Local Redistribution and Local Democracy: Interest Groups and the Courts 9780300171822

Citation preview

LOCAL REDISTRIBUTION AND LOCAL DEMOCRACY

This page intentionally left blank

clayton p. gillette

Local Redistribution and Local Democracy interest groups and the courts

new haven and london

Published with assistance from the Louis Stern Memorial Fund. Copyright © 2011 by Yale University. All rights reserved. This book may not be reproduced, in whole or in part, including illustrations, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers. Yale University Press books may be purchased in quantity for educational, business, or promotional use. For information, please e-mail [email protected] (U.S. office) or [email protected] (U.K. office). Set in Scala and Scala Sans type by IDS Infotech Ltd., Chandigarh, India. Printed in the United States of America. Library of Congress Cataloging-in-Publication Data Gillette, Clayton P. Local redistribution and local democracy : interest groups and the courts / Clayton P. Gillette. p. cm. Includes bibliographical references and index. ISBN 978–0–300–12565–8 (cloth : alk. paper) 1. Distribution (Economic theory) 2. Local government—United States 3. Local finance–United States. I. Title. HB523.G56 2011 339.20973–dc22 2010045038 A catalogue record for this book is available from the British Library. This paper meets the requirements of ANSI/NISO Z39.48–1992 (Permanence of Paper). 10 9 8 7 6 5 4 3 2 1

For Jordan

This page intentionally left blank

CONTENTS

preface

ix

1

The Scope of Local Competence 1

2

Why Does Local Autonomy Matter? 31

3

The Meaning and Scope of Local Redistribution 53

4

Why Do Local Governments Redistribute? 72

5

Local Political Markets and State Constitutional Constraints 106

6

Proxies for Distinguishing Benign and Malign Redistribution 142

7

The Limits of Judicial Intervention 174

8

Conclusion 196

notes index

201 225 vii

This page intentionally left blank

PREFACE

This book attempts to merge various but distinct literatures relating to local governments: urban finance theory, public choice theory, and legal theory. Urban finance theory provides us with an ideal of what local governments should seek to accomplish. In one particular conception, it tells us the conditions under which decentralized governments can efficiently provide local public goods and thus gives us a basis for comparing the capabilities of local governments, centralized governments, and markets. Urban finance theory does not necessarily recommend one form of organization over another. There may be reasons to allocate responsibilities to more centralized governments or nongovernmental institutions, notwithstanding the inefficiencies that those allocations engender. Urban finance theory, at least in the form in which I am interested, is silent on those issues. Public choice theory, however, gives us pause about the teachings of urban finance theory. It tells us that those who are charged with implementing the strategies that allow local governments to achieve what theory tells us they can most efficiently accomplish may deviate from the interests of their constituents. That is, local officials may be unfaithful agents who pursue some personal objective that diverges from pursuit of the public welfare with which they are entrusted. They are aided in this personal effort by collective action problems that induce some parties to ix

x

pr ef ace

be inattentive to local projects that are not worth their costs, while other parties—often a distinct but well-organized minority—invest heavily in attaining those same projects. The problem of special interests is by no means unique to local government. But local governments may be more vulnerable to the difficulties of special interests for reasons that go back at least to James Madison’s observations that decentralized governments are less likely to comprise offsetting factions. Law can mediate between the teachings of urban finance theory and public choice theory. Law has a positive aspect insofar as it creates rules that allocate responsibilities to certain levels of government. It, therefore, can embody urban finance theory by allocating to localities those functions that are most efficiently delivered at decentralized levels of government. Legal allocations can also implement policies that officials believe should trump efficient provision, such as equitable considerations that may be indicative of a more communitarian vision of local life. But law can also incorporate the lessons of public choice theory and constrain the discretion of officials where circumstances increase the likelihood that officials' conduct will diverge from public welfare. Obviously law can do that in cases of blatant divergence, such as bribery. But legal doctrine may also identify the more subtle circumstances in which local officials can benefit by attending to those most invested in a particular project, notwithstanding that the project fails to serve local interests generally. Law could then raise the costs of official defalcation, either by negating local authority to act autonomously in those circumstances, or by exacting a substantial cost for local action, such as requiring evidence that a particular project comports with local interests. In doing so, law explicitly assigns decision-making responsibility to institutions that are most likely to exercise authority in a manner that reflects the interests of local constituents. The ability of law to perform this function, however, is complicated by a practical difficulty. Law can proscribe local action in a broad set of circumstances in which deviation from local interests is plausible. But whether any particular project that arguably fits within that description in fact deviates from local interests will likely be contestable. Failure to investigate more deeply the fit between the conditions of divergence from local interests and individual cases risks either weakening the legal threat (if individual cases are presumed to pass muster) or invalidating

p ref ace

xi

projects that would generate significant local benefit (if individual cases are presumed to be invalid). The task of investigating that fit typically falls to courts. But courts suffer from their own issues of institutional legitimacy and competence. Assigning to courts the task of making the necessary individualized determinations essentially asks them to understand the theories of urban finance and public choice that lead to the legal interventions that judges must implement. That raises questions about both the scope of judicial authority and the accuracy with which they can execute the authority they have. These issues coincide in the puzzle about local government that is central to this book. What I consider here to be the orthodox theory of urban finance predicts that local governments will not engage in the redistribution of wealth. Any effort to do so allegedly will induce those who are net subsidizers of redistribution to exit and will attract those who can benefit from redistribution. Thus, a locality that attempts to redistribute wealth will soon discover that it has lost tax base that it needs to provide services to residents and that it has gained a significant population that cannot pay the tax price that corresponds to the locally provided benefits that the population consumes. Notwithstanding that very basic assumption of urban finance, virtually every major locality engages in the redistribution of wealth. The objective of this book is to explain that apparent contradiction and to indicate the role that law can play in resolving it. Law assumes that role because, as one achieves a richer understanding of local motives in redistributing wealth, the resolution may lie in either a relatively happy story in which the apparent contradiction disappears, or a relatively unhappy story in which, consistent with public choice theory, the contradiction is attributed to divergence between the interests of local residents and of the officials who are charged with serving those interests. Law and legal institutions play a role in distinguishing between these explanations. It is possible, of course, indeed likely, that each of these explanations accounts for some part of the local redistribution puzzle. That is where the practical issues related to law step in. Ideally, law might be able to instantiate the richer understandings of local motives, so that localities that have benevolent reasons for redistribution, emanating from local self-interest or altruism, would be authorized to act, while those that redistribute for more nefarious motives, such as serving the personal goals of officials,

xii

pr eface

would be constrained. Can law effectively play that role? That is, can legal institutions, primarily courts that would be asked to challenge redistributive programs, isolate those situations in which redistribution is inappropriate, while validating those situations that serve local welfare or local preferences? These inquiries motivate this book. I am, therefore, interested in issues of institutional design. Urban finance theory tells us that one level of government rather than another is best situated to serve particular functions. Public choice theory tells us that different institutions may suffer different degrees of agency costs and that one way to control agency costs related to one institution is to subject its decisions to review by an alternative institution. Legal theory allows us to develop rules that optimize decision-making responsibilities and define the circumstances and design of reviewing institutions. This book draws on much of my prior work. Some of the basic ideas were set forth in an article, “Local Redistribution, Living Wage Ordinances, and Judicial Intervention,” published in the Northwestern University Law Review in 2007. I am grateful to the Law Review for publishing that piece and for recognizing that it would subsequently be the subject of further development. I am also grateful to a variety of colleagues who discussed many of the ideas that appear in this book, though none of them should be tarred with the accusation that they agree with me. They include Ed Baker, Lynn Baker, David Barron, Vicki Been, Richard Briffault, Helen Hershkoff, Rick Hills, Marcel Kahan, Lewis Kornhauser, Daryl Levinson, Gerald Lopez, Rick Pildes, Richard Revesz, Julie Roin, Richard Schragger, Dan Shaviro, and participants in workshops at New York University School of Law, the University of Florida College of Law, and Florida State University College of Law. Student assistants at NYU School of Law were also instrumental in gathering some of the material on which I relied. Ben Holzer of the NYU School of Law Class of 2009 was particularly helpful, especially with respect to the history of municipal poor relief. Dean Richard Revesz generously permitted me a sabbatical to put many of these thoughts together. The Filomen D’Agostino and Max E. Greenberg Research Fund at NYU School of Law provided generous research support. This book would not have come to fruition without the encouragement and support of Michael O’Malley of Yale University Press. Ann-Marie Imbornoni’s close reading and edit of the manuscript caused me to correct and clarify

p ref ace

xiii

much of the text. Of course, my greatest gratitude is reserved for Abby Gillette for bearing all the standard obligations and sacrifices that we impose on those to whom we are closest, and then for so much more. Prior to writing this article I served as a consultant to opponents of some of the living wage ordinances mentioned in this article and was a signatory on an amicus brief supporting another living wage ordinance. Perhaps those seemingly inconsistent roles reflect the complexity of the analysis that I discuss herein.

This page intentionally left blank

ch ap t e r o ne

The Scope of Local Competence

in july 2006, the Chicago City Council brought what it likely thought was a climactic conclusion to a controversial proposal that required a few local employers to pay their workers a “living wage.” The 35 to 14 vote increased the minimum wage payable by covered employers to $10 per hour, plus $3 in benefits per hour, by July 2010. The new wage represented a substantial increase over the then-applicable federal minimum of $5.15 per hour and even exceeded the Illinois statewide minimum wage, which was scheduled to reach only $8.25 by the time the ordinance became effective. Primary proponents of the ordinance—including labor organizers; the Grassroots Collaborative, which consisted of local church and labor leaders; and ACORN, a national organizing group—maintained that its enactment would remedy exploitation of the working poor, reverse what advocates labeled as “callous disregard for hard-working people,”1 and fetter corporations whose workplace practices allegedly frustrated employees’ effort to enter the ranks of the middle class. But the city council’s effort to improve the lot of local employees had limited scope. The ordinance would apply only to those employers described in the ordinance as “large retailers.” That term comprised retail stores in the city that occupied ninety thousand square feet or more, and that were part of a company that had total annual gross revenues of $1 billion or more. In short, the ordinance was directed at “big-box” 1

2

t he sco p e o f lo c a l c o m pe te n c e

stores, for which Wal-Mart and Home Depot have served as poster children. While opponents of the proposal contended that living wage ordinances were inherently antibusiness and would cause Chicago to lose badly needed jobs and tax revenue to suburbs more hospitable to national retailers, at least some opponents also questioned what they saw as discrimination against big-box employers and promised legal challenges once the ordinance passed. The immediate targets of the proposal strategically remained on the sidelines during much of the debate, although Wal-Mart had recently indicated an interest in opening a new store on Chicago’s relatively poor South Side. Instead, local proxies ensured that the chains’ views were heard. Local suppliers and small businesses that performed work for the big-box stores testified against the ordinance. After the vote approving the ordinance, however, Wal-Mart officials themselves became more vocal. They indicated that the wage ordinance would “make it hard to invest in Chicago”2 and announced a strategy of serving Chicago residents “from suburban Chicagoland.”3 Target Corp. was reported to have placed “on hold” plans to open additional stores in the city and to be considering closure of its existing stores.4 Nevertheless, the proponents celebrated the vote as a major victory in nationwide efforts to enact living wage ordinances. While some form of minimum wage requirement had been enacted in more than 120 cities and counties, most were smaller localities, and most of the ordinances extended the applicable wage only to municipal employees or employees of suppliers that contracted directly with the locality. An ordinance in a major city that applied to private employers promised to change the terms of the debate. Alas, the proponents’ celebration was short-lived. In mid-September, Mayor Richard M. Daley exercised his veto authority for the first time in his seventeen years in office. In his brief letter of explanation to the city council, Daley contended that the ordinance “would drive jobs and businesses from our city, penalizing neighborhoods that need additional economic activity the most.” Daley concluded that it was his “duty” to veto the ordinance.5 Daley’s response was not unexpected. He had earlier announced his opposition to the ordinance and had lobbied against its adoption. But proponents had declared the margin of victory “veto-proof.” They had

th e s c o pe o f l o cal co mp et ence

3

miscalculated. Largely at Daley’s instigation, three aldermen who had initially supported the measure voted against overriding the veto. One of the converts represented a district on the South Side where Wal-Mart had indicated it might locate its first Chicago store; the other two had close ties to the Daley political organization. Two weeks after the veto, Wal-Mart opened its first Chicago store, located on the relatively poor West Side of the city. The living wage issue did not disappear with the mayor’s victory. In the following year’s city elections, the Chicago Federation of Labor failed to endorse the incumbent mayor for the first time in its history. Instead, organized labor worked to defeat the aldermen who had voted against the measure. In February 2007, eleven incumbent aldermen failed to obtain the majority vote necessary to avoid a runoff election in their districts. Many cited their failure to support the big-box ordinance as a cause of their electoral difficulties. In the subsequent elections, a half dozen aldermen aligned with the mayor on the living wage ordinance were defeated by union-backed candidates. The signal was clear: the living wage ordinance was on the exam in the course of local politics, and those who opposed it had failed. The intensity of the Chicago debate was not unprecedented. In 2001, the city council in Santa Monica, California, had passed a living wage ordinance that required covered employers to pay employees $10.50 per hour plus benefits or $12.25 per hour without benefits. Although the Santa Monica ordinance was not as narrowly drawn as Chicago’s, it did not target all local employers. It applied to the city with respect to its own employees, to any contractor or subcontractor with respect to workers on projects performed under contract with the city, and to any private employer whose business met a gross receipts test and was located in an area of the city designated as the Coastal Zone or Extended Downtown Core. The restrictions on gross receipts and location limited the private employers who were subject to the ordinance primarily to hotels and restaurants along the Santa Monica beachfront. The ordinance also contained a clause that rendered it inapplicable to wage rates established through a collective bargaining agreement, as long as that agreement specified that it was intended to supersede the wages set forth in the ordinance. In heated exchanges before the Santa Monica City Council, proponents of the ordinance defended the need to pay a living wage to

4

t he s co p e o f lo c a l c o m pe te n c e

low-wage workers, many of whom were immigrants and nonresidents of the city. Local business owners contended that the ordinance would cause them to reduce their workforce, thus hurting the very class of individuals proponents purported to assist. Labor activists and union members featured prominently among the advocates of the ordinance, while the affected hotel owners were instrumental in organizing the opposition. After the city council enacted the law, many of the approximately forty firms subject to its private business provision organized a referendum campaign to place the issue in front of the electorate. In the ensuing November 2002 plebiscite, a narrow majority—14,380 to 13,860—rejected the ordinance.6 There the story remained for about two and a half years. Then, in early 2005, the city council enacted another living wage ordinance. This version, however, covered only employees of private contractors who entered into contracts of $50,000 or more with the city itself. Neither city employees nor employees of the private firms that had opposed the earlier ordinance were subject to the new proposal. The modified ordinance became effective on July 1, 2005. Santa Fe, New Mexico, witnessed a similarly contentious path for a living wage ordinance but experienced a different result. In 2003, the city council considered expanding an existing ordinance that required payment of higher than otherwise applicable minimum wages to city employees, contractors doing business with the city, and other businesses directly receiving benefits from the city. As initially introduced, the proposed ordinance both increased the applicable minimum wage and extended the scope of the ordinance to cover all for-profit employers in the city that had more than ten employees and nonprofit businesses that employed more than twenty-five workers. The proposed wage would start at $10.50 per hour in 2008, with subsequent increases tied to the consumer price index for urban wage earners in the western United States. Approximately 75 percent of all employees and 20 percent of all businesses in Santa Fe would have been subject to that version of the ordinance. At a public hearing that lasted into the early morning hours of February 27, economists, members of the Green Party of Santa Fe, labor activists, advocates for the working poor, and workers faced off against small-business owners, representatives of the New Mexico Restaurant Association, officers of nonprofits, and hotel managers. Seventy-three

th e s c o pe o f l o cal co mp et ence

5

people spoke in favor of the ordinance, eighty-five against it. In emotional terms that touched on the political, economic, and social implications of the issue, advocates emphasized the low costs that the ordinance entailed, the moral issues involved in ensuring payment of a living wage, and the economic distress of the working poor. In similarly evocative terms, opponents stressed the risk of shuttered small businesses, the difficulty of finding workers whose productivity warranted higher-thanmarket wages, and the disincentives that the ordinance would create for businesses to locate in Santa Fe.7 City councilors enacted the ordinance by a 7 to 1 vote, but only after it had been narrowed to cover employers with twenty-five or more workers. The amendments reduced the coverage of the ordinance to about 60 percent of the Santa Fe workforce and cut in half the number of small businesses that would be affected. Enactment of the Santa Fe ordinance generated the inevitable lawsuit. The challengers contended that Santa Fe lacked legal authority to adopt a local minimum wage. Under well-established, if controversial, legal principles, local governments lack the capacity to initiate local legislation without prior legislative or state constitutional authority. Local governments are, in the standard language of legal doctrine, merely political subdivisions of the state, which exercises plenary power over them. As a corollary of these principles of limited municipal autonomy, any conflicts between state laws and local ordinances are typically resolved in favor of the former. The challengers contended that Santa Fe’s ordinance exceeded the permissible scope of local legislation. Both the trial court and the court of appeals dismissed the objections.8 Santa Fe was a “home rule” municipality, which, under the state constitution, enabled it to enact legislation relating to “local affairs” without a prior explicit grant of authority from the state. The living wage ordinance, the courts concluded, fell within that protected domain. Even a home rule locality is generally disabled from enacting legislation that directly conflicts with state laws. But, the courts declared, the state’s minimum wage law, which imposed a lower hourly rate than the one dictated by Santa Fe, was intended only to set a floor, not a ceiling. Thus, the local ordinance was neither inconsistent with nor superseded by state law. As evidenced by the broad array of cities and counties that have adopted living wage ordinances, the debates in Chicago, Santa Monica, and Santa Fe do not represent exceptional episodes. The campaigns in

