Donate
Independent, objective, nonpartisan research

R 298MDR

Authors

R 298MDR

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_298MDR.pdf" ["wpmf_size"]=> string(6) "237697" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(181837) "Subsidizing Redevelopment in California Michael Dardia January 1998 Copyright © 1998 Public Policy Institute of California, San Francisco, CA. All rights reserved. PPIC permits short sections of text, not to exceed three paragraphs, to be quoted without written permission, provided that full attribution is given to the source and the above copyright notice is included. Foreword The development and redevelopment of cities has been embroiled in controversy since the beginning of the century, and redevelopment in California is no exception. Over time, most cities in the state have adopted its use, and as that use has grown, so has the controversy. Longstanding issues have included the tension between affordable housing and downtown development, the application of eminent domain for private rather than public purposes, and the degree of blight in project areas. In this report, Michael Dardia addresses the most recent issue to take center stage in the policy debate: Who pays the bill for redevelopment? The logic of the laws governing redevelopment agencies (RDAs) assumes that redevelopment pays for itself through tax increment financing—that is, RDAs receive any increase in property tax revenues (above a 2 percent inflation factor) in project areas because their investment in area improvements is responsible for increasing property values. But what if the increases are also—even primarily—the result of a general increase in real estate values that would have occurred regardless iii of an RDA’s activities? That question is vital to local governments because tax limitation measures such as Proposition 13 have shrunk the tax and fee base that they can tap, but the public’s demand for services remains unchanged. When an RDA receives property tax revenues through tax increment financing, those revenues are lost to other local jurisdictions—the county, schools, and special districts. If the RDAs are not largely responsible for the increase in property values, those jurisdictions are, in effect, subsidizing redevelopment, with no say in how the revenues are used. Subsidizing Redevelopment in California describes a government function that has aggressively broadened from the traditional role of redeveloping blighted, inner-city areas to include the generation of tax revenues for city governments. After correcting for local real estate trends, the author finds that redevelopment projects do not increase property values by enough to account for the tax increment revenues they receive. Overall, the agencies stimulated enough growth to cover just above half of those tax revenues. The rest resulted from local trends and would have gone to other jurisdictions in the absence of redevelopment. These are important findings for state legislators considering the full range of benefits from redevelopment and who should pay for those benefits in California. Continuing the public-finance focus of two earlier PPIC reports, this publication completes the inaugural series of studies first launched by PPIC in the fall of 1995. The research on assessed values in redevelopment project and control areas was data-intensive and time consuming. It was made possible only by the independence PPIC has to ask questions that require such undertakings. The result is a study that is central to a key policy issue in the state, is based on solid empirical iv evidence from local assessment and tax records, and provides findings that clarify the extent of the financial burdens that existing policy creates for local governments and the state. David W. Lyon President and CEO Public Policy Institute of California v Summary In fiscal year 1994–1995—the most recent year for which figures are available—redevelopment agencies (RDAs) received 8 percent of the property tax revenues collected in the state of California, amounting to $1.5 billion. These are revenues that, absent the RDAs, would have gone to other public agencies such as the state and counties. Given that Proposition 13 and subsequent initiatives have severely limited local governments’ ability to raise or levy taxes, this flow of tax revenues to RDAs merits more public attention and understanding than it has historically received. Although the state has 351 RDAs with over 700 projects under way, not many of the state’s citizens know very much about them and why they are the subject of policy debate. To raise general understanding of redevelopment, this report takes a close look at the purposes of redevelopment agencies, their incentives, and how they operate. To illuminate the policy debate, it focuses primarily on the issues surrounding the property tax revenues that RDAs have historically received and how recent reform legislation has affected vii RDAs’ activities and revenues. It examines how the attraction of these property taxes can affect RDAs’ judgment about which areas to label as blighted and how RDAs affect property values in redevelopment project areas. The report does not provide an overall cost-benefit analysis of redevelopment in California. Over its 50 years of existence, redevelopment has pursued a number of goals: alleviation of blight, provision of low- and moderate-income housing, downtown redevelopment, and overall economic development. Some of these goals are ill-defined, and several conflict, making it difficult to assess the benefits of redevelopment quantitatively. Redevelopment’s full costs are also difficult to measure, given the mix of public and private financing and the myriad spillovers and cross-subsidies that exist. The report focuses, instead, on the extent to which other jurisdictions may unintentionally subsidize redevelopment in California and provides the only estimate available of the size of these subsidies. What Do RDAs Do and Why Are They Controversial? Redevelopment agencies were authorized by the state legislature in 1945 to combat urban blight, and since then more than three-fourths of all the cities in California have established RDAs. The definition of blight is vague, allowing for a wide range of neighborhoods to be considered blighted. According to the law, a blighted area is supposed to be a “serious physical and economic burden on the community which cannot reasonably be expected to be reversed or alleviated by private enterprise or governmental action, or both, without redevelopment.” Each city decides which of its areas are blighted, and there is little viii effective state oversight to police these decisions. Virtually all of the redevelopment activity in the state is undertaken by city agencies, although 15 counties also have active RDAs. These RDAs engage in a wide variety of activities, from providing municipal infrastructure such as streets, lighting, and parking structures, to assisting in the provision of affordable housing and renovating downtown commercial areas. Although redevelopment has its roots in federal urban renewal programs established in the 1940s, more than half of the RDAs in California were created since the passage of Proposition 13 in 1978. Fiscal pressures on local governments made the property tax revenues received by RDAs especially attractive to city governments. The mechanism by which the RDAs receive property taxes, called tax increment financing, was created in 1952 to enable cities to match the share provided by federal programs. Under tax increment financing, an RDA receives all the property tax increases in a redevelopment project area for the life of a project—often 30 years or more. Had an area not been designated as a project area, those same property tax revenues would have been split among the counties, cities, school districts, and other special districts. The allotment to each jurisdiction varies by county, but across the state the average share is 11 percent for cities, 21 percent for counties, 51 percent for school districts, and 17 percent for all the special districts (not including the RDA shares). Given the fiscal pressures created by Proposition 13 and the large number of projects created after Proposition 13 passed, it was probably inevitable that the conflict over tax increment financing would heat up. Although school districts have had their property tax losses compensated for by the state government, counties and special districts are not so favored. The rapid rise in RDAs and their project areas meant further ix pressure on counties’ and special districts’ revenues. Tax increment’s share of statewide property taxes grew from 4.6 percent to just under 8 percent over the last ten years; in some faster-growing counties with aggressive RDAs, tax increment revenues now account for as much as 18 percent of all property taxes. To the other jurisdictions, the tax increment revenues that RDAs received were monies diverted from their coffers. In recent years, $300 million that would otherwise have gone to counties, and $800 million that would have gone to school districts, instead went to RDAs. As the RDAs see it, these tax increment revenues came from the growth in property values stimulated by their activities. Recognizing the strain that tax increment financing could cause other local jurisdictions, the state legislature allowed them to negotiate with RDAs to recover some of the tax increment. These payments, called pass-throughs, grew rapidly in the 1980s. Over the last ten years, pass-throughs increased from 4 percent of all tax increment revenues to almost 14 percent; counties collected about two-thirds of this amount. School districts had little incentive to fight for the pass-throughs, since the state reimbursed them for any lost property tax revenue. The cost to the state for this reimbursement grew rapidly to hundreds of millions of dollars per year. Along with perceived abuses in the kinds of areas that were declared blighted, these costs helped lead to reform legislation (AB 1290) in 1993. AB 1290 instituted a number of reforms, including setting a uniform pass-through rate of approximately 33 percent, more than double the current average rate. This was an implicit recognition that the subsidies flowing to RDAs via tax increment financing, of whatever size, were too large. However, no estimates of the size of these subsidies existed at that time, so the size of this adjustment relied more on political judgment than on empirical evidence. x Purpose and Approach of the Study To estimate the size of the subsidies that RDAs receive from tax increment revenues, the study looked at all the redevelopment projects started between 1978 and 1982, the first five years after Proposition 13 passed. Examining projects begun in those years allows more than enough time for the RDAs to have had a significant effect on property values in the area. Assessed values in the project areas were tracked from 1983, just after the last projects were begun, to 1996. The assessed value was obtained for every parcel in the area in 1996, by using assessor’s office and tax collector’s office data gathered from each county. Cost considerations restricted the original sample to those counties with at least three projects begun within the five-year period. Ten counties, with 114 projects, met this criterion. To judge the effect of RDA actions on the increase in property values, each project area was matched to a Census Block Group in the same city which served as a control. The RDAs’ effect was measured as the difference between the growth of assessed value in a project area and that in its matched area. Although 1996 assessed values were available electronically, 1983 values could be obtained only from the microfiche records maintained by each county’s assessor or tax collector. Microfiche records were available from only four of these counties, and the inability to match enough 1996 parcels with their 1983 counterparts led to the final sample of 38 projects from three counties: Los Angeles, San Bernardino, and San Mateo. These 38 projects were very similar to the 114 projects in the original sample of ten counties, however, so the results from the final sample can be reasonably extrapolated to the original 114 projects. xi The study also examined the question of the degree of blight in project areas. The combination of a vague legal definition of blight, little oversight, and the strong financial incentive of tax increment has led some critics to charge that many project areas are not suffering from urban blight. AB 1290 made the blight definition somewhat more restrictive, and prior legislation had banned the inclusion of primarily agricultural land in project areas. Nevertheless, with more than 700 projects already in place, the question of how blighted some of these were remains important. Results Since identification of blight is a necessary condition for establishing a redevelopment project, how blighted were the project areas? In the group of 134 projects, 25 consisted of more than 50 percent vacant land. Such projects appear to have development, rather than redevelopment, as their prime purpose; most would probably have been disallowed had they begun after AB 1290 went into effect. Although one cannot assess the physical conditions in project areas in 1980, the Census provides data on conditions related to the degree of blight in an area at that time. There were noteworthy differences between project areas and their cities. The average vacancy and poverty rates were higher, and median incomes and median rents were lower, in the project areas than in the rest of the city. It is difficult to know in retrospect whether the project areas were the worst in each city, but they were worse off than the average area in these cities. Given that most project areas were more blighted than the rest of their cities, how much of the property tax revenues that RDAs receive was due to their effect on local property values? Any difference between xii what they received and what they generated can be considered an involuntary subsidy from other local jurisdictions. From 1983 to 1996, assessed value in the 38 projects studied grew by an average of 270 percent, compared with 144 percent in the matched areas. Two-thirds of the project areas outgrew their matches in property value, but 11 project areas grew more slowly. Because the RDAs kept more than five-sixths of any increases in property taxes within project areas, they needed to grow much faster than the control group areas to be considered self-financing. After the analysis removed other influences on assessed value growth and gave the projects credit for any pass-through payments made, only four project areas grew fast enough to justify the RDAs’ claim that they alone generated the increase in property taxes. Two other projects generated at least 90 percent of the property taxes received, and two more generated at least 80 percent of their tax revenues. In other words, fewer than onequarter of the projects came close to being responsible for the property tax revenues they received. These projects were also the ones with the most vacant land: The eight fastest-growing projects began with an average of more than 50 percent vacant land, compared with only 14 percent in vacant land for the other 30 projects. This raises again the issue of whether these projects were engaged more in development than in redevelopment. Given the size of each project’s tax increment revenues, how large was the overall subsidy? In dollar value, the projects collectively generated an estimated 51 percent of the tax increment revenues they received in fiscal year 1994–1995. The total across all 38 projects was $78 million in tax increment revenues, of which almost $40 million can be explained by the actions of the RDAs. Thus, the subsidy from other jurisdictions to the RDAs totaled $38 million in that year alone. If the xiii same pattern held for the other 76 projects in the ten counties originally examined, the total subsidy for redevelopment for the 114 projects would be $170 million annually. Had these projects been subject to the higher pass-through rates required under AB 1290, these projects would have generated only 61 percent of the tax revenues they received. Recommendations To date, there have been no estimates of the size of the subsidies provided to redevelopment through the tax increment financing system. The legislature has implied, through changes mandated under AB 1290, that the existing pass-through rates (determined through project-specific negotiations) were too low, and therefore that the subsidies were too large. However, these decisions had to be made with insufficient information. This study has demonstrated that half of the tax increment revenues received by those projects begun between 1978 and 1982 total about $170 million annually. The legislature may well intend subsidies of this size or larger. Redevelopment does provide infrastructure, affordable housing, and commercial development. Given that the state, counties, and special districts provide subsidies of this size, however, it might be worth targeting these monies more narrowly on the most blighted areas. It might also be worth keeping closer supervision on the activities that are being subsidized. With this in mind, the study makes four recommendations: 1. The legislature should formally clarify its goals for redevelopment. 2. Blight conditions need to be aligned with the goals for redevelopment and made more precise. 3. There should be a formal oversight authority to monitor RDA behavior. xiv 4. If the legislature intends redevelopment to be self-financing rather than highly subsidized, the pass-through rate should be increased significantly. xv Contents Foreword..................................... iii Summary..................................... vii Figures ...................................... xix Tables ....................................... xxi Acknowledgments ............................... xxiii 1. INTRODUCTION ........................... What Is Redevelopment and How Is It Financed? ........ Why Is Redevelopment Financing Controversial? ........ Purpose and Approach of This Study ................ Overview of the Report ......................... 1 2 4 6 9 2. THE REDEVELOPMENT PROCESS .............. Redevelopment Agencies ........................ Redevelopment Projects ........................ Project Financing ............................. Size of Redevelopment Activity .................... 11 11 13 15 16 3. POLICY ISSUES ............................. How Blighted Are Project Areas? ................... Do RDAs Generate Enough of the Growth in Project Areas to Earn Their Tax Increment? ................. Who “Loses” Under Tax Increment Financing? ......... Counties ................................. 23 25 28 29 30 xvii State Government ........................... 32 How Other Jurisdictions Get Some Relief: Fiscal Review and Pass-Throughs ........................ 32 4. STUDY DESIGN AND METHODOLOGY .......... Study Design ............................... Sample Project Selection ........................ Matched-Pair Selection ......................... Assessed Value Determination ..................... Parcel Description ............................ Benchmarks for Analysis ........................ Potential Sources of Bias ........................ 39 40 42 45 48 49 52 53 5. RESULTS OF THE STUDY ..................... 57 How Blighted Are Project Areas? ................... 57 Are RDAs Subsidized by Other Governments? .......... 60 6. CONCLUSIONS AND POLICY IMPLICATIONS ...... Conclusions ................................ The Extent of Blight and the Size of Subsidies ......... AB 1290: The Most Recent Reform Attempt ......... Beyond AB 1290: Addressing the Larger Policy Questions ... 69 70 70 72 73 Appendix A. Policy Issues Concerning Affordable Housing in Redevelopment Projects ........................ 79 B. Sales Tax and Redevelopment ..................... 87 C. Detailed Discussion of Study Methodology ............ 91 D. Regression Analysis............................ 99 E. Final Sample Projects .......................... 103 Bibliography .................................. 105 About the Author ............................... 109 Other PPIC Publications........................... 110 xviii Figures 2.1. Statewide RDA Revenues ..................... 2.2. Statewide RDA Expenditures .................. 2.3. Statewide RDA Revenues and Expenditures ......... 2.4. Cumulative Statewide Operating Expenditures, 1984–1985 to 1993–1994 .................... 3.1. Tax Increment’s Share of Property Taxes, by County ... 3.2. Pass-Through Share of Statewide Tax Increment ...... 3.3. Statewide Distribution of Pass-Throughs ........... 3.4. Local Government Pass-Throughs ............... 5.1. Redevelopment Areas as a Percentage of City Land Area .................................. 5.2. Growth in Assessed Value, 1983 to 1996 ........... 5.3. Share of Tax Increments “Earned” in FY 1994–1995, by Project ................................. 5.4. Net Subsidies to RDAs in FY 1994–1995, by Project ... 5.5. RDAs Generated Half of the Tax Increment They Received ................................ 17 18 19 20 31 34 35 37 58 61 63 64 66 xix A.1. Operating Expenses for RDA Housing Activities ...... 84 B.1. Redevelopment Projects and City Sales Taxes ........ 88 xx Tables 3.1. Legal Definitions of Conditions That Cause Blight .... 26 4.1. Original Sample Projects, by Year ................ 43 4.2. Comparison of Original and Final Samples in 1980 .... 45 4.3. Use of Blight Categories to Justify Sample Projects ..... 46 4.4. Comparison of Final Projects, Matched Areas, and the Rest of the City in 1980 ...................... 48 4.5. Summary Characteristics of Parcels in Final Sample .... 50 4.6. Distribution of Use Codes in Project and Matched Areas .................................. 50 5.1. Area Characteristics in 1980 ................... 59 5.2. Project Area Characteristics in 1980 .............. 60 A.1. Redevelopment Housing Activities, by Income Category ............................... 85 C.1. Breakdown of Parcel Changes, 1983–1996, by County .. 95 D.1. Estimation of RDAs’ Effect on AV Growth ......... 101 E.1. Final Sample Projects ....................... 103 xxi Acknowledgments On a project with as long, and as difficult, a gestation as this one, the author cannot help but be deeply indebted to the many people who contributed to the final product. Projects of this scope are possible only with the ongoing insights, comments, and assistance of colleagues, and at PPIC the author has been blessed with an outstanding group of fellow researchers and support staff. The greatest debt is to Laura Mameesh, a superb research assistant at PPIC who has served alternately as programmer, editor, sounding board, temp manager, and sanity check for the past two years. Eileen Roush, now at the California Department of Finance, also assisted in much of the early data collection and interviewing. Those at PPIC who have made significant contributions as formal or informal reviewers include Mark Baldassare, John Ellwood, Paul Lewis, Joyce Peterson, Fred Silva, and Michael Teitz; the usual caveat that remaining errors are the author’s alone certainly applies here. Outside technical reviewers also provided invaluable suggestions. The author xxiii thanks Timothy Bartik of the W.E. Upjohn Institute for Employment Research; Steven Sheffrin, from the University of California, Davis; John Landis, from the University of California, Berkeley; and Cliff Graves, President of Bay Area Life Sciences Alliance (formerly Executive Director, San Francisco Redevelopment Agency) for their time and thoughtful comments. Over the course of this project, the author interviewed many professionals in the redevelopment community, including redevelopment agency staff, state legislative staffers, consultants, lawyers, and government analysts. Several people were particularly important in the early phases of the project, as well as invaluable sources of experience throughout the project: Larry Arceneaux, President of Katz Hollis; Dave DeRoos, then Deputy Director of the California Redevelopment Association; and Peter Detwiler, consultant to the Senate Housing and Land Use Committee. The author thanks them for their generous cooperation. Others who provided valuable assistance on a regular basis were Wayne Beck, at the State Controller’s Office; Marianne O’Malley, of the Legislative Analyst’s Office; and Jennifer Swenson, at the California Research Bureau. The California Redevelopment Association was kind enough to allow the author and research assistants to attend a number of association conferences and provided many detailed comments on an early draft of this report. In particular, William Carlson, Executive Director, and Chris Hart, Legislative Analyst, have been helpful in this regard. This project was difficult and time consuming because of the data collection and validation required to identify and obtain historical parcel records for over 50,000 parcels in four counties. The most heartfelt thanks go to the temporary employees who carefully and tirelessly xxiv entered vast amounts of data from microfiche records: Paul Cheney; Tana Davis; Raul Poza; and Regis St. Louis; and to Phyllis and Bob Stratford, who had the unenviable task of reconciling parcel boundaries by poring over assessors’ maps in each of the four counties. Much assistance was provided by the staff in each county’s assessor’s office, and in the San Mateo Tax Collector’s Office. In particular, the author thanks Garo Megerdichian in Los Angeles, Ruby Murphy in San Bernardino, and Janet Dick in San Mateo for their tireless help; all three exemplified a researcher’s ideal of a civil servant and reflected great credit on their offices. Many more people than can be listed here have helped in one way or another, and the author also extends thanks to those whose names are omitted. xxv 1. Introduction In 1945, California passed legislation authorizing establishment of redevelopment agencies (RDAs) to alleviate “urban blight.” Subsequently, most cities and many counties have formed RDAs. In 1994–1995, the 351 RDA projects then active spent $3.3 billion to build almost 5,000 (primarily low-income) housing units and to rehabilitate almost 6,000 others, to provide infrastructure and parking for downtown commercial areas, and to build other public facilities. RDAs also displaced more than 500 households from their homes, received $1.5 billion in property taxes that would otherwise have gone to other public agencies, and initiated several new projects that consisted of thousands of acres of vacant land. This range of activities and outcomes helps explain why redevelopment generates so much policy controversy. Much of redevelopment’s appeal—and much of the opposition to it—revolves around tax increment financing: Once an RDA forms a project area, any increase in property taxes in that area (above a 2 percent annual inflation 1 factor) goes to the RDA. The rationale for this is that the improvements it has made in the area are responsible for the increase in property tax assessments. However, it is possible that other forces are contributing to a general rise in real estate values across an entire county. If so, diversion of this revenue to the RDA from the county, schools, and other local entities could be characterized as an involuntary subsidy. For example, money that counties could use to support state-mandated health or welfare services is being used instead by the agencies, and counties have no say in how it is used. It may be that higher state goals for redevelopment are well served by this subsidy and that the other taxing agencies can legitimately be expected to provide it. However, debate over this issue has so far operated in the absence of any hard evidence on the size of the subsidy. The purpose of this report is to raise general understanding of redevelopment by looking at the purposes of RDAs, their incentives, and how they operate. More important for the policy debate, it analyzes for the first time how much of the tax revenue RDAs receive is actually the result of their effect on property values and, thus, on property tax revenues. This introduction sets the stage by briefly addressing three questions: What is redevelopment? How is it financed? Why is redevelopment financing controversial? It also describes the purpose and approach of the study and the organization of this report. What Is Redevelopment and How Is It Financed? The California Community Redevelopment Act of 1945 enables any city or county to establish a redevelopment agency to combat urban blight—a power that three-fourths of the cities and a number of counties 2 in the state have since exercised. Over 97 percent of all RDA financial activity takes place in cities’ agencies, making redevelopment primarily a city-level phenomenon.1 Once established, RDAs designate specific redevelopment projects that focus on those portions of their city that they declare to be “blighted.” Within the boundaries of these projects, RDAs will purchase and assemble lots for use by developers, raze or build structures, provide infrastructure such as parking and lighting, and provide affordable housing. Blight is defined in state law as a combination of physical and economic conditions—such as vacant or decrepit buildings, declining property values, poverty, and high crime rates—that prevent private enterprise from developing the area. In areas suffering from such conditions, public intervention is seen as necessary before any private development can occur. It is sometimes the absence or inadequacy of publicly provided infrastructure that is said to deter private developers. To help RDAs overcome blighted conditions, the state gives the agencies some special powers that are controversial. Agencies can assemble property for sale to private parties and can use eminent domain if necessary to acquire private property that they want to sell, often at a discount, to a private developer. This is the only occasion when eminent domain can be used for private rather than public use. RDAs receive some city general funds and federal Community Development Block Grants to plan and to begin work on specific projects. The bulk of ____________ 1Unlike redevelopment authorities in many of the older cities in the East and Midwest, most of the redevelopment activity in California takes place in cities that are themselves suburbs. California has 332 cities with active redevelopment agencies, but only 35 of those cities are classified as central cities by the Census Bureau. Although an imperfect classification, this illustrates the difference between city-based notions of urban blight in California and in older areas of the country. 3 redevelopment activity is financed by issuing debt that is backed by property tax revenues that will accrue to the agency in the future. To facilitate this financing, RDAs are entitled to receive special property tax revenues. Any increase in property taxes that occurs after a project is designated—minus an annual 2 percent inflation adjustment—goes entirely to the RDA, instead of being shared among all the jurisdictions that normally share the area’s property taxes (such as counties and school districts). By law, these payments to RDAs last as long as there is debt outstanding on the project and often go on for 30 or more years. The original logic behind this arrangement—called tax increment financing— is that cities often bear the up-front costs of (re)development and yet have to share the tax revenues stemming from any increases in property values with all the other taxing jurisdictions. To encourage redevelopment2 and more equitably share its costs, tax increment financing was created in 1952 and applies to every redevelopment project. Why Is Redevelopment Financing Controversial? Fifty years after RDAs were authorized, tax increment revenues amount to $1.5 billion per year, or over 8 percent of all property tax revenues collected in the state. As noted above, RDAs are almost exclusively run by cities, and city RDAs receive 97 percent of all the tax increment revenues in the state. Cities receive an average of 11 percent of property taxes, so the tax increment revenues flowing from redevelopment almost double the amount of property taxes received by cities. In a few counties, RDAs receive as much as 18 percent of all ____________ 2The proximate motivation for this law was the matching funds that cities were required to pay to receive federal urban renewal funding. 4 property taxes—almost tripling the cities’ property tax receipts. Such large sums are guaranteed to generate controversy when other local governments are under fiscal stress themselves. The state government is also affected. When property taxes go to RDAs instead of school districts, the state is required to reimburse the school districts for the lost revenue. With school districts receiving an average of 51 percent of property taxes statewide, the state payments could be as large as $800 million per year; the counties’ forgone revenues are approximately another $300 million per year; special districts receive most of the remaining $500 million. Since the property tax constraints imposed by Proposition 13 in 1978, local governments have been searching for new sources of revenue. With the surge in the number of RDAs and redevelopment projects since 1978 have come charges that the fiscal pressures that local governments have experienced, combined with a loose legal definition of blight, create a strong incentive for those governments to identify blight within their boundaries. In practice, blight is a relative rather than an absolute concept. Areas that have been designated as blighted include both the Adams-Normandie neighborhood—site of the L.A. riots—and an affluent residential area in Foster City that needed some infrastructure improvements. Other evidence that the financial incentives from tax increment financing can influence the propensity to declare blight comes from the amount of cities’ land area that is covered by redevelopment projects. Roughly half of California cities have between 11 percent and 30 percent, and another fifth have well over 30 percent, of their land area in redevelopment projects. It seems unlikely that such large fractions of so many cities could be suffering from urban blight. Counties especially think that many cities view their redevelopment projects as a blank check 5 with which to transfer tax revenue from counties and the schools to their own RDAs.3 Although RDAs may provide useful public goods such as streets, parking structures, lighting, and other facilities, other tax jurisdictions would rather that cities use their own general funds to provide these goods. Cities and their RDAs reply to these charges by insisting that without their redevelopment activities, there would be no increase in property values to be taxed. The state legislature recognized and addressed some problems with redevelopment through the passage of AB 1290 in 1993. Among other things, the legislation tightened the definition of blight somewhat and changed the distribution of tax increment revenues. It arrived at a political compromise by requiring that any new RDA projects share approximately a third of their tax increment revenues with the other local taxing jurisdictions. Before this legislation, any such payments (called pass-throughs) were negotiated for each project case by case. These passthroughs averaged 14 percent in fiscal year 1993–1994, and counties received almost two-thirds of these payments. Purpose and Approach of This Study There have been no analyses to date of how much of the property value increases can be ascribed to the efforts of the RDAs rather than to overall real estate trends. In this study, we measured the effect of RDAs on assessed value growth to see whether their effect was large enough to ____________ 3See, for example, testimony by Los Angeles County to the Senate Committee on Local Government on December 7, 1989 (Report 457-S). In recent years, both Santa Clara and Los Angeles Counties have sued redevelopment agencies in the cities of San Jose and Los Angeles over a bond measure and project extension, respectively. (Both suits were ultimately unsuccessful.) San Diego and Riverside Counties dropped lawsuits against redevelopment agencies after negotiating pass-through agreements (see California Legislative Analyst’s Office, 1994). 6 be self-financing—which is the rationale for their existing arrangements. Any shortfall means that at least some of the RDAs’ activities have, in effect, been subsidized by other governments, and the study estimates the size of these subsidies. Because RDA project areas have been judged to be blighted, comparing their assessed value growth to that of the entire city they lie within might be misleading. If RDA project areas really are blighted, their growth prospects may well be worse than those of the rest of the city. To properly account for their initial condition, we compared RDA project areas with other areas in the same city that were similar to the projects in terms of poverty and vacancy rates. We then compared the change in assessed value from 1983 to 1996 for all the properties in each redevelopment project area with the change in its matched area. This growth in assessed value is what determines the amount of tax increment revenues that the RDA receives. We looked at all the redevelopment projects that began in the first five years after Proposition 13 passed in 1978. This allowed the projects between 14 and 19 years— half or more of their expected lifespan—to have produced an effect on the area.4 Out of 114 projects in ten counties, we were able to get sufficient historical records for 38 projects in three counties. These 38 projects were very similar to the original 114 projects, so the results can be considered representative of the larger group. RDAs cannot be considered self-financing simply because assessed values in the project areas grew faster than in their matches. Merely growing faster would demonstrate that the RDAs’ activity had at least some positive effect on local property values, but it is not sufficient proof that they were responsible for all the tax increment they received. The ____________ 4For cost reasons, the sample excluded any projects in counties with fewer than three projects started between 1978 and 1982. 7 sample projects kept 84 percent of the increase in property tax revenues that occurred within the project area, whereas other jurisdictions retained 16 percent. To justify the claim that they are self-financing, project areas therefore needed to grow six times as fast as their matched areas. Our study calculated how many of the projects actually had such rapid growth rates. It also looked at the tax-sharing formula mandated by the reform legislation in light of the actual growth rates. We also examined whether the reform legislation, which increased the amount of new property taxes that RDAs must share with other governments, picked the correct sharing formula if the program is meant to be self-financing. Under the new legislation, future redevelopment projects will keep an average of only 66 percent of any increases in property taxes; the study’s results allow estimation of the size of the subsidies implied by this pass-through rate. It is important to note at the outset what this study does not address. The study estimates the size of state and local government subsidies to redevelopment agencies in California. It does not analyze how effectively the RDAs have spent these subsidies, or whether their benefits have exceeded their costs. Any cost-benefit analysis of redevelopment is complicated by the vague and often conflicting goals embodied in its history. Redevelopment grew out of urban renewal efforts in the 1940s and initially focused on substandard housing; over the next two decades, it grew to encompass downtown renewal as well. The tension between housing and downtown development goals continues to this day, as numerous legislative hearings over the last several years readily attest.5 There have been recent debates over whether general economic ____________ 5See reports of previous Senate Housing and Land Use Committee hearings (1989, 1995, and 1996). 8 development is a legitimate purpose for redevelopment agencies, and legislation has also been passed to allow the use of redevelopment powers in communities affected by earthquakes and military base closures. Given the wide range of competing goals, there is no clear yardstick against which RDAs should be measured, and there is no single group whom they are to serve. Should they be judged by how much they improve the project area or the entire city? Are their achievements better measured by improvements in the lot of local residents or of the local property market? Furthermore, there may be goals for redevelopment that are not quantifiable even when universally agreed upon: Attributes such as “quality of life” or “more vibrant” communities do not readily lend themselves to economic analysis. Ultimately, the value placed on redevelopment projects’ achievements may outweigh concerns about whether RDAs are heavily subsidized. Nevertheless, in this era of constrained fiscal resources for local governments in California, it is important to know how much one level of government is subsidizing another. Any subsidies to RDAs might otherwise be spent by counties and the state on other programs that might also have significant public benefits. Once the size of the subsidies is known, policymakers will be in a better position to decide how they want to allocate those resources. Overview of the Report Chapter 2 provides a more detailed discussion of RDAs, redevelopment projects, how projects are financed, and the size of redevelopment activity in California. These descriptions provide the background necessary for the discussion in Chapter 3 of the policy issues raised by tax increment financing and blight determinations. Chapter 4 then presents the design and methodology of the analysis that addresses 9 these issues. The results and implications of the study are discussed in Chapters 5 and 6, respectively. Several appendices address related issues. Some persistent questions about redevelopment housing activities are discussed in Appendix A and questions about sales taxes are addressed in Appendix B. A more detailed discussion of the study methodology is presented in Appendix C, followed by results of an area-level regression analysis. Finally, Appendix E lists the redevelopment projects included in the final sample. 10 2. The Redevelopment Process Redevelopment is a broad, multifaceted undertaking in California. There is a great deal of variation among the kinds of communities that engage in redevelopment, as well as among the activities undertaken in the hundreds of project areas. The intent in this chapter is to provide background only on the aspects that primarily bear on the issue of tax increment financing. Redevelopment Agencies As discussed in Chapter 1, the California Community Redevelopment Act of 1945 was passed to help local governments eliminate urban blight. (The definition of blight and the controversy surrounding the practice of blight determination are discussed in Chapter 3.) The law allows the governing body of any city or county to establish a redevelopment agency if it declares that there is a need for one. In the first 20 years after the law was passed, only 46 agencies were formed throughout the state; in the next 20 years, 266 were formed. In 11 the middle of that period, Proposition 13 was passed, with its rollback of property taxes and a strong limit on the rate at which property taxes can increase. By 1986, most cities in California had an RDA; since then, only another 76 agencies have been formed. More than three-fourths of the cities and almost half of the counties have formed redevelopment agencies. However, not every local government that has formed an RDA necessarily uses its powers: In 1994–1995, 10 percent of all city agencies and 38 percent of county agencies reported no financial activity. Most of the inactive city agencies were those from cities with a population of less than 25,000; all but one city larger than 100,000 people have active redevelopment agencies. When a city or county forms an agency, a public referendum is required to ratify the decision to create the agency, since once it is formed it is endowed with substantial powers.1 In 90 percent of the agencies, the governing board of the RDA is the city council (and most of the staff are city staff), 6 percent of agencies use the county board of supervisors, and 4 percent have a separate organization.2 The larger city agencies tend to have a separate organization to oversee redevelopment activities—in such cases, the governing board is appointed by the city council. In most other agencies, the city and the redevelopment agency are essentially the same. City staff from the appropriate office carry out the relevant functions—for example, finance or planning—and the RDA’s budget is billed accordingly. RDAs can purchase and sell property, make loans or grants, enter into leases and other contracts, seize ____________ 1Since most RDAs are formed by cities, for clarity’s sake the discussion in the text will treat agencies as if they were city RDAs. Since 97 percent of redevelopment expenditures occur within city agencies, this approach does not distort the current situation. 