstandard- Standardized T- Independent Variable ized Coef. (beta) Coef. Value Log of 1971 population* –.048 –.13 –1.90 % change in population, 1971–91*** –.085 –.21 –3.62 Change in % senior citizens, 1970–90** –.015 –.14 –2.25 Average household size, 1970 –.008 –.01 –0.14 Change in average household size, 1970–90*** –.530 –.52 –6.20 Change in % black, 1970–90 –.001 –.01 –0.13 Change in % Hispanic, 1970–90 –.002 –.07 –0.88 Median family income, 1969 –.000 –.10 –1.47 Change in real median family income, 1969–89 .000 .05 0.70 Change in real aggregate income of county (log of 1990 value minus log of 1970 value) .127 .08 1.63 Urbanized as of 1970 (dummy variable)*** .244 .27 3.55 Changed from rural to urbanized, 1970–90 (dummy variable)* .142 .10 1.96 Central city (dummy variable) –.057 –.04 –0.81 Change in population density (log of 1990 persons per square mile minus log of 1970 persons per square mile)** .125 .11 2.19 Los Angeles region (dummy variable) –.003 –.00 –0.05 San Francisco Bay area (dummy variable) .024 .02 0.39 Central Valley (dummy variable)* –.107 –.10 –1.82 No. of new redevelopment areas designated, 1970–89 .000 .00 0.02 % of own-source general revenues from property tax, 1971 –.109 –.04 –0.83 Initial sales tax “success” (log of per capita sales tax revenues, average of 1971, 1972, 1973)*** –.162 –.27 –5.21 Adjusted R-squared = 0.49 No. of cases = 345 cities NOTES: *p < .1, **p < .05, ***p < .01. Dependent variable is the difference between the log of per capita sales tax revenues as of the early 1990s (average of 1991, 1992, and 1993) and as of the early 1970s (average of 1971, 1972, and 1973), measured in constant 1995 dollars. 61 population gains lost some of their relative standing in per capita sales tax revenues over this period. In this model, however, a higher initial population level also had negative effects on subsequent sales tax performance.12 Central Valley communities have done worse over time relative to other cities. A growth in senior citizen share of the population also led to declines in cities’ relative performance. Finally, it is interesting to note that cities with high levels of initial sales tax success—that is, higher per capita sales tax revenues in 1971– 73—declined over time relative to other cities. This could indicate a “regression to the mean,” as discussed in Chapter 2. Rather than “rich” (in sales tax) cities getting richer, it appears that retail-heavy places lost some of their relative advantages during the 1970s and 1980s.13 We also estimated separate regression models for changes over each decade—the 1970s and 1980s—to see whether patterns differed in these two periods. Some of the results were broadly similar across the two decades—for example, the negative effects of population increases, increases in household size, and initial sales tax success. However, some differences by decade were apparent. Higher-population cities suffered ____________ 12In the earlier cross-sectional model, cities with higher populations are seen as having better performance. In the longitudinal model, cities with initial higher populations do worse over time. These results are not necessarily inconsistent. Cities with large populations have probably always done better than small cities; it is just that their relative dominance has deteriorated since the 1970s. A more anomalous finding is that population density has a negative effect in the cross-sectional model, but increases in density have positive effects in the longitudinal model. It may be that cities that saw notable drops in density over this period either were experiencing neighborhood decay or were annexing outlying, low-density areas that tend to be primarily residential. In either case, sales tax performance would likely suffer. 13Note also that cities with low portions of revenue from the property tax, contrary to the speculation of some observers, have not gained ground in sales tax revenues over time. These cities began the period with high sales tax revenues—indeed, this is what enabled some of them to avoid levying a property tax. The fact that these cities have also ended up with higher-than-average sales tax levels reflects their initial situation, not any changes in their success in the wake of Proposition 13. 62 particular decline during the 1970s but did not decline in the 1980s; and urbanized status was a particular advantage in the latter decade. Also, the relative decline of Central Valley cities occurred entirely in the 1980s. Summary The question posed by the title of this chapter was “Which types of cities have benefited from the situs rule?” We offer the following answers, based on the data analysis in this chapter. We present only the conclusions about which we are most confident. These relationships hold true while “holding constant” for other factors of potential importance to sales tax revenue generation. To summarize, we performed two analyses of sales tax “success.” One examined cities’ sales tax revenues per capita as of 1991–93. We will refer to this as the “cross-sectional” model. The other examined cities’ gains in per capita sales tax revenues, relative to other cities, between the early 1970s and early 1990s. We will refer to this as the “over time” model. The results of the two models indicate that the following types of cities have done better relative to other cities: • Cities with smaller or declining household sizes—generally meaning families with fewer children and a larger proportion of childless single adults (cross-sectional and over time models); • Cities that have not had rapid population growth (cross-sectional and over time); • Cities located outside the 12-county Central Valley region (cross-sectional and over time); • Cities more heavily engaged in redevelopment efforts (crosssectional only); 63 • Cities traversed by interstate highways (measured in crosssectional model only); • Cities that have middle to upper-middle income levels—but not very-high-income communities (cross-sectional only); • Cities with lower shares of black population (cross-sectional only); • Cities with a decreasing share of population aged 65 or older (over time only); • Cities in urbanized areas and those changing from rural to urbanized (over time only); • Cities that began the period with lower-than-average per capita sales tax revenues (over time only); and • Cities with higher populations (cross-sectional model)—but note that those with larger populations as of 1970 lost some of their relative advantage subsequently (over time model). Thus, the situs rule for distributing sales taxes does create distinct revenue winners and losers, though not in any very simple pattern. One might suspect upon first considering the issue that communities with higher socioeconomic status profiles—perhaps growing suburbs—would be most advantaged by a system that returns revenues to jurisdictions where sales occur. As we have seen, however, the picture is more mixed. Although it does appear to be the case that cities with larger black populations are disadvantaged by the rule, it is also true that very wealthy cities are disadvantaged. Central Valley communities—and particularly cities with high household sizes—are disadvantaged, but once we control for these factors there is some evidence (clouded by methodological complexities) that cities with larger Hispanic populations actually do 64 better.14 Moreover, rapidly growing cities appear to be doing worse than communities with more stable populations, and it is demonstrably untrue that the “rich” (cities with high sales taxes per capita) got “richer” between the 1970s and 1990s. In short, if policymakers hope that dismantling the situs rule for distributing the Bradley-Burns sales tax will mainly benefit cities with disadvantaged populations, they would be mistaken. In fact, some of the largest beneficiaries could be the wealthiest communities. We will further discuss the issue of alternatives to the situs-based system in our concluding chapter. The finding that redevelopment effort appears to lead to sales tax advantages for cities does raise the central issue of the fiscalization of land use: To what degree can local land-use policies influence cities’ standing in the sales tax revenue hierarchy? And whether or not such land-use policies are effective, is it true that cities do in fact favor retail over other types of development? These issues are explored in the next two chapters. ____________ 14Since cities with large Hispanic populations also tend to be those with large household sizes, however, this is not of much consolation. 65 4. The Sales Tax and Local Land-Use Decisions Across our region, virtually every city pursues what they call “economic development.” In virtually all cases, what they really mean is “real estate development.” And in the vast majority of cases, what that essentially boils down to is “sales tax development” (Cole, 1998). Thus far, we have examined the fiscal consequences of the situsbased local sales tax, looking at the distribution of revenues across cities. Another major policy controversy regarding this tax, however, concerns allegations that the quest for sales tax dollars has led local governments to bias or distort their development decisions in favor of retail. The result, critics charge, is a “fiscalization” of land use, with negative consequences for regional land-use planning and for the California economy. In this and the next chapter we explore this issue. We begin in this chapter with a conceptual discussion, considering the ways the local sales tax might affect development patterns and the policy considerations this raises. In Chapter 5, we will draw upon the findings of a specially designed survey 67 of city officials, focusing on their land-use strategies, to provide some evidence on the ways the local sales tax affects their decisionmaking. What Is Meant by the “Fiscalization” of Land Use? Kotin and Peiser (1997, p. 1975) define the fiscalization of land use as “the tendency of communities to establish land uses based on the net tax revenues they will generate for the city.” The city planning profession, which pioneered the concept of fiscal impact studies, “succeeded beyond their wildest dreams in making citizens and public officials aware of, and concerned about, the fiscal impacts of development” (Bunnell, 1997, p. 137). Misczynski (1986) made the first major written explication of the argument that California land-use decisions were becoming driven by fiscal considerations. His concern was not primarily the sales tax, however, but rather the growing reliance of local governments on development fees and arcane financing arrangements to provide public facilities such as schools and roads. Such mechanisms have become far more prevalent and important as the state and federal governments have withdrawn from providing funds for local public infrastructure. More recently, however, the term “fiscalization of land use” has been applied largely to local competition for retail businesses, as local governments seek situs-based sales taxes. As we have seen, the local sales tax is one of the few revenue sources with the potential—at least in theory—for substantial growth as a result of city decisions. Plus, its discretionary character as a revenue source makes it particularly attractive. Thus, current critiques of local governance and public finance in California often devote much attention to the quest among localities for sales tax dollars—or so-called “cash-box zoning.” We are not aware 68 of any studies that have systematically examined the effects of sales taxes on the land-use regulation behavior of local governments. Numerous authors, however, have alluded to these issues in California. “To the extent that land use decisions are now driven by their fiscal consequences,” wrote Chapman (1998, p. 11), “fiscalization has occurred.” He points to cities’ recruitment of so-called big-box stores and car dealerships as prime examples of this tendency, noting that “most jurisdictions trying to maximize sales tax revenues choose to encourage these types of development over residential development, which generates sales tax revenue only to the extent that the new residents shop in the same city in which they live” (p. 12). Columnist Neal Peirce sees fiscalized land-use decisions as “meaning a wild scramble for retail base especially for mega-auto malls and big box retailers. The result is not just more suburban sprawl, but real ugliness . . . .” (Peirce, n.d.). Schwartz (1997, p. 184) complains that “the reliance on sales taxes to replace lost property tax revenues has motivated planning and economic development decisions that sacrifice the long-term fiscal and environmental health of communities for short-term gains in sales tax producing land uses.” This dynamic can operate when cities recruit bigbox stores that then deplete the vitality of the existing downtown area. Fulton devotes an entire chapter in his 1997 book on the Los Angeles region, The Reluctant Metropolis, to an examination of what he calls “sales tax canyon”—the retail-dominated landscape that has emerged through competition between the cities of Ventura, Oxnard, and Camarillo in Ventura County. With property taxes restricted and land uses fiscalized around the sales tax, “Proposition 13 has provided the architecture on which Southern California’s urban landscape has been built ever since” (Fulton, 1997b, p. 260). In another important book, 69 Schrag (1998, p. 178) wrote that “since the biggest and cheapest source of local revenue was the sales tax, the planners not only pursued shopping centers, outlet stores, and auto malls; they often did so in preference over employers engaged in light manufacturing, even if they promised better jobs, because their enterprises did not return any sales tax to the local community . . . .” Schrag also points to cities’ aggressive use of redevelopment agencies, with their ability to issue debt, in efforts to attract retail development (p. 180). Tax increments from properties in the redevelopment areas are used in some cases to provide inducements or infrastructure improvements for large-scale retail projects in the area.1 Chapman (1981, p. 75) sees a “close connection” between situs-based sales tax collection and the use of redevelopment, claiming that “to some extent, the combination of these two institutional factors has had an impact on the development patterns of jurisdictions, as can be seen in the proliferation of shopping malls throughout the state.” Kotin and Peiser (1997, p. 1971) write that “redevelopment agencies have been particularly aggressive in pursuing high volume retailers such as Home Base, Price Club, K-Mart, Wal-Mart and Costco because of the sales taxes they generate . . . .” They find that, given increasing competition among redevelopment agencies for these retailers, a greater share of the financial benefits of these deals has gone to the retailers, developers, and particularly landowners. However, they also note that some cities have been able to set performance-based incentives in their deals with ____________ 1The passage of the redevelopment reform law, AB 1290, in 1994 restricts the ability of cities to use redevelopment funds to finance retail projects on vacant land sites. For a broader discussion of the fiscal issues of redevelopment, including AB 1290, see Dardia (1998). 70 developers, which can have the effect of protecting the city’s share of the benefits. Many stories circulate among California policy officials regarding allegedly fiscalized land-use decisions. Some have cited Monrovia’s decision several years ago to turn away a Kodak plant for a vacant site in its community in favor of a big-box retail store. Critics claimed that the city thereby traded well-paying manufacturing jobs for retail jobs that pay roughly half as much (Barber, 1994; Schrag, 1998). Similarly, “the city of Indio sacrificed a historic neighborhood through the use of its powers of eminent domain so that a mall developer could triple the size of the local shopping center” (Schwartz, 1997, p. 200). Others discuss strategic annexations by cities, where, for example, cities may annex commercial strips but not the surrounding residential areas (Senate Local Government Committee, 1989).2 One city official summarized the situation in simple fashion, telling us that although he often sees elected officials cutting ribbons at new retail developments, he never sees them do the same at new residential subdivisions. Thinking Carefully About the Fiscalization Argument Although rarely addressed systematically, the critique against sales tax-driven local land-use decisions seems to rest on certain assumptions or hypotheses, which are often left unstated. These include the following: ____________ 2On land-use conflicts between cities and counties, see Brooks (1988); Senate Local Government Committee (1989); Sokolow (1993); League of California Cities (1996); and Fulton (1997a). 71 • Hypothesis 1: Gaining sales tax revenue is a key goal for local land-use decisions, and thus retail development is given favorable treatment over other types of land uses. • Hypothesis 2: This favoritism toward retail has the effect of retarding residential and industrial development. • Hypothesis 3: The built landscape would look systematically different in the absence of situs-based local sales taxation. We would argue that the term “fiscalization of land use,” when applied to the local sales tax, presumes at a minimum Hypotheses 1 and 2. Some critics appear to imply (the far more adventurous) Hypothesis 3 as well. In thinking about these hypotheses, one must immediately confront one analytic issue. When considering the effects of the local sales tax on land-use decisionmaking, we are implicitly comparing the present system to some alternative system. Let us assume that the “alternative world” would be a California with a sales tax rate that is the same as it is now, but in which revenues were not returned to local governments on a situs basis. Favoring Retail? Let us now consider these assumptions. First, do local governments prefer retail development? This question is an empirical issue which, as of yet, no studies have systematically explored. The next chapter will take up this question in detail. Until now, anecdotal evidence of cash box zoning, and journalistic accounts of development policy in specific communities, have been used to support this argument. But even if all localities desire retail, not all of them are in a good position to get it. Retailers will tend to locate only 72 where there is market demand, reasonable transportation accessibility, and a population of sufficient size and purchasing power for their products and where the market is not already saturated with competitors. For some jurisdictions in remote locations or built-out market areas, significant retail development is unlikely, no matter how much they might desire it. Then, too, there is only so much “potential” retail in any given market area. At a given population, and a given level of income, taxable sales are finite; residents will not respond to every increase in retail “supply” with increased demand. Let us assume that a jurisdiction desires retail, has the potential to attract it, and there is in fact “room for growth” in the local retail market, because the population or income levels of the area are growing. What tools or policy levers does a local government have at its disposal to manifest that preference for retail? California law forbids making a “gift of public funds” to private firms or individuals, and so outright subsidies or donations of land to retailers or developers are theoretically forbidden (though there is legal uncertainty about the exact meaning of this doctrine and there are ways to sidestep it). Other city policies, however, can be used to support retail development, including the following: • Development fees or Community Facility District taxes that might ordinarily be applied to new developments to compensate the city for the impacts of the new growth may be waived or reimbursed, or held to artificially low levels. • Permitting and other bureaucratic procedures for new development might be expedited for retail developers. • “Excessive” amounts of land might be zoned for retail use, in the hopes that an increased portfolio of potential land sites might lure more retail developers. 73 • Local governments might be more inclined to grant a general plan change, or rezoning, to retail developments than to other types of land uses. • Redevelopment agencies might be used to clear land and assemble land parcels of sufficient size for major retailers—at below-market prices. • Redevelopment agencies, and city governments themselves, may provide the resources to build or improve infrastructure and amenities to serve retail developments. • Before recent reforms, some redevelopment agencies shared (“rebated”) city sales tax revenues directly with retail developers (particularly in the case of big box stores). • Cities with Enterprise Zones may use this designation to waive fees and expedite decisions for retailers. • Economic development officials might actively market their jurisdiction to retailers and retail developers.3 The issue then becomes whether these techniques are used so pervasively as to systematically change land-use patterns in the state. Of course, if all localities use such inducements to retailers, then the advantages individual cities might gain from these policies will be “competed away.” Presumably, the end result, in the aggregate, would ____________ 3One other technique that localities might use, at least in theory, is to lower the sales tax rate within their boundaries. This issue is more relevant in some other states where municipalities have a wider degree of discretion in the local sales tax rate they will levy. In California, however, the one-cent Bradley-Burns rate is uniform throughout the state, and so the border problem does not occur. The overall sales tax rate does differ somewhat across counties, because some counties levy additional voter-approved sales taxes, at their option, for special purposes, mainly transportation. It is unlikely that any county government would decide to rescind these add-on sales taxes simply in the hopes of luring more retailing, however, since most of the presumed benefits of the increased retail sales would be enjoyed by cities, not the county, and since “border effects” are less significant at the geographic level of counties. 74 be a shift of resources from local governments to the retail sector. If retail markets are competitive, then any cost advantages retailers enjoy because of these local policies will tend to result in slightly lower prices for retail goods. Thus, some might argue that consumers are in some sense the beneficiaries of the local governments’ largesse. Disfavoring Other Types of Growth? If local governments favor retail—whether or not they can actually hope to get any—what are the implications for residential and industrial development? In answering this question, we can see some possible effects of sales tax competition on land use. If jurisdictions zone “excessive” amounts of land for retail, then on the margins, less land is available for potential industrial or residential development, making such projects somewhat more expensive to build. Similarly, if retailers are more likely than housing or industry to win exemptions from the local zoning code, then there will be somewhat more flexibility and certainty in the retail development market, which will advantage that type of land use, at the margins. And in those (perhaps rare) instances where a retail development proposal competes with an industrial or housing proposal for the same land site, as in the Monrovia example, then the locality will tend to choose the retailer. Overall, then, a general predisposition toward retail by local governments probably will tend to make housing and industry marginally more difficult, uncertain, and expensive to develop. Changing the Landscape? This brings us to Hypothesis 3. Would the built landscape of California, or any of its metropolitan regions, look markedly different in the absence of a situs-based local sales tax? We can consider this question 75 in terms of two separate issues. First, does competition for retail lead to “extra” retail development? Second, does the retail industry in California have different location patterns than it would absent the sales tax? If cities systematically disfavor housing and industrial development in a way that makes it costlier or less predictable to develop, then the logical corollary is that retail development would be slightly cheaper, easier, and more certain. This would tend to lower the business costs of retailers. It is conceivable, but hardly certain, that because of these lower costs, slightly more retailing might exist; California would have more of its land use devoted to retail than would otherwise be the case. According to this line of reasoning, businesses that might otherwise be operating just below their break-even point would now find it possible to stay afloat financially. Alternatively, the same amount of retail business might exist, but it might consume slightly more square footage than it otherwise would, because retail space could be developed more inexpensively. This logic is very tenuous, however, because, as we have noted, the reduced costs of retail development are just as likely to be passed along to landowners or customers, rather than to stores’ profit margins. Although it is difficult to find data that would help evaluate this possibility, a “shopping center census” conducted by a private trade group casts doubt on the idea that California is especially heavily “retailed.” The International Council of Shopping Centers (1996), using data collected by a private firm on shopping centers nationwide, reported that California in 1995 had 19.2 square feet of “gross leasable area” of shopping space per capita. This figure put California only slightly above the national average of 18.9 square feet. The state ranked 19th of the 50 76 states—hardly an unexpectedly high rank given California’s wealth and urbanization. A second way that the sales tax might affect land-use patterns concerns the location of retail centers within metropolitan areas. Does California’s situs rule result in retail location patterns that are systematically different from those we would expect in the absence of a local sales tax? Retailers use certain standard techniques and rules of thumb in deciding where to locate (Davies, 1976; Guy, 1994). These depend on the type of goods that are for sale, the price level of those goods, and the special niche of the retailer (mass merchandise versus specialty, for example). No matter what inducements a local government offers, a Wal-Mart is not likely to open a new store in downtown San Francisco, a luxury car dealership is not likely to move its showroom to a small Central Valley farm town, and a dealer of specialized stereo equipment is not likely to find a location on a Sierra mountaintop optimal. Retailers locate mainly based on factors that are not subject to much control by local governments (Schneider, 1986). Consider the issue of “downtown” location versus “suburban” location, frequently discussed by those interested in urban development policy. In general, chain stores and general merchandise stores, which are among the most sought-after retail land uses, tend to seek middle- or upper-income suburban areas with good highway access for their new locations. “For regional and superregional centers, this often implies proximity to a freeway offramp; for small centers, traffic circulation on adjoining travel routes is important” (Kroll and Marrinan, 1985b, p. 25). This is true in California, as it is throughout the United States. Thus, a given retailer’s potential locations may be limited to “western San 77 Bernardino County along the I-10;” for another, the relevant region might be “the tri-Valley area of Alameda County.” Within such a limited universe, the retailer or developer might then shop around for the most attractive “deal”—consistent with their locational needs. But in any event, the location is likely to be a freeway-oriented suburban parcel, and this would be true whether or not the situs-based sales tax exists. Specific jurisdictions may “win” and “lose” in competing to host a given retailer, but the winner and loser locales are probably fairly similar places within the metropolitan hierarchy. Thus, overall patterns of retail location are not likely to be significantly affected. Implications and Policy Considerations It is probably safe to say that in any system where local governments raise a substantial portion of their funds by taxing or applying fees to activities within their boundaries, land uses will always be fiscalized, in one way or another. In some states with no local sales taxes whatsoever, for example, it is well known that municipalities are picky about the types of development they allow because of the amount of property taxes and service needs different land uses generate. Office and research facilities, for example, might be welcomed, whereas multi-bedroom apartments would be shunned, since the latter cost far more in local services than they provide in property taxes. Local governments, armed with planning studies, have become ever more sophisticated at determining these “fiscal impact hierarchies” (Bunnell, 1997; see also Orfield, 1997, p. 85). In California, the property tax is a very constrained revenue source for local governments, and its allocation is largely outside their control. Some cities receive a very small slice of the property tax dollar generated 78 by any particular parcel, with other portions being allocated to school districts, the county, and special districts. Thus, cities have become increasingly aggressive about the types and amounts of fees and assessments they can levy on new developments, with estimates of such charges for new housing construction in one county estimated at $20,000 to $30,000 per unit (Dresch and Sheffrin, 1997). But the most oft-remarked type of local land-use strategizing has involved competing for land uses that generate local sales taxes. As we saw in earlier chapters, however, overall per capita sales tax revenues available to California cities have been stagnant. If cities are increasingly competing, then they are fighting over a largely fixed pie. We have also seen that the overall hierarchy of cities in terms of per capita sales tax revenues has been relatively stable. Individual cities are not likely to leap ahead of their neighbors in their sales tax “success,” and those that have been “leaders” in the past have tended to perform less well than other cities over time. If, as we have argued, the retail landscape is probably not systematically reshaped by sales tax competition, and if cities are fighting over a fixed pie of retail businesses that would have located in their regions regardless of any inducements, then the main effect of fiscalized land-use decisionmaking in California is probably to shift resources from the public sector to retailers, their developers, and landowners.4 A second likely effect of favoritism toward retail is an underprovision of residential and industrial development, although we cannot measure this effect. It is unlikely that retail space will be “overproduced” or shifted markedly between central cities, suburbs, and small towns. ____________ 4This may, in turn, lead to slightly lower retail prices, so that Californians as taxpayers are to some degree subsidizing themselves as consumers. 79 Note that there may well be a “prisoner’s dilemma” character to sales tax competition, as well. That is, although cities locked into competition for retailers are not assured of any positive results—and lose out in the aggregate—cities that choose not to compete for retail will most likely lose sales tax revenues. It is very difficult for any individual city to unilaterally pull out of the sales tax game. All of these likely results, however, hinge on the assumption that jurisdictions do in fact actively favor retail development. The next chapter will provide empirical evidence to help evaluate this assumption. 80 5. The Preference for Retail Development Concern about fiscalization of land use in California tends to focus on competition between city governments for sales tax revenue. Observers argue that cities tend to bias land-use decisionmaking in favor of retail development to capture sales tax revenue, since this is one of the few sources of general revenues with much growth potential over which local governments have significant discretion. The most fundamental assumption on which this argument is based is that city governments do indeed favor retail development to maximize sales tax revenue. However, no systematic studies have attempted to confirm this hypothesis, and evidence in the literature to date has been largely anecdotal. In this chapter, we provide strong evidence to confirm that city governments in California do systematically favor retail development over other land uses when it comes to new development on vacant land, as well as redevelopment in designated “blighted” areas. These 81 conclusions are based on a recent PPIC mail survey of city managers regarding city development strategies. The survey results confirm that gaining sales tax revenue is the highest priority for city managers when making land-use decisions in relation to new development and redevelopment, and it is near the top of their list of considerations in relation to annexations of new territory. According to survey respondents, proposed retail projects are also the most likely to garner a zoning change or financial incentive from city officials to developers. Although the survey results point to certain regional distinctions in California in regard to attitudes toward development, the preferred status of retail development is generally consistent across all regions in the state. Survey Methodology In August 1998, PPIC mailed a questionnaire to the top administrative official—generally the city manager or city administrator—in each of the 471 cities then in existence in the state. The questions addressed three distinct but related topics: new development on vacant land sites, city-backed redevelopment in designated “blighted” areas, and annexation of new properties outside the city limits. Respondents were asked about their city’s level of activity in regard to these three types of development or redevelopment. If their city was at least somewhat actively engaged in an activity, respondents were then asked about their development priorities in reference to that activity. Regarding new development and redevelopment projects, respondents were asked to rate, on a one to seven scale, the desirability of seven types of land uses and the importance of 18 major strategic considerations in guiding their decisions and responding to development or redevelopment proposals. For new development only, respondents 82 were also asked about the likelihood that they would offer a general plan change or other incentive to attract seven different types of development. Finally, respondents were asked to rate the importance of 12 motivations for annexing new territory. A copy of the questionnaire is included in the appendix at the end of this report, which also provides more detail regarding survey methodology and the respondents. Officials from more than two-thirds (70 percent) of the state’s cities responded to the survey. In terms of population size and regional location, the cities that responded to the survey closely resemble the overall breakdown in the state. Thus, the results should be reasonably representative of overall attitudes and trends. Tables 5.1 and 5.2 show the breakdown of respondent and non-respondent cities by population size and region. Table 5.1 Survey Respondents, by City Population Size (in percent) Population in 1998 Less than 10,000 10,000–24,999 25,000–49,999 50,000–99,999 100,000–249,999 250,000 or more All California cities Respondents 25 21 22 19 10 3 100 n = 330 NonRespondents 36 21 21 15 7 0 100 n = 141 All California Cities 28 21 21 18 9 2 100 n = 471 NOTE: Totals may not add to 100 percent because of rounding. Development Activity in California Cities The overwhelming majority of California cities are able to pursue some form of development or redevelopment activity, as Table 5.3 83 Table 5.2 Survey Respondents, by Region (in percent) Region Los Angeles area Bay area Central Valley Other All California cities Respondents 38 23 18 21 100 n = 330 NonRespondents 37 18 21 25 100 n = 141 All California Cities 38 21 19 22 100 n = 471 NOTE: Totals may not add to 100 percent because of rounding. Table 5.3 New Development and Redevelopment Activity in California Cities Amount of Vacant Land Available for New Development % with “considerable” amount % with “limited” amount % with no vacant land Total Level of Redevelopment Activity Very Not Very Not Active Active Active 19 7 8 18 9 8 13 7 11 49 24 27 NOTE: Totals may not sum exactly because of rounding. SOURCE: PPIC City Managers Survey, 1998. Total 34 35 31 100 reveals. Only 11 percent of all cities surveyed have no vacant land available for development and are not engaged in any redevelopment activity. Over half (53 percent) both have vacant land available for new development and pursue redevelopment activity. Table 5.4 provides information on development patterns by region for the three types of development addressed in the survey: new development on vacant land, redevelopment in designated blighted areas, and annexations of new territory. As the table indicates, about two- 84 Table 5.4 Development Activity, by Region San Los Francisco Central Angeles Bay Valley Amount of vacant land available for new development % with “considerable” amount 27 19 68 % with “limited” amount 32 32 29 % with no vacant land 41 49 3 Level of redevelopment activity % very active 58 48 50 % somewhat active 22 22 27 % not active 20 30 23 Annexation plans within next five years % five or more square miles 13 4 13 % one to five square miles 19 16 52 % less than one square mile 11 32 22 % that cannot annex new territory 24 12 0 % that can annex, but will not 33 36 13 SOURCE: PPIC City Managers Survey, 1998. Other 35 46 19 37 26 37 10 25 32 9 25 State Total 34 34 31 50 24 26 10 26 22 13 28 thirds of surveyed cities have some vacant land available for new development. Of those cities, half report a “considerable” amount, and the other half have only a “limited” amount of vacant land. Cities in the Central Valley have the most vacant land. Two-thirds of them have a considerable amount but only 3 percent have none. San Francisco Bay area cities have the least vacant land. Nearly half have none at all whereas another third have only a limited amount. As Table 5.4 also reveals, three-quarters of respondent cities pursue some redevelopment activity, and half are very active. Los Angeles area cities are especially active; four-fifths are engaged in redevelopment, of which three-quarters engage in it very actively. Even in the rural regions of the state, however, the majority of cities pursue some redevelopment activity. 85 California cities achieve much of their new growth through annexation. Because of conflicts over the distribution of revenues from the annexed areas, annexation has been a major source of friction between cities and counties in recent years. Well over half (58 percent) of respondent cities plan to annex new territory within the next five years, as Table 5.4 indicates. Central Valley cities are the most likely to have annexation plans; nearly all (87 percent) plan to annex, in general, between one and five square miles within the next five years. Los Angeles area cities are the least likely to be planning annexations; less than half have such intentions. Our analysis of attitudes about development based on the survey results is limited to responses from cities that could potentially be pursuing the development activity in question. So, for questions related to new development on vacant land, we analyze responses only from cities that have at least some vacant land available for development. For those questions relating to redevelopment activity, we analyze responses only from cities that are pursuing redevelopment at least somewhat actively. In relation to annexation of new territory, we analyze only those responses from cities planning to annex some land within the next five years. Introduction to Survey Findings The survey of city managers reveals that city governments in California have a fairly consistent set of priorities in regard to development, whether aimed at vacant land, designated redevelopment zones, or newly annexed territory. City policymakers appear to adopt a set of general priorities relating to development that are applied to all types of development in the city. However, development priorities do 86 vary somewhat among cities depending on their location and the level of development activity that is under way. In spite of these distinctions, most cities across all regions agree on one point: Retail development is the preferred land use overall, and maximizing sales tax revenue is the main motivation. Land-Use Preferences Retail projects are the land use most preferred by city governments in California for both new development projects on vacant land and city redevelopment projects, according to our survey respondents. The overall preference for retail development is evident in Figure 5.1, which shows average “desirability scores” for various types of land uses. In fact, three-quarters of cities surveyed rank retail as their highest preference (or Average desirability score 7 6 New development Redevelopment 5 4 3 2 1 Retail Office Mixed- Light Single- Multi- Heavy use industrial family family industrial development residential residential SOURCE: PPIC City Manager Survey, 1998. Figure 5.1—Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers 87 tied for first) for development projects on vacant land, and the share is even higher (80 percent) for redevelopment projects. Office, mixed-use, and light industrial development are also considered quite desirable, as Figure 5.1 indicates. The least favored land uses overall are multifamily housing and heavy industry. Cities across all regions in California prefer retail development most of all. However, in regard to other types of land uses there are distinct attitudes by region. For example, Central Valley cities prize industrial development, especially light industrial, more than cities in other regions in the state. San Francisco Bay area cities, on the other hand, prefer mixed-use development more than cities in other regions. Within regions, land-use preferences are quite similar whether development is considered in relation to new projects on vacant land or redevelopment projects in designated blighted areas. Average desirability scores for a particular land use tend to vary by less than half a point (on a seven-point scale) when comparing new development with redevelopment by region, with one exception: In all regions, singlefamily residential development is considered less desirable for redevelopment areas than for new development projects on vacant land. Factors Influencing City Development Decisions Why do California cities consistently favor retail development over other land uses? Our survey results indicate that maximizing sales tax revenue is the prime motivation guiding development decisions made by city administrations in the state. Table 5.5 lists the average importance scores of 18 motivations for development and redevelopment decisions as ranked by our survey respondents. As the table indicates, maximizing new sales tax revenue is the top consideration overall. Individually, 72 88 Table 5.5 Factors Influencing Development and Redevelopment Decisions Considerations/Motivations New sales tax revenue generated City Council support Eradication of blight Adequacy of infrastructure in project area Likelihood of job creation Cost of municipal services for new development Traffic and other spillovers Conformity with city’s general plan Acceptability of proposal to nearby neighborhoods Project aesthetics, urban design issues New property tax revenue generated Environmental considerations New fee/assessment/enterprise revenue generated Contribution to sound regional economy Support from Chamber of Commerce or other local business interests Meeting affordable housing needs Competition from nearby cities Preservation of agricultural land Nearby cities’ views Views of other local governments Average Importance Score New Development Redevelopment 6.5 6.4 6.3 6.4 n/a 6.2 6.1 5.8 6.0 5.9 5.9 5.5 5.8 5.8 5.7 5.7 5.7 5.7 5.6 5.9 5.4 6.1 5.4 5.4 5.0 4.9 4.8 4.8 4.7 4.8 4.3 4.8 4.3 4.1 3.7 n/a 3.0 n/a n/a 4.0 NOTE: n/a indicates that question was not asked. SOURCE: PPIC City Managers Survey, 1998. percent of cities surveyed rank it as their prime motivation (or tied as the top motivation) in decisions about development on vacant land. Twothirds consider it the prime motivation (or tied as the top motivation) in decisions about redevelopment projects. City council support for development projects is considered almost as important as generating new sales tax revenue in influencing decisions about new development. In the case of redevelopment projects, city council support is actually tied with sales tax revenue as the top 89 motivation. The importance of city council support is not surprising, given that they approve major development projects and hire and fire city managers. But it is perhaps surprising that any other factor should equal city council support as an influence in development decisions. Maximizing sales tax revenue also outranks job creation, service delivery, and infrastructure considerations as a motivation in both development and redevelopment decisions. New sales tax revenue outranks even the main professed goal of redevelopment law—the eradication of blight. These attitudes vary somewhat by region. San Francisco Bay area cities are exceptional in that generating new sales tax revenue is not their top concern overall. Rather, city council support is their foremost priority for both development and redevelopment projects. For new development only, San Francisco Bay area respondents also consider spillovers such as traffic congestion to be more important than generating new sales tax revenue. Central Valley cities, in contrast, tend to emphasize the importance of promoting new job growth. For redevelopment decisions, job growth is tied with sales tax revenue as the top motivation of Central Valley cities, whereas for new development, it follows closely behind. Respondent cities’ priorities are strikingly consistent when new development is compared with redevelopment. The correlation of “importance scores” for 15 out of 16 motivations for development decisions, when comparing factors influencing new development with factors influencing redevelopment, is at or above 0.68. One motivation—generating new property tax revenue—is not correlated as highly, because cities consider it a more important motivation for redevelopment than for new development projects. Through taxincrement financing, cities are able to retain a larger portion of property 90 tax revenue gains in designated redevelopment areas than they can retain from new development on vacant land. Regional Variation in Attitudes About Development Although individual cities adopt consistent attitudes toward different types of development, there are distinct regional patterns in the factors that motivate development decisions in California. These regional patterns affect attitudes about sales tax revenue and help to explain why cities in some parts of the state are less actively engaged in the fiscalization of land use. Table 5.6 lists average importance scores for different development motivations, by region. It reveals that certain motivations are ranked at about the same level of importance by cities in all regions in the state. These include the cost of municipal services for development projects and the adequacy of infrastructure in the project area. However, there is greater regional variation in relation to other motivations. Some fiscal or “pragmatic” concerns are considered more important in rural and less-developed parts of the state. These factors include generating new sales tax revenue, promoting new job growth, and contributing to a sound regional economy. These three factors receive their highest ratings from cities in the Central Valley and other mostly rural portions of the state and are considered less important in the San Francisco Bay area. Another set of factors shows the opposite tendency. These political and “quality of life” concerns are judged to be more important by cities in the more heavily developed parts of the state, most of all in the San 91 Table 5.6 Importance Scores for Development Motivations, by Region New Development Redevelopment Factors Influencing Development Decisions SF Bay L.A. Cent. SF Bay L.A. Cent. Area Area Val. Other Area Area Val. Other Less regional variation (0.4 points or less difference between scores, by region) Adequacy of infrastructure in project area 6.1 5.9 6.1 Cost of municipal services for development 5.8 5.9 5.8 Conformity with city’s general plan 5.9 5.9 5.5 New property tax revenue generated 5.3 5.5 5.6 New fee/assessment/enterprise revenue generated 4.9 4.9 4.9 Support from Chamber of Commerce or business 4.6 4.7 4.5 Eradication of blight n/a n/a n/a Views of other local governments n/a n/a n/a 6.2 6.0 5.6 5.3 5.3 4.8 n/a n/a 5.8 5.7 5.8 5.3 5.6 5.5 5.8 5.7 5.6 6.2 5.9 6.2 4.9 4.9 4.7 4.9 4.9 4.6 6.0 6.3 6.2 4.0 4.1 3.9 5.9 5.5 5.8 6.1 5.0 4.9 6.3 4.0 More regional variation (0.5 points or greater difference between scores, by region, for development or redevelopment) Factors more important to cities in the San Francisco Bay or Los Angeles areas City Council support 6.6 6.4 6.2 6.2 6.7 6.4 6.2 Traffic and other spillovers 6.3 5.8 5.4 5.8 6.1 5.8 5.4 Acceptability of proposal to nearby neighborhoods 6.1 5.9 5.5 5.6 6.0 5.8 5.3 Project aesthetics, urban design issues 5.9 5.6 5.3 5.6 6.2 5.9 5.5 Environmental considerations 5.7 5.3 5.2 5.4 5.7 5.3 5.0 Meeting affordable housing needs 5.0 4.0 4.2 4.4 4.9 4.6 4.9 Competition from nearby cities 4.1 4.7 4.1 4.1 3.8 4.4 3.8 Nearby cities’ views 3.2 3.5 2.6 2.7 n/a n/a n/a 6.4 5.9 5.6 5.8 5.6 5.0 4.0 n/a Factors more important to cities in the Central Valley or “other” areas New sales tax revenue generated 6.2 6.5 6.6 6.7 6.3 6.4 6.5 Likelihood of job creation 5.3 5.7 6.5 6.2 5.3 5.9 6.5 Contribution to sound regional economy 4.3 4.6 5.1 5.2 4.4 4.7 4.9 Preservation of agricultural land 3.8 2.9 4.6 3.6 n/a n/a n/a 6.5 6.1 5.3 n/a NOTE: n/a indicates that question was not asked. SOURCE: PPIC City Managers Survey, 1998. 