6

the s co p e o f lo c a l c o m pe te n c e

these municipalities, however, do demonstrate the passion and politics that local proposals for redistributing wealth can generate. Embodied in these controversies are deeply varying ideologies about the economic effects of wage rates, governmental intervention into labor markets, the widening income disparity between wealthy and poor, and the proper scope of local government. Something less noticed but of equal importance is that the controversies over living wage ordinances reveal the various political forums in which decisions about controversial proposals for local action can occur. While organized proponents may effectively lobby for enactment of the proposals before local legislatures, organized opponents may similarly appeal to local representatives or seek to reverse legislative decisions through referenda, state intervention, or litigation. Different forums, these examples indicate, may generate different results, so that the enactment or defeat of a proposal may ultimately appear to be a consequence of the arena in which the determination was made, rather than of any inexorably correct view of the proper scope of local action. My concern in this book is with the particular set of biases that judicial determinations bring to the process of defining the scope of local autonomy. As the Santa Fe situation reveals, litigation, like war, can be a continuation of politics by other means. Where enactment of living wage ordinances has generated litigation, the judicial responses to the legal issues have been as mixed as legislative responses to the explicitly political ones. The Louisiana Supreme Court invalidated a New Orleans living wage law by finding violations of the same type of state constitutional requirements that the New Mexico courts had found to be satisfied.9 A Missouri appellate court similarly held that state law preempted a St. Louis proposal.10 But a New Jersey appellate court upheld a county living-wage ordinance against challenges on equal protection and vagueness grounds,11 and a deeply divided panel of a federal appellate court rejected a federal constitutional challenge to a living wage ordinance in Berkeley, California.12 The Louisiana Supreme Court decision most thoroughly considers the legal doctrines through which the underlying tensions are displayed. In four different opinions, that court debated the scope of autonomy that home rule municipalities enjoy, the extent to which local wage ordinances generate statewide effects that preclude municipal regulation, the capacity of localities to regulate private economic activity, and the appropriate role

th e s c o pe o f l o c al co mp et ence

7

of a court in venturing into those issues. Although a majority of the court ultimately concluded that state constitutional principles placed wage regulation beyond the municipality’s authority, the breadth of the opinions and the divisions among them suggest conclusions ultimately motivated less by the ambiguities of legal doctrine than by contested views about the merits of local efforts to improve the lot of the working poor. To some extent, the common features of the political and legal debates belie more-subtle distinctions among the ordinances at issue. Although multiple municipalities have deliberated the merits of a proposition denominated as a living wage ordinance, the source and scope of the contested proposals have varied significantly. Some apply only to municipal employees or employees of contractors who deal with the municipality; some, as in the New Jersey case, cover only employees in a specific industry; some, as in Chicago, Berkeley, and Santa Monica, have been limited to a distinct minority of private employers in the locality; some, as in Santa Fe, broadly cover a substantial percentage of local employers. The scope of a particular ordinance may reflect compromise, a sense of what is legally permissible, or a sense of what is politically plausible. These political and legal battles, however, adumbrate a deeper puzzle that I will claim should inform the judicial attitude toward local autonomy. Under the standard theory of urban finance, local governments should not be the focal point of debates about living wages at all. That is because these ordinances, and others like them, implicate local policies that redistribute wealth. I define the term “local redistribution” expansively. Typical discussions of redistribution refer to policies that transfer wealth from the relatively wealthy to the relatively poor. But my concern in this work is to explore the economic, political, and legal circumstances of local decision making where some residents (or nonresidents) subsidize the activities of others. These necessarily include local policies that, at least in the first instance, require redistribution to the relatively wealthy. Such policies are usually defended in nonredistributive terms, as when the expenditure is expected ultimately to produce net economic benefits for the locality as a whole. Subsidies for public stadiums, tax exemptions for businesses that agree to locate within the municipality, or the exercise of eminent domain to take property that will be developed privately all fall into this category. Nevertheless, as I explain in chapter 2, because those policies require that at least some residents

8

the s co p e o f lo c a l c o m pe te n c e

pay for programs that will leave them worse off and thus induce them to avoid the redistributive payments, they raise the same question of crosssubsidies as local redistribution in favor of the poor. The conventional wisdom of urban finance that municipalities should play little role in fulfilling the redistributive functions of government is predicated on a simple and compelling premise. Residents and firms, at least those that are mobile, that bear the burden of local redistribution can too easily exit to neighboring jurisdictions that impose only benefit-based exactions of the sort that pays for goods and services that the payers themselves utilize. Local governments, for example, can effectively tax to pave local roads or maintain local parks that are used by substantial portions of the local population. But efforts to redistribute wealth to a small subgroup of residents will induce those who pay the subsidies to relocate outside the offending municipality’s boundaries. What is worse from the municipal perspective is that the emigrants are likely to comprise the relatively wealthy, who bear a disproportionate share of the redistributive burden and thus have incentives to find alternative residence in order to avoid taxation from which they receive no tangible benefit. Residents who migrate to geographically proximate jurisdictions will continue to be able to enjoy the (dwindling) benefits of employment or entertainment in the redistributing locality, while escaping the associated redistributive burdens. The resulting exodus, according to the orthodox theory, means that the redistributing locality loses out in the ruthless interjurisdictional competition for residents and tax base. As net payers exit, the redistributive burden falls increasingly on those who remain, exacerbating incentives for them to emigrate as well. Indeed, on some accounts, redistributive municipalities will not only suffer a loss of potential net payers but will attract additional beneficiaries, increasing the demand and cost for redistributive services. The predicted consequence is a downward financial spiral as fewer payers are left to fund more subsidies.13 Indeed, the orthodox theory predicts more dire social consequences of local redistribution. Relatively wealthy emigrants who exit for geographically proximate jurisdictions that promise fewer redistributive taxes, but who continue to work and shop in the redistributing locality, may also distort the provision of other public goods. Those who reside in benefit-based jurisdictions will likely demand more roads and other services that facilitate access to employment, culture, or economic

th e s c o pe o f l o cal co mp et ence

9

activity in the locality from which they moved. Ultimately, dissatisfaction with increased commuting times and increased taxes may cause employers to emigrate from the redistributive locality as well, causing further deterioration of the locality’s tax base and imposing deadweight losses as funds that could be dedicated to productive enterprises are swallowed by the costs of transition to new localities. Debates about living wage ordinances in places like Chicago, Santa Monica, and Santa Fe offer quintessential examples of the orthodox theory. Conferring above-market wages on employees arguably reallocates to employees the surplus between an employer’s revenues and costs that would otherwise be captured either by employers, in the form of higher profits, or by customers, in the form of lower prices. Opponents of the ordinances predicted that mobile employers and their customers would respond by taking their business to neighboring jurisdictions or by hiring fewer workers and only those whose net marginal product justified the higher mandated wages. The net effect of local redistribution, on this story, means sacrificing the very jobholders that the living wage purports to assist. Notwithstanding the sincere and altruistic motivations of those who endorse local redistribution, therefore, orthodox theory predicts that their efforts will ultimately be self-defeating. Importantly, the orthodox theory of urban finance is not an argument against redistribution. Rather, it is an argument about which level of government in a federal system is best suited to redistribute wealth. If two levels of government can redistribute the same amount of money, but taxation for redistributive purposes by one would cause less undesirable activity than taxation by the other, it makes sense to assign the redistributive role to the former. In a federal system, the theory implies, redistribution by more-centralized levels of government will cause fewer dislocations than redistribution by decentralized governments. While taxpayers required to subsidize redistribution may easily move from one locality to a neighboring one to escape redistributive taxes, even interlocally mobile capital may find the costs of exiting more-centralized jurisdictions (states or the nation in a federal system) too high to bear. Thus, they will pay redistributive taxes assessed by those jurisdictions rather than bear the higher costs of emigrating to less redistributive states or nations. As a result, the theory goes, while local governments can and should adequately allocate wealth in a manner consistent with

10

t h e sco p e o f lo c a l c o m pe te n c e

the preferences of their residents, they should cede the redistributive function of government to others. If living wage ordinances represented an isolated counterexample to the conventional wisdom about local redistribution, we might explain their presence as a function of some idiosyncratic characteristic, such as a unique coalition of supporters or their salience as a social issue. But these ordinances serve as only one example, albeit a highly visible and controversial one, of pervasive local redistributive programs that orthodox theory predicts should be rare. For instance, the most recently available figures from the Census of Governments reveal that municipal governments paid $11.3 billion in direct expenditures for public welfare in 2006–7 (not reimbursed by federal or state grants) and an additional $20.8 billion for hospitals and health care.14 These expenditures represent about 7 percent of all municipal direct expenditures.15 These figures do not include indirect local redistribution such as progressive local income taxes, exemptions from local sales taxes on goods that are used disproportionately by the poor, or property tax limits that apply to the poorest of homeowners. Nor do they include broad mandates that favor poor residents, such as local regulations—including living wage obligations and rent control ordinances—imposed either on the municipality itself or on private individuals to make expenditures or accept prices that differ from those dictated by markets. If the orthodox theory of local redistribution is correct, if local redistribution is inconsistent with local welfare, then why do we see so much of it? Indeed, why are local redistributive programs proposed at all? The core argument of this book is that the local redistributive puzzle is solved by one of two plausible but radically different explanations. The first is relatively benign. It posits that the problem lies with the orthodox theory itself; that is, the conventional wisdom ignores the extent to which local redistribution may deliver highly localized benefits that could not be replicated by redistribution at more-centralized levels of government. The orthodox view implies that preferences for redistribution are homogeneous throughout the society. If that assumption were true, redistribution solely at the central level would be sensible, as local efforts would add costs without increasing the socially desirable amount of redistribution. If, however, preferences for redistribution are heterogeneous, perhaps because different localities get different benefits from different levels of

th e s c o pe o f l o c a l co mp et ence

11

redistribution, then local redistribution would allow a larger number of individuals to satisfy their preferences for a specific level of redistribution. As I will suggest, there are multiple reasons for a locality to prefer an above-average level of redistribution, even for the poor. Certainly one could imagine that informed residents of a particular locality would share a preference for redistributive policies out of a sense of altruism for the local poor, notwithstanding that their municipality will subsequently lag in interlocal competition for tax base or residents. As I suggest in chapter 3, some localities hold themselves out as embracing redistributive policies. Indeed, if localities have disincentives against redistribution but nevertheless implement such policies, then shouldn’t we assume that the programs that we see represent either altruistic behavior that enhances social welfare for the larger community or the expectation of some discrete local benefit that belies the orthodox theory? These rationales suggest that local redistribution deserves at least cautious support both within the municipality and, at least to the extent that negative external effects are minimal, outside it. Failure to permit such programs would dilute the scope of local autonomy by denying to municipalities the capacity to enact policies consistent with the interests of local residents. Moreover, to the extent that redistribution means both that local residents bear the costs of the program and that they enjoy the related benefits, support for local subsidies imposes few external burdens and is therefore consistent with even weak endorsements of municipal autonomy. If a locality enacts inefficient or undesirable policies but hurts only its own residents in the process, then why not allow it to proceed and correct its errors through the political process rather than by ex ante restrictions on local competence? On any of these benign rationales, courts would have little reason to resolve any ambiguities about legal authority against the enactment of local redistribution projects. But the presence of redistributive activity, whether to the poor or the wealthy, does not necessarily denote the presence of an altruistic population or of expected unique local benefits. It is also plausible to provide an explanation that is more “malign,” a concept on which I expand in chapter 4 but that essentially involves motivations for local action that are not linked to the interests of constituents. Given the potential for local redistributive activity to confer wealth on a subgroup within the locality, its presence could also portend the presence of a coalition that can, at low

12

t h e sco p e o f lo c a l c o m pe te n c e

personal cost, successfully pursue policies that deviate from the interests of local residents in general. Presumably, even advocates of local autonomy would be reluctant to endorse local redistributive activity that was inconsistent with residents’ interests. Local autonomy, that is, does not justify decision making by those who happen to serve as political officials, notwithstanding that their entrenched or monopolistic positions permit pursuit of policies that distort the interests of their constituents. Rather, the justifications for local autonomy assume that local policies substantially overlap with residents’ interests. Indeed, I will suggest that a major objective of legal doctrines that regulate local governments is to neutralize the tendency of local government officials to pursue for personal reasons interests that deviate from those of residents, and thus to increase the capacity of local governments to realize the objectives for which local governments are uniquely suited. Can courts distinguish between the benign and the malign in evaluating the validity of locally redistributive programs? Should they attempt to do so? Do they attempt to do so? Those are the issues with which I am concerned in this book. One could respond to the possibility that local officials will deviate from their constituents’ interests by reducing local discretion generally. That would mean allocating decision-making responsibility for localities to some more-centralized jurisdiction or reverting to rules of statutory construction that narrowly define the scope of legislative delegations to local governments. But that is not the tack I take in this book. Allowing external substantive review threatens invalidation of a local program with which the reviewing entity disagrees, notwithstanding that the locality has a significant interest in pursuing it. That possibility would significantly reduce the scope of local autonomy. Rather, I argue, evaluation of a municipality’s legal authority to redistribute wealth ought to consider the process by which the redistributive project was adopted to ensure that the full panoply of local interests affected by the decision— both those who benefit from redistribution and those who bear the subsidy—were sufficiently represented to conclude that the project can be explained as a benign rather than malign instantiation of local redistribution. The underlying assumption is that flawed political processes are more likely to generate malign redistributive programs, while processes in which monopolistic interest groups are neutralized are more likely to produce benign redistributions.

th e s c o pe o f l o c a l co mp et ence

13

Any reliance on process to separate benign from malign redistribution shifts the responsibility for defining the scope of local autonomy from the locality itself to the entity that is charged with determining the adequacy of the local political process. When redistributive programs are challenged as being beyond the scope of local authority, courts ultimately become the arbiters of the proper scope of local autonomy. But no legal doctrine explicitly assigns to courts the obligation to review the process by which an ordinance was enacted and to validate it if the process was sufficiently pure but invalidate it if the process was substantially flawed. Instead, courts decide whether to uphold a local project by applying doctrines that speak in more ambiguous terms about the proper scope of local autonomy. For instance, courts may nullify an ordinance as dealing with something other than “municipal affairs” that fall within the municipality’s jurisdiction or as preempted by state law even if the ordinance does involve a municipal function. That was the strategy of those members of the Louisiana Supreme Court who found that New Orleans had overstepped its authority in enacting a living wage ordinance. Alternatively, courts may take a more acquiescent approach to the scope of local autonomy. They may interpret constitutional and legislative grants of local autonomy expansively and thus defer to the political judgments made within the locality itself. In such cases, courts will play a minimal role and will entertain a broad presumption of validity for municipal action. Courts that adopt this perspective, such as the New Mexico Supreme Court that reviewed the validity of the Santa Fe living wage ordinance, are less likely to find a conflict between a local redistributive project and a state law or to find that the ordinance has extraterritorial effects that transcend the appropriate scope of home rule. In the following chapters, I argue that the appropriate answers to both the question of the propriety of local redistribution and the question of judicial intervention are significantly intertwined around the issue of political process and the likelihood that challenged redistributive programs evolved from a desire to advance the interests of local residents. In effect, my claim is that the arguments that favor decentralized government require that localities be able to exercise significant latitude to decide the ideal mix of goods and services that they wish to offer. I consider this argument to be largely deferential to local autonomy. But my regard for local autonomy extends only insofar as the use of