2From California Redevelopment Association (1996), Chapter 2. 12 property by eminent domain and transfer it to other private owners, and construct or rehabilitate structures and improvements. The RDA will often work with the city manager or mayor, as well as the planning and other offices, to decide the priorities of the agency and to integrate the agency’s plans with those of the city as a whole. Redevelopment Projects To undertake any redevelopment activity, an RDA must form a project area—a specific geographic area that the RDA identifies as blighted and where it will undertake redevelopment activities. Most agencies have only one or two projects, but the larger agencies can have half a dozen or more projects active at any given time. In 1994, there were 731 redevelopment projects throughout the state. Project areas range from as small as two acres to as large as 22,000 acres, with most projects covering between 100 and 2,500 acres. An RDA must take a number of required steps in the “plan adoption process” by which a project area is formed. The city council adopts a resolution to designate a survey area to study for possible use as a redevelopment project.3 The agency will often have a particular kind of project in mind at this point—such as a shopping mall or housing development—but may not have worked out the details or selected a developer. The overall plan might envision several stages of development occurring over a 20- or 30-year period. Redevelopment projects’ stated objectives are roughly equally divided among commercial, residential, and public uses, and most plans entail a mix of these activities. ____________ 3The final project area may be smaller than the survey area but it must be within the survey area. For more details, see California Redevelopment Association (1996) and California Debt Advisory Commission (1995). 13 Redevelopment of a declining downtown area is one of the most common goals of RDAs, but agencies have undertaken a wide variety of projects. Examples of past redevelopment projects include constructing downtown hotel and parking structures, developing multiplex cinemas and retail malls, building or rehabilitating affordable housing units, and subsidizing freeway interchanges and access roads for industrial and commercial complexes. For a project area to be adopted, the RDA must complete a Preliminary Report (as well as an Environmental Impact Report and a report to the city council). The Preliminary Report establishes the project area’s boundaries and provides the evidence of blight conditions;4 describes the activities the agency intends to undertake to eliminate the blight conditions, including housing activities; and outlines the project’s financing structure. The plan also sets time limits on the life of the project and when it can undertake debt and sets the maximum amount of debt that can be outstanding at any given time. Some of the constraints on project formation and behavior changed under AB 1290, reform legislation passed in 1993. Details of this legislation are discussed in Chapter 3. The Preliminary Report is sent to other local governments that might be affected fiscally by the project and must be discussed at public hearings. If low- or moderate-income residents will be displaced by planned redevelopment activity, the RDA must form a Project Area Committee (PAC), consisting of persons with an interest in the project area. (In practice, PACs are formed in most project areas even if they are not required.) PACs are meant to provide an opportunity for a community to participate in its redevelopment plans. After this input, ____________ 4Blight definitions and related policy issues are discussed in Chapter 3. 14 the city council enacts an ordinance that formally creates the project area; in some cases a referendum is required as well. The planning and adoption process can take as little as several months to as long as several years; commonly, it takes one to two years from the initial steps to formal adoption. Once a project area is formed, the RDA will contract with one or more developers to help implement the redevelopment plan. Although this process sometimes involves open bidding, developer selection is more likely to take place through a request for qualifications (RFQ) or a request for proposals (RFP). To encourage the developer to produce the project it has in mind, the RDA will often provide incentives in the form of land write-downs (where agency-owned land is sold below the market price) and agency-provided infrastructure. The details differ markedly from one project to another, depending on market conditions and the nature of the project’s specific requirements. The RDA generally acquires and prepares the land, engages in cleanup if necessary, and constructs the necessary infrastructure (frequently including parking structures). The improved land is then transferred to the developer as part of the deal. Project Financing Most of the funding in the early years of redevelopment came from federal grants and loans, which typically required local matching funds to cover 25 percent or more of the project cost. To help local governments meet these matching requirements, the state legislature created tax increment financing—by which RDAs keep most of any future increases in property taxes in a project area. This was meant to allow RDAs to become self-financing, but it created a financial incentive for city-based 15 RDAs to define blight more broadly than they otherwise might to capture the future property tax revenues. (Chapter 3 addresses the question of how blighted redevelopment project areas were when they were formed.) The initial planning phase of a project is usually funded by a combination of city general funds (often in the form of loans), Community Development Block Grants, and sometimes funds transferred from other project areas run by the same agency. Occasionally, the developers involved with a project will lend money to the RDA to allow the deal to proceed quickly; they are then repaid out of subsequent tax increment revenues that the agency receives as assessed values grow. Once the project has progressed enough to be considered a reasonable investment risk by outside investors, the RDA will issue longterm debt to finance its activities. These tax allocation bonds are secured by the flow of tax increment revenues over the life of the project area. Lease revenue bonds are sometimes used to construct parking facilities, with parking fees pledged to service the debt. Other common sources of revenues that agencies receive from project areas include sales and use tax from retail sales, transient occupancy tax from hotels within the project area, any profits from the sale of land purchased by the RDA, and lease revenue (typically from parking structures provided as part of the development plan). Size of Redevelopment Activity Redevelopment in California is very big business. Statewide, redevelopment agencies received total revenues of $1.2 billion5 in 1984– ____________ 5All dollar figures are real (i.e., inflation-adjusted) 1993 dollars unless otherwise noted. 16 Revenue (billions of 1993$) 1985 (see Figure 2.1); this increased to almost $2.2 billion in 1993– 1994. Tax increment revenues rose from just under half of total revenues to 69 percent over the period. As happened with other intergovernmental programs, grant revenues declined by 22 percent whereas interest proceeds and “other” revenues were basically flat. Absolute tax increment revenues, however, rose by 64 percent, even after adjustment for inflation. Some of this increase reflects the strong real estate markets in the 1980s, and some comes from the increase in the number of active redevelopment projects under way. All other 2.5 Interest proceeds Grant revenues Tax increment 2.0 1.5 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.1—Statewide RDA Revenues 17 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Expenditures (billions of 1993$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Redevelopment expenditures went from $1.9 billion in 1984–1985 to $3.5 billion in 1993–1994 (see Figure 2.2). The largest share, approximately 40 percent, went to pay principal and interest on outstanding debt. The shares of most expenditure categories were fairly stable,6 although expenditures on real estate purchases fell sharply (to only 4 percent of expenditures) in the most recent year. This decline was All other Planning and design Administration costs 4.0 Interest and debt payments Real estate purchases 3.5 Improvement/construction 3.0 2.5 2.0 1.5 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.2—Statewide RDA Expenditures ____________ 6“Other” includes rehabilitation and relocation costs as well as the “All other” category in the controller’s office reports. No detail is given for the “All other” category, where the sharp increase occurred. 18 Revenues and expenditures (bIllions of 1994$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 most likely due to the effects of the recession on real estate markets in the state, as well as to concerns about reduced tax increment payments resulting from downward reassessments. Several agencies experienced fiscal distress during the recent recession with regard to meeting their debt service payments, and a number of agencies have cut back on new activity until their financial situation stabilizes. The revenues and expenditures imply a large annual deficit in RDA budgets (see Figure 2.3). This seems to be a result of the combination of financial and operational activity in their financial statements. As described above, RDAs sell bonds early in each project and use the proceeds to engage in development; along the way, unused portions of 4.0 3.5 Total expenditures 3.0 2.5 2.0 1.5 Total revenues 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.3—Statewide RDA Revenues and Expenditures 19 the principal can earn interest for the agencies.7 When the projects mature, more tax increment comes in and is used to meet debt service as well as to fund additional development. To get a sense of how agencies spend their money on actual project activities, an “operational” budget, which neglects debt finance expenditures, is shown in Figure 2.4. This is the sum of all nondebt service expenditures from 1984–1985 to 1993–1994 (again in 1993 dollars). All other 30% Improvement/ construction 38% Planning and design 5% Administration costs 13% Real estate purchases 14% SOURCE: California State Controller’s Office (annual). Figure 2.4—Cumulative Statewide Operating Expenditures, 1984–1985 to 1993–1994 ____________ 7Federal regulations prevent agencies from earning higher interest on investments than on debt. 20 The largest operating expense is for project improvement—site clearance, improvement (infrastructure), construction, and building rehabilitation (Figure 2.4). Real estate purchases were only 4 percent of total operating expenditures in 1993–1994, down from almost 30 percent of operating expenses in 1984–1985 and an average of 14 percent over the entire period. The earlier figures may have been higher than usual because of the large number of new agencies and projects begun in the early 1980s. RDAs generally purchase land at the outset of a project although, because of the successive stages of some projects, there may be several periods of land accumulation over the 30-year life of the project. The recent figure is further depressed because many agencies, particularly in Southern California, suffered financial pressures from the recession. These RDAs have cut back their activities substantially to be able to make their debt service payments. These are the basic operating facts about RDAs that are relevant to an examination of the extent to which redevelopment is subsidized in California. The tax increment financing system that underpins redevelopment has raised several critical policy issues, which are discussed in Chapter 3. 21 3. Policy Issues Redevelopment has been the focus of criticism over much of its life. The critics of redevelopment fall into one of several camps.1 Some critics are adamantly opposed to the use of government powers—such as eminent domain—which they view as amounting to involuntary transfers of private property from one owner to another. Some in this camp also view land write-downs—when RDAs sell land to developers below the agency’s cost of acquiring it—as something that should be considered “gifts of public funds.” Gifts of public funds are unconstitutional, and this ban underlies the lack of property tax abatements to specific firms in California (a practice that is widely used in other states).2 ____________ 1This categorization of redevelopment’s critics is based on the author’s interviews with many professionals involved in redevelopment, as well as on testimony given at the Senate Housing Committee hearings in October 1995 (particularly that of William Fulton, editor of California Planning and Development Report). Other summaries of the debate between RDAs and their critics can be found in the proceedings from recent state legislative hearings (California Legislature, 1989, 1995, and 1996). 2The courts have ruled, however, that the practice of land write-downs is constitutional. 23 More to the point of this study, other critics, consisting primarily of other taxing entities, view RDAs’ use of tax increment financing as a diversion of property taxes that are properly theirs. This group often notes that the biggest surge in redevelopment agency and project formation followed the passage of Proposition 13. In their view, RDAs are using tax increment financing as a way around the political obstacles to raising taxes. Lastly, some observers generally support the concept of redevelopment but oppose what they see as glaring abuses in recent years, such as the designation of primarily vacant land as blighted; the focus on attracting retail firms that generate sales tax revenues, such as car dealerships, into project areas;3 and the reluctance of many agencies to provide affordable housing.4 The last set of critics tend to be concerned that some RDAs have lost sight of their obligation to alleviate urban blight in favor of maximizing local tax revenues, thus endangering the status of redevelopment in general. This report focuses on the concerns of critics who claim that the large tax increment revenues that flow to the RDAs are not justified by their effect on assessed values. Implicitly, any property tax revenues received in excess of those revenues generated by the RDAs’ actions amount to involuntary subsidies from other governments. There are currently no estimates of the size of these subsidies, so no one knows how much the current system of redevelopment costs the state and other local governments. Related to this issue is the question of whether the ____________ 3A brief discussion of the role of sales tax revenues in redevelopment can be found in Appendix B. 4Although not a focus of this report, affordable housing has been central to the policy debates about redevelopment since the program’s inception. Appendix A summarizes the history of affordable housing under redevelopment and examines the policy issues that are commonly discussed. 24 financial incentives created by tax increment financing (together with the vague definitions of blight in existing law) cause RDAs to declare more areas to be blighted than actually are. The most recent chapter in redevelopment reform was the state legislature’s enactment of the CRL Reform Act of 1993 (AB 1290). AB 1290 was designed to address several problems with redevelopment that had not been corrected by previous legislation.5 These are the bill’s primary provisions: penalties for RDAs that fail to meet the required level of effort for providing affordable housing, somewhat stricter blight definitions required in setting project boundaries and implementation plans, prohibitions on RDAs’ receiving a portion of sales taxes, specific time limits on projects’ ability to incur debt, and institutionalizing the previously informal transfer payments between RDAs and other taxing entities affected by redevelopment projects. The critics’ practical charges that are examined in this report can be boiled down to two key questions, both of which were addressed by AB 1290: 1. How blighted are redevelopment project areas? 2. Can RDAs justify the tax increment they receive by their effect on growth, or are they subsidized by other governments? How Blighted Are Project Areas? As noted in Chapter 2, RDAs must make a finding of blight to use their powers of eminent domain, tax increment, etc. This flows from redevelopment’s roots in urban renewal programs. The past urban ____________ 5For more details about the problems AB 1290 was meant to address, refer to the reports on recent legislative hearings cited in footnote 1 of this chapter. 25 renewal focus meant that the definition of blight originally concerned residential—specifically slum—areas. If an area was blighted, it was assumed that private industry acting on its own would not invest in the area; concerted government action was therefore required. According to state law,6 a blighted area is one . . . in which the combination of conditions set forth in Section 33031 [listed in Table 3.1] is so prevalent and so substantial that it causes a reduction of, or lack of, proper utilization of an area that it constitutes a serious physical and economic burden on the community which cannot reasonably be expected to be reversed or alleviated by private enterprise or governmental action, or both, without redevelopment. This legal definition of blight is not precise enough to allow a definitive assessment of whether a given area is or is not truly blighted; blight is also supposed to be a characteristic of an entire project area and is not necessarily true of every property in the area. Table 3.1 Legal Definitions of Conditions That Cause Blight Physical Conditions That Cause Blight Unsafe or unhealthy buildings Factors that prevent economically viable use of buildings or lots Adjacent uses that are incompatible and prevent economic use Subdivided lots of irregular form and shape or of inadequate size Economic Conditions That Cause Blight Depreciated or stagnant property values Abnormally high business vacancies Lack of necessary commercial facilities Residential overcrowding that leads to public safety and welfare problems High crime rate SOURCE: Drawn from Health and Safety Code Section 33031. ____________ 6This definition comes from Health and Safety Code Section 33030 (b), as reprinted in Beatty et al. (1995). 26 The possibility of 30 years of tax increment revenues can create a temptation in some cities to be rather liberal in defining blight. As one local official put it in an interview, “blight is might.” The financial incentives to test the limits of the blight concept grew much stronger after the passage of Proposition 13 in 1978. Financially strapped cities could earn a greater share of the property taxes generated within their boundaries by setting up redevelopment projects. Since there is no oversight authority to police redevelopment agencies, the only check on agencies’ behavior is the threat of a lawsuit by some person or jurisdiction claiming to be hurt by an agency’s actions.7 Lawsuits must be filed within 60 days of a plan’s adoption, which can limit the ability of residents and other taxing agencies to block an adopted plan that is objectionable. The broad definitions of blight in the law provide quite a bit of latitude for an agency’s designation of project area boundaries. Table 3.1 lists the main definitions of blight conditions in current state law (i.e., AB 1290). Their breadth raises the question of whether an area is blighted merely because it is a low-income area, since many places throughout California could arguably have at least one of each of these conditions. If all low-income areas may be considered blighted, blight requirements provide little guidance in targeting redevelopment efforts. Before AB 1290 was passed, either physical or economic conditions could deem an area as blighted for purposes of project designation. Under AB 1290, however, both physical and economic blight conditions ____________ 7It could be argued that the threat of action by the state legislature—such as the restrictions imposed by AB 1290—also acts as a check on RDAs’ behavior. Such a check works more through the California Redevelopment Association, with its concern for the collective health of redevelopment in the state, than through any individual agency that has little reason to be concerned with any repercussions to other cities. 27 must be documented to classify an area as legally blighted. AB 1290 no longer permits projects that are not “predominantly (i.e., more than 80 percent) urban,”8 but the inclusion of largely vacant land in project areas was legitimate under state law in earlier years. The new law also requires that project plans for the first five years demonstrate how the agency’s actions would eliminate the specific blight conditions that were used to justify the formation of the project area. AB 1290 also curtails the use of “inadequate infrastructure” and other conditions involving shortcomings in public improvements. The law rescinds agencies’ ability to receive sales tax revenues from firms located in project areas—a post-Proposition 13 authority that was widely used to give rebates to retail stores and auto malls. This change was aimed at stopping the numerous cases of vacant (sometimes agricultural) land being declared blighted to attract sales-taxgenerating businesses within a given city’s borders. Do RDAs Generate Enough of the Growth in Project Areas to Earn Their Tax Increment? Although California was the first state to adopt tax increment financing, it is now used in at least 32 states.9 In theory, tax increment financing can be seen as desirable from a state’s perspective, since cities bear the entire initial cost of redevelopment but, in the absence of tax increment, have to share the future property tax benefits of that ____________ 8Note that “urban” does not necessarily imply no vacant land, although it is assumed to mean not undeveloped. Most of the project areas with large percentages of vacant land apparently were undeveloped. 9See Huddleston (1986) and Stauber and Wyatt (1990) for a more detailed discussion of the theory of tax increment financing and its use in Wisconsin; Stinson (1991) discusses its use in Minnesota. Davis (1989) and Klemanski (1990) discuss the incentives that tax increment financing creates for different levels of government and the political dynamics and lack of accountability that affect its use. 28 redevelopment with other taxing entities. Tax increment financing, therefore, allows for cost-sharing among all the entities that will benefit from subsequent redevelopment. This theory assumes that in the absence of some form of tax-sharing, there will be too little development, since the local city will not recoup enough of its costs through future increases in property tax revenue. In practice, whether tax increment schemes are actually self-financing depends on how much RDAs contribute to local growth. If the increase in property values in project areas is solely the result of the agencies’ actions, there is no real diversion of funds—as the RDAs themselves argue. If some of the increase is due to general trends in real estate markets, however, any tax increment revenue generated by these trends and received by RDAs is essentially an involuntary subsidy from counties and school districts (and ultimately the state) to the cities. Such a subsidy is neither budgeted for, nor approved by, the governments providing it, and it comes at the expense of whatever public programs those governments would otherwise have spent those funds on. Who are these governments and what relief have they received to date? Who “Loses” Under Tax Increment Financing? Use of tax increment financing of redevelopment is one of the biggest sources of contention between different levels of government. Counties (and to some extent the state) see tax increment revenues as property taxes that are diverted from other taxing entities to the cities via redevelopment agencies. 29 Counties Because of both an increase in redevelopment activities and the sharp increase in property values in the 1980s, tax increment revenues in redevelopment project areas grew rapidly as a share of total property taxes. Tax increment’s share was 4.6 percent of all property taxes collected throughout the state in 1984–1985, and grew to just under 8 percent in 1993–1994.10 This varied greatly by county: The share of property taxes accounted for by tax increment revenues ranged from zero in 15 counties to 18 percent in Riverside County. To put the tax increment share into perspective, the statewide average share of property taxes going to cities was 11 percent in 1993–1994, and the overall county share was 21 percent. Each county has its own allocation formula for how property tax revenues are to be shared among all the taxing entities within that county, so the overall tax distribution is unique to each county. Given how tight municipal finances have been, such a realignment of tax revenues would be expected to generate conflict between the affected governments. Figure 3.1 shows the growth of tax increment’s share of property taxes in the top ten financially active counties.11 Although all the counties showed an upward trend, Riverside and San Bernardino Counties stand out with more than a tripling of tax increment’s share of property taxes over the last ten years.12 ____________ 10This overstates the total somewhat, since the totals include amounts transferred to other jurisdictions as pass-throughs (described below). The overall pattern across counties and over time is unaffected. 11Counties were ranked by total 1993–1994 revenues received by the county and its cities, school districts, special districts, and redevelopment agencies. The source was California Department of Finance (1995). 12The uniform decline in increment shares in 1992–1993 is due to absolute declines in tax increment that year, presumably due to payments to the Educational Revenue Augmentation Fund (ERAF); overall property tax revenue still increased. 30 Percentage 1982–1983 1983–1984 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Riverside San Bernardino Los Angeles Santa Clara Contra Costa 20 Orange 18 Alameda San Diego 16 Sacramento 14 Fresno 12 10 8 6 4 2 0 SOURCE: California State Controller’s Office (annual). Legend lists counties in order of tax increment share in 1993–1994. Figure 3.1—Tax Increment’s Share of Property Taxes, by County The California Redevelopment Association (CRA) notes that some counties themselves have active agencies and asserts that some of the variation in tax increment by county represents activity in county-level agencies. A review of county RDA activities shows that this has not been the case, however. Despite the extent of its redevelopment activity, Riverside County’s agency, for example, accounts for only 5 percent of the tax increment revenue generated by all redevelopment projects inside the county. San Bernardino, the county with the second-highest fraction of property taxes diverted to tax increment, generated none. This reflects 31 the general insignificance of county agencies with respect to tax increment: Statewide, county agencies account for an average of only 2.7 percent of all the tax increment revenue generated within the county. Only 11 county agencies generated any tax increment revenues at all in FY 1993–1994. State Government After Proposition 13 passed, the state took on the obligation to reimburse school districts for property tax revenues lost to other taxing entities. Estimating this burden on the state is difficult, but a ceiling on its value is probably $800 million annually. This number assumes that redevelopment affects school districts uniformly across counties and applies the statewide average receipts of 51 percent of property taxes, less schools’ 2 percent pass-through share of total increment. A more accurate accounting would have to examine each county separately and subtract whatever share of the increased assessed valuation is due to redevelopment activities (the subject of this report’s statistical analysis), as well as the money received through the ERAF requirement. It would also need to account for any in-kind contributions, such as buildings and infrastructure, that RDAs might provide to some school districts. How Other Jurisdictions Get Some Relief: Fiscal Review and Pass-Throughs In 1976, AB 3674 gave counties the responsibility to convene fiscal review committees, allowing them to negotiate a return of some of the tax increment revenues generated by projects within their boundaries. After Proposition 13 drastically reduced property taxes, the loss of funds from tax increment became more acute and local governments fought for 32 every source of funds. In 1983, other taxing entities were also given fiscal review powers in recognition of the potentially large effect that tax increment might have on them. In practice, these fiscal reviews have mainly been performed by counties and often result in an agreement by RDAs to share some tax increment with them. Since other taxing entities were given fiscal review authority over proposed redevelopment projects, RDAs started to share some of the tax increment with the affected entities. These payments (pass-throughs) were the result of project-specific negotiations between the RDAs and some of the other entities. The share of tax increment paid back to counties and other entities as pass-throughs has increased significantly over the last ten years, growing from 3.5 percent of total tax increment in 1984–1985 to 14 percent in 1993–1994 (Figure 3.2). This does not necessarily mean that RDAs are agreeing to share more and more of their tax increment in each year’s new projects. Although some of the increase in pass-throughs comes from higher rates, some comes from the big increases in property values that many project areas have enjoyed. Since most of the older projects did not involve pass-throughs, the maturing of projects that did have pass-throughs means that their effects are now starting to show up in financial statements.13 For projects begun after December 31, 1993, AB 1290 fixed the pass-through rate using a formula that was uniform across all projects. The financial data underlying Figure 3.2 do not include any noncash transfers that occur between RDAs and other entities. Some passthrough agreements allow for construction of infrastructure or other ____________ 13Financial data on the sample projects (those started between 1978 and 1982) show that the share of tax increment revenues paid out as pass-throughs does not change much over the life of projects. 33 Percentage 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 AB 1290 35 30 25 20 15 10 5 0 SOURCES: California State Controller’s Office (annual); AB 1290 estimate from author’s calculations. Figure 3.2—Pass-Through Share of Statewide Tax Increment facilities in lieu of, or in addition to, any cash transfers. The CRA asserts that these noncash contributions are a large (but unspecified) amount, and that schools in particular benefit from these arrangements. Schools do have an incentive to receive noncash rather than cash transfers, since the state will not take the former into account when determining how much to reimburse school districts for property taxes lost to redevelopment. In the absence of any source of information about these agreements, it is impossible to estimate how much larger the total passthroughs would be if the noncash transfers were included. In interviews with numerous redevelopment project managers and consultants, none were able to document the value of these capital projects. In addition, 34 Pass-throughs (billions of 1994$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 none thought that the value of the noncash contributions would approach even half the value of the reported (i.e., cash) pass-throughs. The CRA also asserts that pass-throughs are likely to rise (by prior agreement between RDAs and other entities) over the life of a project, since there is little tax increment generated in the early years of a project. Among the sample projects studied for this report, the average passthrough over the 1985–1994 period rose only from 10 percent to 12 percent. Although it rose in 47 of the projects, it also fell in 27 others, so the trend is neither large nor uniform across projects. Figure 3.3 shows how pass-throughs were distributed under the existing pattern of individually negotiated agreements for each project. Counties received 71 percent of all pass-throughs in 1984–1985, 2.5 Special districts 2.0 School & CC districts Cities 1.5 Counties 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 3.3—Statewide Distribution of Pass-Throughs 35 although their share fell somewhat—to 62 percent in 1993–1994. As Figure 3.3 shows, the falling county share was largely due to the growth in school and community college (CC) districts’ pass-through payments. Despite this falling share, however, the absolute dollars that counties received in pass-throughs increased from under $20 million to almost $150 million over the last ten years. The formula under AB 1290 made the pass-through rate uniform across all projects. The formula uses a sliding scale, with pass-throughs starting at 25 percent of the increment;14 another 21 percent is added to increases in assessed values after the tenth year, and another 14 percent is added for any increases after the thirtieth year. The total of the passthroughs over the life of a project depends on how much and how quickly assessed values rise. Using historical assessed values and inflation rates as an estimate of future trends, the total pass-through under AB 1290 is estimated to amount to over 33 percent of the tax increment. This estimate differs from other estimates, since it expresses the passthrough share in present-value terms. Such estimates are very sensitive to the choice of the rate at which future payments are discounted to present value. In this case, the historical average inflation rate from 1970 to 1992 of 5.8 percent was used as the discount rate. Since most analyses use a real (i.e., inflation-adjusted) discount rate greater than zero, this is a very conservative estimate, which tends to overestimate the pass-through share of tax increment when expressed in present-value terms.15 ____________ 14The pass-through percentages are calculated after contributions to the Low and Moderate Income Housing Fund have been subtracted from gross tax increment revenue. As discussed in Appendix A, RDAs are required to place at least 20 percent of their tax increment revenues into a fund for the provision of affordable housing. 15A model created by the California Redevelopment Association has calculated the average pass-through under AB 1290 to be 37 percent. 36 School districts get the majority of property taxes statewide, but they have had little incentive to negotiate cash pass-throughs with the local RDA, since the state made up for any revenue losses. As noted above, some of the difference might be the fact that schools are said to be more likely to receive noncash than cash transfers. Counties, on the other hand, had no such recourse. Counties rely on property taxes for 15 percent of their total revenue (43 percent of their own-source revenue)16 and, as was seen in Figure 3.3, have been the most aggressive local governments in seeking pass-through agreements. AB 1290 required that the pass-throughs be shared among all affected tax entities according to the same formula that applies to the base property taxes. Figure 3.4 Pass-throughs (billions of 1994$) 6.0 Other 5.0 Schools Counties 4.0 Cities 3.0 2.0 1.0 0 FY94 AB 1290 SOURCES: California State Controller’s Office (annual) and author’s calculations. Figure 3.4—Local Government Pass-Throughs ____________ 16Local government finance data were taken from the California Budget Project (1996). 37 shows the dollar values of this effect on each of the types of jurisdictions affected. Although the overall fraction of tax increment to be passed through to other tax entities will increase, counties will be even worse off than they were before the passage of AB 1290. Their share of passthroughs now averages 62 percent, but their share of pass-throughs from new projects will be the same 21 percent that they now receive from all other property taxes. Under the current rules, counties receive 8.7 percent of all tax increment revenues (62 percent times 14 percent); under AB 1290 rules, they will receive 6.9 percent (21 percent times 33 percent) of all increment over the life of the projects. Given counties’ financial straits and the fact that they are increasingly called upon to provide social services, it is no surprise that tax increment financing causes conflict. Different levels of government perform different services, so such subsidies can change the distribution of public expenditures. For example, counties spend relatively more on public assistance, courts, and jails, whereas cities spend more on public utilities and transportation. With the financial straits that cities themselves are in, it is also not surprising that redevelopment—with its promise of increased property taxes without a required vote—is a popular tool. Chapter 4 describes the methodology employed in this study to estimate how much of increased property values can be ascribed to the actions of the redevelopment agency, as opposed to general real estate market trends and other factors. 38 4. Study Design and Methodology This study addresses the issue of how blighted redevelopment project areas are. It also tackles the central point of contention between levels of government over redevelopment: Whose money is being used for redevelopment? Unless the RDAs cause enough growth in assessed values to cover the tax increment revenues they receive, such revenues are, in effect, subsidies from other governments to each project. RDAs defend their claim on tax increment revenue by asserting that most of the increases in property values in project areas are due to their redevelopment activities and not to generally improving economic conditions. As discussed in Chapter 3, AB 1290 significantly increased the effective pass-through rate from the current average of 14 percent to an average of roughly one-third for new projects. There was no empirical basis for this particular allocation, since there were no estimates of the RDAs’ effect on property values. The uniform pass-through rate set by the legislature was a political compromise, primarily between the RDAs 39 and the state, which faced the largest financial liability because of its obligation to reimburse school districts for any lost tax revenues. Without any independent estimates to resolve this debate, no better allocation of tax increment was possible. Study Design What we really want to know is how fast each project area would have grown in the absence of the redevelopment project. Since that is not possible, the best way to address this question is with a control group. Project areas are supposed to be chosen because they are blighted, which means that they are unable to grow without deliberate government involvement. If this is true, comparing project area growth to that of the entire city may be misleading: The comparison will likely underestimate the amount of growth that redevelopment caused. To properly credit project areas for their initial condition, this study uses a matched-pairs design. (This type of study is also referred to as a case-control study in the biomedical literature.) Each project area examined is compared with another area in the same city that is most similar to the project area in terms of blight conditions. The measure of property value used here is the assessed value of each parcel. Economic studies generally focus on market value, but in the context of estimating property tax revenues, the assessed value is more relevant. Changes in assessed value, not market value, determine the actual amount of tax increment revenues a project area generates. In post-Proposition 13 California, in contrast to most other states, assessed value increases at only a 2 percent annual “inflation” rate unless a property is sold or major construction is done on the site. Assessed value and market value are therefore equal only when properties are sold or 40 constructed or when a large decline in property values causes a property to be reassessed. Changes in assessed value from 1983 to 1996 for every parcel in both the project area and its match were compared to estimate how much of the growth in the project area was caused by the action of the RDA. 1983 was chosen as a base year, since it was the first year that all the projects in the sample had been formed. 1996 was chosen as the final year, since it was the most recent year for which assessed value was available when the study began. There was some concern about the effect of the recession on project areas’ growth relative to their matches’ growth. Since assessed value changes tend to lag market value changes by about two years, 1993 was a good comparison point to use for the recent recession. Assessed value declines were, in fact, more common in project areas than in matches from 1993 to 1996, but not by nearly enough to affect any of the study’s conclusions. The median change in assessed value from 1993 to 1996 was nearly identical (an increase of 4.4 percent) between the projects and matches, as well as across counties. Over this time period, the total assessed value in the state increased by over 150 percent, so that the increases in the sample project areas were expected to be large enough to be detected (even in blighted areas). This increase can be compared with the 29 percent increase that would have occurred if no properties had been sold or undergone major construction—assessed values would then have increased solely by the 2 percent annual inflation factor.1 For each project, the growth rate of its matched area was subtracted from the project’s own growth rate to estimate the portion of the growth in assessed value that was due to the RDA’s actions. The amount of tax ____________ 1The 29 percent figure is the result of compounding the 2 percent annual increases for 13 years. 41 increment revenues generated by this growth rate was compared with the amount of tax revenues the project area received, after accounting for the pass-through rate of the project and the 2 percent annual inflation factor. The share of property tax growth due to redevelopment was estimated for each project, showing the range of subsidies or extra revenues that occurred in these projects. Summed across all projects, the difference between the amount of tax increment generated and the amount received shows how large the subsidies to RDAs from other jurisdictions have been. The results also provided an estimate of what the pass-through rate for new projects should be to have RDAs be self-financing in the long run. If the legislature wants redevelopment to be self-financing,2 the optimal pass-through rate should be the rate at which the tax increment that was self-generated (averaged across all projects) was equal to the amount of tax increment that was received. Sample Project Selection It is impractical to measure the effect of every redevelopment project on assessed values. Recent projects have not had enough time to affect property values; and historical assessed value data are not widely available more than ten years back. Cost considerations also limit the size of study that can be undertaken. To fairly assess the effect of redevelopment on assessed values, this study examined a sample of projects that began long enough ago to have had an effect but recently enough to obtain historical assessed values from early in each project’s life. The original sample ____________ 2Recall from the discussion in Chapter 3 that the theoretical basis for tax increment financing assumed that it equalizes the costs and benefits of development across local governments over a project’s lifetime. 