92 Francisco Bay area. These factors include city council support, spillovers such as traffic congestion, aesthetics and other design issues, and the acceptability of development proposals to nearby neighborhoods. Another motivation—competition with nearby cities—is distinctive in that Los Angeles area cities consider it more important than cities in other regions of the state. These regional patterns help to explain why generating new sales tax revenue is not the most important consideration for San Francisco Bay area cities, in contrast to cities in the rest of the state. Further investigation reveals that the attitudes of Bay area cities vary considerably depending on their level of development activity. The attitudes of Bay area cities that are more actively engaged in new development or redevelopment are closer to attitudes in the rest of the state. In contrast, attitudes among Bay area cities that are less actively pursuing development or redevelopment are quite distinct. San Francisco Bay area cities, regardless of their level of development or redevelopment activity, consider several “quality of life” and political factors to be more important in guiding development decisions than do cities elsewhere in the state. These motivations include city council support for development projects, traffic congestion and other negative spillovers, and project aesthetics and other urban design issues. However, when it comes to “pragmatic” factors, such as sales tax and property tax revenue, infrastructure and service considerations, and promoting new job growth, Bay area cities with “considerable” amounts of vacant land tend to rate these factors higher in importance than Bay area cities with “limited” amounts of vacant land. Most of these distinctions are also evident in regard to redevelopment, although less dramatically. Bay area cities that are “very actively” engaged in 93 redevelopment rate new sales and property tax revenue and promoting job growth as more important considerations than do Bay area cities that are “not very actively” engaged in redevelopment projects. Sales tax revenue is considered very important among San Francisco Bay area cities with a “considerable” amount of vacant land left to develop. Not only do these cities rank sales tax revenue as the most important consideration in relation to new development, they rate it as more important on average than any other group in the state, when measured by region and development level. In relation to redevelopment decisions, only the support of the city council is considered more important than new sales tax revenue for cities that are “very actively” engaged. In sharp contrast, Bay area cities with a “limited” amount of vacant land left to develop and those that are “not very actively” engaged in redevelopment consider a number of motivations, mostly “quality of life” factors, to be of equal or greater importance than generating new sales tax revenue. In considering this regional distinctiveness, keep in mind that San Francisco Bay area cities tend to be more “built out” than cities elsewhere in the state. That is, relatively fewer Bay area cities still have vacant land available for development than cities in other regions. In addition, relatively fewer Bay area cities are engaged in redevelopment than cities in the Los Angeles area or the Central Valley, and a higher share have elected not to pursue annexation. One could surmise from these findings that certain Bay area cities are less actively pursuing sales tax revenue (and thereby fiscalizing their land use decisions) because they are more fully developed, and therefore do not have the same opportunities to alter land-use patterns. They may be more concerned about the negative 94 consequences of past growth and therefore also more concerned about ensuring adequate planning for future growth. Factors Influencing Annexation Decisions When it comes to the factors that guide city administrations in decisions to annex new territory, maximizing sales tax revenue is not the top motivation. Instead, controlling the development of surrounding areas with the intention of ensuring consistency with city plans is the most important consideration, according to our survey respondents. However, generating new sales tax revenue follows closely as the second most important consideration, as Table 5.7 indicates. Table 5.7 Factors Influencing Annexation Decisions, by Region Factors Influencing Annexation Decisions To control development of surrounding areas To gain sales tax revenue To create jobs More efficient service provision To gain property tax revenue Agreements with county To provide greenbelt or open space To gain fee/assessment/enterprise revenue To meet housing needs Direction from LAFCO Agreements with other cities To prevent annexations by other cities Importance Scores, by Region Los San Angeles Francisco Central Area Bay Area Valley Other 5.9 6.0 5.8 5.3 5.4 4.4 6.0 5.6 4.8 3.9 6.0 5.3 5.2 4.0 5.3 4.5 4.7 4.1 4.9 4.8 4.7 4.0 4.8 4.3 4.6 5.2 4.2 3.7 4.2 3.5 4.7 4.4 3.8 3.6 4.4 3.8 4.4 3.6 3.9 3.5 3.6 3.3 3.1 2.6 3.2 3.2 2.5 2.0 SOURCE: PPIC City Managers Survey, 1998. All Cities 5.7 5.4 5.1 4.8 4.7 4.5 4.4 4.3 3.9 3.9 3.1 2.7 95 This overall trend masks certain distinctions by region and level of annexation activity. Controlling the development of surrounding areas is considered more important than generating new sales tax revenue in the more urbanized areas of the state, specifically in the Los Angeles and San Francisco Bay areas. It is also considered more important to certain Central Valley cities—those with the most ambitious annexation plans, namely, plans to annex one or more square miles of territory within the next five years. However, for Central Valley cities overall, and for cities in other mostly rural parts of the state, sales tax revenue is considered more important. San Francisco Bay area cities are unique in that providing a greenbelt or open space is ranked higher than generating sales tax revenue as a motivation for annexation. In contrast, Central Valley cities are unique in that future job growth is considered equally as important as gaining new sales tax revenue when it comes to annexation decisions. Do City Governments Act on Their Development Preferences? Just because city governments in California desire retail development above other land uses does not mean that they are in a position to get it. However, the survey data indicate that they are at least prepared to try. These results confirm the common contention that local land-use decisionmaking in California tends to be biased in favor of retail development. Retail was ranked by survey respondents as the type of land use for which they would be most likely to provide a general plan change or a financial incentive to a developer of new projects on vacant land. On a scale from one to seven, where one meant they were “very unlikely” and 96 seven meant they were “very likely” to provide an incentive for that type of development, more than two-thirds of respondents gave retail development a score above four. More than half of respondents gave retail development a score of six or seven, which can be taken to mean they would be quite likely to offer such an inducement. As with land-use preferences and motivations for development, certain regional distinctions underlie this general finding. The regional distinctions are completely consistent with previous findings. Central Valley cities are more likely to offer an incentive for light industrial development than for any other type of land use, whereas certain San Francisco Bay area cities, namely, those with a “limited” amount of vacant land, are most willing to offer an incentive for mixed-use development. Retail development follows closely behind in both cases, however. Which Types of Cities Favor Retail Development the Most? The preceding discussion described general trends in attitudes about development by region and level of development activity. However, we cannot decipher from those results whether there are certain characteristics of cities—characteristics that may vary by region—that could help to explain development priorities. In this section we explore more systematically whether any particular characteristics of cities are associated with more favorable attitudes toward retail development. We estimate statistical models to help account for the variations between different cities in California in their attitudes about retail development and generating new sales tax revenue. 97 Interest in Retail/Sales Tax: Dependent Variables We test 12 dependent variables in turn. Six of them are the actual desirability or importance scores—rated on a scale ranging from one to seven—that each survey respondent gave to retail development and sales tax revenue. These scores measure the absolute importance of retail development and sales tax revenue according to a particular respondent. The other six are variations intended to gauge the relative importance to each survey respondent of retail development and sales tax revenue as compared with other options for land use and other development motivations. These scores are constructed as the difference between each respondent’s desirability score for retail (or his/her importance score for maximizing sales tax revenue), and the average desirability score for other land use options (or the average importance score for other development motivations). The dependent variables are listed in Table 5.8. City Characteristics: Independent Variables In examining which city characteristics are related to heavy interest in retail development and sales taxes, we tested most of the independent variables described in Chapter 3. We used ordinary least-squares regression to assess whether they are associated with the 12 dependent variables described above. The independent variables employed in the model are listed and described in Table 5.9. Results of Model Estimation Table 5.10 presents the signs and statistical significance levels for the variables in the regression models. As the table indicates, only one independent variable proves to be significant in almost all of the 98 Table 5.8 Dependent Variables in the Regression Model 1. The desirability score of retail for new development projects on vacant land. 2. The difference between #1 and the average desirability score for six other types of land uses (besides retail) for new development projects on vacant land. 3. The desirability score of retail for redevelopment project areas. 4. The difference between #3 and the average desirability score for six other types of land uses (besides retail) for redevelopment project areas. 5. The importance score of maximizing new sales tax revenue as a consideration in attracting new development and responding to development proposals. 6. The difference between #5 and the average importance score for 15 other motivations (besides maximizing new sales tax revenue) for new development projects on vacant land. 7. The importance score of maximizing new sales tax revenue as a consideration in evaluating redevelopment project proposals. 8. The difference between #7 and the average importance score for 15 other motivations (besides maximizing new sales tax revenue) for redevelopment project decisions. 9. The importance score of maximizing new sales tax revenue as a consideration in decisions to annex new territory. 10. The difference between #9 and the average importance score for 11 other motivations (besides maximizing new sales tax revenue) for annexation decisions. 11. The likelihood that the respondent’s city would provide a general plan change (rezoning) or a financial incentive to the developer of a retail project on vacant land. 12. The difference between #11 and the average likelihood score for six other types of land use (besides retail). 99 Table 5.9 Independent Variables in the Regression Model 1. Bay area city: dummy variable for San Francisco Bay area location. 2. L. A. area city: dummy variable for Los Angeles area location. 3. Central Valley city: dummy variable for Central Valley location. 4. Development level: dummy coded 1 for cities with “considerable” amounts of vacant land left to develop, 0 for all other cities. 5. Redevelopment level: dummy coded 1 for cities “very actively” involved in redevelopment projects, 0 for all other cities. 6. Population density: the natural logarithm of city population per square mile in 1990, according to the U.S. Census. This was included in place of population and area because of the evident interaction between them. 7. Tourism: the natural logarithm of a county-level estimate of 1992 travel-related expenditures, from the Division of Tourism. 8. Median household income: median household income in 1989, from the U.S. Census. 9. Median household income squared: variable #8, squared. 10. Household size: average household size in 1990, from the U.S. Census. 11. Black population share: the percentage of blacks in the population in 1990, from the U.S. Census. 12. Hispanic population share: the percentage of Hispanics in the population in 1990, from the U.S. Census. 13. Property tax share: the percentage of own-source general revenue made up of property tax revenue in fiscal year 1992–93, from the Department of Finance. 14. Sales tax share: the percentage of own-source revenue made up of sales tax revenue in fiscal year 1992–93, from the Department of Finance. 15. Urbanized: a dummy variable to indicate whether the city was included as part of an urbanized area in 1990, from the U.S. Census. 100 Table 5.9 (continued) 16. Central city: a dummy variable to indicate whether the city was considered a core city in a metropolitan area in 1990, from the U.S. Census. 17. Population growth: percentage change in population, 1991 to 1998, from the Department of Finance. 18. Age of housing: median age of the housing stock as of 1990, according to the U.S. Census. 19. Isolated: a dummy variable indicating that the city has no neighboring cities within five miles of its boundary. 20. Neighboring cities: number of neighboring cities within five miles of city boundaries. 21. Interstates: a dummy variable indicating that one or more Interstate highways fall within city boundaries. 22. Own-source revenue: the natural logarithm of per capita general-purpose ownsource revenue (total revenue minus intergovernmental transfers minus public service enterprise revenue), from the California State Controller. 23. Years incorporated: the number of years since the city incorporated. 24. Annexation level: dummies for the number of square miles the city plans to annex & within the next five years—one to five square miles or more than five square miles. 25. These are considered in relation to a reference category of cities planning to annex less than one square mile. (These variables apply only to regressions relating to annexation.) regressions. This variable is central city, which is negatively associated with greater desire for retail development or sales tax revenue in 10 of the 12 regressions. Two other variables are significantly associated with the sales tax/retail orientation about half the time (in five or six regressions). These are household size, which is positively associated, and urbanized, which is positively associated with desire for retail for new development and redevelopment projects but negatively associated with desire for retail development for newly annexed territory. 101 Table 5.10 Results of Regression Models: City Characteristics Associated with Preference for Retail Development Dependent Variable 1. Absolute score: retail for new development 2. Relative score: retail for new development 3. Absolute score: retail for redevelopment 4. Relative score: retail for redevelopment 5. Absolute score: sales tax for new development 6. Relative score: sales tax for new development Independent Variables with Significant Association Development level Household size Hispanic population share Central city Central Valley city Hispanic population share Urbanized Central city Redevelopment level Tourism Household size Hispanic population share Sales tax share Urbanized Central city Population growth Own-source revenue Years incorporated Central Valley city Black population share Sales tax share Urbanized Central city Bay area city LA area city Median household income Median household income squared Household size Property tax share Central city Bay area city LA area city Median household income Sign and Signif. of Relationship +** +* –* –*** –* –* +* –*** +* –** +** –** –*** +*** –*** –* –** +* –* –** –** +*** –** –** –** +*** –*** +** –* –*** –** –** +** Model No. of Adj. R- ObserSquared vations 0.10 208 0.12 208 0.11 228 0.17 228 0.22 209 0.15 209 102 Table 5.10 (continued) Dependent Variable 7. Absolute score: sales tax for redevelopment Independent Variables with Significant Associations Median household income squared Household size Property tax share Central city Isolated Tourism Central city Sign and Signif. of Relationship –*** +** –** –*** +* –* –*** Model No. of Adj. R- ObserSquared vations 0.05 230 8. Relative score: sales tax for redevelopment Central city Isolated –*** 0.05 230 +* 9. Absolute score: sales tax for annexation Household size Hispanic population share Property tax share Urbanized Annexation level: five miles or more 10. Relative score: sales tax for annexation Bay area city Household size Urbanized Years incorporated 11. Absolute score: incentive for retail Median household income Median household income squared Household size Property tax share Central city Population growth Own-source revenue 12. Relative score: Central city incentive for retail Age of housing NOTE: *p < .1, **p < .05, ***p < .01. +*** –* –* –** +** –* +** –** +* +* –** +* –* –** –** –* –*** –* 0.30 175 0.19 175 0.15 206 0.09 206 103 Factors Related to Fiscalization: Interpreting the Results The poor overall fit of the model (the adjusted R-square is less than 0.3 for all regressions but one) indicates that there is no very distinct pattern of characteristics that helps predict which cities prefer retail development the most. Although certain trends are apparent, they cannot be called decisive. It seems that the desire for retail development is so pervasive that it cuts across cities with very different characteristics. Cities in urbanized parts of the state are more interested in retail than cities in rural areas when it comes to new development and redevelopment projects. This may reflect the greater degree of competition for retail dollars that cities in urbanized areas are facing. Cities in rural parts of the state, on the other hand, are particularly interested in retail development for newly annexed territory. Rural cities are far more likely to have annexation plans in the first place, and they may tend to direct retail development to areas near highways or other roads near town. California’s 36 central cities prove to be the major exception when it comes to the general trend for urbanized communities to value retail most highly. Central cities, which have lost per capita sales tax revenue since the 1970s, do not appear to be alarmed by that trend; they are less interested in garnering new sales tax revenue than other cities in the state. Instead, central cities seem to be more interested in balancing their various land-use goals. Communities with larger average household sizes, which as we saw in Chapter 3 are less successful at generating per capita sales tax revenue, appear to be interested in reversing that trend. 104 Characterizing Regional Patterns in Development Priorities Regression analysis enables us to explore in greater depth the factors that underlie the regional patterns in attitudes about development described above. We are able to assess whether certain characteristics of cities in California—characteristics that vary by region—can help explain regional distinctions in development priorities. We noted above that maximizing sales tax revenue is considered somewhat more important to cities in the Central Valley and other mostly rural parts of the state than in the San Francisco Bay and Los Angeles areas. Likewise we noted that certain “quality of life” factors are judged to be more important by cities in the more heavily developed parts of the state. These counteracting trends help to explain why certain San Francisco Bay area cities, in contrast to other cities in the state, place a number of priorities above maximizing sales tax revenue when making development decisions. In this section we investigate whether any particular traits of California cities can help explain these patterns. First we employ a technique called factor analysis to delineate patterns in survey responses. We identify two main “factors,” or patterns, that characterize responses to the questions regarding the various motivations for development and redevelopment decisions. These patterns are surprisingly similar in relation to new development and redevelopment. In both cases, the first pattern involves high importance scores (orthogonally rotated factor loadings above 0.5) assigned to certain planning and “quality of life” motivations, namely, urban design issues, traffic and other spillovers, the conformity of development projects with the city’s general plan, and environmental considerations. The second pattern involves high importance scores 105 assigned to certain “pragmatic” motivations, namely, generating new sales tax revenue, property tax revenue, and revenue from fees and assessments. The regional distinctions that we noted above are confirmed by the factor analysis. San Francisco Bay area cities tend to score highest on the “quality of life” factor, and Central Valley cities score lowest on average. Cities located in the Central Valley and other mostly rural parts of the state tend to score highest on the “pragmatic,” revenue-oriented factor. We estimated ordinary least squares regressions to test which of the independent variables employed above are associated with each city’s factor scores. In this way we can determine which characteristics define cities that follow each of the two main patterns of development priorities. The results of these regressions are displayed in Table 5.11. Although the results are again somewhat murky, if we look only to the most consistent results we can conclude that “quality of life” cities—or cities that score high on the first factor—are characterized especially by smaller household sizes and larger proportions of Hispanics in the population. Cities that follow the “pragmatic” pattern—or cities with high scores on the second factor—tend not to be central cities and to earn fewer tourism dollars. Summary One main assumption underlying the common critique that cities in California have fiscalized land-use decisions is that they bias development decisions in favor of retail projects. Our survey results provide strong empirical confirmation for this assertion. Respondents from the vast majority of cities across the state indicate that their administrations favor retail development above other land uses for both new development and 106 Table 5.11 Results of Regression Models: City Characteristics Associated with Two Patterns of Development Priorities Dependent Independent Variables with Variable Significant Association 1. Score on quality Tourism of life factor for Median household income new development Household size Hispanic population share Isolated Sign and Signif. of Relationship –** +** –** +** –* Model Adj. RSquared 0.15 No. of Observations 201 2. Score on quality of life factor for redevelopment Household size Hispanic population share Interstates –*** +* –* 0.13 226 3. Score on prag- Bay area city matic factor for Los Angeles area city new development Redevelopment level Tourism Median household income Median household income squared Central city Isolated –* –* +** –* +*** –*** –** +* 0.18 201 4. Score on prag- Tourism matic factor for Sales tax share new development Central city NOTE: *p < .1, **p < .05, ***p < .01. –* 0.03 226 –* –** redevelopment, and they do this to maximize new sales tax revenue. However, office, light industrial, and mixed-use development are also generally considered attractive. Although some cities that tend especially to be located in the San Francisco Bay area place certain quality of life considerations above maximizing sales tax revenue, these cities are the exception rather than 107 the rule. The state’s 36 central cities also have more varied motivations for their development decisions. Otherwise, however, few characteristics can be said to distinguish cities that are more actively pursuing retail development. Instead, the practice is so common that it can be called nearly ubiquitous. 108 6. Policy Considerations In this chapter, we pull together the findings of the study and discuss options for changing the distribution of the local sales tax. As we will show, although there are certainly reasons to criticize the situs-based sales tax, other options for distributing the tax have drawbacks as well. Policymakers concerned about the land-use and distributional consequences of the situs-based sales tax may instead wish to consider more broad-ranging reforms of local public finance and the state/local fiscal relationship. Summarizing the Results The Bradley-Burns sales tax is an important source of funds for California cities. Although not a huge share of local revenues, the local sales tax is important because it may be used in a discretionary way by cities, as opposed to being earmarked for a specific purpose. And unlike the property tax, it has the potential for substantial revenue growth if 109 cities make supportive land-use decisions. Or rather, cities tend to behave as if the sales tax is subject to major growth. As we have seen, however, the sales tax has been a steady or somewhat stagnant source of revenue in inflation-adjusted, per capita terms. This is due to the relatively fixed nature of retail spending per capita, the increasing share of consumer spending going to transactions that are not covered by the sales tax, and the many exemptions that state law provides for sales taxation. This stagnant revenue trend conceals a great degree of variation among cities in the degree to which they benefit from the tax. The overall hierarchy of cities, in terms of their sales tax revenue “success,” has been fairly stable over time; most cities, whatever their efforts in recruiting retail, have not gained or lost a great deal relative to their competitors. In spite of the meager overall prospects for growth in real per capita sales tax revenues, city governments have often been quite entrepreneurial at pursuing sales-tax-generating developments. We can conclude from the survey results in Chapter 5 that city officials do favor retail development over other land uses and are motivated by sales tax concerns as strongly as any other factor. Nevertheless, office, light industrial, and mixed-use projects also are viewed quite favorably, and most cities do not appear one-dimensional in their growth priorities— although housing is rarely a preferred land use. To the extent that cities are competing for retail—and many are prepared to offer concessions to get it—they are fighting over a relatively fixed pie, at least in per capita terms. Retailers, given the locational constraints of the market areas they operate in, are probably likely to locate in certain types of settings, concessions or not. Thus, it is not a 110 great leap to conclude that the major effect of fiscalized land-use decisionmaking is a transfer of resources from local government to retailers, developers, and landowners. With all of the favorable attention that cities show to retail, it is also likely that residential and industrial development are somewhat more difficult and more expensive to develop than would be the case in the absence of a situs-based sales tax. These land-use issues, along with the vast disparities in sales tax revenues that exist among cities, often lead reformers to urge a change in the situs basis for sales tax distribution. In the remainder of this chapter, we explore this issue. Policy Debates over Sales Tax Revenue Sharing A Los Angeles Times columnist has complained about the “absurdity” of sales tax revenues going only to the jurisdiction in which a sale takes place, arguing that, carried to a logical extreme, loyal local residents should limit all their purchases to stores in their home towns. “That way lies madness—and an undermining of the common good and the larger community” (Flanagan, 1998). Examining this “absurd” situation, some have proposed sharing the sales tax among local governments. Such proposals have existed for at least 20 years (Senate Local Government Committee, 1989, p. B-5). Although local tax base sharing is rare in the United States, it is not unknown. Most national discussions of tax sharing center on the property tax, which is a much more important share of local revenues in most parts of the country than in California. The Twin Cities region of Minnesota has practiced a program of property tax base sharing since 1971. Under the Minnesota fiscal disparities system, 40 percent of the growth in property tax revenue from commercial and industrial 111 properties is placed in a pool that is shared among the cities, counties, and school districts of the state’s major metropolitan area. These shared revenues, which totaled $367 million in 1996, are distributed to jurisdictions that are below average in their commercial/industrial property wealth per capita. Orfield (1997, p. 65), an advocate of greater sharing in Minnesota, finds that “even after this reallocation, disparities in property wealth remain enormous.” Nevertheless, tax base sharing has succeeded in reducing tax base disparities among communities in the Twin Cities by a very significant degree (Ross and Levine, 1996, p. 333; Orfield, 1997, p. 87). California policymakers have periodically proposed tax-sharing arrangements for the local sales tax. In 1994, Assemblymember Valerie Brown introduced a bill that attempted to revise sales tax allocation within each county as a way to address the fiscalization of land use. Under this plan, although all cities and counties would be guaranteed at least their existing levels of sales tax proceeds, the growth in sales tax revenue within each county would be distributed among jurisdictions mainly on the basis of population. Supporters claimed that this approach would better match revenues to service needs and would stop jurisdictions from artificially favoring retail over residential and industrial growth. Opponents argued that the proposal would lead to cities being disinterested in recruiting new businesses, would reward anti-growth jurisdictions, and would lead to ill-considered residential sprawl (because counties, seeking capitated sales tax revenues, might seek population growth for their unincorporated areas). The passage of Proposition 11, on the ballot in the November 1998 general election, amended the California constitution so as to allow cities and counties to enter into revenue-sharing agreements regarding revenues 112 obtained under the Bradley-Burns local sales tax law. Unlike the popular vote requirement under previous provisions, this measure allows jurisdictions to enter into such contracts with a two-thirds vote of the city councils or county boards of supervisors involved. There has been movement toward such a sharing arrangement in recent months between Modesto and its surrounding county, Stanislaus (tied to Modesto’s annexation of county lands). Although isolated instances of such “bilateral disarmament” from the sales tax race will probably occur, overall it seems unlikely that Proposition 11 will greatly reduce sales tax competition, since “winner” jurisdictions will have few incentives to enter into such agreements. Arguments For and Against Moving Toward a Population-Based Distribution of Sales Tax Revenues A system for distributing the sales tax that is based more on local populations than on their position in the retail hierarchy does have serious arguments in its favor. However, one can also make a good case against replacing all or part of the situs-based system with a populationbased system. Let us consider some of the arguments. In favor of the population-based system, one can make the following points: • The population-based system would presumably better reflect the revenue needs of each jurisdiction. Most public services are provided to residents, not to retail businesses. Some of the very extreme current disparities in revenues under the situs-based approach are fairly astounding; recall that the range is from $2 to $56,892 per capita. Many would likely find these disparities as lacking serious policy justification. 113 • Awarding Bradley-Burns revenues based on population might lead local governments to view housing development in a new light. Such a reform would create an incentive to increase residential populations. At present, housing is seen as a drain on local finances, and most cities (as the survey results confirm) do not especially encourage its development. Since many areas of California have some of the least affordable housing in the country, making housing development more attractive to local governments seems like a worthwhile policy goal. • Moving away from the situs-based system would mean that cities would no longer have an incentive to “chase” retail development. Given that retail growth is not the firmest basis for overall economic development, and that retail location is something of a zero-sum competition among localities in any given region, this would presumably be better for the state economy. Transfers of resources from government to retailers and developers would be reduced, and locational distortions in the land market presumably would be reduced as well. These are all important considerations. However, strong arguments can also be made against the move to a population-based distribution of local sales tax revenues: • Cities, to some degree, function as competitive economic units within an open economy. Much of city politics is rightly geared at maximizing revenues and economic vitality (see Peterson, 1981). Taking funds away from city governments that have managed through hard work or entrepreneurial leadership to gain a sales tax advantage seems to punish cities for their success. • Retail-heavy cities have special needs for public revenues. In effect, cities that are retail centers perform a service for the rest of their regions. Cities with many stores have large “visitor” and “worker” populations, even if their residential population is far 114 lower. This daytime population surge requires public goods and services such as larger roads, more trash collection, and more sophisticated police and fire protection than in equally sized towns that lack stores. Cities that are tourist destinations are special examples of this phenomenon, with daytime populations often many times larger than residential populations. In particular, the presence of stores and shopping generates high levels of public safety demands. As Figure 6.1 illustrates, retailheavy cities (as measured by their sales tax “success”) clearly have higher crime rates, on average, than other cities. The situs system compensates them for these externalities of retail.1 15,000 FBI Uniform Crime Index (1990) 12,000 9,000 6,000 3,000 0 0123456 Log of adjusted per capita sales tax revenues (average of 1991, 1992, 1993) SOURCES: FBI Uniform Crime Index; California State Controller, Financial Transactions Concerning Cities of California (annual). 7 Figure 6.1—Relationship Between Retail Concentrations and Crime in California Cities ____________ 1The situs system does not, however, compensate nearby communities for the additional traffic and other effects that they must wrestle with. 115 • As we showed in Chapters 2 and 3, many of the “least successful” cities in per capita sales tax revenues are very highincome residential communities, such as Hillsborough and Rolling Hills Estates. Switching toward a more populationbased system would seem to have the perverse effect of rewarding high-status communities for their “not in my backyard” approach to commercial development.2 (We will have more to say about the distributional effects of a population-based sales tax distribution in the next section.) • Finally, a broader critique of moving away from the situs rule toward a population-based approach is that it removes the one major incentive that California cities now have to pursue growth and development. One might argue that the way toward more balanced land uses would be to create a local revenue system that provides incentives for other types of development—not to simply make retail as disfavored as residential and industrial development. In short, scrapping the situs-based system would not solve the broader incentive problems with California’s system of local public finance.3 ____________ 2Orfield (1997, p. 87) sees a similar problem with Minnesota’s system for sharing commercial/industrial property tax revenues, in that wealthy residential enclave suburbs have been net recipients of revenues from the tax-sharing pool. 3An additional concern about shifting away from a situs basis for local sales taxes reflects legal considerations. Courts generally have held that there must be a “nexus” between the jurisdiction levying a tax and the activity that is subject to the tax. In the case of the local sales tax, as currently practiced, local governments levy a sales tax on an activity occurring within their boundaries—retailers must remit the tax revenues to the jurisdiction where the sale takes place. If, by contrast, taxes were locally levied but subject to sharing across jurisdictional boundaries, courts might consider the nexus requirement violated. One way around this problem would be to make the local sales tax a state-levied tax that is then distributed to localities in a revenue-sharing arrangement. This approach, however, would place local governments in the position of claimants on a state-generated subvention, which would leave many home-rule advocates unhappy. In theory, the nexus requirement could be satisfied by having a local tax levied upon the consumer according to his or her jurisdiction of residence, rather than one collected by the retailer based upon point of sale. This approach, however, can be dismissed as administratively unworkable. 116 Which Local Governments Would Benefit from a Population-Based Distribution? To further explore the distributional consequences of moving away from the situs system toward a population-based system for distributing local sales tax revenues, we examined such a scenario more systematically. Using 1994 data, the most recent year for which complete city financial data were available,4 we performed two types of analyses. In both, we asked the following question: What if the situs system were completely abolished and the Bradley-Burns revenues were instead distributed solely on the basis of population—city population in the case of cities and unincorporated-area population in the case of counties? It is very unlikely that reformers would ever succeed in suddenly abolishing the situs system, but making this strong assumption allows us to evaluate the general pattern of revenue gains and losses among jurisdictions. If policymakers decided instead to base revenues half on population and half on the location of the sale, the same pattern of gainers and losers would emerge, but the immediate gains and losses would simply be half the magnitude as our results show here. In the first analysis, we assume that cities and counties receive a share of statewide Bradley-Burns revenues based on their relative share of statewide population. In the second analysis, which is closer to some of the reforms that have been proposed in the past, we assume that BradleyBurns revenues remain within the county where the sale took place. In this analysis, each jurisdiction receives a share of the countywide total of local sales tax revenues based on its share of the countywide population. ____________ 4Several cities failed to submit financial records to the California State Controller for fiscal years 1995 and 1996. 