14

t he s co p e o f lo c a l c o m pe te n c e

local discretion serves the objectives for which we create decentralized government in the first instance. Local autonomy has no claim of priority when it is exercised in a manner that deviates from the very purposes that justify it. That is, to the extent that local autonomy make sense as a means for allowing potential residents to sort themselves into jurisdictions that appeal to different preferences, claims of local autonomy cannot easily justify policies that deviate from those preferences. What may appear to be an obvious platitude, however, turns out to be a remarkably complicated issue as one attempts to decipher local interests against a background of mixed motives of local officials and those who influence them, political bargaining, and imperfect political processes. Living wage ordinances serve as exemplars of both the process that I suggest underlies the value of local government and the difficulty of reverse engineering local decisions to determine whether those values have, in fact, been realized. The stories about living wage ordinances in Chicago, Santa Monica, and Santa Fe initially look like examples of political bargaining in which all sides are well represented. Thus, there seems little reason for judicial intervention to second-guess the local decisions. The New Mexico courts that upheld the Santa Fe ordinance seem to have gotten it right in refusing to invoke relatively flexible doctrines concerning home rule and state preemption to reverse a local decision that appeared to reflect a wellbalanced process that incorporated a variety of affected interests. The local debates that preceded the ordinance look, at first cut, like local democracy at its best. If the New Orleans ordinance sparked a similarly robust debate, the Louisiana Supreme Court’s intervention initially appears to have dealt an unnecessarily harsh blow to local autonomy. But that first cut at interpreting local living wage decisions obscures important details about the incentives of localities to engage in malign redistribution, the incentives of those who would benefit from and be burdened by local redistribution, and the capacity of courts to detect malign or benign programs. Ultimately, I maintain, the objective of the legal system is to promote a process that makes those incentives and their consequences more transparent, and thus increases the likelihood that some rough approximation of local interests will emerge by neutralizing influences that distort municipal decision making. In the context of redistributive legislation, those influences typically mean interest groups that seek either to capture a share of the municipal budget for

th e s c o pe o f l o c al co mp et ence

15

objectives not shared by local residents generally, or to commandeer the local authority to mandate private activity that diverges from the preferences of local residents. As I hope to develop, it is unclear whether legal systems, and courts in particular, can or should attempt to constrain interest groups. A substantial literature argues against such efforts. But I argue that the relevant literature has arisen in a context that is more distant from the case of local redistribution than might appear. That literature primarily involves the complex relationship between federal courts and Congress, and deals primarily with principles of statutory interpretation that courts should apply in light of that relationship. I contend that the relationship between state courts, which pass on the scope of local autonomy, and the state and local legislatures that authorize or implement local redistributive programs varies from the federal model in important doctrinal and institutional ways. Whether state courts have greater authority to reverse engineer local redistributive decisions than their federal cousins have to revisit congressional legislation, however, does not address the propriety of the exercise of that authority. Even a conclusion that courts have the authority to intervene in local decision making is not necessarily a recommendation of the practice. If judicial inquiries into the nature of local political processes are themselves fraught with error, they are no more likely to represent the exercise of local discretion that justifies local autonomy than are legislative decisions that evolve from a flawed political process. Thus, in later chapters, I address the issue of whether courts can engage in the practice of reviewing local political processes with a degree of accuracy sufficient to justify the effort. LOCAL GOVERNMENT AND INTEREST GROUPS: THE PROBLEM OF COLLECTIVE ACTION

The problem of local redistribution with which I am concerned can be elucidated with a rudimentary account of the theory of collective action. Economic theories of local government begin with the understanding that a primary objective of local governments is to supply local public goods, that is, goods and services that are substantially nonrival and nonexcludable within a geographical range approximating the boundaries of the locality. Their nonrival nature means that the use of the good by one

16

t he s co p e o f lo c a l c o m pe te n c e

person does not preclude simultaneous use by another person. Or, it may be possible to exclude an additional user, but there is little social reason to do so since that use does not interfere with the enjoyment of the initial user. Their nonexcludable nature means that once the good is provided, it is difficult to preclude others from using it. Think of controlling mosquitoes, or avoiding depletion of the ozone layer. Any action one person takes toward these ends will benefit others, regardless of the latter’s failure to contribute to the effort. Goods with these features have important consequences. Those who could benefit from their provision may seek to avoid paying for them because they can obtain the same benefits without incurring any personal cost if others who would similarly benefit are willing to pay. Think of a city street that might need paving. Abutting users might all agree that paving the road would return individual benefit in excess of the necessary per capita contribution. But each abutter would also understand that he could not easily be precluded from using the road along with the others if he failed to contribute and the others absorbed his share of the costs. If no one can require individual contribution, each abutter has an incentive to free ride off the efforts of others. That does not mean that no one will cooperate. Individuals who live within close proximity of each other have incentives to cooperate even without being required to do so. Moreover, there are limits to the “publicness” of local public goods. At some point a city street will become sufficiently congested that it will take on the features of rivalness; access to the road can be limited by the placement of intersections, and pricing by toll booths is theoretically possible (although the marginal cost of adding an additional user may be so small as to make the collection of that cost inefficient). The benefits to providing a service with the characteristics of a local public good may be so great to one individual that he is willing to bear the related costs, even without receiving contributions from other beneficiaries. Nevertheless, the standard conclusion that follows from the analysis of public goods in general and local public goods in particular is that they will be undersupplied relative to the demand for them because people will understate their desire for them in order to avoid the related costs. Local government provides the standard solution to the problem of local public goods. Armed with the capacity to impose and collect taxes and fees, local governments can overcome the collective action problem

th e s c o pe o f l o c a l co mp et ence

17

by receiving signals (primarily through voting for officials who promise to provide a particular bundle) of the public goods that constituents desire, providing those goods, and then charging constituents for them. As long as the goods provided basically reflect their preferences, constituents have no incentive to understate their desire for a good because they know that it will not be provided unless a substantial number of residents agree, and that if it is provided, all residents (in the case of taxes) or all users (in the case of fees) will be required to pay. Local government essentially solves the dilemma of free riding. Or so the theory goes. The problem is that local governance itself—the process of determining what local governments should do and then implementing those decisions—also possesses the characteristics of a public good. At least within the limits of a local budget, the decision to spend tax dollars or impose fees for one project does not preclude the imposition of taxes or fees for another. Any decision that purports to serve the interest of constituents applies to all residents; none is easily excluded from the same benefits of government. These characteristics affect the decisions that are made by local governments and the identity of the people who make them. I noted above that government officials, unlike their private market counterparts, determine what goods and services constituents desire by examining the expression of preferences registered through voting or through direct contact with constituents who urge provision of a particular local public good. But there are some difficulties with the assumption that constituents reveal their preferences. First, the publicgood nature of political participation means that many people might not bother to register them at all. If constituents believe that others will provide the appropriate signals and incur the costs of doing so, then they have no more incentive to bear the personal costs of assisting in governance than they did in providing the underlying goods directly. Voting in particular may not be worthwhile for individual citizens because it provides a relatively imprecise mechanism for registering preferences. Elections tend to be occasional, binary affairs in which residents vote on a bundle of goods rather than on a service-by-service basis. A resident who is unhappy with sidewalk repair and public parks may still vote for incumbents if she is satisfied with the quality of public schools and public safety. Elected officials may have no easy way of disaggregating votes to distinguish between those services that are satisfactory and those

18

t he s co p e o f lo c a l c o m pe te n c e

that residents desire to see improved. Nor does voting register the intensity of agreement or disagreement; voting tends to be an on-off switch (I vote for or against), so that if I favor candidate A very strongly, my vote can be canceled by the vote of someone with only the mildest preference for candidate B. Second, electoral checks arise only in the presence of political competition, and there is significant evidence that local politics are likely to be less competitive and less partisan than politics at more-centralized levels of government. Local officials who enjoy a monopoly obviously have broader discretion to impose taxes for projects that serve their own purposes, even though those projects would not survive a direct vote by residents. Jessica Trounstine demonstrates that monopolistic local officials target spending toward their supporters more than nonmonopolists.16 This possibility can be extended by models that suggest how local officials can shroud decision making about expenditures and thus enhance the budget for favored projects by controlling the spending agenda. Romer and Rosenthal show how binary voting permits an agenda setter to increase spending for individual projects beyond what constituents prefer.17 Assume, for instance, a political body—a city council— that has a monopoly over proposing expenditures. Assume that the city council proposes an expenditure of $100,000 for a project that most voters would prefer, although they would prefer to fund it for no more than $80,000. But given that the $100,000 expenditure is closer to the preference of most voters than no expenditure, voters without the power to amend the proposal will likely approve the excessive expenditure. Taken to its extreme, an agenda-setting monopolist may adopt a strategy of redistributing disproportionate local benefits to one group in an explicit effort to induce net payers to exit, thus increasing the probability that the monopolist will continue to enjoy the support of those who remain. While the locality as a whole may deteriorate, incumbents may receive sufficient personal benefits from their monopoly to make the strategy personally rewarding.18 Third, even well-meaning public officials may misinterpret the public’s preferences, and less well-meaning officials may ignore those preferences. Interpreting the meaning of votes is complicated in an environment where cycling is possible, information is imperfect, choices are limited, and different groups face different organizational advantages and

th e s c o pe o f l o c a l co mp et ence

19

disadvantages.19 Hence, public monitoring of officials’ conduct is necessary to ensure that the public’s preferences are implemented as well as registered. But monitoring conduct is also a nonrival and nonexclusive activity. My costly review of an official’s conduct does not preclude others from similar behavior and confers benefits even on those who contribute nothing to my effort. The result is that there is little incentive for any one person to undertake the tasks of governing, registering preferences for governors to follow, and monitoring those officials to ensure that they act in a manner consistent with the preferences that are registered. Those who engage in these activities, therefore, may do so because they have idiosyncratically high interests in the outcome or have interests in particular issues. Like the person who may provide a local public good directly, even without obtaining contributions from others, because his personal benefit from obtaining it is sufficiently high to offset the personal cost, those who participate in local government may have a stake in the outcome that is not representative of the interests of most constituents. The public-good nature of government, therefore, provides an advantage to those who can overcome the collective action problem and convince local officials to adopt the policies that the organized group advocates. If the benefits of political action are concentrated among a small group, and if that group is able to coalesce and control its membership, then the benefits to individual members of participating in collective action may exceed the costs. Thus, those who are left to vote, or to lobby for passage or defeat of proposed legislation, will be those with above-average amounts to gain or lose from the ultimate decision. It is, as a consequence, by no means clear that the interests of those who participate reflect the interests of local constituents at large. At the extreme, a group able to organize and lobby successfully for a favored governmental activity may not seek a public good at all. Rather, it may seek to have the government provide that group with a benefit that is largely private, but the cost of which is subsidized by the local treasury. The small size of the individual contribution through tax dollars or fees deters even those constituents who receive no benefit from the expenditure from organizing in opposition to it. Redistributive ordinances are, almost by definition, vulnerable to political processes that increase the risk of deviation from constituent preferences. The standard formula for collective action problems that

20

t h e sco p e o f lo c a l c o m pe te n c e

cause the political distortions with which I am concerned is a proposal for government action that will benefit a small, concentrated group at a cost that, albeit significant in the aggregate, is too small on a per capita basis to be salient to individual members of the relatively large and diffuse group that subsidizes it. This formula is also characteristic of many local redistributive programs, since their objective is to provide benefits to a relatively small subgroup of the locality. These programs deviate from projects paid for through benefit-based mechanisms such as user fees or service charges in which each resident gets what he or she pays for and pays for what he or she receives. Indeed, I will argue that groups that stand to benefit from redistribution often exploit the characteristics that deter collective action among prospective opponents. Redistributive benefits could be funded through any of a variety of mechanisms, ranging from imposing the payment obligation on a single taxpayer to funding it from the general treasury of the locality. Prospective beneficiaries of local redistribution have incentives to structure the payment obligation in a manner that minimizes the organizational success and political influence of subsidizers. Note, for instance, the various distinctions among different ordinances that are lumped together under the category of “living wage,” as reflected in the Chicago, Santa Monica, and Santa Fe versions. Chicago’s limitation to “big-box” stores, Santa Monica’s limitation to businesses frequented by nonresidents, and Santa Fe’s ultimate extension to businesses with twenty-five or more employees can all be explained in terms of minimizing political opposition rather than striking an optimal balance between those who can benefit from and those who can afford to subsidize an above-market living wage. More difficult to establish is whether that same effort, by diluting the participation of prospective opponents in local political battles, is also relevant to the determination that a locality has the authority to enact redistributive ordinances or that the proposed program serves the interests of the locality. All legislation creates winners (those who benefit from the legislation in excess of the costs they incur) and losers (those who pay more than they benefit). Each group has incentives to participate in the political process by which officials determine whether to enact proposed programs. Ideally, political processes at the local level, where broad participation is plausible, allow accurate measurement of residents’ preferences. Moreover, as the orthodox theory implies, the market for residence limits

th e s c o pe o f l o c al co mp et ence

21

the capacity of any group to exploit mobile nonparticipants because net losers can emigrate from redistributive jurisdictions at relatively low cost. But the disincentives that prospective opponents face to become involved in debates about legislation of which they disapprove complicate any effort to infer majority will from enacted legislation. Explicitly redistributive legislation appears to be particularly susceptible to these democratic difficulties. Think of a tax abatement given to a housing developer, where the effect is to require all residents either to tolerate fewer services or to pay nominally higher taxes to compensate for the forgone municipal tax revenues. The result predicted by standard collective action theory is that the beneficiary that obtains such a benefit is intensely interested in its award, but the numerous and diffuse group members who pay the de minimis individual cost remain relatively disinterested in the proposed transfer. Even if aggregate tax increases exceed the abatement's value enjoyed by the beneficiary, it is not worth the while of any payer to object, given the small individual payment. The beneficiary of the entire subsidy, however, will find it worthwhile to lobby for the payment. Sole (or small groups of) beneficiaries cannot rely on the efforts of others to procure the benefit and thus have incentives to lobby for it. Meanwhile, the large number of payers means that each is likely to free ride on the efforts of others to forestall the transfer. Of course, the result of mutual free riding is that few payers will actively oppose the proposal. As a result, local officials will hear a skewed version about the propriety of the inefficient abatement. In this situation, even well-meaning political officials are likely to act in a manner that misrepresents local preferences because the content of those preferences is insufficiently registered. I will refer to such a situation as one in which the interest group that seeks the benefit is able to dominate the debate about its provision. The problem is obviously exacerbated by the public-good nature of governing. Just as those who become involved in the task of registering preferences or monitoring local officials may have interests that deviate from those of the average resident, so might those who make the effort to become public officials. Theorists have posited a variety of objectives for public officials that vary from the pure representation of constituents— maximizing the public budget that could be spent, maximizing the chances of reelection, maximizing the chances of election to higher office, or maximizing opportunities for post–public service employment. None of these

22

th e sco p e o f lo c a l c o m pe te n c e

implicate bribery in its most invidious sense. But each of them implicates the possibility that those who control local expenditures may redistribute public wealth to those who promise certain personal benefits in return, and the public-good nature of monitoring officials dilutes the incentives of those who must pay for nonpreferred services to detect and deter their provision. None of this means that local decision making is inherently flawed relative to other forums for collective decision making. To the contrary, local government is often seen as a focal point for self-governance where individuals and groups that could not gain access to more-centralized political forums can participate in public decision making. Groups may form on all sides of an issue at the local level and generate a debate that presumably results in a decision that reflects the breadth of local interests. Think, for instance, of rent control, which requires landlords, but not the population at large, to subsidize rents. Landlords may be relatively small in number, but the adverse affect of their being unable to charge market rents for their properties creates significant incentives to engage actively in the debates about rent control. In addition, the fact that they already deal with local legislators on a variety of issues, such as zoning and building codes, means that they face lower marginal costs in obtaining access to the political process on the rent control issue. Local resident tenants will be more numerous, but that does not mean that they will eschew the political battle. Unlike the case of the tax abatement, the value of rent control to any one of them will likely be sufficient to warrant participation in the debate as well. Not all will participate, of course. Even those who will gain from enactment of a rent control ordinance need not participate in order to obtain those benefits, as long as enough others who are similarly situated bear the related costs of organization and lobbying. But the incentive to free ride on the labors of others is somewhat reduced by the capacity of tenants to identify their neighbors and cajole them into joining the collective effort. The result is that both those who favor rent control and those who oppose it have incentives to be heard on the issue. If the paying group (landlords) is much smaller than the benefited group (tenants), then the former essentially has limited capacity to influence the outcome. But would we identify that as a problem of collective action or of democratic decision making? The presence of a well-organized group

th e s c o pe o f l o c al co mp et ence

23

that has the ability to have its views considered, albeit rejected, seems far from a process in which proposals escape the scrutiny of democratic deliberation because the interests of the group are not represented at all. It is only the latter case that raises the specter of a process in which decision making is likely to distort constituent preferences. Again, any concern that redistributive efforts will inexorably be subject to a process dominated by discrete interest groups is at least tested by the very fact that Chicago, Santa Monica, and Santa Fe hosted rigorous debates within city councils and ostensibly within the broader community. The emotional comments at the Santa Fe hearing may not be the local equivalent of debates at the Constitutional Convention. But they do suggest the presence of a deliberative process that presumably underlies democratic decision making, the prospect of which is the basis for identifying local governments with participatory democracy. The invocation of legal constraints on local autonomy in such instances seems to represent a misplaced distrust of local decision making, and an inappropriate allocation to courts to decide the proper scope of local activity. T H E L O C A L I T Y A S A B AT T L E G R O U N D F O R I N T E R E S T GROUP COMPETITION