42 examined consists of all projects started between 1978 and 1982.3 This time period captures projects that began right after Proposition 13, and which have had enough time to develop property to have a detectable effect on aggregate assessed value in the project area. In addition, there was no significant change in redevelopment law over this period, so all of the projects operated under similar rules. There were 134 projects begun in this time period, representing 18 percent of the 669 projects that were active statewide in 1994. In that year, they accounted for approximately $666 million in expenditures (also 18 percent), out of the statewide total of $3.8 billion. Table 4.1 shows the number of projects and their average acreage, by the starting year of the project. The size of the project areas ranges from 3 acres to 9,000 acres, with a mean of over 1,000 acres. Given the one very large project, the median size of almost 500 acres gives a more accurate sense of the typical size of these projects. The comparable statewide median Table 4.1 Original Sample Projects, by Year Starting Year 1978 1979 1980 1981 1982 Projects 18 21 14 30 51 Average Acres 631 1,247 1,113 1,477 838 SOURCE: California State Controller’s Office (annual). ____________ 3Given the need to travel to each county’s assessor’s office to resolve any missing parcel data, projects in counties with two or fewer projects were omitted for cost reasons. This affected 20 projects in 13 counties. Other projects were deleted later for various technical reasons (see California Redevelopment Association, 1996, and Appendix C). 43 size for all redevelopment project areas was 441 acres in 1994. These projects were started during a period of rapid increase in the number of redevelopment agencies and projects. From 1976 to 1985, 154 new redevelopment agencies were established, doubling the existing number; and a total of 300 new project areas were formed (a 150 percent increase).4 Of these original sample projects, 37 were the first projects undertaken by their redevelopment agency. Southern California counties are slightly more represented in the sample (73 percent) than they are among all redevelopment projects (63 percent), reflecting the region’s faster growth in the 1980s as well as the more aggressive use of redevelopment in some cities in the region. These data are reassuring in that the projects in the original sample were similar to the greater population of redevelopment projects. The results obtained from examining this group cannot be used to estimate the effect of projects from other time periods, however, since existing redevelopment law as well as overall economic conditions have changed over time and cannot be adequately controlled for. But they can explain what happened to one-fifth of the redevelopment projects in the state, and those projects were quite similar to the rest of the state’s projects in terms of size and expenditures. Unfortunately, not every project in the original sample could be used in the study—the biggest obstacle was the availability of historical assessed value data. After eliminating counties that had fewer than three projects begun between 1978 and 1982, 114 projects located in ten counties remained in the group, but historical microfiche data could be obtained for only four counties with 67 projects begun in that time ____________ 4California Redevelopment Association (1996). 44 period. The inability to find matches because of several remaining data problems left a final sample of 38 projects and their matches.5 The three counties in the final sample are Los Angeles, San Bernardino, and San Mateo; these counties are a good cross-section of the state and the projects in them are representative of the original sample of 114 projects. Table 4.2 shows the relationship between project areas and the rest of the city that they lie within, for the original sample and the final sample. In the original sample, the vacancy rate in the project areas was 64 percent higher than in the rest of the city, whereas in the final sample it was 28 percent higher; the median household income was about 12 percent lower in both the original and final samples; and median rents in the original sample were 91 percent of that in the rest of the city, while in the final sample they were 96 percent. Table 4.2 Comparison of Original and Final Samples in 1980 Projects Rest of city Projects Rest of city Vacancy Median Rate, % Income, $ Original Sample 7.8 25,337 4.7 29,198 Final Sample 5.5 26,044 4.3 29,291 Median Rent, $ 422 462 437 454 SOURCE: 1980 Census Block Group data, STF-3A file. Matched-Pair Selection The goal is to choose an area in each city that in effect “could have been” a redevelopment project, at least by the stated definitions of ____________ 5More detail on why the projects were dropped is provided in Appendix C. 45 blight.6 Since cities have different (and unmeasured) tax and regulatory structures that might affect property values, each project area is compared with another area within the same city that is selected for its similarity to the project area in 1980. By selecting matches within each city, any omitted factors that vary across cities and affect property values are controlled for. Other studies that have examined the effects of local development programs, most notably enterprise zones, demonstrate the importance of making comparisons between areas within the same city. To determine the characteristics on which to match the pairs, we looked at the blight definitions that were used to justify each of the 114 projects in the original sample and grouped them into seven general categories. Table 4.3 shows the general categories of blight definitions and the number of projects that cited them in their plans. The three most commonly cited categories—underutilized property, inadequate infrastructure, and deteriorating buildings—all involved property characteristics. Table 4.3 Use of Blight Categories to Justify Sample Projects Blight Category Underutilized/vacant property Inadequate infrastructure Deteriorating buildings Unemployment, crime, poverty Parking/traffic problems Other economic problems Aesthetic reasons No. of Projects Citing 55 65 58 35 20 52 16 SOURCE: Individual project planning documents. ____________ 6Before making this choice, any other redevelopment project areas in each of these 80 cities were removed from consideration. 46 We then chose those variables in the 1980 Census (at the Block Group level) that best approximated some of the blight conditions cited by the redevelopment agencies. Census Block Groups are generally areas comprising between six and eight city blocks; in less populated areas, Block Groups can encompass much larger areas of land. Regardless, Block Group areas are defined so that population counts are roughly comparable across groups. The Census variables initially selected as the best proxies available for the blight conditions were percentage of units vacant, percentage of structures built more than 30 years earlier (i.e., before 1949), median rent, and percentage of individuals in poverty. The characteristics for which there are no good proxies in the Census data were cited less frequently, except for inadequate infrastructure. Given the fact that “inadequate infrastructure” was highly correlated with the other blight conditions, the lack of a Census proxy was not considered a serious problem.7 Although the vacancy and poverty rates are obvious proxies for blight, some reviewers of the matching methodology viewed the age of structures and median rent as less reliable in capturing blight conditions than vacancy and poverty rates. There is no simple relationship between the age of a structure and its value—while some periods of rapid growth produced low-quality buildings that did not retain value well, some very old building stock is considered quite desirable. Use of median rent could underweight the experience of the low-income portion of the population. The final matching formula, therefore, used the two-variable version, with vacancy and poverty rates.8 In several cities that had more ____________ 7AB 1290 removed inadequate infrastructure as sufficient evidence of blight. 8To check the sensitivity of our matching to the use of the two- or four-variable formula, the matching program was run for both definitions. In all but six cases, the 47 than one project in the final sample, the same match was selected for more than one project. In these cases, one of the projects’ second choices was selected as the match. Once again, the overall match function value in each case changed by only a very small amount. Table 4.4 compares the vacancy rate, median household income, and median rent for the final sample, the matches, and the rest of the cities. By averaging across all the pairs, these figures do not completely capture how closely each project and matched area compared. Nevertheless, they do show that using matched pairs probably improves on a comparison of project areas to the rest of the city. Table 4.4 Comparison of Final Projects, Matched Areas, and the Rest of the City in 1980 Final sample projects Matched areas Rest of city Vacancy Rate, % 5.5 5.0 4.3 Poverty Rate, % 14.5 15.5 14.9 Median Income, $ 26,044 28,470 29,291 SOURCE: 1980 Census Block Group data, STF-3A file. Median Rent, $ 437 440 454 Assessed Value Determination Once matching census block groups were selected, every parcel in each Block Group was identified and downloaded from the Dataquick property database.9 This database contains the parcel number, 1996 assessed value, tax rate area codes, and other characteristics of each parcel ____________________________________________________ Block Group chosen as the closest match was the same in both cases; in those six cases the absolute value of the match function from the two-variable formula was very close to that from the four-variable formula. 9The process of finding the corresponding 1983 assessed value for each parcel in the 1996 database is described in more detail in Appendix C. 48 from databases maintained by county assessors’ and tax collectors’ offices. Parcels in redevelopment projects were identified by their tax rate area codes. The list of parcel numbers in 1996 was used to obtain the assessed value in 1983 from each county’s historical microfiche. In those cases where a parcel’s physical boundaries were changed between 1983 and 1996, the assessor’s office map books were used to find the corresponding parcel number(s) in 1983. If fewer than 90 percent of the parcels (or parcels with less than 90 percent of the total 1996 assessed value) in a project or match area could be reconciled, the pair was dropped from the analysis. Parcel Description There were over 47,000 parcels in the final sample containing the 38 project pairs. Because project areas tended to be larger than the typical Census Block Group, almost two-thirds of these parcels were in project areas. The characteristics of the project parcels and matched-area parcels are summarized in Table 4.5. Compared with those in matched areas, the parcels in project areas had a larger mean assessed value (AV) in 1983 ($88,460 compared with $66,230), were less developed in 1983 (their mean percentage improvement then was 56 percent, compared with 66 percent in the matches), and had a building stock that was newer in 1996 (mean year built was 1963 compared with 1960).10 These differences are all statistically significant. ____________ 10As noted in Table 4.5, percentage improved and year built were weighted by 1983 assessed value. The year built as of 1983 was not available for all the parcels that changed, so the averages are not reported here. For the majority of parcels that did not change, the difference was not as great. See Table C.1 for the distribution of parcel changes between 1983 and 1996 in the project areas and matches. 49 Table 4.5 Summary Characteristics of Parcels in Final Sample No. of parcels Mean assessed value in 1983 Mean % improved in 1983 Mean year built Projects 30,160 $88,460 56% 1963 Matches 16,931 $66,230 66% 1960 t-Value of Difference t = –5.51 t = 59.80 t = –33.96 SOURCE: Author’s calculations from Dataquick database. NOTE: Percentage improved and year built were weighted by 1983 assessed value. There were also noteworthy differences in the composition of parcels by use codes (Table 4.6).11 Consistent with the experiences of many redevelopment consultants and RDA project managers, project areas had fewer residential parcels and more commercial and industrial parcels than the matched areas. Almost 88 percent of the matched parcels were residential, compared to 74 percent of the project parcels. In place of residential properties, the project areas had twice as many commercial Table 4.6 Distribution of Use Codes in Project and Matched Areas Use Code Residential Commercial Industrial Government/utility/nonprofit Vacant Project Parcels, % 73 12 6 2 7 SOURCE: Dataquick property database. Match Parcels, % 88 5 1 1 6 ____________ 11Use codes are assigned by each county’s assessor’s office to define what type of use a parcel is primarily for. For the purposes of this report, parcels are classified as either residential, commercial, industrial, government/utility/nonprofit, vacant, or other. Fewer than 1 percent of the parcels in either project or matched areas were “other,” so they are not reported in Table 4.5. In each county, the actual use codes are more detailed than those in these reported categories. 50 parcels as the matches did. They also had many more industrial parcels than the matches, although the industrial parcels were concentrated in a few project areas. As might be expected in areas with targeted redevelopment projects under way, parcels in project areas were more likely to physically change over the 13-year period than were those in the matched areas. One-third of the project parcels were combined or split between 1983 and 1996, almost three times the rate for the matches. It is often said that an important service provided by RDAs is to combine small or oddly shaped parcels into larger lots to make them more marketable. In fact, only 186 parcels out of the 9,000 project parcels that physically changed between 1983 and 1996 (just over 2 percent) were combined from several parcels into one. RDAs were much more likely to divide one or several parcels into many parcels, usually for condominiums or new single-family subdevelopments. Only 8 percent of the parcels in the matched areas were created from larger parcels, but over 26 percent of the project parcels in 1996 originally came from larger parcels. Across all 38 pairs, the mean growth of assessed value in all the project parcels was 204 percent; parcels in all of the matched areas saw their assessed value grow by 145 percent. This difference, too, was statistically significant at the 99 percent level. (If no parcels had been sold or had major construction take place, the 2 percent annual inflation factor under Proposition 13 would have caused assessed values to grow by 29 percent over these 13 years.) Growth at the parcel level might be misleading, however, since without matching by project area we cannot control for local factors that can affect assessed values. To make full use of the power of the matched-pair methodology, we must compare each project with its match and see what the relative growth rates were for 51 each project. This not only restricts the comparison to the matched area for each project but also avoids having the problem that the largest project areas will dominate the results if we pool all the projects and matches together. The growth in assessed value in the matched area of each pair is subtracted from the total assessed value growth in the project area to estimate the RDA’s effect on the growth in that project area.12 The size and direction of any subsidies that might have occurred were estimated by comparing the total amount of assessed value that the RDAs created to the amount of tax increment revenue they actually received. Benchmarks for Analysis In many economic studies of the effect of specific government programs, the yardstick used is whether the program had any effect at all. RDAs themselves have argued in reply to critics of tax increment financing that, were it not for agency activities, there would be no increase in assessed value to be taxed. In measuring the size of the subsidies that redevelopment agencies receive under the tax increment financing system, it is not sufficient to show that RDAs had some effect on growth in the project areas. Since tax increment financing has been justified as a self-financing mechanism, the benchmark is more stringent: Project areas must generate enough growth in assessed value to cover the tax increment revenues the RDAs receive (less any pass-through payments they make). Given that, how high a standard does this set? Overall, the projects in this sample retained almost 84 percent of the tax increment revenues that were generated within their boundaries; the ____________ 12The actual formulas used in these calculations are reported in Appendix C. 52 other taxing jurisdictions received just over 16 percent. This means that, in the aggregate, they need to have grown more than six times as fast as their matches to be justified in the claim made by RDAs that the tax increment they receive was due to their actions. Anything less than that implies that the projects as a whole are subsidized; anything more means that they are not only self-financing but that they actually produce additional revenues for other governments. Not every project needed to grow six times faster than its matched area, since some projects had much higher pass-through rates than the average for all projects. In Chapter 5, the results are presented both for the individual projects (using their specific pass-through rates) and for the sample overall. This will demonstrate the range of success and failure of projects in terms of being self-financing, and will permit an estimate of the size of any subsidies that flow from other taxing jurisdictions to RDAs. Potential Sources of Bias The use of matched areas as a control group avoids some of the common problems in estimating the effect of a program on some economic outcome. By restricting the matched area to the same city as the project, we can also control for political, regulatory, and other factors that could distort the results. However, even with this kind of control group approach there are sources of bias in the results that cannot be controlled for. There are several sources of potential bias in the study, and they act in opposite directions. If the RDAs actually chose the worst area of the city as a project area, then by definition no possible match can be quite as bad off as the project area. The effect of this selection bias should be to cause us to 53 underestimate the effect of RDAs on assessed value growth. This is only true if the RDAs did in fact choose the worst area of the city (with “worst” defined in terms of the potential of the area to increase in property value if unassisted). As discussed in Chapter 3, there are strong financial incentives to declare as blighted those areas that can be quickly (re)developed; therefore, at least some of the project areas are probably not the worst area of the city. Working in the opposite direction is the strategic behavior that some agencies engage in when determining the boundaries of project areas. According to a number of consultants, lawyers, and analysts familiar with redevelopment practices, many RDAs draw at least some of their project area boundaries to include properties they know will soon be sold or developed. This enables the agency to capture tax increment revenues early in the project’s life. Where this occurs, there will be a bias toward overestimating the effect of RDAs on growth in assessed value, since the agency gets credit for growth that it did not influence. Also working to overestimate the effect of RDAs on assessed value growth is the way that Proposition 13 changed the practice of reassessing properties. Since Proposition 13 passed, assessed values increase only 2 percent per year unless a property is sold. If parcels in project areas are more likely to turn over than parcels in matched areas, their assessed values will tend to be closer to market value than those in the matched areas. Since parcels in project areas were in fact more likely to turn over between 1983 and 1996, this phenomenon has the effect of exaggerating the RDAs’ effect on assessed values. There is no practical way to measure the presence, let alone the relative size, of these effects. In interpreting the results though, the fact that there are forces working in opposite directions provides some protection against serious errors in estimation. 54 For any one of these biases to alter the conclusions, they would need to be large enough—and to arise systematically across most projects—to overcome the effect of the opposing source of bias. 55 5. Results of the Study This study sought to answer two questions raised by the critics of redevelopment: How blighted are project areas, and how much are RDAs subsidized by tax increment financing? How Blighted Are Project Areas? To consider the effect of financial incentives on blight determination more closely, the study looked at a sample consisting of the 134 redevelopment projects that were started in the first five years after Proposition 13 passed (between 1978 and 1982). Project plans, maps, and financial statements as well as Census data were analyzed to get a sense of the condition of these projects. In their own reports that demonstrated the existing blight conditions, the sample projects collectively cited physical conditions (underutilized/vacant property, inadequate infrastructure, and deteriorating buildings) more frequently than economic and social conditions (such as high unemployment, 57 crime, and poverty) in their plans.1 In 25 of the 134 projects, more than 50 percent of the project area was vacant land in 1986–1987—the first year these kinds of data were reported in the annual Controller’s Report. Across all agencies, 19 cities still had more than half of the land area in all of their projects vacant as of 1986–1987. Some RDAs have declared that most of their city is blighted. Figure 5.1 shows the number of cities by the amount of their land area covered by redevelopment projects.2 Ten cities were more than half Number of cities 90 80 70 60 50 40 30 20 10 0 0–10 11–30 31–50 51–100 Percentage of city land area SOURCES: U.S. Census Bureau (1994); California State Controller’s Office (annual). Figure 5.1.—Redevelopment Areas as a Percentage of City Land Area ____________ 1Blight conditions in the sample projects are discussed in more detail in Chapter 4. 2The data on city land area come from the U.S. Bureau of the Census (1994) and do not include information on cities with fewer than 25,000 residents. Thus, some cities well-known for being almost all redevelopment project by land area (such as Emeryville and Industry) are not included in this figure. 58 redevelopment project by land area. These top ten are Cathedral City, Chico, Fontana, Huntington Park, Manteca, Pittsburg, Ridgecrest, San Marcos, San Pablo, and West Sacramento. Another 25 cities were between one-third and one-half redevelopment project by land area. At this point, it is no longer possible to compare the sample project areas’ actual physical condition at the time of their formation with the condition of the rest of the city. However, the sample project areas can be compared with the rest of the cities that they lie within, using the available Census data.3 Table 5.1 shows the vacancy rate, median household income, and median rent in 1980. As should be expected if project areas were chosen because they are doing less well than the cities as a whole, vacancy rates were higher—and incomes and rents were lower—in the project areas than in the rest of the city. The average differences are not large, although the differences within each city vary widely. These comparisons mask a large amount of variation among the projects, and in some cases the project areas in one city had better 1980 statistics than other cities’ nonproject areas. This brings up another important conceptual issue: whether blight is meant to be an absolute Table 5.1 Area Characteristics in 1980 (current dollars) Vacancy rate Median income Median rent Sample Projects 8% $25,337 $422 Rest of City 5% $29,198 $462 SOURCE: 1980 STF 3-A Census files. ____________ 3In this section, the rest of the city excludes any other redevelopment projects that lie within the city regardless of their start dates. 59 concept or if it should be judged solely in relative terms within a given city. As currently practiced and under the legal definitions cited above, blight is defined strictly on a relative basis. This can be seen in Table 5.2, which shows the 1980 characteristics of three project areas that were started between 1978 and 1982. Note that these figures are for the project area itself, not the entire city in which the project is located. As long as every city is empowered to establish a redevelopment agency, and each agency can establish project areas wherever it can demonstrate blight, there will be project areas in almost every city regardless of how their degree of “blightedness” might differ from city to city. Table 5.2 Project Area Characteristics in 1980 Characteristic Year agency formed No. of acres in project % vacant land 1980 household income Median home value Agency and Project Area Los Angeles: Foster City: Adams-Normandie Foster City 1948 1981 400 855 35 58 $14,915 $40,164 $93,363 $234,637 SOURCES: annual Controller’s Reports; STF-3A Census files. Victorville: Bear Valley 1981 1,248 95 $17,907 $112,514 Are RDAs Subsidized by Other Governments? The other, and more critical, question this study sought to answer was whether RDAs are actually self-financing or whether they receive subsidies from the state and other local governments. The benchmark for evaluating this question was presented in the previous chapter: In the aggregate, assessed value in the project areas needed to grow, on average, six times faster than in their matches for the projects to be self-financing. 60 Since each project area has a different pass-through rate, the actual ratio needed to claim self-sufficiency is different for every project. The growth in assessed value for each project and its match is shown in Figure 5.2.4 Virtually all of the project areas and matches had rapid assessed value growth from 1983 to 1996. All but six of the project areas and four of the matched areas at least doubled their assessed value over the 13-year period. This performance is of a similar magnitude to general assessed value changes: Total assessed value in the three counties in the final sample grew by more than 150 percent over the same period. Percentage growth 1,800 1,600 1,400 Project Match 1,200 1,000 800 600 400 200 0 Project/match SOURCE: Author’s calculations from Dataquick database. Figure 5.2.—Growth in Assessed Value, 1983 to 1996 ____________ 4The projects in Figures 5.2 through 5.4 are displayed in the same order in all three charts—in order of increasing net subsidies. 61 The fastest growth was in project areas—six projects grew by over 400 percent, but no matched area did; 15 projects more than tripled their assessed value, when only two matches did. The mean assessed value growth in project areas was 270 percent, compared with an average 144 percent growth in the matched areas. Two-thirds of the projects—27 out of the 38 pairs—actually had faster growth in assessed value than their matches did. This means that, in most cases, redevelopment did have a positive effect on assessed value growth in the project areas. However, the true effect of redevelopment on assessed value is only that growth over and above the growth in the matched area. To evaluate how much of each project’s retained tax increment revenues was generated by the RDA, we must take into account not only the relative growth between the project area and its match, but also the size of the pass-through payments that the project made to other jurisdictions (and the 2 percent annual inflation factor that the projects do not receive). A project may have grown much faster than its matched area, but because of a very low pass-through rate it still might not have grown fast enough to cover all the revenues it received. In contrast, a project may have outgrown its matched area by only a small amount but if it passed through most of its tax increment revenues, it might have been self-financing. The formulas used to calculate the RDAs’ share of the tax increment revenue growth are described in Appendix C. Figure 5.3 shows what share of tax increment revenue each project was responsible for. If a project has a 100 percent share, it means that the project generated exactly enough growth in assessed value so that none of the tax increment revenues it received were diverted from other governments. 62 Percentage earned 200 100 0 –100 –200 –300 –400 SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.3—Share of Tax Increments “Earned” in FY 1994–1995, by Project (Throughout this chapter, “tax increment received” refers to the retained tax increment. This is the tax increment revenue left after subtracting any payments to other jurisdictions in the form of pass-throughs.) Only four projects generated more tax increment revenues than they actually received in FY 1994–1995, which was the latest year for which financial data were available. Two projects generated over 90 percent of their tax increment revenues, and two other projects generated between 80 percent and 90 percent of their revenues. Therefore, fewer than a quarter of all the projects were actually self-financing—or at least close to it—as far as their tax increment revenues were concerned. Another quarter of the projects had such combinations of slow relative growth and low pass-through rates that they generated none of the tax increment revenues that they received (those shown with negative shares). The 63 other 19 projects generated between 2 percent and 75 percent of the revenues they received. The eight projects that generated at least 80 percent of the tax increment revenues they received had an average of 51 percent vacant land in their areas in 1986–1987; in contrast, the other 30 project areas had an average of only 14 percent vacant land. It seems that the project areas that outgrew their matches by the greatest amount were engaged more in development than in redevelopment. What all this means for the size of subsidies that pass from counties and the state (via payments to the school districts) depends on the size of the projects. If all the best performers are large projects, their effect could more than offset the subsidies received by smaller projects and the net subsidy could be quite low. But if the reverse were true, the net subsidy could be quite large. Figure 5.4 shows the net subsidy that each 1 0 –1 –2 –3 –4 SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.4—Net Subsidies to RDAs in FY 1994–1995, by Project 64 Transfers (millions of $) project received in FY 1994–1995. Positive numbers indicate that a project generated even more tax revenue growth in the area than it received in (retained) tax increment revenue; negative numbers show the size of the subsidy a project received. As seen in Figure 5.4, only four projects generated enough growth in assessed value to produce a surplus of tax increment revenues. In twothirds of the projects, the financial transfers either way amount to less than $1 million annually. In the remaining 13 projects, the annual subsidies received from other jurisdictions exceeded $1 million for each project area, and in three of those project areas, the other taxing jurisdictions lost over $3 million in property taxes to the RDA. Summed across all 38 projects, the RDAs collectively accounted for 51 percent of the tax increment they received in FY 1994–1995. Figure 5.5 shows the amount of tax increment generated and the amount received in the sample projects. These projects received a total of over $78 million in tax increment revenues (after pass-throughs), and generated just under $40 million of those revenues by their effect on the growth in assessed value. This implies total annual subsidies flowing from all the other taxing jurisdictions to these redevelopment projects of over $38 million.5 This subsidy estimate would be reduced by whatever amount, if any, that these projects provided in the form of noncash contributions to the other taxing jurisdictions. These findings can also give a sense of how large the annual subsidy is for the entire sample of project areas that were formed between 1978 ____________ 5The data and estimation approach used in this study do not permit a calculation of the total subsidies received over the lifetime of these projects. Such a calculation would require the annual relative changes in assessed value between the project and match, as well as any pass-through payments made; it would also require tracking any physical changes in parcels (i.e., those combined or split) on an annual basis. 65 FY 1995 tax increment (millions of $) 80 70 60 50 40 30 20 10 0 Received Generated SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.5—RDAs Generated Half of the Tax Increment They Received and 1982.6 The final three counties used were not selected according to any characteristics of the counties—they are the only counties for which 1983 assessed value data were available. Within those counties, the projects in the final sample were those for which historical assessed values could be found for at least 90 percent of the 1996 parcels and assessed value. The only systematic difference between the projects included in the final sample and all the projects started between 1978 and 1982 was that in the omitted projects there were either no possible matches or map ____________ 6These results cannot be extrapolated to projects that were begun either before 1978 or after 1982. The legal environment, real estate markets, fiscal pressures on cities, and a host of other factors were all significantly different from those facing this group of projects. 66 data were missing. The results for the final 38 projects can be extrapolated to the 114 projects in the ten counties in which more than three projects started between 1978 and 1982. If the other 76 projects had a similar distribution of outcomes as the final sample, the total subsidies to this group—constituting one-sixth of all redevelopment projects in California—are very large. These 114 projects received $347 million in tax increment revenues in FY 1994–1995, and if they generated the same share of the assessed value growth as the final sample projects did, the subsidies to this cohort of projects was $170 million in the last fiscal year alone. Appendix D shows regression results using pair-level observations. Pair-level regressions more closely approximate the use of matched pairs, but parcel-level analysis permits the use of information such as the assessed value in 1983 and degree of improvement on the property. Both sets of estimates show RDA effects of similar magnitude. 67 6. Conclusions and Policy Implications By virtue of the nature and size of redevelopment agencies’ financing, as well as their extensive powers and minimal state oversight, redevelopment is very controversial. There are disagreements over other issues, such as housing, but the biggest battle is over the $1.5 billion per year that RDAs receive in property tax revenues that would otherwise go to the state, counties, and special districts. How much of this $1.5 billion is a subsidy from other governments depends on how much of the assessed value growth can be attributed to redevelopment projects, rather than to general real estate trends. Policy debates about redevelopment have thus far proceeded without any estimate of the size of the subsidies, so there has been little, if any, debate as to how large these subsidies should be. How comfortable policymakers are with the level of subsidies partly depends on the extent of blight in project areas. This study used matched pairs of project areas and similar nonproject areas in the same city to estimate RDAs’ effect on assessed values and examined 69 characteristics of project areas and their cities to see how much worse the project areas were than the rest of the city. Conclusions The Extent of Blight and the Size of Subsidies Assessing the extent of blight is hampered by the lack of any quantitative definition of blight. Of the 186 cities in California with populations above 25,000, 35 cities had more than 30 percent of their entire land area under redevelopment. Another 82 cities had between 10 percent and 30 percent of their land area in redevelopment projects. The Legislative Analyst’s Office (LAO) has reported that RDAs placed about 100 square miles of land in new project areas in 1993 alone—more than double the average amount placed in the preceding three years.1 The report assumed that this increase was a response to new restrictions contained in the reform legislation enacted in 1993 (AB 1290). However, it gives a sense of the large areas being declared blighted on an ongoing basis. It seems implausible that these areas are all suffering from such a combination of blighting conditions “that it constitutes a serious physical and economic burden on the community.” Analysis of 1980 Census data for the projects that were formed between 1978 and 1982 shows that the project areas were somewhat worse off than the rest of their cities, but there is wide variation among those cities in income and other measures. An area that is worse off than the rest of an affluent city is not necessarily a “serious physical and economic burden” to that city. Assessing the size of the subsidies for redevelopment from other taxing agencies is more straightforward. From the matched-pair analysis, ____________ 1See California Legislative Analyst’s Office (1994). 70 we found that two-thirds of the sample project areas grew faster than their matches did, implying that redevelopment did have some positive effect on assessed values. However, given how much of the tax increment revenues the RDAs actually kept, they would have had to grow much faster than their matches to have generated all the property tax revenues they received. The results show that only four of the 38 projects examined generated enough growth to cover all the tax increment revenues they received. Four other projects generated at least 80 percent of the revenues they received. The remaining 30 projects received subsidies ranging from several thousand dollars to almost $4 million in fiscal year 1994–1995. Across all the projects, the RDAs generated just 51 percent of their tax increment revenues in fiscal year 1994–1995. This means that other governments—the state, counties, and special districts—provided subsidies to these RDAs of over $38 million that year (after accounting for pass-through payments). If the 76 other projects in the original sample had the same effect on growth in assessed value as these projects did—and it is reasonable to assume that they did—the subsidies from other governments just for redevelopment projects begun between 1978 and 1982 would be over $170 million annually. Some RDAs make noncash contributions to school districts and other governments, and the value of these contributions should also be subtracted from the estimated subsidy. No estimates of the size of these contributions exist, however, and none of the redevelopment experts interviewed believed that these contributions were even half the size of the reported cash pass-throughs. Given existing information, this report’s estimate of the subsidies to redevelopment in California is the most accurate estimate possible. 71 The eight least-subsidized projects—those that were responsible for at least 80 percent of their tax increment revenues—had an average of more than 51 percent vacant land when they began. This extent of vacant land would most likely have disqualified them as blighted under AB 1290, and it supports the critics’ contention that many of the fastergrowing projects were engaging more in development than in redevelopment. The remaining 30 projects had an average of only 14 percent vacant land. In fact, the biggest predictor of growth in assessed value in an area was how much of the value of a property was from improvements, rather than from land. In other words, the less developed a property was, the more its value grew. In the eight fastest-growing projects, 38 percent of their 1983 assessed values derived from improvements on the land. In contrast, in the other projects, 62 percent of their assessed values came from improvements, which is comparable to the average of 64 percent for the matched areas. AB 1290: The Most Recent Reform Attempt Concern about the size of the subsidies and the incentives provided by the loose definition of blight are evidenced by the passage of AB 1290. That it was seen as a significant revision to existing redevelopment law is suggested by the many RDAs that rushed to adopt projects before the law came into effect in January 1994. Nevertheless, the blight definitions are still very subjective and leave many areas of the state eligible for incorporation into redevelopment projects, if the definitions are taken at face value. By setting the new, uniform pass-through rate well above the average rate for current projects, the legislature (at least implicitly) decided that the existing subsidies, of unknown size, were too large. Although the 72 new pass-through rate does increase the percentage of tax increment revenues that must be shared with other governments, it will probably result in continuing large subsidies for RDAs. If the pass-through rates mandated by AB 1290 had been in effect for the projects examined in this report, those projects collectively would have generated 61 percent (instead of 51 percent) of the tax increment revenues they received. Ironically, AB 1290’s allocation formula will make the counties even worse off financially than they were under the old system. Although overall pass-through payments will increase, the counties’ share of those payments will shrink. If all existing projects had been under AB 1290 rules in 1994–1995, counties would have received $28 million less in property tax revenues. These changes will happen at the same time that counties come under increased fiscal pressure from welfare reform, as well as from demands for health care and other social services. Beyond AB 1290: Addressing the Larger Policy Questions The fact that the state, counties, and special districts subsidize RDAs by as much as hundreds of millions of dollars each year is not necessarily undesirable. Many RDAs play a vital role in improving blighted areas and, collectively, the RDAs provide a large number of affordable housing units throughout the state. Many of the policy problems concerning redevelopment stem from the lack of a clearly stated and unambiguous goal for redevelopment. In past hearings and discussions on this topic, the most frequently invoked goals were alleviating urban blight, providing affordable housing, downtown development, and economic development. In the absence of a legislatively specified goal, it is difficult to assess the redevelopment program and to provide substantive policy 73 recommendations. The definitions of blight in current law reflect this ambiguity of purpose. These problems are exacerbated by the weak and decentralized oversight provided by the state. There is no agency that systematically reviews RDAs’ plans and activities for compliance with the law; enforcement requires court challenges by local residents or other taxing agencies, generally within 60 days or less of plan adoption. These are the key policy questions that remain to be answered, even after AB 1290: What is the state trying to accomplish with redevelopment? Should the state, counties, and special districts subsidize RDAs’ pursuit of these goals? If so, how large should the subsidies be? Where should this money be spent? Who should decide on the money’s allocation and monitor the RDAs’ performance? This report’s recommendations follow from the fact that these questions still need to be asked more than 50 years after the California Community Development Act of 1945 was first enacted. 