117 In each of these two analyses, our basic questions were the same: How many jurisdictions would be better or worse off, in terms of their per capita sales tax revenues, compared to the amount they actually received in 1994? What share of the state’s population lives in jurisdictions that would be better off? What are the demographic characteristics of the potential gainer and loser cities under each of these plans? Answering these questions should help policymakers consider the politics of sales tax reform as well as the distributional issues involved. Statewide Redistribution As shown in Table 6.1, under a statewide redistribution of Bradley- Burns revenues in 1994, 55.5 percent of the 470 cities then in existence would have received more revenue than they did under the situs-based system. However, these “gainer” cities are smaller in population on average than those that would have lost revenue under the population- Table 6.1 Number of Jurisdictions That Would Be Better or Worse Off Under a Statewide, Population-Based Distribution of the Local Sales Tax, 1994 Type of Jurisdiction Cities that would be better off Cities that would be worse off No. of Jurisdictions 261 209 % of Total Total PopuCities or lation of Such Counties Jurisdictions 55.5 12,543,375 44.5 12,893,360 % of State Population 39.6 40.7 Counties that would be better off Counties that would be worse off 51 6 89.5 5,240,000 16.6 10.5 984,680 3.1 NOTES: Population for counties refers to unincorporated areas only. City and County of San Francisco is counted as a city. SOURCES: Authors’ calculations using data from California State Controller, Financial Transactions Concerning Cities of California (1994); annual city population estimates from the California Department of Finance, Demographic Research Unit. 118 based scenario. Thus, the overall population of cities that stood to gain was slightly less than those that stood to lose. Among counties, the vast majority (51 out of 57, not counting San Francisco) would have improved their fiscal position under the population-based system. The gainer counties had more than five times the unincorporated-area population as the counties that stood to lose. Overall, counting both cities and unincorporated areas, 56.2 percent of the state’s population lived in jurisdictions that stood to gain from a statewide reallocation of the Bradley-Burns tax. Focusing strictly on cities, Table 6.2 provides a portrait of the characteristics of cities that would gain and lose under this scenario. As one can see, cities in the Central Valley and Los Angeles area would disproportionately gain, whereas those in the Bay area and the remainder of the state would tend to lose. Nonurbanized cities would be especially Table 6.2 Characteristics of Cities Gaining and Losing Revenues Under a Statewide, Population-Based Distribution of the Local Sales Tax No. in Los Angeles area No. in San Francisco Bay area No. in Central Valley No. in rest of state No. urbanized No. nonurbanized No. central cities Mean per capita income, 1989 Mean household size, 1990 Mean % black, 1990 Mean % Hispanic, 1990 Mean % senior citizens, 1990 % of own-source local revenue from sales taxes, 1993 Gaining Cities 108 48 57 48 152 100 9 $17,090 3.0 4.3 28.5 11.0 18 SOURCES: U.S. Census (1990); authors’ calculations. Losing Cities 69 52 32 56 141 63 27 $17,854 2.6 3.4 19.9 13.0 28 119 likely to be better off, but three-quarters of the state’s 36 central cities would be losers. In terms of socioeconomic characteristics, the cities that would stand to gain tend to be slightly poorer, have higher household sizes, and higher proportions of Hispanics in their populations. These outcomes reflect the broadly redistributive nature of the statewide proposal. Within-County Redistribution If, instead, sales tax revenues were reallocated within each county on the basis of population, outcomes would be somewhat different, as shown in Table 6.3. In this case, a slim percentage majority of cities Table 6.3 Number of Jurisdictions That Would Be Better or Worse Off Under an Intracounty, Population-Based Distribution of the Local Sales Tax, 1994 Type of Jurisdiction Cities that would be better off Cities that would be worse off Cities with no change No. of Jurisdictions 240 229 1 % of Total Cities or Counties 51.1 48.7 0.2 Total Population of Such Jurisdictions 13,539,135 11,145,500 752,100 % of State Population 42.8 35.2 2.4 Counties that would be better off Counties that would be worse off Counties with no change 48 6 3 84.2 5,298,900 16.7 10.5 895,500 2.8 5.3 30,280 0.1 NOTES: Population for counties refers to unincorporated areas only. The City and County of San Francisco is counted as a city, and thus without a surrounding county would experience no change. The three counties with no change are those with no incorporated cities. SOURCES: Authors’ calculations using data from California State Controller, Financial Transactions Concerning Cities of California (1994); annual city population estimates from the California Department of Finance, Demographic Research Unit. 120 would have been better off in 1994 than under the situs approach. However, the gainer cities in this case have higher populations. Among counties, 48 would have received more sales tax revenues, six less, and three without any incorporated cities would have experienced no change. Overall, under the countywide population-based distribution of the Bradley-Burns tax, 59.5 percent of the state’s population lived in units that stood to gain, 38 percent in jurisdictions that would have lost revenues, and 2.5 percent in jurisdictions that would have experienced no change. Thus, it would appear that there would be more potential political weight behind the countywide redistribution than the statewide approach. Table 6.4 shows, moreover, that the intracounty redistribution of sales tax revenues would benefit a majority of cities in the Los Angeles Table 6.4 Characteristics of Cities Gaining and Losing Revenues Under an Intracounty, Population-Based Distribution of the Local Sales Tax No. in Los Angeles area No. in San Francisco Bay area No. in Central Valley No. in rest of state No. urbanized No. nonurbanized No. central cities Mean per capita income, 1989 Mean household size, 1990 Mean % black, 1990 Mean % Hispanic, 1990 Mean % senior citizens, 1990 % of own-source local revenue from sales taxes, 1993 Gaining Cities 99 55 48 38 152 79 11 $17,993 3.0 4.4 28.3 10.5 17 SOURCES: U.S. Census (1990); authors’ calculations. Losing Cities 78 44 41 66 140 84 24 $16,843 2.7 3.3 21.0 13.2 28 121 area, Bay area, and Central Valley alike. However, most cities in other parts of the state would lose, and urbanized communities would do better than nonurbanized ones. Cities that stood to gain under this proposal also tend to have higher per capita incomes than those that would lose (although the “winning” cities also had higher minority populations). Thus, this proposal would not necessarily advance socialequity goals. Broader Remedies: Balancing Fiscal Rewards for Growth in California Communities Thus, moving to redistribute the local sales tax—or the growth in that revenue source—on a population basis might, in a narrow sense, work to the fiscal advantage of the majority of jurisdictions, covering the majority of population in the state. But such a shift not only would be politically unpopular among the numerous local governments that stand to lose but would also fail to address the wider problem of making nonretail forms of development more fiscally viable. Perhaps the way around this dilemma is to widen our focus beyond the sales tax—which is, after all, merely one, relatively modest component of local government revenues. Rather, policymakers might do better to examine the overall context of local public finance. As it stands, retail is looked upon with favor not merely because it is associated with the sales tax but because other forms of development are perceived simply to not pay their own way. Cities often are particularly unenthusiastic about housing because they receive such a small slice of the property tax dollar. The city manager of Long Beach, for example, complained that a typical owner-occupied home produces only $210 per year in property tax revenue, but it costs the city $350 per year in 122 services. As a result, Long Beach has turned more attention toward retail development (Flanagan, 1998). Counties, which receive a larger share of the property tax in unincorporated areas and provide less in the way of municipal-type services than cities do, often tend to be more welcoming of residential growth. As a result, critics charge, “starter homes and other tax ‘losers’ have been relegated to distant locations on the metropolitan fringe, often in unincorporated areas” (Fulton, 1998, p. 1). Cities often have reasons to prefer retail to industrial or office development as well. Kotin (n.d.) claims that even attractive modern office campuses, with highly paid employees, are likely to produce a fiscal deficit in many cities with low shares of the property tax distribution in their area. He concludes, “A system which essentially prohibits taxation in relation to services to property or to employed population, combined with a system that limits changes in property tax assessments until a property changes hands, virtually dooms a city to spending more to accommodate new employees than it will receive in benefits.” As a result, one columnist has argued, “by focusing virtually all their attention on where wealth is spent, cities are ignoring the real source of economic development: the creation, not the consumption, of wealth” (Cole, 1998). Most thoughtful observers probably would view cities’ disincentive to seek residential or industrial development as negative for the state of California. But note that this result is not the “fault” of the local sales tax. Even in the absence of a situs-based sales tax, industrial, office, and residential development would still often fail to “pay their own way.” Thus, the solution to this problem would seem to involve adjusting local finances to provide a greater incentive to take on such development— rather than adjusting the sales tax to provide less incentives to develop 123 retail.5 Some local officials argue that cities and counties are fighting over “bones in the road,” arguing that the real need is for a redirection of property taxes back to local governments. For such reasons, in 1994 the nonpartisan Legislative Analyst’s Office proposed a plan under which the one-cent Bradley-Burns sales tax would be turned over to the state treasury. In return, the state would allocate more of the property tax to cities and counties and less to school districts (with the state covering the loss of school funds with its increased sales tax revenues). This proposal goes further in actually ending the situs-based local sales tax, as the Legislative Analyst’s Office viewed the land-use effects of sales tax competition as particularly pernicious (1994, pp. 123–125). The San Diego Association of Governments (SANDAG) has proposed an exchange of $4.8 billion between state- and local-controlled revenues. This would involve cities and counties exchanging statecontrolled funds for additional property taxes and an additional one cent of the sales tax rate. SANDAG would also give local governments authority over the half-cent public safety sales tax (currently collected by the state and delivered to counties on the basis of population, and to local governments within counties by formula). The local governments within each county would decide collaboratively on how to distribute the extra 1 percent sales tax and new property tax revenues amongst themselves (with certain “hold harmless” provisions). More recently, the California Governance Consensus Project, which is a broad-based effort at resolving intergovernmental conflicts, has ____________ 5It is worth noting, however, that adjusting the Bradley-Burns sales tax to a more population-based distribution rule would give cities a somewhat greater impetus to develop housing, since more housing means more population and thus more revenue. 124 discussed proposals to reshuffle certain sources of state and local public funds. The aim has been a reallocation of funding sources such that the resulting change would be revenue-neutral to the state and to local governments but would provide better incentives for balanced land use. One option discussed involves dedicating a considerably larger share of property taxes to the support of (non-school) local governments. To offset the increased share of K–14 school spending that the state’s general fund would have to bear after this property tax reallocation, local governments would give up various other revenues to the state— including vehicle license fees and a half cent of the Bradley-Burns sales tax. Although the Consensus Project has apparently since abandoned this specific proposal, it represents one of the most comprehensive attempts to address local fiscal problems in a revenue-neutral context. Conclusion Clearly, the issues involved in such a reallocation of revenue sources would be complex. But this complexity should not make the challenge insurmountable. California currently finds itself with a local fiscal system that promotes a competition for one type of development—retail. Such competition has significant costs, in terms of the comparative lack of attention to broader forms of economic and community development as well as the cost of incentives to the public treasury. Moreover, for most cities, the competition for a greater share of sales taxes in their region is likely to be ineffective. The amount of sales tax revenue per capita in California has been stagnant, most cities have roughly maintained the same relative position in the retail hierarchy, and retailers make location decisions based on factors that are mainly outside the control of local governments. 125 In short, in the post-Proposition 13 environment, the local sales tax is a motivation for certain types of local government land-use behaviors that—although unlikely to systematically alter broad patterns of retail development—do have drawbacks for the state as a whole. No one has seriously suggested simply doing away with the situs-based sales tax as a response to these problems, since doing so would only create a gaping hole in city finances. And distributing the local sales tax on a per capita basis would fail to provide incentives for other types of development such as housing and industry. Fortunately, as the proposals noted above show, alternatives with the potential to broaden cities’ interest in pursuing balanced growth are available to policymakers. These options involve giving local policymakers control over a larger proportion of property tax revenues, in exchange for returning other, narrower revenue bases to the state. It will be difficult politically to change the existing system, since those who gain from the status quo could stand to lose much if it were changed, whereas the potential gains are more broadly diffused and lack a highly organized and motivated constituency. Nevertheless, the state government, by setting the fiscal rules that influence local land-use decisions, holds key leverage over statewide economic development strategy. The state has the capability to create a local fiscal system that would be more encouraging of the residential and industrial development that will be important to California’s continued prosperity. 126 Appendix The Mail Survey of City Development Strategies Our discussion in Chapter 5 of the relationship between city landuse policies and sales tax considerations is based upon evidence from a survey conducted in 1998. In August of that year, an eight-page questionnaire, titled “Development Strategies in California Cities,” was sent to the city manager or city administrator of each of the 471 cities then existing in California. Where the city lacked a city manager or city administrator, the survey was sent to the top administrative official, as identified by the League of California Cities, typically the city clerk. In some cases, the person to whom the survey mailing was addressed passed the survey along to another city official to complete—typically a planning or community development professional. The self-reported job titles of respondents are reported in Table A.1. As the table shows, 76 percent of the respondents were city managers or administrators and an additional 7 percent were assistant city managers 127 Table A.1 Survey Respondents, by Title Title City Manager Assistant City Manager City Administrator Assistant City Administrator Director of Economic/Community Development Director of Planning Planner Management Analyst City Clerk Other No. of Respondents 211 19 36 2 22 7 11 5 8 5 % of Respondents 64.7 5.8 11.0 0.6 6.7 2.1 3.4 1.5 2.5 1.5 Total responses to question Did not answer question Total surveys received 326 100 4 330 or assistant city administrators. The other respondents were mainly a mix of planning officials, community development directors, and city clerks. Our response rate was 70 percent. We followed standard mail survey methodologies designed to enhance response rates (Dillman, 1978; Jobber, 1986; Salant and Dillman, 1994; Sudman, 1985)—a particularly important concern among busy professionals such as city managers. A cover letter discussed the questionnaire and its goals. We also were fortunate in having the cooperation of the League of California Cities, whose executive director provided a memo for the mailing encouraging respondents to fill out the questionnaire. Survey recipients were sent a reminder postcard encouraging them to respond, and non-respondents were sent a new copy of the questionnaire three weeks after the original mail-out. A stamped envelope was provided for return mailing. 128 Responses were tracked using identification numbers for each city. However, respondents were assured anonymity, and thus no responses that could be identified with specific cities are reported in our research. The survey asked respondents a set of closed-ended questions regarding new development on vacant land, city-supported redevelopment, and annexation policy in their cities. A “screen” question for each of these three main sections weeded out cities for which the questions were not applicable. The survey questions covered a broad variety of topics and asked respondents about the importance of numerous factors that might influence city land use policy. The sales tax and retail development were not the major focus. In fact, we wished to avoid telegraphing that this topic was our main immediate concern, as we feared doing so might make respondents “think too hard” about providing the “right” answer. We hope to examine some of the other aspects of the survey responses in future research publications. A copy of the survey is reproduced on the pages that follow. 129 Please return the completed questionnaire in the enclosed stamped envelope. Public Policy Institute of California 500 Washington St., Suite 800 San Francisco, CA 94111 (415) 291-4400 phone (415) 291-4401 fax Project Directors: Paul Lewis and Elisa Barbour Code: ___________ Note: The individual responses to this survey are confidential. The code is used to track the returning surveys. In reporting the survey results, responses of individual cities will not be identified. 130 Q-1. Which of the following statements best applies to your city? (check one answer) ___ There is considerable vacant land available for new development. ___ There is a limited amount of vacant land available for new development. ___ There is little or no vacant land available; the city is “built out.” If you chose “little or no vacant land available”please skip ahead to question Q-5. Your responses to the other questions are very important to us. Q-2 – Q-4. The following questions relate to private-sector development proposals for currently vacant land sites in your city. Q-2. Given your city’s overall strategies and plans for land use and future development, how desirable to your city administration would each of these types of new development be? In other words, how sought-after are these types of development in your city, in general? Please rank each of the following: (Circle a number between 1, which is “very undesirable,” and 7, which is “very desirable”) 1 Single-family residential 2 Multifamily residential 3 Light industrial 4 Heavy industrial 5 Retail 6 Office Very undesirable Very desirable 1 2345 67 1 2345 67 1 2345 67 1 2345 67 1 2345 67 1 2345 67 7 Mixed-use development 1 2345 67 Q-3. For each of the following types of new development, please indicate how likely your city would be to provide a general plan change (rezoning) or a financial incentive to the developer or builder of the project: (Circle a number between 1, which is “very unlikely,” and 7, which is “very likely”) 1 Single-family residential 2 Multifamily residential Very unlikely Very likely 1 2 3 45 6 7 1 2 3 45 6 7 3 Light industrial 1 2 3 45 6 7 131 4 Heavy industrial 5 Retail 6 Office 7 Mixed-use development Very Very unlikely likely 1 2345 67 1 2345 67 1 2345 67 1 2345 67 Q-4. Generally speaking, how important are the following considerations to your city administration’s strategies in attracting new development and responding to development proposals? Please review each item below and indicate how important it is. (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Cost of municipal services for the new development 1 2345 67 2 New property tax revenues generated 1 2345 67 3 New sales tax revenues generated 1 2 3 4 5 6 7 4 New fee/assessment/enterprise revenues generated 1 2345 67 5 Adequacy of infrastructure in area of project 1 2345 67 6 Likelihood of job creation 1 2345 67 7 Conformity with city’s general plan 1 2 3 4 5 6 7 8 Acceptability of proposal to nearby neighborhoods 1 2345 67 9 City council support for project 1 2 3 4 5 6 7 10 Support of Chamber of Commerce or other local business interests for project 1 2345 67 11 Nearby cities’ views on the project 1 2 3 4 5 6 7 12 Project aesthetics, urban design issues 1 2345 67 13 Traffic and other spillovers 1 2345 67 14 Contribution to sound regional economy 1 2345 67 15 Meeting affordable housing needs in area 1 2345 67 16 Preservation of agricultural land 1 2 3 4 5 6 7 17 Environmental considerations 1 2 3 4 56 7 18 Competition from nearby cities 1 2 3 4 5 6 7 132 Q-5 – Q-7. Next, we’d like to ask you about redevelopment projects supported by the city government and/or Redevelopment Agency (RDA) in your city. Q-5. Is your city engaged in redevelopment? (please check the best answer) _____ Yes, very actively _____ Yes, but not very actively _____ Not currently engaged in redevelopment If you chose this answer, please skip ahead to question Q-8. Q-6. Given your city’s overall strategies and plans for redevelopment, how desirable to your city administration would each of these types of projects be in your redevelopment areas? In other words, how sought-after are these types of projects in your city’s redevelopment areas? (Circle a number between 1, which is “very undesirable,” and 7, which is “very desirable”) 1 Single-family residential 2 Multifamily residential 3 Light industrial 4 Heavy industrial 5 Retail 6 Office 7 Mixed-use development Very undesirable Very desirable 12 34 56 7 12 34 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 133 Q-7. Again, in considering your city’s redevelopment areas, in general how important are the following considerations to your city administration’s strategies in choosing which types of projects and land uses are appropriate? (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Cost of municipal services for the new development 1 2345 67 2 New property tax revenues generated (including increment retained by Redevelopment Agency) 1 2 3 4 5 6 7 3 New sales tax revenues generated 1 2 3 4 5 6 7 4 New fee/assessment/enterprise revenues generated 1 2345 67 5 Adequacy of infrastructure in area of project 1 2345 67 6 Likelihood of job creation 1 2345 67 7 Conformity with city’s general plan 1 2 3 4 5 6 7 8 Acceptability of proposal to nearby neighborhoods 1 2345 67 9 City council support for project 1 2 3 4 5 6 7 10 Support of Chamber of Commerce or other local business interests for project 1 2345 67 11 Views of other local governments such as school district or county 1 2 3 4 5 6 7 12 Environmental considerations 1 2345 67 13 Project aesthetics, urban design issues 1 2345 67 14 Traffic and other spillovers 1 2345 67 15 Eradication of blight 1 2345 67 16 Contribution to sound regional economy 1 2345 67 17 Meeting area’s affordable housing needs 1 2 34 56 7 18 Competition from nearby cities 1 2 3 4 5 6 7 19 Other goal(s) (please specify): 1 2345 67 134 Q-8 – Q-11. Related to city development strategies is the issue of annexation. We’d like to learn about your city’s plans in this area. Q-8. What is the current square mileage of your city (if known)? ________ Q-9. Many California cities, working with Local Agency Formation Commissions, have identified “spheres of influence” beyond their current boundaries. How many square miles are in your city’s sphere of influence (if known)? (Do not include area within current city boundaries, just the additional area in the sphere of influence. If your city does not have a sphere of influence, answer zero. If answer is unknown, leave blank.) _________ Q-10. In your estimation, what are your city’s plans relating to annexation over the next five years? (Check the best response) _____ Plan to annex more than five square miles of land _____ Plan to annex about one to five square miles of land _____ Plan to annex some land, but less than one square mile _____ Cannot annex; my city does not border any unincorporated areas _____ Can annex, but do not plan to do so If your city cannot or does not plan to annex, please skip to question Q-12. 135 Q-11. How important are the following possible motivations for annexation? In other words, how do these factors affect your plans regarding whether to annex and which properties to annex? (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Land for future development to meet housing needs 1 2345 67 2 Land for future development to create jobs 1 2345 67 3 More efficient service provision 1 2 3 4 5 6 7 4 Gain future property tax revenues 1 2 3 4 5 6 7 5 Gain future sales tax revenues 1 2345 67 6 Gain future fee/assessment/ enterprise revenues 1 2345 67 7 Control development of surrounding areas to ensure consistency with city plans 1 2345 67 8 Agreements with county 1 2345 67 9 Agreements with other cities 1 2345 67 10 Direction provided by LAFCO 1 2 3 4 5 6 7 11 Provide greenbelt or open space 1 2 3 4 5 6 7 12 Prevent annexations by other cities 1 2 3 4 5 6 7 13 Other motivation(s) (please specify): 1 2345 67 General background questions: Q-12. Within the boundaries of your city, approximately how many interchanges are there that provide entry to a freeway? (Freeway = interstate highway or limited-access state highway.) _________________________________ Q-13. Please indicate your job title: _________________________________ Q-14. Please check here if you would be willing to participate in a brief telephone interview regarding these topics: __________ 136 Thank you very much for your participation. Your response is crucial to help inform policymakers about the growth challenges facing California and the strategies cities use to deal with these challenges. Please return the survey in the enclosed postagepaid envelope. You will receive a future mailing that will discuss the results of this survey. We welcome your comments on these topics, and comments regarding the questionnaire itself. You may include any written comments below, or on a separate sheet. The Public Policy Institute of California is a private, nonprofit organization dedicated to independent, nonpartisan research on economic, social, and political issues that affect the lives of Californians. For further information, including the full text of PPIC research reports, please visit our World Wide Web site at www.ppic.org. 137 Bibliography Advisory Commission on Intergovernmental Relations, Local Revenue Diversification: Local Sales Taxes, Advisory Commission on Intergovernmental Relations, Washington, D.C., 1989. Barber, Mary Beth, “Local Government Hits the Wall,” California Government and Politics, 1994–95 Annual, California Journal Press, Sacramento, 1994, pp. 82–84. Benson, Mitchel, “Do Exemptions Threaten State’s Sales-Tax Base?” Wall Street Journal, Western edition, May 7, 1997. Berry, Brian J. L., et al., Market Centers and Retail Location: Theory and Applications, Prentice Hall, Englewood Cliffs, New Jersey, 1988. Brooks, Lea, “The Tax Wars,” Golden State Report, June 1988, pp. 32– 38. Bunnell, Gene, “Fiscal Impact Studies as Advocacy and Story Telling,” Journal of Planning Literature, Vol. 12, No. 2, 1997, pp. 136–151. California State Controller, Financial Transactions Concerning Cities of California, annual publication, State Controller’s Office, Sacramento (publication renamed Cities Annual Report for 1995–96 fiscal year). 139 Chapman, Jeffrey I., Proposition 13 and Land Use, Lexington Books, Lexington, Massachusetts, 1981. Chapman, Jeffrey I., Proposition 13: Some Unintended Consequences, Occasional Paper, Public Policy Institute of California, San Francisco, 1998. Cole, Rick, “Sales Tax War Hurts Civic Economies,” Pasadena StarNews, February 22, 1998. Coleman, Michael, Financing Cities: City Financing in the Decades after Proposition 13, League of California Cities, Sacramento, 1998. Dardia, Michael, Subsidizing Redevelopment in California, Public Policy Institute of California, San Francisco, 1998. Davies, Ross L., Marketing Geography, Retailing and Planning Associates, Corbridge, England, 1976. Dillman, Don A., Mail and Telephone Surveys: The Total Design Method, John Wiley & Sons, New York, 1978. Division of Tourism, California Trade and Commerce Agency, California Travel Impacts by County, 1992–1996, Prepared by Dean Runyan Associates, Portland, Oregon, 1998. Dresch, Marla, and Steven M. Sheffrin, Who Pays for Development Fees and Exactions? Public Policy Institute of California, San Francisco, 1997. Due, John F., and John L. Mikesell, Sales Taxation: State and Local Structure and Administration, 2nd ed., Urban Institute Press, Washington, D.C., 1994. Flanagan, James, “The Problem Is Larger Than Housing,” Los Angeles Times, August 30, 1998. Fulton, William, “Cities, Counties Seek to Resolve Annexations Fight,” California Planning & Development Report, Vol. 12, No. 6, 1997a, pp. 1, 9. 140 Fulton, William, The Reluctant Metropolis: The Politics of Urban Growth in Los Angeles, Solano Press Books, Point Arena, California, 1997b. Fulton, William, “Twenty Years of Proposition 13: Tax-Cutting Initiative Shaped Planning and Development in State,” California Planning & Development Report, Vol. 13, No. 6, 1998, pp. 1, 8–10. Ghaus, A. F. Aisha, “Optimal Local Sales Tax,” Urban Studies, Vol. 32, No. 8, 1995, pp. 1369–1381. Guy, Clifford, The Retail Development Process: Location, Property, and Planning, Routledge, London, 1994. International Council of Shopping Centers, “NRB/Shopping Centers Today 1995 Shopping Center Census,” International Council of Shopping Centers, New York, 1996. Jobber, David, “Improving Response Rates in Industrial Mail Surveys,” Industrial Marketing Management, Vol. 15, 1986, pp. 183–195. Kotin, Allan, “Economic Development Can Be Bad for Your Community’s Fiscal Health,” Metropolitan Forum Project, web page http://www.democracynet.org/metro/new/kotinapr.html, n.d. Kotin, Allan, and Richard Peiser, “Public-Private Joint Ventures for High Volume Retailers: Who Benefits?” Urban Studies, Vol. 34, No. 12, 1997, pp. 1971–1986. Krmenec, Andrew J., “Sales Tax as Property Tax Relief? The Shifting Onus of Local Revenue Generation,” Professional Geographer, Vol. 43, No. 1, 1991, pp. 60–67. Kroll, Cynthia, and Eileen Marrinan, “California Retail Sales: Growth Patterns and Locational Tendencies,” Center for Real Estate and Urban Economics, Working Paper 85-95, University of California, Berkeley, 1985a. Kroll, Cynthia, and Eileen Marrinan, “Siting Retail Space in the 1980s: A California Perspective,” Center for Real Estate and Urban Economics, Working Paper 85-96, University of California, Berkeley, 1985b. 141 League of California Cities, Conflicts at the City’s Edge: Annexations, Development and Tax Negotiations, prepared by Crawford Multari & Starr, San Luis Obispo, California, 1996. League of California Cities, “Facts about City Dollars,” pamphlet, n.d. Legislative Analyst’s Office, State of California, Making Government Make Sense, Sacramento, July 21, 1994. Legislative Analyst’s Office, State of California, Why County Revenues Vary: State Laws and Local Conditions Affecting County Finance, Sacramento, May 7, 1998. Lewis, Paul G., Shaping Suburbia: How Political Institutions Organize Urban Development, University of Pittsburgh Press, Pittsburgh, 1996. Lewis, Paul G., Deep Roots: Local Government Structure in California, Public Policy Institute of California, San Francisco, 1998. Miller, Gary J., Cities by Contract: The Politics of Municipal Incorporation, MIT Press, Cambridge, Massachusetts, 1981. Misczynski, Dean J., “The Fiscalization of Land Use,” in John J. Kirlin and Donald R. Winkler (eds.), California Policy Choices, Vol. 3, University of Southern California, Los Angeles, 1986. Newman, Morris, “Legislature Questions Retailing Subsidies,” California Planning & Development Report, Vol. 13, No. 4, 1998, pp. 1–2. Orfield, Myron, Metropolitics: A Regional Agenda for Community and Stability, Brookings Institution Press, Washington, D.C., 1997. Peirce, Neal R., “Neal Peirce on Prop 13 Terrors: They’d Do It All Again!” Metropolitan Forum Project, Worldwide Web site http://www.democracynet.org/metro/new/peircejuly.html, n.d. Peterson, Paul E., City Limits, University of Chicago Press, Chicago, 1981. Ross, Bernard H., and Myron A. Levine, Urban Politics: Power in Metropolitan America, F. E. Peacock Publishers, Itasca, Illinois, 1996. 142 Salant, Priscilla, and Don A. Dillman, How to Conduct Your Own Survey, John Wiley and Sons, New York, 1994. Schneider, Mark, “The Market for Local Economic Development: The Growth of Suburban Retail Trade, 1972–1982,” Urban Affairs Quarterly, Vol. 22, No. 1, 1986, pp. 24–41. Schrag, Peter, Paradise Lost: California’s Experience, America’s Future, New Press, New York, 1998. Schwartz, Jonathan, “Prisoners of Proposition 13: Sales Taxes, Property Taxes, and the Fiscalization of Municipal Land Use Decisions,” Southern California Law Review, Vol. 71, No. 1, 1997, pp. 183–217. Senate Local Government Committee and Assembly Local Government Committee, State of California, Land Use and Local Revenue-Sharing: Playing the Zero-Sum Game, Summary Report of a Joint Interim Hearing, Sacramento, November 17, 1989. Shires, Michael A., Patterns in California Government Revenues Since Proposition 13, Public Policy Institute of California, San Francisco, 1999. Shuit, Douglas P., “The Retail Wags the Dog,” Los Angeles Times, July 17, 1998, p. B-2. Sokolow, Alvin D., “State Rules and the County-City Arena: Competition for Land and Taxes in California’s Central Valley,” Publius, Vol. 23, Winter 1993, pp. 53–69. State Board of Equlization, 1997–98 Annual Report, Sacramento, 1998. Sudman, Seymour, “Mail Surveys of Reluctant Professionals,” Evaluation Review, Vol. 9, No. 3, 1985, pp. 349–360. Vercoutere, Michele E., “The Local Sales Tax in California: Its Significance to the Urban Structure,” Master’s Thesis, School of Architecture and Urban Planning, University of California at Los Angeles, 1979. 143 About the Authors PAUL G. LEWIS Paul Lewis is a research fellow at the Public Policy Institute of California. He studies local government, urban and suburban growth issues, and transportation. He is also interested in the nature and causes of long-term policy change. Among his recent publications are Deep Roots: Local Government Structure in California (Public Policy Institute of California, 1998), Federal Transportation Policy and the Role of Metropolitan Planning Organizations in California (Public Policy Institute of California, 1997), and Shaping Suburbia: How Political Institutions Organize Urban Development (University of Pittsburgh Press, 1996). He holds an A.B. in political science from Indiana University and an M.A. and Ph.D. in politics from Princeton. ELISA BARBOUR Elisa Barbour is a research assistant at the Public Policy Institute of California who works on projects dealing with California governance issues. Before coming to PPIC, she did research examining local labor markets, welfare reform, and urban poverty. She also worked for a number of community and advocacy organizations. She holds a B.A. in political science from Oberlin College and an M.A. in City Planning from the University of California, Berkeley. 145" } ["___content":protected]=> string(102) "

R 799PLR

" ["_permalink":protected]=> string(84) "https://www.ppic.org/publication/california-cities-and-the-local-sales-tax/r_799plr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8124) ["ID"]=> int(8124) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:35:03" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3235) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 799PLR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_799plr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_799PLR.pdf" ["wpmf_size"]=> string(7) "1027884" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(257769) "California Cities and the Local Sales Tax ••• Paul G. Lewis Elisa Barbour 1999 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data Lewis, Paul George, 1966– California cities and the local sales tax / Paul G. Lewis, Elisa Barbour. p. cm. Includes bibliographical references. ISBN: 1-58213-010-8 (alk. paper) 1. Sales tax—California. 2. California—Economic policy. I. Barbour, Elisa, 1961— II. Title. HJ5715.U62C25 1999 336.2'713'09794—dc21 99-23683 CIP Copyright © 1999 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 the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword As visitors are quick to learn, California’s sales tax rate is higher than average. Very few residents, let alone visitors, are aware that the rate varies throughout the state, depending on local preferences. Nor are they aware that 1 percent of the overall rate of 7.25 to 8.5 percent is allocated to local governments on a point-of-sale basis. This 1 percent rate is a “diamond in the rough” for cities and counties. It is particularly a “diamond” for cities because it is a major source of their discretionary revenues. Moreover, sales tax revenues have the potential to grow as a city expands and its commercial land uses generate more sales. The “rough” is the restricted arena for raising local revenues after the passage of Proposition 13 in 1978. The authors of this report, Paul Lewis and Elisa Barbour, take a thoughtful look at the consequences of a sales tax regime that is so important to local governments. They address two key questions: What are the effects of the point-of-sale or situs-based sales tax on land-use iii decisions? and How do California cities vary in the benefits they receive from the tax? In conducting this study, the authors first surveyed top administrative officials in 330 cities throughout the state. They then integrated their findings with a thorough analysis of recorded revenue distributions over a 24 year period from 1971 to 1995. Their survey supports the argument, advanced by Dean Misczynski in the mid-1980s, that there has been a “fiscalization" of land-use decisions: Retail land uses are preferred over residential and heavy industrial uses by a substantial margin. City policymakers pursue land uses that generate high levels of sales tax revenues—shopping centers and auto malls. At the same time, the authors demonstrate that despite all the competition for revenue-generating retail businesses, the hierarchy among cities in their sales tax success has changed little over the study period. In other words, the competition is intense and the incentives offered to developers can be substantial, but the pattern of winners and losers remains largely the same. They argue that there is a relatively fixed amount of retail activity that can be supported by the market in a region at any given level of population and that, in the broadest context, business decisions about where to locate are not greatly affected by local incentives. The downside of each city’s giving priority to retail uses, however, is that residential and industrial uses may become more difficult and expensive to site. The authors conclude the study with an examination of various options for reducing distortions in land-use preferences stemming from sales tax policy. They suggest that one promising approach might be to reallocate a greater share of local property taxes to the cities, perhaps in exchange for returning some of the sales tax revenues to the state. Whatever policy correction is chosen, it is clear that cities compete iv actively for retail sales, that the consequences only marginally affect overall ranking in revenues, and that there are real costs to city budgets, even though the returns are negligible. This volume is one of a series that the Public Policy Institute of California is publishing on the status of public finance in California. The policy issues involved in understanding public finance issues today, including the question of equity, the fiscal relationship between state and local governments, and the level and quality of government services are large and important concerns that need to be analyzed with an objective and independent eye. They are exactly the kinds of issues that PPIC was founded to study. We trust that this growing body of research and findings on local government finance will reduce the level of disagreement and set the stage for a more informed reformulation of public policy. David W. Lyon President and CEO Public Policy Institute of California v Summary Although it is somewhat hidden from public view, controversy surrounds the local sales tax in California. A 1 percent local sales tax is collected by the state as part of the larger sales and use tax levied on most items for sale. It generates revenues that are returned to the local government in which the sale occurred (see Table S.1). This return of funds to the point-of-sale jurisdiction is known as the situs rule. The situs rule has tremendous implications, since it gives cities a special incentive to promote the location of retail businesses within their boundaries—an incentive that does not exist for residential or industrial development. For California’s cities, the significance of the local sales tax is much greater than one might suspect, given its relatively small share of their overall revenues. In recent years, the situs-based sales tax has constituted about one-tenth of total city revenues. Nevertheless, the sales tax—along with the property tax—is one of the major sources of discretionary revenues for cities. Revenue from most other major sources, by contrast, vii Table S.1 Components of California’s Overall Sales Tax Rate Rate, % 6.00 5.00 0.50 0.50 1.25 1.00 0.25 Up to 1.25 7.25 to 8.50 Purpose State sales tax, consisting of: State general fund Local Revenue Fund—distributed to counties for health and welfare responsibilities Public Safety Fund—distributed to counties, some cities Bradley-Burns sales tax, consisting of: Local sales tax—directed to general fund of jurisdiction where sale occurred Local transportation tax—directed to county where sale occurred Local special taxes, generally for transportation— optional, require voter approval, used in 24 counties and a few cities (note: in most counties, the maximum rate authorized is 1.50 percent). Total rate SOURCES: Adapted from State Board of Equalization (1998), p. 25; California State Controller (1995–96), p. ix; and the California Revenue and Taxation Code, Section 7251. is earmarked or restricted for certain functions. Moreover, city policymakers perceive that the sales tax has the potential for substantial revenue growth in good economic times—if supportive land-use decisions are made. In the post-Proposition 13 world, these factors— budget flexibility and growth potential—make it a highly prized revenue source for the state’s municipalities. This report focuses on two of the major issues arising from cities’ heavy interest in sales tax revenues. First, what are the effects of the situs-based sales tax on cities’ land-use decisionmaking? Are localities systematically favoring retail development over other types of growth? If they are, this is potentially a negative scenario for economic development in the state as a whole. Second, how do California cities vary in the viii benefits they receive from the local sales tax? Which types of communities are doing better or worse in their quest for sales tax revenues? The Sales Tax and Local Development Decisions Many critics, pointing to local government efforts to recruit retail businesses, have worried about public-sector “giveaways” to retailers or developers. Critics also complain that the “fiscalization of land use”— development decisions favoring tax-generating activities—has retarded housing and other non-retail development. But the evidence of fiscalized land-use decisionmaking has thus far been largely anecdotal. To shed light on this question, PPIC conducted a mail survey of local officials on the topic of “Development Strategies in California Cities” in late 1998. The survey was sent to the city manager, or other top administrative official, in each of the 471 cities then in existence in the state. The response rate was 70 percent, and the sample was quite representative. The survey results provide strong evidence that city governments do systematically favor retail development over other land uses when it comes to new development on vacant land, as well as for redevelopment in designated “blighted” areas. Figure S.1 shows how cities ranked retail and six other land-use categories on a seven-point scale of desirability. (The responses were screened to include only those cities with vacant land available for development, and only those saying they engage in redevelopment.) Retail is the most favored category, although office, mixed-use, and light industrial development are also considered quite desirable. Respondents also indicated that retail projects were the most ix Average desirability score 7 6 New development Redevelopment 5 4 3 2 1 Retail Office Mixed- Light Single- Multi- Heavy use industrial family family industrial development residential residential SOURCE: PPIC City Manager Survey, 1998. Figure S.1—Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers likely to receive a general plan change or financial incentive from their cities. Why the favor toward retail? Our survey also asked respondents to rank the importance of 18 motivations for evaluating development and redevelopment projects, including such factors as job creation, property tax generation, environmental effects, and effects on neighborhoods. Of all these factors, “new sales tax revenues generated” was tied with “city council support for the project” as the top consideration for redevelopment. In the case of new development, sales taxes actually exceeded city council support in importance. Respondents whose cities plan to annex land in the next five years also indicated that gaining new sales tax revenues was the second most important consideration of 12 x factors in deciding which lands to annex, after the desire to control development patterns in surrounding areas. There are regional variations in the degree of importance accorded to sales tax considerations and the favor shown to retail. A set of “pragmatic” factors regarding land use—including sales tax and other fiscal concerns—are considered most important in Central Valley cities and in other mostly rural portions of the state but less important in the San Francisco Bay area. The 36 cities designated as “central cities” by the U.S. Census Bureau tend to have more balanced land-use perspectives than other communities, with retail and sales tax motivations ranked lower relative to other concerns. In general, however, few characteristics can be said to distinguish cities that are more actively pursuing retail development. Instead, the practice is so common that it can be called nearly ubiquitous. Disparities in City Revenues Under the Local Sales Tax City success in the sales tax game varies a great deal. In the 1995–96 fiscal year, the amount of local sales tax revenues per resident received by California cities ranged from $2.25 to $56,891.84. In short, “winners” and “losers” emerge in the local sales tax revenue game. Critics complain that these inequities in revenues do not reflect communities’ social needs for government services, nor do they correspond directly to the amount of effort cities have devoted to economic development. As retail patterns have shifted—given the upswing in suburbanization, shopping malls, and new types of retail-market activity—certain local governments have benefited and others have lost ground. As Figure S.2 shows, the 36 central cities have seen lower levels xi Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Central cities All other cities 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.2—Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 of sales tax revenues per capita over the period since 1971 relative to other municipalities. Also in this period, sales tax revenues per capita in the San Francisco Bay area moved well ahead of those in the greater Los Angeles region. Multivariate statistical analysis indicates that a city’s success in receiving per capita sales tax revenues (as of fiscal years 1990–91 through 1992–93) can be partially predicted based on certain city characteristics. Not surprisingly, the findings reveal that sales tax success is associated with certain market characteristics—cities with higher populations, lower densities, and fewer persons per household. Cities devoting more of their land to redevelopment projects also performed better. But demographic characteristics of cities do not relate to sales tax success in a totally xii straightforward fashion: Cities with very low incomes and higher shares of blacks in their populations had less success, but so did cities with high population growth rates and very high incomes. This last finding emerges in part because many of the “least successful” sales-taxgenerating cities tend to be very wealthy residential suburbs that eschew commercial development. It indicates that although the situs rule does systematically favor some cities, it is not necessarily biased in favor of high-status communities. Sales tax disparities among cities have not been