Initially, much of the collective action story appears to be consistent with James Madison’s justification for centralized decision making in Federalist #10. Centralization was a response to what we may refer to as “asymmetrical access,” in which those on one side of an issue have the capacity to influence decision making, while those on the other side face organizational obstacles of the type that I suggested inherently impede collective action. Madison famously advocated a strong central government on the grounds that it would best be able to forestall the negative consequences that materialize when a single “faction,” what today we would call an interest group, dominates the debate on an issue of public importance. Madison appears to have been concerned primarily about majority factions that could run roughshod over the interests of minorities. But he was at least attentive to the capacity of minority groups to dominate political debate. Madison spoke of the need to neutralize “the cabals of a few” as a justification for having a large number of representatives. His conclusion was that dominant groups, those that are unopposed in public debate, were more likely to arise in smaller bodies (by which he meant

24

t he s co p e o f lo c a l c o m pe te n c e

state legislatures) than in those of more-centralized governments. Of themselves, interest groups are not necessarily problematic. Indeed, Madisonian Nirvana may be achieved when interest groups arise on all sides of an issue, as long as the result is interest group competition rather than division of the spoils among groups.20 Under those circumstances, we would anticipate robust debate and a greater likelihood that what emerges reflects the interest of the public at large. The antidote to dominant interest groups is what other theorists have referred to as countervailing forces.21 Factional influence on any given issue would best be contested by the emergence of a competing faction that represented an alternative side of the debate. For Madison, therefore, the solution to factional dominance was to shift decision making to that level of government that was likely to be a forum in which competing factions would emerge. The centralized federal government, Madison claimed, was systematically more congenial to factional competition than were relatively decentralized states. Any given decentralized jurisdiction might host an insufficient number of advocates on multiple sides of an issue to ensure competitive deliberation. If, for instance, the beneficiaries of proposed action were concentrated in one jurisdiction, while net payers were geographically diffuse, the former jurisdiction would face pressure to take the action, while the net payers would have greater difficulty coalescing and would likely go unheard. But even among the diffuse payers, there might exist a few in various jurisdictions that would have sufficiently intense preferences about the issue to incur the costs of organizing. If those parties could identify like-minded individuals from other jurisdictions, then they may be able to coalesce at the central level of government in which all are represented, even though their small numbers would preclude an effective coalition in any of their local jurisdictions. Here is Madison’s more dramatic language: The smaller the society, the fewer probably will be the distinct parties and interests composing it; the fewer the distinct parties and interests, the more frequently will a majority be found of the same party; and the smaller the number of individuals composing a majority, and the smaller the compass within which they are placed, the more easily will they concert and execute their plans of oppression. Extend the sphere and you take in a

th e s c o pe o f l o c al co mp et ence

25

greater variety of parties and interests; you make it less probable that a majority of the whole will have a common motive to invade the rights of other citizens; or if such a common motive exists, it will be more difficult for all who feel it to discover their own strength and to act in unison with each other. If Madison were systematically correct, his reasoning might supply an additional rationale for deferring much of the control over local legislation to more-centralized governments. But much as Madison’s analysis presages modern collective action analysis, interest group competition is not systematically more prominent in centralized rather than decentralized jurisdictions. The capacity of groups to coalesce and to lobby is a function of a variety of factors, some of which are internal to the group, and some of which are a function of the organizational structure of the decision-making body that is the focus of lobbying. Within the group, the costs of identifying, communicating with, and obtaining contributions from like-minded individuals will determine the capacity to organize. Once organized, the group must have the capacity to attain access to and influence the decision-making forum. Groups that are repeat players before a legislature may be willing to invest more in creating a relationship than groups that cannot amortize that investment over a series of projects. Not all groups will have the same advantages at the same level of government. In some situations, the costs of organization and access may favor formation of interest groups at the local level; in some instances they may favor formation at more-centralized levels. I will argue in chapter 5 that some groups may be able to coalesce at both centralized and decentralized levels, while other groups that would serve as a countervailing force can effectively coalesce at only one of the levels. As a result, Madison’s general claim about asymmetrical access appears correct: we may sometimes see a group emerge unopposed at one level of government although it confronts organized opposition at another level. But that general claim does not support a reason to agree with Madison that the local level will systematically suffer from asymmetrical access while the centralized level will necessarily host competing groups. The result is that the question of which forum is best suited to interest group competition is more complicated than Madison’s equation of small jurisdictional size with tendencies for factional dominance

26

t h e sco p e o f lo c a l c o m pe te n c e

suggests. On some issues, the local forum is more likely to generate the kind of democratic, deliberative debate that Madison implied would lead to publicly interested results. On other issues, Madison’s fear of factional dominance at the decentralized level is more likely to materialize. But if the choice among decision-making jurisdictions is predicated on the possibility of factional dominance, it is necessary to consider, with respect to any particular issue, the likelihood that each side in the debate will be able to coalesce effectively and to have a fair opportunity both to be heard and to convince the decision-making body. At times, that description will best fit a relatively decentralized decision-making body, and at other times, it will best fit a relatively centralized body. If we apply this general principle to the issue of local redistribution, it follows that we should be more tolerant of local autonomy with respect to those decisions that evolve out of situations in which countervailing forces are likely to participate, and less willing to recognize local efforts to redistribute wealth where it is likely that, relative to more-centralized decision makers, the local decision has been dominated by one side of the debate. In short, the invocation of political process basically means that local decisions did not suffer a comparative disadvantage in ensuring the participation of affected interests. The legal doctrines that assign authority among jurisdictions, however, neither explicitly invoke this standard nor much of any other. Instead, they are stated in inherently ambiguous language such as “conflict” with state legislation, “municipal affairs,” or “public purpose.” The scope of jurisdictional authority must then be clarified, typically by judicial resolution of a challenge to a locality’s exercise of authority under an ambiguous constitutional or statutory clause. When a court interprets a grant of authority to a locality narrowly, it effectively determines that the locality lacks the capacity to engage in activities that would be permitted under a more liberal interpretation of the same constitutional or statutory language. When, for instance, a court invalidates a living wage ordinance as being in conflict with state law, it is implicitly concluding that the state is a more appropriate decision maker with respect to wage restrictions than the locality. Allowing courts to interpret the proper scope of local autonomy by asking whether countervailing forces were or plausibly could have been involved in the local debate is certainly consistent with the large body of local government law that can best be understood as a means

th e s c o pe o f l o c al co mp et ence

27

of neutralizing the organizational advantages of groups that could otherwise obtain a disproportionate share of public funds. State constitutional doctrines such as expenditure limitations, debt limitations, prohibitions on unfunded mandates and “special legislation,” and “single subject” requirements—doctrines that have no federal constitutional analogue— appear to have a common objective of constraining the capacity of interest groups to dominate the decision-making apparatus of the state. These provisions, however, are not self-enforcing. Either they are aspirational only, or they authorize judicial intervention to determine when legislative acts have complied with or violated the constitutional mandates. Given the inherent ambiguity of the language of these clauses, and their ostensible objective of constraining special interests, is it appropriate for courts to take into account the structure of the political process that generated the challenged legislation? That is, can courts incorporate the quality of political competition into their interpretive task? The imperfections and institutional limits on judicial analysis explain much of why a political process explanation for legal outcomes is not in vogue, notwithstanding John Hart Ely’s efforts to revive it in his influential work Democracy and Distrust. Courts are susceptible to their own institutional biases that may cause their decisions to deviate from constituent preferences with greater frequency than do the decisions of local officials, and with less vulnerability to political correctives. Throughout this book, however, I suggest that political process review requires different considerations when the process under review involves the state or local level and the federal level, with which Ely and most of the debate have been concerned. The position that I take on the propriety of local redistribution and judicial responses to it necessarily entails a normative view that recognizes distinct features of local as opposed to more-centralized levels of government. The structure of state constitutional law and the decision-making process at different levels of government may warrant different relationships between the judiciary and the legislature. The very existence of legal restrictions on local autonomy implies that there are valid reasons to exclude certain decisions from the range of subjects appropriate for local decision making or at least to be more suspicious of them than of political decisions made at the state or federal level. After all, the denial of local authority to engage in redistribution is not a denial of government competence to deal with the issue

28

th e s co p e o f lo c a l c o m pe te n c e

generally. Instead, it reflects a desire to avoid local decision making on the subject. If process concerns inform the decision to assign specific tasks to state or local bodies, that may bolster the case for considering process in the evaluation of specific redistributive proposals. The desire for pluralism also responds to an alternative critique of the position that I expound here. Throughout this book, I refer to the preferences of constituents. That perhaps unfortunate phrase might be interpreted to mean that I believe that all constituents of a locality have the same preferences, or that their preferences can be distilled into some sort of general local will. While intramural preferences of any locality may be more homogeneous than those of the society at large, especially in the large municipalities that are responsible for much of the redistribution with which I am concerned, it would be difficult to identify any such singular set of resident preferences. To ask the question, what do residents of New York or Baltimore want? as if there were an answer that would satisfy all residents is to deny the diversity of desires that makes urban life interesting. But if that is the case, then how can it make sense to speak of residents’ preferences as if something other than the result of intergroup bargaining defined the result? Here, again, process is relevant. Defining residents’ preferences may mean nothing other than attention to the bargains struck within political debates. But those debates can be expected to generate the kinds of robust conversation and the kinds of results that vindicate broad local autonomy only if they are characterized by an exchange free of dominant interests. My argument, therefore, is one that permits broad local autonomy, but autonomy to be exercised when the locality enjoys an institutional advantage over alternative forums for making collective decisions about the same issue. THE DARK SIDE OF DEMOCRACY?

Before heading into the analysis, a cautionary word is in order. For some, examination of the political process to discern and correct observed political failures diminishes the value of democratic decision making and governmental action in general. It represents a dark picture of purely selfinterested political actors trading (selling or buying) votes in a market that is indistinguishable from a market for commodities. Values that we anticipate underlie democracy, such as community, other-regarding behavior,

th e s c o pe o f l o c al co mp et ence

29

and deliberation and reflection, are—on this view—subordinated to narrow economic interests. The use of such an analysis, for some commentators, is loaded with a political agenda of forestalling government intervention in the marketplace by demonstrating that government action is inherently and systematically corrupted by special interests. Thus, government action is disfavored relative to presumably anonymous and neutral market forces. The lesson that I extract comparing the capacities of centralized and decentralized jurisdictions to host interest group competition, however, is not intended as a recommendation of dark political behavior or even an admission of its prevalence. On the contrary, it is intended to present a relatively rosy picture of the utility of legal doctrine in implementing the teachings of institutional design. Throughout this book, I will expand on the rather simple model of collective action and local decision making that I have presented here to suggest how law can neutralize the incentives of local officials who wish to deviate from constituents’ interests. More to the point, recognition that political actors may respond to selfinterest does not constitute a recommendation of that behavior or suggest that it is universal. But there is little doubt that some political actors are guided by self-interest, and that groups with idiosyncratic interests exploit opportunities presented by political actors who implicitly or explicitly are willing to provide benefits in return for electoral support or other personal benefits. The more optimistic lesson that one takes from public choice is consistent with Madison’s observation in Federalist #51 that “[i]f men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary.” That aphorism is itself perhaps influenced by David Hume’s recommendation that “in contriving any system of government, and fixing the several checks and controuls of the constitution, every man ought to be supposed a knave, and to have no other end, in all his actions, than private interest.”22 Law can provide the constraints required by these admonitions. In chapter 5, for instance, I discuss a variety of state constitutional provisions that I suggest were inserted in order to frustrate efforts by interest groups to capture government decision makers. Indeed, one objective of this book is to identify mechanisms that can strengthen those provisions and thus advance local democratic decision-making. But designing

30

t he s co p e o f lo c a l c o m pe te n c e

institutions that enhance democracy requires recognition of the variety of ways in which local democratic decision-making worthy of the name can go awry. This is not necessarily, to use William Eskridge’s appealing phrase, a view of politics without romance. Indeed, by demonstrating flaws in democratic processes, public choice lays a predicate for defining legal interventions that can improve decision making. The constitutional clauses on which I will rely, therefore, reveal a significant faith in the ability of law to constrain dominant interest groups. The strategy could backfire. For instance, raising the costs of obtaining enactment could mean that fewer groups form, but those that do are less likely to face organized opposition. I return to this problem later, since it implicates the issue of whether one level of government is more likely than another to be populated by dominant interest groups.

ch ap t e r t w o

Why Does Local Autonomy Matter?

in this chapter, I discuss the purposes of local government. I then relate those purposes to the propriety of local redistribution and to the acceptable conditions for a working political process. My objective is to lay the basis for claims in later chapters that the orthodox theory of urban finance may understate the extent to which a locality characterized by a well-working political process would be willing to adopt local redistributive programs. In contrast, the absence of such a process could, consistent with the orthodox theory, serve as a signal of inappropriate local redistribution. Determining the proper scope of local autonomy is largely a question of institutional design. Assigning substantial authority to decentralized governments makes sense insofar as they are able to implement specified objectives better than alternative public or private institutions such as more-centralized governments or markets. The traditional objectives for which local governments are favored include fostering democratic governance, efficient signaling of preferences about local public goods, and sorting of those with different preferences to promote provision of a broader array of public goods. The democratizing functions of decentralization are typically tied to the greater opportunities that local governance provides for participation in decision making about matters of public policy than more-centralized 31

32

why d o es lo cal a uto n o m y m a tte r ?

governments, and the greater accountability that public officials have by virtue of their proximity to constituents. Even outside the idealized world of the New England town meeting, in which residents are presumed to engage in deliberative conversation that will produce conciliation and consensus, the underlying assumption is that individuals have beneficial contact with and influence over political officials at the local level that they cannot exercise with respect to state or federal officials. One of the more compelling stories about local government involves the ability of individuals to participate in democratic governance through activities as diverse as running for office or making a personal phone call to a local councilman in order to express support for or disapproval of local action. There are limits to the policies that individual citizens can influence directly, since local officials are unlikely to have jurisdiction over issues of state or national importance. But as a practical matter, individuals may be more invested in decisions about property taxes or zoning changes than national or global policies that have less immediate effects on their daily lives. Moreover, there is mixed empirical evidence on the issue of whether even relatively small jurisdictions engender a robust civic life. Presumably, participation in debates about issues of purely local significance also produces educative effects that encourage local residents to develop more-informed opinions and engage in political debate even over issues that they cannot influence directly.1 Local government is also tied to democratic principles insofar as it permits closer monitoring of officials than constituents can provide in more-centralized jurisdictions. As I indicated in the previous chapter, monitoring local officials constitutes a public good, in that all residents benefit from the monitoring performed by any one of them. The relatively low cost of monitoring at the local level, where the effects of policies on matters such as zoning, street maintenance, or school quality are more tangible, may enhance any incentive to scrutinize public officials, public-good effects notwithstanding. As I emphasize below, in theory local governments attract residents based on an offer of distinct bundles of local public goods and services. The effect is that relative to the larger, more populous jurisdiction of which it is a political subdivision, any given local government will attract a population with homogeneous preferences about the governmental services that it desires and is willing to pay for. Where residents’ preferences are heterogeneous, monitoring officials

w h y d o e s l o c a l a utono my mat t er?