1. The legislature should formally clarify its goal(s) for redevelopment. The various goals espoused for redevelopment not only compete for available resources, they often imply conflicting choices of project areas and expenditures. If the historical purpose of alleviating blight were the goal, the project area would be the most destitute neighborhood in the city and the project would involve affordable housing, provision of social services, anti-crime efforts, etc. If downtown development were the goal, the core commercial area would be chosen and project activities would be more oriented toward retail and commercial development and provision of parking lots and related infrastructure. If economic development were the goal, RDAs would have to focus much less on property markets and any particular area, and more on labor markets and training. In this case, most of their existing tools are poorly suited to the task. 74 There may be instances where these goals overlap or are complementary, and an overly narrow focus may unduly constrain the activities of some projects. That said, without a more specific statement of intent by the legislature, the major tensions surrounding redevelopment will continue. 2. Blight conditions need to be aligned with the goal(s) of redevelopment and should be made more precise. With the financial incentive of tax increment revenues, blight conditions cannot remain in the eye of the beholder if redevelopment efforts are meant to target the most serious cases of blight. This means that they must be more like the quantitative criteria used to determine eligibility for enterprise zones in some states. Criteria such as a poverty rate of at least 20 percent of the population, 20 percent population loss in recent years, x percent of the buildings or assessed value abandoned, y percent of property taxes in arrearage, or the crime rate z times the state average. If the redevelopment subsidies are to be targeted, blight must be judged more on an absolute than on a relative basis. It is not enough that a project area be worse than the rest of the city. In a poor city, the average area may be sufficiently blighted to warrant assistance; in an affluent city, its worst area might still not qualify. One way to address this might be to qualify cities as to whether they are eligible to receive tax increment revenues or to provide lower pass-through rates for more disadvantaged cities. However it is done, redevelopment subsidies need to be better targeted. Asking the legislature to compile a list of cities that qualify for tax increment revenues is a very tall task. Even if there were the political desire and discipline to do so, there would likely be a gradual expansion of the eligible list, as happened with the Model Cities program of the 1960s and as has happened with enterprise zones.2 Even assigning different ranges of pass-through rates to cities ____________ 2Since the first ten zones were designated in 1986, the number of enterprise zones has grown to 25, and nine Employment and Economic Incentive Areas have been added. 75 according to degree of need might be too difficult politically. Unless these kinds of steps can be taken, however, redevelopment will remain a melange of blight remediation and commercial development aimed at increasing property tax revenues. 3. There should be some form of oversight authority to monitor RDA behavior and to avoid the need for frequent legislative interventions. Both the LAO and the State Auditor, citing the problems stemming from the lack of meaningful oversight, have suggested designating one state agency as the primary authority over RDAs.3 Creation of an oversight authority with suitable enforcement powers would allow a more systematic evaluation of whether a proposed project complied with existing law. Such an authority could request clarifications of ambiguities in legislation, as well as improved data on project activities and financial statements. However, an oversight authority would face its own problems. None of the state agencies proposed for the oversight role are eager to assume this task, and opponents of government expansion would resist the creation of a new bureaucracy if a separate entity were proposed instead. Any oversight authority would also be subject to strong political pressures from the same groups that have been involved in legislative reform debates. Although an authority could ensure compliance with existing law, it would be unable to target redevelopment activities to the most blighted areas without a more objective and restrictive definition of blight. Absent any legislative guidance on the matter, it would also be unable to set the size of financial assistance for redevelopment. 4. If the legislature intends redevelopment to be self-financing, the passthrough rate should be increased significantly. Raising the pass-through rate is difficult to do retroactively, since debt has been issued with ____________ 3See California Legislative Analyst’s Office (1994) and California State Auditor (1996). The LAO recommended that the Attorney General review every project plan for compliance with the law; the State Auditor did not recommend a particular agency. Other agencies that have been suggested for the oversight authority role are the Departments of Finance, and Housing and Community Development. 76 future tax increment revenues pledged for debt payments. With over 700 projects already in place, the level of their subsidies would not change quickly, no matter what pass-through rate was mandated for new projects. Over time, however, a higher rate would decrease the subsidies paid to RDAs. The legislature may well support the idea of spending large subsidies on redevelopment, and there are good reasons to do so. A targeted level of subsidies would put redevelopment subsidies into competition with other appropriations, where its merits (noneconomic as well as economic) could be weighed against other public uses for those funds. Policymaking would be greatly improved if an explicit level, or at least a range, of subsidies were ratified by the legislature. As long as redevelopment relies on tax increment financing for its funding rather than being directly subsidized, there will be conflicts between state goals and local interests. At either the pre- or post-AB 1290 pass-through rates, there will also be large annual subsidies passed from the state, counties, and special districts to RDAs. Until the legislature expresses explicit goals and subsidy levels for redevelopment, no amount of tinkering with the detailed rules is likely to end the conflict over redevelopment in California. 77 Appendix A Policy Issues Concerning Affordable Housing in Redevelopment Projects The financial incentives created by the prospect of tax increment revenues sometimes lead agencies to be more focused on activities that will generate increases in assessed valuation than on providing affordable housing. Providing affordable housing has been viewed by many as a core activity since redevelopment began, and the state legislature has repeatedly strengthened the affordable housing requirements in redevelopment law. Do RDAs Produce Enough Low-Income Housing? History of Housing Under Redevelopment Housing activities, especially those targeted toward low-income residents, were the original motivation for the federal urban renewal 79 programs of the 1940s. The Federal Housing Act of 1949 provided funds for cities to engage in urban renewal projects that originally stressed slum clearance. Federal aid was available as long as the project area was residential, regardless of what replaced the residential buildings. The two groups that had supported the urban renewal programs—real estate developers and housing advocates—had competing visions for the use of redevelopment funds, visions that continue to clash in the present. In 1954, the Federal Housing Act was amended to allow 10 percent of federal grants to be used in nonresidential areas; in 1959 it was 20 percent; in 1961 it was raised to 30 percent; and in 1965 it became 35 percent.1 The combination of residents’ increasing resistance to largescale evictions and growing interest in downtown revitalization led to the decline of slum clearance as a redevelopment goal by the late 1960s. The tension between developers and housing advocates led to the creation in 1976 of the Low and Moderate Income Housing Fund (AB 3674, Montoya). This required all new projects to set aside 20 percent of their tax increment revenues to increase the supply of affordable housing and required that any demolished low-income housing in project areas be replaced one for one. In 1986, the set-aside and replacement requirements were extended to all project areas, regardless of start date. Among the major changes in AB 1290 (the 1993 reform legislation) was the creation of a “use it or lose it” penalty for low- and moderate-income housing funds. Agencies with an excess surplus— defined as more than $1 million—in unused 20 percent housing setaside funds within a certain time period must turn the funds over to their respective county housing authority. Agencies are just now approaching ____________ 1Discussion of federal urban renewal programs draws heavily from Frieden and Sagalyn (1989). 80 these deadlines, and there may still be delays before the eventual disbursement of these funds. Because of the rise in tax increment revenue and the decline in federal housing funds, redevelopment agencies’ low- and moderateincome housing funds are one of the largest pools of public money available for local housing projects. Activities can include building new low- and moderate-income2 (i.e., affordable) housing, rehabilitating and purchasing existing housing, and subsidizing rent for low-income households. In 1993–1994, the state’s RDAs built 4,521 housing units and had another 10,000 in progress. They also rehabilitated an additional 4,539 units, although it is not clear how extensively the units were rehabilitated.3 A number of observers have commented that much of the reported rehabilitation is minor and mainly cosmetic, resulting in no real increase in affordable housing in the area. AB 1290 addressed this issue by formally defining substantial modification. Over the last eight years, the annual number of units built and rehabilitated has ranged from 5,000 to 9,000. Approximately 50,000 housing units have been constructed or rehabilitated since 1987–1988. Although this pales in comparison to the 1.2 million building permits issued for housing ____________ 2Households are categorized into four categories for the purposes of this law. Very Low Income refers to households with less than 50 percent of the countywide median household income; Low Income refers to households with greater than 50 percent and less than 80 percent of median income; Moderate Income covers households with income between 80 percent and 120 percent of median income; and Other Households have more than 120 percent of area median income. Countywide median income by household size is determined by the U.S. Department of Housing and Urban Development (HUD). 3Most agencies do not actually build housing themselves but instead provide financing to other organizations to produce housing. 81 statewide4 over that time, it is purported to be a large share of all the affordable housing created in the state. The key complaints by housing critics are that the RDAs do not spend the required 20 percent minimum of tax increment on low- and moderate-income housing, too much of what they do spend is on administrative costs, and not enough of the housing is targeted to the lowest-income groups. These questions are addressed in the following subsections. Do RDAs Spend the Required 20 Percent Minimum of Increment on Low- and Moderate-Income Housing? Statewide tax increments added to low- and moderate-income funds totaled $255 million in 1993–1994, which was 16 percent of total tax increments and almost 19 percent of retained increments (i.e., after passthroughs). Although in the aggregate RDAs set aside only 18.5 percent of their tax increment in housing funds over the past five years,5 the 30 largest agencies set aside just under the 20 percent requirement. Smaller agencies are less likely to meet their required amounts, and some agencies have done little to increase the supply of low-income housing. The aggregate shortfall in the actual amounts set aside in the last five years is only 1.5 percent of the tax increment received statewide; yet this amounts to over $107 million in funds. Agencies can set aside less than 20 percent of their increment if there is no need for affordable housing in the community or if less than 20 percent is sufficient for affordable housing activities. Agencies that were originally grandfathered from the ____________ 4Building permit data are from the Construction Industry Research Board, reported in California Department of Finance (1994), Table I-8. 5California Legislature (1996). 82 20 percent requirement were allowed to defer payments if they had already committed the funds to existing programs, and other agencies are granted exceptions to the rules for various reasons.6 Do RDAs Spend Too Much of Their Housing Funds on Administrative Costs? Figure A.1 summarizes the operating expenditures of RDAs’ housing activities in 1993–1994.7 Planning and administration expenses account for a similar share of housing activities as they do for the rest of RDA operations, so unless agencies are misclassifying spending categories, there is no support for the assertion that RDAs as a group are frittering away housing funds on administration. This is not to say that all agencies are not spending too much on administration—42 agencies spent 100 percent of their money on planning and administration costs and 36 spent over 50 percent.8 In some cases, for example, in new project areas or in projects starting new housing programs, such high levels of planning costs might be reasonable. For example, of the 42 agencies that spent 100 percent of their housing funds on planning and administration in 1993–1994, 12 were still spending more than half of their funds in those categories two years later. In 19 of the 36 agencies that were spending more than half of their funds on administrative costs in 1993–1994, administrative costs were still near or above 50 percent two years later. Although the number of potential abusive cases is much ____________ 6As of July 1, 1996, any grandfathered projects were required to start repaying any deferred amounts (unless they had an exemption from the requirements altogether). This should raise the average share of increment set aside for housing funds in future years. 7Housing data are from the California Housing and Community Development Agency (1996), unless otherwise noted. 8California Legislature (1996). 83 Rental subsidies 10% Other 6% Site improvements 13% Rehabilitation 13% Planning, administration 20% New construction 18% Housing purchase 5% Land purchase 15% SOURCE: California State Controller’s Office (annual). Figure A.1—Operating Expenses for RDA Housing Activities smaller when examining costs over a three-year period, almost 10 percent of the state’s agencies still seem to be consistently spending their housing funds on overhead rather than on the production of low-income housing. Is Enough RDA Housing Targeted to Low-Income Households? RDAs also face inclusionary housing requirements, which stipulate the minimum shares of affordable housing units built or rehabilitated. Of agency-developed housing, at least 30 percent of all new and rehabilitated units must be affordable to low- or moderate-income households, and half of that must be affordable to very low-income 84 households. If housing is not agency-developed, the inclusionary requirement drops to 15 percent. No precise definition of “agencydeveloped” exists, and it is common practice for agencies to claim that new housing is not agency-developed, to meet the smaller requirement.9 Table A.1 shows the number of units produced by the income category of the unit. Of all newly built and rehabilitated units produced by RDAs in 1993–1994, 48 percent were affordable to very low-income households (46 percent of all built units and 50 percent of all rehabilitated units).10 From these figures, it seems that RDAs are easily exceeding the requirements for the production of very low- and lowincome housing units. Another common cause of resistance to redevelopment activity comes over the subject of displaced households. Such concern has its Table A.1 Redevelopment Housing Activities, by Income Category No. of Units: Built Rehabilitated Acquired Subsidized Total Very Low 2,081 2,280 315 1,769 6,445 Low 1,250 1,756 159 807 3,972 Moderate 1,168 488 61 773 2,490 Other 22 15 4 45 86 Total 4,521 4,539 539 3,394 12,993 % Very Low 46 50 58 52 50 SOURCE: California Department of Housing and Community Development (1994). ____________ 9CRA notes that “agency-developed” housing is defined as units that an RDA contracts to build directly, whereas units for which an RDA provides financial assistance are “agency-assisted” housing. The Department of Housing and Community Development has sought clarifying legislation on this issue. 10AB 1290 requires that RDAs that meet these inclusionary requirements with units located outside the project area must do so on a two-for-one basis. 85 roots in the large-scale relocations that occurred during the peak years of the federal slum clearance programs. In 1993–1994, 579 households were displaced by redevelopment actions, of which 130 were outside project areas. Also, 422 net housing units were lost from redevelopment projects, and 697 other units were created to replace units destroyed in earlier years. There was an imbalance in income categories, however— 73 percent of units lost were very low-income whereas 53 percent of units replaced were affordable to low-income and above. Since the units replaced went to offset units lost in earlier years, though, the discrepancy might not represent a real imbalance. Given the scale of both prior relocations and of the overall number of units built and in production, these do not seem like large amounts of displacement overall. 86 Appendix B Sales Tax and Redevelopment After Proposition 13, sales tax became a more important source of revenue for local governments in California. Some redevelopment projects are also oriented toward sales-tax generation, and in 1995–1996, 43 projects received sales-tax revenue. Four have become inactive since 1996, and an additional seven projects that used to receive sales tax became inactive before 1995. Most of these projects received a tax rate of 1 percent and were started between 1986 and 1994. Authority to receive sales taxes was granted to redevelopment agencies soon after Proposition 13 and was revoked by AB 1290 commencing in 1994 (perhaps by coincidence, 17 of the 43 projects were started on July 1, 1994). How important is the sales tax generated within redevelopment projects to the city? Figure B.1 shows the fraction of total city sales taxes that the redevelopment project represents. In most of the cities, the redevelopment projects’ sales taxes are not a large fraction of total sales 87 Ratio 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 AlhamAbnraahBeaiAmldzwusinaBPeaarukmoCnatrsonChinCooltConoronaDEalvMisonFtiellmFoHroeunntatinnagGtoanlKtPinagrksLbaurMgMiraoMdriaelpniotaVsalNleeywPaarlmkPadsaaledenPaPiceorrRisSivaeTnreadmCpiltey CTWiotyrerasnt cCeovina SOURCES: Board of Equalization, 1995 data on redevelopment agencies; California State Controller’s Office (annual). Figure B.1—Redevelopment Projects and City Sales Taxes taxes. In six cities, though, redevelopment agencies bring in almost half of the cities’ sales taxes and, in one case, the agency brought in twothirds. For these cities, redevelopment agencies generate a significant amount of sales tax and demonstrate what can happen when fiscal stress and redevelopment’s wide latitude are combined. A related question is how important tax increment is relative to sales taxes. Some observers think that sales tax is an important element in RDA planning, even for projects not concentrating solely on retail activity. This question is very difficult to answer with available data. In cities with specific sales-tax agreements, the tax increment generated within a project area can be compared with the sales-tax revenue raised in the same area. Data were available for 25 of the 43 project areas that 88 have special sales- and use-tax arrangements.1 The amount of tax increment was less than the sales tax generated within the project in nine cases and was more than triple the sales tax in 11 projects. Since these were projects with specific sales-tax agreements, they can be assumed to be much more sales-tax-oriented than the average project. Given that, it seems safe to say that tax increment is a more important determinant of RDA incentives than sales-tax revenues are. ____________ 1In seven of the other cases, the sales tax went to specific businesses instead of to the projects. 89 Appendix C Detailed Discussion of Study Methodology Matching Algorithm The matching method used in this study is similar to the logic behind regression, i.e., minimizing the sum of squared differences (weighted by the variance of the potential matches). This number is related to the Mahalanobis distance, which has been frequently used in regional economic studies of this type; the difference is that the Mahalanobis distance is weighted by the variance-covariance matrix rather than just the variance. For a more involved discussion of this technique, see Rephann and Isserman (1994). If there are three Census variables to be matched on (say, X, Y, and Z), then the matching Block Groups are those in each city that minimize the expression 91 ∑i n =1   (X RDA − X nonRDA σ 2X ,nonRDA )2 + (YRDA − YnonRDA )2 σ 2Y ,nonRDA + (ZRDA − ZnonRDA )2 σ2Z,nonRDA   , where n = all non-RDA Block Groups in a given city. Conceptually, this approach selects the Block Group that minimizes the distance in XYZ space between each project area and its control. There is a tradeoff between using the same matching variables for every city and choosing the variables that match blight conditions for each city individually. The methodology used here follows generic matching. Using the same variables for matching across all the projects allows for better generalization of the results to those projects in the original sample that were lost because historical data were unavailable. Loss of Project Areas from the Beginning Sample After dropping counties with fewer than three projects that began between 1978 and 1982, the sample had 114 projects located in ten counties. Assessed value data for 1983 could be obtained for only four counties, which together had 67 projects that began between 1978 and 1982. In 25 cases, the projects had to be dropped because of circumstances that caused problems in using the matched-pairs methodology. At this point, some unresolved data issues precluded the use of the four Solano County projects. These data problems—along with the losses outlined below—left us with the final sample of 38 projects in the counties of Los Angeles, San Bernardino, and San Mateo. Five projects were dropped because we were unable to get maps of the project areas in the respective cities; this made it impossible to eliminate other project areas when searching for possible matches. Two projects were dropped because they consisted of a single parcel each. 92 Three projects were lost because the project area covered at least part of every Census Block Group in the city. Ten additional projects were lost because every Census Block Group was covered by either the sample project area or other project areas in the same city. Two projects were dropped because the closest Census Block Group selected by the matching algorithm was too dissimilar to the project area (i.e., the function value was much larger than the range in all the other pairs examined). One project was dropped because of large inconsistencies between assessor’s office and controller’s office data. Last, two projects had to be dropped at the very end because even after consulting historical parcel map books, we were unable to match enough parcels to achieve the required thresholds of 90 percent of total parcels and total 1996 assessed value. Imputing Assessed Value for Changed Parcels As discussed in Chapter 4, the parcels used in the analysis were first selected from the 1996 secured roll for the entire state, as collected by Dataquick from county assessors’ offices and tax collectors (and other private sources). Project parcels were identified by the tax rate area codes that uniquely identify all the parcels within a project area.1 A file that linked current tax rate area codes and the jurisdictions receiving taxes from each tax rate area was purchased from the State Board of Equalization. Parcels in the matched areas were identified by their ____________ 1A redevelopment project area can contain multiple tax rate areas; however, a tax rate area is not split between a project area and the outlying area. In other words, a project area will contain whole tax rate areas, each with a unique listing of parcel numbers that constitute the base on which the tax increment is calculated. When a new tax rate area is being created where a project area is currently located, two tax rate areas in effect are created (one inside the project area, one outside) so as to not split the tax rate area. Tax rate areas can be prolific. Los Angeles County has roughly 10,000 tax rate areas. 93 Census Block Group. This set of parcel numbers formed the basic file for the rest of the study. Once the parcel numbers were identified, the 1983 microfiche was searched for those parcel numbers. If present, the 1983 assessed value and use code was appended to the 1996 record. Almost 16 percent of the parcel numbers could not be found in the 1983 microfiche. In the project areas, 34 percent of the parcels could not be found because the parcel had changed; for the matches, the figure was 12 percent. To resolve these missing cases, trips were made to each county’s assessor’s office to track the parcel changes. Parcel map books were consulted to find which 1983 parcel(s) each 1996 parcel was derived from. In 38 of the 42 project-match pairs, we were able to identify 1983 assessed values for more than 90 percent of the parcels and 90 percent of the assessed value in the 1996 database for either the project area or the matched area.2 The pairs for which we could not get such a complete accounting were dropped from the analysis (but the final results are not affected much whether those pairs are included or not). For the rest of the pairs, the corresponding 1983 parcel was categorized by the type of change that occurred between 1983 and 1996. If the change was simply from one parcel number to another, the 1983 assessed value was added to the 1996 record. If one 1996 parcel came from several 1983 parcels, the 1983 values were summed to create the correct 1983 value. If several 1996 parcels came from a single 1983 parcel, the 1983 value was evenly apportioned to all the succeeding 1996 parcels. In those cases where several 1983 parcels became several 1996 parcels, the total values were summed in each year and apportioned ____________ 2The thresholds for deciding which projects had sufficient data to be used were determined before examining the parcel-level breakdowns. 94 equally among all the 1996 parcels. Table C.1 shows the distribution of the sample parcels by type of changes, if any, between 1983 and 1996. There was quite a bit of variation between counties (and between projects) in the extent of parcel changes in the last 13 years. In each county, a greater share of parcels changed in project areas than in matches—although this was not universally true among the projectmatch pairs. What was universally true was how rare it was for numerous parcels to be assembled into one larger parcel: Only in the projects in Los Angeles County did this occur in over one-half of one percent of the parcels. It was much more likely that one large parcel was subdivided into many smaller parcels, or that several large parcels were combined and then split into many parcels. Overall, we were not able to identify 3 percent of the parcels in the 1983 microfiche, and another 1 percent were tax-exempt parcels for which there was no assessed value recorded in the 1983 secured-roll microfiche. Table C.1 Breakdown of Parcel Changes, 1983–1996, by County (in percent) Missing, Missing, No Many: 1: Many: Nongov- Govern- Change 1:1 1 Many Many ernment ment Los Angeles Project 70 3 1 9 10 5 2 Match 88 20 42 2 2 San Bernardino Project 54 5 0 10 27 3 1 Match 88 10 54 1 0 San Mateo Project 68 2 0 18 9 2 1 Match 95 10 40 0 0 Total Project Match 66 89 3 1 11 13 4 10 53 2 2 1 SOURCES: Dataquick property database (1996); individual county assessor’s office and/or tax collector data (1983). 95 Calculation of RDAs’ Share of Assessed Value Growth and Size of Tax Transfers RDAs receive property taxes based on the increase in assessed value for parcels within the project area, less the 2 percent annual inflation factor as decreed by Proposition 13. The tax increment revenue they receive in FY 1995–1996 is calculated as the change in aggregate assessed value (minus the compounded inflation factor), times the property tax rate, multiplied by the share of tax increment the RDA retains. Tax increment received = (1 − pass-through) * (tax rate) * (AV96 − (1 − .0213) * AV83). AVnn is the total assessed value in all the parcels in the project area in year nn. The pass-through rates for each project come from the annual financial self-reports made by each RDA to the state controller’s office. Because these figures are subject to error and annual fluctuations, the pass-through rate used in these calculations is the average reported by each project area over the last 11 years.3 To calculate how much tax increment the project generated itself, the growth rate in the matched area rather than the 2 percent inflation factor is subtracted from the change in aggregate assessed value. The growth rate in the matched area is calculated as the growth in assessed value between 1996 and 1983, divided by the assessed value in 1983. The inflation factor is not subtracted in this case because in nonRDA areas that amount is also shared among all the taxing jurisdictions. ____________ 3The state controller’s office annual reports on redevelopment agencies began in FY 1985, and the latest year published was FY 1995, so only 11 years of data were available. This is also why there is a one-year difference between the assessed value data used in the calculations (FY 1996) and the dollar amounts used from the controller’s report (FY 1995). 96 Growth rate in matched area (Rm) = (AV96 − AV83) . AV 83 The growth generated by each project is defined as the total growth in assessed value in the project area minus the growth rate of the matched area. This growth is multiplied by the tax rate to get the amount of tax increment revenue generated in the project area. Tax increment generated = (tax rate) * (AV96 − (1 + Rm ) * AV83); RDA share of tax increment = tax increment generated ; tax increment received Estimated tax transfer = RDA share of tax increment * net FY 1996 tax increment . The share of the tax increment revenues each project generated is simply the ratio of the tax increment the project generated to the amount it received. These figures were calculated using the parcels and their values as described above. The comparison was therefore made between consistent groups of parcels in 1983 and 1996. To estimate the actual amount of transfers that occurred in FY 1995, though, the calculated ratio was multiplied not by our estimate of the tax increment received, but rather by the amount of tax increment reported in the FY 1994– 1995 controller’s report. On average, our calculation of the 1996 total assessed value in the project areas was 77 percent of the value reported in the controller’s report. Approximately 8–10 percent of this difference results because the RDAs report both the secured and unsecured rolls, whereas this study used only the secured rolls. Another 5 percent of the difference results because we were able to obtain the historical assessed value only for 97 parcels totaling 95 percent of the total 1996 assessed value in the project areas. The remaining difference remains unexplained. Conversations with officials in the controller’s office indicate that the annual figures in RDAs’ self-reports are not without errors, and that a discrepancy in the totals on the order of 7–10 percent is not cause for concern. For example, the reported increment assessed value (what we would call AV95–AV85) was 18 percent larger than the increase in total assessed value between the two years. Some of this discrepancy may be due to mergers that increased the size of project areas, but some may also be due to the inaccuracies cited by the controller’s office. 98 Appendix D Regression Analysis With a relatively small number of projects in the final sample, regression analysis can demonstrate that the differences in growth rates are nevertheless statistically significant. It also allows the two matching variables to be controlled for. The matched-pairs analysis can be replicated in a regression format by using each area—project or match— as an observation, with 38 dummy variables to control for the pairs and a separate dummy variable for whether the observation is a project or a matched area. The regression can also be run with poverty and vacancy rates as regressors, to check whether the observable differences between the projects and their matches change the estimated size of the RDA effect. These regressions are not meant as a complete analysis of these phenomena; they are provided to demonstrate that the differences in growth rates between projects and matches are statistically significant, 99 despite the small sample size. Further econometric work in a forthcoming working paper will examine the factors—including parcel turnover and type of expenditures—that affect project areas’ growth in more detail. The equation to be estimated is: ln(AVgrowth) = βRDADRDA + βvac VAC + βpovPOV + β1D1 + ... + β38D38 + e , where ln(AVgrowth) is the logarithm of the percentage change in assessed value from 1983 to 1996, βRDA is the estimated effect of the RDA on assessed value growth in the project area ( DRDA takes the value of 1 if the area is a project and 0 if the area is a match), VAC is the vacancy rate according to the 1980 Census, and POV is the poverty rate. Each area dummy takes the value of 1 for that project area and its match, and 0 for all other project/match pairs. This retains the information contained in the pairing and allows the coefficient on DRDA to estimate the redevelopment growth effect across all pairs. Table D.1 shows the results from this regression. It also reports regression results when vacancy rate and poverty rate are included, with and without the controls for matched pairs. When controlling for matching without vacancy and poverty rates, project areas are estimated to grow 42 percent faster than their matched areas; when poverty and vacancy rates are controlled for as well, the RDA effect increases to 45 percent faster.1 The estimated RDA effect is only slighter lower when matching is ignored. The estimated RDA effect from these equations is similar to that obtained (41 percent) from the simple comparison of mean growth rates ____________ 1With the dependent variable in logarithmic form, the elasticity is interpreted as e β − 1. In this case, the coefficient is βRDA , called the “RDA effect” in Table D.1. 100 Table D.1 Estimation of RDAs’ Effect on AV Growth [Dependent Variable: log(AV growth)] Equation 1 Equation 2 Equation 3 Equation 4 RDA effect Poverty rate Vacancy rate Pair dummies Adjusted R2 No. of observations 0.335 (2.56) — — — 0.069 76 0.305 (2.36) –0.116 (–0.158) 1.437 (2.02) — 0.096 76 0.335 (2.57) — — Included 0.453 76 0.369 (2.65) 2.205 (0.72) –2.854 (–0.879) Included 0.435 76 NOTE: Equations with pair dummies were estimated without a constant term. T-statistics are reported in parentheses. between project and match parcels. With or without poverty and vacancy rate controls, adding controls for the matched pairs increases the predictive power of the estimation significantly (note adjusted R2 values). Parcel-level regression results (not reported here) confirm the presence and approximate size of the project effect; they also demonstrate that the largest predictor of the difference in assessed value growth between the project and matched areas is the percentage improvement in the project area’s parcels in 1983. This is consistent with the observation made in Chapter 5 that the projects that generated the largest share of their tax increment had much more vacant land in their areas than the other projects did. Regressions that were run separately for projects and matches also indicate that the initial percentage improvement is a stronger predictor of growth in the project areas than it is in the matched areas. 101 Appendix E Final Sample Projects County Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Table E.1 Final Sample Projects Redevelopment Agency Azusa Baldwin Park Baldwin Park Downey El Monte El Monte Huntington Park Inglewood La Verne Los Angeles Los Angeles Los Angeles Los Angeles Maywood Maywood Montebello Pasadena Redevelopment Project West End Merged Project Areas Central Business District Project Area West Ramona Boulevard Project Area Downey Project Area El Monte Plaza Project Area Plaza El Monte Project Area North Project Area Century Project Area Project Area 1 Adams Normandie Project Area Chinatown Project Area North Hollywood Project Area Rodeo-La Cienega Project Area #2 Westside Project Area Economic Revitalization Project Area Lake Washington Project Area 103 County Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Mateo San Mateo San Mateo San Mateo San Mateo San Mateo Redevelopment Agency Pomona Pomona Pomona Pomona Pomona Pomona Torrance Whittier Fontana Ontario Ontario Ontario Rialto San Bernardino Victorville Belmont Foster City Menlo Park Redwood City San Mateo San Mateo Redevelopment Project Arrow-Towne Project Area Holt Ave/Indian Hill Project Area Mission/Corona Business Center Project Area Reservoir Street Industrial Project Area Southwest Pomona Project Area West Holt Project Area Downtown Project Area Whittier Boulevard Project Area North Fontana Project Area Project Area No. 1 Project Area No. 2 Cimarron Project Area Project Area A Northwest Project Area Bear Valley Road Project Area Los Castanos Project Area Foster City Project Area Las Pulgas Community Development Project Area No. 2 Project Area Downtown District Project Area Shoreline District Project Area 104 Bibliography Anderson, John E., “Tax Increment Financing: Municipal Adoption and Growth,” National Tax Journal, Vol. XLIII, No. 2, 1990, pp. 155– 163. Beatty, David F., et al., Redevelopment in California, 2nd Edition, Solano Press Books, Point Arena, California, 1995. Boarnet, M. G., and W. T. Bogart, “Enterprise Zones and Employment: Evidence from New Jersey,” Journal of Urban Economics, Vol. 40, 1996, pp. 198–215. California Budget Project, State and Local Government Finance in California: A Primer, Sacramento, California, 1996. California Debt Advisory Commission, Recommended Practices for California Redevelopment Agencies, Report CDAC-5, Sacramento, California, 1995. California Department of Finance, California Statistical Abstract, 1994, Sacramento, California, 1994. California Department of Finance, California Statistical Abstract, 1995, Sacramento, California, 1995. California Housing and Community Development Agency, Redevelopment Housing Activities in California, 1991–92 to 1995–96, Sacramento, California, annual. 105 California Legislative Analyst’s Office, Redevelopment After Reform: A Preliminary Look, Sacramento, California, 1994. California Legislature, Senate Committee on Housing and Land Use and Senate Select Committee on Redevelopment, Redevelopment and Blight, Report 831-S, Sacramento, California, 1995. California Legislature, Senate Committee on Housing and Land Use, End or Means? Redevelopment Agencies’ Housing Programs, Report 888-S, Sacramento, California, November 13, 1996. California Legislature, Senate Committee on Local Government, Redeveloping California: Finding the Agenda for the 1990’s, Report 457-S, Sacramento, California, 1989. California Redevelopment Association, 1996 Introduction to Redevelopment, Sacramento, California, 1996. California State Auditor, Statewide Redevelopment Agencies, Sacramento, California, 1996. California State Controller’s Office, Financial Transactions Concerning Community Redevelopment Agencies, Fiscal Years 1984–85 to 1993–94, Sacramento, California, annual. California Trade and Commerce Agency, State Auditor, The Effectiveness of the Employment and Economic Incentive and Enterprise Zone Programs Cannot Be Determined, Sacramento, California, November 1995. Chapman, Jeff, “Tax-Increment Finance as a Tool of Redevelopment,” in Helen Ladd (ed.), Local Government Tax and Land Use Policy, Edward Elgar Press, 1998 (forthcoming). Davis, Don, “Tax Increment Financing,” Public Budgeting and Finance, Spring 1989, pp. 63–73. Dowall, David E., et al., “Evaluation of California’s Enterprise Zone and Employment and Economic Incentive Programs,” Institute of Urban and Regional Development, University of California at Berkeley, Working Paper 614, 1994. Dye, Richard F., “A Comparative Analysis of Tax Increment Financing in Northeastern Illinois”; Executive Summary from Assessing the Impact of Tax Increment Financing in Northeastern Illinois: Empirical 106 Analysis and Case Studies, The Civic Federation, Chicago, Illinois, 1997. Frieden, Bernard J., and Lynne B. Sagalyn, Downtown, Inc.: How America Rebuilds Cities, MIT Press, Cambridge, Massachusetts, 1989. Huddleston, Jack R., “Intrametropolitan Financial Flows Under Tax Increment Financing,” Policy Sciences, Vol. 19, 1986, pp. 143–161. Huddleston, Jack R., “Tax Increment as a State Development Policy,” Growth and Change, Vol. 15, No. 2, 1984, pp. 11–17. Huddleston, Jack R., “Variations in Development Subsidies under Tax Increment Financing,” Land Economics, Vol. 57, No. 3, 1981, pp. 373–384. Isserman, A., and T. Rephann, “The Economic Effects of the Appalachian Regional Commission,” APA Journal, Summer 1995, pp. 345–364. Klemanski, John S., “Using Tax Increment Financing for Urban Redevelopment Projects,” Economic Development Quarterly, Vol. 4, No. 1, 1990, pp. 23–28. Papke, Leslie E., “Tax Policy and Urban Development: Evidence from the Indiana Enterprise Zone Program,” Journal of Public Economics, Vol. 54, 1994, pp. 37–49. Rephann, T., and A. Isserman, “New Highways as Economic Development Tools: An Evaluation Using Quasi-Experimental Matching Methods,” Regional Science and Urban Economics, Vol. 24, 1994, pp. 723–751. Stauber, R., and M. Wyatt, “Property Tax Relief and Tax Incremental Financing,” in Robert H. Haveman and Jack Huddleston (eds.), Dollars and Sense: Policy Choices and the Wisconsin Budget, Robert M. LaFollette School of Public Affairs, University of Wisconsin-Madison, 1990. Stinson, Thomas F., “Subsidizing Local Economic Development Through Tax Increment Financing,” Policy Studies Journal, Vol. 20, No. 2, 1991, pp. 241–248. U.S. Bureau of the Census, County and City Data Book 1994, Washington, D.C., 1994. 107 About the Author MICHAEL DARDIA Michael Dardia is a research fellow at the Public Policy Institute of California. His research interests include redevelopment, regional economics, and industrial restructuring. In previous work, he has examined defense conversion and the effects of defense cuts on the California economy. He is the author of several publications on defense spending and the aerospace industry. Dardia holds a B.S. in biological sciences from the State University of New York at Stony Brook, an M.S. in policy analysis and management from the university’s W. Averell Harriman College of Management and Policy Analysis, and a Ph.D. in public policy analysis from the RAND Graduate School of Policy Studies. 109" } ["___content":protected]=> string(102) "

R 298MDR