increasing over the past two decades. In fact, the cities that were the most successful in receiving per capita sales tax revenues as of the early 1970s tended to have significantly lower gains over time, relative to other cities. Over the entire two-decade period, most communities experienced a relatively stable position in the sales tax hierarchy, as Figure S.3 illustrates. Fighting for Slices of a Fixed-Size Pie Overall, from the mid-1970s through the mid-1990s, sales taxes, measured in real dollars per capita, were a fairly stagnant source of funds for California cities, allowing for business-cycle swings (see Figure S.4). This is due to the relatively fixed nature of retail spending per capita, the many exemptions that state law provides from sales taxation, and the increasing share of consumer spending going to transactions that are not subject to the sales tax (for example, personal services and catalog sales). Cities are clearly trying to attract retail development. Despite this, the hierarchy among cities in their sales tax success has not changed much. And since per capita sales tax collections are steady or declining overall, it is likely that cities are competing over a relatively fixed amount xiii Log of per capita sales tax revenues, 1995 8 7 6 5 4 3 2 1 0 012345678 Log of per capita sales tax revenues, 1971 (in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.3—Cities’ Sales Tax Revenues per Capita, 1971 and 1995 of per capita revenue. There is only a certain amount of retail activity that can be supported in a region at any given level of population. This makes the growth prospects of retail different from industries in which regions can experience indigenous growth and local gains can lead to broader economic benefits outside the host city. Cities that succeed in recruiting retail businesses within their borders, by contrast, can generally be viewed as simply shifting retail sales geographically within their market region. The issue then becomes whether cities’ efforts to attract retail could systematically affect land-use patterns in the state. First, the actions of cities will not induce “extra” retail development to occur, if the additional retailers cannot hope to make a profit. There is no evidence xiv Revenue (constant 1995 dollars) 140 130 120 110 100 90 80 70 60 1971 1976 1981 1986 1991 1996 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure S.4—Sales Tax Revenues per Capita for All California Cities that California has substantially more retail space per person than other states. Second, it is also questionable whether all of the favor cities shine on retail leads to major differences in retail location. Retailers generally locate in relation to their customer base, transportation accessibility, suppliers, and competitors. Specific jurisdictions may win and lose in competing for a given retailer, but the winner and loser locales are likely to be fairly similar places. Third, however, the competition for sales taxes may have a broader effect. If cities favor retail development substantially over housing and industrial development, then those land uses will likely be somewhat more difficult, uncertain, and expensive to develop. xv If all localities use similar policy efforts as inducements to retailers, then the advantages that individual cities might gain from these policies will likely be “competed away.” Presumably, the aggregate result would be a shift of resources from local governments to the retail sector (store owners, developers, landowners—and potentially even customers). Considerations for State Policymakers The effects of the sales tax on land-use decisions, along with the vast disparities in sales tax revenues among cities, often lead reformers to urge a change in the situs basis for sales tax distribution. The most significant recent reform was Proposition 11, passed by the voters in November 1998. This constitutional amendment allows any two or more local governments to negotiate sales tax revenue-sharing agreements, with a two-thirds vote of their governing boards. It will enable some experiments in cooperation but is likely to have only a modest overall effect, since local governments that are doing well at attracting retailers will have little incentive to “disarm” in the sales tax race. An alternative to the situs rule often discussed would be to switch toward a more population-based system for distributing sales tax revenues among local governments. Using 1993–94 fiscal data, we calculated that 56 percent of cities would have improved their fiscal situation if sales tax revenues were redistributed statewide on a population basis; 51 percent of cities would have improved their position if the revenues were redistributed by population within each county. The vast majority of county governments would have received more funds under either scenario. A large majority of the California population lives in jurisdictions that would have been better off under each of these population-based approaches to distributing sales tax revenues (56.2 xvi percent and 59.5 percent, respectively). However, the intra-county redistribution of revenues would not advance social equity goals, in the sense of providing more funds to lower-income communities. Distribution of sales tax revenue by population has the advantages of better reflecting fiscal needs, potentially encouraging housing development, and minimizing local “chasing” of retail development. However, the population-based approach has potential disadvantages as well. These include a possible decline in entrepreneurial economic development policy among cities as well as difficult-to-justify windfall gains for wealthy residential enclave communities. The reform also could shortchange some cities with very large retail sectors, which often have greater need for public revenues. Because of the influx of shoppers and workers, their “daytime populations” are far higher than their resident populations, causing burdens for roads and infrastructure, public safety, and other local public services. A more important critique of a population-based distribution of the sales tax is that it fails to address the broader problem of providing fiscal incentives for balanced development. The way around this dilemma is to widen our focus beyond the sales tax—which is merely one component of local revenues. As it stands, retail is looked upon with favor not merely because it is associated with the sales tax but because other forms of development are perceived to simply not pay their own way. Cities often are particularly unenthusiastic about housing because they receive a relatively small slice of the property tax dollar and a fairly large service burden. The solution to this problem would involve adjusting local finances to provide a greater incentive to take on such development— rather than merely adjusting the sales tax to provide less incentive to develop retail. Reallocating a substantially greater share of local property xvii taxes to local governments—perhaps in exchange for a withdrawal of some state aid and a portion of the local sales tax—is one promising approach to this complex problem. xviii Contents Foreword..................................... iii Summary..................................... vii Figures ...................................... xxiii Tables ....................................... xxv Acknowledgments ............................... xxvii 1. INTRODUCTION ........................... Historical Overview of Local Sales Taxes in the United States ................................. Local Sales Taxes in California .................... Contemporary Importance of the Sales Tax to Cities ...... Policy Issue I: Disparities Among Communities in Revenues ............................... Policy Issue II: Accusations of Fiscalized Land-Use Decisions ............................... Recent State Legislative Proposals to Address Sales Tax Concerns ............................... Organization of This Report ...................... 2. TRENDS IN CITY SALES TAX REVENUES ......... Sales Taxes as a Component of Cities’ Revenue Streams .... Revenue Gainers and Losers ...................... Changes in the Sales Tax Hierarchy over Time .......... Regional Trends ............................. 1 3 4 8 10 11 13 14 17 18 22 25 28 xix Trends in Central Cities ........................ The Experience of Newly Incorporated Cities ........... Summary .................................. 3. WHICH TYPES OF CITIES HAVE BENEFITED FROM THE SITUS RULE?........................... Variations Across Cities: The Extremes .............. Variations Across Cities: A Cross-Sectional Model ....... Measuring Sales Tax Revenues: The Dependent Variable ... City Characteristics: Independent Variables ............ Results of Model Estimation ...................... Factors Related to Sales Tax Success: Interpreting the Results ................................ Cities’ Relative Gains over Time: A Longitudinal Model ... Summary .................................. 4. THE SALES TAX AND LOCAL LAND-USE DECISIONS ............................... What Is Meant by the “Fiscalization” of Land Use? ....... Thinking Carefully About the Fiscalization Argument ..... Favoring Retail? ............................ Disfavoring Other Types of Growth? ............... Changing the Landscape? ...................... Implications and Policy Considerations ............... 5. THE PREFERENCE FOR RETAIL DEVELOPMENT ... Survey Methodology ........................... Development Activity in California Cities ............. Introduction to Survey Findings ................... Land-Use Preferences .......................... Factors Influencing City Development Decisions ........ Regional Variation in Attitudes About Development ...... Factors Influencing Annexation Decisions ............. Do City Governments Act on Their Development Preferences? ............................. Which Types of Cities Favor Retail Development the Most? ................................. Interest in Retail/Sales Tax: Dependent Variables ...... City Characteristics: Independent Variables .......... 33 36 39 41 41 45 45 46 51 52 58 63 67 68 71 72 75 75 78 81 82 83 86 87 88 91 95 96 97 98 98 xx Results of Model Estimation .................... 98 Factors Related to Fiscalization: Interpreting the Results .. 104 Characterizing Regional Patterns in Development Priorities .. 105 Summary .................................. 106 6. POLICY CONSIDERATIONS ................... 109 Summarizing the Results ........................ 109 Policy Debates over Sales Tax Revenue Sharing .......... 111 Arguments For and Against Moving Toward a Population- Based Distribution of Sales Tax Revenues .......... 113 Which Local Governments Would Benefit from a Population-Based Distribution? ................ 117 Statewide Redistribution ....................... 118 Within-County Redistribution ................... 120 Broader Remedies: Balancing Fiscal Rewards for Growth in California Communities ..................... 122 Conclusion................................. 125 Appendix The Mail Survey of City Development Strategies ......... 127 Bibliography .................................. 139 About the Authors ............................... 145 xxi Figures S.1. Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers ............................... S.2. Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ............ S.3. Cities’ Sales Tax Revenues per Capita, 1971 and 1995 .. S.4. Sales Tax Revenues per Capita for All California Cities .................................. 2.1. Revenue to California Cities, by Source ............ 2.2. Sales Tax Revenues per Capita for All California Cities .................................. 2.3. Dispersion of Adjusted per Capita Sales Tax Revenues Among California Cities, 1971–1995 ............. 2.4. Cities’ Sales Tax Revenues per Capita, 1971 and 1995 .. 2.5. Relationship Between Cities’ Sales Tax “Success” in the Early 1970s and Subsequent Gains .............. 2.6. Relationship Between Changes in Cities’ Sales Tax “Success” over the 1970s and 1980s .............. 2.7. Sales Tax Revenues per Capita in Major Metropolitan Areas .................................. x xii xiv xv 18 20 25 26 27 29 30 xxiii 2.8. Per Capita Sales Tax Revenues in the Los Angeles Area, 1995 .................................. 31 2.9. Per Capita Sales Tax Revenues in the San Francisco Bay Area, 1995 .............................. 32 2.10. Changes in per Capita Sales Tax Revenues in the Los Angeles Area ............................. 33 2.11. Changes in per Capita Sales Tax Revenues in the San Francisco Bay Area ......................... 34 2.12. Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ............ 35 2.13. Sales Tax Revenues per Capita: Cities Incorporated after Proposition 13 Compared to Other Cities .......... 37 3.1. Relationship Between Average Household Size and City Sales Tax “Success” ......................... 54 3.2. Relationship Between Hispanic Share of Population and City Sales Tax “Success”...................... 56 3.3. Relationship Between Income Level and City Sales Tax “Success” ............................... 57 5.1. Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers ............................... 87 6.1. Relationship Between Retail Concentrations and Crime in California Cities ......................... 115 xxiv Tables S.1. Components of California’s Overall Sales Tax Rate .... viii 1.1. Components of California’s Overall Sales Tax Rate .... 7 1.2. Cities with Highest and Lowest Sales Tax Revenues per Capita, 1996 ............................. 10 2.1. Changes in Real per Capita Sales Tax Revenues Between 1970s and 1990s Among California Cities .......... 22 3.1. Summary Statistics for Cities with Extreme Levels of per Capita Sales Tax Revenues as of 1995 ............. 43 3.2. Regression Model of City Sales Tax “Success” as of the Early 1990s .............................. 53 3.3. Regression Model of Changes in City Sales Tax “Success” Between Early 1970s and Early 1990s ...... 61 5.1. Survey Respondents, by City Population Size ........ 83 5.2. Survey Respondents, by Region ................. 84 5.3. New Development and Redevelopment Activity in California Cities ........................... 84 5.4. Development Activity, by Region ................ 85 5.5. Factors Influencing Development and Redevelopment Decisions ............................... 89 xxv 5.6. Importance Scores for Development Motivations, by Region ................................. 92 5.7. Factors Influencing Annexation Decisions, by Region ... 95 5.8. Dependent Variables in the Regression Model ........ 99 5.9. Independent Variables in the Regression Model ....... 100 5.10. Results of Regression Models: City Characteristics Associated with Preference for Retail Development..... 102 5.11. Results of Regression Models: City Characteristics Associated with Two Patterns of Development Priorities................................ 107 6.1. Number of Jurisdictions That Would Be Better or Worse Off Under a Statewide, Population-Based Distribution of the Local Sales Tax, 1994 ................... 118 6.2. Characteristics of Cities Gaining and Losing Revenues Under a Statewide, Population-Based Distribution of the Local Sales Tax............................ 119 6.3. Number of Jurisdictions That Would Be Better or Worse Off Under an Intracounty, Population-Based Distribution of the Local Sales Tax, 1994 ........... 120 6.4. Characteristics of Cities Gaining and Losing Revenues Under an Intracounty, Population-Based Distribution of the Local Sales Tax ......................... 121 A.1. Survey Respondents, by Title .................. 128 xxvi Acknowledgments This project proved much more complicated than we initially imagined. Through the help of numerous people, we were able to get a better grasp of the facts and issues, complete the research in a more timely fashion, and improve our writing and presentation. Seven people read the draft report in its entirety and provided many helpful comments: Michael Coleman, Michael Dardia, Peter Detwiler, Max Neiman, Joyce Peterson, Peter Schaafsma, and Michael Teitz. The reviews were clearly done with great care. Although we were not able to follow through on all of the suggestions these able reviewers made, their help resulted in significant improvements to the report. At an early stage of the research, a group of city managers, finance directors, and PPIC staff met with us to discuss the project and offered very constructive comments and suggestions. City officials present included Kenneth Blackman, Pamela Gibson, Rod Gould, Brad Grant, Derek Hanway, and Kathleen Millison. We thank Fred Silva for his help in organizing the session and for his continuing advice. Also present at xxvii the meeting was Don Benninghoven, the executive director (now retired) of the League of California Cities. We thank Mr. Benninghoven and the League’s staff for their help in reviewing our mail survey of local officials before its release and encouraging recipients of the survey to respond. Don Blubaugh also made useful suggestions on question wording. The mail survey drew responses from 330 cities and we would like to thank each of the busy professionals who took time to respond. At PPIC, Sun Kim did a great deal of work in organizing the administration of the survey; we could not have made this step work without his help. Sun also provided mapping analysis that proved important to the research and found ways to overcome software limitations along the way. In addition to the reviewers listed above, Mark Baldassare, Kim Rueben, Michael Shires, and other colleagues provided useful research advice. Mark Butala, a summer intern at PPIC in 1997, got the ball rolling by collecting much of the sales tax data and performing some preliminary data analysis. Others at PPIC who provided key help at various stages included Gary Bjork, Meghan Cashman, Arabella Cureton, Shannon Goecke, Kathy Kayhour, Judi Leonor, and Heather McMann. Participants in a colloquium session at California State University at Sacramento’s Graduate Program in Public Policy and Administration provided helpful feedback. In particular, Laurie Watson helped save us from an error. Although all of these persons helped improve the report, any errors, omissions, or other shortcomings are the authors’ responsibility. xxviii 1. Introduction In recent years, observers of California local government have given increasing amounts of attention to the local sales tax. This interest stems from two major concerns relating to the sales tax as an outcome and also as a cause of important economic and geographic trends in the state. First, the differences in levels of sales tax revenue flowing to the various local governments are an outcome of retailers’ location decisions in California. This is because of California’s so-called situs rule, which returns local sales tax revenue—although collected by the state—to the jurisdiction in which the sale occurred. As retail patterns in California and the nation have changed in recent decades—given the upswing in suburbanization, shopping malls, and new types of retailing such as power centers and “big box” stores—certain local governments have benefited and others have lost revenues. In some cases, neighboring cities receive vastly different flows of sales tax funds per capita. Some state and local officials and other observers have complained about these inequities among communities, arguing that such differences do not reflect 1 communities’ social needs or their economic development effort, and that the relative lack of sales tax receipts can complicate local budgeting. The second area of concern regarding the local sales tax is a natural outgrowth of the first issue. Since cities do not want to be “losers” in the quest for local revenues, they increasingly compete for and recruit retail land uses and provide incentives, according to many observers. This behavior may in turn cause land-use and development patterns to be different than what might exist under a more “neutral” system of local finance. That is, local land-use decisions may be biased or distorted in favor of retail—and thus against residential or industrial projects that might otherwise occupy the same properties. Moreover, local governments may engage in a difficult game of enticing retailers to relocate from other places in their market area, thus redistributing the location of retail. These strategies are often referred to as the “fiscalization of land use” and are generally thought to have become more pronounced since the passage of Proposition 13 in 1978, which limited local governments’ access to property tax revenues. This report examines both of these issues—disparities across communities in sales tax revenues and the possibility of sales-tax-driven land-use decisionmaking. We analyze the state’s municipalities— generally all referred to as cities in California—since these governments are by far the biggest beneficiaries of the local sales tax.1 The issues are complex, and some questions, particularly those related to the fiscalization of land use, cannot be answered definitively. Nevertheless, ____________ 1In 1998, when most of our research was completed, there were 471 cities in California. The incorporation of two new cities, Oakley and Laguna Woods, occurred after our study period. Incorporation attempts are currently under way in several other communities. 2 we seek to provide some useful basic analysis of these issues, which have been subject to much controversy but little serious empirical study. Questions to be addressed include: Which types of cities are helped or hindered by the situs rule? How might a different rule for distributing local sales tax revenues affect various types of communities? Which types of cities have gained or lost ground in receiving such revenues in recent decades? Do city officials seek retail development more than other types of land uses? How important is the sales tax as a motivation for city land-use strategies, as compared to other factors? This introductory chapter provides background on the local sales tax and some of the controversies related to it. Historical Overview of Local Sales Taxes in the United States2 Today, Californians pay about 4 percent of their personal income as state and local sales taxes, with the state claiming three-quarters of that amount.3 But the sales tax has not always been part of the public finance system in the state. California, and many other states and localities, began using sales taxes during the Great Depression, when existing revenue sources proved drastically insufficient to meet the extreme needs of their populations. During the Depression, not only did the traditional property-based revenues that local governments relied upon drop as a result of free-falling property values and foreclosures, but also social welfare activities were funded overwhelmingly at the state and local level. ____________ 2This section draws upon Advisory Commission on Intergovernmental Relations (ACIR) (1989); Due and Mikesell (1994); and Vercoutere (1979). 3See the state-by-state comparison table in Governing magazine’s State and Local Sourcebook 1998, p. 34. 3 California began its state sales tax (at a 2.5 percent rate) in 1933 (Benson, 1997). By 1936, 25 states had sales taxes. It is believed that New York City, in 1934, was the first city to make use of a local sales tax; New Orleans followed in 1936. Once the substantial revenue possibilities inherent in local sales taxes became apparent, the innovation spread, and even after the Depression emergency, hundreds of other local governments turned to this source. Policymakers were attracted to the local sales tax because of its relative invisibility, compared to property tax bills, and its responsiveness to inflation. In the 1940s, California and Illinois joined the list of states employing municipal sales taxes. Another set of states turned toward local sales taxes in the 1960s, and the share of local revenues accounted for by sales taxes thereby approximately doubled in the two decades after 1966 (ACIR, 1989). Nationwide, the 1970s saw a trend toward substituting local sales tax revenues for local property tax revenues (Krmenec, 1991). In fiscal year 1991, 16 percent of total tax revenues4 received by U.S. municipalities came from sales taxes, making it the second leading source of tax revenue behind the property tax (Due and Mikesell 1994, p. 277). By 1994, 33 states and the District of Columbia had local sales taxes. Local Sales Taxes in California When California’s economy picked up after the Great Depression, the state lowered its state sales tax from 3 percent to 2.5 percent in 1943. Soon after this reduction, numerous cities around the state followed up ____________ 4That is, revenues excluding enterprise fees, service charges, intergovernmental grants, and so on. 4 by imposing a locally levied sales tax—a propitious time to do so, since consumers probably did not notice the imposition (Vercoutere, 1979, pp. 22–26). Postwar population growth, infrastructure needs, and the declining property values of the Depression had stressed city budgets, which were only beginning to recover. Beginning with San Bernardino in 1944, cities around the state began to levy their own sales taxes (typically at 0.5 percent), with 35 doing so by the end of 1946. “By FY 1954–55, 175 (56%) California cities had a sales tax ordinance. These taxing cities were the residences of more than 87% of the state’s population” (Vercoutere, 1979, p. 32). Cities instituting the tax at this time tended not only to be larger than average in population size but also were typically located near other cities that also levied sales taxes. Ultimately, a number of perceived problems with locally administered sales taxes led the state legislature to pass the Bradley-Burns Uniform Sales and Use Tax Law in 1955 (taking effect January 1, 1956). These difficulties included inefficiencies in local administration, complications for retailers in complying with varied local rates and ordinances, worries about locational distortions for businesses because of varying rates, and variations in the exemptions from sales taxes granted by localities for certain goods. The Bradley-Burns law created a uniform local sales tax rate of 1 percent throughout California among the cities and counties choosing to levy the tax. All have so chosen. The law also required that sales taxes be collected by the state and be distributed on a situs basis, meaning that the jurisdiction hosting the retail facility gets the entire local share of sales taxes for that establishment. Counties receive the situs revenues only for sales in unincorporated areas. The distribution is different for manufacturers selling to businesses and for mail-order sales and auto 5 leases, where sales and use tax revenues are distributed into a shared county pool. Today, many local government observers refer to the onecent local sales tax simply as the “Bradley-Burns tax.” Thus, the state collects a statewide sales tax and also collects the 1 percent local tax that is levied by local governments. Although the state collection and distribution superficially give this the appearance of a revenue-sharing scheme, it is important to keep in mind that BradleyBurns is actually a locally levied tax. Additional sales taxes subsequently enacted and earmarked for specific purposes have complicated the picture further. Table 1.1 illustrates the way sales tax revenues are distributed in California. The state’s overall sales tax rate varies between 7.25 and 8.5 percent, depending on where the sale takes place. Of this amount, only a small to medium share goes to local governments. One cent for each dollar in sales goes to the general fund of the city in which the sale occurred—or, in unincorporated areas, to the county. A one-quarter cent share is dedicated to transportation development funds (mostly targeted to mass transit) in each county, under the provisions of the Transportation Development Act of 1972.5 One-half cent is distributed to counties (and some cities) for public safety programs, under the provisions of Proposition 172, adopted in 1993.6 An additional half cent goes to ____________ 5Technically, the 0.25-cent transportation sales tax is also considered a part of the Bradley-Burns tax. However, since it is distributed to counties and dedicated to a single purpose, we do not consider it in our analysis. 6Counties are allowed to shift other funds away from public safety, so Proposition 172 money is more akin to a general revenue (Legislative Analyst’s Office, 1998, p. 12). Counties are required to direct a small share of Proposition 172 funds (currently about 6 percent) to cities as a small way of making up some of the state’s early-1990s property tax shift. Overall, however, Proposition 172 amounts are not subject to the situs rule within cities and are not dealt with in this report. 6 Table 1.1 Components of California’s Overall Sales Tax Rate Rate, % 6.00 5.00 0.50 0.50 1.25 1.00 0.25 Up to 1.25 7.25 to 8.50 Purpose State sales tax, consisting of: State general fund Local Revenue Fund—distributed to counties for health and welfare responsibilities Public Safety Fund—distributed to counties, some cities Bradley-Burns sales tax, consisting of: Local sales tax—directed to general fund of jurisdiction where sale occurred Local transportation tax—directed to county where sale occurred Local special taxes, generally for transportation— optional, require voter approval, used in 24 counties and a few cities (note: in most counties, the maximum rate authorized is 1.50 percent). Total rate SOURCES: Adapted from State Board of Equalization (1998), p. 25; California State Controller (1995–96), p. ix; and the California Revenue and Taxation Code, Section 7251. counties to help carry out their health and welfare functions. Finally, some local governments levy an additional sales tax of up to 1 percent for special purposes—mainly transportation—subject to the approval of their voters (1.25 percent in the unified City and County of San Francisco). In this study, we focus upon the uniform local sales tax rate of 1 percent. For this principal portion of the Bradley-Burns tax, revenues return to the jurisdiction of sale. In a legal sense, the city sales tax is considered a credit against the countywide 1 percent rate. If a city chose to forgo this credit, the funds collected would remain with the county government. Not surprisingly, all cities in the state take advantage of this option, although some cities receive slightly less than the full 1 percent 7 rate because of tax-sharing agreements with their counties. This 1 percent sales tax is particularly attractive to cities because the use of the resulting revenues is discretionary—not earmarked by the state for specific, required activities. Contemporary Importance of the Sales Tax to Cities The passage of the Proposition 13 property tax limitation in 1978 compelled local governments to take a long look at alternative sources of revenues beyond the property tax. Later, the state’s massive shift of property taxes from cities and counties to school districts in fiscal years 1992–93 and 1993–94 during the state’s own budget crisis only hastened this search. The loss of control over property tax revenues has been particularly difficult for cities because those revenues represented a major source of discretionary, general-fund revenues. Today, city officials lack discretion over the spending of much of their revenue base. The League of California Cities (n.d., p. 2) has estimated that, as of 1994–95, only 36 percent of city revenues were not restricted by law to specific purposes. Using a different methodology, Shires (1999, pp. 32–34) concluded that cities in 1995 had discretion over 44 percent of their revenues (excluding bond proceeds), down from 49 percent in 1978. In either case, the message is the same: Cities lack flexibility over how more than half of their revenues are spent. Cities’ three main options for augmenting revenues in the post-1978 years have been to impose new general taxes, service-related fees and charges (including those on developers), and to encourage sales-taxoriented retail development. Growth in general taxes such as business license taxes, franchise taxes, utility users’ taxes, and transient lodging taxes was indeed significantly greater than growth in property or sales tax 8 revenue in the period after Proposition 13 (Shires, 1999, p. 46). The growth in this revenue stream, however, may have been limited by the passage of Propositions 62 and 218, which require majority voter approval for new or increased general taxes. The second option, fees and charges, has also grown in relative importance. However, fees and charges typically are limited by law to funding the cost of providing the service in question or mitigating the costs of development—although with some creativity they can sometimes be used to subsidize a city’s general-fund operations. The limitations on these two revenue options have left the local sales tax as one of the few general-purpose revenue sources with much growth potential to remain under city officials’ control.7 “Control” here, however, is indirect, operating through a city’s land-use policies. Sales taxes in California amounted to 9.6 percent of city revenues in fiscal year 1995–96, although this figure differs substantially from one jurisdiction to another. The vast majority of counties received far less than the average city in per capita sales tax revenues. Recall that counties are allocated only the sales taxes generated by sales in unincorporated areas, which typically are less urbanized than cities. Therefore, counties receive only 1.4 percent of their overall revenues from the sales tax, although such revenues are in many cases important to counties for the provision of “municipal”-type services in the unincorporated areas— services that many observers argue are underfunded. Given that cities rather than counties are most heavily invested in the competition for sales tax revenues, we have chosen in this report to focus on the former. In doing so, we concentrate on two main policy issues: the disparities in ____________ 7Cities may also turn to redevelopment to maximize the amount of property tax increment flowing to the city, but this revenue flow is earmarked. 9 sales tax revenues across cities and the effects of local sales tax competition on land-use decisionmaking. Policy Issue I: Disparities Among Communities in Revenues Sales taxes are important sources of revenues for city governments. But some cities have done better than others at capturing taxable sales. Under the situs rule, there are benefits to jurisdictions that manage to capture large shares of the taxable sales in their market area. The variation among cities in sales tax revenues is enormous, ranging from $2.25 per capita in Rolling Hills to $56,891.84 per capita in Vernon in fiscal year 1995–96. Table 1.2 lists the top and bottom ten cities in this sales tax “hierarchy.” Cities that serve as retailing centers in their region frequently enjoy far higher per capita collections of local sales taxes than ordinary jurisdictions do. Historically, central cities, with their downtown Table 1.2 Cities with Highest and Lowest Sales Tax Revenues per Capita, 1996 Ten Cities with Highest Revenues Vernon Industry Sand City Colma Irwindale Santa Fe Springs Signal Hill Commerce Emeryville Brisbane Per Capita Sales Tax Revenues, $ 56,891.84 30,130.96 6,304.87 4,400.14 2,002.12 1,172.58 841.73 824.05 799.93 515.94 Ten Cities with Lowest Revenues Rolling Hills Bradbury Tehama Hidden Hills Monte Sereno Hillsborough Canyon Lake Atherton La Habra Heights Avenal Per Capita Sales Tax Revenues, $ 2.25 2.57 4.19 4.54 5.14 5.31 5.71 5.92 6.68 9.71 SOURCES: Calculated from sales tax revenue data in California State Controller (1995–96), and annual city population estimates of the California Department of Finance, Demographic Research Unit. 10 shopping districts, served as the commercial centerpieces of metropolitan areas. Today, however, it is frequently suburban cities with major shopping malls, such as Irvine or Pleasanton, that are the focus of the retail market. In addition, some relatively small cities that specialize in retail activities—or host manufacturers selling equipment to other businesses—collect vast sums of sales taxes per capita, such as Irwindale, Vernon, Colma, and Sand City.8 These retail-centered economies can, in turn, lead to further concentrations of retail in such places. In short, revenue “winners” and “losers” have emerged in the local sales tax game.9 A major task of this report, therefore, will be to examine the disparities among cities in per capita sales tax revenues. What factors influence these disparities? What types of cities are advantaged and disadvantaged by the situs rule? Policy Issue II: Accusations of Fiscalized Land-Use Decisions In the wake of voter initiatives such as Propositions 13 and 218, which put serious constraints on the local property tax, the sales tax has been one of the few major revenue sources subject to at least partial control of local officials. By influencing land-use decisions, cities and counties capture greater or lesser shares of regional retail trade—and thus, greater or lesser shares of taxable sales transactions. Local ____________ 8Miller (1981, Chap. 3) shows how some of these types of cities, such as Industry and Commerce, were founded in the 1950s and 1960s as property tax havens for business landowners. Out-of-town consumers paying sales taxes provided disproportionate support for the local treasury. 9As we will discuss in Chapter 6, however, retail concentrations typically increase the amount of public services that local government must offer, because of the added “daytime population” and activity associated with retailing. Although we cannot measure these additional costs in this report, it is important to keep them in mind when evaluating the distribution of sales tax revenues across cities. 11 governments may favor commercial over other types of development and may extend services and boundaries considerably to encompass actual or potential retail growth. In short, overall city development strategies and plans may be heavily affected by the search for land uses that are lucrative in the amounts of revenue they contribute. These considerations suggest an emphasis on the “supply side” of the local land market, as opposed to the demand-side responses of customers and retailers. Local governments can use their land-use powers of zoning and subdivision regulation, as well as their ability to build infrastructure, to reward some types of growth and penalize others. In this fashion, land use is determined politically as well as economically (Lewis, 1996). The presence of local sales taxes may encourage competition among local governments for retail developments, may lead to the costs of retail developments being externalized onto local governments, and may lead cities to favor retail over other types of development that would not generate sales taxes (LAO, 1994, p. 125). A former legislative representative of the California State Association of Counties has argued, “It is . . . clear that the situs allocation of sales tax greatly encourages cities to pursue commercial development over both industrial and residential development. This incentive is so strong that cities and counties actually attempt to ‘steal’ sales tax generators from other localities” (Senate Local Government Committee, 1989). Or as a Los Angeles Times reporter put it, cities are “pursuing sales tax dollars at a fever pitch” (Shuit, 1998). Most observers stress that this is not a good long-term economic strategy, as local governments focus their recruitment efforts on businesses like auto malls and big box retail stores that do little to expand the economy. Developers and retailers have been 12 known to play one jurisdiction off against another, seeking the most advantageous “deal” for their projects. This issue is the so-called fiscalization of land-use decisionmaking. It is the second policy focus of this report. Recent State Legislative Proposals to Address Sales Tax Concerns At least three recent proposals from the state legislature have attempted concrete reforms of the sales tax allocation system. Assemblymember Valerie Brown introduced AB 3505 in 1994, which would have altered the distribution of the local sales tax by basing it more on cities’ relative population, as opposed to a strictly situs-based system. Although the bill did not pass, it did generate a dialogue about the possible merits of sales tax sharing. AB 1835, a 1998 bill sponsored by Assemblymember Tom Torlakson, attempted a ban on the sales tax “wars.” The bill would have prohibited the use of public funds by local governments attempting to lure retail businesses from other localities within any given market area. This bill was in part a response to events in Martinez, in Contra Costa County, which reportedly received an ultimatum from a large local discount retailer that the store would move to a nearby city if Martinez did not pay $2 million in costs connected with the planned expansion of the store property. The retailer also reportedly sought $200,000 in rent per year from the city for the privilege of using its parking lot during hours when the store was closed (Newman, 1998, p. 2). This bill passed the Assembly, but died in the state Senate, where the League of California Cities as well as auto dealers lobbied against it (Shuit, 1998). 13 In 1999, Torlakson offered a similar proposal, AB 178, which is pending as of this writing. Proposition 11, which appeared on the November 1998 ballot, was an attempt to help cities stave off ruinous competition. A constitutional amendment authored by Assemblymember George Runner, the proposition allows any two or more cities or counties to negotiate sales tax revenue-sharing agreements, with a two-thirds vote of their governing boards.10 Previously, local governments were permitted to engage in such revenue-sharing only with an affirmative popular vote, a requirement seen as stymieing possible cooperative efforts. The proposition, which passed with 53 percent of the vote, will enable some experiments in cooperation among local governments, but many informed observers say that its effects are likely to be modest. Local governments that are “doing well” in generating sales tax revenues will have little incentive to engage in such agreements. Organization of This Report The remaining chapters of this report provide evidence that should be essential to any further discussions about reforming the local sales tax. Chapter 2 looks at general trends in local sales tax collections and the tax’s place in cities’ overall revenue picture. Chapter 3 presents statistical analyses of sales tax data that allow us to draw conclusions about the types of cities that are doing best and worst under the situs rule. The subsequent two chapters address the issue of fiscalized land-use decisions. Chapter 4 reviews the allegations that have been made against city ____________ 10There are useful analyses of Torlakson’s AB 1835 and Runner’s ACA 10 by the Senate Local Government Committee and other committees. These may be accessed at http://www.leginfo.ca.gov/. 14 development decisions regarding the sales tax and considers the conceptual issues involved. Chapter 5 discusses results from a survey of city officials regarding the degree to which they seek retail development as opposed to other types of land uses, and the degree to which sales tax motivations influence their views toward development, redevelopment, and annexation. Finally, Chapter 6 considers the policy implications of our findings and briefly highlights some alternative policy proposals that have been discussed by those who would reform the situs-based local sales tax. 15 2. Trends in City Sales Tax Revenues How important is the local sales tax to city budgets? Clearly, the answer is different for each city. The sales tax is an almost trivial revenue source to a few, a leading source for others, and a significant but not overwhelming source for most. In this chapter, we use data calculated from the California State Controller’s annual volume, Financial Transactions Concerning Cities of California, to help readers get a more informed sense of the relative importance of the sales tax over the period since 1970. This time period allows us to examine patterns before and after Proposition 13, generally considered a watershed in local public finance. We present data on overall sales tax collections, the disparity among cities, and the patterns in specific subtypes of cities. 17 Sales Taxes as a Component of Cities’ Revenue Streams Aggregating revenue data across all cities, we can get a better sense of the overall role that the local sales tax plays. Figure 2.1 charts the various categories of municipal revenues over the period 1970 to 1995 in percentage terms, looking at the various revenue sources as shares of total revenue in all California cities. Sales taxes are clearly not an overwhelming source of city funds in comparison to other types of revenue, but they are a significant slice of the revenue pie. And despite the increased attention to the sales tax since the Proposition 13 property tax limitation in 1978, the sales tax has been a remarkably consistent share of overall revenues. Public service enterprises Licenses, fees, and assessments Subventions Property taxes Sales taxes 40 30 Percentage 20 10 0 1970 1975 1980 1985 1990 SOURCE: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual). 1995 Figure 2.1—Revenue to California Cities, by Source (as a percentage of total revenue) 18 Note that after the passage of the Proposition 13 property tax limitation in 1978, sales taxes surpassed property taxes as a share of city revenues—a situation that persisted until 1990. In the early 1990s, the sales tax once again approached parity with property taxes. This was due in part to the state’s shift of property taxes away from cities to schools in this period and also in part to a deep recession. The recession held down increases in property taxes, as fewer new structures were built or changed hands and many property owners won downward reassessments. The importance of sales taxes is all the more clear when one considers the issue of spending discretion associated with each revenue source. Many other leading revenue sources have their proceeds restricted to certain types of uses. For example, city enterprise revenues—the rates and service charges paid by users of city-provided water, electricity, or other such utility-like functions—normally must flow back into the operation of the enterprise in question. Similarly, many state subventions and federal grants are earmarked for particular functions or programs. Thus, sales and property taxes (along with Vehicle License Fee revenues) are the major single sources of discretionary income that may be used for general purposes by cities. This makes them particularly attractive and sought after forms of revenue. Since the passage of Proposition 13, however, cities have been very limited in their ability to raise new revenues from the property tax. This is because property tax rates have a ceiling of 1 percent, properties are reassessed only when sold, and the division of the property tax dollar among local governments is set by a state formula—one that often provides a relatively small share to cities.1 Thus, the sales tax has taken ____________ 1In aggregate, about 14 percent of the property taxes paid by California residents goes to cities, with the rest going to schools, special districts, and county governments. 