33

becomes complicated because the overall objective of the municipality is unclear. Different constituents may monitor for different goods and services, but it will be difficult for any monitor to know whether any deficiencies that she discovers are the result of rent-seeking by discrete groups or are required by trade-offs with other groups and that are necessary to obtain support for other objectives that the monitor shares. Homogeneity surely has some drawbacks for monitoring (as well as for other social values). For example, the homogeneous nature of localities may enhance free riding, since residents with similar preferences are less likely to need to engage in individual monitoring in order to ensure that personal preferences are being served. But relative homogeneity of preferences for services implies that fewer residents will advocate an agenda that deviates significantly from that of the median voter or from the bundle of goods and services that a majority of residents prefer. As a result, those deviations that do occur are more likely to be salient, as officials will be less able to explain those projects as consistent with the interests of a dissident minority whose presence is essential to the welfare of the locality as a whole. As a consequence, local activity that is inconsistent with the desires of residents will be more apparent without costly investigation. In theory, the democratizing argument for local government has little relationship to the orthodox view of local government. The democratizing argument does not provide a motivation for redistributive programs. But to the extent that the argument promotes a strong view of local autonomy, in which localities have broad discretion to decide which local objectives to pursue, there is certainly no inconsistency between local democracy and local redistribution. The democratizing characteristics of local government assume that localities can pursue virtually any objectives that constituents desire, constrained only by the external consequences that restrict the scope of local autonomy generally. There is no reason why local residents could not have a preference for redistribution or that decisions to redistribute should vary in their validity from local decisions concerning which schools to close, where to locate the new municipal jail, or how many police officers should be hired. Each of these decisions imposes costs that will be borne primarily by local residents (minimal externalities) and will emanate from a decision-making process that internalizes the interests of all residents.

34

wh y d o es lo cal a uto n o m y m a tte r ?

SIGNALING AND SORTING THEORIES OF LOCAL GOVERNMENT

Democratic governance is not the only objective that commentators posit for local government, however. The standard economic justification for decentralized government is related to preference satisfaction. This argument is inextricably linked to Charles Tiebout’s observation that, under highly stylized conditions in which potential residents have perfect mobility and choice among a large number of jurisdictions whose activities impose no negative externalities on neighbors, informed individuals will sort themselves into units that offer bundles of local public goods and services preferred by residents at a tax price they are willing to pay. Under these assumptions, jurisdictions can compete for prospective residents and tax base by offering specific goods and service bundles. In-migration by prospective residents who prefer the bundle of goods and services that the locality offers, and out-migration by residents who prefer an alternative, provide signals to local officials of what alterations to the status quo are necessary to attract and maintain an optimal population. Local officials who are publicly interested, in the restricted sense that they seek to achieve an optimal population size given the cost of the preferred local bundle of goods and services, respond to those signals by adopting tax rates and public-provision policies that maximize the welfare of residents. In the phrase that has become associated with Tiebout, the interaction between local public provision and migration means that potential residents vote with their feet, rendering politics a less-important mechanism for registering collective preferences. Instead, perfect mobility induces efficient delivery of municipal services, as prospective residents gravitate to jurisdictions that they find compatible given their willingness to pay the full costs of the local public goods that they receive. One may (many do) resist drawing any normative inferences from the results generated by satisfaction of the Tiebout assumptions, either because preference satisfaction does not equate with municipal interests (e.g., residents may be myopic) or because municipal interests may be inconsistent with broader social interests. Thus, recognizing the force of Tiebout as an explanation for local government is not necessarily a recommendation that local governments pursue an optimal population at a set tax price. But unless we are willing to regulate mobility formally, some substantial degree of residential organization along the lines of preferences is inevitable.2 Assume, for instance, that a majority of the residents of

w h y d o e s l o c a l a utono my mat t er?

35

community A share a preference for service x. Assume that the policy of the municipality clearly supports x. Assume, however, that a majority of residents of community A come to the demonstrably correct conclusion that x is undesirable for the future of community A and that they are able to elect a majority of the city council that adopts a new policy, not-x. Assume, finally, that there is another community, B, that is identical to community A in all respects other than that it retains the policy x. Mobile residents and potential residents who prefer x will emigrate from community A to community B, notwithstanding that community A will ultimately be better off as a result of the enactment of not-x. Nonresidents who have a preference for not-x and who currently live in a community that offers x will migrate to community A. As a positive matter, even if not as a normative one, therefore, the presence of a significant number of decentralized jurisdictions, each with the ability to formulate a policy about which goods and services it will provide, and each facing a mobile population, will tend to attract populations that sort according to preferences. Even where the heroic assumptions of the Tiebout model cannot be satisfied, localities will tend to attract populations that prefer the goods and services that local officials provide, as those who are mobile migrate. In a world in which Tiebout’s assumptions can more closely be approximated, social organization along these lines makes sense insofar as it maximizes the ability of individuals to obtain the governmentally provided goods and services that they desire, as evidenced by their willingness to pay for them and to avoid paying for goods they do not want. The welfarist implications of the Tiebout model can be seen most easily by positing a society that comprises groups with different preferences for a particular public good. All suffer a welfare loss if they are aggregated in a single jurisdiction that offers a level of the good that corresponds to the preferences of the median voter. Assume, for instance, that there is only one potential public good, a public playground funded by property taxes paid by all residents. If there is a community of one thousand voters, six hundred of whom prefer public playgrounds and four hundred of whom do not, the former will outvote the latter, and the minority will end up paying for a service they do not desire. This may make the winners happy, since they can impose some of the cost on the losers. Indeed, given the binary nature of the vote, the winners can impose on the losers a much more costly playground than the winners would be willing to pay

36

wh y d o es lo cal a uto n o m y m a tte r ?

for because some of the costs will be subsidized by the losers. The result also leaves the losers with less ability to fund activities that they prefer, since there are likely some practical, if not legal, limits to the locality’s ability to raise taxes. If there were two communities that sorted according to their desire for publicly funded playgrounds, the six hundred proponents will live in one community and fund a playground of the size that they are willing to pay for, and the four hundred opponents will live in the other community without a playground. The preferences of all will be satisfied, with each community paying for the services that it values. In this sense, decentralized sorting facilitates the efficient provision of local public services because residents only get services if they are willing to pay the related costs, and they only pay for the services that they desire.3 The example is oversimplistic given the large number of goods and services that localities provide and the subsequent impossibility of ideal sorting about each one. But the very fact that localities vary in the bundle of goods and services they provide suggests that there is something to the point. The interlocal competition implicit in a Tiebout model generates additional positive effects, some of which are highly intertwined with the democratizing effects of decentralized government. Some evidence suggests that jurisdictional fragmentation constrains bureaucratic budgets otherwise vulnerable to the demands of budget-maximizing agencies. Recent empirical work suggests that lower taxes and superior services in some jurisdictions generate efficiency-generating responses in the tax rates and services of neighboring jurisdictions, and that the presence of multiple jurisdictions in a metropolitan area constrains the size of government.4 Competition, of course, is not necessarily costless. Interjurisdictional competition decreases reliance on ability-to-pay taxes and thus can reduce tendencies for local redistribution even where it is otherwise desirable.5 Tax competition may lead to a race to the bottom, rather than to the top, as jurisdictions offer low tax rates by cannibalizing services useful to those who are relatively immobile. But at the very least, the constraints that competition imposes on municipal waste must be weighed against the traditional drawbacks of political fragmentation. Regardless of its normative consequences, Tiebout’s model does appear to have substantial explanatory force. Empirical work supports the proposition that the variety of local governments and mobility of potential residents and capital create significant sorting and interlocal competition along

w h y d o e s l o c a l a utono my mat t er?

37

the lines that Tiebout suggests. Early work by Wallace Oates concluded that median property value was positively influenced by a community’s public spending, a result that is at least consistent with the assumption that individuals migrate to communities that provide valued services. Jan Brueckner found that Massachusetts communities exhibited no systematic tendency to over- or underprovide public goods. One might expect Tiebout to be realized in relatively small jurisdictions or those that offer relatively few local public goods so that residents can become better informed. William Fischel’s study of “homevoters” in New England towns bears this out. But there also seems to be some evidence that larger cities are unique in a manner that is consistent with Tiebout sorting. It is unlikely that pure coincidence explains why San Francisco is more responsive to the needs of the homeless than other California localities, why Sun City Center finances more services for the elderly than communities that are not actively seeking retirees, or why Boston attracts more colleges and universities than Omaha. Even if these distinctions did not develop from deliberative decisions, once unique characteristics are entrenched, relatively mobile residents and potential residents emigrate from and immigrate to these localities based on their sympathy with those characteristics.6 Moreover, it is a mistake to dismiss preference satisfaction as a means of treating cities “simply as objects of consumption,” notwithstanding Tiebout’s reference to “consumer-voters.”7 Individuals may have residential preferences for characteristics that are not easily classified as consumable, such as neighbors who are socioeconomically diverse, close-knit identity groups, proximity to cultural events, or open-air spaces. But unless we think that acting on preferences will impose too many negative effects, there seems little reason to interfere with allowing individuals to sort themselves residentially along the lines of preferences. Notwithstanding my embrace of the positive effects of constituent preferences, it would be inappropriate to base all local policies on that principle. The conditions required for a pure Tiebout world are, of course, fanciful. People are constrained by employment opportunities and wealth in deciding where to reside. People have incomplete information concerning variations among localities. Local action does produce negative external effects. Some commentators fear that interlocal competition for mobile capital will generate a race to the bottom, as localities ignore the consequences that their fiscal policies have for jurisdictions from which

38

wh y d o es lo cal a uto n o m y m a tte r ?

residents and mobile capital are attracted.8 Nor will Tieboutian signaling work as well as advertised. Exit, when it occurs, provides imperfect, muddy signals. Emigrants do not give exit interviews in which they indicate the extent to which their departure was motivated by dissatisfaction with city services. Many of those most dissatisfied with local services may be unable to move because they are tied to jobs, family, or friends or lack financial resources. Those residents are vulnerable to exploitation, either by being taxed for services they do not desire or by being unable to secure the services they do desire. Even if we could achieve something like a Tiebout world, it is not clear that we would want to do so. There is much to be said about the poverty of preferences and of policies that are committed solely to their satisfaction. Individuals likely do not form their residential preferences after intense, objective consideration and deliberation over alternatives or free of distorting influences. Adaptive preferences or inertia may dictate more residential decisions than deep reflective equilibrium. Residential choices may be path dependent, habitual, or arbitrary. A story in the New York Times after Hurricane Katrina revealed that when New Orleans residents returned from temporary homes in cities that operated more efficiently, they began questioning the services that they had always accepted in their home city because they had theretofore been unaware that those services were so inferior to analogues in other jurisdictions.9 Choices may be further constrained by history, geography, or socioeconomic characteristics of current residents. Some residents will tolerate local characteristics only because they cannot afford their preferred alternatives, suffer from information gaps about available options, or act irrationally in the face of preferences. Where one grows up, what one’s peers consider a good place to live, limitations imposed by employment, and socially constructed values about the desirability of urban or suburban life will all affect residential decisions. Some preferences may be odious, such as a preference to exclude from the community individuals who belong to a group that has historically been the subject of discrimination and that, if permitted, would be excluded by so many localities as to significantly constrain the residential choices of that group’s members. Other preferences may impose too many negative external effects, such as a preference to maximize tax revenues and minimize costs by siting industries that produce both tax revenues and pollution at the local boundary.

w h y d o e s l o c a l a utono my mat t er?

39

As a practical matter, attributing preferences to a locality cannot imply that every member of the entity shares identical preferences or that preferences are perfectly expressed and formed. Even in highly homogeneous communities, residents are likely to disagree over allocations in municipal budgets, ordinances, and governing procedures. Those who prevail may represent those who can best organize rather than those who represent majority preferences or those with the most intense interest in the issue. Localities offer a bundle of services, some of which will be desired by any given resident, some of which will be ignored, some of which will be disfavored.10 The impossibility of disaggregating the bundle of goods and services that any locality offers reduces the likelihood that any part of it accurately reflects constituent preferences. Relatively immobile residents will, by definition, tolerate much more deviation between the local characteristics that they confront and those that they prefer before deciding to vote with their feet. Even if we reject an objective of preference satisfaction, however, Tiebout explains something important about the role of local government. Preference satisfaction may not be everything. But it also is not irrelevant. Most preferences about the provision about local public goods do not implicate controversial positions that generate debate about the permissible scope of local action, such as rights to discriminate on the basis of ethnicity, sexual orientation, or social class. Rather, most preferences deal with more mundane services about which reasonable people can disagree and about which we would likely permit interlocal diversity, such as frequency of garbage collection, tax rates, and school class size. Tiebout explains our inability to rely solely on the market for residence to resolve these run-of-the-mill debates by revealing how far we live from a world in which an unhindered market would do all the work necessary to ensure preference satisfaction. Just as Coase reveals how positive transactions costs can affect the ultimate assignment of a legal entitlement by assuming a world of no transactions costs,11 so does Tiebout explain why politics and law are necessary to smooth out the delivery of goods and services within a locality by positing the unrealizable conditions under which those factors would be superfluous. Assuming (a big assumption for some) that localities should pursue at least some of the preferences of their constituents, the Tiebout model isolates a variety of factors that facilitate that objective. By recognizing the

40

why d o es lo ca l a uto n o m y m a tte r ?

obstacles to achieving that objective, the Tiebout model clarifies the kinds of interventions that are necessary to compensate for market failures. In the world as we know it, some individuals will face fewer obstacles than others to voting with their feet when local goods and services deviate from their preferences. The fact that the former group can exercise an exit option will have implications for local welfare. To the extent that local officials desire to encourage those individuals to remain as residents, their preferences will exercise disproportionate influence on local decision making. The theory that local governments compete for residents by offering different bundles of public goods has two implications for local redistribution. First, it suggests that opportunities for redistribution should be more limited at the local than at more-centralized levels of government. If prospective residents sort themselves according to preferences, then one would anticipate that local governments would be characterized by greater homogeneity than we would see among residents in morecentralized governments. If those who are attracted to the same basket of goods and services, and who can afford to pay the related tax price, are of similar socioeconomic status, then there would be little basis for redistribution within the locality. Schneider does find that homogeneous communities tend to tax themselves more and to spend more on governmental services, but he also finds that redistributive service levels decline with income and more slowly in the highest-income communities.12 That result is consistent with what one would expect from sorting theory, since residents with homogeneous preferences could realize significant economies by allowing government provision of uniformly desired goods, rather than negotiating individual contracts. The result also raises a questionable effect of the sorting theory in that wealthy individuals who prefer living in socioeconomically homogeneous environments will migrate to jurisdictions that have neither the desire nor the need to redistribute wealth, thus imposing more of a redistributive burden on other jurisdictions. That effect, however, can be minimized by redistributing at more-centralized levels of government. But the probability that relatively homogeneous localities are less likely to engage in redistribution arguably supports a presumption that on those occasions when we do see local redistribution, something malign must be going on. That is, the very presence of unexpected local redistribution

w h y d o e s l o c a l a utono my mat t er?

41

suggests that the beneficiaries of redistributive spending have dominated the political process to procure expenditures that are disfavored by most residents. On reflection, however, no such conclusion follows. Indeed, the opposite presumption may be appropriate. The same homogeneity that makes redistribution unnecessary also diminishes the probability that a dominant interest group could arise to force through an agenda that was inconsistent with majority views of local welfare. Since homogeneity entails that residents will hold a common view about local goods and services, there should be only a low probability that a group could emerge that is both sufficiently idiosyncratic to place it at odds with the balance of residents and sufficiently influential to capture the local legislative process. Thus, when we see local redistribution, we might actually presume that it is consistent with local preferences. Local redistribution to the relatively poor within a homogeneous community may reveal support for residents who share the preferences of payers but who need assistance in obtaining the mutually preferred goods and services. Payers may be more willing to support these services because the beneficiaries of redistribution will presumably not seek to alter the character of the municipality. Instead, they seek only to take advantage of the bundle of public goods that the locality offers. The second implication of sorting theory is that there is no reason to exclude a preference for local redistribution from the variables among which prospective residents might sort themselves. Initially, the Tiebout model appears to underpin the orthodox theory about the limited role of local redistribution. Redistributive localities will lose in the interlocal competition for residents and mobile capital simply because net payers can exit to alternative locations (although it is plausible that net payers who wanted to live in a redistributive locality would gravitate to a locality that held itself out as being redistributive). Different individuals might have different preferences for redistribution, just as they may have different preferences for publicly funded education, public parks, high-density living, proximity to shopping or green spaces, or any of the other items in the basket of goods and services that localities aggregate. Certainly if we look at patterns of redistribution, there appears to be significant variation among localities. Sharp and Maynard-Moody report that their study of 250 cities revealed that “per thousand” residents, expenditures for welfare, net of intergovernmental transfers, were zero in 56 percent of the cities. Within the remaining cities, expenditures per thousand residents

42

why d o es lo cal a uto n o m y m a tte r ?

ranged from $16.90 to $148,646.13 If we believe that localities will tend to pursue the objectives of their constituents, that variation suggests that some localities expressly embrace a reputation for being more redistributive than others and that potential residents use that variable in selecting among jurisdictions. Some of that variation may be attributable to different fiscal capacity rather than to any differences in preferences. That is, even if all localities had homogeneous preferences for redistribution, wealthier localities would be better able to satisfy those preferences than poorer ones. Peterson, for instance, found that median family income and property value—his proxies for fiscal capacity—were more strongly correlated with local expenditures on redistribution than, for instance, need or political power (though the next chapter discusses Peterson’s narrow conception of redistribution).14 Sharp and Maynard-Moody find support for the fiscal capacity explanation when they use per capital total spending as the predictor. But fiscal capacity, at best, explains the “how” of local redistribution. It tells us that wealthier localities that desire to redistribute have greater ability to do so and thus might explain variations if all localities had similar preferences for redistribution. Fiscal capacity does not answer the “why” of local redistribution and thus cannot disprove that there is variation among localities in the preference for redistribution or explain variations from the conventional wisdom about the consequences of municipal benevolence. Certainly nothing in the wide variance reported by Sharp and Maynard-Moody suggests that the relationship between fiscal capacity and redistribution levels is sufficiently direct to transform capacity into the primary explanatory force for local redistribution. Fiscal capacity is, moreover, most relevant to gauge the level of redistributive expenditures that are funded through municipal budgets. But as living wage ordinances demonstrate, local redistribution can also take the form of mandates that are funded by targets of the regulation rather than through direct municipal funding. Recall that some living wage ordinances apply to the locality itself, either by requiring the locality to pay mandated wage rates to its own employees or by requiring contractors doing business with the locality to pay those rates, which presumably will raise the costs to the locality of doing business with the contractors. Living wage ordinances of that type may allow some inference about the relationship between fiscal capacity and redistribution. Other living wage ordinances

w h y d o e s l o c a l a utono my mat t er?