" ["_permalink":protected]=> string(82) "https://www.ppic.org/publication/subsidizing-redevelopment-in-california/r_298mdr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8089) ["ID"]=> int(8089) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:34:48" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3180) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 298MDR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_298mdr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_298MDR.pdf" ["wpmf_size"]=> string(6) "237697" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(181837) "Subsidizing Redevelopment in California Michael Dardia January 1998 Copyright © 1998 Public Policy Institute of California, San Francisco, CA. All rights reserved. PPIC permits short sections of text, not to exceed three paragraphs, to be quoted without written permission, provided that full attribution is given to the source and the above copyright notice is included. Foreword The development and redevelopment of cities has been embroiled in controversy since the beginning of the century, and redevelopment in California is no exception. Over time, most cities in the state have adopted its use, and as that use has grown, so has the controversy. Longstanding issues have included the tension between affordable housing and downtown development, the application of eminent domain for private rather than public purposes, and the degree of blight in project areas. In this report, Michael Dardia addresses the most recent issue to take center stage in the policy debate: Who pays the bill for redevelopment? The logic of the laws governing redevelopment agencies (RDAs) assumes that redevelopment pays for itself through tax increment financing—that is, RDAs receive any increase in property tax revenues (above a 2 percent inflation factor) in project areas because their investment in area improvements is responsible for increasing property values. But what if the increases are also—even primarily—the result of a general increase in real estate values that would have occurred regardless iii of an RDA’s activities? That question is vital to local governments because tax limitation measures such as Proposition 13 have shrunk the tax and fee base that they can tap, but the public’s demand for services remains unchanged. When an RDA receives property tax revenues through tax increment financing, those revenues are lost to other local jurisdictions—the county, schools, and special districts. If the RDAs are not largely responsible for the increase in property values, those jurisdictions are, in effect, subsidizing redevelopment, with no say in how the revenues are used. Subsidizing Redevelopment in California describes a government function that has aggressively broadened from the traditional role of redeveloping blighted, inner-city areas to include the generation of tax revenues for city governments. After correcting for local real estate trends, the author finds that redevelopment projects do not increase property values by enough to account for the tax increment revenues they receive. Overall, the agencies stimulated enough growth to cover just above half of those tax revenues. The rest resulted from local trends and would have gone to other jurisdictions in the absence of redevelopment. These are important findings for state legislators considering the full range of benefits from redevelopment and who should pay for those benefits in California. Continuing the public-finance focus of two earlier PPIC reports, this publication completes the inaugural series of studies first launched by PPIC in the fall of 1995. The research on assessed values in redevelopment project and control areas was data-intensive and time consuming. It was made possible only by the independence PPIC has to ask questions that require such undertakings. The result is a study that is central to a key policy issue in the state, is based on solid empirical iv evidence from local assessment and tax records, and provides findings that clarify the extent of the financial burdens that existing policy creates for local governments and the state. David W. Lyon President and CEO Public Policy Institute of California v Summary In fiscal year 1994–1995—the most recent year for which figures are available—redevelopment agencies (RDAs) received 8 percent of the property tax revenues collected in the state of California, amounting to $1.5 billion. These are revenues that, absent the RDAs, would have gone to other public agencies such as the state and counties. Given that Proposition 13 and subsequent initiatives have severely limited local governments’ ability to raise or levy taxes, this flow of tax revenues to RDAs merits more public attention and understanding than it has historically received. Although the state has 351 RDAs with over 700 projects under way, not many of the state’s citizens know very much about them and why they are the subject of policy debate. To raise general understanding of redevelopment, this report takes a close look at the purposes of redevelopment agencies, their incentives, and how they operate. To illuminate the policy debate, it focuses primarily on the issues surrounding the property tax revenues that RDAs have historically received and how recent reform legislation has affected vii RDAs’ activities and revenues. It examines how the attraction of these property taxes can affect RDAs’ judgment about which areas to label as blighted and how RDAs affect property values in redevelopment project areas. The report does not provide an overall cost-benefit analysis of redevelopment in California. Over its 50 years of existence, redevelopment has pursued a number of goals: alleviation of blight, provision of low- and moderate-income housing, downtown redevelopment, and overall economic development. Some of these goals are ill-defined, and several conflict, making it difficult to assess the benefits of redevelopment quantitatively. Redevelopment’s full costs are also difficult to measure, given the mix of public and private financing and the myriad spillovers and cross-subsidies that exist. The report focuses, instead, on the extent to which other jurisdictions may unintentionally subsidize redevelopment in California and provides the only estimate available of the size of these subsidies. What Do RDAs Do and Why Are They Controversial? Redevelopment agencies were authorized by the state legislature in 1945 to combat urban blight, and since then more than three-fourths of all the cities in California have established RDAs. The definition of blight is vague, allowing for a wide range of neighborhoods to be considered blighted. According to the law, a blighted area is supposed to be a “serious physical and economic burden on the community which cannot reasonably be expected to be reversed or alleviated by private enterprise or governmental action, or both, without redevelopment.” Each city decides which of its areas are blighted, and there is little viii effective state oversight to police these decisions. Virtually all of the redevelopment activity in the state is undertaken by city agencies, although 15 counties also have active RDAs. These RDAs engage in a wide variety of activities, from providing municipal infrastructure such as streets, lighting, and parking structures, to assisting in the provision of affordable housing and renovating downtown commercial areas. Although redevelopment has its roots in federal urban renewal programs established in the 1940s, more than half of the RDAs in California were created since the passage of Proposition 13 in 1978. Fiscal pressures on local governments made the property tax revenues received by RDAs especially attractive to city governments. The mechanism by which the RDAs receive property taxes, called tax increment financing, was created in 1952 to enable cities to match the share provided by federal programs. Under tax increment financing, an RDA receives all the property tax increases in a redevelopment project area for the life of a project—often 30 years or more. Had an area not been designated as a project area, those same property tax revenues would have been split among the counties, cities, school districts, and other special districts. The allotment to each jurisdiction varies by county, but across the state the average share is 11 percent for cities, 21 percent for counties, 51 percent for school districts, and 17 percent for all the special districts (not including the RDA shares). Given the fiscal pressures created by Proposition 13 and the large number of projects created after Proposition 13 passed, it was probably inevitable that the conflict over tax increment financing would heat up. Although school districts have had their property tax losses compensated for by the state government, counties and special districts are not so favored. The rapid rise in RDAs and their project areas meant further ix pressure on counties’ and special districts’ revenues. Tax increment’s share of statewide property taxes grew from 4.6 percent to just under 8 percent over the last ten years; in some faster-growing counties with aggressive RDAs, tax increment revenues now account for as much as 18 percent of all property taxes. To the other jurisdictions, the tax increment revenues that RDAs received were monies diverted from their coffers. In recent years, $300 million that would otherwise have gone to counties, and $800 million that would have gone to school districts, instead went to RDAs. As the RDAs see it, these tax increment revenues came from the growth in property values stimulated by their activities. Recognizing the strain that tax increment financing could cause other local jurisdictions, the state legislature allowed them to negotiate with RDAs to recover some of the tax increment. These payments, called pass-throughs, grew rapidly in the 1980s. Over the last ten years, pass-throughs increased from 4 percent of all tax increment revenues to almost 14 percent; counties collected about two-thirds of this amount. School districts had little incentive to fight for the pass-throughs, since the state reimbursed them for any lost property tax revenue. The cost to the state for this reimbursement grew rapidly to hundreds of millions of dollars per year. Along with perceived abuses in the kinds of areas that were declared blighted, these costs helped lead to reform legislation (AB 1290) in 1993. AB 1290 instituted a number of reforms, including setting a uniform pass-through rate of approximately 33 percent, more than double the current average rate. This was an implicit recognition that the subsidies flowing to RDAs via tax increment financing, of whatever size, were too large. However, no estimates of the size of these subsidies existed at that time, so the size of this adjustment relied more on political judgment than on empirical evidence. x Purpose and Approach of the Study To estimate the size of the subsidies that RDAs receive from tax increment revenues, the study looked at all the redevelopment projects started between 1978 and 1982, the first five years after Proposition 13 passed. Examining projects begun in those years allows more than enough time for the RDAs to have had a significant effect on property values in the area. Assessed values in the project areas were tracked from 1983, just after the last projects were begun, to 1996. The assessed value was obtained for every parcel in the area in 1996, by using assessor’s office and tax collector’s office data gathered from each county. Cost considerations restricted the original sample to those counties with at least three projects begun within the five-year period. Ten counties, with 114 projects, met this criterion. To judge the effect of RDA actions on the increase in property values, each project area was matched to a Census Block Group in the same city which served as a control. The RDAs’ effect was measured as the difference between the growth of assessed value in a project area and that in its matched area. Although 1996 assessed values were available electronically, 1983 values could be obtained only from the microfiche records maintained by each county’s assessor or tax collector. Microfiche records were available from only four of these counties, and the inability to match enough 1996 parcels with their 1983 counterparts led to the final sample of 38 projects from three counties: Los Angeles, San Bernardino, and San Mateo. These 38 projects were very similar to the 114 projects in the original sample of ten counties, however, so the results from the final sample can be reasonably extrapolated to the original 114 projects. xi The study also examined the question of the degree of blight in project areas. The combination of a vague legal definition of blight, little oversight, and the strong financial incentive of tax increment has led some critics to charge that many project areas are not suffering from urban blight. AB 1290 made the blight definition somewhat more restrictive, and prior legislation had banned the inclusion of primarily agricultural land in project areas. Nevertheless, with more than 700 projects already in place, the question of how blighted some of these were remains important. Results Since identification of blight is a necessary condition for establishing a redevelopment project, how blighted were the project areas? In the group of 134 projects, 25 consisted of more than 50 percent vacant land. Such projects appear to have development, rather than redevelopment, as their prime purpose; most would probably have been disallowed had they begun after AB 1290 went into effect. Although one cannot assess the physical conditions in project areas in 1980, the Census provides data on conditions related to the degree of blight in an area at that time. There were noteworthy differences between project areas and their cities. The average vacancy and poverty rates were higher, and median incomes and median rents were lower, in the project areas than in the rest of the city. It is difficult to know in retrospect whether the project areas were the worst in each city, but they were worse off than the average area in these cities. Given that most project areas were more blighted than the rest of their cities, how much of the property tax revenues that RDAs receive was due to their effect on local property values? Any difference between xii what they received and what they generated can be considered an involuntary subsidy from other local jurisdictions. From 1983 to 1996, assessed value in the 38 projects studied grew by an average of 270 percent, compared with 144 percent in the matched areas. Two-thirds of the project areas outgrew their matches in property value, but 11 project areas grew more slowly. Because the RDAs kept more than five-sixths of any increases in property taxes within project areas, they needed to grow much faster than the control group areas to be considered self-financing. After the analysis removed other influences on assessed value growth and gave the projects credit for any pass-through payments made, only four project areas grew fast enough to justify the RDAs’ claim that they alone generated the increase in property taxes. Two other projects generated at least 90 percent of the property taxes received, and two more generated at least 80 percent of their tax revenues. In other words, fewer than onequarter of the projects came close to being responsible for the property tax revenues they received. These projects were also the ones with the most vacant land: The eight fastest-growing projects began with an average of more than 50 percent vacant land, compared with only 14 percent in vacant land for the other 30 projects. This raises again the issue of whether these projects were engaged more in development than in redevelopment. Given the size of each project’s tax increment revenues, how large was the overall subsidy? In dollar value, the projects collectively generated an estimated 51 percent of the tax increment revenues they received in fiscal year 1994–1995. The total across all 38 projects was $78 million in tax increment revenues, of which almost $40 million can be explained by the actions of the RDAs. Thus, the subsidy from other jurisdictions to the RDAs totaled $38 million in that year alone. If the xiii same pattern held for the other 76 projects in the ten counties originally examined, the total subsidy for redevelopment for the 114 projects would be $170 million annually. Had these projects been subject to the higher pass-through rates required under AB 1290, these projects would have generated only 61 percent of the tax revenues they received. Recommendations To date, there have been no estimates of the size of the subsidies provided to redevelopment through the tax increment financing system. The legislature has implied, through changes mandated under AB 1290, that the existing pass-through rates (determined through project-specific negotiations) were too low, and therefore that the subsidies were too large. However, these decisions had to be made with insufficient information. This study has demonstrated that half of the tax increment revenues received by those projects begun between 1978 and 1982 total about $170 million annually. The legislature may well intend subsidies of this size or larger. Redevelopment does provide infrastructure, affordable housing, and commercial development. Given that the state, counties, and special districts provide subsidies of this size, however, it might be worth targeting these monies more narrowly on the most blighted areas. It might also be worth keeping closer supervision on the activities that are being subsidized. With this in mind, the study makes four recommendations: 1. The legislature should formally clarify its goals for redevelopment. 2. Blight conditions need to be aligned with the goals for redevelopment and made more precise. 3. There should be a formal oversight authority to monitor RDA behavior. xiv 4. If the legislature intends redevelopment to be self-financing rather than highly subsidized, the pass-through rate should be increased significantly. xv Contents Foreword..................................... iii Summary..................................... vii Figures ...................................... xix Tables ....................................... xxi Acknowledgments ............................... xxiii 1. INTRODUCTION ........................... What Is Redevelopment and How Is It Financed? ........ Why Is Redevelopment Financing Controversial? ........ Purpose and Approach of This Study ................ Overview of the Report ......................... 1 2 4 6 9 2. THE REDEVELOPMENT PROCESS .............. Redevelopment Agencies ........................ Redevelopment Projects ........................ Project Financing ............................. Size of Redevelopment Activity .................... 11 11 13 15 16 3. POLICY ISSUES ............................. How Blighted Are Project Areas? ................... Do RDAs Generate Enough of the Growth in Project Areas to Earn Their Tax Increment? ................. Who “Loses” Under Tax Increment Financing? ......... Counties ................................. 23 25 28 29 30 xvii State Government ........................... 32 How Other Jurisdictions Get Some Relief: Fiscal Review and Pass-Throughs ........................ 32 4. STUDY DESIGN AND METHODOLOGY .......... Study Design ............................... Sample Project Selection ........................ Matched-Pair Selection ......................... Assessed Value Determination ..................... Parcel Description ............................ Benchmarks for Analysis ........................ Potential Sources of Bias ........................ 39 40 42 45 48 49 52 53 5. RESULTS OF THE STUDY ..................... 57 How Blighted Are Project Areas? ................... 57 Are RDAs Subsidized by Other Governments? .......... 60 6. CONCLUSIONS AND POLICY IMPLICATIONS ...... Conclusions ................................ The Extent of Blight and the Size of Subsidies ......... AB 1290: The Most Recent Reform Attempt ......... Beyond AB 1290: Addressing the Larger Policy Questions ... 69 70 70 72 73 Appendix A. Policy Issues Concerning Affordable Housing in Redevelopment Projects ........................ 79 B. Sales Tax and Redevelopment ..................... 87 C. Detailed Discussion of Study Methodology ............ 91 D. Regression Analysis............................ 99 E. Final Sample Projects .......................... 103 Bibliography .................................. 105 About the Author ............................... 109 Other PPIC Publications........................... 110 xviii Figures 2.1. Statewide RDA Revenues ..................... 2.2. Statewide RDA Expenditures .................. 2.3. Statewide RDA Revenues and Expenditures ......... 2.4. Cumulative Statewide Operating Expenditures, 1984–1985 to 1993–1994 .................... 3.1. Tax Increment’s Share of Property Taxes, by County ... 3.2. Pass-Through Share of Statewide Tax Increment ...... 3.3. Statewide Distribution of Pass-Throughs ........... 3.4. Local Government Pass-Throughs ............... 5.1. Redevelopment Areas as a Percentage of City Land Area .................................. 5.2. Growth in Assessed Value, 1983 to 1996 ........... 5.3. Share of Tax Increments “Earned” in FY 1994–1995, by Project ................................. 5.4. Net Subsidies to RDAs in FY 1994–1995, by Project ... 5.5. RDAs Generated Half of the Tax Increment They Received ................................ 17 18 19 20 31 34 35 37 58 61 63 64 66 xix A.1. Operating Expenses for RDA Housing Activities ...... 84 B.1. Redevelopment Projects and City Sales Taxes ........ 88 xx Tables 3.1. Legal Definitions of Conditions That Cause Blight .... 26 4.1. Original Sample Projects, by Year ................ 43 4.2. Comparison of Original and Final Samples in 1980 .... 45 4.3. Use of Blight Categories to Justify Sample Projects ..... 46 4.4. Comparison of Final Projects, Matched Areas, and the Rest of the City in 1980 ...................... 48 4.5. Summary Characteristics of Parcels in Final Sample .... 50 4.6. Distribution of Use Codes in Project and Matched Areas .................................. 50 5.1. Area Characteristics in 1980 ................... 59 5.2. Project Area Characteristics in 1980 .............. 60 A.1. Redevelopment Housing Activities, by Income Category ............................... 85 C.1. Breakdown of Parcel Changes, 1983–1996, by County .. 95 D.1. Estimation of RDAs’ Effect on AV Growth ......... 101 E.1. Final Sample Projects ....................... 103 xxi Acknowledgments On a project with as long, and as difficult, a gestation as this one, the author cannot help but be deeply indebted to the many people who contributed to the final product. Projects of this scope are possible only with the ongoing insights, comments, and assistance of colleagues, and at PPIC the author has been blessed with an outstanding group of fellow researchers and support staff. The greatest debt is to Laura Mameesh, a superb research assistant at PPIC who has served alternately as programmer, editor, sounding board, temp manager, and sanity check for the past two years. Eileen Roush, now at the California Department of Finance, also assisted in much of the early data collection and interviewing. Those at PPIC who have made significant contributions as formal or informal reviewers include Mark Baldassare, John Ellwood, Paul Lewis, Joyce Peterson, Fred Silva, and Michael Teitz; the usual caveat that remaining errors are the author’s alone certainly applies here. Outside technical reviewers also provided invaluable suggestions. The author xxiii thanks Timothy Bartik of the W.E. Upjohn Institute for Employment Research; Steven Sheffrin, from the University of California, Davis; John Landis, from the University of California, Berkeley; and Cliff Graves, President of Bay Area Life Sciences Alliance (formerly Executive Director, San Francisco Redevelopment Agency) for their time and thoughtful comments. Over the course of this project, the author interviewed many professionals in the redevelopment community, including redevelopment agency staff, state legislative staffers, consultants, lawyers, and government analysts. Several people were particularly important in the early phases of the project, as well as invaluable sources of experience throughout the project: Larry Arceneaux, President of Katz Hollis; Dave DeRoos, then Deputy Director of the California Redevelopment Association; and Peter Detwiler, consultant to the Senate Housing and Land Use Committee. The author thanks them for their generous cooperation. Others who provided valuable assistance on a regular basis were Wayne Beck, at the State Controller’s Office; Marianne O’Malley, of the Legislative Analyst’s Office; and Jennifer Swenson, at the California Research Bureau. The California Redevelopment Association was kind enough to allow the author and research assistants to attend a number of association conferences and provided many detailed comments on an early draft of this report. In particular, William Carlson, Executive Director, and Chris Hart, Legislative Analyst, have been helpful in this regard. This project was difficult and time consuming because of the data collection and validation required to identify and obtain historical parcel records for over 50,000 parcels in four counties. The most heartfelt thanks go to the temporary employees who carefully and tirelessly xxiv entered vast amounts of data from microfiche records: Paul Cheney; Tana Davis; Raul Poza; and Regis St. Louis; and to Phyllis and Bob Stratford, who had the unenviable task of reconciling parcel boundaries by poring over assessors’ maps in each of the four counties. Much assistance was provided by the staff in each county’s assessor’s office, and in the San Mateo Tax Collector’s Office. In particular, the author thanks Garo Megerdichian in Los Angeles, Ruby Murphy in San Bernardino, and Janet Dick in San Mateo for their tireless help; all three exemplified a researcher’s ideal of a civil servant and reflected great credit on their offices. Many more people than can be listed here have helped in one way or another, and the author also extends thanks to those whose names are omitted. xxv 1. Introduction In 1945, California passed legislation authorizing establishment of redevelopment agencies (RDAs) to alleviate “urban blight.” Subsequently, most cities and many counties have formed RDAs. In 1994–1995, the 351 RDA projects then active spent $3.3 billion to build almost 5,000 (primarily low-income) housing units and to rehabilitate almost 6,000 others, to provide infrastructure and parking for downtown commercial areas, and to build other public facilities. RDAs also displaced more than 500 households from their homes, received $1.5 billion in property taxes that would otherwise have gone to other public agencies, and initiated several new projects that consisted of thousands of acres of vacant land. This range of activities and outcomes helps explain why redevelopment generates so much policy controversy. Much of redevelopment’s appeal—and much of the opposition to it—revolves around tax increment financing: Once an RDA forms a project area, any increase in property taxes in that area (above a 2 percent annual inflation 1 factor) goes to the RDA. The rationale for this is that the improvements it has made in the area are responsible for the increase in property tax assessments. However, it is possible that other forces are contributing to a general rise in real estate values across an entire county. If so, diversion of this revenue to the RDA from the county, schools, and other local entities could be characterized as an involuntary subsidy. For example, money that counties could use to support state-mandated health or welfare services is being used instead by the agencies, and counties have no say in how it is used. It may be that higher state goals for redevelopment are well served by this subsidy and that the other taxing agencies can legitimately be expected to provide it. However, debate over this issue has so far operated in the absence of any hard evidence on the size of the subsidy. The purpose of this report is to raise general understanding of redevelopment by looking at the purposes of RDAs, their incentives, and how they operate. More important for the policy debate, it analyzes for the first time how much of the tax revenue RDAs receive is actually the result of their effect on property values and, thus, on property tax revenues. This introduction sets the stage by briefly addressing three questions: What is redevelopment? How is it financed? Why is redevelopment financing controversial? It also describes the purpose and approach of the study and the organization of this report. What Is Redevelopment and How Is It Financed? The California Community Redevelopment Act of 1945 enables any city or county to establish a redevelopment agency to combat urban blight—a power that three-fourths of the cities and a number of counties 2 in the state have since exercised. Over 97 percent of all RDA financial activity takes place in cities’ agencies, making redevelopment primarily a city-level phenomenon.1 Once established, RDAs designate specific redevelopment projects that focus on those portions of their city that they declare to be “blighted.” Within the boundaries of these projects, RDAs will purchase and assemble lots for use by developers, raze or build structures, provide infrastructure such as parking and lighting, and provide affordable housing. Blight is defined in state law as a combination of physical and economic conditions—such as vacant or decrepit buildings, declining property values, poverty, and high crime rates—that prevent private enterprise from developing the area. In areas suffering from such conditions, public intervention is seen as necessary before any private development can occur. It is sometimes the absence or inadequacy of publicly provided infrastructure that is said to deter private developers. To help RDAs overcome blighted conditions, the state gives the agencies some special powers that are controversial. Agencies can assemble property for sale to private parties and can use eminent domain if necessary to acquire private property that they want to sell, often at a discount, to a private developer. This is the only occasion when eminent domain can be used for private rather than public use. RDAs receive some city general funds and federal Community Development Block Grants to plan and to begin work on specific projects. The bulk of ____________ 1Unlike redevelopment authorities in many of the older cities in the East and Midwest, most of the redevelopment activity in California takes place in cities that are themselves suburbs. California has 332 cities with active redevelopment agencies, but only 35 of those cities are classified as central cities by the Census Bureau. Although an imperfect classification, this illustrates the difference between city-based notions of urban blight in California and in older areas of the country. 3 redevelopment activity is financed by issuing debt that is backed by property tax revenues that will accrue to the agency in the future. To facilitate this financing, RDAs are entitled to receive special property tax revenues. Any increase in property taxes that occurs after a project is designated—minus an annual 2 percent inflation adjustment—goes entirely to the RDA, instead of being shared among all the jurisdictions that normally share the area’s property taxes (such as counties and school districts). By law, these payments to RDAs last as long as there is debt outstanding on the project and often go on for 30 or more years. The original logic behind this arrangement—called tax increment financing— is that cities often bear the up-front costs of (re)development and yet have to share the tax revenues stemming from any increases in property values with all the other taxing jurisdictions. To encourage redevelopment2 and more equitably share its costs, tax increment financing was created in 1952 and applies to every redevelopment project. Why Is Redevelopment Financing Controversial? Fifty years after RDAs were authorized, tax increment revenues amount to $1.5 billion per year, or over 8 percent of all property tax revenues collected in the state. As noted above, RDAs are almost exclusively run by cities, and city RDAs receive 97 percent of all the tax increment revenues in the state. Cities receive an average of 11 percent of property taxes, so the tax increment revenues flowing from redevelopment almost double the amount of property taxes received by cities. In a few counties, RDAs receive as much as 18 percent of all ____________ 2The proximate motivation for this law was the matching funds that cities were required to pay to receive federal urban renewal funding. 4 property taxes—almost tripling the cities’ property tax receipts. Such large sums are guaranteed to generate controversy when other local governments are under fiscal stress themselves. The state government is also affected. When property taxes go to RDAs instead of school districts, the state is required to reimburse the school districts for the lost revenue. With school districts receiving an average of 51 percent of property taxes statewide, the state payments could be as large as $800 million per year; the counties’ forgone revenues are approximately another $300 million per year; special districts receive most of the remaining $500 million. Since the property tax constraints imposed by Proposition 13 in 1978, local governments have been searching for new sources of revenue. With the surge in the number of RDAs and redevelopment projects since 1978 have come charges that the fiscal pressures that local governments have experienced, combined with a loose legal definition of blight, create a strong incentive for those governments to identify blight within their boundaries. In practice, blight is a relative rather than an absolute concept. Areas that have been designated as blighted include both the Adams-Normandie neighborhood—site of the L.A. riots—and an affluent residential area in Foster City that needed some infrastructure improvements. Other evidence that the financial incentives from tax increment financing can influence the propensity to declare blight comes from the amount of cities’ land area that is covered by redevelopment projects. Roughly half of California cities have between 11 percent and 30 percent, and another fifth have well over 30 percent, of their land area in redevelopment projects. It seems unlikely that such large fractions of so many cities could be suffering from urban blight. Counties especially think that many cities view their redevelopment projects as a blank check 5 with which to transfer tax revenue from counties and the schools to their own RDAs.3 Although RDAs may provide useful public goods such as streets, parking structures, lighting, and other facilities, other tax jurisdictions would rather that cities use their own general funds to provide these goods. Cities and their RDAs reply to these charges by insisting that without their redevelopment activities, there would be no increase in property values to be taxed. The state legislature recognized and addressed some problems with redevelopment through the passage of AB 1290 in 1993. Among other things, the legislation tightened the definition of blight somewhat and changed the distribution of tax increment revenues. It arrived at a political compromise by requiring that any new RDA projects share approximately a third of their tax increment revenues with the other local taxing jurisdictions. Before this legislation, any such payments (called pass-throughs) were negotiated for each project case by case. These passthroughs averaged 14 percent in fiscal year 1993–1994, and counties received almost two-thirds of these payments. Purpose and Approach of This Study There have been no analyses to date of how much of the property value increases can be ascribed to the efforts of the RDAs rather than to overall real estate trends. In this study, we measured the effect of RDAs on assessed value growth to see whether their effect was large enough to ____________ 3See, for example, testimony by Los Angeles County to the Senate Committee on Local Government on December 7, 1989 (Report 457-S). In recent years, both Santa Clara and Los Angeles Counties have sued redevelopment agencies in the cities of San Jose and Los Angeles over a bond measure and project extension, respectively. (Both suits were ultimately unsuccessful.) San Diego and Riverside Counties dropped lawsuits against redevelopment agencies after negotiating pass-through agreements (see California Legislative Analyst’s Office, 1994). 6 be self-financing—which is the rationale for their existing arrangements. Any shortfall means that at least some of the RDAs’ activities have, in effect, been subsidized by other governments, and the study estimates the size of these subsidies. Because RDA project areas have been judged to be blighted, comparing their assessed value growth to that of the entire city they lie within might be misleading. If RDA project areas really are blighted, their growth prospects may well be worse than those of the rest of the city. To properly account for their initial condition, we compared RDA project areas with other areas in the same city that were similar to the projects in terms of poverty and vacancy rates. We then compared the change in assessed value from 1983 to 1996 for all the properties in each redevelopment project area with the change in its matched area. This growth in assessed value is what determines the amount of tax increment revenues that the RDA receives. We looked at all the redevelopment projects that began in the first five years after Proposition 13 passed in 1978. This allowed the projects between 14 and 19 years— half or more of their expected lifespan—to have produced an effect on the area.4 Out of 114 projects in ten counties, we were able to get sufficient historical records for 38 projects in three counties. These 38 projects were very similar to the original 114 projects, so the results can be considered representative of the larger group. RDAs cannot be considered self-financing simply because assessed values in the project areas grew faster than in their matches. Merely growing faster would demonstrate that the RDAs’ activity had at least some positive effect on local property values, but it is not sufficient proof that they were responsible for all the tax increment they received. The ____________ 4For cost reasons, the sample excluded any projects in counties with fewer than three projects started between 1978 and 1982. 7 sample projects kept 84 percent of the increase in property tax revenues that occurred within the project area, whereas other jurisdictions retained 16 percent. To justify the claim that they are self-financing, project areas therefore needed to grow six times as fast as their matched areas. Our study calculated how many of the projects actually had such rapid growth rates. It also looked at the tax-sharing formula mandated by the reform legislation in light of the actual growth rates. We also examined whether the reform legislation, which increased the amount of new property taxes that RDAs must share with other governments, picked the correct sharing formula if the program is meant to be self-financing. Under the new legislation, future redevelopment projects will keep an average of only 66 percent of any increases in property taxes; the study’s results allow estimation of the size of the subsidies implied by this pass-through rate. It is important to note at the outset what this study does not address. The study estimates the size of state and local government subsidies to redevelopment agencies in California. It does not analyze how effectively the RDAs have spent these subsidies, or whether their benefits have exceeded their costs. Any cost-benefit analysis of redevelopment is complicated by the vague and often conflicting goals embodied in its history. Redevelopment grew out of urban renewal efforts in the 1940s and initially focused on substandard housing; over the next two decades, it grew to encompass downtown renewal as well. The tension between housing and downtown development goals continues to this day, as numerous legislative hearings over the last several years readily attest.5 There have been recent debates over whether general economic ____________ 5See reports of previous Senate Housing and Land Use Committee hearings (1989, 1995, and 1996). 8 development is a legitimate purpose for redevelopment agencies, and legislation has also been passed to allow the use of redevelopment powers in communities affected by earthquakes and military base closures. Given the wide range of competing goals, there is no clear yardstick against which RDAs should be measured, and there is no single group whom they are to serve. Should they be judged by how much they improve the project area or the entire city? Are their achievements better measured by improvements in the lot of local residents or of the local property market? Furthermore, there may be goals for redevelopment that are not quantifiable even when universally agreed upon: Attributes such as “quality of life” or “more vibrant” communities do not readily lend themselves to economic analysis. Ultimately, the value placed on redevelopment projects’ achievements may outweigh concerns about whether RDAs are heavily subsidized. Nevertheless, in this era of constrained fiscal resources for local governments in California, it is important to know how much one level of government is subsidizing another. Any subsidies to RDAs might otherwise be spent by counties and the state on other programs that might also have significant public benefits. Once the size of the subsidies is known, policymakers will be in a better position to decide how they want to allocate those resources. Overview of the Report Chapter 2 provides a more detailed discussion of RDAs, redevelopment projects, how projects are financed, and the size of redevelopment activity in California. These descriptions provide the background necessary for the discussion in Chapter 3 of the policy issues raised by tax increment financing and blight determinations. Chapter 4 then presents the design and methodology of the analysis that addresses 9 these issues. The results and implications of the study are discussed in Chapters 5 and 6, respectively. Several appendices address related issues. Some persistent questions about redevelopment housing activities are discussed in Appendix A and questions about sales taxes are addressed in Appendix B. A more detailed discussion of the study methodology is presented in Appendix C, followed by results of an area-level regression analysis. Finally, Appendix E lists the redevelopment projects included in the final sample. 10 2. The Redevelopment Process Redevelopment is a broad, multifaceted undertaking in California. There is a great deal of variation among the kinds of communities that engage in redevelopment, as well as among the activities undertaken in the hundreds of project areas. The intent in this chapter is to provide background only on the aspects that primarily bear on the issue of tax increment financing. Redevelopment Agencies As discussed in Chapter 1, the California Community Redevelopment Act of 1945 was passed to help local governments eliminate urban blight. (The definition of blight and the controversy surrounding the practice of blight determination are discussed in Chapter 3.) The law allows the governing body of any city or county to establish a redevelopment agency if it declares that there is a need for one. In the first 20 years after the law was passed, only 46 agencies were formed throughout the state; in the next 20 years, 266 were formed. In 11 the middle of that period, Proposition 13 was passed, with its rollback of property taxes and a strong limit on the rate at which property taxes can increase. By 1986, most cities in California had an RDA; since then, only another 76 agencies have been formed. More than three-fourths of the cities and almost half of the counties have formed redevelopment agencies. However, not every local government that has formed an RDA necessarily uses its powers: In 1994–1995, 10 percent of all city agencies and 38 percent of county agencies reported no financial activity. Most of the inactive city agencies were those from cities with a population of less than 25,000; all but one city larger than 100,000 people have active redevelopment agencies. When a city or county forms an agency, a public referendum is required to ratify the decision to create the agency, since once it is formed it is endowed with substantial powers.1 In 90 percent of the agencies, the governing board of the RDA is the city council (and most of the staff are city staff), 6 percent of agencies use the county board of supervisors, and 4 percent have a separate organization.2 The larger city agencies tend to have a separate organization to oversee redevelopment activities—in such cases, the governing board is appointed by the city council. In most other agencies, the city and the redevelopment agency are essentially the same. City staff from the appropriate office carry out the relevant functions—for example, finance or planning—and the RDA’s budget is billed accordingly. RDAs can purchase and sell property, make loans or grants, enter into leases and other contracts, seize ____________ 1Since most RDAs are formed by cities, for clarity’s sake the discussion in the text will treat agencies as if they were city RDAs. Since 97 percent of redevelopment expenditures occur within city agencies, this approach does not distort the current situation. 