19 on increasing significance since 1978, despite its relatively flat trend in revenues per capita. Figure 2.2 takes a closer look at this trend for selected years since 1970–71. The bottom line on the graph, with sales taxes expressed in nominal dollars per capita (that is, unadjusted for inflation) shows almost continual growth in revenues until the early-1990s recession caused the retail sector to slow. In constant 1995 dollars, however (the top line), we can see that sales tax revenues have been a somewhat stagnant source of funds over this period, generally hovering between $100 and $120 per Revenue (dollars) 140 120 100 80 60 Constant (1995) dollars 40 Nominal dollars 20 1971 1976 1981 1986 1991 1996 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.2—Sales Tax Revenues per Capita for All California Cities 20 capita.2 Although these revenues grew throughout the 1970s, the early1980s recession saw declining sales per capita—and thus declining sales tax revenues. There was only a muted recovery by 1985, and another decline in the period since. These trends illustrate the sales tax’s sensitivity to broader economic conditions and trends. It is well known that “the retail sector is among the most cyclically sensitive segments of the economy” (Kroll and Marrinan, 1985a, p. 1). A factor that may be of even more long-term significance than the business cycle is the changing nature of consumer purchases themselves. Many analysts have pointed out that an increasing share of Americans’ disposable income in recent years has been spent on items not typically subject to local sales taxation—personal services, mailorder sales, and purchases via the Internet. In fact, the California Budget Project has found that taxable sales as a proportion of personal income in the state fell by more than a third between 1950 and 1995. In addition, the state legislature has carved out numerous exemptions to the sales tax for various products over the years, ranging from diaper services to cattle feed to magazine subscriptions (see Benson, 1997). These trends have further constrained California’s local governments’ ability to squeeze any growth from this desirable revenue source. Not surprisingly, the ratio of local sales tax revenues to personal income in cities has declined since the 1970s (Coleman, 1998, p. 24). Thus, the battle among cities for sales tax dollars is highly competitive. Not only are retail sales for a given population relatively fixed in amount, but the overall amount of sales per capita subject to the ____________ 2Throughout this report, dollar figures are adjusted for inflation by using the California Consumer Price Index, which is calculated by the California Department of Labor and Industrial Relations, Division of Labor Statistics Research. 21 sales tax levy appears to be in long-term decline. Thus, cities that “succeed” in recruiting retail businesses within their borders can generally be viewed as simply shifting retail sales geographically within a region. In short, the competition for retail appears to be a zero-sum game (Senate Local Government Committee, 1989). Revenue Gainers and Losers As Figure 2.2 shows, experiences with the local sales tax clearly were quite different in the 1970s from the 1980s, with overall per capita growth in revenues in the earlier decade and stagnation in the latter decade. How many cities gained and lost ground over this period? How much commonality was there in city experiences? Table 2.1 shows the number of cities increasing and decreasing their real per capita sales tax revenues in the period from the early 1970s to the early 1980s, in the decade that followed, and over the entire period. In this case, we have used three-year averages of sales tax revenues and population for each city: fiscal years 1970–71 through 1972–73, Table 2.1 Changes in Real per Capita Sales Tax Revenues Between 1970s and 1990s Among California Cities No. of cities gaining Median amount gained No. of cities losing Median amount lost 1971–73 to 1981–83 315 $19.04 88 $8.32 1981–83 to 1991–93 181 $14.59 245 $17.93 1971–73 to 1991–93 234 $30.83 169 $15.22 NOTE: Revenues measured in constant 1995 dollars. SOURCES: Calculated from California State Controller, Financial Transactions Concerning Cities of California (annual); and annual city population estimates from the California Department of Finance, Demographic Research Unit. 22 1980–81 through 1982–83, and 1990–91 through 1992–93. (For the sake of simplicity, henceforth in this report, we will refer to fiscal year 1970–71 simply as 1971, 1994–95 as 1995, and so on. California’s fiscal years actually run from July 1 to June 30.) We have chosen to take three-year averages because individual cities tend to have a fair amount of variation in sales tax revenues per capita from year to year; a three-year average is a more stable and reliable measure than a single-year measure and is less sensitive to business-cycle swings. Focusing on the time periods immediately after the decennial federal Census (1970, 1980, 1990) will allow us to use Census data to analyze city characteristics. As the table indicates, the vast majority of cities—315 of 403, or 78 percent—were better off in the early 1980s than a decade earlier, by this measure. Moreover, the average amount of increase in per capita sales tax revenues for the revenue gainers was far greater than the median amount lost by the revenue losers. In the following decade, however, most cities reversed course, with 245 of 426 (58 percent) losing ground—and the losing cities were losing more, on average, than the winners were gaining. Over the entire 20-year period, most cities—58 percent—gained some ground.3 It is possible that these numbers mask a growing disparity among cities. That is, it may be the case that as competition for retail businesses heated up, the sales tax “winners” became significantly better off than the sales tax “losers.”4 In attempting to measure this phenomenon, however, ____________ 3To be counted in the table, the city must have been in existence both at the beginning and at the end of the relevant period. 4Although we sometimes use the terms “winners” and “losers” in this report as a convenient shorthand for cities’ levels of per capita sales tax revenues, we do not mean to imply that the former group of cities has employed “better” policies than the latter. Indeed, one can readily conceive of circumstances under which cities make short-term- 23 a key issue is to define the winners and losers. Some traditional measures of variance—highest versus lowest, or standard deviations, for example— are potentially misleading in this context, because they are highly sensitive to outliers or extreme observations. To correct for this problem, we calculated a statistic called the interquartile range (IQR) among cities for each year measured. The IQR simply measures the distance between the 75th percentile of a distribution and the 25th percentile—in this case, the gap between the lowest city in the top quarter of the sales tax success list and the highest city in the bottom quarter. Figure 2.3 displays the IQR over the period 1971 to 1995. It shows a widening dispersion (or “spread”) among cities through the 1970s, with a flatter trend in the period since. Thus, cities were becoming more unequal in their success at attracting sales tax revenues until about the time of Proposition 13 (1978), with no lasting increase in city inequality in the period since.5 Note, too, that the amount of dispersion decreased during the recessions of the early 1980s and 1990s. One possible interpretation for the convergence during these periods is the following: When the economy is expanding, consumers are willing to buy more expensive and luxury goods (automobiles, jewelry, and so on). Sales tax revenues from such purchases may tend to accrue to certain cities specializing in those retail activities. During recessionary periods, ________________________________________________________ oriented, even ill-conceived, land-use decisions that might help them land in the “winner” group. 5As a check of these conclusions, we also examined the percentage of each city’s own-source, general-purpose revenues accounted for by the sales tax in 1973, 1983, and 1993. We then calculated the standard deviation in this share across all cities. The standard deviation was 15 percent in 1973, 13 percent in 1983, and 12 percent in 1993, again indicating that disparities among cities in sales tax reliance were level or slightly decreasing. 24 Interquartile range (in 1995 dollars) 100 90 80 70 60 50 1971 1974 1977 1980 1983 1986 1989 1992 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.3—Dispersion of Adjusted per Capita Sales Tax Revenues Among California Cities, 1971–1995 however, consumers defer making such purchases, and thus the fortunes of these “winner” cities decline relative to other places, making the overall distribution more equal. It is also conceivable that the overall flatness of the IQR measure since 1978 may indicate that the enhanced competition for retailers among cities since Proposition 13 (if the enhanced competition in fact exists) is driving worse-off cities to offer more incentives to retailers and developers. If so, these actions might shift retail sales, at the margin, to places that were falling behind. Changes in the Sales Tax Hierarchy over Time Scatterplot diagrams may also be used to illustrate cities’ performances over time. These have the advantage of showing each 25 Log of per capita sales tax revenues, 1995 individual city’s performance, rather than aggregating them together. In this series of scatterplots, we have used the natural logarithm of sales taxes per capita, rather than the actual amounts of dollars per capita, as a way to improve the graphic presentation of the data.6 Figure 2.4 plots each city’s position in 1971 on the horizontal axis and its 1995 position on the vertical axis. What emerges quite clearly from this picture is the relative stability of the sales tax hierarchy. That is, cities that enjoyed high receipts at the beginning of our study period 8 7 6 5 4 3 2 1 0 012345678 Log of per capita sales tax revenues, 1971 (in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.4—Cities’ Sales Tax Revenues per Capita, 1971 and 1995 ____________ 6Because some cities are extreme outliers, boasting thousands of dollars of sales tax revenues per capita, and most other cities are in the lower ranges, with dozens of dollars per capita, it would be difficult—and not very helpful—to show them all on the same graph. Taking the natural log of these values, a standard technique in the social sciences for transforming skewed data, has the effect of “pulling in” very high values to form a more normal distribution. 26 were still in an advantageous position in the early 1990s, and the lowerranking cities tended to stay in the low ranges. Using this technique to examine the data, then, it would appear that cities can generally expect to maintain their relative positions in retail markets, if past trends hold. This is not to say, however, that there are no factors systematically affecting the capacity of cities to improve or worsen their positions in attracting retail sales. Chapter 3 will illustrate the combination of factors that influence cities’ performance over time. Were cities that started the sales tax “race” in an advantageous position able to gain more ground over time relative to the other cities? No, as we have seen in the discussion of the IQR. Figure 2.5 is a Change in log of per capita sales tax revenues (log of 1991–93 average minus log of 1971–73 average, calculated in 1995 dollars) 3 2 1 0 –1 –2 –3 –4 01 2 3 4 5 6 7 Log of per capita sales tax revenues (average of 1971–73, calculated in 1995 dollars) 8 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.5—Relationship Between Cities’ Sales Tax “Success” in the Early 1970s and Subsequent Gains 27 scatterplot presenting the same basic idea in a different manner. On the horizontal axis, we plot the log of each city’s per capita sales tax revenues in the 1971–73 three-year average period. On the vertical axis, we plot each city’s “success” over time—that is, the log of its early-1990s revenues minus the log of its early-1970s revenues. (Note that some cities have negative values on this measure; they worsened over time in per capita sales tax collections, in real-dollar terms.) The data points do not show an increasing relationship between a high standing in the early 1970s period and subsequent gains. If anything, the scatterplot appears to show a negative relationship between early success and subsequent gains—although there is a great deal of variation. Moreover, cities’ performances through the 1970s do not appear to be related to their performances through the 1980s, as Figure 2.6 reveals. This scatterplot graphs 1970s gains or losses on the horizontal axis and 1980s gains or losses on the vertical axis, again using changes in natural logs for illustrative purposes. The data points appear as a random cloud, indicating that there is no discernible relationship between individual cities’ performances in the two decades. Regional Trends Sales tax patterns among different types or categories of California cities may also be compared. One such comparison is the distinction between municipalities in the two major metropolitan areas of the state. Figure 2.7 displays per capita sales tax collections for cities within the five-county region centered on Los Angeles and the nine-county San Francisco Bay area. Here we can see a clear divergence in the fortunes of these two regions, with the Bay area overtaking the Southland in the early 1980s, and a widening disparity between the two areas in the period 28 1980s changes (log of 1991–93 average per capita sales tax revenues minus log of 1981–83 average per capita sales tax revenues, calculated in 1995 dollars) 2 1.5 1 .5 0 –.5 –1 –1.5 –2 –1 –.5 0 .5 1 1.5 2 1970s changes (log of 1981–83 average per capita sales tax revenues minus log of 1971–73 average per capita sales tax revenues, calculated in 1995 dollars) SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population data from the California Department of Finance, Demographic Research Unit. Figure 2.6—Relationship Between Changes in Cities’ Sales Tax “Success” over the 1970s and 1980s since. The Los Angeles area experienced deeper declines in retail sales per capita during the two major recessions of this period, and the Bay area was also quicker to recover from these troughs in the business cycle. Overall, the improving fortunes of the Bay area relative to Los Angeles during this period can largely be attributed to the former region’s significantly increasing level of per capita income, relative to Los Angeles. The decline of the aerospace industry in Southern California in the 1990s also may have particularly depressed the sales of goods from that region. 29 Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Los Angeles region San Francisco Bay area 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.7—Sales Tax Revenues per Capita in Major Metropolitan Areas Although the graph shows the differing long-term trends in these two regions, it is important to emphasize that the experience of individual cities within each region varies a great deal. Figures 2.8 and 2.9 are maps of the central portions of the Los Angeles region and Bay area, showing each city’s 1995 level of per capita sales tax revenues. Although the maps illustrate that there are proportionally more Bay area cities with high levels of sales tax revenues, both regions exhibit much variation in the performance of individual cities. Another interesting impression one can gain from the maps—a not unexpected finding—is that virtually all of the “top performing” sales tax cities in these two regions are located along major highways, which are displayed in the maps. The accessibility advantages of the freeway can help make an area ripe for modern auto-oriented retail. However, it would be premature to 30 ,,,,,PacificOcean ,,,Lessthan$50 $50 to $99 $100 to $149 ,$150orgreater Figure 2.8—Per Capita Sales Tax Revenues in the Los Angeles Area, 1995 conclude from these maps that freeways “cause” higher levels of retail sales relative to other areas without highway access. After all, it may well be the case that highways were routed in such a way as to connect areas of heavy commerce, so that to some degree highways resulted from retail clusters, as opposed to stores clustering around highways. In Figures 2.10 and 2.11, we present complementary maps that show each city’s change in per capita sales tax revenues between the early 1970s (average of 1971, 1972, and 1973) and the period 20 years later. 31 ,,,PacificOcean San Francisco Bay ,,,Lessthan$50 $50 to $99 $100 to $149 , ,$150orgreater Figure 2.9—Per Capita Sales Tax Revenues in the San Francisco Bay Area, 1995 Again, the Bay area has more cities with gains over this period (darker shaded areas), particularly in the South Bay region, where wealth increased markedly as Silicon Valley industries developed. And once again, there are noticeable clusters of “gainer” cities along some of the major highways. 32 ,PacificOcean ,,,Lessthan–$15 –$15 to 0 $1 to $14 ,,$15to$44 $45 or greater Figure 2.10—Changes in per Capita Sales Tax Revenues in the Los Angeles Area (Average of 1971 Through 1973 Subtracted from Average of 1991 Through 1993) Trends in Central Cities Another comparison of interest involves the relative fortunes of older central cities in relation to other types of jurisdictions. As California and the nation have increasingly suburbanized—both in population 33 ,,,,,,PacificOcean San Francisco Bay ,,,,,Lessthan–$15 –$15 to 0 $1 to $14 $15 to $44 , ,$45orgreater Figure 2.11—Changes in per Capita Sales Tax Revenues in the San Francisco Bay Area (Average of 1971 Through 1973 Subtracted from Average of 1991 Through 1993) and economic functions —we might expect the traditional business centers to experience a relative decline in their retail strength. A difficulty involved in this comparison involves a definitional issue: Which cities qualify as “central” cities? To avoid making our own subjective judgments, we relied upon the designation of central cities by the U.S. Census Bureau. There are 36 such cities in California. Some 34 might object that the list is overly broad, since it includes cities such as Petaluma and Seaside along with more traditional older cities, but the Census Bureau does at least use consistent reasoning in identifying such places. The Bureau has identified one or more traditional central places—historic business and population centers of long standing— within each designated metropolitan statistical area of the state.7 Figure 2.12 compares per capita sales tax revenues for these 36 cities to the other cities of the state. Here again, we see the changing fortunes Revenue (in 1995 constant dollars) 140 130 120 110 100 90 80 Central cities All other cities 70 60 1970 1975 1980 1985 1990 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.12—Sales Tax Revenues per Capita: Central Cities Compared to Other Cities, 1971–1995 ____________ 7The central cities are Anaheim, Bakersfield, Chico, Fairfield, Fresno, Lompoc, Long Beach, Los Angeles, Merced, Modesto, Monterey, Napa, Oakland, Oxnard, Petaluma, Porterville, Redding, Riverside, Sacramento, Salinas, San Bernardino, San Diego, San Francisco, San Jose, Santa Ana, Santa Barbara, Santa Cruz, Santa Maria, Santa Rosa, Seaside, Stockton, Tulare, Vallejo, Ventura, Visalia, and Yuba City. 35 of different categories of communities. The central cities began the period with slightly higher sales tax receipts per capita than other places but were overtaken in 1973. The disparity has grown wider in the period since. Central cities appear to have been less robust than other cities in their emergence from the two recessions. Overall, the movement of commerce outward from traditional downtowns toward suburban freeway locations is probably at the root of these trends.8 The Experience of Newly Incorporated Cities A number of observers have alleged that many new cities in California have incorporated as a way to capture sales tax revenues in their area, for local use—thus denying these revenues to the broader unincorporated county area. Although some examples of this cityformation-for-profit phenomenon undoubtedly exist (Miller, 1981), overall there has been no noticeable upsurge in incorporations over the past 30 years (Lewis, 1998). In Figure 2.13, we compare the per-capita sales tax revenues for 1981–1995 of cities incorporated since the passage of Proposition 13 in 1978 with those of cities existing before 1978. (The number of post1978 cities, which grows over time, is displayed on the appropriate line on the graph.) With property tax rates limited by that proposition—and disadvantageous property tax distribution formulas typical for new ____________ 8It is worth noting, however, that “central cities” cannot be equated with the fortunes of downtowns. Many California central cities are quite extensive in territory and include a large amount of suburban-style shopping strips and centers. The inclusion of cities like Anaheim or Fairfield that many observers would label suburbs—these cities have only a small traditional “downtown” presence, in comparison with their more modern commercial areas—makes the relative decline of the 36 central cities all the more noteworthy. 36 Revenue (in 1995 constant dollars) 140 120 100 80 15 60 11 40 9 cities 20 1981 1983 21 30 40 48 51 53 53 Post-Proposition 13 cities Other 417 cities 1985 1987 1989 1991 1993 1995 SOURCES: Calculated from the California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues); and annual city population estimates from the California Department of Finance, Demographic Research Unit. Figure 2.13—Sales Tax Revenues per Capita: Cities Incorporated after Proposition 13 Compared to Other Cities cities—we might expect new cities in this period to pay especially close attention to sales tax revenues. But as the graph shows, the new cities are significantly worse off in sales taxes, on average, than the pre-1978 cities. Clearly, creating a new city is no guarantee of fiscal bliss—even though incorporation proponents pay close and careful attention to the new city’s boundaries, typically trying to include as much taxable retail as possible. Several factors militate against high sales tax levels in these new cities. The first fact to consider is the typical position of these cities within their regional context. Typically, inner areas, which would be more likely to have developed a longstanding retail base, are already incorporated, whereas newly incorporated cities tend to be at the outer fringe of 37 metropolitan areas. These fringe areas tend to be heavily residential, with commercial properties few in number and a limited population base in the areas surrounding the new city. Thus, we would expect them to have less retail sales per resident. Another way of expressing this phenomenon is that areas that could experience a windfall gain of tax revenues from incorporation presumably would have already incorporated some time ago; thus, the areas remaining unincorporated in recent years will tend to be more marginal retail locations. Other factors, more related to intergovernmental politics, may also help explain the relatively poor sales tax success of new cities. One is the strategic response of county governments (and to a lesser degree, existing cities) to the proposed formation of a new city. Representatives of counties and existing cities, acting through the Local Agency Formation Commission (LAFCO) in their county, have some veto powers over the formation of new local governments. Incorporations that would seriously weaken the county fiscally (and thus potentially hinder county service provision even in existing incorporated areas) may be vigorously opposed by county delegates on LAFCO boards. Thus, some incorporation proposals that would result in large shifts of sales tax revenues to new cities are likely to be quashed. Moreover, state law dictates that the new city must cement a mutually agreeable tax distribution agreement with its county before a LAFCO can approve an incorporation proposal. As counties have become considerably more fiscally stressed since Proposition 13 and the early-1990s property tax shift, these negotiations are often protracted. The negotiations have 38 resulted in significant concessions by the new cities, whereby a large share of sales taxes in the new city will be passed on to the county.9 Summary Sales taxes are a fairly stagnant source of revenues, overall, for California cities. Nevertheless, some cities have done much better than others. Central cities have tailed off in per capita sales tax revenues relative to other communities, for example, and Bay area cities have continued to gain relative to Los Angeles area cities. There is, however, no evidence of an increase in overall disparities among cities over time and nothing to support the idea that cities that were leading the pack in sales tax collections a decade or two ago have pulled still farther ahead since then. Nor can one make the case that cities newly incorporated after Proposition 13 have an advantageous position with respect to sales taxes. In the next chapter, we will look more systematically at the characteristics of cities that are associated with “success” in the competition for sales tax revenues. ____________ 9Similar agreements are required when a city wishes to annex land. The city and county must work out a tax-sharing arrangement. These negotiations have become increasingly contentious in recent years. 39 3. Which Types of Cities Have Benefited from the Situs Rule? We have no property tax, we have no lighting tax, we have no library tax, we have no utility users tax . . . all because of the sales tax and the economic base we have developed (City manager of Cerritos, quoted in Shuit, 1998). In Chapter 2, we looked at some general patterns relating to cities’ “success” in receiving sales tax revenues. This chapter looks more systematically at the relationship in a recent period between various city characteristics and levels of per capita sales tax collection. We also examine the factors related to cities’ changing fortunes over time, looking at the period from the early 1970s to the early 1990s. Variations Across Cities: The Extremes At the time our data were gathered, the most recent and complete local fiscal data available from the State Controller’s office were from Fiscal Year 1994–95. (Again, for convenience we will refer to this as 1995, and so on for other fiscal years.) In that year, sales tax revenues for 41 California cities ranged from $3.12 per capita in Hillsborough, a residential Bay area enclave, to $50,999.52 per capita in Vernon, a commercial city in Los Angeles county with just 80 residents in that year. The median or “typical” city received $85.52 in sales taxes per capita. Counties, meanwhile, which receive situs-based sales tax revenues only for sales in their unincorporated areas, fared worse than most cities. This finding reflects the relatively weak presence of retail in outlying areas. The median county raised $42.32 per unincorporated area resident, with a range from $19.99 in Modoc County to $174.53 in San Mateo County.1 This tremendous range of “success” in generating sales tax revenues is interesting in and of itself, but it also raises a basic question: What types of cities tended to be the most or least successful? One way to examine this question is to look at the “extremes” of the distribution—those places with very low or very high per capita sales tax revenues. Table 3.1 provides some summary statistics regarding demographic and other characteristics of these communities. From the table, we can conclude that, on average, the cities that have been most successful at getting sales tax revenues tend to be older (in terms of incorporation date) and are more likely to be urbanized than the cities with extremely low sales tax revenues per capita. The “top 20” have lower household sizes, a smaller percentage of children in the population, and perhaps surprisingly, a higher percentage of Hispanics. It also may be surprising that the ____________ 1The City and County of San Francisco is treated as a city for purposes of analysis in this report. All population data are from the official annual estimates made by the California Department of Finance, Demographic Research Unit. We do not rely on the annual population totals listed for each city in the California State Controller’s Financial Transactions volumes, as these proved to be inaccurate and not consistent with the Department of Finance population data. 42 Table 3.1 Summary Statistics for Cities with Extreme Levels of per Capita Sales Tax Revenues as of 1995 Population, 1995 Incorporation date Median year housing built Area (square miles), 1990 Per capita income, 1989 Aggregate income in city and 10- mile buffer around it Household size, 1990 % black, 1990 % Hispanic, 1990 % non-Hispanic white, 1990 % under age 18, 1990 % over age 64, 1990 % central cities % in urbanized areas % in Central Valley % in Bay area % in Los Angeles area % of own-source general revenues from property taxes, 1993 % of own-source general revenues from sales taxes, 1993 Average of Top 20 Cities Bottom 20 Cities 12,020 7,460 1929 1944 1961 1965 5.6 13.4 $17,569 $34,277 $24.2 billion 2.7 3.6 29.6 71.5 22.4 13.4 0 85 5 20 50 $12.0 billion 3.1 4.6 22.7 76.0 27.0 11.6 0 58 35 30 30 9.1 25.6 35.9 5.2 SOURCES: U.S. Census (1990); California State Controller, Financial Transactions Concerning Cities of California (1993, 1995). “bottom 20” cities have much higher income levels. Note, however, that the bottom 20 also tend to be located in market areas with much lower levels of aggregate income—that is, purchasing power. A partial explanation for these findings is that several of the “bottom 20” cities are wealthy bedroom communities. These include cities such as Hillsborough and Atherton in the Bay area, and Rolling Hills and Hidden Hills in the Los Angeles area, all of which had per capita incomes 43 of over $63,000 in the 1990 Census. (By comparison, the statewide per capita income was $16,409.) These are relatively low-density suburbs with very little commercial development. We might surmise that the wealthy residents of these areas have preferred to “zone out” retailing to preserve their community character. This is an important point, since it means that the situs rule does not necessarily work to the advantage of the state’s wealthiest communities. It should be pointed out, however, that the “bottom 20” also includes a number of very poor cities with depressed economies, many of which are in the Central Valley. Examples include Avenal, McFarland, Parlier, and Biggs, each of which had a per capita income of less than $8,600 in 1989—about half the statewide average. Many of the “top 20” cities also fall into certain types. At least seven are commercial or industrial enclaves with small populations. These include the top five cities—Vernon, Industry, Colma, Sand City, and Irwindale—none of which had a population above 1,220 in 1995. Others are the commercial centers of regions of tourism with high daytime populations, notably Carmel, Capitola, and Sonora. It is difficult to generalize about the remaining cities in the top 20, other than to say that most of them have fairly small populations. The only very wealthy city in this group is Beverly Hills. It is dangerous to generalize to all California cities from this set of outliers. Nevertheless, differences between the top and bottom 20 do begin to provide some clues about factors that might be important to the overall group. The next section examines the issue more systematically. 44 Variations Across Cities: A Cross-Sectional Model A more thorough approach is to estimate statistical models that can help account for the variations among cities in per capita sales tax revenues. We do not seek to “explain” sales tax revenue-raising success, per se; rather, we wish to discover community characteristics associated with such success. In devising such a model, it is useful to consider the factors that we might expect to influence retailer location decisions. There are two major ways to analyze retail location. One is from the viewpoint of the practitioner, looking at optimal locations for new stores within market areas. National retail chains tend to have their own store location appraisal divisions, and typically have sets of criteria that any location must meet. A second perspective on retail location is that of the academic urban geographer, concerned with the spatial hierarchies of retailing across the landscape (Berry et al., 1988; Davies, 1976; Guy, 1994; see also Schneider, 1986). We reviewed this literature and, where possible, located data that would help replicate these studies for California (see also Kroll and Marrinan (1985a, 1985b). Except where indicated, the source of our data is the U.S. Census. We describe each variable in the analysis below. Readers who are less interested in the estimation strategy may wish to skip ahead to the summary at the end of this chapter. Measuring Sales Tax Revenues: The Dependent Variable Given the skewness of the data, we use the natural log of per capita sales tax revenue in each city as the dependent variable. We have averaged 45 these data for 1991, 1992, and 1993 to improve reliability, since individual cities tend to vary upward and downward a fair amount from year to year. The early-1990s data also have the advantage of being closer in time to the independent variables, most of which are drawn from 1990 Census data.2 We eliminate from the analysis four cities that are extreme outliers—Vernon, Industry, Colma, and Sand City. Each has a tiny population and tremendously high per capita sales tax revenues. City Characteristics: Independent Variables A number of demographic variables are employed to help illuminate cities’ levels of per capita sales tax revenues. One is the log of the 1991 population. A larger population in a city increases its potential purchasing power but also increases the “denominator” for our measure of sales tax revenues per capita. A city may have many residents, but they may or may not shop locally. Thus, the expected effect of population is unclear. Population change between 1985 and 1991 (in percent) is also included, since we felt that high population growth levels may spur additional retail sales because of the need for housing-related products. Household characteristics of the population are potentially important determinants of retail sales per capita. We include as a variable the percentage of senior citizens in each city, since elderly persons typically consume less than younger adults do. Related is the issue of average household size. Larger household sizes imply cities with more children, who tend to consume less than adults. Moreover, some retail goods—refrigerators and furniture, for example—tend to be bought on a ____________ 2However, performing the same analysis on 1995 per capita sales tax data yields extremely similar results. 46 “per household” basis. The literature also suggests that areas with small household sizes may tend to have a greater share of disposable income and be more oriented to consumption goods. Thus, we expect household size to be negatively related to per capita sales tax revenues. (A variable representing the percentage of children in the population also was considered but not included in the model because of its very high collinearity with household size.) A city’s racial characteristics also may be important. We include percentage black and percentage Hispanic as of 1990. Schneider notes that disproportionately black suburbs, in addition to having lower incomes on average, “are also less prestigious than white suburbs, and firms located in black areas will have difficulty attracting white shoppers, who constitute the bulk of consumers in most suburban areas” (1986, p. 27). We have not found any literature suggesting a relationship between Hispanic population shares and retail sales, but we include this variable for its potential importance in the California context. We also include variables related to the potential purchasing power of the city in question. One is the city’s median household income (as of 1989). As we have discussed, cities with higher incomes have more potential purchasing power—but very high-income people often seek “exclusive” residential environments devoid of retailing and other businesses. This relationship between income and sales tax success thus appears nonlinear—that is, increasing incomes should generally add to sales tax success, but beyond some threshold of wealth, cities may have lower sales tax revenues. For this reason, we specify a quadratic form, using household income and household income squared as variables. Potentially just as important is the log of the aggregate income of the market area around each city—a variable specially constructed for this 47 study. Using a computerized mapping program that allowed us to overlay city boundaries on 1990 Census information, we drew a 10-mile wide “buffer zone” around each city, and calculated the aggregate income of the population within this line—that is, in the city and its surrounding buffer zone. This aggregate income measure, which is reasonably though not perfectly estimated, can be thought of as the potential purchasing power of residents who live within a given travel radius around each city.3 Simply put, we expect retailers to gravitate toward locations where there is sufficient buying power to keep their stores afloat. A final factor related to purchasing power is tourism. We include a county-level estimate of 1992 travel-related expenditures (again in log form because of skewed data among the state’s counties). These data are from the Division of Tourism (1998, p. 9). ____________ 3Data complications led us to substitute a slightly different measure of the aggregate income around each city. Our mapping program was unable to construct geographic “buffers” for a number of cities. We developed an alternative method in which a circle was drawn, centered around each city’s centroid. The radius of each circle was determined by adding 10 miles (the width of the buffer in the earlier method) to the square root of the city’s square mileage divided by π. (In short, we made the simplifying assumption that cities were circular in shape and used the equation area = π x radius2 to calculate an approximate city radius, then added 10 miles to that radius figure.) We then used the mapping program to calculate the aggregate income of the population within this circle. As it turned out, using this method gives results that are extremely closely related (r = .9992) to the preferred “buffer” method, in those cities where both could be computed. There is one further caveat to note. The mapping program apportions populations within each census tract to our circle area based simply on the percentage of the tract’s land area that falls within the circle. This procedure is not generally a problem in most cities (particularly in metropolitan areas) where tracts are quite small. It is, however, a potential problem in a small number of rural cities, where tracts surrounding the city can be huge and sparsely settled (thus leading to an unrealistically low apportionment of the population to the circle). After some diagnostic checks, it was decided to drop all observations where the city’s aggregate income (within its actual boundaries) comprised more than 60 percent of the aggregate income in the wider circle area. We concluded that this deleted the “problem” cases. As a result, 21 cities were dropped from our model estimation. Please contact the authors for any further information on the construction of the aggregate income variable. 48 Another set of variables concerns cities’ locations and their “age”—or stage of urbanization. Using the mapping program, we were able to determine whether each city has an interstate highway within its borders as of 1990, and we include a dummy variable measuring the presence of one of these major freeways. Highway accessibility is often thought to be vital for many high-volume retailers. We also include two dummy variables—urbanization and central city—that relate to whether the Census Bureau defined each city, as of 1990, as being part of an urbanized area, or as being a core city in a metropolitan area. Urbanized areas present more of a “critical mass” of shoppers that can lead to extensive retail competition and support specialty or luxury stores. Chain stores tend to seek agglomeration economies, sometimes avoiding smaller regions which are more difficult to service logistically from centralized warehousing and distribution networks (Kroll and Marrinan, 1985b, pp. 25–28). Research also shows that metropolitan areas experience more internal specialization in their retailing geography than rural areas (Berry et al., 1988, p. 164). Central cities, in particular, are traditional centers of commerce and are often the most accessible location in their respective regions. Thus, one might expect these variables to be positively associated with sales tax revenues per capita. We also include population density, because of the possibility that retailers might be attracted by concentrated populations—or, alternatively, repelled by the congestion and high land costs associated with such concentrations. Cities with larger land areas per person (holding other factors constant) may be those that have “cherry-picked” retail properties on their periphery by annexation. On the other hand, low-density cities may include more outlying rural areas or exurban 49 residential development, which are less likely to include retail facilities. Thus, the effect of density on sales tax success is uncertain. In addition, we include dummy variables representing location in three key, distinctive regions of California: the five-county Los Angeles area, the nine-county San Francisco Bay area, and the 12-county Central Valley region. (The reference category is cities in other parts of the state.) Through use of mapping software, we also were able to identify geographically isolated cities—those with no neighboring cities within five miles of the city boundary. We also include this geographic variable, given our suspicion that cities without nearby “competitors” might be in a better position to attract and retain retail centers. In terms of cities’ degree of development, we include measures of the number of years incorporated—that is, the length of time since the community officially became a city—and the median age of housing structures in the city (among occupied units only). “Older” places, as measured by incorporation dates, are probably more likely to have established business districts. However, older housing may indicate a filtering-down of households and thus perhaps imply lower spending on retail. We therefore expect a positive relationship between sales tax revenues and number of years incorporated but perhaps a negative relationship between sales taxes and age of housing. Finally, we include two variables that measure local policies relevant to sales tax revenues. One is the percentage of sales taxes collected within the city that are shared with the county under a recognized agreement. Many cities in the state have such agreements with their counties to automatically pass through a percentage share of their sales tax revenues 50 to the county.4 Such sharing would have the effect, of course, of reducing a city’s per capita revenues. Second, we turn to the question of city redevelopment policies. Cities frequently employ redevelopment in their quest to attract or retain retail stores and shopping centers. We therefore include a variable measuring the percentage of each city’s land area taken up by public redevelopment projects. This serves as a measure of local effort devoted to redevelopment, which is a widely used policy instrument in California cities (Dardia, 1998). If it is true that redevelopment effort is related to success in attracting retail development, then this variable should be positively associated with per capita sales tax revenues. Results of Model Estimation An ordinary least-squares regression model was estimated to calculate the relationship between these independent variables and per capita sales tax revenues. It is important to note that the regression technique measures the effect of each variable, holding each of the other variables constant. Thus, although cities in the Central Valley tend to have low per capita incomes, the model is able to show the association between Central Valley location and sales tax revenues, independent of each city’s per capita income. We should also note that because of data limitations, the model includes 398 cities as observations, rather than the full 457 ____________ 4We rely on the sales-tax-sharing arrangements listed in the 1992 volume of the California State Controller, Financial Transactions Concerning Counties of California (p. 142). Many other cities have negotiated ad hoc tax-sharing agreements with their counties concerning specific properties that the cities have annexed. There is no source that would allow us to account for such arrangements. In addition, a small number of cities have special agreements to share Bradley-Burns revenues with their redevelopment agencies. We do not expect that these infrequent arrangements would affect our results here in any substantial way. 51 existing and reporting sales tax revenues as of 1991.5 Table 3.2 provides the results of the regression model, which accounts for 42 percent of the variation in per capita sales tax revenues. Of the 22 independent variables, 12 have a statistically significant relationship with per capita sales tax revenues, generally in the expected direction. The variables positively associated with sales tax “success” are population, household income, redevelopment effort, the presence of an interstate highway, and, more surprisingly, Hispanic share of the population. Other city characteristics are negatively associated with sales tax “success.” These include most strongly average household size as well as population density, black share of population, household income squared, location in the Central Valley or (weakly) San Francisco Bay area, and, perhaps surprisingly, population change. None of the other variables are statistically significant. Factors Related to Sales Tax Success: Interpreting the Results What can we make of these findings? Many are not particularly surprising. First, cities with large populations—thus, a larger potential customer base that tends to attract retailers—have more sales per capita. Another factor that helps bring in shoppers—the presence within a city of a major highway—is also clear in the results. Cities with large average household sizes, on the other hand, tend to have less sales tax revenues ____________ 5Including the population change (1985–1991) variable means that only the 441 cities incorporated before 1985 can be included. Missing data relating to the redevelopment variable result in several other cities being dropped. In addition, as discussed above, certain small rural cities were eliminated when calculating aggregate income in the city market area; and four tiny commercial enclaves with huge sales tax revenues per capita were dropped because they skew the dependent variable. 