43

apply only to private employers and do not have direct effects on municipal expenditures. (Of course, if the critiques of living wage ordinances are correct, even mandates on private employers will have negative implications for future municipal tax revenues.) One might expect, therefore, that if fiscal capacity determines the level of local redistribution, localities with high fiscal capacity that adopt living wage ordinances would enact the versions that apply to both public and private employers, while those with low fiscal capacity would adopt the more restrictive versions. One would also expect to see relatively poor localities enacting other regulatory mandates with redistributive consequences, such as linkage ordinances that require builders to contribute to municipal trust funds for affordable housing, and programs that restrict condominium conversion or that require developers to set aside a percentage of units for low- or moderate-income families. I am not aware of any work that tests these relationships. Nevertheless, even the plausible explanations for why relatively homogeneous localities would embrace redistribution do not support a presumption that all local redistribution is benign. The relative homogeneity of cities is just that—relative. Cities may be more homogeneous than the states of which they are political subdivisions, but constraints on mobility and other failures to realize the idealized assumptions of the Tiebout model mean that the preferences among residents of any sizeable locality will fall within a substantial range. As a result, even within large cities, there is likely to be substantial variation in preferences and willingness to pay for a variety of local public goods, as evidenced by the efforts of residential associations, business improvement districts, users of private protective services, or private schools to exceed the governmentally provided baseline of government services.15 The significant degree of variation in preferences in large cities suggests that local redistribution may reflect more than just efforts by local residents to subsidize those who share their service preferences but who cannot afford the cost. Finally, the Tiebout model provides one of the strongest justifications (though, again, not a motivation) for at least some forms of local redistribution. The model begins with an assumption that the local government is a monopolist with respect to the bundle of public goods that it provides. Those who vote with their feet do so because they cannot otherwise obtain the bundle they desire; the public nature of the goods limits their availability in private market transactions. Outside a pure Tiebout world,

44

wh y d o es lo cal a uto n o m y m a tte r ?

the immobility of some constrains their ability to attain their preferred bundle of goods and services, or their ability to afford even the marginal cost prices that monopolist government providers presumably would charge for basic services. If this same group also lacks the political power to compensate for its lack of mobility in advocating delivery of a preferred service package, then perhaps there is a rationale for intervention so that otherwise immobile individuals can satisfy their preferences. If the function of local government is to compensate for market failure, and if individuals choose to live in a specific locality in part to take refuge from the failures of the private market, then it would be somewhat perverse to permit a locality to offer disparate levels of the same public goods to different members of its constituency depending on their ability to pay. Arguably, such a municipality will have abdicated the very reason for its existence, that is, to solve market failures. Those who find themselves unable to register satisfaction with service levels through mobility or politics may or may not prefer the bundle of public goods offered by the locality in which they reside. But the fact that they have settled within a community that offers a particular bundle of goods and services makes it somewhat anomalous to permit the locality to deny them access to parts of that bundle. If residents who have sought refuge from the failure of the private market through the collective service provider can neither receive those services nor relocate to a more cordial jurisdiction, then some ameliorative measure is necessary. This is the effect of the “equal services” doctrine of local government law, which imposes on localities the obligation to make a service available to all comers on a nondiscriminatory basis. But outside the Tiebout world, “equality” cannot be achieved by allowing all residents the equal right to pay for essential services where the capacity to pay is itself distributed unequally. While centralized redistribution, perhaps in the form of cash payments, might similarly enhance the ability of individuals to migrate to the local jurisdiction of their choice, local redistributive programs may be justified to ensure access more closely targeted to the bundle of goods that a locality offers, and thus to promote the efficiency-enhancing benefits associated with the Tiebout model. This is not to say that localities must offer any particular service. Rather it is to say that once a locality does provide a specific service it must do so in a manner that presumably would have resulted if the Tiebout ideal were attainable, that is, it should provide the service to all who request it.

w h y d o e s l o c a l a utono my mat t er?

45

A sorting theory of local government, therefore, provides some explanations or justifications for local redistribution. But, of itself, the theory tells us little about the motivation behind local redistribution. Moreover, if there is a plausible argument that local redistribution can follow benign or malign patterns, justifications for local redistribution do not provide a means for determining when we have either one. Thus, the question remains, why do localities engage in redistribution? LOCAL AUTONOMY AND LOCAL DEMOCRACY

Achieving either the democratizing or the efficient service-delivery objective of local government places a premium on the local decision-making process. The democratizing consequences of local self-government can be realized only if two conditions are satisfied. First, individuals must have opportunities to engage in the type of political participation and monitoring of public officials that self-government assumes will materialize. If venues in which residents engage each other and their political representatives do not exist or cannot be accessed, then neither the educational nor the political advantages of democratic governance can materialize. Second, individuals who participate in the political process must be able to influence the outcome. This does not mean that individuals get the quality and quantity of local public goods and services that they prefer. As I suggested in the previous chapter, democracy assumes winners and losers. It demands only that those who lose have an opportunity to play the game on the same basis as winners. General-purpose local governments provide a wide range of services—education, sanitation, public safety, entertainment, leisure, and transportation, to name a few. Nothing in my argument suggests that any of these is provided at a level that all residents would endorse. As I develop in later chapters, a major strength of cities lies in their diversity, and that diversity will necessarily translate into heterogeneous preferences about which public goods and services the city should provide. The demands of different groups within the city will necessarily generate trade-offs among preferences and compromises that fail to reflect the budgetary allocation desired by any single group. Moreover, the polycentric nature of municipal budgetary decisions complicates the allocation that might be demanded with respect to any service looked at individually. More spending on police may reduce the demand for spending on fire protection. More spending on educational programs

46

w h y d o es lo ca l a uto n o m y m a tte r ?

may entail less spending on parks. In this context, a well-working political process means little more than that residents with different objectives have an opportunity to participate in the inevitable intergroup bargaining and thus to have meaningful input into the ultimate deal that is struck. It is less important that each group wins an equal or even pro rata share of the municipal budget or of municipal decisions. It does mean that the competing views are presented and addressed, even if not adopted. The relationship between efficiency or preference-satisfaction rationales for local government and political process is somewhat more complicated. At first, any such relationship would appear oxymoronic. As I indicated above, in a pure Tiebout world, politics become essentially irrelevant. Sufficient signals are sent through the migration of atomistic individuals that collective expressions of preferences are superfluous. There is no need to contact local officials or to participate in robust political debate in order to achieve one’s preferences for local public goods. Instead, one can costlessly move to a jurisdiction that offers the preferred bundle. Indeed, for those who value the democratic aspects of local government, the absence of politics in the Tiebout model is one of its most significant detractions. Outside of that “romantic” arena, however, some alternative mechanism is necessary to register collective preferences and translate them into an optimal basket of goods and services. In Albert Hirschman’s felicitous phrase, voice, the expression of dissatisfaction with particular goods and services, may substitute for costly exit. Voice may take the form of individual expression to peers or to those in charge of the decision that the speaker opposes. Or voice may take the form of collective action through politics. But if voice in general, and politics in particular, are intended to serve as alternatives to the signaling effects of exit, their effective use requires that the signals be transmitted and received in a manner that informs local officials of their content and induces those officials to react as they would to similar messages conveyed by perfectly mobile residents who vote with their feet. If the political process fails to amplify or distorts those signals, then neither the efficient provision nor the sorting that underlies Tiebout can occur. Transmission of signals does not necessarily mean that local officials treat them as instructions to be executed under all circumstances. There is something to be said for the conception of representatives who, by virtue of

w h y d o e s l o c a l a utono my mat t er?

47

their deliberations or access to information, “refine and enlarge the public views by passing them through the medium of a chosen body of citizens, whose wisdom may best discern the true interest of [their constituents].”16 I briefly take up in chapter 4 the possibility that political officials may legitimately advance the interests of the locality even where doing so entails deviating from constituents’ preferences. For the moment, it is sufficient to note that refinement or enlargement of (even where it translates into deviating from) constituents’ views is not the same as ignoring those preferences. As I have suggested above, the positive argument that those who can afford to do so will sort themselves among various local governments according to preferences does not necessarily recommend that objective for local governments. Nevertheless, it would be odd to contend that those who seek an alternative objective for local governments can ignore the implications of their decisions for mobile residents who disagree. Officials who seek to calculate the consequences of deviating from constituent preferences, or of altering those preferences, must start with knowledge of the collective preferences that exist. LOCAL AUTONOMY AND LOCAL POLITICAL PROCESSES

Implicit in all justifications for local autonomy, therefore, lie some assumptions about a political process in which different perspectives about desired local public goods and services are effectively registered. Local autonomy assumes that a subgroup within the locality at least has an opportunity to have proposals adopted, either by building a majority coalition with other subgroups or by convincing local officials that acceptance of its proposals will benefit the community at large. What this position requires is a local decision-making process in which competitive proposals can be presented and in which choice among them is likely to be made by reference to local interests rather than by means of something that distorts local interests, such as the organizational ability of proponents or the capacity of proponents to support personal objectives of officials. Redistribution and externalization of costs The notion that those affected by local policies should be able to register preferences in a meaningful way implicates one particular manner in which local redistribution may be problematic. I have been speaking to this point as if redistribution occurs intramurally. But redistribution may

48

wh y d o es lo cal a uto n o m y m a tte r ?

also involve interlocal arrangements, and where that possibility arises, local officials may serve their constituents’ interests too well, at least from a broader social perspective. Local officials have incentives to favor programs that confer local benefits but impose costs on nonresidents. Local officials have less incentive to consider the costs of local activities that are borne by nonresidents, or to balance those costs against the local benefits of the activity. A locality, for instance, may attempt to capture the tax benefits of polluting industries while ignoring the costs imposed on neighbors. Indeed, strategic localities will attempt to exploit the different incidence of costs and benefits by siting undesirable land uses at the local boundary. Even where the internal benefits exceed the external costs, there is a serious issue about whether local residents are entitled to be subsidized by nonresidents, or whether some transfer payment to nonresidents who receive no benefits from the local action is appropriate. The problem is exacerbated where external costs exceed local benefits, so that even a transfer payment would not cure the inefficiency of the local program. As one might expect, common legal constraints on local activity deter the strategic imposition of external costs. There are few examples of state restrictions on local autonomy that explicitly state thou shalt not impose external costs, although mandates to provide extraterritorial services in a nondiscriminatory manner could qualify. Indeed, Richard Briffault has demonstrated that localities have long benefitted from express grants of extraterritorial authority, so that one might infer that the greater fear is that localities with such power will otherwise suffer the imposition of costs by neighboring jurisdictions.17 But for the most part, legal doctrine allocates governmental responsibilities in ways that require a jurisdiction to internalize the costs and benefits of its activities. When municipal action threatens significant external costs, approval from some morecentralized government is usually required. Consistent with those principles, courts frequently invoke the presence or absence of external effects to determine whether the legislature or constitutional drafters intended to permit a particular local activity or to allow state law to trump local ordinances where the answer is ambiguous. Similarly, state legislatures often explicitly preempt the threat that unhampered local governments will externalize costs by limiting the capacity of localities to deal with issues such as pollution or local zoning at the urban fringe.

w h y d o e s l o c a l a uto no my mat t er?

49

Concerns for extraterritorial effects are occasionally incorporated into debates about local redistribution. During living wage debates, for example, opponents have frequently contended that enactment of the proposal would affect wage rates in surrounding jurisdictions. The Supreme Court of Louisiana concluded that wage regulation is a matter of statewide concern in part out of an expressed concern that the New Orleans living wage ordinance would disrupt state labor markets. The court noted that the state legislature had made specific findings that “local variation in legally required minimum wage rates would . . . be detrimental to the business environment of the state and to the citizens, businesses, and governments of the various local jurisdictions as well as the local labor market.” The legislature, the court believed, thus concluded that preservation of business competition and employee stability mandated “a uniform environment with respect to minimum wages rates.” Other local redistributive proposals have been subjected to similar responses. In holding that an ordinance enacted in Telluride, Colorado, that required developers to set aside a certain amount of housing for relatively low-income individuals conflicted with, and was thus subordinate to, state law, the Colorado Supreme Court declared the state’s interest in ensuring uniform access to statewide residential markets, and decreed that Telluride’s ambitious experiment “can change the dynamics of supply and demand in an important sector of the [state’s] economy—the housing market.”18 The externality constraint on local action, however, only weakly explains legal resistance to local redistribution. Virtually any local action will have some external consequences. Road construction within one locality will alter traffic patterns elsewhere, just as will the placement of shopping malls at the urban fringe. The degree of impact, however, may vary dramatically. One might imagine that higher wages in New Orleans would attract workers from other parts of the state as long as higher living costs in that city did not erase the wage gains. Employers in other jurisdictions may be obligated to increase the wages of employees whom they did not want to lose. But would that effect be more like a locality imposing pollution costs on a neighbor, or more like the negligible wear and tear of limited road repair on the tires of nonresidents? The court didn’t say; none of its conclusions referred to findings about the significance, as opposed to the presence, of an external cost.

50

why d o es lo cal a uto n o m y m a tte r ?

For the most part, however, state legislatures have not expressed concern about the extramural consequences of local intervention into market prices, and some state legislatures have explicitly permitted local adjustments to state minimum wages.19 Typically, as in the pollution example, the concern for external effects lies in the fear that unconstrained localities will enact measures that generate local benefits but impose external costs. But local redistribution to the poor is different in ways that suggest that such ordinances will produce fewer negative external effects than the pollution analogy suggests. The locality that pollutes may be able to secure jobs and economic development by attracting polluting firms while shifting many of the costs to downwind neighbors. But even if local redistribution to the poor raises costs to other jurisdictions, the redistributing locality cannot escape the bulk of the costs that its largesse confers. As the literature on “welfare magnets” suggests, redistribution tends to attract more of the relatively poor, who are not typically in high demand as local residents. As a result, even the subsidizing locality that externalizes some costs is likely to be internalizing the majority of costs, as well as the benefits, of its programs. The redistributing locality thus appears to occupy a superior position to compare the costs and benefits of any proposed redistributive program. Although the locality will still ignore the costs it externalizes, the fact that it simultaneously internalizes much of the cost related to its program suggests that its interests are pretty well aligned with those of neighboring localities. Indeed, recall that the orthodox theory predicts that localities will not redistribute at all, not too much, since redistributive expenditures threaten to attract indigent households that require expenditures in excess of the tax benefits they confer. The concern for externalizing costs, therefore, does not seem to be the major obstacle to the enactment or validation of local redistribution. Redistribution and internal political processes The story of malign local redistribution, then, must primarily depend on the possibility that the local political process will be defective in ways that permit officials to redistribute wealth in a manner inconsistent with the interests of their own constituents. As Robin Einhorn suggests, latenineteenth-century municipal theorists feared that municipal forays into redistributive ventures would necessarily entail redistribution motivated by political power rather than by municipal interest. That does not mean

w h y d o e s l o c a l a utono my mat t er?