2From California Redevelopment Association (1996), Chapter 2. 12 property by eminent domain and transfer it to other private owners, and construct or rehabilitate structures and improvements. The RDA will often work with the city manager or mayor, as well as the planning and other offices, to decide the priorities of the agency and to integrate the agency’s plans with those of the city as a whole. Redevelopment Projects To undertake any redevelopment activity, an RDA must form a project area—a specific geographic area that the RDA identifies as blighted and where it will undertake redevelopment activities. Most agencies have only one or two projects, but the larger agencies can have half a dozen or more projects active at any given time. In 1994, there were 731 redevelopment projects throughout the state. Project areas range from as small as two acres to as large as 22,000 acres, with most projects covering between 100 and 2,500 acres. An RDA must take a number of required steps in the “plan adoption process” by which a project area is formed. The city council adopts a resolution to designate a survey area to study for possible use as a redevelopment project.3 The agency will often have a particular kind of project in mind at this point—such as a shopping mall or housing development—but may not have worked out the details or selected a developer. The overall plan might envision several stages of development occurring over a 20- or 30-year period. Redevelopment projects’ stated objectives are roughly equally divided among commercial, residential, and public uses, and most plans entail a mix of these activities. ____________ 3The final project area may be smaller than the survey area but it must be within the survey area. For more details, see California Redevelopment Association (1996) and California Debt Advisory Commission (1995). 13 Redevelopment of a declining downtown area is one of the most common goals of RDAs, but agencies have undertaken a wide variety of projects. Examples of past redevelopment projects include constructing downtown hotel and parking structures, developing multiplex cinemas and retail malls, building or rehabilitating affordable housing units, and subsidizing freeway interchanges and access roads for industrial and commercial complexes. For a project area to be adopted, the RDA must complete a Preliminary Report (as well as an Environmental Impact Report and a report to the city council). The Preliminary Report establishes the project area’s boundaries and provides the evidence of blight conditions;4 describes the activities the agency intends to undertake to eliminate the blight conditions, including housing activities; and outlines the project’s financing structure. The plan also sets time limits on the life of the project and when it can undertake debt and sets the maximum amount of debt that can be outstanding at any given time. Some of the constraints on project formation and behavior changed under AB 1290, reform legislation passed in 1993. Details of this legislation are discussed in Chapter 3. The Preliminary Report is sent to other local governments that might be affected fiscally by the project and must be discussed at public hearings. If low- or moderate-income residents will be displaced by planned redevelopment activity, the RDA must form a Project Area Committee (PAC), consisting of persons with an interest in the project area. (In practice, PACs are formed in most project areas even if they are not required.) PACs are meant to provide an opportunity for a community to participate in its redevelopment plans. After this input, ____________ 4Blight definitions and related policy issues are discussed in Chapter 3. 14 the city council enacts an ordinance that formally creates the project area; in some cases a referendum is required as well. The planning and adoption process can take as little as several months to as long as several years; commonly, it takes one to two years from the initial steps to formal adoption. Once a project area is formed, the RDA will contract with one or more developers to help implement the redevelopment plan. Although this process sometimes involves open bidding, developer selection is more likely to take place through a request for qualifications (RFQ) or a request for proposals (RFP). To encourage the developer to produce the project it has in mind, the RDA will often provide incentives in the form of land write-downs (where agency-owned land is sold below the market price) and agency-provided infrastructure. The details differ markedly from one project to another, depending on market conditions and the nature of the project’s specific requirements. The RDA generally acquires and prepares the land, engages in cleanup if necessary, and constructs the necessary infrastructure (frequently including parking structures). The improved land is then transferred to the developer as part of the deal. Project Financing Most of the funding in the early years of redevelopment came from federal grants and loans, which typically required local matching funds to cover 25 percent or more of the project cost. To help local governments meet these matching requirements, the state legislature created tax increment financing—by which RDAs keep most of any future increases in property taxes in a project area. This was meant to allow RDAs to become self-financing, but it created a financial incentive for city-based 15 RDAs to define blight more broadly than they otherwise might to capture the future property tax revenues. (Chapter 3 addresses the question of how blighted redevelopment project areas were when they were formed.) The initial planning phase of a project is usually funded by a combination of city general funds (often in the form of loans), Community Development Block Grants, and sometimes funds transferred from other project areas run by the same agency. Occasionally, the developers involved with a project will lend money to the RDA to allow the deal to proceed quickly; they are then repaid out of subsequent tax increment revenues that the agency receives as assessed values grow. Once the project has progressed enough to be considered a reasonable investment risk by outside investors, the RDA will issue longterm debt to finance its activities. These tax allocation bonds are secured by the flow of tax increment revenues over the life of the project area. Lease revenue bonds are sometimes used to construct parking facilities, with parking fees pledged to service the debt. Other common sources of revenues that agencies receive from project areas include sales and use tax from retail sales, transient occupancy tax from hotels within the project area, any profits from the sale of land purchased by the RDA, and lease revenue (typically from parking structures provided as part of the development plan). Size of Redevelopment Activity Redevelopment in California is very big business. Statewide, redevelopment agencies received total revenues of $1.2 billion5 in 1984– ____________ 5All dollar figures are real (i.e., inflation-adjusted) 1993 dollars unless otherwise noted. 16 Revenue (billions of 1993$) 1985 (see Figure 2.1); this increased to almost $2.2 billion in 1993– 1994. Tax increment revenues rose from just under half of total revenues to 69 percent over the period. As happened with other intergovernmental programs, grant revenues declined by 22 percent whereas interest proceeds and “other” revenues were basically flat. Absolute tax increment revenues, however, rose by 64 percent, even after adjustment for inflation. Some of this increase reflects the strong real estate markets in the 1980s, and some comes from the increase in the number of active redevelopment projects under way. All other 2.5 Interest proceeds Grant revenues Tax increment 2.0 1.5 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.1—Statewide RDA Revenues 17 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Expenditures (billions of 1993$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Redevelopment expenditures went from $1.9 billion in 1984–1985 to $3.5 billion in 1993–1994 (see Figure 2.2). The largest share, approximately 40 percent, went to pay principal and interest on outstanding debt. The shares of most expenditure categories were fairly stable,6 although expenditures on real estate purchases fell sharply (to only 4 percent of expenditures) in the most recent year. This decline was All other Planning and design Administration costs 4.0 Interest and debt payments Real estate purchases 3.5 Improvement/construction 3.0 2.5 2.0 1.5 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.2—Statewide RDA Expenditures ____________ 6“Other” includes rehabilitation and relocation costs as well as the “All other” category in the controller’s office reports. No detail is given for the “All other” category, where the sharp increase occurred. 18 Revenues and expenditures (bIllions of 1994$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 most likely due to the effects of the recession on real estate markets in the state, as well as to concerns about reduced tax increment payments resulting from downward reassessments. Several agencies experienced fiscal distress during the recent recession with regard to meeting their debt service payments, and a number of agencies have cut back on new activity until their financial situation stabilizes. The revenues and expenditures imply a large annual deficit in RDA budgets (see Figure 2.3). This seems to be a result of the combination of financial and operational activity in their financial statements. As described above, RDAs sell bonds early in each project and use the proceeds to engage in development; along the way, unused portions of 4.0 3.5 Total expenditures 3.0 2.5 2.0 1.5 Total revenues 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 2.3—Statewide RDA Revenues and Expenditures 19 the principal can earn interest for the agencies.7 When the projects mature, more tax increment comes in and is used to meet debt service as well as to fund additional development. To get a sense of how agencies spend their money on actual project activities, an “operational” budget, which neglects debt finance expenditures, is shown in Figure 2.4. This is the sum of all nondebt service expenditures from 1984–1985 to 1993–1994 (again in 1993 dollars). All other 30% Improvement/ construction 38% Planning and design 5% Administration costs 13% Real estate purchases 14% SOURCE: California State Controller’s Office (annual). Figure 2.4—Cumulative Statewide Operating Expenditures, 1984–1985 to 1993–1994 ____________ 7Federal regulations prevent agencies from earning higher interest on investments than on debt. 20 The largest operating expense is for project improvement—site clearance, improvement (infrastructure), construction, and building rehabilitation (Figure 2.4). Real estate purchases were only 4 percent of total operating expenditures in 1993–1994, down from almost 30 percent of operating expenses in 1984–1985 and an average of 14 percent over the entire period. The earlier figures may have been higher than usual because of the large number of new agencies and projects begun in the early 1980s. RDAs generally purchase land at the outset of a project although, because of the successive stages of some projects, there may be several periods of land accumulation over the 30-year life of the project. The recent figure is further depressed because many agencies, particularly in Southern California, suffered financial pressures from the recession. These RDAs have cut back their activities substantially to be able to make their debt service payments. These are the basic operating facts about RDAs that are relevant to an examination of the extent to which redevelopment is subsidized in California. The tax increment financing system that underpins redevelopment has raised several critical policy issues, which are discussed in Chapter 3. 21 3. Policy Issues Redevelopment has been the focus of criticism over much of its life. The critics of redevelopment fall into one of several camps.1 Some critics are adamantly opposed to the use of government powers—such as eminent domain—which they view as amounting to involuntary transfers of private property from one owner to another. Some in this camp also view land write-downs—when RDAs sell land to developers below the agency’s cost of acquiring it—as something that should be considered “gifts of public funds.” Gifts of public funds are unconstitutional, and this ban underlies the lack of property tax abatements to specific firms in California (a practice that is widely used in other states).2 ____________ 1This categorization of redevelopment’s critics is based on the author’s interviews with many professionals involved in redevelopment, as well as on testimony given at the Senate Housing Committee hearings in October 1995 (particularly that of William Fulton, editor of California Planning and Development Report). Other summaries of the debate between RDAs and their critics can be found in the proceedings from recent state legislative hearings (California Legislature, 1989, 1995, and 1996). 2The courts have ruled, however, that the practice of land write-downs is constitutional. 23 More to the point of this study, other critics, consisting primarily of other taxing entities, view RDAs’ use of tax increment financing as a diversion of property taxes that are properly theirs. This group often notes that the biggest surge in redevelopment agency and project formation followed the passage of Proposition 13. In their view, RDAs are using tax increment financing as a way around the political obstacles to raising taxes. Lastly, some observers generally support the concept of redevelopment but oppose what they see as glaring abuses in recent years, such as the designation of primarily vacant land as blighted; the focus on attracting retail firms that generate sales tax revenues, such as car dealerships, into project areas;3 and the reluctance of many agencies to provide affordable housing.4 The last set of critics tend to be concerned that some RDAs have lost sight of their obligation to alleviate urban blight in favor of maximizing local tax revenues, thus endangering the status of redevelopment in general. This report focuses on the concerns of critics who claim that the large tax increment revenues that flow to the RDAs are not justified by their effect on assessed values. Implicitly, any property tax revenues received in excess of those revenues generated by the RDAs’ actions amount to involuntary subsidies from other governments. There are currently no estimates of the size of these subsidies, so no one knows how much the current system of redevelopment costs the state and other local governments. Related to this issue is the question of whether the ____________ 3A brief discussion of the role of sales tax revenues in redevelopment can be found in Appendix B. 4Although not a focus of this report, affordable housing has been central to the policy debates about redevelopment since the program’s inception. Appendix A summarizes the history of affordable housing under redevelopment and examines the policy issues that are commonly discussed. 24 financial incentives created by tax increment financing (together with the vague definitions of blight in existing law) cause RDAs to declare more areas to be blighted than actually are. The most recent chapter in redevelopment reform was the state legislature’s enactment of the CRL Reform Act of 1993 (AB 1290). AB 1290 was designed to address several problems with redevelopment that had not been corrected by previous legislation.5 These are the bill’s primary provisions: penalties for RDAs that fail to meet the required level of effort for providing affordable housing, somewhat stricter blight definitions required in setting project boundaries and implementation plans, prohibitions on RDAs’ receiving a portion of sales taxes, specific time limits on projects’ ability to incur debt, and institutionalizing the previously informal transfer payments between RDAs and other taxing entities affected by redevelopment projects. The critics’ practical charges that are examined in this report can be boiled down to two key questions, both of which were addressed by AB 1290: 1. How blighted are redevelopment project areas? 2. Can RDAs justify the tax increment they receive by their effect on growth, or are they subsidized by other governments? How Blighted Are Project Areas? As noted in Chapter 2, RDAs must make a finding of blight to use their powers of eminent domain, tax increment, etc. This flows from redevelopment’s roots in urban renewal programs. The past urban ____________ 5For more details about the problems AB 1290 was meant to address, refer to the reports on recent legislative hearings cited in footnote 1 of this chapter. 25 renewal focus meant that the definition of blight originally concerned residential—specifically slum—areas. If an area was blighted, it was assumed that private industry acting on its own would not invest in the area; concerted government action was therefore required. According to state law,6 a blighted area is one . . . in which the combination of conditions set forth in Section 33031 [listed in Table 3.1] is so prevalent and so substantial that it causes a reduction of, or lack of, proper utilization of an area that it constitutes a serious physical and economic burden on the community which cannot reasonably be expected to be reversed or alleviated by private enterprise or governmental action, or both, without redevelopment. This legal definition of blight is not precise enough to allow a definitive assessment of whether a given area is or is not truly blighted; blight is also supposed to be a characteristic of an entire project area and is not necessarily true of every property in the area. Table 3.1 Legal Definitions of Conditions That Cause Blight Physical Conditions That Cause Blight Unsafe or unhealthy buildings Factors that prevent economically viable use of buildings or lots Adjacent uses that are incompatible and prevent economic use Subdivided lots of irregular form and shape or of inadequate size Economic Conditions That Cause Blight Depreciated or stagnant property values Abnormally high business vacancies Lack of necessary commercial facilities Residential overcrowding that leads to public safety and welfare problems High crime rate SOURCE: Drawn from Health and Safety Code Section 33031. ____________ 6This definition comes from Health and Safety Code Section 33030 (b), as reprinted in Beatty et al. (1995). 26 The possibility of 30 years of tax increment revenues can create a temptation in some cities to be rather liberal in defining blight. As one local official put it in an interview, “blight is might.” The financial incentives to test the limits of the blight concept grew much stronger after the passage of Proposition 13 in 1978. Financially strapped cities could earn a greater share of the property taxes generated within their boundaries by setting up redevelopment projects. Since there is no oversight authority to police redevelopment agencies, the only check on agencies’ behavior is the threat of a lawsuit by some person or jurisdiction claiming to be hurt by an agency’s actions.7 Lawsuits must be filed within 60 days of a plan’s adoption, which can limit the ability of residents and other taxing agencies to block an adopted plan that is objectionable. The broad definitions of blight in the law provide quite a bit of latitude for an agency’s designation of project area boundaries. Table 3.1 lists the main definitions of blight conditions in current state law (i.e., AB 1290). Their breadth raises the question of whether an area is blighted merely because it is a low-income area, since many places throughout California could arguably have at least one of each of these conditions. If all low-income areas may be considered blighted, blight requirements provide little guidance in targeting redevelopment efforts. Before AB 1290 was passed, either physical or economic conditions could deem an area as blighted for purposes of project designation. Under AB 1290, however, both physical and economic blight conditions ____________ 7It could be argued that the threat of action by the state legislature—such as the restrictions imposed by AB 1290—also acts as a check on RDAs’ behavior. Such a check works more through the California Redevelopment Association, with its concern for the collective health of redevelopment in the state, than through any individual agency that has little reason to be concerned with any repercussions to other cities. 27 must be documented to classify an area as legally blighted. AB 1290 no longer permits projects that are not “predominantly (i.e., more than 80 percent) urban,”8 but the inclusion of largely vacant land in project areas was legitimate under state law in earlier years. The new law also requires that project plans for the first five years demonstrate how the agency’s actions would eliminate the specific blight conditions that were used to justify the formation of the project area. AB 1290 also curtails the use of “inadequate infrastructure” and other conditions involving shortcomings in public improvements. The law rescinds agencies’ ability to receive sales tax revenues from firms located in project areas—a post-Proposition 13 authority that was widely used to give rebates to retail stores and auto malls. This change was aimed at stopping the numerous cases of vacant (sometimes agricultural) land being declared blighted to attract sales-taxgenerating businesses within a given city’s borders. Do RDAs Generate Enough of the Growth in Project Areas to Earn Their Tax Increment? Although California was the first state to adopt tax increment financing, it is now used in at least 32 states.9 In theory, tax increment financing can be seen as desirable from a state’s perspective, since cities bear the entire initial cost of redevelopment but, in the absence of tax increment, have to share the future property tax benefits of that ____________ 8Note that “urban” does not necessarily imply no vacant land, although it is assumed to mean not undeveloped. Most of the project areas with large percentages of vacant land apparently were undeveloped. 9See Huddleston (1986) and Stauber and Wyatt (1990) for a more detailed discussion of the theory of tax increment financing and its use in Wisconsin; Stinson (1991) discusses its use in Minnesota. Davis (1989) and Klemanski (1990) discuss the incentives that tax increment financing creates for different levels of government and the political dynamics and lack of accountability that affect its use. 28 redevelopment with other taxing entities. Tax increment financing, therefore, allows for cost-sharing among all the entities that will benefit from subsequent redevelopment. This theory assumes that in the absence of some form of tax-sharing, there will be too little development, since the local city will not recoup enough of its costs through future increases in property tax revenue. In practice, whether tax increment schemes are actually self-financing depends on how much RDAs contribute to local growth. If the increase in property values in project areas is solely the result of the agencies’ actions, there is no real diversion of funds—as the RDAs themselves argue. If some of the increase is due to general trends in real estate markets, however, any tax increment revenue generated by these trends and received by RDAs is essentially an involuntary subsidy from counties and school districts (and ultimately the state) to the cities. Such a subsidy is neither budgeted for, nor approved by, the governments providing it, and it comes at the expense of whatever public programs those governments would otherwise have spent those funds on. Who are these governments and what relief have they received to date? Who “Loses” Under Tax Increment Financing? Use of tax increment financing of redevelopment is one of the biggest sources of contention between different levels of government. Counties (and to some extent the state) see tax increment revenues as property taxes that are diverted from other taxing entities to the cities via redevelopment agencies. 29 Counties Because of both an increase in redevelopment activities and the sharp increase in property values in the 1980s, tax increment revenues in redevelopment project areas grew rapidly as a share of total property taxes. Tax increment’s share was 4.6 percent of all property taxes collected throughout the state in 1984–1985, and grew to just under 8 percent in 1993–1994.10 This varied greatly by county: The share of property taxes accounted for by tax increment revenues ranged from zero in 15 counties to 18 percent in Riverside County. To put the tax increment share into perspective, the statewide average share of property taxes going to cities was 11 percent in 1993–1994, and the overall county share was 21 percent. Each county has its own allocation formula for how property tax revenues are to be shared among all the taxing entities within that county, so the overall tax distribution is unique to each county. Given how tight municipal finances have been, such a realignment of tax revenues would be expected to generate conflict between the affected governments. Figure 3.1 shows the growth of tax increment’s share of property taxes in the top ten financially active counties.11 Although all the counties showed an upward trend, Riverside and San Bernardino Counties stand out with more than a tripling of tax increment’s share of property taxes over the last ten years.12 ____________ 10This overstates the total somewhat, since the totals include amounts transferred to other jurisdictions as pass-throughs (described below). The overall pattern across counties and over time is unaffected. 11Counties were ranked by total 1993–1994 revenues received by the county and its cities, school districts, special districts, and redevelopment agencies. The source was California Department of Finance (1995). 12The uniform decline in increment shares in 1992–1993 is due to absolute declines in tax increment that year, presumably due to payments to the Educational Revenue Augmentation Fund (ERAF); overall property tax revenue still increased. 30 Percentage 1982–1983 1983–1984 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 Riverside San Bernardino Los Angeles Santa Clara Contra Costa 20 Orange 18 Alameda San Diego 16 Sacramento 14 Fresno 12 10 8 6 4 2 0 SOURCE: California State Controller’s Office (annual). Legend lists counties in order of tax increment share in 1993–1994. Figure 3.1—Tax Increment’s Share of Property Taxes, by County The California Redevelopment Association (CRA) notes that some counties themselves have active agencies and asserts that some of the variation in tax increment by county represents activity in county-level agencies. A review of county RDA activities shows that this has not been the case, however. Despite the extent of its redevelopment activity, Riverside County’s agency, for example, accounts for only 5 percent of the tax increment revenue generated by all redevelopment projects inside the county. San Bernardino, the county with the second-highest fraction of property taxes diverted to tax increment, generated none. This reflects 31 the general insignificance of county agencies with respect to tax increment: Statewide, county agencies account for an average of only 2.7 percent of all the tax increment revenue generated within the county. Only 11 county agencies generated any tax increment revenues at all in FY 1993–1994. State Government After Proposition 13 passed, the state took on the obligation to reimburse school districts for property tax revenues lost to other taxing entities. Estimating this burden on the state is difficult, but a ceiling on its value is probably $800 million annually. This number assumes that redevelopment affects school districts uniformly across counties and applies the statewide average receipts of 51 percent of property taxes, less schools’ 2 percent pass-through share of total increment. A more accurate accounting would have to examine each county separately and subtract whatever share of the increased assessed valuation is due to redevelopment activities (the subject of this report’s statistical analysis), as well as the money received through the ERAF requirement. It would also need to account for any in-kind contributions, such as buildings and infrastructure, that RDAs might provide to some school districts. How Other Jurisdictions Get Some Relief: Fiscal Review and Pass-Throughs In 1976, AB 3674 gave counties the responsibility to convene fiscal review committees, allowing them to negotiate a return of some of the tax increment revenues generated by projects within their boundaries. After Proposition 13 drastically reduced property taxes, the loss of funds from tax increment became more acute and local governments fought for 32 every source of funds. In 1983, other taxing entities were also given fiscal review powers in recognition of the potentially large effect that tax increment might have on them. In practice, these fiscal reviews have mainly been performed by counties and often result in an agreement by RDAs to share some tax increment with them. Since other taxing entities were given fiscal review authority over proposed redevelopment projects, RDAs started to share some of the tax increment with the affected entities. These payments (pass-throughs) were the result of project-specific negotiations between the RDAs and some of the other entities. The share of tax increment paid back to counties and other entities as pass-throughs has increased significantly over the last ten years, growing from 3.5 percent of total tax increment in 1984–1985 to 14 percent in 1993–1994 (Figure 3.2). This does not necessarily mean that RDAs are agreeing to share more and more of their tax increment in each year’s new projects. Although some of the increase in pass-throughs comes from higher rates, some comes from the big increases in property values that many project areas have enjoyed. Since most of the older projects did not involve pass-throughs, the maturing of projects that did have pass-throughs means that their effects are now starting to show up in financial statements.13 For projects begun after December 31, 1993, AB 1290 fixed the pass-through rate using a formula that was uniform across all projects. The financial data underlying Figure 3.2 do not include any noncash transfers that occur between RDAs and other entities. Some passthrough agreements allow for construction of infrastructure or other ____________ 13Financial data on the sample projects (those started between 1978 and 1982) show that the share of tax increment revenues paid out as pass-throughs does not change much over the life of projects. 33 Percentage 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 AB 1290 35 30 25 20 15 10 5 0 SOURCES: California State Controller’s Office (annual); AB 1290 estimate from author’s calculations. Figure 3.2—Pass-Through Share of Statewide Tax Increment facilities in lieu of, or in addition to, any cash transfers. The CRA asserts that these noncash contributions are a large (but unspecified) amount, and that schools in particular benefit from these arrangements. Schools do have an incentive to receive noncash rather than cash transfers, since the state will not take the former into account when determining how much to reimburse school districts for property taxes lost to redevelopment. In the absence of any source of information about these agreements, it is impossible to estimate how much larger the total passthroughs would be if the noncash transfers were included. In interviews with numerous redevelopment project managers and consultants, none were able to document the value of these capital projects. In addition, 34 Pass-throughs (billions of 1994$) 1984–1985 1985–1986 1986–1987 1987–1988 1988–1989 1989–1990 1990–1991 1991–1992 1992–1993 1993–1994 none thought that the value of the noncash contributions would approach even half the value of the reported (i.e., cash) pass-throughs. The CRA also asserts that pass-throughs are likely to rise (by prior agreement between RDAs and other entities) over the life of a project, since there is little tax increment generated in the early years of a project. Among the sample projects studied for this report, the average passthrough over the 1985–1994 period rose only from 10 percent to 12 percent. Although it rose in 47 of the projects, it also fell in 27 others, so the trend is neither large nor uniform across projects. Figure 3.3 shows how pass-throughs were distributed under the existing pattern of individually negotiated agreements for each project. Counties received 71 percent of all pass-throughs in 1984–1985, 2.5 Special districts 2.0 School & CC districts Cities 1.5 Counties 1.0 0.5 0 SOURCE: California State Controller’s Office (annual). Figure 3.3—Statewide Distribution of Pass-Throughs 35 although their share fell somewhat—to 62 percent in 1993–1994. As Figure 3.3 shows, the falling county share was largely due to the growth in school and community college (CC) districts’ pass-through payments. Despite this falling share, however, the absolute dollars that counties received in pass-throughs increased from under $20 million to almost $150 million over the last ten years. The formula under AB 1290 made the pass-through rate uniform across all projects. The formula uses a sliding scale, with pass-throughs starting at 25 percent of the increment;14 another 21 percent is added to increases in assessed values after the tenth year, and another 14 percent is added for any increases after the thirtieth year. The total of the passthroughs over the life of a project depends on how much and how quickly assessed values rise. Using historical assessed values and inflation rates as an estimate of future trends, the total pass-through under AB 1290 is estimated to amount to over 33 percent of the tax increment. This estimate differs from other estimates, since it expresses the passthrough share in present-value terms. Such estimates are very sensitive to the choice of the rate at which future payments are discounted to present value. In this case, the historical average inflation rate from 1970 to 1992 of 5.8 percent was used as the discount rate. Since most analyses use a real (i.e., inflation-adjusted) discount rate greater than zero, this is a very conservative estimate, which tends to overestimate the pass-through share of tax increment when expressed in present-value terms.15 ____________ 14The pass-through percentages are calculated after contributions to the Low and Moderate Income Housing Fund have been subtracted from gross tax increment revenue. As discussed in Appendix A, RDAs are required to place at least 20 percent of their tax increment revenues into a fund for the provision of affordable housing. 15A model created by the California Redevelopment Association has calculated the average pass-through under AB 1290 to be 37 percent. 36 School districts get the majority of property taxes statewide, but they have had little incentive to negotiate cash pass-throughs with the local RDA, since the state made up for any revenue losses. As noted above, some of the difference might be the fact that schools are said to be more likely to receive noncash than cash transfers. Counties, on the other hand, had no such recourse. Counties rely on property taxes for 15 percent of their total revenue (43 percent of their own-source revenue)16 and, as was seen in Figure 3.3, have been the most aggressive local governments in seeking pass-through agreements. AB 1290 required that the pass-throughs be shared among all affected tax entities according to the same formula that applies to the base property taxes. Figure 3.4 Pass-throughs (billions of 1994$) 6.0 Other 5.0 Schools Counties 4.0 Cities 3.0 2.0 1.0 0 FY94 AB 1290 SOURCES: California State Controller’s Office (annual) and author’s calculations. Figure 3.4—Local Government Pass-Throughs ____________ 16Local government finance data were taken from the California Budget Project (1996). 37 shows the dollar values of this effect on each of the types of jurisdictions affected. Although the overall fraction of tax increment to be passed through to other tax entities will increase, counties will be even worse off than they were before the passage of AB 1290. Their share of passthroughs now averages 62 percent, but their share of pass-throughs from new projects will be the same 21 percent that they now receive from all other property taxes. Under the current rules, counties receive 8.7 percent of all tax increment revenues (62 percent times 14 percent); under AB 1290 rules, they will receive 6.9 percent (21 percent times 33 percent) of all increment over the life of the projects. Given counties’ financial straits and the fact that they are increasingly called upon to provide social services, it is no surprise that tax increment financing causes conflict. Different levels of government perform different services, so such subsidies can change the distribution of public expenditures. For example, counties spend relatively more on public assistance, courts, and jails, whereas cities spend more on public utilities and transportation. With the financial straits that cities themselves are in, it is also not surprising that redevelopment—with its promise of increased property taxes without a required vote—is a popular tool. Chapter 4 describes the methodology employed in this study to estimate how much of increased property values can be ascribed to the actions of the redevelopment agency, as opposed to general real estate market trends and other factors. 38 4. Study Design and Methodology This study addresses the issue of how blighted redevelopment project areas are. It also tackles the central point of contention between levels of government over redevelopment: Whose money is being used for redevelopment? Unless the RDAs cause enough growth in assessed values to cover the tax increment revenues they receive, such revenues are, in effect, subsidies from other governments to each project. RDAs defend their claim on tax increment revenue by asserting that most of the increases in property values in project areas are due to their redevelopment activities and not to generally improving economic conditions. As discussed in Chapter 3, AB 1290 significantly increased the effective pass-through rate from the current average of 14 percent to an average of roughly one-third for new projects. There was no empirical basis for this particular allocation, since there were no estimates of the RDAs’ effect on property values. The uniform pass-through rate set by the legislature was a political compromise, primarily between the RDAs 39 and the state, which faced the largest financial liability because of its obligation to reimburse school districts for any lost tax revenues. Without any independent estimates to resolve this debate, no better allocation of tax increment was possible. Study Design What we really want to know is how fast each project area would have grown in the absence of the redevelopment project. Since that is not possible, the best way to address this question is with a control group. Project areas are supposed to be chosen because they are blighted, which means that they are unable to grow without deliberate government involvement. If this is true, comparing project area growth to that of the entire city may be misleading: The comparison will likely underestimate the amount of growth that redevelopment caused. To properly credit project areas for their initial condition, this study uses a matched-pairs design. (This type of study is also referred to as a case-control study in the biomedical literature.) Each project area examined is compared with another area in the same city that is most similar to the project area in terms of blight conditions. The measure of property value used here is the assessed value of each parcel. Economic studies generally focus on market value, but in the context of estimating property tax revenues, the assessed value is more relevant. Changes in assessed value, not market value, determine the actual amount of tax increment revenues a project area generates. In post-Proposition 13 California, in contrast to most other states, assessed value increases at only a 2 percent annual “inflation” rate unless a property is sold or major construction is done on the site. Assessed value and market value are therefore equal only when properties are sold or 40 constructed or when a large decline in property values causes a property to be reassessed. Changes in assessed value from 1983 to 1996 for every parcel in both the project area and its match were compared to estimate how much of the growth in the project area was caused by the action of the RDA. 1983 was chosen as a base year, since it was the first year that all the projects in the sample had been formed. 1996 was chosen as the final year, since it was the most recent year for which assessed value was available when the study began. There was some concern about the effect of the recession on project areas’ growth relative to their matches’ growth. Since assessed value changes tend to lag market value changes by about two years, 1993 was a good comparison point to use for the recent recession. Assessed value declines were, in fact, more common in project areas than in matches from 1993 to 1996, but not by nearly enough to affect any of the study’s conclusions. The median change in assessed value from 1993 to 1996 was nearly identical (an increase of 4.4 percent) between the projects and matches, as well as across counties. Over this time period, the total assessed value in the state increased by over 150 percent, so that the increases in the sample project areas were expected to be large enough to be detected (even in blighted areas). This increase can be compared with the 29 percent increase that would have occurred if no properties had been sold or undergone major construction—assessed values would then have increased solely by the 2 percent annual inflation factor.1 For each project, the growth rate of its matched area was subtracted from the project’s own growth rate to estimate the portion of the growth in assessed value that was due to the RDA’s actions. The amount of tax ____________ 1The 29 percent figure is the result of compounding the 2 percent annual increases for 13 years. 41 increment revenues generated by this growth rate was compared with the amount of tax revenues the project area received, after accounting for the pass-through rate of the project and the 2 percent annual inflation factor. The share of property tax growth due to redevelopment was estimated for each project, showing the range of subsidies or extra revenues that occurred in these projects. Summed across all projects, the difference between the amount of tax increment generated and the amount received shows how large the subsidies to RDAs from other jurisdictions have been. The results also provided an estimate of what the pass-through rate for new projects should be to have RDAs be self-financing in the long run. If the legislature wants redevelopment to be self-financing,2 the optimal pass-through rate should be the rate at which the tax increment that was self-generated (averaged across all projects) was equal to the amount of tax increment that was received. Sample Project Selection It is impractical to measure the effect of every redevelopment project on assessed values. Recent projects have not had enough time to affect property values; and historical assessed value data are not widely available more than ten years back. Cost considerations also limit the size of study that can be undertaken. To fairly assess the effect of redevelopment on assessed values, this study examined a sample of projects that began long enough ago to have had an effect but recently enough to obtain historical assessed values from early in each project’s life. The original sample ____________ 2Recall from the discussion in Chapter 3 that the theoretical basis for tax increment financing assumed that it equalizes the costs and benefits of development across local governments over a project’s lifetime. 