52 Table 3.2 Regression Model of City Sales Tax “Success” as of the Early 1990s Independent Variable Log of 1991 population** % change in population, 1985–91*** % senior citizens Average household size*** % black*** % Hispanic*** Median household income, 1989*** Median household income squared, 1989*** Log of aggregate income in city and adjoining market area Log of travel-related spending (tourism) in county, 1992 Urbanized (dummy variable) Central city (dummy variable) Log of population density** Interstate highway through city (dummy variable)*** Los Angeles region (dummy variable) San Francisco Bay area (dummy variable)* Central Valley (dummy variable)*** No other cities within 5-mile radius (dummy variable) Years since city incorporated Years since median housing unit built % of sales tax within city allocated to county, 1992 Redevelopment areas (1994) as % of 1990 land area*** Adjusted R-squared = 0.42 No. of cases = 398 cities Unstandard- Standardized ized Coef. (beta) Coef. .103 .17 –.467 –.15 –.008 –.06 –1.060 –.66 –.016 –.13 .014 .36 .000 .59 –.000 –.83 .040 .09 –.054 .033 .057 –.125 –.09 .02 .02 –.12 .310 –.106 –.229 –.317 .18 –.06 –.12 –.14 .077 .001 –.003 .03 .03 –.03 .009 .02 .728 .19 T-Value 2.12 –2.95 –1.01 –6.67 –2.80 3.90 2.87 –4.77 0.77 –1.14 0.24 0.38 –1.97 3.75 –0.86 –1.69 –2.80 0.66 0.57 –0.57 0.42 4.43 NOTES: *p < .1, **p < .05, ***p < .01. Dependent variable is the log of per capita sales tax revenues (average of 1991, 1992, and 1993), measured in constant 1995 dollars. All independent variables measured as of 1990, except where noted. 53 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) per capita. Household size is, in fact, the most statistically significant variable in this and every other model that was estimated. Figure 3.1 is a scatterplot diagram illustrating the negative bivariate relationship between household size and sales tax success. Some less obvious outcomes emerge as well. Cities engaging in a great deal of redevelopment effort—as measured by the percentage of land area in redevelopment project areas—tend to have more successful sales tax results. Thus, it may well be that the great amount of effort California cities have devoted to redevelopment has shown some results, at least in the narrow fiscal sense of sales taxes received. (Dardia, 1998, shows the limitations of redevelopment as a broader economic 8 7 6 5 4 3 2 1 0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Average number of persons per household, 1990 SOURCES: U.S. Census (household size); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.1—Relationship Between Average Household Size and City Sales Tax “Success” 54 development tool.) Or, perhaps it is the case that cities with larger retail districts tend to be most predisposed to use redevelopment. Cities with lower population densities (controlling for their urban/rural status) also have more retail sales per capita, perhaps because retailers may seek less-congested highway-oriented locations. Cities located in Central Valley counties—and surprisingly, the Bay area—tend to have less success in generating sales tax revenues, holding constant for socioeconomic and other factors. The Valley effect is considerably stronger.6 Cities’ ethno-racial characteristics also may have a role. Localities with higher shares of blacks in the population have lower sales per capita (again, this is controlling for income and the other variables in the model). One might conclude that retailers and shoppers avoid such locations. However, cities’ Hispanic population shares are positively related to sales tax receipts. It is not obvious why this is so. It is important to note, however, that among California cities, Hispanic population share is highly correlated with average household size (r = .82). This high degree of collinearity means that we should not be too confident in interpreting the sign and coefficient of the Hispanic variable (although the variable never failed to attain high levels of statistical significance in many specifications of the model). This issue illustrates some of the limitations of a regression model—where all other factors are held constant—if one is seeking policy implications. One should keep in mind that, because cities that are heavily Hispanic tend to have large household sizes, such cities actually tend to receive lower per capita sales tax revenues, as shown in Figure 3.2. ____________ 6The Bay area effect is only marginally statistically significant and is not robust to alternative specifications of the model. 55 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) 8 7 6 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Percentage of city population Hispanic, 1990 SOURCES: U.S. Census (percentage Hispanic); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.2—Relationship Between Hispanic Share of Population and City Sales Tax “Success” The relationship between cities’ median household income and per capita sales tax revenues reflects the finding discussed above—that many of the lowest-ranked sales tax cities are very wealthy bedroom communities. The bivariate relationship between per capita income and sales tax success is shown in Figure 3.3. The graph reveals that the relationship between city income and sales taxes is generally (though not dramatically) positive at lower- and middle-income levels, but a negative relationship emerges fairly clearly at upper-income levels. The regression results—income positive, income-squared negative—bear this out. City population change, though expected to increase sales because of housing-related purchases, actually is associated with lower sales taxes per 56 Log of per capita sales tax revenues (average of 1991–93, in 1995 dollars) 8 7 6 5 4 3 2 1 0 0 50,000 100,000 150,000 200,000 250,000 Median household income, 1989 (dollars) SOURCES: U.S. Census (household income); California Department of Finance, Demographic Research Unit (city population); California State Controller, Financial Transactions Concerning Cities of California (annual) (sales tax revenues). Figure 3.3—Relationship Between Income Level and City Sales Tax “Success” capita. This may be the case because the fastest-growing cities tend to be at the fringes of metropolitan areas—places that are relatively heavy in housing and low in commercial land uses. In short, the retailers may simply have not caught up with population movement, with the effect being that rapidly growing cities are “under-retailed,” with residents traveling to more established business centers to do their shopping. One of the most puzzling results is that the aggregate income of a city’s wider market area does not seem to influence its sales tax success. 57 Although the relationship is in the expected positive direction, it is not statistically significant in this specification.7 Cities’ Relative Gains over Time: A Longitudinal Model We turn next to a related question: Which types of cities have gained and lost ground in sales tax revenues relative to other jurisdictions in recent years? To address this issue, we examine each city’s change in per capita sales tax revenues from the early 1970s to the early 1990s (in inflation-adjusted 1995 dollars). Specifically, we take the average of the per capita sales tax revenues for each city for 1971, 1972, and 1973, and again for 1991, 1992, and 1993. To deal with the extremely high per capita sales taxes some cities exhibit, we again use logarithms, subtracting the log of the average per capita sales tax revenues in the early 1970s from the log of the per capita sales tax revenues in the early 1990s. For this analysis, we are interested in how changes in the various demographic and other characteristics of each city are related to changes in sales tax receipt. Most of the variables are analogous to those in the cross-sectional model above, except that they reflect longitudinal change. We do include the initial (1970 or 1971) levels of a few potentially important variables, such as population size and urbanization status, because we might expect these initial characteristics to influence subsequent changes in cities’ retail market activity as well. Because ____________ 7We also estimated a regression model with county-level dummy variables for all counties containing two or more cities, to control for any unobserved characteristics of each county that might affect sales tax success. Other independent variables were the same as above, except that the tourism, Bay area, Los Angeles area, and Central Valley variables were eliminated as they were subsumed by the county dummies. The results of this model were essentially the same as those reported above. The county dummies lacked (joint) significance, and the explanatory power of the model was unchanged. 58 certain data (specifically data on Hispanic population shares and on land area) were unavailable for cities with populations below 2,500 in the 1970 Census, we include in the model only cities with 1970 populations of 2,500 or more.8 This also has the benefit of eliminating the outlier cities of Colma, Industry, Sand City, and Vernon and substantially improving the fit of the model. Another data limitation is that our measure of the aggregate income, or purchasing power, of each city’s wider market area cannot be constructed from the 1970 and 1980 data. As an (inferior) alternative, we use the change in aggregate income of the entire county in which each city is located.9 Another difficulty involves measuring changes in redevelopment effort, since we do not know the share of city land area accounted for by redevelopment projects in the 1970s. Instead, we use a variable that measures the number of new redevelopment sites designated in each city between 1970 and 1989. In addition to the “change” variables, we include dummy variables for cities considered central cities by the Census Bureau, and those that were considered urbanized as of 1970. We also include a dummy for cities experiencing a change from rural to urbanized status between 1970 and 1990. The model also takes into account the initial population level, ____________ 8The Census Bureau used a somewhat different concept of Hispanic populations in the 1970 Census than in later years. Thus, strictly speaking, cities’ “percent Hispanic” are not fully comparable across decades. However, cities’ shares of Hispanic population relative to one another are comparable in each Census, and thus cities’ relative change in Hispanic share—the variable we use here—should be only minimally affected by the definitional change. In any event, the Hispanic variable turns out to be unimportant for the longitudinal analysis. 9A model using county-level dummy variables, and therefore absorbing the variation of the county aggregate income variable and region variables, produces very similar results to those presented here. 59 average household size, and median family income of each city.10 (Unfortunately, we lack data on the presence of an interstate highway in each city as of 1970, so this variable is not included.) Finally, we have two measures of each city’s “initial position” with respect to fiscal policy. One is its property tax revenue percentage of own-source general revenues in 1971. We might anticipate that low property tax cities sought sales tax revenues more vigorously—although we cannot predict whether they were successful in attracting retail. The second fiscal measure is each city’s sales tax “success” as of 1971–73. This variable is the log of each city’s average per capita sales tax revenues for 1971, 1972, and 1973. We include this variable on the thought that perhaps “the rich got richer”— that is, initially successful cities might have consolidated their advantages as retail centers over time. Table 3.3 displays the results of the model estimated, which accounts for about half of the variation among cities. Nine of the 20 variables show statistically significant relationships with the change-in-sales-tax variable. Three of these were positively associated. Cities that were urbanized in 1970 did better relative to their nonurbanized counterparts. So did those changing their status from nonurban to urbanized, and those with increases in population density.11 Six variables are negatively associated with sales tax “success” over time. Some of these are not surprising, given our earlier results. For example, a growth in average household size is negatively related to changes in per capita sales tax revenues. And once again, cities seeing ____________ 10Household income is not available for 1970, so we use family income instead. 11Cities with greater gains in county aggregate income—that is, more dollars of potential spending money within broad commuting range—did better over time at gaining sales tax revenues, but this result was just below the level of statistical significance (p < .104). 60 Table 3.3 Regression Model of Changes in City Sales Tax “Success” Between Early 1970s and Early 1990s Unstandard- Standardized T- Independent Variable ized Coef. (beta) Coef. Value Log of 1971 population* –.048 –.13 –1.90 % change in population, 1971–91*** –.085 –.21 –3.62 Change in % senior citizens, 1970–90** –.015 –.14 –2.25 Average household size, 1970 –.008 –.01 –0.14 Change in average household size, 1970–90*** –.530 –.52 –6.20 Change in % black, 1970–90 –.001 –.01 –0.13 Change in % Hispanic, 1970–90 –.002 –.07 –0.88 Median family income, 1969 –.000 –.10 –1.47 Change in real median family income, 1969–89 .000 .05 0.70 Change in real aggregate income of county (log of 1990 value minus log of 1970 value) .127 .08 1.63 Urbanized as of 1970 (dummy variable)*** .244 .27 3.55 Changed from rural to urbanized, 1970–90 (dummy variable)* .142 .10 1.96 Central city (dummy variable) –.057 –.04 –0.81 Change in population density (log of 1990 persons per square mile minus log of 1970 persons per square mile)** .125 .11 2.19 Los Angeles region (dummy variable) –.003 –.00 –0.05 San Francisco Bay area (dummy variable) .024 .02 0.39 Central Valley (dummy variable)* –.107 –.10 –1.82 No. of new redevelopment areas designated, 1970–89 .000 .00 0.02 % of own-source general revenues from property tax, 1971 –.109 –.04 –0.83 Initial sales tax “success” (log of per capita sales tax revenues, average of 1971, 1972, 1973)*** –.162 –.27 –5.21 Adjusted R-squared = 0.49 No. of cases = 345 cities NOTES: *p < .1, **p < .05, ***p < .01. Dependent variable is the difference between the log of per capita sales tax revenues as of the early 1990s (average of 1991, 1992, and 1993) and as of the early 1970s (average of 1971, 1972, and 1973), measured in constant 1995 dollars. 61 population gains lost some of their relative standing in per capita sales tax revenues over this period. In this model, however, a higher initial population level also had negative effects on subsequent sales tax performance.12 Central Valley communities have done worse over time relative to other cities. A growth in senior citizen share of the population also led to declines in cities’ relative performance. Finally, it is interesting to note that cities with high levels of initial sales tax success—that is, higher per capita sales tax revenues in 1971– 73—declined over time relative to other cities. This could indicate a “regression to the mean,” as discussed in Chapter 2. Rather than “rich” (in sales tax) cities getting richer, it appears that retail-heavy places lost some of their relative advantages during the 1970s and 1980s.13 We also estimated separate regression models for changes over each decade—the 1970s and 1980s—to see whether patterns differed in these two periods. Some of the results were broadly similar across the two decades—for example, the negative effects of population increases, increases in household size, and initial sales tax success. However, some differences by decade were apparent. Higher-population cities suffered ____________ 12In the earlier cross-sectional model, cities with higher populations are seen as having better performance. In the longitudinal model, cities with initial higher populations do worse over time. These results are not necessarily inconsistent. Cities with large populations have probably always done better than small cities; it is just that their relative dominance has deteriorated since the 1970s. A more anomalous finding is that population density has a negative effect in the cross-sectional model, but increases in density have positive effects in the longitudinal model. It may be that cities that saw notable drops in density over this period either were experiencing neighborhood decay or were annexing outlying, low-density areas that tend to be primarily residential. In either case, sales tax performance would likely suffer. 13Note also that cities with low portions of revenue from the property tax, contrary to the speculation of some observers, have not gained ground in sales tax revenues over time. These cities began the period with high sales tax revenues—indeed, this is what enabled some of them to avoid levying a property tax. The fact that these cities have also ended up with higher-than-average sales tax levels reflects their initial situation, not any changes in their success in the wake of Proposition 13. 62 particular decline during the 1970s but did not decline in the 1980s; and urbanized status was a particular advantage in the latter decade. Also, the relative decline of Central Valley cities occurred entirely in the 1980s. Summary The question posed by the title of this chapter was “Which types of cities have benefited from the situs rule?” We offer the following answers, based on the data analysis in this chapter. We present only the conclusions about which we are most confident. These relationships hold true while “holding constant” for other factors of potential importance to sales tax revenue generation. To summarize, we performed two analyses of sales tax “success.” One examined cities’ sales tax revenues per capita as of 1991–93. We will refer to this as the “cross-sectional” model. The other examined cities’ gains in per capita sales tax revenues, relative to other cities, between the early 1970s and early 1990s. We will refer to this as the “over time” model. The results of the two models indicate that the following types of cities have done better relative to other cities: • Cities with smaller or declining household sizes—generally meaning families with fewer children and a larger proportion of childless single adults (cross-sectional and over time models); • Cities that have not had rapid population growth (cross-sectional and over time); • Cities located outside the 12-county Central Valley region (cross-sectional and over time); • Cities more heavily engaged in redevelopment efforts (crosssectional only); 63 • Cities traversed by interstate highways (measured in crosssectional model only); • Cities that have middle to upper-middle income levels—but not very-high-income communities (cross-sectional only); • Cities with lower shares of black population (cross-sectional only); • Cities with a decreasing share of population aged 65 or older (over time only); • Cities in urbanized areas and those changing from rural to urbanized (over time only); • Cities that began the period with lower-than-average per capita sales tax revenues (over time only); and • Cities with higher populations (cross-sectional model)—but note that those with larger populations as of 1970 lost some of their relative advantage subsequently (over time model). Thus, the situs rule for distributing sales taxes does create distinct revenue winners and losers, though not in any very simple pattern. One might suspect upon first considering the issue that communities with higher socioeconomic status profiles—perhaps growing suburbs—would be most advantaged by a system that returns revenues to jurisdictions where sales occur. As we have seen, however, the picture is more mixed. Although it does appear to be the case that cities with larger black populations are disadvantaged by the rule, it is also true that very wealthy cities are disadvantaged. Central Valley communities—and particularly cities with high household sizes—are disadvantaged, but once we control for these factors there is some evidence (clouded by methodological complexities) that cities with larger Hispanic populations actually do 64 better.14 Moreover, rapidly growing cities appear to be doing worse than communities with more stable populations, and it is demonstrably untrue that the “rich” (cities with high sales taxes per capita) got “richer” between the 1970s and 1990s. In short, if policymakers hope that dismantling the situs rule for distributing the Bradley-Burns sales tax will mainly benefit cities with disadvantaged populations, they would be mistaken. In fact, some of the largest beneficiaries could be the wealthiest communities. We will further discuss the issue of alternatives to the situs-based system in our concluding chapter. The finding that redevelopment effort appears to lead to sales tax advantages for cities does raise the central issue of the fiscalization of land use: To what degree can local land-use policies influence cities’ standing in the sales tax revenue hierarchy? And whether or not such land-use policies are effective, is it true that cities do in fact favor retail over other types of development? These issues are explored in the next two chapters. ____________ 14Since cities with large Hispanic populations also tend to be those with large household sizes, however, this is not of much consolation. 65 4. The Sales Tax and Local Land-Use Decisions Across our region, virtually every city pursues what they call “economic development.” In virtually all cases, what they really mean is “real estate development.” And in the vast majority of cases, what that essentially boils down to is “sales tax development” (Cole, 1998). Thus far, we have examined the fiscal consequences of the situsbased local sales tax, looking at the distribution of revenues across cities. Another major policy controversy regarding this tax, however, concerns allegations that the quest for sales tax dollars has led local governments to bias or distort their development decisions in favor of retail. The result, critics charge, is a “fiscalization” of land use, with negative consequences for regional land-use planning and for the California economy. In this and the next chapter we explore this issue. We begin in this chapter with a conceptual discussion, considering the ways the local sales tax might affect development patterns and the policy considerations this raises. In Chapter 5, we will draw upon the findings of a specially designed survey 67 of city officials, focusing on their land-use strategies, to provide some evidence on the ways the local sales tax affects their decisionmaking. What Is Meant by the “Fiscalization” of Land Use? Kotin and Peiser (1997, p. 1975) define the fiscalization of land use as “the tendency of communities to establish land uses based on the net tax revenues they will generate for the city.” The city planning profession, which pioneered the concept of fiscal impact studies, “succeeded beyond their wildest dreams in making citizens and public officials aware of, and concerned about, the fiscal impacts of development” (Bunnell, 1997, p. 137). Misczynski (1986) made the first major written explication of the argument that California land-use decisions were becoming driven by fiscal considerations. His concern was not primarily the sales tax, however, but rather the growing reliance of local governments on development fees and arcane financing arrangements to provide public facilities such as schools and roads. Such mechanisms have become far more prevalent and important as the state and federal governments have withdrawn from providing funds for local public infrastructure. More recently, however, the term “fiscalization of land use” has been applied largely to local competition for retail businesses, as local governments seek situs-based sales taxes. As we have seen, the local sales tax is one of the few revenue sources with the potential—at least in theory—for substantial growth as a result of city decisions. Plus, its discretionary character as a revenue source makes it particularly attractive. Thus, current critiques of local governance and public finance in California often devote much attention to the quest among localities for sales tax dollars—or so-called “cash-box zoning.” We are not aware 68 of any studies that have systematically examined the effects of sales taxes on the land-use regulation behavior of local governments. Numerous authors, however, have alluded to these issues in California. “To the extent that land use decisions are now driven by their fiscal consequences,” wrote Chapman (1998, p. 11), “fiscalization has occurred.” He points to cities’ recruitment of so-called big-box stores and car dealerships as prime examples of this tendency, noting that “most jurisdictions trying to maximize sales tax revenues choose to encourage these types of development over residential development, which generates sales tax revenue only to the extent that the new residents shop in the same city in which they live” (p. 12). Columnist Neal Peirce sees fiscalized land-use decisions as “meaning a wild scramble for retail base especially for mega-auto malls and big box retailers. The result is not just more suburban sprawl, but real ugliness . . . .” (Peirce, n.d.). Schwartz (1997, p. 184) complains that “the reliance on sales taxes to replace lost property tax revenues has motivated planning and economic development decisions that sacrifice the long-term fiscal and environmental health of communities for short-term gains in sales tax producing land uses.” This dynamic can operate when cities recruit bigbox stores that then deplete the vitality of the existing downtown area. Fulton devotes an entire chapter in his 1997 book on the Los Angeles region, The Reluctant Metropolis, to an examination of what he calls “sales tax canyon”—the retail-dominated landscape that has emerged through competition between the cities of Ventura, Oxnard, and Camarillo in Ventura County. With property taxes restricted and land uses fiscalized around the sales tax, “Proposition 13 has provided the architecture on which Southern California’s urban landscape has been built ever since” (Fulton, 1997b, p. 260). In another important book, 69 Schrag (1998, p. 178) wrote that “since the biggest and cheapest source of local revenue was the sales tax, the planners not only pursued shopping centers, outlet stores, and auto malls; they often did so in preference over employers engaged in light manufacturing, even if they promised better jobs, because their enterprises did not return any sales tax to the local community . . . .” Schrag also points to cities’ aggressive use of redevelopment agencies, with their ability to issue debt, in efforts to attract retail development (p. 180). Tax increments from properties in the redevelopment areas are used in some cases to provide inducements or infrastructure improvements for large-scale retail projects in the area.1 Chapman (1981, p. 75) sees a “close connection” between situs-based sales tax collection and the use of redevelopment, claiming that “to some extent, the combination of these two institutional factors has had an impact on the development patterns of jurisdictions, as can be seen in the proliferation of shopping malls throughout the state.” Kotin and Peiser (1997, p. 1971) write that “redevelopment agencies have been particularly aggressive in pursuing high volume retailers such as Home Base, Price Club, K-Mart, Wal-Mart and Costco because of the sales taxes they generate . . . .” They find that, given increasing competition among redevelopment agencies for these retailers, a greater share of the financial benefits of these deals has gone to the retailers, developers, and particularly landowners. However, they also note that some cities have been able to set performance-based incentives in their deals with ____________ 1The passage of the redevelopment reform law, AB 1290, in 1994 restricts the ability of cities to use redevelopment funds to finance retail projects on vacant land sites. For a broader discussion of the fiscal issues of redevelopment, including AB 1290, see Dardia (1998). 70 developers, which can have the effect of protecting the city’s share of the benefits. Many stories circulate among California policy officials regarding allegedly fiscalized land-use decisions. Some have cited Monrovia’s decision several years ago to turn away a Kodak plant for a vacant site in its community in favor of a big-box retail store. Critics claimed that the city thereby traded well-paying manufacturing jobs for retail jobs that pay roughly half as much (Barber, 1994; Schrag, 1998). Similarly, “the city of Indio sacrificed a historic neighborhood through the use of its powers of eminent domain so that a mall developer could triple the size of the local shopping center” (Schwartz, 1997, p. 200). Others discuss strategic annexations by cities, where, for example, cities may annex commercial strips but not the surrounding residential areas (Senate Local Government Committee, 1989).2 One city official summarized the situation in simple fashion, telling us that although he often sees elected officials cutting ribbons at new retail developments, he never sees them do the same at new residential subdivisions. Thinking Carefully About the Fiscalization Argument Although rarely addressed systematically, the critique against sales tax-driven local land-use decisions seems to rest on certain assumptions or hypotheses, which are often left unstated. These include the following: ____________ 2On land-use conflicts between cities and counties, see Brooks (1988); Senate Local Government Committee (1989); Sokolow (1993); League of California Cities (1996); and Fulton (1997a). 71 • Hypothesis 1: Gaining sales tax revenue is a key goal for local land-use decisions, and thus retail development is given favorable treatment over other types of land uses. • Hypothesis 2: This favoritism toward retail has the effect of retarding residential and industrial development. • Hypothesis 3: The built landscape would look systematically different in the absence of situs-based local sales taxation. We would argue that the term “fiscalization of land use,” when applied to the local sales tax, presumes at a minimum Hypotheses 1 and 2. Some critics appear to imply (the far more adventurous) Hypothesis 3 as well. In thinking about these hypotheses, one must immediately confront one analytic issue. When considering the effects of the local sales tax on land-use decisionmaking, we are implicitly comparing the present system to some alternative system. Let us assume that the “alternative world” would be a California with a sales tax rate that is the same as it is now, but in which revenues were not returned to local governments on a situs basis. Favoring Retail? Let us now consider these assumptions. First, do local governments prefer retail development? This question is an empirical issue which, as of yet, no studies have systematically explored. The next chapter will take up this question in detail. Until now, anecdotal evidence of cash box zoning, and journalistic accounts of development policy in specific communities, have been used to support this argument. But even if all localities desire retail, not all of them are in a good position to get it. Retailers will tend to locate only 72 where there is market demand, reasonable transportation accessibility, and a population of sufficient size and purchasing power for their products and where the market is not already saturated with competitors. For some jurisdictions in remote locations or built-out market areas, significant retail development is unlikely, no matter how much they might desire it. Then, too, there is only so much “potential” retail in any given market area. At a given population, and a given level of income, taxable sales are finite; residents will not respond to every increase in retail “supply” with increased demand. Let us assume that a jurisdiction desires retail, has the potential to attract it, and there is in fact “room for growth” in the local retail market, because the population or income levels of the area are growing. What tools or policy levers does a local government have at its disposal to manifest that preference for retail? California law forbids making a “gift of public funds” to private firms or individuals, and so outright subsidies or donations of land to retailers or developers are theoretically forbidden (though there is legal uncertainty about the exact meaning of this doctrine and there are ways to sidestep it). Other city policies, however, can be used to support retail development, including the following: • Development fees or Community Facility District taxes that might ordinarily be applied to new developments to compensate the city for the impacts of the new growth may be waived or reimbursed, or held to artificially low levels. • Permitting and other bureaucratic procedures for new development might be expedited for retail developers. • “Excessive” amounts of land might be zoned for retail use, in the hopes that an increased portfolio of potential land sites might lure more retail developers. 73 • Local governments might be more inclined to grant a general plan change, or rezoning, to retail developments than to other types of land uses. • Redevelopment agencies might be used to clear land and assemble land parcels of sufficient size for major retailers—at below-market prices. • Redevelopment agencies, and city governments themselves, may provide the resources to build or improve infrastructure and amenities to serve retail developments. • Before recent reforms, some redevelopment agencies shared (“rebated”) city sales tax revenues directly with retail developers (particularly in the case of big box stores). • Cities with Enterprise Zones may use this designation to waive fees and expedite decisions for retailers. • Economic development officials might actively market their jurisdiction to retailers and retail developers.3 The issue then becomes whether these techniques are used so pervasively as to systematically change land-use patterns in the state. Of course, if all localities use such inducements to retailers, then the advantages individual cities might gain from these policies will be “competed away.” Presumably, the end result, in the aggregate, would ____________ 3One other technique that localities might use, at least in theory, is to lower the sales tax rate within their boundaries. This issue is more relevant in some other states where municipalities have a wider degree of discretion in the local sales tax rate they will levy. In California, however, the one-cent Bradley-Burns rate is uniform throughout the state, and so the border problem does not occur. The overall sales tax rate does differ somewhat across counties, because some counties levy additional voter-approved sales taxes, at their option, for special purposes, mainly transportation. It is unlikely that any county government would decide to rescind these add-on sales taxes simply in the hopes of luring more retailing, however, since most of the presumed benefits of the increased retail sales would be enjoyed by cities, not the county, and since “border effects” are less significant at the geographic level of counties. 74 be a shift of resources from local governments to the retail sector. If retail markets are competitive, then any cost advantages retailers enjoy because of these local policies will tend to result in slightly lower prices for retail goods. Thus, some might argue that consumers are in some sense the beneficiaries of the local governments’ largesse. Disfavoring Other Types of Growth? If local governments favor retail—whether or not they can actually hope to get any—what are the implications for residential and industrial development? In answering this question, we can see some possible effects of sales tax competition on land use. If jurisdictions zone “excessive” amounts of land for retail, then on the margins, less land is available for potential industrial or residential development, making such projects somewhat more expensive to build. Similarly, if retailers are more likely than housing or industry to win exemptions from the local zoning code, then there will be somewhat more flexibility and certainty in the retail development market, which will advantage that type of land use, at the margins. And in those (perhaps rare) instances where a retail development proposal competes with an industrial or housing proposal for the same land site, as in the Monrovia example, then the locality will tend to choose the retailer. Overall, then, a general predisposition toward retail by local governments probably will tend to make housing and industry marginally more difficult, uncertain, and expensive to develop. Changing the Landscape? This brings us to Hypothesis 3. Would the built landscape of California, or any of its metropolitan regions, look markedly different in the absence of a situs-based local sales tax? We can consider this question 75 in terms of two separate issues. First, does competition for retail lead to “extra” retail development? Second, does the retail industry in California have different location patterns than it would absent the sales tax? If cities systematically disfavor housing and industrial development in a way that makes it costlier or less predictable to develop, then the logical corollary is that retail development would be slightly cheaper, easier, and more certain. This would tend to lower the business costs of retailers. It is conceivable, but hardly certain, that because of these lower costs, slightly more retailing might exist; California would have more of its land use devoted to retail than would otherwise be the case. According to this line of reasoning, businesses that might otherwise be operating just below their break-even point would now find it possible to stay afloat financially. Alternatively, the same amount of retail business might exist, but it might consume slightly more square footage than it otherwise would, because retail space could be developed more inexpensively. This logic is very tenuous, however, because, as we have noted, the reduced costs of retail development are just as likely to be passed along to landowners or customers, rather than to stores’ profit margins. Although it is difficult to find data that would help evaluate this possibility, a “shopping center census” conducted by a private trade group casts doubt on the idea that California is especially heavily “retailed.” The International Council of Shopping Centers (1996), using data collected by a private firm on shopping centers nationwide, reported that California in 1995 had 19.2 square feet of “gross leasable area” of shopping space per capita. This figure put California only slightly above the national average of 18.9 square feet. The state ranked 19th of the 50 76 states—hardly an unexpectedly high rank given California’s wealth and urbanization. A second way that the sales tax might affect land-use patterns concerns the location of retail centers within metropolitan areas. Does California’s situs rule result in retail location patterns that are systematically different from those we would expect in the absence of a local sales tax? Retailers use certain standard techniques and rules of thumb in deciding where to locate (Davies, 1976; Guy, 1994). These depend on the type of goods that are for sale, the price level of those goods, and the special niche of the retailer (mass merchandise versus specialty, for example). No matter what inducements a local government offers, a Wal-Mart is not likely to open a new store in downtown San Francisco, a luxury car dealership is not likely to move its showroom to a small Central Valley farm town, and a dealer of specialized stereo equipment is not likely to find a location on a Sierra mountaintop optimal. Retailers locate mainly based on factors that are not subject to much control by local governments (Schneider, 1986). Consider the issue of “downtown” location versus “suburban” location, frequently discussed by those interested in urban development policy. In general, chain stores and general merchandise stores, which are among the most sought-after retail land uses, tend to seek middle- or upper-income suburban areas with good highway access for their new locations. “For regional and superregional centers, this often implies proximity to a freeway offramp; for small centers, traffic circulation on adjoining travel routes is important” (Kroll and Marrinan, 1985b, p. 25). This is true in California, as it is throughout the United States. Thus, a given retailer’s potential locations may be limited to “western San 77 Bernardino County along the I-10;” for another, the relevant region might be “the tri-Valley area of Alameda County.” Within such a limited universe, the retailer or developer might then shop around for the most attractive “deal”—consistent with their locational needs. But in any event, the location is likely to be a freeway-oriented suburban parcel, and this would be true whether or not the situs-based sales tax exists. Specific jurisdictions may “win” and “lose” in competing to host a given retailer, but the winner and loser locales are probably fairly similar places within the metropolitan hierarchy. Thus, overall patterns of retail location are not likely to be significantly affected. Implications and Policy Considerations It is probably safe to say that in any system where local governments raise a substantial portion of their funds by taxing or applying fees to activities within their boundaries, land uses will always be fiscalized, in one way or another. In some states with no local sales taxes whatsoever, for example, it is well known that municipalities are picky about the types of development they allow because of the amount of property taxes and service needs different land uses generate. Office and research facilities, for example, might be welcomed, whereas multi-bedroom apartments would be shunned, since the latter cost far more in local services than they provide in property taxes. Local governments, armed with planning studies, have become ever more sophisticated at determining these “fiscal impact hierarchies” (Bunnell, 1997; see also Orfield, 1997, p. 85). In California, the property tax is a very constrained revenue source for local governments, and its allocation is largely outside their control. Some cities receive a very small slice of the property tax dollar generated 78 by any particular parcel, with other portions being allocated to school districts, the county, and special districts. Thus, cities have become increasingly aggressive about the types and amounts of fees and assessments they can levy on new developments, with estimates of such charges for new housing construction in one county estimated at $20,000 to $30,000 per unit (Dresch and Sheffrin, 1997). But the most oft-remarked type of local land-use strategizing has involved competing for land uses that generate local sales taxes. As we saw in earlier chapters, however, overall per capita sales tax revenues available to California cities have been stagnant. If cities are increasingly competing, then they are fighting over a largely fixed pie. We have also seen that the overall hierarchy of cities in terms of per capita sales tax revenues has been relatively stable. Individual cities are not likely to leap ahead of their neighbors in their sales tax “success,” and those that have been “leaders” in the past have tended to perform less well than other cities over time. If, as we have argued, the retail landscape is probably not systematically reshaped by sales tax competition, and if cities are fighting over a fixed pie of retail businesses that would have located in their regions regardless of any inducements, then the main effect of fiscalized land-use decisionmaking in California is probably to shift resources from the public sector to retailers, their developers, and landowners.4 A second likely effect of favoritism toward retail is an underprovision of residential and industrial development, although we cannot measure this effect. It is unlikely that retail space will be “overproduced” or shifted markedly between central cities, suburbs, and small towns. ____________ 4This may, in turn, lead to slightly lower retail prices, so that Californians as taxpayers are to some degree subsidizing themselves as consumers. 79 Note that there may well be a “prisoner’s dilemma” character to sales tax competition, as well. That is, although cities locked into competition for retailers are not assured of any positive results—and lose out in the aggregate—cities that choose not to compete for retail will most likely lose sales tax revenues. It is very difficult for any individual city to unilaterally pull out of the sales tax game. All of these likely results, however, hinge on the assumption that jurisdictions do in fact actively favor retail development. The next chapter will provide empirical evidence to help evaluate this assumption. 