51

that mere errors by local officials in calculating municipal interests justify external constraints on municipal autonomy. If officials pursue local welfare, local autonomy entails a right on the part of local officials to “get it wrong,” as long as local residents bear the cost of their misjudgments. This is the lesson of John Stuart Mill, who, notwithstanding antipathy toward policy making by decentralized governments that would attract “officers . . . of inferior qualifications,” concluded that “localities may be allowed to mismanage their own interests, but not to prejudice those of others.”20 Mill’s view reflects a belief that local constituents are best positioned to determine the proper scope of local action. Should they determine that local officials have acted inappropriately, redress lies within the realm of local political competition rather than appeal for centralized regulation of local activity or judicial invalidation of municipal initiatives. Even malign redistribution, one might argue, should remain a matter of local concern, its scope to be determined by local political processes, as long as its effects are limited to the locality itself. Local officials may miscalculate the costs and benefits of redistribution and enact programs that turn out either to generate excessive costs or not to generate the expected benefits. But miscalculations of that sort are inherent in the speculative business of attempting to formulate a local strategy. But Mill’s appeal to electoral correctives itself assumes a working political process. As decades of study of the problem of collective action have made clear, strategic behavior by groups of residents can distort the decisions made by public officials so that the goods and services are provided in a manner that is preferred by those distinct groups but not by residents as a whole. This phenomenon occurs because those who would benefit from the redistributive effort find it worthwhile to organize and lobby for its enactment, while those who would pay for the program would face individual costs too great to justify their opposition, notwithstanding that aggregate costs of the program exceed aggregate benefits. To say that a municipality is acting according to residents’ preferences essentially excludes decision-making processes in which dominant interest groups defeat residents who have an interest in the allocation of goods and services but who are, as a practical matter, precluded from participating in the debates about the ultimate allocation of local goods and services. These imperfections of democratic governance cannot be eliminated. But, perhaps one role of legal doctrine is to recognize their existence

52

wh y d o es lo cal a uto n o m y m a tte r ?

and to restrict their consequences. I do not propose legal intervention to second-guess the substantive redistributive decisions of local governments. Doing so flies in the face of the market for residence that underlies the argument for local autonomy. But because the success of that market depends on achieving residents’ preferences, there is a stronger case for legal intervention where the political market fails by virtue of the dictates of dominant interest groups. The difficulty lies in distinguishing those situations from situations characterized by a working political market. If situations in which there is a greater risk of divergence between the conduct of officials and the preferences of constituents share observable characteristics, then it might be appropriate to invite more rigorous legal scrutiny of local action when those characteristics are apparent.

ch ap t e r t h re e

The Meaning and Scope of Local Redistribution

government statistics on local redistribution and the orthodox theory are both concerned with the reallocation of resources from the “haves” to the “have nots.” But local redistribution looks even more pervasive once we include redistribution to the relatively wealthy. The orthodox theory should apply with equal force to municipal expenditures that favor subsidies for the wealthy and that confer direct benefits on few residents. Although municipal subsidies for sports stadiums, big-box retailers, or industrial firms are typically defended on the grounds that they return net financial benefits to the sponsoring locality by increasing local economic activity, employment, and tax base, even if those benefits materialize, they will not be shared equally throughout the community, and some residents will suffer net harms as a consequence. Just like residents who are net payers of subsidies to the relatively poor, residents who are net payers of subsidies to the wealthy may still enjoy access to shopping, employment, or entertainment in the city if they migrate to outlying areas. In these cases as well, therefore, the orthodox theory predicts that those who pay redistributive taxes in excess of personal benefits will exit to less redistributive jurisdictions. Nevertheless, rough estimates of state and local government expenditures for economic development fall in the range of $20 to $30 billion annually.1 Obviously, the scope of local redistribution varies with the definition of the term. Virtually any legislation has some redistributive effect to the 53

54

th e mean in g an d s c o pe o f l o c a l r e d i s tr ib ut io n

extent that it imposes on an individual taxpayer a personal cost in excess of the value that the taxpayer receives from the expenditure. Thus, municipal funding of a homeless shelter would qualify, since the immediate beneficiaries will likely not have paid much toward the service. But so would municipal funding from general tax revenues of a public golf course frequented by a small percentage of the local population, or a tax abatement intended to induce an employer to migrate into or remain within the jurisdiction. Even a road-paving project that all residents recognize as beneficial to the locality as a whole could still be classified as redistributive if some residents paid individual costs in excess of individual benefits. But to classify the provision of widely used and desired local public goods as redistributive simply renders the question of whether local governments should engage in redistribution meaningless. Attempts to provide a more precise definition of redistribution are inherently problematic. Take, for instance, the definition used by Paul Peterson in his influential book City Limits. Peterson limits the term in two ways. First, he restricts it to redistribution from the better off to the less well off segments in the society. Thus, he excludes tax expenditures and abatements that are intended to promote local economic development. Second, he includes only redistributive policies that have negative effects on local economies. Thus, he excludes development policies that assist the poor but that have net positive local effects, such as job retraining programs or programs that attract needed low-income workers.2 I include “redistribution” both to the wealthy and to the poor and thus reject Peterson’s first restriction. The concern of the orthodox theory is not simply that wealthy taxpayers will exit from redistributive jurisdictions; rather, the theory contends that any mismatch between benefits and burdens causes residents who make net tax payments to emigrate to jurisdictions that impose only benefit-based exactions. While this scenario is typically associated with redistribution to the relatively poor, a resident who obtains insufficient benefit (financial, psychic, civic, or otherwise) from a municipally funded sports stadium will be no more enamored of paying taxes for that project than the resident who objects to municipal funding of low-cost housing. Since my argument concerns the capacity of groups to redirect municipal expenditures in ways that exacerbate the gap between the services a locality provides and those its residents prefer, it would unduly restrict the analysis to omit expenditures

t h e mean i n g a n d s c o pe o f l o c a l redis t rib ut io n

55

that benefit wealthy groups. Indeed, those groups are likely to exercise considerable political power relative to the poor, and thus they may be able to distort expenditures more than advocates for the poor. Under Peterson’s second restriction, a policy is redistributive only if it has a negative impact on the local economy, notwithstanding that it entails transfer payments from the wealthy to the poor. (Note, however, that a municipally funded subsidy for a new automobile plant that costs the locality amounts in excess of the economic benefits conferred would still not qualify as “redistributive” for Peterson because the expenditure did not flow from the wealthy to the poor.) Under Peterson’s rationale, a living wage ordinance would constitute a developmental, nonredistributive policy and thus be appropriate for municipal adoption, if it ultimately increased worker income without increasing unemployment, and if it induced productivity gains that also benefited employers. But if the same policy reduced employment and generated no productivity gains, so that it constituted a wealth transfer from employers to employees and shrank the ranks of the employed, it would be redistributive and undesirable. Peterson’s second restriction implicitly recognizes that some limitation is necessary, lest every municipal activity be classified as redistributive. Peterson’s restriction to projects that generate net costs seems appropriate insofar as it excludes those projects that, from a municipal rather than an individual perspective, fit the benefit-based municipal expenditures that stand as the natural alternative to redistributive expenditures. Few, for instance, would classify as redistributive a municipal project that is undertaken to provide goods that residents actually desire but cannot obtain because of collective action problems or some alternative market failure, notwithstanding that government provision of the goods entails transferring wealth to (that is, buying them from) private providers, such as private contractors who perform road paving. Nevertheless, Peterson’s restriction to projects that return net benefits to the locality is implausible as a tool for analyzing the propriety of proposed municipal activity. We want to know whether the municipality is engaged in a legitimate activity when the proposal is adopted. At that time, however, we often can’t know the effects of a proposal on the local economy. The sports stadium may turn out to fuel the creation of adjacent businesses that swell the municipal coffers with tax revenues. It may turn out to be nothing more than a sop to sports team owners, as dollars spent

56

t he mean in g an d s c o pe o f l o c a l r e d i s tr ib ut io n

for tickets simply substitute for alternative entertainment expenditures in the locality. Indeed, it may be that even ex post the net effects of the project are contested. The empirical work to date on living wage ordinances illustrates the point. Studies of economic effects of the ordinances reach inconsistent and controversial conclusions, made more ambiguous by the apparent political associations of some of the researchers. One study of the effects of the Santa Fe ordinance concluded that the introduction of a living wage increased unemployment (+3.2 percent), especially for the least skilled and least educated workers, and decreased hours for those able to keep their jobs by 1.6 hours per week.3 That study was countered by others that purported to show that the ordinance generated little or no negative impact on employment levels.4 In a series of papers, Scott Adams and David Neumark found that living wage ordinances increased both the income of low-wage workers and unemployment among the least skilled workers, though these effects appear to vary with the scope of the ordinance and the presence of similar ordinances in neighboring localities.5 In testing the effects of living wage policies within the confines of San Francisco International Airport, however, Michael Reich, Peter Hall, and Ken Jacobs found no unemployment effect. Instead, turnover rates declined, worker performance and morale improved in quality, and absenteeism and disciplinary issues decreased.6 Robert Pollin, who has been active in numerous living wage campaigns, summarized his research in a 2005 article that predicted living wage ordinances would cause relatively small modifications for firms or adopting cities. He concluded that the effect of living wage ordinances on most covered firms amounts to be between 1 percent and 2 percent of their total production costs or sales, and he hypothesized that firms would absorb the costs with relatively modest adjustments, such as by raising prices and productivity or by redistributing the firm’s income downward in a minor way.7 The ex ante uncertainty as to whether a municipal policy will generate net benefits undermines the utility of that standard as a basis for classifying a proposal as redistributive. We might condemn local decisions that, at the time of implementation, are predicted to have little chance of achieving any reasonable municipal objective. But it is unhelpful to say of a local action that initially constituted a reasonable gamble that it was

t h e mean i n g a n d s c o pe o f l o c a l redis t rib ut io n

57

valid or appropriate if it turns out to generate net benefits but invalid if it turns out to generate net costs. Local officials, or their constituents, may come to regret having made the gamble. But, assuming that their decision was made through a process that allowed reliable prediction and evaluation of the probable results, regret does not transform the validity or propriety of the initial decision. I use the term “redistributive” to refer to any local policy that confers on a subgroup of residents benefits that are substantially disproportionate to the related local costs that the same subgroup bears. That definition denominates the conditions under which the strategies of dominant interest groups with which I am concerned are likely to arise. For example, expenditures for road paving would presumably not qualify as redistributive because most residents sufficiently use or benefit from well-maintained roads, and individual costs do not vary substantially from individual tax payments dedicated to road maintenance. The fact that some residents use roads more than others would not transform the expenditure into a redistributive one, on these terms, because those who do use the roads constitute a sufficiently large group and bear a sufficiently large proportion of the costs, and nonusers face little risk of the kind of exploitation that underlies concerns about political process. If, by contrast, only a single road in the municipality was paved, but the costs were borne by all residents, it would be easier to classify that expenditure as redistributive. If that same road project was funded through a special assessment imposed on abutting landowners, the project again would likely not fall within the redistributive category. A municipally financed homeless shelter would qualify as redistributive, since those who pay the costs receive few of the direct benefits related to its provision. A municipally financed sports stadium would similarly qualify as redistributive on the assumption that stadium owners receive municipal benefits far in excess of any burden they bear, even if their activity generates net benefits to the locality as a whole. Spillover benefits may also limit the range of redistribution under my definition. Educational expenditures from the general treasury may not qualify as redistributive, notwithstanding that those families with children in the public school obtain benefits not shared by families without children, because the latter enjoy significant benefits to the extent that (1) education costs are capitalized into the value of homes and businesses, and (2) education provides a better workforce

58

t he mean in g an d s c o pe o f l o c a l r e d i s tr ib ut io n

for residents as a group. Whether we should apply the term “redistributive” to the expenditure, therefore, depends on whether the unique benefits obtained by families with children in the system are significantly disproportionate to the costs that they bear, but the answer to that question does not depend only on whether a particular taxpayer has children within the system. I limit the disproportionate costs to “local costs,” however. If a more centralized entity is willing to subsidize the costs of a project, as in the case of a state or federal grant to a locality, then the typical expectation underlying the grant is that some local group (perhaps the locality as a whole) will obtain a benefit that it could not otherwise enjoy, but that the project produces sufficient spillover benefits that it is worth undertaking from the broader social perspective. The analysis of that project entails fascinating issues of fiscal federalism but does not entail my concern about distortions from local preferences. Indeed, if the locality can obtain “free” money from the central government, it more likely would be in the interest and consistent with the preferences of local residents to pursue it. My concern here lies with the unique and controversial role of intralocal redistribution, so I exclude examples of extramural funding for local projects. The problems of collective action to which I alluded in chapter 1 reveal the relationship between this definition of redistribution and the incentives of dominant interest groups. That discussion suggested that residents burdened with individually small payments for redistributive expenditures are unlikely to incur the time and costs required to register their objections, while those who will benefit disproportionately to their personal costs have significant incentives to advocate the expenditures. Thus, the imposition of disproportionate costs, and a definition of redistribution that is characterized by that factor, sets up the very scenario that I want to explore: the possibility that local decision makers will face a set of interests that monopolize the debate about a proposed expenditure, not because those interests are shared by a majority of local residents, but because they represent the views of those who have organized to participate in the debate. Suspicions about the cross-subsidies that this definition of redistribution entails are reflected throughout the history of American municipal finance, though much of the municipal largesse involves subsidies

t h e mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

59

to the wealthy. During the period of nineteenth-century city boosterism, municipal expenditures tended to be concerned primarily with internal improvements necessary to attract business and with protective services such as police or fire prevention. As in the case of prodevelopment subsidies today, city officials justified these subsidies largely on the theory that all residents would benefit from the ensuing growth, and thus all should contribute toward it. Prodevelopment strategies were consistent with the vision of municipal corporations as essentially private entities chartered by the state to fulfill specified obligations. Development was consistent with a conception of municipal government occupied solely with “public” functions that would not commingle public and private provision of goods and services. Failure to respect the division between public and private left the city vulnerable to the charge that rather than advancing the public good, officials were “serving private interests or groups.”8 Government interventions to provide goods and services tended to be justified in terms that today would be classified as part of the allocation function of government that compensates for failures of the private market, rather than redistributive efforts to ensure that a specific service was available even to those who could not afford market prices. Hence, Hartog’s conclusion that New York City’s eighteenth-century municipal government was expected to be neither a competitor nor a direct partner with private enterprise in the development of the city’s economy. To the contrary, its most important function was to create a predictable and consistent environment within which that private-market economy would flourish.9 Any redistributive effect of legislative action that conferred a benefit on private actors would have been considered short term or incidental, since the subsequent increase in tax revenues meant that development projects would ultimately pay for themselves. Certainly that was true with respect to municipal indebtedness incurred to support railroads, defaults on which led to the promulgation of new restrictions on state and municipal expenditures. The strategy was often successful, as in the case of New York City’s municipal waterworks, a project that drastically increased the city’s debt but that became self-supporting shortly after installation.10 The fact that special groups or special interests would disproportionately benefit from these projects produced no barrier to their subsidization. As Oscar and Mary Flug Handlin long ago explained in their investigation of the intertwined relationship of antebellum Massachusetts and its private