42 examined consists of all projects started between 1978 and 1982.3 This time period captures projects that began right after Proposition 13, and which have had enough time to develop property to have a detectable effect on aggregate assessed value in the project area. In addition, there was no significant change in redevelopment law over this period, so all of the projects operated under similar rules. There were 134 projects begun in this time period, representing 18 percent of the 669 projects that were active statewide in 1994. In that year, they accounted for approximately $666 million in expenditures (also 18 percent), out of the statewide total of $3.8 billion. Table 4.1 shows the number of projects and their average acreage, by the starting year of the project. The size of the project areas ranges from 3 acres to 9,000 acres, with a mean of over 1,000 acres. Given the one very large project, the median size of almost 500 acres gives a more accurate sense of the typical size of these projects. The comparable statewide median Table 4.1 Original Sample Projects, by Year Starting Year 1978 1979 1980 1981 1982 Projects 18 21 14 30 51 Average Acres 631 1,247 1,113 1,477 838 SOURCE: California State Controller’s Office (annual). ____________ 3Given the need to travel to each county’s assessor’s office to resolve any missing parcel data, projects in counties with two or fewer projects were omitted for cost reasons. This affected 20 projects in 13 counties. Other projects were deleted later for various technical reasons (see California Redevelopment Association, 1996, and Appendix C). 43 size for all redevelopment project areas was 441 acres in 1994. These projects were started during a period of rapid increase in the number of redevelopment agencies and projects. From 1976 to 1985, 154 new redevelopment agencies were established, doubling the existing number; and a total of 300 new project areas were formed (a 150 percent increase).4 Of these original sample projects, 37 were the first projects undertaken by their redevelopment agency. Southern California counties are slightly more represented in the sample (73 percent) than they are among all redevelopment projects (63 percent), reflecting the region’s faster growth in the 1980s as well as the more aggressive use of redevelopment in some cities in the region. These data are reassuring in that the projects in the original sample were similar to the greater population of redevelopment projects. The results obtained from examining this group cannot be used to estimate the effect of projects from other time periods, however, since existing redevelopment law as well as overall economic conditions have changed over time and cannot be adequately controlled for. But they can explain what happened to one-fifth of the redevelopment projects in the state, and those projects were quite similar to the rest of the state’s projects in terms of size and expenditures. Unfortunately, not every project in the original sample could be used in the study—the biggest obstacle was the availability of historical assessed value data. After eliminating counties that had fewer than three projects begun between 1978 and 1982, 114 projects located in ten counties remained in the group, but historical microfiche data could be obtained for only four counties with 67 projects begun in that time ____________ 4California Redevelopment Association (1996). 44 period. The inability to find matches because of several remaining data problems left a final sample of 38 projects and their matches.5 The three counties in the final sample are Los Angeles, San Bernardino, and San Mateo; these counties are a good cross-section of the state and the projects in them are representative of the original sample of 114 projects. Table 4.2 shows the relationship between project areas and the rest of the city that they lie within, for the original sample and the final sample. In the original sample, the vacancy rate in the project areas was 64 percent higher than in the rest of the city, whereas in the final sample it was 28 percent higher; the median household income was about 12 percent lower in both the original and final samples; and median rents in the original sample were 91 percent of that in the rest of the city, while in the final sample they were 96 percent. Table 4.2 Comparison of Original and Final Samples in 1980 Projects Rest of city Projects Rest of city Vacancy Median Rate, % Income, $ Original Sample 7.8 25,337 4.7 29,198 Final Sample 5.5 26,044 4.3 29,291 Median Rent, $ 422 462 437 454 SOURCE: 1980 Census Block Group data, STF-3A file. Matched-Pair Selection The goal is to choose an area in each city that in effect “could have been” a redevelopment project, at least by the stated definitions of ____________ 5More detail on why the projects were dropped is provided in Appendix C. 45 blight.6 Since cities have different (and unmeasured) tax and regulatory structures that might affect property values, each project area is compared with another area within the same city that is selected for its similarity to the project area in 1980. By selecting matches within each city, any omitted factors that vary across cities and affect property values are controlled for. Other studies that have examined the effects of local development programs, most notably enterprise zones, demonstrate the importance of making comparisons between areas within the same city. To determine the characteristics on which to match the pairs, we looked at the blight definitions that were used to justify each of the 114 projects in the original sample and grouped them into seven general categories. Table 4.3 shows the general categories of blight definitions and the number of projects that cited them in their plans. The three most commonly cited categories—underutilized property, inadequate infrastructure, and deteriorating buildings—all involved property characteristics. Table 4.3 Use of Blight Categories to Justify Sample Projects Blight Category Underutilized/vacant property Inadequate infrastructure Deteriorating buildings Unemployment, crime, poverty Parking/traffic problems Other economic problems Aesthetic reasons No. of Projects Citing 55 65 58 35 20 52 16 SOURCE: Individual project planning documents. ____________ 6Before making this choice, any other redevelopment project areas in each of these 80 cities were removed from consideration. 46 We then chose those variables in the 1980 Census (at the Block Group level) that best approximated some of the blight conditions cited by the redevelopment agencies. Census Block Groups are generally areas comprising between six and eight city blocks; in less populated areas, Block Groups can encompass much larger areas of land. Regardless, Block Group areas are defined so that population counts are roughly comparable across groups. The Census variables initially selected as the best proxies available for the blight conditions were percentage of units vacant, percentage of structures built more than 30 years earlier (i.e., before 1949), median rent, and percentage of individuals in poverty. The characteristics for which there are no good proxies in the Census data were cited less frequently, except for inadequate infrastructure. Given the fact that “inadequate infrastructure” was highly correlated with the other blight conditions, the lack of a Census proxy was not considered a serious problem.7 Although the vacancy and poverty rates are obvious proxies for blight, some reviewers of the matching methodology viewed the age of structures and median rent as less reliable in capturing blight conditions than vacancy and poverty rates. There is no simple relationship between the age of a structure and its value—while some periods of rapid growth produced low-quality buildings that did not retain value well, some very old building stock is considered quite desirable. Use of median rent could underweight the experience of the low-income portion of the population. The final matching formula, therefore, used the two-variable version, with vacancy and poverty rates.8 In several cities that had more ____________ 7AB 1290 removed inadequate infrastructure as sufficient evidence of blight. 8To check the sensitivity of our matching to the use of the two- or four-variable formula, the matching program was run for both definitions. In all but six cases, the 47 than one project in the final sample, the same match was selected for more than one project. In these cases, one of the projects’ second choices was selected as the match. Once again, the overall match function value in each case changed by only a very small amount. Table 4.4 compares the vacancy rate, median household income, and median rent for the final sample, the matches, and the rest of the cities. By averaging across all the pairs, these figures do not completely capture how closely each project and matched area compared. Nevertheless, they do show that using matched pairs probably improves on a comparison of project areas to the rest of the city. Table 4.4 Comparison of Final Projects, Matched Areas, and the Rest of the City in 1980 Final sample projects Matched areas Rest of city Vacancy Rate, % 5.5 5.0 4.3 Poverty Rate, % 14.5 15.5 14.9 Median Income, $ 26,044 28,470 29,291 SOURCE: 1980 Census Block Group data, STF-3A file. Median Rent, $ 437 440 454 Assessed Value Determination Once matching census block groups were selected, every parcel in each Block Group was identified and downloaded from the Dataquick property database.9 This database contains the parcel number, 1996 assessed value, tax rate area codes, and other characteristics of each parcel ____________________________________________________ Block Group chosen as the closest match was the same in both cases; in those six cases the absolute value of the match function from the two-variable formula was very close to that from the four-variable formula. 9The process of finding the corresponding 1983 assessed value for each parcel in the 1996 database is described in more detail in Appendix C. 48 from databases maintained by county assessors’ and tax collectors’ offices. Parcels in redevelopment projects were identified by their tax rate area codes. The list of parcel numbers in 1996 was used to obtain the assessed value in 1983 from each county’s historical microfiche. In those cases where a parcel’s physical boundaries were changed between 1983 and 1996, the assessor’s office map books were used to find the corresponding parcel number(s) in 1983. If fewer than 90 percent of the parcels (or parcels with less than 90 percent of the total 1996 assessed value) in a project or match area could be reconciled, the pair was dropped from the analysis. Parcel Description There were over 47,000 parcels in the final sample containing the 38 project pairs. Because project areas tended to be larger than the typical Census Block Group, almost two-thirds of these parcels were in project areas. The characteristics of the project parcels and matched-area parcels are summarized in Table 4.5. Compared with those in matched areas, the parcels in project areas had a larger mean assessed value (AV) in 1983 ($88,460 compared with $66,230), were less developed in 1983 (their mean percentage improvement then was 56 percent, compared with 66 percent in the matches), and had a building stock that was newer in 1996 (mean year built was 1963 compared with 1960).10 These differences are all statistically significant. ____________ 10As noted in Table 4.5, percentage improved and year built were weighted by 1983 assessed value. The year built as of 1983 was not available for all the parcels that changed, so the averages are not reported here. For the majority of parcels that did not change, the difference was not as great. See Table C.1 for the distribution of parcel changes between 1983 and 1996 in the project areas and matches. 49 Table 4.5 Summary Characteristics of Parcels in Final Sample No. of parcels Mean assessed value in 1983 Mean % improved in 1983 Mean year built Projects 30,160 $88,460 56% 1963 Matches 16,931 $66,230 66% 1960 t-Value of Difference t = –5.51 t = 59.80 t = –33.96 SOURCE: Author’s calculations from Dataquick database. NOTE: Percentage improved and year built were weighted by 1983 assessed value. There were also noteworthy differences in the composition of parcels by use codes (Table 4.6).11 Consistent with the experiences of many redevelopment consultants and RDA project managers, project areas had fewer residential parcels and more commercial and industrial parcels than the matched areas. Almost 88 percent of the matched parcels were residential, compared to 74 percent of the project parcels. In place of residential properties, the project areas had twice as many commercial Table 4.6 Distribution of Use Codes in Project and Matched Areas Use Code Residential Commercial Industrial Government/utility/nonprofit Vacant Project Parcels, % 73 12 6 2 7 SOURCE: Dataquick property database. Match Parcels, % 88 5 1 1 6 ____________ 11Use codes are assigned by each county’s assessor’s office to define what type of use a parcel is primarily for. For the purposes of this report, parcels are classified as either residential, commercial, industrial, government/utility/nonprofit, vacant, or other. Fewer than 1 percent of the parcels in either project or matched areas were “other,” so they are not reported in Table 4.5. In each county, the actual use codes are more detailed than those in these reported categories. 50 parcels as the matches did. They also had many more industrial parcels than the matches, although the industrial parcels were concentrated in a few project areas. As might be expected in areas with targeted redevelopment projects under way, parcels in project areas were more likely to physically change over the 13-year period than were those in the matched areas. One-third of the project parcels were combined or split between 1983 and 1996, almost three times the rate for the matches. It is often said that an important service provided by RDAs is to combine small or oddly shaped parcels into larger lots to make them more marketable. In fact, only 186 parcels out of the 9,000 project parcels that physically changed between 1983 and 1996 (just over 2 percent) were combined from several parcels into one. RDAs were much more likely to divide one or several parcels into many parcels, usually for condominiums or new single-family subdevelopments. Only 8 percent of the parcels in the matched areas were created from larger parcels, but over 26 percent of the project parcels in 1996 originally came from larger parcels. Across all 38 pairs, the mean growth of assessed value in all the project parcels was 204 percent; parcels in all of the matched areas saw their assessed value grow by 145 percent. This difference, too, was statistically significant at the 99 percent level. (If no parcels had been sold or had major construction take place, the 2 percent annual inflation factor under Proposition 13 would have caused assessed values to grow by 29 percent over these 13 years.) Growth at the parcel level might be misleading, however, since without matching by project area we cannot control for local factors that can affect assessed values. To make full use of the power of the matched-pair methodology, we must compare each project with its match and see what the relative growth rates were for 51 each project. This not only restricts the comparison to the matched area for each project but also avoids having the problem that the largest project areas will dominate the results if we pool all the projects and matches together. The growth in assessed value in the matched area of each pair is subtracted from the total assessed value growth in the project area to estimate the RDA’s effect on the growth in that project area.12 The size and direction of any subsidies that might have occurred were estimated by comparing the total amount of assessed value that the RDAs created to the amount of tax increment revenue they actually received. Benchmarks for Analysis In many economic studies of the effect of specific government programs, the yardstick used is whether the program had any effect at all. RDAs themselves have argued in reply to critics of tax increment financing that, were it not for agency activities, there would be no increase in assessed value to be taxed. In measuring the size of the subsidies that redevelopment agencies receive under the tax increment financing system, it is not sufficient to show that RDAs had some effect on growth in the project areas. Since tax increment financing has been justified as a self-financing mechanism, the benchmark is more stringent: Project areas must generate enough growth in assessed value to cover the tax increment revenues the RDAs receive (less any pass-through payments they make). Given that, how high a standard does this set? Overall, the projects in this sample retained almost 84 percent of the tax increment revenues that were generated within their boundaries; the ____________ 12The actual formulas used in these calculations are reported in Appendix C. 52 other taxing jurisdictions received just over 16 percent. This means that, in the aggregate, they need to have grown more than six times as fast as their matches to be justified in the claim made by RDAs that the tax increment they receive was due to their actions. Anything less than that implies that the projects as a whole are subsidized; anything more means that they are not only self-financing but that they actually produce additional revenues for other governments. Not every project needed to grow six times faster than its matched area, since some projects had much higher pass-through rates than the average for all projects. In Chapter 5, the results are presented both for the individual projects (using their specific pass-through rates) and for the sample overall. This will demonstrate the range of success and failure of projects in terms of being self-financing, and will permit an estimate of the size of any subsidies that flow from other taxing jurisdictions to RDAs. Potential Sources of Bias The use of matched areas as a control group avoids some of the common problems in estimating the effect of a program on some economic outcome. By restricting the matched area to the same city as the project, we can also control for political, regulatory, and other factors that could distort the results. However, even with this kind of control group approach there are sources of bias in the results that cannot be controlled for. There are several sources of potential bias in the study, and they act in opposite directions. If the RDAs actually chose the worst area of the city as a project area, then by definition no possible match can be quite as bad off as the project area. The effect of this selection bias should be to cause us to 53 underestimate the effect of RDAs on assessed value growth. This is only true if the RDAs did in fact choose the worst area of the city (with “worst” defined in terms of the potential of the area to increase in property value if unassisted). As discussed in Chapter 3, there are strong financial incentives to declare as blighted those areas that can be quickly (re)developed; therefore, at least some of the project areas are probably not the worst area of the city. Working in the opposite direction is the strategic behavior that some agencies engage in when determining the boundaries of project areas. According to a number of consultants, lawyers, and analysts familiar with redevelopment practices, many RDAs draw at least some of their project area boundaries to include properties they know will soon be sold or developed. This enables the agency to capture tax increment revenues early in the project’s life. Where this occurs, there will be a bias toward overestimating the effect of RDAs on growth in assessed value, since the agency gets credit for growth that it did not influence. Also working to overestimate the effect of RDAs on assessed value growth is the way that Proposition 13 changed the practice of reassessing properties. Since Proposition 13 passed, assessed values increase only 2 percent per year unless a property is sold. If parcels in project areas are more likely to turn over than parcels in matched areas, their assessed values will tend to be closer to market value than those in the matched areas. Since parcels in project areas were in fact more likely to turn over between 1983 and 1996, this phenomenon has the effect of exaggerating the RDAs’ effect on assessed values. There is no practical way to measure the presence, let alone the relative size, of these effects. In interpreting the results though, the fact that there are forces working in opposite directions provides some protection against serious errors in estimation. 54 For any one of these biases to alter the conclusions, they would need to be large enough—and to arise systematically across most projects—to overcome the effect of the opposing source of bias. 55 5. Results of the Study This study sought to answer two questions raised by the critics of redevelopment: How blighted are project areas, and how much are RDAs subsidized by tax increment financing? How Blighted Are Project Areas? To consider the effect of financial incentives on blight determination more closely, the study looked at a sample consisting of the 134 redevelopment projects that were started in the first five years after Proposition 13 passed (between 1978 and 1982). Project plans, maps, and financial statements as well as Census data were analyzed to get a sense of the condition of these projects. In their own reports that demonstrated the existing blight conditions, the sample projects collectively cited physical conditions (underutilized/vacant property, inadequate infrastructure, and deteriorating buildings) more frequently than economic and social conditions (such as high unemployment, 57 crime, and poverty) in their plans.1 In 25 of the 134 projects, more than 50 percent of the project area was vacant land in 1986–1987—the first year these kinds of data were reported in the annual Controller’s Report. Across all agencies, 19 cities still had more than half of the land area in all of their projects vacant as of 1986–1987. Some RDAs have declared that most of their city is blighted. Figure 5.1 shows the number of cities by the amount of their land area covered by redevelopment projects.2 Ten cities were more than half Number of cities 90 80 70 60 50 40 30 20 10 0 0–10 11–30 31–50 51–100 Percentage of city land area SOURCES: U.S. Census Bureau (1994); California State Controller’s Office (annual). Figure 5.1.—Redevelopment Areas as a Percentage of City Land Area ____________ 1Blight conditions in the sample projects are discussed in more detail in Chapter 4. 2The data on city land area come from the U.S. Bureau of the Census (1994) and do not include information on cities with fewer than 25,000 residents. Thus, some cities well-known for being almost all redevelopment project by land area (such as Emeryville and Industry) are not included in this figure. 58 redevelopment project by land area. These top ten are Cathedral City, Chico, Fontana, Huntington Park, Manteca, Pittsburg, Ridgecrest, San Marcos, San Pablo, and West Sacramento. Another 25 cities were between one-third and one-half redevelopment project by land area. At this point, it is no longer possible to compare the sample project areas’ actual physical condition at the time of their formation with the condition of the rest of the city. However, the sample project areas can be compared with the rest of the cities that they lie within, using the available Census data.3 Table 5.1 shows the vacancy rate, median household income, and median rent in 1980. As should be expected if project areas were chosen because they are doing less well than the cities as a whole, vacancy rates were higher—and incomes and rents were lower—in the project areas than in the rest of the city. The average differences are not large, although the differences within each city vary widely. These comparisons mask a large amount of variation among the projects, and in some cases the project areas in one city had better 1980 statistics than other cities’ nonproject areas. This brings up another important conceptual issue: whether blight is meant to be an absolute Table 5.1 Area Characteristics in 1980 (current dollars) Vacancy rate Median income Median rent Sample Projects 8% $25,337 $422 Rest of City 5% $29,198 $462 SOURCE: 1980 STF 3-A Census files. ____________ 3In this section, the rest of the city excludes any other redevelopment projects that lie within the city regardless of their start dates. 59 concept or if it should be judged solely in relative terms within a given city. As currently practiced and under the legal definitions cited above, blight is defined strictly on a relative basis. This can be seen in Table 5.2, which shows the 1980 characteristics of three project areas that were started between 1978 and 1982. Note that these figures are for the project area itself, not the entire city in which the project is located. As long as every city is empowered to establish a redevelopment agency, and each agency can establish project areas wherever it can demonstrate blight, there will be project areas in almost every city regardless of how their degree of “blightedness” might differ from city to city. Table 5.2 Project Area Characteristics in 1980 Characteristic Year agency formed No. of acres in project % vacant land 1980 household income Median home value Agency and Project Area Los Angeles: Foster City: Adams-Normandie Foster City 1948 1981 400 855 35 58 $14,915 $40,164 $93,363 $234,637 SOURCES: annual Controller’s Reports; STF-3A Census files. Victorville: Bear Valley 1981 1,248 95 $17,907 $112,514 Are RDAs Subsidized by Other Governments? The other, and more critical, question this study sought to answer was whether RDAs are actually self-financing or whether they receive subsidies from the state and other local governments. The benchmark for evaluating this question was presented in the previous chapter: In the aggregate, assessed value in the project areas needed to grow, on average, six times faster than in their matches for the projects to be self-financing. 60 Since each project area has a different pass-through rate, the actual ratio needed to claim self-sufficiency is different for every project. The growth in assessed value for each project and its match is shown in Figure 5.2.4 Virtually all of the project areas and matches had rapid assessed value growth from 1983 to 1996. All but six of the project areas and four of the matched areas at least doubled their assessed value over the 13-year period. This performance is of a similar magnitude to general assessed value changes: Total assessed value in the three counties in the final sample grew by more than 150 percent over the same period. Percentage growth 1,800 1,600 1,400 Project Match 1,200 1,000 800 600 400 200 0 Project/match SOURCE: Author’s calculations from Dataquick database. Figure 5.2.—Growth in Assessed Value, 1983 to 1996 ____________ 4The projects in Figures 5.2 through 5.4 are displayed in the same order in all three charts—in order of increasing net subsidies. 61 The fastest growth was in project areas—six projects grew by over 400 percent, but no matched area did; 15 projects more than tripled their assessed value, when only two matches did. The mean assessed value growth in project areas was 270 percent, compared with an average 144 percent growth in the matched areas. Two-thirds of the projects—27 out of the 38 pairs—actually had faster growth in assessed value than their matches did. This means that, in most cases, redevelopment did have a positive effect on assessed value growth in the project areas. However, the true effect of redevelopment on assessed value is only that growth over and above the growth in the matched area. To evaluate how much of each project’s retained tax increment revenues was generated by the RDA, we must take into account not only the relative growth between the project area and its match, but also the size of the pass-through payments that the project made to other jurisdictions (and the 2 percent annual inflation factor that the projects do not receive). A project may have grown much faster than its matched area, but because of a very low pass-through rate it still might not have grown fast enough to cover all the revenues it received. In contrast, a project may have outgrown its matched area by only a small amount but if it passed through most of its tax increment revenues, it might have been self-financing. The formulas used to calculate the RDAs’ share of the tax increment revenue growth are described in Appendix C. Figure 5.3 shows what share of tax increment revenue each project was responsible for. If a project has a 100 percent share, it means that the project generated exactly enough growth in assessed value so that none of the tax increment revenues it received were diverted from other governments. 62 Percentage earned 200 100 0 –100 –200 –300 –400 SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.3—Share of Tax Increments “Earned” in FY 1994–1995, by Project (Throughout this chapter, “tax increment received” refers to the retained tax increment. This is the tax increment revenue left after subtracting any payments to other jurisdictions in the form of pass-throughs.) Only four projects generated more tax increment revenues than they actually received in FY 1994–1995, which was the latest year for which financial data were available. Two projects generated over 90 percent of their tax increment revenues, and two other projects generated between 80 percent and 90 percent of their revenues. Therefore, fewer than a quarter of all the projects were actually self-financing—or at least close to it—as far as their tax increment revenues were concerned. Another quarter of the projects had such combinations of slow relative growth and low pass-through rates that they generated none of the tax increment revenues that they received (those shown with negative shares). The 63 other 19 projects generated between 2 percent and 75 percent of the revenues they received. The eight projects that generated at least 80 percent of the tax increment revenues they received had an average of 51 percent vacant land in their areas in 1986–1987; in contrast, the other 30 project areas had an average of only 14 percent vacant land. It seems that the project areas that outgrew their matches by the greatest amount were engaged more in development than in redevelopment. What all this means for the size of subsidies that pass from counties and the state (via payments to the school districts) depends on the size of the projects. If all the best performers are large projects, their effect could more than offset the subsidies received by smaller projects and the net subsidy could be quite low. But if the reverse were true, the net subsidy could be quite large. Figure 5.4 shows the net subsidy that each 1 0 –1 –2 –3 –4 SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.4—Net Subsidies to RDAs in FY 1994–1995, by Project 64 Transfers (millions of $) project received in FY 1994–1995. Positive numbers indicate that a project generated even more tax revenue growth in the area than it received in (retained) tax increment revenue; negative numbers show the size of the subsidy a project received. As seen in Figure 5.4, only four projects generated enough growth in assessed value to produce a surplus of tax increment revenues. In twothirds of the projects, the financial transfers either way amount to less than $1 million annually. In the remaining 13 projects, the annual subsidies received from other jurisdictions exceeded $1 million for each project area, and in three of those project areas, the other taxing jurisdictions lost over $3 million in property taxes to the RDA. Summed across all 38 projects, the RDAs collectively accounted for 51 percent of the tax increment they received in FY 1994–1995. Figure 5.5 shows the amount of tax increment generated and the amount received in the sample projects. These projects received a total of over $78 million in tax increment revenues (after pass-throughs), and generated just under $40 million of those revenues by their effect on the growth in assessed value. This implies total annual subsidies flowing from all the other taxing jurisdictions to these redevelopment projects of over $38 million.5 This subsidy estimate would be reduced by whatever amount, if any, that these projects provided in the form of noncash contributions to the other taxing jurisdictions. These findings can also give a sense of how large the annual subsidy is for the entire sample of project areas that were formed between 1978 ____________ 5The data and estimation approach used in this study do not permit a calculation of the total subsidies received over the lifetime of these projects. Such a calculation would require the annual relative changes in assessed value between the project and match, as well as any pass-through payments made; it would also require tracking any physical changes in parcels (i.e., those combined or split) on an annual basis. 65 FY 1995 tax increment (millions of $) 80 70 60 50 40 30 20 10 0 Received Generated SOURCES: Author’s calculations from Dataquick database; California State Controller’s Office (annual). Figure 5.5—RDAs Generated Half of the Tax Increment They Received and 1982.6 The final three counties used were not selected according to any characteristics of the counties—they are the only counties for which 1983 assessed value data were available. Within those counties, the projects in the final sample were those for which historical assessed values could be found for at least 90 percent of the 1996 parcels and assessed value. The only systematic difference between the projects included in the final sample and all the projects started between 1978 and 1982 was that in the omitted projects there were either no possible matches or map ____________ 6These results cannot be extrapolated to projects that were begun either before 1978 or after 1982. The legal environment, real estate markets, fiscal pressures on cities, and a host of other factors were all significantly different from those facing this group of projects. 66 data were missing. The results for the final 38 projects can be extrapolated to the 114 projects in the ten counties in which more than three projects started between 1978 and 1982. If the other 76 projects had a similar distribution of outcomes as the final sample, the total subsidies to this group—constituting one-sixth of all redevelopment projects in California—are very large. These 114 projects received $347 million in tax increment revenues in FY 1994–1995, and if they generated the same share of the assessed value growth as the final sample projects did, the subsidies to this cohort of projects was $170 million in the last fiscal year alone. Appendix D shows regression results using pair-level observations. Pair-level regressions more closely approximate the use of matched pairs, but parcel-level analysis permits the use of information such as the assessed value in 1983 and degree of improvement on the property. Both sets of estimates show RDA effects of similar magnitude. 67 6. Conclusions and Policy Implications By virtue of the nature and size of redevelopment agencies’ financing, as well as their extensive powers and minimal state oversight, redevelopment is very controversial. There are disagreements over other issues, such as housing, but the biggest battle is over the $1.5 billion per year that RDAs receive in property tax revenues that would otherwise go to the state, counties, and special districts. How much of this $1.5 billion is a subsidy from other governments depends on how much of the assessed value growth can be attributed to redevelopment projects, rather than to general real estate trends. Policy debates about redevelopment have thus far proceeded without any estimate of the size of the subsidies, so there has been little, if any, debate as to how large these subsidies should be. How comfortable policymakers are with the level of subsidies partly depends on the extent of blight in project areas. This study used matched pairs of project areas and similar nonproject areas in the same city to estimate RDAs’ effect on assessed values and examined 69 characteristics of project areas and their cities to see how much worse the project areas were than the rest of the city. Conclusions The Extent of Blight and the Size of Subsidies Assessing the extent of blight is hampered by the lack of any quantitative definition of blight. Of the 186 cities in California with populations above 25,000, 35 cities had more than 30 percent of their entire land area under redevelopment. Another 82 cities had between 10 percent and 30 percent of their land area in redevelopment projects. The Legislative Analyst’s Office (LAO) has reported that RDAs placed about 100 square miles of land in new project areas in 1993 alone—more than double the average amount placed in the preceding three years.1 The report assumed that this increase was a response to new restrictions contained in the reform legislation enacted in 1993 (AB 1290). However, it gives a sense of the large areas being declared blighted on an ongoing basis. It seems implausible that these areas are all suffering from such a combination of blighting conditions “that it constitutes a serious physical and economic burden on the community.” Analysis of 1980 Census data for the projects that were formed between 1978 and 1982 shows that the project areas were somewhat worse off than the rest of their cities, but there is wide variation among those cities in income and other measures. An area that is worse off than the rest of an affluent city is not necessarily a “serious physical and economic burden” to that city. Assessing the size of the subsidies for redevelopment from other taxing agencies is more straightforward. From the matched-pair analysis, ____________ 1See California Legislative Analyst’s Office (1994). 70 we found that two-thirds of the sample project areas grew faster than their matches did, implying that redevelopment did have some positive effect on assessed values. However, given how much of the tax increment revenues the RDAs actually kept, they would have had to grow much faster than their matches to have generated all the property tax revenues they received. The results show that only four of the 38 projects examined generated enough growth to cover all the tax increment revenues they received. Four other projects generated at least 80 percent of the revenues they received. The remaining 30 projects received subsidies ranging from several thousand dollars to almost $4 million in fiscal year 1994–1995. Across all the projects, the RDAs generated just 51 percent of their tax increment revenues in fiscal year 1994–1995. This means that other governments—the state, counties, and special districts—provided subsidies to these RDAs of over $38 million that year (after accounting for pass-through payments). If the 76 other projects in the original sample had the same effect on growth in assessed value as these projects did—and it is reasonable to assume that they did—the subsidies from other governments just for redevelopment projects begun between 1978 and 1982 would be over $170 million annually. Some RDAs make noncash contributions to school districts and other governments, and the value of these contributions should also be subtracted from the estimated subsidy. No estimates of the size of these contributions exist, however, and none of the redevelopment experts interviewed believed that these contributions were even half the size of the reported cash pass-throughs. Given existing information, this report’s estimate of the subsidies to redevelopment in California is the most accurate estimate possible. 71 The eight least-subsidized projects—those that were responsible for at least 80 percent of their tax increment revenues—had an average of more than 51 percent vacant land when they began. This extent of vacant land would most likely have disqualified them as blighted under AB 1290, and it supports the critics’ contention that many of the fastergrowing projects were engaging more in development than in redevelopment. The remaining 30 projects had an average of only 14 percent vacant land. In fact, the biggest predictor of growth in assessed value in an area was how much of the value of a property was from improvements, rather than from land. In other words, the less developed a property was, the more its value grew. In the eight fastest-growing projects, 38 percent of their 1983 assessed values derived from improvements on the land. In contrast, in the other projects, 62 percent of their assessed values came from improvements, which is comparable to the average of 64 percent for the matched areas. AB 1290: The Most Recent Reform Attempt Concern about the size of the subsidies and the incentives provided by the loose definition of blight are evidenced by the passage of AB 1290. That it was seen as a significant revision to existing redevelopment law is suggested by the many RDAs that rushed to adopt projects before the law came into effect in January 1994. Nevertheless, the blight definitions are still very subjective and leave many areas of the state eligible for incorporation into redevelopment projects, if the definitions are taken at face value. By setting the new, uniform pass-through rate well above the average rate for current projects, the legislature (at least implicitly) decided that the existing subsidies, of unknown size, were too large. Although the 72 new pass-through rate does increase the percentage of tax increment revenues that must be shared with other governments, it will probably result in continuing large subsidies for RDAs. If the pass-through rates mandated by AB 1290 had been in effect for the projects examined in this report, those projects collectively would have generated 61 percent (instead of 51 percent) of the tax increment revenues they received. Ironically, AB 1290’s allocation formula will make the counties even worse off financially than they were under the old system. Although overall pass-through payments will increase, the counties’ share of those payments will shrink. If all existing projects had been under AB 1290 rules in 1994–1995, counties would have received $28 million less in property tax revenues. These changes will happen at the same time that counties come under increased fiscal pressure from welfare reform, as well as from demands for health care and other social services. Beyond AB 1290: Addressing the Larger Policy Questions The fact that the state, counties, and special districts subsidize RDAs by as much as hundreds of millions of dollars each year is not necessarily undesirable. Many RDAs play a vital role in improving blighted areas and, collectively, the RDAs provide a large number of affordable housing units throughout the state. Many of the policy problems concerning redevelopment stem from the lack of a clearly stated and unambiguous goal for redevelopment. In past hearings and discussions on this topic, the most frequently invoked goals were alleviating urban blight, providing affordable housing, downtown development, and economic development. In the absence of a legislatively specified goal, it is difficult to assess the redevelopment program and to provide substantive policy 73 recommendations. The definitions of blight in current law reflect this ambiguity of purpose. These problems are exacerbated by the weak and decentralized oversight provided by the state. There is no agency that systematically reviews RDAs’ plans and activities for compliance with the law; enforcement requires court challenges by local residents or other taxing agencies, generally within 60 days or less of plan adoption. These are the key policy questions that remain to be answered, even after AB 1290: What is the state trying to accomplish with redevelopment? Should the state, counties, and special districts subsidize RDAs’ pursuit of these goals? If so, how large should the subsidies be? Where should this money be spent? Who should decide on the money’s allocation and monitor the RDAs’ performance? This report’s recommendations follow from the fact that these questions still need to be asked more than 50 years after the California Community Development Act of 1945 was first enacted. 