80 5. The Preference for Retail Development Concern about fiscalization of land use in California tends to focus on competition between city governments for sales tax revenue. Observers argue that cities tend to bias land-use decisionmaking in favor of retail development to capture sales tax revenue, since this is one of the few sources of general revenues with much growth potential over which local governments have significant discretion. The most fundamental assumption on which this argument is based is that city governments do indeed favor retail development to maximize sales tax revenue. However, no systematic studies have attempted to confirm this hypothesis, and evidence in the literature to date has been largely anecdotal. In this chapter, we provide strong evidence to confirm that city governments in California do systematically favor retail development over other land uses when it comes to new development on vacant land, as well as redevelopment in designated “blighted” areas. These 81 conclusions are based on a recent PPIC mail survey of city managers regarding city development strategies. The survey results confirm that gaining sales tax revenue is the highest priority for city managers when making land-use decisions in relation to new development and redevelopment, and it is near the top of their list of considerations in relation to annexations of new territory. According to survey respondents, proposed retail projects are also the most likely to garner a zoning change or financial incentive from city officials to developers. Although the survey results point to certain regional distinctions in California in regard to attitudes toward development, the preferred status of retail development is generally consistent across all regions in the state. Survey Methodology In August 1998, PPIC mailed a questionnaire to the top administrative official—generally the city manager or city administrator—in each of the 471 cities then in existence in the state. The questions addressed three distinct but related topics: new development on vacant land sites, city-backed redevelopment in designated “blighted” areas, and annexation of new properties outside the city limits. Respondents were asked about their city’s level of activity in regard to these three types of development or redevelopment. If their city was at least somewhat actively engaged in an activity, respondents were then asked about their development priorities in reference to that activity. Regarding new development and redevelopment projects, respondents were asked to rate, on a one to seven scale, the desirability of seven types of land uses and the importance of 18 major strategic considerations in guiding their decisions and responding to development or redevelopment proposals. For new development only, respondents 82 were also asked about the likelihood that they would offer a general plan change or other incentive to attract seven different types of development. Finally, respondents were asked to rate the importance of 12 motivations for annexing new territory. A copy of the questionnaire is included in the appendix at the end of this report, which also provides more detail regarding survey methodology and the respondents. Officials from more than two-thirds (70 percent) of the state’s cities responded to the survey. In terms of population size and regional location, the cities that responded to the survey closely resemble the overall breakdown in the state. Thus, the results should be reasonably representative of overall attitudes and trends. Tables 5.1 and 5.2 show the breakdown of respondent and non-respondent cities by population size and region. Table 5.1 Survey Respondents, by City Population Size (in percent) Population in 1998 Less than 10,000 10,000–24,999 25,000–49,999 50,000–99,999 100,000–249,999 250,000 or more All California cities Respondents 25 21 22 19 10 3 100 n = 330 NonRespondents 36 21 21 15 7 0 100 n = 141 All California Cities 28 21 21 18 9 2 100 n = 471 NOTE: Totals may not add to 100 percent because of rounding. Development Activity in California Cities The overwhelming majority of California cities are able to pursue some form of development or redevelopment activity, as Table 5.3 83 Table 5.2 Survey Respondents, by Region (in percent) Region Los Angeles area Bay area Central Valley Other All California cities Respondents 38 23 18 21 100 n = 330 NonRespondents 37 18 21 25 100 n = 141 All California Cities 38 21 19 22 100 n = 471 NOTE: Totals may not add to 100 percent because of rounding. Table 5.3 New Development and Redevelopment Activity in California Cities Amount of Vacant Land Available for New Development % with “considerable” amount % with “limited” amount % with no vacant land Total Level of Redevelopment Activity Very Not Very Not Active Active Active 19 7 8 18 9 8 13 7 11 49 24 27 NOTE: Totals may not sum exactly because of rounding. SOURCE: PPIC City Managers Survey, 1998. Total 34 35 31 100 reveals. Only 11 percent of all cities surveyed have no vacant land available for development and are not engaged in any redevelopment activity. Over half (53 percent) both have vacant land available for new development and pursue redevelopment activity. Table 5.4 provides information on development patterns by region for the three types of development addressed in the survey: new development on vacant land, redevelopment in designated blighted areas, and annexations of new territory. As the table indicates, about two- 84 Table 5.4 Development Activity, by Region San Los Francisco Central Angeles Bay Valley Amount of vacant land available for new development % with “considerable” amount 27 19 68 % with “limited” amount 32 32 29 % with no vacant land 41 49 3 Level of redevelopment activity % very active 58 48 50 % somewhat active 22 22 27 % not active 20 30 23 Annexation plans within next five years % five or more square miles 13 4 13 % one to five square miles 19 16 52 % less than one square mile 11 32 22 % that cannot annex new territory 24 12 0 % that can annex, but will not 33 36 13 SOURCE: PPIC City Managers Survey, 1998. Other 35 46 19 37 26 37 10 25 32 9 25 State Total 34 34 31 50 24 26 10 26 22 13 28 thirds of surveyed cities have some vacant land available for new development. Of those cities, half report a “considerable” amount, and the other half have only a “limited” amount of vacant land. Cities in the Central Valley have the most vacant land. Two-thirds of them have a considerable amount but only 3 percent have none. San Francisco Bay area cities have the least vacant land. Nearly half have none at all whereas another third have only a limited amount. As Table 5.4 also reveals, three-quarters of respondent cities pursue some redevelopment activity, and half are very active. Los Angeles area cities are especially active; four-fifths are engaged in redevelopment, of which three-quarters engage in it very actively. Even in the rural regions of the state, however, the majority of cities pursue some redevelopment activity. 85 California cities achieve much of their new growth through annexation. Because of conflicts over the distribution of revenues from the annexed areas, annexation has been a major source of friction between cities and counties in recent years. Well over half (58 percent) of respondent cities plan to annex new territory within the next five years, as Table 5.4 indicates. Central Valley cities are the most likely to have annexation plans; nearly all (87 percent) plan to annex, in general, between one and five square miles within the next five years. Los Angeles area cities are the least likely to be planning annexations; less than half have such intentions. Our analysis of attitudes about development based on the survey results is limited to responses from cities that could potentially be pursuing the development activity in question. So, for questions related to new development on vacant land, we analyze responses only from cities that have at least some vacant land available for development. For those questions relating to redevelopment activity, we analyze responses only from cities that are pursuing redevelopment at least somewhat actively. In relation to annexation of new territory, we analyze only those responses from cities planning to annex some land within the next five years. Introduction to Survey Findings The survey of city managers reveals that city governments in California have a fairly consistent set of priorities in regard to development, whether aimed at vacant land, designated redevelopment zones, or newly annexed territory. City policymakers appear to adopt a set of general priorities relating to development that are applied to all types of development in the city. However, development priorities do 86 vary somewhat among cities depending on their location and the level of development activity that is under way. In spite of these distinctions, most cities across all regions agree on one point: Retail development is the preferred land use overall, and maximizing sales tax revenue is the main motivation. Land-Use Preferences Retail projects are the land use most preferred by city governments in California for both new development projects on vacant land and city redevelopment projects, according to our survey respondents. The overall preference for retail development is evident in Figure 5.1, which shows average “desirability scores” for various types of land uses. In fact, three-quarters of cities surveyed rank retail as their highest preference (or Average desirability score 7 6 New development Redevelopment 5 4 3 2 1 Retail Office Mixed- Light Single- Multi- Heavy use industrial family family industrial development residential residential SOURCE: PPIC City Manager Survey, 1998. Figure 5.1—Desirability of Various Land Uses for Development and Redevelopment Projects, as Viewed by California City Managers 87 tied for first) for development projects on vacant land, and the share is even higher (80 percent) for redevelopment projects. Office, mixed-use, and light industrial development are also considered quite desirable, as Figure 5.1 indicates. The least favored land uses overall are multifamily housing and heavy industry. Cities across all regions in California prefer retail development most of all. However, in regard to other types of land uses there are distinct attitudes by region. For example, Central Valley cities prize industrial development, especially light industrial, more than cities in other regions in the state. San Francisco Bay area cities, on the other hand, prefer mixed-use development more than cities in other regions. Within regions, land-use preferences are quite similar whether development is considered in relation to new projects on vacant land or redevelopment projects in designated blighted areas. Average desirability scores for a particular land use tend to vary by less than half a point (on a seven-point scale) when comparing new development with redevelopment by region, with one exception: In all regions, singlefamily residential development is considered less desirable for redevelopment areas than for new development projects on vacant land. Factors Influencing City Development Decisions Why do California cities consistently favor retail development over other land uses? Our survey results indicate that maximizing sales tax revenue is the prime motivation guiding development decisions made by city administrations in the state. Table 5.5 lists the average importance scores of 18 motivations for development and redevelopment decisions as ranked by our survey respondents. As the table indicates, maximizing new sales tax revenue is the top consideration overall. Individually, 72 88 Table 5.5 Factors Influencing Development and Redevelopment Decisions Considerations/Motivations New sales tax revenue generated City Council support Eradication of blight Adequacy of infrastructure in project area Likelihood of job creation Cost of municipal services for new development Traffic and other spillovers Conformity with city’s general plan Acceptability of proposal to nearby neighborhoods Project aesthetics, urban design issues New property tax revenue generated Environmental considerations New fee/assessment/enterprise revenue generated Contribution to sound regional economy Support from Chamber of Commerce or other local business interests Meeting affordable housing needs Competition from nearby cities Preservation of agricultural land Nearby cities’ views Views of other local governments Average Importance Score New Development Redevelopment 6.5 6.4 6.3 6.4 n/a 6.2 6.1 5.8 6.0 5.9 5.9 5.5 5.8 5.8 5.7 5.7 5.7 5.7 5.6 5.9 5.4 6.1 5.4 5.4 5.0 4.9 4.8 4.8 4.7 4.8 4.3 4.8 4.3 4.1 3.7 n/a 3.0 n/a n/a 4.0 NOTE: n/a indicates that question was not asked. SOURCE: PPIC City Managers Survey, 1998. percent of cities surveyed rank it as their prime motivation (or tied as the top motivation) in decisions about development on vacant land. Twothirds consider it the prime motivation (or tied as the top motivation) in decisions about redevelopment projects. City council support for development projects is considered almost as important as generating new sales tax revenue in influencing decisions about new development. In the case of redevelopment projects, city council support is actually tied with sales tax revenue as the top 89 motivation. The importance of city council support is not surprising, given that they approve major development projects and hire and fire city managers. But it is perhaps surprising that any other factor should equal city council support as an influence in development decisions. Maximizing sales tax revenue also outranks job creation, service delivery, and infrastructure considerations as a motivation in both development and redevelopment decisions. New sales tax revenue outranks even the main professed goal of redevelopment law—the eradication of blight. These attitudes vary somewhat by region. San Francisco Bay area cities are exceptional in that generating new sales tax revenue is not their top concern overall. Rather, city council support is their foremost priority for both development and redevelopment projects. For new development only, San Francisco Bay area respondents also consider spillovers such as traffic congestion to be more important than generating new sales tax revenue. Central Valley cities, in contrast, tend to emphasize the importance of promoting new job growth. For redevelopment decisions, job growth is tied with sales tax revenue as the top motivation of Central Valley cities, whereas for new development, it follows closely behind. Respondent cities’ priorities are strikingly consistent when new development is compared with redevelopment. The correlation of “importance scores” for 15 out of 16 motivations for development decisions, when comparing factors influencing new development with factors influencing redevelopment, is at or above 0.68. One motivation—generating new property tax revenue—is not correlated as highly, because cities consider it a more important motivation for redevelopment than for new development projects. Through taxincrement financing, cities are able to retain a larger portion of property 90 tax revenue gains in designated redevelopment areas than they can retain from new development on vacant land. Regional Variation in Attitudes About Development Although individual cities adopt consistent attitudes toward different types of development, there are distinct regional patterns in the factors that motivate development decisions in California. These regional patterns affect attitudes about sales tax revenue and help to explain why cities in some parts of the state are less actively engaged in the fiscalization of land use. Table 5.6 lists average importance scores for different development motivations, by region. It reveals that certain motivations are ranked at about the same level of importance by cities in all regions in the state. These include the cost of municipal services for development projects and the adequacy of infrastructure in the project area. However, there is greater regional variation in relation to other motivations. Some fiscal or “pragmatic” concerns are considered more important in rural and less-developed parts of the state. These factors include generating new sales tax revenue, promoting new job growth, and contributing to a sound regional economy. These three factors receive their highest ratings from cities in the Central Valley and other mostly rural portions of the state and are considered less important in the San Francisco Bay area. Another set of factors shows the opposite tendency. These political and “quality of life” concerns are judged to be more important by cities in the more heavily developed parts of the state, most of all in the San 91 Table 5.6 Importance Scores for Development Motivations, by Region New Development Redevelopment Factors Influencing Development Decisions SF Bay L.A. Cent. SF Bay L.A. Cent. Area Area Val. Other Area Area Val. Other Less regional variation (0.4 points or less difference between scores, by region) Adequacy of infrastructure in project area 6.1 5.9 6.1 Cost of municipal services for development 5.8 5.9 5.8 Conformity with city’s general plan 5.9 5.9 5.5 New property tax revenue generated 5.3 5.5 5.6 New fee/assessment/enterprise revenue generated 4.9 4.9 4.9 Support from Chamber of Commerce or business 4.6 4.7 4.5 Eradication of blight n/a n/a n/a Views of other local governments n/a n/a n/a 6.2 6.0 5.6 5.3 5.3 4.8 n/a n/a 5.8 5.7 5.8 5.3 5.6 5.5 5.8 5.7 5.6 6.2 5.9 6.2 4.9 4.9 4.7 4.9 4.9 4.6 6.0 6.3 6.2 4.0 4.1 3.9 5.9 5.5 5.8 6.1 5.0 4.9 6.3 4.0 More regional variation (0.5 points or greater difference between scores, by region, for development or redevelopment) Factors more important to cities in the San Francisco Bay or Los Angeles areas City Council support 6.6 6.4 6.2 6.2 6.7 6.4 6.2 Traffic and other spillovers 6.3 5.8 5.4 5.8 6.1 5.8 5.4 Acceptability of proposal to nearby neighborhoods 6.1 5.9 5.5 5.6 6.0 5.8 5.3 Project aesthetics, urban design issues 5.9 5.6 5.3 5.6 6.2 5.9 5.5 Environmental considerations 5.7 5.3 5.2 5.4 5.7 5.3 5.0 Meeting affordable housing needs 5.0 4.0 4.2 4.4 4.9 4.6 4.9 Competition from nearby cities 4.1 4.7 4.1 4.1 3.8 4.4 3.8 Nearby cities’ views 3.2 3.5 2.6 2.7 n/a n/a n/a 6.4 5.9 5.6 5.8 5.6 5.0 4.0 n/a Factors more important to cities in the Central Valley or “other” areas New sales tax revenue generated 6.2 6.5 6.6 6.7 6.3 6.4 6.5 Likelihood of job creation 5.3 5.7 6.5 6.2 5.3 5.9 6.5 Contribution to sound regional economy 4.3 4.6 5.1 5.2 4.4 4.7 4.9 Preservation of agricultural land 3.8 2.9 4.6 3.6 n/a n/a n/a 6.5 6.1 5.3 n/a NOTE: n/a indicates that question was not asked. SOURCE: PPIC City Managers Survey, 1998. 92 Francisco Bay area. These factors include city council support, spillovers such as traffic congestion, aesthetics and other design issues, and the acceptability of development proposals to nearby neighborhoods. Another motivation—competition with nearby cities—is distinctive in that Los Angeles area cities consider it more important than cities in other regions of the state. These regional patterns help to explain why generating new sales tax revenue is not the most important consideration for San Francisco Bay area cities, in contrast to cities in the rest of the state. Further investigation reveals that the attitudes of Bay area cities vary considerably depending on their level of development activity. The attitudes of Bay area cities that are more actively engaged in new development or redevelopment are closer to attitudes in the rest of the state. In contrast, attitudes among Bay area cities that are less actively pursuing development or redevelopment are quite distinct. San Francisco Bay area cities, regardless of their level of development or redevelopment activity, consider several “quality of life” and political factors to be more important in guiding development decisions than do cities elsewhere in the state. These motivations include city council support for development projects, traffic congestion and other negative spillovers, and project aesthetics and other urban design issues. However, when it comes to “pragmatic” factors, such as sales tax and property tax revenue, infrastructure and service considerations, and promoting new job growth, Bay area cities with “considerable” amounts of vacant land tend to rate these factors higher in importance than Bay area cities with “limited” amounts of vacant land. Most of these distinctions are also evident in regard to redevelopment, although less dramatically. Bay area cities that are “very actively” engaged in 93 redevelopment rate new sales and property tax revenue and promoting job growth as more important considerations than do Bay area cities that are “not very actively” engaged in redevelopment projects. Sales tax revenue is considered very important among San Francisco Bay area cities with a “considerable” amount of vacant land left to develop. Not only do these cities rank sales tax revenue as the most important consideration in relation to new development, they rate it as more important on average than any other group in the state, when measured by region and development level. In relation to redevelopment decisions, only the support of the city council is considered more important than new sales tax revenue for cities that are “very actively” engaged. In sharp contrast, Bay area cities with a “limited” amount of vacant land left to develop and those that are “not very actively” engaged in redevelopment consider a number of motivations, mostly “quality of life” factors, to be of equal or greater importance than generating new sales tax revenue. In considering this regional distinctiveness, keep in mind that San Francisco Bay area cities tend to be more “built out” than cities elsewhere in the state. That is, relatively fewer Bay area cities still have vacant land available for development than cities in other regions. In addition, relatively fewer Bay area cities are engaged in redevelopment than cities in the Los Angeles area or the Central Valley, and a higher share have elected not to pursue annexation. One could surmise from these findings that certain Bay area cities are less actively pursuing sales tax revenue (and thereby fiscalizing their land use decisions) because they are more fully developed, and therefore do not have the same opportunities to alter land-use patterns. They may be more concerned about the negative 94 consequences of past growth and therefore also more concerned about ensuring adequate planning for future growth. Factors Influencing Annexation Decisions When it comes to the factors that guide city administrations in decisions to annex new territory, maximizing sales tax revenue is not the top motivation. Instead, controlling the development of surrounding areas with the intention of ensuring consistency with city plans is the most important consideration, according to our survey respondents. However, generating new sales tax revenue follows closely as the second most important consideration, as Table 5.7 indicates. Table 5.7 Factors Influencing Annexation Decisions, by Region Factors Influencing Annexation Decisions To control development of surrounding areas To gain sales tax revenue To create jobs More efficient service provision To gain property tax revenue Agreements with county To provide greenbelt or open space To gain fee/assessment/enterprise revenue To meet housing needs Direction from LAFCO Agreements with other cities To prevent annexations by other cities Importance Scores, by Region Los San Angeles Francisco Central Area Bay Area Valley Other 5.9 6.0 5.8 5.3 5.4 4.4 6.0 5.6 4.8 3.9 6.0 5.3 5.2 4.0 5.3 4.5 4.7 4.1 4.9 4.8 4.7 4.0 4.8 4.3 4.6 5.2 4.2 3.7 4.2 3.5 4.7 4.4 3.8 3.6 4.4 3.8 4.4 3.6 3.9 3.5 3.6 3.3 3.1 2.6 3.2 3.2 2.5 2.0 SOURCE: PPIC City Managers Survey, 1998. All Cities 5.7 5.4 5.1 4.8 4.7 4.5 4.4 4.3 3.9 3.9 3.1 2.7 95 This overall trend masks certain distinctions by region and level of annexation activity. Controlling the development of surrounding areas is considered more important than generating new sales tax revenue in the more urbanized areas of the state, specifically in the Los Angeles and San Francisco Bay areas. It is also considered more important to certain Central Valley cities—those with the most ambitious annexation plans, namely, plans to annex one or more square miles of territory within the next five years. However, for Central Valley cities overall, and for cities in other mostly rural parts of the state, sales tax revenue is considered more important. San Francisco Bay area cities are unique in that providing a greenbelt or open space is ranked higher than generating sales tax revenue as a motivation for annexation. In contrast, Central Valley cities are unique in that future job growth is considered equally as important as gaining new sales tax revenue when it comes to annexation decisions. Do City Governments Act on Their Development Preferences? Just because city governments in California desire retail development above other land uses does not mean that they are in a position to get it. However, the survey data indicate that they are at least prepared to try. These results confirm the common contention that local land-use decisionmaking in California tends to be biased in favor of retail development. Retail was ranked by survey respondents as the type of land use for which they would be most likely to provide a general plan change or a financial incentive to a developer of new projects on vacant land. On a scale from one to seven, where one meant they were “very unlikely” and 96 seven meant they were “very likely” to provide an incentive for that type of development, more than two-thirds of respondents gave retail development a score above four. More than half of respondents gave retail development a score of six or seven, which can be taken to mean they would be quite likely to offer such an inducement. As with land-use preferences and motivations for development, certain regional distinctions underlie this general finding. The regional distinctions are completely consistent with previous findings. Central Valley cities are more likely to offer an incentive for light industrial development than for any other type of land use, whereas certain San Francisco Bay area cities, namely, those with a “limited” amount of vacant land, are most willing to offer an incentive for mixed-use development. Retail development follows closely behind in both cases, however. Which Types of Cities Favor Retail Development the Most? The preceding discussion described general trends in attitudes about development by region and level of development activity. However, we cannot decipher from those results whether there are certain characteristics of cities—characteristics that may vary by region—that could help to explain development priorities. In this section we explore more systematically whether any particular characteristics of cities are associated with more favorable attitudes toward retail development. We estimate statistical models to help account for the variations between different cities in California in their attitudes about retail development and generating new sales tax revenue. 97 Interest in Retail/Sales Tax: Dependent Variables We test 12 dependent variables in turn. Six of them are the actual desirability or importance scores—rated on a scale ranging from one to seven—that each survey respondent gave to retail development and sales tax revenue. These scores measure the absolute importance of retail development and sales tax revenue according to a particular respondent. The other six are variations intended to gauge the relative importance to each survey respondent of retail development and sales tax revenue as compared with other options for land use and other development motivations. These scores are constructed as the difference between each respondent’s desirability score for retail (or his/her importance score for maximizing sales tax revenue), and the average desirability score for other land use options (or the average importance score for other development motivations). The dependent variables are listed in Table 5.8. City Characteristics: Independent Variables In examining which city characteristics are related to heavy interest in retail development and sales taxes, we tested most of the independent variables described in Chapter 3. We used ordinary least-squares regression to assess whether they are associated with the 12 dependent variables described above. The independent variables employed in the model are listed and described in Table 5.9. Results of Model Estimation Table 5.10 presents the signs and statistical significance levels for the variables in the regression models. As the table indicates, only one independent variable proves to be significant in almost all of the 98 Table 5.8 Dependent Variables in the Regression Model 1. The desirability score of retail for new development projects on vacant land. 2. The difference between #1 and the average desirability score for six other types of land uses (besides retail) for new development projects on vacant land. 3. The desirability score of retail for redevelopment project areas. 4. The difference between #3 and the average desirability score for six other types of land uses (besides retail) for redevelopment project areas. 5. The importance score of maximizing new sales tax revenue as a consideration in attracting new development and responding to development proposals. 6. The difference between #5 and the average importance score for 15 other motivations (besides maximizing new sales tax revenue) for new development projects on vacant land. 7. The importance score of maximizing new sales tax revenue as a consideration in evaluating redevelopment project proposals. 8. The difference between #7 and the average importance score for 15 other motivations (besides maximizing new sales tax revenue) for redevelopment project decisions. 9. The importance score of maximizing new sales tax revenue as a consideration in decisions to annex new territory. 10. The difference between #9 and the average importance score for 11 other motivations (besides maximizing new sales tax revenue) for annexation decisions. 11. The likelihood that the respondent’s city would provide a general plan change (rezoning) or a financial incentive to the developer of a retail project on vacant land. 12. The difference between #11 and the average likelihood score for six other types of land use (besides retail). 99 Table 5.9 Independent Variables in the Regression Model 1. Bay area city: dummy variable for San Francisco Bay area location. 2. L. A. area city: dummy variable for Los Angeles area location. 3. Central Valley city: dummy variable for Central Valley location. 4. Development level: dummy coded 1 for cities with “considerable” amounts of vacant land left to develop, 0 for all other cities. 5. Redevelopment level: dummy coded 1 for cities “very actively” involved in redevelopment projects, 0 for all other cities. 6. Population density: the natural logarithm of city population per square mile in 1990, according to the U.S. Census. This was included in place of population and area because of the evident interaction between them. 7. Tourism: the natural logarithm of a county-level estimate of 1992 travel-related expenditures, from the Division of Tourism. 8. Median household income: median household income in 1989, from the U.S. Census. 9. Median household income squared: variable #8, squared. 10. Household size: average household size in 1990, from the U.S. Census. 11. Black population share: the percentage of blacks in the population in 1990, from the U.S. Census. 12. Hispanic population share: the percentage of Hispanics in the population in 1990, from the U.S. Census. 13. Property tax share: the percentage of own-source general revenue made up of property tax revenue in fiscal year 1992–93, from the Department of Finance. 14. Sales tax share: the percentage of own-source revenue made up of sales tax revenue in fiscal year 1992–93, from the Department of Finance. 15. Urbanized: a dummy variable to indicate whether the city was included as part of an urbanized area in 1990, from the U.S. Census. 100 Table 5.9 (continued) 16. Central city: a dummy variable to indicate whether the city was considered a core city in a metropolitan area in 1990, from the U.S. Census. 17. Population growth: percentage change in population, 1991 to 1998, from the Department of Finance. 18. Age of housing: median age of the housing stock as of 1990, according to the U.S. Census. 19. Isolated: a dummy variable indicating that the city has no neighboring cities within five miles of its boundary. 20. Neighboring cities: number of neighboring cities within five miles of city boundaries. 21. Interstates: a dummy variable indicating that one or more Interstate highways fall within city boundaries. 22. Own-source revenue: the natural logarithm of per capita general-purpose ownsource revenue (total revenue minus intergovernmental transfers minus public service enterprise revenue), from the California State Controller. 23. Years incorporated: the number of years since the city incorporated. 24. Annexation level: dummies for the number of square miles the city plans to annex & within the next five years—one to five square miles or more than five square miles. 25. These are considered in relation to a reference category of cities planning to annex less than one square mile. (These variables apply only to regressions relating to annexation.) regressions. This variable is central city, which is negatively associated with greater desire for retail development or sales tax revenue in 10 of the 12 regressions. Two other variables are significantly associated with the sales tax/retail orientation about half the time (in five or six regressions). These are household size, which is positively associated, and urbanized, which is positively associated with desire for retail for new development and redevelopment projects but negatively associated with desire for retail development for newly annexed territory. 101 Table 5.10 Results of Regression Models: City Characteristics Associated with Preference for Retail Development Dependent Variable 1. Absolute score: retail for new development 2. Relative score: retail for new development 3. Absolute score: retail for redevelopment 4. Relative score: retail for redevelopment 5. Absolute score: sales tax for new development 6. Relative score: sales tax for new development Independent Variables with Significant Association Development level Household size Hispanic population share Central city Central Valley city Hispanic population share Urbanized Central city Redevelopment level Tourism Household size Hispanic population share Sales tax share Urbanized Central city Population growth Own-source revenue Years incorporated Central Valley city Black population share Sales tax share Urbanized Central city Bay area city LA area city Median household income Median household income squared Household size Property tax share Central city Bay area city LA area city Median household income Sign and Signif. of Relationship +** +* –* –*** –* –* +* –*** +* –** +** –** –*** +*** –*** –* –** +* –* –** –** +*** –** –** –** +*** –*** +** –* –*** –** –** +** Model No. of Adj. R- ObserSquared vations 0.10 208 0.12 208 0.11 228 0.17 228 0.22 209 0.15 209 102 Table 5.10 (continued) Dependent Variable 7. Absolute score: sales tax for redevelopment Independent Variables with Significant Associations Median household income squared Household size Property tax share Central city Isolated Tourism Central city Sign and Signif. of Relationship –*** +** –** –*** +* –* –*** Model No. of Adj. R- ObserSquared vations 0.05 230 8. Relative score: sales tax for redevelopment Central city Isolated –*** 0.05 230 +* 9. Absolute score: sales tax for annexation Household size Hispanic population share Property tax share Urbanized Annexation level: five miles or more 10. Relative score: sales tax for annexation Bay area city Household size Urbanized Years incorporated 11. Absolute score: incentive for retail Median household income Median household income squared Household size Property tax share Central city Population growth Own-source revenue 12. Relative score: Central city incentive for retail Age of housing NOTE: *p < .1, **p < .05, ***p < .01. +*** –* –* –** +** –* +** –** +* +* –** +* –* –** –** –* –*** –* 0.30 175 0.19 175 0.15 206 0.09 206 103 Factors Related to Fiscalization: Interpreting the Results The poor overall fit of the model (the adjusted R-square is less than 0.3 for all regressions but one) indicates that there is no very distinct pattern of characteristics that helps predict which cities prefer retail development the most. Although certain trends are apparent, they cannot be called decisive. It seems that the desire for retail development is so pervasive that it cuts across cities with very different characteristics. Cities in urbanized parts of the state are more interested in retail than cities in rural areas when it comes to new development and redevelopment projects. This may reflect the greater degree of competition for retail dollars that cities in urbanized areas are facing. Cities in rural parts of the state, on the other hand, are particularly interested in retail development for newly annexed territory. Rural cities are far more likely to have annexation plans in the first place, and they may tend to direct retail development to areas near highways or other roads near town. California’s 36 central cities prove to be the major exception when it comes to the general trend for urbanized communities to value retail most highly. Central cities, which have lost per capita sales tax revenue since the 1970s, do not appear to be alarmed by that trend; they are less interested in garnering new sales tax revenue than other cities in the state. Instead, central cities seem to be more interested in balancing their various land-use goals. Communities with larger average household sizes, which as we saw in Chapter 3 are less successful at generating per capita sales tax revenue, appear to be interested in reversing that trend. 104 Characterizing Regional Patterns in Development Priorities Regression analysis enables us to explore in greater depth the factors that underlie the regional patterns in attitudes about development described above. We are able to assess whether certain characteristics of cities in California—characteristics that vary by region—can help explain regional distinctions in development priorities. We noted above that maximizing sales tax revenue is considered somewhat more important to cities in the Central Valley and other mostly rural parts of the state than in the San Francisco Bay and Los Angeles areas. Likewise we noted that certain “quality of life” factors are judged to be more important by cities in the more heavily developed parts of the state. These counteracting trends help to explain why certain San Francisco Bay area cities, in contrast to other cities in the state, place a number of priorities above maximizing sales tax revenue when making development decisions. In this section we investigate whether any particular traits of California cities can help explain these patterns. First we employ a technique called factor analysis to delineate patterns in survey responses. We identify two main “factors,” or patterns, that characterize responses to the questions regarding the various motivations for development and redevelopment decisions. These patterns are surprisingly similar in relation to new development and redevelopment. In both cases, the first pattern involves high importance scores (orthogonally rotated factor loadings above 0.5) assigned to certain planning and “quality of life” motivations, namely, urban design issues, traffic and other spillovers, the conformity of development projects with the city’s general plan, and environmental considerations. The second pattern involves high importance scores 105 assigned to certain “pragmatic” motivations, namely, generating new sales tax revenue, property tax revenue, and revenue from fees and assessments. The regional distinctions that we noted above are confirmed by the factor analysis. San Francisco Bay area cities tend to score highest on the “quality of life” factor, and Central Valley cities score lowest on average. Cities located in the Central Valley and other mostly rural parts of the state tend to score highest on the “pragmatic,” revenue-oriented factor. We estimated ordinary least squares regressions to test which of the independent variables employed above are associated with each city’s factor scores. In this way we can determine which characteristics define cities that follow each of the two main patterns of development priorities. The results of these regressions are displayed in Table 5.11. Although the results are again somewhat murky, if we look only to the most consistent results we can conclude that “quality of life” cities—or cities that score high on the first factor—are characterized especially by smaller household sizes and larger proportions of Hispanics in the population. Cities that follow the “pragmatic” pattern—or cities with high scores on the second factor—tend not to be central cities and to earn fewer tourism dollars. Summary One main assumption underlying the common critique that cities in California have fiscalized land-use decisions is that they bias development decisions in favor of retail projects. Our survey results provide strong empirical confirmation for this assertion. Respondents from the vast majority of cities across the state indicate that their administrations favor retail development above other land uses for both new development and 106 Table 5.11 Results of Regression Models: City Characteristics Associated with Two Patterns of Development Priorities Dependent Independent Variables with Variable Significant Association 1. Score on quality Tourism of life factor for Median household income new development Household size Hispanic population share Isolated Sign and Signif. of Relationship –** +** –** +** –* Model Adj. RSquared 0.15 No. of Observations 201 2. Score on quality of life factor for redevelopment Household size Hispanic population share Interstates –*** +* –* 0.13 226 3. Score on prag- Bay area city matic factor for Los Angeles area city new development Redevelopment level Tourism Median household income Median household income squared Central city Isolated –* –* +** –* +*** –*** –** +* 0.18 201 4. Score on prag- Tourism matic factor for Sales tax share new development Central city NOTE: *p < .1, **p < .05, ***p < .01. –* 0.03 226 –* –** redevelopment, and they do this to maximize new sales tax revenue. However, office, light industrial, and mixed-use development are also generally considered attractive. Although some cities that tend especially to be located in the San Francisco Bay area place certain quality of life considerations above maximizing sales tax revenue, these cities are the exception rather than 107 the rule. The state’s 36 central cities also have more varied motivations for their development decisions. Otherwise, however, few characteristics can be said to distinguish cities that are more actively pursuing retail development. Instead, the practice is so common that it can be called nearly ubiquitous. 108 6. Policy Considerations In this chapter, we pull together the findings of the study and discuss options for changing the distribution of the local sales tax. As we will show, although there are certainly reasons to criticize the situs-based sales tax, other options for distributing the tax have drawbacks as well. Policymakers concerned about the land-use and distributional consequences of the situs-based sales tax may instead wish to consider more broad-ranging reforms of local public finance and the state/local fiscal relationship. Summarizing the Results The Bradley-Burns sales tax is an important source of funds for California cities. Although not a huge share of local revenues, the local sales tax is important because it may be used in a discretionary way by cities, as opposed to being earmarked for a specific purpose. And unlike the property tax, it has the potential for substantial revenue growth if 109 cities make supportive land-use decisions. Or rather, cities tend to behave as if the sales tax is subject to major growth. As we have seen, however, the sales tax has been a steady or somewhat stagnant source of revenue in inflation-adjusted, per capita terms. This is due to the relatively fixed nature of retail spending per capita, the increasing share of consumer spending going to transactions that are not covered by the sales tax, and the many exemptions that state law provides for sales taxation. This stagnant revenue trend conceals a great degree of variation among cities in the degree to which they benefit from the tax. The overall hierarchy of cities, in terms of their sales tax revenue “success,” has been fairly stable over time; most cities, whatever their efforts in recruiting retail, have not gained or lost a great deal relative to their competitors. In spite of the meager overall prospects for growth in real per capita sales tax revenues, city governments have often been quite entrepreneurial at pursuing sales-tax-generating developments. We can conclude from the survey results in Chapter 5 that city officials do favor retail development over other land uses and are motivated by sales tax concerns as strongly as any other factor. Nevertheless, office, light industrial, and mixed-use projects also are viewed quite favorably, and most cities do not appear one-dimensional in their growth priorities— although housing is rarely a preferred land use. To the extent that cities are competing for retail—and many are prepared to offer concessions to get it—they are fighting over a relatively fixed pie, at least in per capita terms. Retailers, given the locational constraints of the market areas they operate in, are probably likely to locate in certain types of settings, concessions or not. Thus, it is not a 110 great leap to conclude that the major effect of fiscalized land-use decisionmaking is a transfer of resources from local government to retailers, developers, and landowners. With all of the favorable attention that cities show to retail, it is also likely that residential and industrial development are somewhat more difficult and more expensive to develop than would be the case in the absence of a situs-based sales tax. These land-use issues, along with the vast disparities in sales tax revenues that exist among cities, often lead reformers to urge a change in the situs basis for sales tax distribution. In the remainder of this chapter, we explore this issue. Policy Debates over Sales Tax Revenue Sharing A Los Angeles Times columnist has complained about the “absurdity” of sales tax revenues going only to the jurisdiction in which a sale takes place, arguing that, carried to a logical extreme, loyal local residents should limit all their purchases to stores in their home towns. “That way lies madness—and an undermining of the common good and the larger community” (Flanagan, 1998). Examining this “absurd” situation, some have proposed sharing the sales tax among local governments. Such proposals have existed for at least 20 years (Senate Local Government Committee, 1989, p. B-5). Although local tax base sharing is rare in the United States, it is not unknown. Most national discussions of tax sharing center on the property tax, which is a much more important share of local revenues in most parts of the country than in California. The Twin Cities region of Minnesota has practiced a program of property tax base sharing since 1971. Under the Minnesota fiscal disparities system, 40 percent of the growth in property tax revenue from commercial and industrial 111 properties is placed in a pool that is shared among the cities, counties, and school districts of the state’s major metropolitan area. These shared revenues, which totaled $367 million in 1996, are distributed to jurisdictions that are below average in their commercial/industrial property wealth per capita. Orfield (1997, p. 65), an advocate of greater sharing in Minnesota, finds that “even after this reallocation, disparities in property wealth remain enormous.” Nevertheless, tax base sharing has succeeded in reducing tax base disparities among communities in the Twin Cities by a very significant degree (Ross and Levine, 1996, p. 333; Orfield, 1997, p. 87). California policymakers have periodically proposed tax-sharing arrangements for the local sales tax. In 1994, Assemblymember Valerie Brown introduced a bill that attempted to revise sales tax allocation within each county as a way to address the fiscalization of land use. Under this plan, although all cities and counties would be guaranteed at least their existing levels of sales tax proceeds, the growth in sales tax revenue within each county would be distributed among jurisdictions mainly on the basis of population. Supporters claimed that this approach would better match revenues to service needs and would stop jurisdictions from artificially favoring retail over residential and industrial growth. Opponents argued that the proposal would lead to cities being disinterested in recruiting new businesses, would reward anti-growth jurisdictions, and would lead to ill-considered residential sprawl (because counties, seeking capitated sales tax revenues, might seek population growth for their unincorporated areas). The passage of Proposition 11, on the ballot in the November 1998 general election, amended the California constitution so as to allow cities and counties to enter into revenue-sharing agreements regarding revenues 112 obtained under the Bradley-Burns local sales tax law. Unlike the popular vote requirement under previous provisions, this measure allows jurisdictions to enter into such contracts with a two-thirds vote of the city councils or county boards of supervisors involved. There has been movement toward such a sharing arrangement in recent months between Modesto and its surrounding county, Stanislaus (tied to Modesto’s annexation of county lands). Although isolated instances of such “bilateral disarmament” from the sales tax race will probably occur, overall it seems unlikely that Proposition 11 will greatly reduce sales tax competition, since “winner” jurisdictions will have few incentives to enter into such agreements. Arguments For and Against Moving Toward a Population-Based Distribution of Sales Tax Revenues A system for distributing the sales tax that is based more on local populations than on their position in the retail hierarchy does have serious arguments in its favor. However, one can also make a good case against replacing all or part of the situs-based system with a populationbased system. Let us consider some of the arguments. In favor of the population-based system, one can make the following points: • The population-based system would presumably better reflect the revenue needs of each jurisdiction. Most public services are provided to residents, not to retail businesses. Some of the very extreme current disparities in revenues under the situs-based approach are fairly astounding; recall that the range is from $2 to $56,892 per capita. Many would likely find these disparities as lacking serious policy justification. 113 • Awarding Bradley-Burns revenues based on population might lead local governments to view housing development in a new light. Such a reform would create an incentive to increase residential populations. At present, housing is seen as a drain on local finances, and most cities (as the survey results confirm) do not especially encourage its development. Since many areas of California have some of the least affordable housing in the country, making housing development more attractive to local governments seems like a worthwhile policy goal. • Moving away from the situs-based system would mean that cities would no longer have an incentive to “chase” retail development. Given that retail growth is not the firmest basis for overall economic development, and that retail location is something of a zero-sum competition among localities in any given region, this would presumably be better for the state economy. Transfers of resources from government to retailers and developers would be reduced, and locational distortions in the land market presumably would be reduced as well. These are all important considerations. However, strong arguments can also be made against the move to a population-based distribution of local sales tax revenues: • Cities, to some degree, function as competitive economic units within an open economy. Much of city politics is rightly geared at maximizing revenues and economic vitality (see Peterson, 1981). Taking funds away from city governments that have managed through hard work or entrepreneurial leadership to gain a sales tax advantage seems to punish cities for their success. • Retail-heavy cities have special needs for public revenues. In effect, cities that are retail centers perform a service for the rest of their regions. Cities with many stores have large “visitor” and “worker” populations, even if their residential population is far 114 lower. This daytime population surge requires public goods and services such as larger roads, more trash collection, and more sophisticated police and fire protection than in equally sized towns that lack stores. Cities that are tourist destinations are special examples of this phenomenon, with daytime populations often many times larger than residential populations. In particular, the presence of stores and shopping generates high levels of public safety demands. As Figure 6.1 illustrates, retailheavy cities (as measured by their sales tax “success”) clearly have higher crime rates, on average, than other cities. The situs system compensates them for these externalities of retail.1 15,000 FBI Uniform Crime Index (1990) 12,000 9,000 6,000 3,000 0 0123456 Log of adjusted per capita sales tax revenues (average of 1991, 1992, 1993) SOURCES: FBI Uniform Crime Index; California State Controller, Financial Transactions Concerning Cities of California (annual). 7 Figure 6.1—Relationship Between Retail Concentrations and Crime in California Cities ____________ 1The situs system does not, however, compensate nearby communities for the additional traffic and other effects that they must wrestle with. 115 • As we showed in Chapters 2 and 3, many of the “least successful” cities in per capita sales tax revenues are very highincome residential communities, such as Hillsborough and Rolling Hills Estates. Switching toward a more populationbased system would seem to have the perverse effect of rewarding high-status communities for their “not in my backyard” approach to commercial development.2 (We will have more to say about the distributional effects of a population-based sales tax distribution in the next section.) • Finally, a broader critique of moving away from the situs rule toward a population-based approach is that it removes the one major incentive that California cities now have to pursue growth and development. One might argue that the way toward more balanced land uses would be to create a local revenue system that provides incentives for other types of development—not to simply make retail as disfavored as residential and industrial development. In short, scrapping the situs-based system would not solve the broader incentive problems with California’s system of local public finance.3 ____________ 2Orfield (1997, p. 87) sees a similar problem with Minnesota’s system for sharing commercial/industrial property tax revenues, in that wealthy residential enclave suburbs have been net recipients of revenues from the tax-sharing pool. 3An additional concern about shifting away from a situs basis for local sales taxes reflects legal considerations. Courts generally have held that there must be a “nexus” between the jurisdiction levying a tax and the activity that is subject to the tax. In the case of the local sales tax, as currently practiced, local governments levy a sales tax on an activity occurring within their boundaries—retailers must remit the tax revenues to the jurisdiction where the sale takes place. If, by contrast, taxes were locally levied but subject to sharing across jurisdictional boundaries, courts might consider the nexus requirement violated. One way around this problem would be to make the local sales tax a state-levied tax that is then distributed to localities in a revenue-sharing arrangement. This approach, however, would place local governments in the position of claimants on a state-generated subvention, which would leave many home-rule advocates unhappy. In theory, the nexus requirement could be satisfied by having a local tax levied upon the consumer according to his or her jurisdiction of residence, rather than one collected by the retailer based upon point of sale. This approach, however, can be dismissed as administratively unworkable. 116 Which Local Governments Would Benefit from a Population-Based Distribution? To further explore the distributional consequences of moving away from the situs system toward a population-based system for distributing local sales tax revenues, we examined such a scenario more systematically. Using 1994 data, the most recent year for which complete city financial data were available,4 we performed two types of analyses. In both, we asked the following question: What if the situs system were completely abolished and the Bradley-Burns revenues were instead distributed solely on the basis of population—city population in the case of cities and unincorporated-area population in the case of counties? It is very unlikely that reformers would ever succeed in suddenly abolishing the situs system, but making this strong assumption allows us to evaluate the general pattern of revenue gains and losses among jurisdictions. If policymakers decided instead to base revenues half on population and half on the location of the sale, the same pattern of gainers and losers would emerge, but the immediate gains and losses would simply be half the magnitude as our results show here. In the first analysis, we assume that cities and counties receive a share of statewide Bradley-Burns revenues based on their relative share of statewide population. In the second analysis, which is closer to some of the reforms that have been proposed in the past, we assume that BradleyBurns revenues remain within the county where the sale took place. In this analysis, each jurisdiction receives a share of the countywide total of local sales tax revenues based on its share of the countywide population. ____________ 4Several cities failed to submit financial records to the California State Controller for fiscal years 1995 and 1996. 117 In each of these two analyses, our basic questions were the same: How many jurisdictions would be better or worse off, in terms of their per capita sales tax revenues, compared to the amount they actually received in 1994? What share of the state’s population lives in jurisdictions that would be better off? What are the demographic characteristics of the potential gainer and loser cities under each of these plans? Answering these questions should help policymakers consider the politics of sales tax reform as well as the distributional issues involved. Statewide Redistribution As shown in Table 6.1, under a statewide redistribution of Bradley- Burns revenues in 1994, 55.5 percent of the 470 cities then in existence would have received more revenue than they did under the situs-based system. However, these “gainer” cities are smaller in population on average than those that would have lost revenue under the population- Table 6.1 Number of Jurisdictions That Would Be Better or Worse Off Under a Statewide, Population-Based Distribution of the Local Sales Tax, 1994 Type of Jurisdiction Cities that would be better off Cities that would be worse off No. of Jurisdictions 261 209 % of Total Total PopuCities or lation of Such Counties Jurisdictions 55.5 12,543,375 44.5 12,893,360 % of State Population 39.6 40.7 Counties that would be better off Counties that would be worse off 51 6 89.5 5,240,000 16.6 10.5 984,680 3.1 NOTES: Population for counties refers to unincorporated areas only. City and County of San Francisco is counted as a city. SOURCES: Authors’ calculations using data from California State Controller, Financial Transactions Concerning Cities of California (1994); annual city population estimates from the California Department of Finance, Demographic Research Unit. 118 based scenario. Thus, the overall population of cities that stood to gain was slightly less than those that stood to lose. Among counties, the vast majority (51 out of 57, not counting San Francisco) would have improved their fiscal position under the population-based system. The gainer counties had more than five times the unincorporated-area population as the counties that stood to lose. Overall, counting both cities and unincorporated areas, 56.2 percent of the state’s population lived in jurisdictions that stood to gain from a statewide reallocation of the Bradley-Burns tax. Focusing strictly on cities, Table 6.2 provides a portrait of the characteristics of cities that would gain and lose under this scenario. As one can see, cities in the Central Valley and Los Angeles area would disproportionately gain, whereas those in the Bay area and the remainder of the state would tend to lose. Nonurbanized cities would be especially Table 6.2 Characteristics of Cities Gaining and Losing Revenues Under a Statewide, Population-Based Distribution of the Local Sales Tax No. in Los Angeles area No. in San Francisco Bay area No. in Central Valley No. in rest of state No. urbanized No. nonurbanized No. central cities Mean per capita income, 1989 Mean household size, 1990 Mean % black, 1990 Mean % Hispanic, 1990 Mean % senior citizens, 1990 % of own-source local revenue from sales taxes, 1993 Gaining Cities 108 48 57 48 152 100 9 $17,090 3.0 4.3 28.5 11.0 18 SOURCES: U.S. Census (1990); authors’ calculations. Losing Cities 69 52 32 56 141 63 27 $17,854 2.6 3.4 19.9 13.0 28 119 likely to be better off, but three-quarters of the state’s 36 central cities would be losers. In terms of socioeconomic characteristics, the cities that would stand to gain tend to be slightly poorer, have higher household sizes, and higher proportions of Hispanics in their populations. These outcomes reflect the broadly redistributive nature of the statewide proposal. Within-County Redistribution If, instead, sales tax revenues were reallocated within each county on the basis of population, outcomes would be somewhat different, as shown in Table 6.3. In this case, a slim percentage majority of cities Table 6.3 Number of Jurisdictions That Would Be Better or Worse Off Under an Intracounty, Population-Based Distribution of the Local Sales Tax, 1994 Type of Jurisdiction Cities that would be better off Cities that would be worse off Cities with no change No. of Jurisdictions 240 229 1 % of Total Cities or Counties 51.1 48.7 0.2 Total Population of Such Jurisdictions 13,539,135 11,145,500 752,100 % of State Population 42.8 35.2 2.4 Counties that would be better off Counties that would be worse off Counties with no change 48 6 3 84.2 5,298,900 16.7 10.5 895,500 2.8 5.3 30,280 0.1 NOTES: Population for counties refers to unincorporated areas only. The City and County of San Francisco is counted as a city, and thus without a surrounding county would experience no change. The three counties with no change are those with no incorporated cities. SOURCES: Authors’ calculations using data from California State Controller, Financial Transactions Concerning Cities of California (1994); annual city population estimates from the California Department of Finance, Demographic Research Unit. 120 would have been better off in 1994 than under the situs approach. However, the gainer cities in this case have higher populations. Among counties, 48 would have received more sales tax revenues, six less, and three without any incorporated cities would have experienced no change. Overall, under the countywide population-based distribution of the Bradley-Burns tax, 59.5 percent of the state’s population lived in units that stood to gain, 38 percent in jurisdictions that would have lost revenues, and 2.5 percent in jurisdictions that would have experienced no change. Thus, it would appear that there would be more potential political weight behind the countywide redistribution than the statewide approach. Table 6.4 shows, moreover, that the intracounty redistribution of sales tax revenues would benefit a majority of cities in the Los Angeles Table 6.4 Characteristics of Cities Gaining and Losing Revenues Under an Intracounty, Population-Based Distribution of the Local Sales Tax No. in Los Angeles area No. in San Francisco Bay area No. in Central Valley No. in rest of state No. urbanized No. nonurbanized No. central cities Mean per capita income, 1989 Mean household size, 1990 Mean % black, 1990 Mean % Hispanic, 1990 Mean % senior citizens, 1990 % of own-source local revenue from sales taxes, 1993 Gaining Cities 99 55 48 38 152 79 11 $17,993 3.0 4.4 28.3 10.5 17 SOURCES: U.S. Census (1990); authors’ calculations. Losing Cities 78 44 41 66 140 84 24 $16,843 2.7 3.3 21.0 13.2 28 121 area, Bay area, and Central Valley alike. However, most cities in other parts of the state would lose, and urbanized communities would do better than nonurbanized ones. Cities that stood to gain under this proposal also tend to have higher per capita incomes than those that would lose (although the “winning” cities also had higher minority populations). Thus, this proposal would not necessarily advance socialequity goals. Broader Remedies: Balancing Fiscal Rewards for Growth in California Communities Thus, moving to redistribute the local sales tax—or the growth in that revenue source—on a population basis might, in a narrow sense, work to the fiscal advantage of the majority of jurisdictions, covering the majority of population in the state. But such a shift not only would be politically unpopular among the numerous local governments that stand to lose but would also fail to address the wider problem of making nonretail forms of development more fiscally viable. Perhaps the way around this dilemma is to widen our focus beyond the sales tax—which is, after all, merely one, relatively modest component of local government revenues. Rather, policymakers might do better to examine the overall context of local public finance. As it stands, retail is looked upon with favor not merely because it is associated with the sales tax but because other forms of development are perceived simply to not pay their own way. Cities often are particularly unenthusiastic about housing because they receive such a small slice of the property tax dollar. The city manager of Long Beach, for example, complained that a typical owner-occupied home produces only $210 per year in property tax revenue, but it costs the city $350 per year in 122 services. As a result, Long Beach has turned more attention toward retail development (Flanagan, 1998). Counties, which receive a larger share of the property tax in unincorporated areas and provide less in the way of municipal-type services than cities do, often tend to be more welcoming of residential growth. As a result, critics charge, “starter homes and other tax ‘losers’ have been relegated to distant locations on the metropolitan fringe, often in unincorporated areas” (Fulton, 1998, p. 1). Cities often have reasons to prefer retail to industrial or office development as well. Kotin (n.d.) claims that even attractive modern office campuses, with highly paid employees, are likely to produce a fiscal deficit in many cities with low shares of the property tax distribution in their area. He concludes, “A system which essentially prohibits taxation in relation to services to property or to employed population, combined with a system that limits changes in property tax assessments until a property changes hands, virtually dooms a city to spending more to accommodate new employees than it will receive in benefits.” As a result, one columnist has argued, “by focusing virtually all their attention on where wealth is spent, cities are ignoring the real source of economic development: the creation, not the consumption, of wealth” (Cole, 1998). Most thoughtful observers probably would view cities’ disincentive to seek residential or industrial development as negative for the state of California. But note that this result is not the “fault” of the local sales tax. Even in the absence of a situs-based sales tax, industrial, office, and residential development would still often fail to “pay their own way.” Thus, the solution to this problem would seem to involve adjusting local finances to provide a greater incentive to take on such development— rather than adjusting the sales tax to provide less incentives to develop 123 retail.5 Some local officials argue that cities and counties are fighting over “bones in the road,” arguing that the real need is for a redirection of property taxes back to local governments. For such reasons, in 1994 the nonpartisan Legislative Analyst’s Office proposed a plan under which the one-cent Bradley-Burns sales tax would be turned over to the state treasury. In return, the state would allocate more of the property tax to cities and counties and less to school districts (with the state covering the loss of school funds with its increased sales tax revenues). This proposal goes further in actually ending the situs-based local sales tax, as the Legislative Analyst’s Office viewed the land-use effects of sales tax competition as particularly pernicious (1994, pp. 123–125). The San Diego Association of Governments (SANDAG) has proposed an exchange of $4.8 billion between state- and local-controlled revenues. This would involve cities and counties exchanging statecontrolled funds for additional property taxes and an additional one cent of the sales tax rate. SANDAG would also give local governments authority over the half-cent public safety sales tax (currently collected by the state and delivered to counties on the basis of population, and to local governments within counties by formula). The local governments within each county would decide collaboratively on how to distribute the extra 1 percent sales tax and new property tax revenues amongst themselves (with certain “hold harmless” provisions). More recently, the California Governance Consensus Project, which is a broad-based effort at resolving intergovernmental conflicts, has ____________ 5It is worth noting, however, that adjusting the Bradley-Burns sales tax to a more population-based distribution rule would give cities a somewhat greater impetus to develop housing, since more housing means more population and thus more revenue. 124 discussed proposals to reshuffle certain sources of state and local public funds. The aim has been a reallocation of funding sources such that the resulting change would be revenue-neutral to the state and to local governments but would provide better incentives for balanced land use. One option discussed involves dedicating a considerably larger share of property taxes to the support of (non-school) local governments. To offset the increased share of K–14 school spending that the state’s general fund would have to bear after this property tax reallocation, local governments would give up various other revenues to the state— including vehicle license fees and a half cent of the Bradley-Burns sales tax. Although the Consensus Project has apparently since abandoned this specific proposal, it represents one of the most comprehensive attempts to address local fiscal problems in a revenue-neutral context. Conclusion Clearly, the issues involved in such a reallocation of revenue sources would be complex. But this complexity should not make the challenge insurmountable. California currently finds itself with a local fiscal system that promotes a competition for one type of development—retail. Such competition has significant costs, in terms of the comparative lack of attention to broader forms of economic and community development as well as the cost of incentives to the public treasury. Moreover, for most cities, the competition for a greater share of sales taxes in their region is likely to be ineffective. The amount of sales tax revenue per capita in California has been stagnant, most cities have roughly maintained the same relative position in the retail hierarchy, and retailers make location decisions based on factors that are mainly outside the control of local governments. 125 In short, in the post-Proposition 13 environment, the local sales tax is a motivation for certain types of local government land-use behaviors that—although unlikely to systematically alter broad patterns of retail development—do have drawbacks for the state as a whole. No one has seriously suggested simply doing away with the situs-based sales tax as a response to these problems, since doing so would only create a gaping hole in city finances. And distributing the local sales tax on a per capita basis would fail to provide incentives for other types of development such as housing and industry. Fortunately, as the proposals noted above show, alternatives with the potential to broaden cities’ interest in pursuing balanced growth are available to policymakers. These options involve giving local policymakers control over a larger proportion of property tax revenues, in exchange for returning other, narrower revenue bases to the state. It will be difficult politically to change the existing system, since those who gain from the status quo could stand to lose much if it were changed, whereas the potential gains are more broadly diffused and lack a highly organized and motivated constituency. Nevertheless, the state government, by setting the fiscal rules that influence local land-use decisions, holds key leverage over statewide economic development strategy. The state has the capability to create a local fiscal system that would be more encouraging of the residential and industrial development that will be important to California’s continued prosperity. 126 Appendix The Mail Survey of City Development Strategies Our discussion in Chapter 5 of the relationship between city landuse policies and sales tax considerations is based upon evidence from a survey conducted in 1998. In August of that year, an eight-page questionnaire, titled “Development Strategies in California Cities,” was sent to the city manager or city administrator of each of the 471 cities then existing in California. Where the city lacked a city manager or city administrator, the survey was sent to the top administrative official, as identified by the League of California Cities, typically the city clerk. In some cases, the person to whom the survey mailing was addressed passed the survey along to another city official to complete—typically a planning or community development professional. The self-reported job titles of respondents are reported in Table A.1. As the table shows, 76 percent of the respondents were city managers or administrators and an additional 7 percent were assistant city managers 127 Table A.1 Survey Respondents, by Title Title City Manager Assistant City Manager City Administrator Assistant City Administrator Director of Economic/Community Development Director of Planning Planner Management Analyst City Clerk Other No. of Respondents 211 19 36 2 22 7 11 5 8 5 % of Respondents 64.7 5.8 11.0 0.6 6.7 2.1 3.4 1.5 2.5 1.5 Total responses to question Did not answer question Total surveys received 326 100 4 330 or assistant city administrators. The other respondents were mainly a mix of planning officials, community development directors, and city clerks. Our response rate was 70 percent. We followed standard mail survey methodologies designed to enhance response rates (Dillman, 1978; Jobber, 1986; Salant and Dillman, 1994; Sudman, 1985)—a particularly important concern among busy professionals such as city managers. A cover letter discussed the questionnaire and its goals. We also were fortunate in having the cooperation of the League of California Cities, whose executive director provided a memo for the mailing encouraging respondents to fill out the questionnaire. Survey recipients were sent a reminder postcard encouraging them to respond, and non-respondents were sent a new copy of the questionnaire three weeks after the original mail-out. A stamped envelope was provided for return mailing. 128 Responses were tracked using identification numbers for each city. However, respondents were assured anonymity, and thus no responses that could be identified with specific cities are reported in our research. The survey asked respondents a set of closed-ended questions regarding new development on vacant land, city-supported redevelopment, and annexation policy in their cities. A “screen” question for each of these three main sections weeded out cities for which the questions were not applicable. The survey questions covered a broad variety of topics and asked respondents about the importance of numerous factors that might influence city land use policy. The sales tax and retail development were not the major focus. In fact, we wished to avoid telegraphing that this topic was our main immediate concern, as we feared doing so might make respondents “think too hard” about providing the “right” answer. We hope to examine some of the other aspects of the survey responses in future research publications. A copy of the survey is reproduced on the pages that follow. 129 Please return the completed questionnaire in the enclosed stamped envelope. Public Policy Institute of California 500 Washington St., Suite 800 San Francisco, CA 94111 (415) 291-4400 phone (415) 291-4401 fax Project Directors: Paul Lewis and Elisa Barbour Code: ___________ Note: The individual responses to this survey are confidential. The code is used to track the returning surveys. In reporting the survey results, responses of individual cities will not be identified. 130 Q-1. Which of the following statements best applies to your city? (check one answer) ___ There is considerable vacant land available for new development. ___ There is a limited amount of vacant land available for new development. ___ There is little or no vacant land available; the city is “built out.” If you chose “little or no vacant land available”please skip ahead to question Q-5. Your responses to the other questions are very important to us. Q-2 – Q-4. The following questions relate to private-sector development proposals for currently vacant land sites in your city. Q-2. Given your city’s overall strategies and plans for land use and future development, how desirable to your city administration would each of these types of new development be? In other words, how sought-after are these types of development in your city, in general? Please rank each of the following: (Circle a number between 1, which is “very undesirable,” and 7, which is “very desirable”) 1 Single-family residential 2 Multifamily residential 3 Light industrial 4 Heavy industrial 5 Retail 6 Office Very undesirable Very desirable 1 2345 67 1 2345 67 1 2345 67 1 2345 67 1 2345 67 1 2345 67 7 Mixed-use development 1 2345 67 Q-3. For each of the following types of new development, please indicate how likely your city would be to provide a general plan change (rezoning) or a financial incentive to the developer or builder of the project: (Circle a number between 1, which is “very unlikely,” and 7, which is “very likely”) 1 Single-family residential 2 Multifamily residential Very unlikely Very likely 1 2 3 45 6 7 1 2 3 45 6 7 3 Light industrial 1 2 3 45 6 7 131 4 Heavy industrial 5 Retail 6 Office 7 Mixed-use development Very Very unlikely likely 1 2345 67 1 2345 67 1 2345 67 1 2345 67 Q-4. Generally speaking, how important are the following considerations to your city administration’s strategies in attracting new development and responding to development proposals? Please review each item below and indicate how important it is. (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Cost of municipal services for the new development 1 2345 67 2 New property tax revenues generated 1 2345 67 3 New sales tax revenues generated 1 2 3 4 5 6 7 4 New fee/assessment/enterprise revenues generated 1 2345 67 5 Adequacy of infrastructure in area of project 1 2345 67 6 Likelihood of job creation 1 2345 67 7 Conformity with city’s general plan 1 2 3 4 5 6 7 8 Acceptability of proposal to nearby neighborhoods 1 2345 67 9 City council support for project 1 2 3 4 5 6 7 10 Support of Chamber of Commerce or other local business interests for project 1 2345 67 11 Nearby cities’ views on the project 1 2 3 4 5 6 7 12 Project aesthetics, urban design issues 1 2345 67 13 Traffic and other spillovers 1 2345 67 14 Contribution to sound regional economy 1 2345 67 15 Meeting affordable housing needs in area 1 2345 67 16 Preservation of agricultural land 1 2 3 4 5 6 7 17 Environmental considerations 1 2 3 4 56 7 18 Competition from nearby cities 1 2 3 4 5 6 7 132 Q-5 – Q-7. Next, we’d like to ask you about redevelopment projects supported by the city government and/or Redevelopment Agency (RDA) in your city. Q-5. Is your city engaged in redevelopment? (please check the best answer) _____ Yes, very actively _____ Yes, but not very actively _____ Not currently engaged in redevelopment If you chose this answer, please skip ahead to question Q-8. Q-6. Given your city’s overall strategies and plans for redevelopment, how desirable to your city administration would each of these types of projects be in your redevelopment areas? In other words, how sought-after are these types of projects in your city’s redevelopment areas? (Circle a number between 1, which is “very undesirable,” and 7, which is “very desirable”) 1 Single-family residential 2 Multifamily residential 3 Light industrial 4 Heavy industrial 5 Retail 6 Office 7 Mixed-use development Very undesirable Very desirable 12 34 56 7 12 34 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 1 2 3 4 56 7 133 Q-7. Again, in considering your city’s redevelopment areas, in general how important are the following considerations to your city administration’s strategies in choosing which types of projects and land uses are appropriate? (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Cost of municipal services for the new development 1 2345 67 2 New property tax revenues generated (including increment retained by Redevelopment Agency) 1 2 3 4 5 6 7 3 New sales tax revenues generated 1 2 3 4 5 6 7 4 New fee/assessment/enterprise revenues generated 1 2345 67 5 Adequacy of infrastructure in area of project 1 2345 67 6 Likelihood of job creation 1 2345 67 7 Conformity with city’s general plan 1 2 3 4 5 6 7 8 Acceptability of proposal to nearby neighborhoods 1 2345 67 9 City council support for project 1 2 3 4 5 6 7 10 Support of Chamber of Commerce or other local business interests for project 1 2345 67 11 Views of other local governments such as school district or county 1 2 3 4 5 6 7 12 Environmental considerations 1 2345 67 13 Project aesthetics, urban design issues 1 2345 67 14 Traffic and other spillovers 1 2345 67 15 Eradication of blight 1 2345 67 16 Contribution to sound regional economy 1 2345 67 17 Meeting area’s affordable housing needs 1 2 34 56 7 18 Competition from nearby cities 1 2 3 4 5 6 7 19 Other goal(s) (please specify): 1 2345 67 134 Q-8 – Q-11. Related to city development strategies is the issue of annexation. We’d like to learn about your city’s plans in this area. Q-8. What is the current square mileage of your city (if known)? ________ Q-9. Many California cities, working with Local Agency Formation Commissions, have identified “spheres of influence” beyond their current boundaries. How many square miles are in your city’s sphere of influence (if known)? (Do not include area within current city boundaries, just the additional area in the sphere of influence. If your city does not have a sphere of influence, answer zero. If answer is unknown, leave blank.) _________ Q-10. In your estimation, what are your city’s plans relating to annexation over the next five years? (Check the best response) _____ Plan to annex more than five square miles of land _____ Plan to annex about one to five square miles of land _____ Plan to annex some land, but less than one square mile _____ Cannot annex; my city does not border any unincorporated areas _____ Can annex, but do not plan to do so If your city cannot or does not plan to annex, please skip to question Q-12. 135 Q-11. How important are the following possible motivations for annexation? In other words, how do these factors affect your plans regarding whether to annex and which properties to annex? (Circle a number between 1, which is “not important,” and 7, which is “very important”) Not Very important important 1 Land for future development to meet housing needs 1 2345 67 2 Land for future development to create jobs 1 2345 67 3 More efficient service provision 1 2 3 4 5 6 7 4 Gain future property tax revenues 1 2 3 4 5 6 7 5 Gain future sales tax revenues 1 2345 67 6 Gain future fee/assessment/ enterprise revenues 1 2345 67 7 Control development of surrounding areas to ensure consistency with city plans 1 2345 67 8 Agreements with county 1 2345 67 9 Agreements with other cities 1 2345 67 10 Direction provided by LAFCO 1 2 3 4 5 6 7 11 Provide greenbelt or open space 1 2 3 4 5 6 7 12 Prevent annexations by other cities 1 2 3 4 5 6 7 13 Other motivation(s) (please specify): 1 2345 67 General background questions: Q-12. Within the boundaries of your city, approximately how many interchanges are there that provide entry to a freeway? (Freeway = interstate highway or limited-access state highway.) _________________________________ Q-13. Please indicate your job title: _________________________________ Q-14. Please check here if you would be willing to participate in a brief telephone interview regarding these topics: __________ 136 Thank you very much for your participation. Your response is crucial to help inform policymakers about the growth challenges facing California and the strategies cities use to deal with these challenges. Please return the survey in the enclosed postagepaid envelope. You will receive a future mailing that will discuss the results of this survey. We welcome your comments on these topics, and comments regarding the questionnaire itself. You may include any written comments below, or on a separate sheet. The Public Policy Institute of California is a private, nonprofit organization dedicated to independent, nonpartisan research on economic, social, and political issues that affect the lives of Californians. For further information, including the full text of PPIC research reports, please visit our World Wide Web site at www.ppic.org. 137 Bibliography Advisory Commission on Intergovernmental Relations, Local Revenue Diversification: Local Sales Taxes, Advisory Commission on Intergovernmental Relations, Washington, D.C., 1989. Barber, Mary Beth, “Local Government Hits the Wall,” California Government and Politics, 1994–95 Annual, California Journal Press, Sacramento, 1994, pp. 82–84. Benson, Mitchel, “Do Exemptions Threaten State’s Sales-Tax Base?” Wall Street Journal, Western edition, May 7, 1997. Berry, Brian J. L., et al., Market Centers and Retail Location: Theory and Applications, Prentice Hall, Englewood Cliffs, New Jersey, 1988. Brooks, Lea, “The Tax Wars,” Golden State Report, June 1988, pp. 32– 38. Bunnell, Gene, “Fiscal Impact Studies as Advocacy and Story Telling,” Journal of Planning Literature, Vol. 12, No. 2, 1997, pp. 136–151. California State Controller, Financial Transactions Concerning Cities of California, annual publication, State Controller’s Office, Sacramento (publication renamed Cities Annual Report for 1995–96 fiscal year). 139 Chapman, Jeffrey I., Proposition 13 and Land Use, Lexington Books, Lexington, Massachusetts, 1981. Chapman, Jeffrey I., Proposition 13: Some Unintended Consequences, Occasional Paper, Public Policy Institute of California, San Francisco, 1998. Cole, Rick, “Sales Tax War Hurts Civic Economies,” Pasadena StarNews, February 22, 1998. Coleman, Michael, Financing Cities: City Financing in the Decades after Proposition 13, League of California Cities, Sacramento, 1998. Dardia, Michael, Subsidizing Redevelopment in California, Public Policy Institute of California, San Francisco, 1998. Davies, Ross L., Marketing Geography, Retailing and Planning Associates, Corbridge, England, 1976. Dillman, Don A., Mail and Telephone Surveys: The Total Design Method, John Wiley & Sons, New York, 1978. Division of Tourism, California Trade and Commerce Agency, California Travel Impacts by County, 1992–1996, Prepared by Dean Runyan Associates, Portland, Oregon, 1998. Dresch, Marla, and Steven M. Sheffrin, Who Pays for Development Fees and Exactions? Public Policy Institute of California, San Francisco, 1997. Due, John F., and John L. Mikesell, Sales Taxation: State and Local Structure and Administration, 2nd ed., Urban Institute Press, Washington, D.C., 1994. Flanagan, James, “The Problem Is Larger Than Housing,” Los Angeles Times, August 30, 1998. Fulton, William, “Cities, Counties Seek to Resolve Annexations Fight,” California Planning & Development Report, Vol. 12, No. 6, 1997a, pp. 1, 9. 140 Fulton, William, The Reluctant Metropolis: The Politics of Urban Growth in Los Angeles, Solano Press Books, Point Arena, California, 1997b. Fulton, William, “Twenty Years of Proposition 13: Tax-Cutting Initiative Shaped Planning and Development in State,” California Planning & Development Report, Vol. 13, No. 6, 1998, pp. 1, 8–10. Ghaus, A. F. Aisha, “Optimal Local Sales Tax,” Urban Studies, Vol. 32, No. 8, 1995, pp. 1369–1381. Guy, Clifford, The Retail Development Process: Location, Property, and Planning, Routledge, London, 1994. International Council of Shopping Centers, “NRB/Shopping Centers Today 1995 Shopping Center Census,” International Council of Shopping Centers, New York, 1996. Jobber, David, “Improving Response Rates in Industrial Mail Surveys,” Industrial Marketing Management, Vol. 15, 1986, pp. 183–195. Kotin, Allan, “Economic Development Can Be Bad for Your Community’s Fiscal Health,” Metropolitan Forum Project, web page http://www.democracynet.org/metro/new/kotinapr.html, n.d. Kotin, Allan, and Richard Peiser, “Public-Private Joint Ventures for High Volume Retailers: Who Benefits?” Urban Studies, Vol. 34, No. 12, 1997, pp. 1971–1986. Krmenec, Andrew J., “Sales Tax as Property Tax Relief? The Shifting Onus of Local Revenue Generation,” Professional Geographer, Vol. 43, No. 1, 1991, pp. 60–67. Kroll, Cynthia, and Eileen Marrinan, “California Retail Sales: Growth Patterns and Locational Tendencies,” Center for Real Estate and Urban Economics, Working Paper 85-95, University of California, Berkeley, 1985a. Kroll, Cynthia, and Eileen Marrinan, “Siting Retail Space in the 1980s: A California Perspective,” Center for Real Estate and Urban Economics, Working Paper 85-96, University of California, Berkeley, 1985b. 141 League of California Cities, Conflicts at the City’s Edge: Annexations, Development and Tax Negotiations, prepared by Crawford Multari & Starr, San Luis Obispo, California, 1996. League of California Cities, “Facts about City Dollars,” pamphlet, n.d. Legislative Analyst’s Office, State of California, Making Government Make Sense, Sacramento, July 21, 1994. Legislative Analyst’s Office, State of California, Why County Revenues Vary: State Laws and Local Conditions Affecting County Finance, Sacramento, May 7, 1998. Lewis, Paul G., Shaping Suburbia: How Political Institutions Organize Urban Development, University of Pittsburgh Press, Pittsburgh, 1996. Lewis, Paul G., Deep Roots: Local Government Structure in California, Public Policy Institute of California, San Francisco, 1998. Miller, Gary J., Cities by Contract: The Politics of Municipal Incorporation, MIT Press, Cambridge, Massachusetts, 1981. Misczynski, Dean J., “The Fiscalization of Land Use,” in John J. Kirlin and Donald R. Winkler (eds.), California Policy Choices, Vol. 3, University of Southern California, Los Angeles, 1986. Newman, Morris, “Legislature Questions Retailing Subsidies,” California Planning & Development Report, Vol. 13, No. 4, 1998, pp. 1–2. Orfield, Myron, Metropolitics: A Regional Agenda for Community and Stability, Brookings Institution Press, Washington, D.C., 1997. Peirce, Neal R., “Neal Peirce on Prop 13 Terrors: They’d Do It All Again!” Metropolitan Forum Project, Worldwide Web site http://www.democracynet.org/metro/new/peircejuly.html, n.d. Peterson, Paul E., City Limits, University of Chicago Press, Chicago, 1981. Ross, Bernard H., and Myron A. Levine, Urban Politics: Power in Metropolitan America, F. E. Peacock Publishers, Itasca, Illinois, 1996. 142 Salant, Priscilla, and Don A. Dillman, How to Conduct Your Own Survey, John Wiley and Sons, New York, 1994. Schneider, Mark, “The Market for Local Economic Development: The Growth of Suburban Retail Trade, 1972–1982,” Urban Affairs Quarterly, Vol. 22, No. 1, 1986, pp. 24–41. Schrag, Peter, Paradise Lost: California’s Experience, America’s Future, New Press, New York, 1998. Schwartz, Jonathan, “Prisoners of Proposition 13: Sales Taxes, Property Taxes, and the Fiscalization of Municipal Land Use Decisions,” Southern California Law Review, Vol. 71, No. 1, 1997, pp. 183–217. Senate Local Government Committee and Assembly Local Government Committee, State of California, Land Use and Local Revenue-Sharing: Playing the Zero-Sum Game, Summary Report of a Joint Interim Hearing, Sacramento, November 17, 1989. Shires, Michael A., Patterns in California Government Revenues Since Proposition 13, Public Policy Institute of California, San Francisco, 1999. Shuit, Douglas P., “The Retail Wags the Dog,” Los Angeles Times, July 17, 1998, p. B-2. Sokolow, Alvin D., “State Rules and the County-City Arena: Competition for Land and Taxes in California’s Central Valley,” Publius, Vol. 23, Winter 1993, pp. 53–69. State Board of Equlization, 1997–98 Annual Report, Sacramento, 1998. Sudman, Seymour, “Mail Surveys of Reluctant Professionals,” Evaluation Review, Vol. 9, No. 3, 1985, pp. 349–360. Vercoutere, Michele E., “The Local Sales Tax in California: Its Significance to the Urban Structure,” Master’s Thesis, School of Architecture and Urban Planning, University of California at Los Angeles, 1979. 143 About the Authors PAUL G. LEWIS Paul Lewis is a research fellow at the Public Policy Institute of California. He studies local government, urban and suburban growth issues, and transportation. He is also interested in the nature and causes of long-term policy change. Among his recent publications are Deep Roots: Local Government Structure in California (Public Policy Institute of California, 1998), Federal Transportation Policy and the Role of Metropolitan Planning Organizations in California (Public Policy Institute of California, 1997), and Shaping Suburbia: How Political Institutions Organize Urban Development (University of Pittsburgh Press, 1996). He holds an A.B. in political science from Indiana University and an M.A. and Ph.D. in politics from Princeton. ELISA BARBOUR Elisa Barbour is a research assistant at the Public Policy Institute of California who works on projects dealing with California governance issues. Before coming to PPIC, she did research examining local labor markets, welfare reform, and urban poverty. She also worked for a number of community and advocacy organizations. She holds a B.A. in political science from Oberlin College and an M.A. in City Planning from the University of California, Berkeley. 145" ["post_date_gmt"]=> string(19) "2017-05-20 09:35:03" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_799plr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:35:03" ["post_modified_gmt"]=> string(19) "2017-05-20 09:35:03" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_799PLR.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) }