1. The legislature should formally clarify its goal(s) for redevelopment. The various goals espoused for redevelopment not only compete for available resources, they often imply conflicting choices of project areas and expenditures. If the historical purpose of alleviating blight were the goal, the project area would be the most destitute neighborhood in the city and the project would involve affordable housing, provision of social services, anti-crime efforts, etc. If downtown development were the goal, the core commercial area would be chosen and project activities would be more oriented toward retail and commercial development and provision of parking lots and related infrastructure. If economic development were the goal, RDAs would have to focus much less on property markets and any particular area, and more on labor markets and training. In this case, most of their existing tools are poorly suited to the task. 74 There may be instances where these goals overlap or are complementary, and an overly narrow focus may unduly constrain the activities of some projects. That said, without a more specific statement of intent by the legislature, the major tensions surrounding redevelopment will continue. 2. Blight conditions need to be aligned with the goal(s) of redevelopment and should be made more precise. With the financial incentive of tax increment revenues, blight conditions cannot remain in the eye of the beholder if redevelopment efforts are meant to target the most serious cases of blight. This means that they must be more like the quantitative criteria used to determine eligibility for enterprise zones in some states. Criteria such as a poverty rate of at least 20 percent of the population, 20 percent population loss in recent years, x percent of the buildings or assessed value abandoned, y percent of property taxes in arrearage, or the crime rate z times the state average. If the redevelopment subsidies are to be targeted, blight must be judged more on an absolute than on a relative basis. It is not enough that a project area be worse than the rest of the city. In a poor city, the average area may be sufficiently blighted to warrant assistance; in an affluent city, its worst area might still not qualify. One way to address this might be to qualify cities as to whether they are eligible to receive tax increment revenues or to provide lower pass-through rates for more disadvantaged cities. However it is done, redevelopment subsidies need to be better targeted. Asking the legislature to compile a list of cities that qualify for tax increment revenues is a very tall task. Even if there were the political desire and discipline to do so, there would likely be a gradual expansion of the eligible list, as happened with the Model Cities program of the 1960s and as has happened with enterprise zones.2 Even assigning different ranges of pass-through rates to cities ____________ 2Since the first ten zones were designated in 1986, the number of enterprise zones has grown to 25, and nine Employment and Economic Incentive Areas have been added. 75 according to degree of need might be too difficult politically. Unless these kinds of steps can be taken, however, redevelopment will remain a melange of blight remediation and commercial development aimed at increasing property tax revenues. 3. There should be some form of oversight authority to monitor RDA behavior and to avoid the need for frequent legislative interventions. Both the LAO and the State Auditor, citing the problems stemming from the lack of meaningful oversight, have suggested designating one state agency as the primary authority over RDAs.3 Creation of an oversight authority with suitable enforcement powers would allow a more systematic evaluation of whether a proposed project complied with existing law. Such an authority could request clarifications of ambiguities in legislation, as well as improved data on project activities and financial statements. However, an oversight authority would face its own problems. None of the state agencies proposed for the oversight role are eager to assume this task, and opponents of government expansion would resist the creation of a new bureaucracy if a separate entity were proposed instead. Any oversight authority would also be subject to strong political pressures from the same groups that have been involved in legislative reform debates. Although an authority could ensure compliance with existing law, it would be unable to target redevelopment activities to the most blighted areas without a more objective and restrictive definition of blight. Absent any legislative guidance on the matter, it would also be unable to set the size of financial assistance for redevelopment. 4. If the legislature intends redevelopment to be self-financing, the passthrough rate should be increased significantly. Raising the pass-through rate is difficult to do retroactively, since debt has been issued with ____________ 3See California Legislative Analyst’s Office (1994) and California State Auditor (1996). The LAO recommended that the Attorney General review every project plan for compliance with the law; the State Auditor did not recommend a particular agency. Other agencies that have been suggested for the oversight authority role are the Departments of Finance, and Housing and Community Development. 76 future tax increment revenues pledged for debt payments. With over 700 projects already in place, the level of their subsidies would not change quickly, no matter what pass-through rate was mandated for new projects. Over time, however, a higher rate would decrease the subsidies paid to RDAs. The legislature may well support the idea of spending large subsidies on redevelopment, and there are good reasons to do so. A targeted level of subsidies would put redevelopment subsidies into competition with other appropriations, where its merits (noneconomic as well as economic) could be weighed against other public uses for those funds. Policymaking would be greatly improved if an explicit level, or at least a range, of subsidies were ratified by the legislature. As long as redevelopment relies on tax increment financing for its funding rather than being directly subsidized, there will be conflicts between state goals and local interests. At either the pre- or post-AB 1290 pass-through rates, there will also be large annual subsidies passed from the state, counties, and special districts to RDAs. Until the legislature expresses explicit goals and subsidy levels for redevelopment, no amount of tinkering with the detailed rules is likely to end the conflict over redevelopment in California. 77 Appendix A Policy Issues Concerning Affordable Housing in Redevelopment Projects The financial incentives created by the prospect of tax increment revenues sometimes lead agencies to be more focused on activities that will generate increases in assessed valuation than on providing affordable housing. Providing affordable housing has been viewed by many as a core activity since redevelopment began, and the state legislature has repeatedly strengthened the affordable housing requirements in redevelopment law. Do RDAs Produce Enough Low-Income Housing? History of Housing Under Redevelopment Housing activities, especially those targeted toward low-income residents, were the original motivation for the federal urban renewal 79 programs of the 1940s. The Federal Housing Act of 1949 provided funds for cities to engage in urban renewal projects that originally stressed slum clearance. Federal aid was available as long as the project area was residential, regardless of what replaced the residential buildings. The two groups that had supported the urban renewal programs—real estate developers and housing advocates—had competing visions for the use of redevelopment funds, visions that continue to clash in the present. In 1954, the Federal Housing Act was amended to allow 10 percent of federal grants to be used in nonresidential areas; in 1959 it was 20 percent; in 1961 it was raised to 30 percent; and in 1965 it became 35 percent.1 The combination of residents’ increasing resistance to largescale evictions and growing interest in downtown revitalization led to the decline of slum clearance as a redevelopment goal by the late 1960s. The tension between developers and housing advocates led to the creation in 1976 of the Low and Moderate Income Housing Fund (AB 3674, Montoya). This required all new projects to set aside 20 percent of their tax increment revenues to increase the supply of affordable housing and required that any demolished low-income housing in project areas be replaced one for one. In 1986, the set-aside and replacement requirements were extended to all project areas, regardless of start date. Among the major changes in AB 1290 (the 1993 reform legislation) was the creation of a “use it or lose it” penalty for low- and moderate-income housing funds. Agencies with an excess surplus— defined as more than $1 million—in unused 20 percent housing setaside funds within a certain time period must turn the funds over to their respective county housing authority. Agencies are just now approaching ____________ 1Discussion of federal urban renewal programs draws heavily from Frieden and Sagalyn (1989). 80 these deadlines, and there may still be delays before the eventual disbursement of these funds. Because of the rise in tax increment revenue and the decline in federal housing funds, redevelopment agencies’ low- and moderateincome housing funds are one of the largest pools of public money available for local housing projects. Activities can include building new low- and moderate-income2 (i.e., affordable) housing, rehabilitating and purchasing existing housing, and subsidizing rent for low-income households. In 1993–1994, the state’s RDAs built 4,521 housing units and had another 10,000 in progress. They also rehabilitated an additional 4,539 units, although it is not clear how extensively the units were rehabilitated.3 A number of observers have commented that much of the reported rehabilitation is minor and mainly cosmetic, resulting in no real increase in affordable housing in the area. AB 1290 addressed this issue by formally defining substantial modification. Over the last eight years, the annual number of units built and rehabilitated has ranged from 5,000 to 9,000. Approximately 50,000 housing units have been constructed or rehabilitated since 1987–1988. Although this pales in comparison to the 1.2 million building permits issued for housing ____________ 2Households are categorized into four categories for the purposes of this law. Very Low Income refers to households with less than 50 percent of the countywide median household income; Low Income refers to households with greater than 50 percent and less than 80 percent of median income; Moderate Income covers households with income between 80 percent and 120 percent of median income; and Other Households have more than 120 percent of area median income. Countywide median income by household size is determined by the U.S. Department of Housing and Urban Development (HUD). 3Most agencies do not actually build housing themselves but instead provide financing to other organizations to produce housing. 81 statewide4 over that time, it is purported to be a large share of all the affordable housing created in the state. The key complaints by housing critics are that the RDAs do not spend the required 20 percent minimum of tax increment on low- and moderate-income housing, too much of what they do spend is on administrative costs, and not enough of the housing is targeted to the lowest-income groups. These questions are addressed in the following subsections. Do RDAs Spend the Required 20 Percent Minimum of Increment on Low- and Moderate-Income Housing? Statewide tax increments added to low- and moderate-income funds totaled $255 million in 1993–1994, which was 16 percent of total tax increments and almost 19 percent of retained increments (i.e., after passthroughs). Although in the aggregate RDAs set aside only 18.5 percent of their tax increment in housing funds over the past five years,5 the 30 largest agencies set aside just under the 20 percent requirement. Smaller agencies are less likely to meet their required amounts, and some agencies have done little to increase the supply of low-income housing. The aggregate shortfall in the actual amounts set aside in the last five years is only 1.5 percent of the tax increment received statewide; yet this amounts to over $107 million in funds. Agencies can set aside less than 20 percent of their increment if there is no need for affordable housing in the community or if less than 20 percent is sufficient for affordable housing activities. Agencies that were originally grandfathered from the ____________ 4Building permit data are from the Construction Industry Research Board, reported in California Department of Finance (1994), Table I-8. 5California Legislature (1996). 82 20 percent requirement were allowed to defer payments if they had already committed the funds to existing programs, and other agencies are granted exceptions to the rules for various reasons.6 Do RDAs Spend Too Much of Their Housing Funds on Administrative Costs? Figure A.1 summarizes the operating expenditures of RDAs’ housing activities in 1993–1994.7 Planning and administration expenses account for a similar share of housing activities as they do for the rest of RDA operations, so unless agencies are misclassifying spending categories, there is no support for the assertion that RDAs as a group are frittering away housing funds on administration. This is not to say that all agencies are not spending too much on administration—42 agencies spent 100 percent of their money on planning and administration costs and 36 spent over 50 percent.8 In some cases, for example, in new project areas or in projects starting new housing programs, such high levels of planning costs might be reasonable. For example, of the 42 agencies that spent 100 percent of their housing funds on planning and administration in 1993–1994, 12 were still spending more than half of their funds in those categories two years later. In 19 of the 36 agencies that were spending more than half of their funds on administrative costs in 1993–1994, administrative costs were still near or above 50 percent two years later. Although the number of potential abusive cases is much ____________ 6As of July 1, 1996, any grandfathered projects were required to start repaying any deferred amounts (unless they had an exemption from the requirements altogether). This should raise the average share of increment set aside for housing funds in future years. 7Housing data are from the California Housing and Community Development Agency (1996), unless otherwise noted. 8California Legislature (1996). 83 Rental subsidies 10% Other 6% Site improvements 13% Rehabilitation 13% Planning, administration 20% New construction 18% Housing purchase 5% Land purchase 15% SOURCE: California State Controller’s Office (annual). Figure A.1—Operating Expenses for RDA Housing Activities smaller when examining costs over a three-year period, almost 10 percent of the state’s agencies still seem to be consistently spending their housing funds on overhead rather than on the production of low-income housing. Is Enough RDA Housing Targeted to Low-Income Households? RDAs also face inclusionary housing requirements, which stipulate the minimum shares of affordable housing units built or rehabilitated. Of agency-developed housing, at least 30 percent of all new and rehabilitated units must be affordable to low- or moderate-income households, and half of that must be affordable to very low-income 84 households. If housing is not agency-developed, the inclusionary requirement drops to 15 percent. No precise definition of “agencydeveloped” exists, and it is common practice for agencies to claim that new housing is not agency-developed, to meet the smaller requirement.9 Table A.1 shows the number of units produced by the income category of the unit. Of all newly built and rehabilitated units produced by RDAs in 1993–1994, 48 percent were affordable to very low-income households (46 percent of all built units and 50 percent of all rehabilitated units).10 From these figures, it seems that RDAs are easily exceeding the requirements for the production of very low- and lowincome housing units. Another common cause of resistance to redevelopment activity comes over the subject of displaced households. Such concern has its Table A.1 Redevelopment Housing Activities, by Income Category No. of Units: Built Rehabilitated Acquired Subsidized Total Very Low 2,081 2,280 315 1,769 6,445 Low 1,250 1,756 159 807 3,972 Moderate 1,168 488 61 773 2,490 Other 22 15 4 45 86 Total 4,521 4,539 539 3,394 12,993 % Very Low 46 50 58 52 50 SOURCE: California Department of Housing and Community Development (1994). ____________ 9CRA notes that “agency-developed” housing is defined as units that an RDA contracts to build directly, whereas units for which an RDA provides financial assistance are “agency-assisted” housing. The Department of Housing and Community Development has sought clarifying legislation on this issue. 10AB 1290 requires that RDAs that meet these inclusionary requirements with units located outside the project area must do so on a two-for-one basis. 85 roots in the large-scale relocations that occurred during the peak years of the federal slum clearance programs. In 1993–1994, 579 households were displaced by redevelopment actions, of which 130 were outside project areas. Also, 422 net housing units were lost from redevelopment projects, and 697 other units were created to replace units destroyed in earlier years. There was an imbalance in income categories, however— 73 percent of units lost were very low-income whereas 53 percent of units replaced were affordable to low-income and above. Since the units replaced went to offset units lost in earlier years, though, the discrepancy might not represent a real imbalance. Given the scale of both prior relocations and of the overall number of units built and in production, these do not seem like large amounts of displacement overall. 86 Appendix B Sales Tax and Redevelopment After Proposition 13, sales tax became a more important source of revenue for local governments in California. Some redevelopment projects are also oriented toward sales-tax generation, and in 1995–1996, 43 projects received sales-tax revenue. Four have become inactive since 1996, and an additional seven projects that used to receive sales tax became inactive before 1995. Most of these projects received a tax rate of 1 percent and were started between 1986 and 1994. Authority to receive sales taxes was granted to redevelopment agencies soon after Proposition 13 and was revoked by AB 1290 commencing in 1994 (perhaps by coincidence, 17 of the 43 projects were started on July 1, 1994). How important is the sales tax generated within redevelopment projects to the city? Figure B.1 shows the fraction of total city sales taxes that the redevelopment project represents. In most of the cities, the redevelopment projects’ sales taxes are not a large fraction of total sales 87 Ratio 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 AlhamAbnraahBeaiAmldzwusinaBPeaarukmoCnatrsonChinCooltConoronaDEalvMisonFtiellmFoHroeunntatinnagGtoanlKtPinagrksLbaurMgMiraoMdriaelpniotaVsalNleeywPaarlmkPadsaaledenPaPiceorrRisSivaeTnreadmCpiltey CTWiotyrerasnt cCeovina SOURCES: Board of Equalization, 1995 data on redevelopment agencies; California State Controller’s Office (annual). Figure B.1—Redevelopment Projects and City Sales Taxes taxes. In six cities, though, redevelopment agencies bring in almost half of the cities’ sales taxes and, in one case, the agency brought in twothirds. For these cities, redevelopment agencies generate a significant amount of sales tax and demonstrate what can happen when fiscal stress and redevelopment’s wide latitude are combined. A related question is how important tax increment is relative to sales taxes. Some observers think that sales tax is an important element in RDA planning, even for projects not concentrating solely on retail activity. This question is very difficult to answer with available data. In cities with specific sales-tax agreements, the tax increment generated within a project area can be compared with the sales-tax revenue raised in the same area. Data were available for 25 of the 43 project areas that 88 have special sales- and use-tax arrangements.1 The amount of tax increment was less than the sales tax generated within the project in nine cases and was more than triple the sales tax in 11 projects. Since these were projects with specific sales-tax agreements, they can be assumed to be much more sales-tax-oriented than the average project. Given that, it seems safe to say that tax increment is a more important determinant of RDA incentives than sales-tax revenues are. ____________ 1In seven of the other cases, the sales tax went to specific businesses instead of to the projects. 89 Appendix C Detailed Discussion of Study Methodology Matching Algorithm The matching method used in this study is similar to the logic behind regression, i.e., minimizing the sum of squared differences (weighted by the variance of the potential matches). This number is related to the Mahalanobis distance, which has been frequently used in regional economic studies of this type; the difference is that the Mahalanobis distance is weighted by the variance-covariance matrix rather than just the variance. For a more involved discussion of this technique, see Rephann and Isserman (1994). If there are three Census variables to be matched on (say, X, Y, and Z), then the matching Block Groups are those in each city that minimize the expression 91 ∑i n =1   (X RDA − X nonRDA σ 2X ,nonRDA )2 + (YRDA − YnonRDA )2 σ 2Y ,nonRDA + (ZRDA − ZnonRDA )2 σ2Z,nonRDA   , where n = all non-RDA Block Groups in a given city. Conceptually, this approach selects the Block Group that minimizes the distance in XYZ space between each project area and its control. There is a tradeoff between using the same matching variables for every city and choosing the variables that match blight conditions for each city individually. The methodology used here follows generic matching. Using the same variables for matching across all the projects allows for better generalization of the results to those projects in the original sample that were lost because historical data were unavailable. Loss of Project Areas from the Beginning Sample After dropping counties with fewer than three projects that began between 1978 and 1982, the sample had 114 projects located in ten counties. Assessed value data for 1983 could be obtained for only four counties, which together had 67 projects that began between 1978 and 1982. In 25 cases, the projects had to be dropped because of circumstances that caused problems in using the matched-pairs methodology. At this point, some unresolved data issues precluded the use of the four Solano County projects. These data problems—along with the losses outlined below—left us with the final sample of 38 projects in the counties of Los Angeles, San Bernardino, and San Mateo. Five projects were dropped because we were unable to get maps of the project areas in the respective cities; this made it impossible to eliminate other project areas when searching for possible matches. Two projects were dropped because they consisted of a single parcel each. 92 Three projects were lost because the project area covered at least part of every Census Block Group in the city. Ten additional projects were lost because every Census Block Group was covered by either the sample project area or other project areas in the same city. Two projects were dropped because the closest Census Block Group selected by the matching algorithm was too dissimilar to the project area (i.e., the function value was much larger than the range in all the other pairs examined). One project was dropped because of large inconsistencies between assessor’s office and controller’s office data. Last, two projects had to be dropped at the very end because even after consulting historical parcel map books, we were unable to match enough parcels to achieve the required thresholds of 90 percent of total parcels and total 1996 assessed value. Imputing Assessed Value for Changed Parcels As discussed in Chapter 4, the parcels used in the analysis were first selected from the 1996 secured roll for the entire state, as collected by Dataquick from county assessors’ offices and tax collectors (and other private sources). Project parcels were identified by the tax rate area codes that uniquely identify all the parcels within a project area.1 A file that linked current tax rate area codes and the jurisdictions receiving taxes from each tax rate area was purchased from the State Board of Equalization. Parcels in the matched areas were identified by their ____________ 1A redevelopment project area can contain multiple tax rate areas; however, a tax rate area is not split between a project area and the outlying area. In other words, a project area will contain whole tax rate areas, each with a unique listing of parcel numbers that constitute the base on which the tax increment is calculated. When a new tax rate area is being created where a project area is currently located, two tax rate areas in effect are created (one inside the project area, one outside) so as to not split the tax rate area. Tax rate areas can be prolific. Los Angeles County has roughly 10,000 tax rate areas. 93 Census Block Group. This set of parcel numbers formed the basic file for the rest of the study. Once the parcel numbers were identified, the 1983 microfiche was searched for those parcel numbers. If present, the 1983 assessed value and use code was appended to the 1996 record. Almost 16 percent of the parcel numbers could not be found in the 1983 microfiche. In the project areas, 34 percent of the parcels could not be found because the parcel had changed; for the matches, the figure was 12 percent. To resolve these missing cases, trips were made to each county’s assessor’s office to track the parcel changes. Parcel map books were consulted to find which 1983 parcel(s) each 1996 parcel was derived from. In 38 of the 42 project-match pairs, we were able to identify 1983 assessed values for more than 90 percent of the parcels and 90 percent of the assessed value in the 1996 database for either the project area or the matched area.2 The pairs for which we could not get such a complete accounting were dropped from the analysis (but the final results are not affected much whether those pairs are included or not). For the rest of the pairs, the corresponding 1983 parcel was categorized by the type of change that occurred between 1983 and 1996. If the change was simply from one parcel number to another, the 1983 assessed value was added to the 1996 record. If one 1996 parcel came from several 1983 parcels, the 1983 values were summed to create the correct 1983 value. If several 1996 parcels came from a single 1983 parcel, the 1983 value was evenly apportioned to all the succeeding 1996 parcels. In those cases where several 1983 parcels became several 1996 parcels, the total values were summed in each year and apportioned ____________ 2The thresholds for deciding which projects had sufficient data to be used were determined before examining the parcel-level breakdowns. 94 equally among all the 1996 parcels. Table C.1 shows the distribution of the sample parcels by type of changes, if any, between 1983 and 1996. There was quite a bit of variation between counties (and between projects) in the extent of parcel changes in the last 13 years. In each county, a greater share of parcels changed in project areas than in matches—although this was not universally true among the projectmatch pairs. What was universally true was how rare it was for numerous parcels to be assembled into one larger parcel: Only in the projects in Los Angeles County did this occur in over one-half of one percent of the parcels. It was much more likely that one large parcel was subdivided into many smaller parcels, or that several large parcels were combined and then split into many parcels. Overall, we were not able to identify 3 percent of the parcels in the 1983 microfiche, and another 1 percent were tax-exempt parcels for which there was no assessed value recorded in the 1983 secured-roll microfiche. Table C.1 Breakdown of Parcel Changes, 1983–1996, by County (in percent) Missing, Missing, No Many: 1: Many: Nongov- Govern- Change 1:1 1 Many Many ernment ment Los Angeles Project 70 3 1 9 10 5 2 Match 88 20 42 2 2 San Bernardino Project 54 5 0 10 27 3 1 Match 88 10 54 1 0 San Mateo Project 68 2 0 18 9 2 1 Match 95 10 40 0 0 Total Project Match 66 89 3 1 11 13 4 10 53 2 2 1 SOURCES: Dataquick property database (1996); individual county assessor’s office and/or tax collector data (1983). 95 Calculation of RDAs’ Share of Assessed Value Growth and Size of Tax Transfers RDAs receive property taxes based on the increase in assessed value for parcels within the project area, less the 2 percent annual inflation factor as decreed by Proposition 13. The tax increment revenue they receive in FY 1995–1996 is calculated as the change in aggregate assessed value (minus the compounded inflation factor), times the property tax rate, multiplied by the share of tax increment the RDA retains. Tax increment received = (1 − pass-through) * (tax rate) * (AV96 − (1 − .0213) * AV83). AVnn is the total assessed value in all the parcels in the project area in year nn. The pass-through rates for each project come from the annual financial self-reports made by each RDA to the state controller’s office. Because these figures are subject to error and annual fluctuations, the pass-through rate used in these calculations is the average reported by each project area over the last 11 years.3 To calculate how much tax increment the project generated itself, the growth rate in the matched area rather than the 2 percent inflation factor is subtracted from the change in aggregate assessed value. The growth rate in the matched area is calculated as the growth in assessed value between 1996 and 1983, divided by the assessed value in 1983. The inflation factor is not subtracted in this case because in nonRDA areas that amount is also shared among all the taxing jurisdictions. ____________ 3The state controller’s office annual reports on redevelopment agencies began in FY 1985, and the latest year published was FY 1995, so only 11 years of data were available. This is also why there is a one-year difference between the assessed value data used in the calculations (FY 1996) and the dollar amounts used from the controller’s report (FY 1995). 96 Growth rate in matched area (Rm) = (AV96 − AV83) . AV 83 The growth generated by each project is defined as the total growth in assessed value in the project area minus the growth rate of the matched area. This growth is multiplied by the tax rate to get the amount of tax increment revenue generated in the project area. Tax increment generated = (tax rate) * (AV96 − (1 + Rm ) * AV83); RDA share of tax increment = tax increment generated ; tax increment received Estimated tax transfer = RDA share of tax increment * net FY 1996 tax increment . The share of the tax increment revenues each project generated is simply the ratio of the tax increment the project generated to the amount it received. These figures were calculated using the parcels and their values as described above. The comparison was therefore made between consistent groups of parcels in 1983 and 1996. To estimate the actual amount of transfers that occurred in FY 1995, though, the calculated ratio was multiplied not by our estimate of the tax increment received, but rather by the amount of tax increment reported in the FY 1994– 1995 controller’s report. On average, our calculation of the 1996 total assessed value in the project areas was 77 percent of the value reported in the controller’s report. Approximately 8–10 percent of this difference results because the RDAs report both the secured and unsecured rolls, whereas this study used only the secured rolls. Another 5 percent of the difference results because we were able to obtain the historical assessed value only for 97 parcels totaling 95 percent of the total 1996 assessed value in the project areas. The remaining difference remains unexplained. Conversations with officials in the controller’s office indicate that the annual figures in RDAs’ self-reports are not without errors, and that a discrepancy in the totals on the order of 7–10 percent is not cause for concern. For example, the reported increment assessed value (what we would call AV95–AV85) was 18 percent larger than the increase in total assessed value between the two years. Some of this discrepancy may be due to mergers that increased the size of project areas, but some may also be due to the inaccuracies cited by the controller’s office. 98 Appendix D Regression Analysis With a relatively small number of projects in the final sample, regression analysis can demonstrate that the differences in growth rates are nevertheless statistically significant. It also allows the two matching variables to be controlled for. The matched-pairs analysis can be replicated in a regression format by using each area—project or match— as an observation, with 38 dummy variables to control for the pairs and a separate dummy variable for whether the observation is a project or a matched area. The regression can also be run with poverty and vacancy rates as regressors, to check whether the observable differences between the projects and their matches change the estimated size of the RDA effect. These regressions are not meant as a complete analysis of these phenomena; they are provided to demonstrate that the differences in growth rates between projects and matches are statistically significant, 99 despite the small sample size. Further econometric work in a forthcoming working paper will examine the factors—including parcel turnover and type of expenditures—that affect project areas’ growth in more detail. The equation to be estimated is: ln(AVgrowth) = βRDADRDA + βvac VAC + βpovPOV + β1D1 + ... + β38D38 + e , where ln(AVgrowth) is the logarithm of the percentage change in assessed value from 1983 to 1996, βRDA is the estimated effect of the RDA on assessed value growth in the project area ( DRDA takes the value of 1 if the area is a project and 0 if the area is a match), VAC is the vacancy rate according to the 1980 Census, and POV is the poverty rate. Each area dummy takes the value of 1 for that project area and its match, and 0 for all other project/match pairs. This retains the information contained in the pairing and allows the coefficient on DRDA to estimate the redevelopment growth effect across all pairs. Table D.1 shows the results from this regression. It also reports regression results when vacancy rate and poverty rate are included, with and without the controls for matched pairs. When controlling for matching without vacancy and poverty rates, project areas are estimated to grow 42 percent faster than their matched areas; when poverty and vacancy rates are controlled for as well, the RDA effect increases to 45 percent faster.1 The estimated RDA effect is only slighter lower when matching is ignored. The estimated RDA effect from these equations is similar to that obtained (41 percent) from the simple comparison of mean growth rates ____________ 1With the dependent variable in logarithmic form, the elasticity is interpreted as e β − 1. In this case, the coefficient is βRDA , called the “RDA effect” in Table D.1. 100 Table D.1 Estimation of RDAs’ Effect on AV Growth [Dependent Variable: log(AV growth)] Equation 1 Equation 2 Equation 3 Equation 4 RDA effect Poverty rate Vacancy rate Pair dummies Adjusted R2 No. of observations 0.335 (2.56) — — — 0.069 76 0.305 (2.36) –0.116 (–0.158) 1.437 (2.02) — 0.096 76 0.335 (2.57) — — Included 0.453 76 0.369 (2.65) 2.205 (0.72) –2.854 (–0.879) Included 0.435 76 NOTE: Equations with pair dummies were estimated without a constant term. T-statistics are reported in parentheses. between project and match parcels. With or without poverty and vacancy rate controls, adding controls for the matched pairs increases the predictive power of the estimation significantly (note adjusted R2 values). Parcel-level regression results (not reported here) confirm the presence and approximate size of the project effect; they also demonstrate that the largest predictor of the difference in assessed value growth between the project and matched areas is the percentage improvement in the project area’s parcels in 1983. This is consistent with the observation made in Chapter 5 that the projects that generated the largest share of their tax increment had much more vacant land in their areas than the other projects did. Regressions that were run separately for projects and matches also indicate that the initial percentage improvement is a stronger predictor of growth in the project areas than it is in the matched areas. 101 Appendix E Final Sample Projects County Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Table E.1 Final Sample Projects Redevelopment Agency Azusa Baldwin Park Baldwin Park Downey El Monte El Monte Huntington Park Inglewood La Verne Los Angeles Los Angeles Los Angeles Los Angeles Maywood Maywood Montebello Pasadena Redevelopment Project West End Merged Project Areas Central Business District Project Area West Ramona Boulevard Project Area Downey Project Area El Monte Plaza Project Area Plaza El Monte Project Area North Project Area Century Project Area Project Area 1 Adams Normandie Project Area Chinatown Project Area North Hollywood Project Area Rodeo-La Cienega Project Area #2 Westside Project Area Economic Revitalization Project Area Lake Washington Project Area 103 County Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Mateo San Mateo San Mateo San Mateo San Mateo San Mateo Redevelopment Agency Pomona Pomona Pomona Pomona Pomona Pomona Torrance Whittier Fontana Ontario Ontario Ontario Rialto San Bernardino Victorville Belmont Foster City Menlo Park Redwood City San Mateo San Mateo Redevelopment Project Arrow-Towne Project Area Holt Ave/Indian Hill Project Area Mission/Corona Business Center Project Area Reservoir Street Industrial Project Area Southwest Pomona Project Area West Holt Project Area Downtown Project Area Whittier Boulevard Project Area North Fontana Project Area Project Area No. 1 Project Area No. 2 Cimarron Project Area Project Area A Northwest Project Area Bear Valley Road Project Area Los Castanos Project Area Foster City Project Area Las Pulgas Community Development Project Area No. 2 Project Area Downtown District Project Area Shoreline District Project Area 104 Bibliography Anderson, John E., “Tax Increment Financing: Municipal Adoption and Growth,” National Tax Journal, Vol. XLIII, No. 2, 1990, pp. 155– 163. Beatty, David F., et al., Redevelopment in California, 2nd Edition, Solano Press Books, Point Arena, California, 1995. Boarnet, M. G., and W. T. Bogart, “Enterprise Zones and Employment: Evidence from New Jersey,” Journal of Urban Economics, Vol. 40, 1996, pp. 198–215. California Budget Project, State and Local Government Finance in California: A Primer, Sacramento, California, 1996. California Debt Advisory Commission, Recommended Practices for California Redevelopment Agencies, Report CDAC-5, Sacramento, California, 1995. California Department of Finance, California Statistical Abstract, 1994, Sacramento, California, 1994. California Department of Finance, California Statistical Abstract, 1995, Sacramento, California, 1995. California Housing and Community Development Agency, Redevelopment Housing Activities in California, 1991–92 to 1995–96, Sacramento, California, annual. 105 California Legislative Analyst’s Office, Redevelopment After Reform: A Preliminary Look, Sacramento, California, 1994. California Legislature, Senate Committee on Housing and Land Use and Senate Select Committee on Redevelopment, Redevelopment and Blight, Report 831-S, Sacramento, California, 1995. California Legislature, Senate Committee on Housing and Land Use, End or Means? Redevelopment Agencies’ Housing Programs, Report 888-S, Sacramento, California, November 13, 1996. California Legislature, Senate Committee on Local Government, Redeveloping California: Finding the Agenda for the 1990’s, Report 457-S, Sacramento, California, 1989. California Redevelopment Association, 1996 Introduction to Redevelopment, Sacramento, California, 1996. California State Auditor, Statewide Redevelopment Agencies, Sacramento, California, 1996. California State Controller’s Office, Financial Transactions Concerning Community Redevelopment Agencies, Fiscal Years 1984–85 to 1993–94, Sacramento, California, annual. California Trade and Commerce Agency, State Auditor, The Effectiveness of the Employment and Economic Incentive and Enterprise Zone Programs Cannot Be Determined, Sacramento, California, November 1995. Chapman, Jeff, “Tax-Increment Finance as a Tool of Redevelopment,” in Helen Ladd (ed.), Local Government Tax and Land Use Policy, Edward Elgar Press, 1998 (forthcoming). Davis, Don, “Tax Increment Financing,” Public Budgeting and Finance, Spring 1989, pp. 63–73. Dowall, David E., et al., “Evaluation of California’s Enterprise Zone and Employment and Economic Incentive Programs,” Institute of Urban and Regional Development, University of California at Berkeley, Working Paper 614, 1994. Dye, Richard F., “A Comparative Analysis of Tax Increment Financing in Northeastern Illinois”; Executive Summary from Assessing the Impact of Tax Increment Financing in Northeastern Illinois: Empirical 106 Analysis and Case Studies, The Civic Federation, Chicago, Illinois, 1997. Frieden, Bernard J., and Lynne B. Sagalyn, Downtown, Inc.: How America Rebuilds Cities, MIT Press, Cambridge, Massachusetts, 1989. Huddleston, Jack R., “Intrametropolitan Financial Flows Under Tax Increment Financing,” Policy Sciences, Vol. 19, 1986, pp. 143–161. Huddleston, Jack R., “Tax Increment as a State Development Policy,” Growth and Change, Vol. 15, No. 2, 1984, pp. 11–17. Huddleston, Jack R., “Variations in Development Subsidies under Tax Increment Financing,” Land Economics, Vol. 57, No. 3, 1981, pp. 373–384. Isserman, A., and T. Rephann, “The Economic Effects of the Appalachian Regional Commission,” APA Journal, Summer 1995, pp. 345–364. Klemanski, John S., “Using Tax Increment Financing for Urban Redevelopment Projects,” Economic Development Quarterly, Vol. 4, No. 1, 1990, pp. 23–28. Papke, Leslie E., “Tax Policy and Urban Development: Evidence from the Indiana Enterprise Zone Program,” Journal of Public Economics, Vol. 54, 1994, pp. 37–49. Rephann, T., and A. Isserman, “New Highways as Economic Development Tools: An Evaluation Using Quasi-Experimental Matching Methods,” Regional Science and Urban Economics, Vol. 24, 1994, pp. 723–751. Stauber, R., and M. Wyatt, “Property Tax Relief and Tax Incremental Financing,” in Robert H. Haveman and Jack Huddleston (eds.), Dollars and Sense: Policy Choices and the Wisconsin Budget, Robert M. LaFollette School of Public Affairs, University of Wisconsin-Madison, 1990. Stinson, Thomas F., “Subsidizing Local Economic Development Through Tax Increment Financing,” Policy Studies Journal, Vol. 20, No. 2, 1991, pp. 241–248. U.S. Bureau of the Census, County and City Data Book 1994, Washington, D.C., 1994. 107 About the Author MICHAEL DARDIA Michael Dardia is a research fellow at the Public Policy Institute of California. His research interests include redevelopment, regional economics, and industrial restructuring. In previous work, he has examined defense conversion and the effects of defense cuts on the California economy. He is the author of several publications on defense spending and the aerospace industry. Dardia holds a B.S. in biological sciences from the State University of New York at Stony Brook, an M.S. in policy analysis and management from the university’s W. Averell Harriman College of Management and Policy Analysis, and a Ph.D. in public policy analysis from the RAND Graduate School of Policy Studies. 109" ["post_date_gmt"]=> string(19) "2017-05-20 09:34:48" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_298mdr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:34:48" ["post_modified_gmt"]=> string(19) "2017-05-20 09:34:48" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_298MDR.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) }