60

t h e mean in g a n d s c o pe o f l o c a l r e d i s trib ut io n

businesses, “not all privileges, grants, or monopolies were odious. . . . So long as that general interest of the community was the object in view, aid to special groups was beneficial to all.”11 Subsidies during this period did not necessarily take the form of expenditures from the public treasury. Rather, governments subsidized firms by providing them with security from market competition or liabilities that would otherwise attach. The efforts to shore up the state economy by supporting private enterprises included grants of franchises to utilities, such as toll bridges or ferries, that agreed to serve all comers at regulated prices; grants of immunity to millers whose activities inundated adjacent land, but who were subject to price controls; and, notwithstanding limited state funds for direct assistance (“bounties”), occasional loans and tax abatements to selected industries and firms that encouraged commerce. Indeed, government largesse in the form of grants of privilege became more entrenched to the extent that recipients contended that any subsequent interference with the initial monopoly, even when competition could redound to the public benefit, was constitutionally prohibited. It took judicial intervention to reject allegedly implicit grants of exclusivity or to uphold reservations of legislative power to alter or repeal corporate charters when necessary for the common good.12 This is not to say that the state was obligated to intervene in the economy. In the mid-nineteenth-century era the Handlins describe as represented by “liberal humanitarianism,” concern for state-conferred privilege caused a reaction against grants of state largesse. But the retreat was predicated on the political and economic propriety of grants rather than on some essentialist view that the prior grants had exceeded the scope of governmental competence. Governmental redistribution to the poor, especially at the local level, was a different matter. Until the post-Depression era, municipal redistribution to the poor had been limited and generally followed patterns of aid established under the English Poor Laws. Nineteenth-century state statutes imposed responsibility for poor relief on local governments. Enacted in the early nineteenth century in almost every state, these laws gave local authorities broad discretion to direct relief efforts, the boundaries of which were circumscribed by more-centralized policies. Many statutes delegated responsibility to county commissioners or appointed

t he mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

61

overseers who were authorized to spend local funds and often to institute local taxes.13 These laws largely followed either of two administrative schemes: New England’s township system and Pennsylvania’s county system. The township system frequently resulted in jurisdictional disputes, with towns located in the same county arguing over responsibility for the poor. The county system made the county the primary unit for poor relief, thus eliminating these disputes. Vermont’s approach exemplified the township system. In 1839, Vermont required every town to support their poor. Towns were explicitly allowed to open poorhouses in conjunction with other towns. Tennessee exemplifed the county system. In 1797, Tennessee made county justices of the peace and the county court jointly responsible for the poor, replete with the authority to levy an annual county tax on property to fund benefits.14 Almost all poor laws differentiated between the deserving and undeserving poor. New York’s law, following principles in place since colonial times, made towns responsible for those unable to work and legally residing in the state. The able-bodied poor could be expelled, and begging was illegal. Ohio similarly provided aid to those in a “suffering condition and unable to support themselves.” Kentucky aided those poor who were “by personal disability or otherwise. . . incapable of procuring a livelihood” and prosecuted others for vagrancy, punishable by up to nine months at forced labor.15 Four poor-relief methods were customarily used: direct relief (known as outdoor relief), poorhouses (known as indoor relief), auctions of the poor to caretakers, and use of the poor as contract labor. Outdoor relief comprised a range of aid, from cash payments to in-kind grants. For example, in 1833, Philadelphia distributed $20,598 in cash, $5,525 in wood, and $1,537 in medicine and “leechery.” In 1860, however, no cash was given, while $17,098 in groceries, $13,911 in coal, $4,718 in wood, $1,476 in medicines, and $470 in burials were provided. Poorhouses centralized relief, providing residential services in exchange for labor and good behavior. Towns that had an insufficient number of poor to make an almshouse practical would auction their care to the lowest bidder. The auction system was condemned by those who advocated additional care for the destitute, since those who “won” the auction had incentives to chisel on expenditures from the public outlays they received for the benefit of their charges. Critics of poor laws pointed to abuses of the system

62

t h e mean in g a n d s c o pe o f l o c a l r e d i s tr ib ut io n

in which families bid for their own members and received public funds, thus subsidizing a “modestly comfortable life for dependent people with their kin.” Local committees or agencies who worked on behalf of the poor also negotiated contracts whereby they sold the services of the poor. No combination of these methods was standard, and policies varied widely by region and jurisdiction.16 Financing arrangements for these efforts also varied. In 1823, every county in New York State imposed a poor tax on real estate within its jurisdiction, and many towns included poor relief in their budgets. Philadelphia also funded poor relief from its real estate tax. James County, Virginia, used revenues from its poll tax. Baltimore drew money from city and county funds but lacked a dedicated poor tax. Boston received aid from the state for the care of poor foreigners and paid the remaining relief balance from its own funds.17 Services for the poor had limited reach. For example, in Tippecanoe County, Indiana, the mandated county relief program “served only a minority of the sick-poor.” A study of the county’s public poor relief from 1827 until 1846 found a “basic system” that “combined outdoor relief in the community with indoor relief at the poor farm supervised by a county physician.” The Tippecanoe efforts were funded by a county tax on townships, and expenditures were requested by appointed overseers with nominal oversight from elected county commissioners. Overseers were often politically connected and personally benefited from poor-relief contracts.18 By the 1820s, concern that traditional relief efforts were failing led to the embrace of the poorhouse as a panacea. Poverty was increasing rapidly as wage laborers and factories replaced stable artisans. Workers now moved from town to town and job to job, and many became indigent transients. State-commissioned reports in Massachusetts, New York, Pennsylvania, and New Hampshire noted that poor-relief costs were increasing at a rapid rate. New systems of poorhouses were proposed to reduce public expenses, deter dependency, and eliminate the worst abuses of the old system, specifically the auctioning of the poor and the practice of moving the poor from town to town to avoid relief expenses.19 Advocates praised the poorhouse system’s fiscal advantages. Poorhouses promised to consolidate duplicative outdoor-relief systems and to employ the poor at productive labor. Massachusetts claimed that

t he mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

63

every town with a poorhouse had reduced relief expenses. New York’s report claimed that poorhouses could decrease costs by almost one-third compared to outdoor relief.20 A committee in Newport, Rhode Island, believed a poor farm would be self-supporting due to profits from agricultural cultivation.21 In many cases municipalities appear to have been responsible for the related costs. In 1826, Boston paid $20,892 to support its poorhouse and received only $9,640 in state aid for the support of foreigners.22 A report from the secretary of state of New York indicated that the fiscal burden for improvements in poor relief fell on counties, stating that “the expenses of erecting and completing each house of employment, be paid by the county, and raised by tax.”23 Many of the benefits were premised on the belief that the conditions of poorhouse labor would transform the character of the poor. By banning alcohol and requiring work, poorhouses restricted “the expense of pauperism through cheaper care and by deterring people from applying for relief,” specifically those looking for a handout. As one official from Pepperell, Massachusetts, noted, once in a poorhouse the “lazy and indolent” would “quit their station and shift for themselves.” Poorhouse residents were employed at a variety of agricultural and manufacturing tasks. In one year, inmates in Salem, Massachusetts, produced 4,391 pounds of pork, 1,000 bushels of turnips, and 2,700 bushels of potatoes and also worked at spinning, weaving, and woodworking.24 Some poorhouses attracted private contributions to supplement public funds. The New York City almshouse was funded by a $25,000 lottery run in 1795. Similarly, Burke County, Georgia, ran a $10,000 poorhouse lottery in 1826. In New London, Connecticut, 116 wealthy citizens gave a total of $1,000 for a poorhouse constructed in 1800. A merchant in Providence, Rhode Island, Ebenezer Knight Dexter, donated $60,000 of land to the city in 1824, which was used to build the Dexter Asylum for an additional $43,000 in public funding. In Savannah, Georgia, private lotteries and donations were mixed with public appropriations to run a private poorhouse from the turn of the century until after the Civil War.25 By mid-century, local poorhouses proliferated. Most states passed laws enabling but not requiring localities to run their own facilities,26 though New York and Delaware did mandate county maintenance of a poorhouse.27 Klebaner reports that by 1860, two out of every three towns in Massachusetts had an almshouse, and larger towns were even more

64

th e mean in g a n d s c o pe o f l o c a l r e d i s tr ib ut io n

likely to have them.28 Every state with poorhouse legislation also authorized towns and counties to join together to build and operate facilities.29 Some cities also supplemented poorhouses and relief with specialized institutions. In 1814, New Orleans took over the privately run Charity Hospital and ran it for the benefit of the poor and the insane. This was particularly notable as Louisiana, a civil law state unconnected to the English Poor Laws, lacked a poor-relief law until after the Civil War.30 New York City housed the private City Dispensary in its municipal building and provided 40 percent of its funding.31 Local governments also continued to provide outdoor relief. In 1853, five major eastern cities made significant outdoor relief expenditures. Baltimore spent $1,200, Boston $30,000, New York $82,000, Philadelphia $75,000, and Providence $5,700.32 In Boston, these expenditures constituted between one-fourth and one-third of the total relief expenditures.33 By mid-century, however, reformers targeted publicly provided outdoor relief as corruption-prone and wasteful and promoted its replacement with private charity.34 In the 1850s, new humanitarian relief organizations in New York City openly opposed outdoor relief. Appalled at what they saw as counterproductive politicization, they founded their own private institutions to serve the poor. The Prison Association created a legal-aid service to keep people out of prison, and the Children’s Aid Society developed a new system to place children outside of asylums.35 By the 1870s, these activities coalesced into the scientific charity movement, based on the idea that private philanthropy more effectively aided the able-bodied than outdoor relief. Scientific charity aimed to help the poor without creating dependency. The movement did, however, continue to support the poorhouse, provided, in the words of one prominent reformer, that it instilled a “distinct moral and physical improvement on the part of all those . . . forced to have recourse to it.”36 Exemplifying these twin trends was the campaign by Brooklyn’s reformist Republican mayor Seth Low to end outdoor relief in the city. In Brooklyn, outdoor relief was administered by the Commissioners of Charity and funded by the County Board of Supervisors. The city’s poorrelief benefits were particularly generous, and reformers called Brooklyn “the paupers’ paradise.” According to Michael B. Katz, Low, an independently wealthy man whose grandfather had founded a prominent private charity, was not motivated by pure altruism toward the poor. Rather, he

t h e mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

65

believed that the commissioners’ outdoor relief expenditures provided an opportunity to weaken the Democratic machine. Low’s allies in private charity testified that the commissioners’ program of outdoor relief was “superfluous, wasteful, corrupting, and, under New York Law, illegal.” Reformers further claimed the city’s high tax rates were the result of relief spending and urged the supervisors to end the program. When the commissioners resisted, investigations and allegations were launched to undermine their public standing—including a lawsuit in which it was alleged that the commissioners had inappropriately provided poorhouse inmates with Thanksgiving turkeys. By 1878, outdoor relief in Brooklyn was eliminated and responsibility transferred to volunteer organizations. The voluntarists claimed that no needy person suffered due to the end of outdoor relief. Nevertheless, vagrancy rates increased by 50 percent and the number of children placed with the Catholic Orphan Asylum nearly doubled.37 Still, by the end of the nineteenth century, public relief spending rivaled private charity. In 1899 Frederic Almy, secretary of the Buffalo Charity Organization Society, published a study of public and private outdoor relief in the forty largest cities in the United States.38 In twenty-six cities the per capita expenditure for public outdoor relief exceeded the per capita rate of private charity. Of the remaining fourteen cities, ten had entirely abolished outdoor relief. Poorhouse spending—unaccounted for in Almy’s study—also remained sizable. By the 1890s, the amount spent on poorhouses in New York State greatly eclipsed that spent on outdoor relief. Whereas in 1860 the total spent on outdoor relief in the state was $524,943 and the amount on poorhouses $351,820, by 1890 outdoor relief totaled only $737,671 while poorhouse costs rose to $3,175,775. Spending priorities, however, varied widely by county. In 1890, Erie County served 6,360 people through temporary relief and 2,279 through poorhouses, while King County served none through temporary relief and 11,164 through poorhouses.39 Furthermore, public/private distinctions were somewhat artificial. Katz notes that “throughout the country, relief was, as it usually has been in America, a public/private venture,” with “no sharp line divid[ing] the two spheres.”40 Indeed, cities occasionally appropriated money directly to private charities. In 1860, New York City provided, among other expenditures, $6,000 grants to five dispensaries, $3,972 to the Colored Orphan

66

t he mean in g an d s c o pe o f l o c a l r e d i s tr ib ut io n

Asylum, $1,000 to the Home for Deaf Mutes, and $250 to the Inebriates’ Society.41 By the late 1890s, the city spent more than $3 million on 220 private charitable societies.42 The largest share went to poor children, but “virtually every charity in the city received a handout from the municipality,” including a group dedicated to supporting impoverished actors.43 Outside of New York City, aid was irregular. New Orleans began subsidizing private orphan asylums due to periodic epidemics, and by 1890, spent $22,550 subsidizing private charities.44 No other municipality spent more than $250,000 on private charities, and Boston, Cleveland, Cincinnati, and Milwaukee spent nothing at all.45 By the end of the nineteenth century local governments could choose from an array of poor-relief methods beyond the English Poor Law traditions. Traditional relief was supplemented by vastly expanded poorhouses and newly enthusiastic private charities. For example, from 1878 to 1895, Cleveland was governed by the highly progressive “Populist Regime,” which implemented a range of programs. The city supported poorhouses, outdoor relief, and a city infirmary. Spending for poor relief grew to between $200,000 and $250,000 annually in the late 1880s.46 The traditional limitations on poor relief to safety nets such as poorhouses or in-kind assistance, rather than compensating for market failures through redistributive taxation or the imposition of regulatory mandates, reflect both a vision of the poor as potentially responsible for their lot and a vision of the limited role of municipalities. As Robin Einhorn has argued, nineteenth-century conceptions of municipal finance were dominated by an ethos of matching benefits and burdens. Access to internal improvements was limited to those who paid for them. The principle that the financial burdens of municipal improvements should be borne by those who would reap the related benefits implicitly precluded the commitment of municipal resources to the poor and, in ways that I will suggest should inform contemporary legal interpretation of local legislation, frustrated efforts of interest groups to capture the municipal treasury. As Einhorn concludes, “By refusing to let government redistribute wealth, [nineteenth-century Americans] also refused to let politicians use city money as a source of political power. By keeping taxes low, budgets small, and decision-making power in private hands, they denied aspiring “bosses” access to political patronage. Politicians who distributed few favors, who made few decisions that redistributed

t he mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

67

wealth, could do little to reward friends or punish enemies. They certainly could not build “machines.”47 The primary mechanism by which this ethos was realized was the special assessment. Taxation could be used for expenditures on public goods that were likely to have similar effects throughout the municipality, such as civil defense, payment of officials’ salaries, and construction of public buildings.48 But the benefit principle dictated that residents who received a discrete advantage from public works pay the related costs. Indeed, municipal improvements were initially implemented on the petition of those beneficiaries, whose expressed willingness to bear the related costs signaled the efficiency of the project. Payment through special assessment was perceived not primarily in terms of ensuring efficient allocation of municipal resources; rather, it was perceived as a means of instantiating a conception of fairness inherent in a matching of benefits and burdens. Much of the criticism of special assessments was directed not at the principle that beneficiaries alone should bear the burdens of municipal improvements, but at the fact that the municipality was involved at all in projects that conferred a unique benefit on only a subset of the local population.49 Implicitly, this vision of the benefit-based city was entirely consistent with the orthodox theory, which warns of adverse consequences should municipalities seek to redistribute wealth. This structure ultimately proved inadequate in the face of widespread and serious economic distress. The surge of unemployment during the depression of 1893 overwhelmed local agencies that served the poor. During 1893 and 1894 the jobless rate peaked at 35 percent in New York and at 43.6 percent in Michigan. Unable to provide sufficient aid, municipalities initiated work relief programs. Cities sponsored public works and private associations collected money for local governments to spend. Projects ranged from street sweeping to sewer construction. Mayor Hazen Pingree of Detroit even offered small plots of land to poor families for vegetable gardens, an approach that spread to other cities.50 This experience revealed the weakness of a localized relief system. As Katz writes, “state government did not dispense relief, and city governments, the depression underscored, lacked the administrative capacity to respond effectively to economic crisis. Small, weak, lacking effective central direction, city government nearly collapsed under the urban explosion that transformed America.”51 Of course, adherents of the orthodox

68

th e mean in g a n d s c o pe o f l o c a l r e d i s tr ib ut io n

theory of urban finance might add that cities also had little incentive to display generosity, since any city that exhibited philanthropic tendencies would attract more of the mobile poor. The Progressives of the early twentieth century modernized relief efforts by emphasizing professionalism within agencies that served the poor. A new class of civil servants, allied with the sophisticated urban reforms of trained settlement workers, assumed management of urban welfare programs. In 1916, a professional association of public welfare boards, the National Public Welfare League, was formed, and fifty city boards became affiliated. The emergence of empirical methods and development of sociological data allowed managers to measure agency performance. Cities increasingly affiliated with private ventures. Aid to charity rose 79 percent between 1912 and 1930, and “by the late 1920s, most cities spent three times more money on private agencies than outdoor relief.”52 Cities also built new infrastructure to care for the poor. By 1920, New York City ran sixty-eight Baby Health Stations, which taught mothers about hygiene and child care.53 Public baths were run by almost every large city, and by 1922 more than forty cities operated baths year round. Baltimore had eleven public baths, Boston twelve, Chicago twenty, and New York City twenty-five. Even western and southern cities, including Omaha, Salt Lake City, Dallas, and Mobile, had their own baths. Poorhouses were gradually supplanted by dedicated institutions for the mentally ill, the aged, and orphans.54 Accordingly, municipal governments became essential to public relief efforts. Katz estimates that per capita federal welfare spending rose from $.09 to $.25 between 1913 and 1929, while per capita state welfare spending rose from $.78 to $1.85, and per capita city welfare spending rose from $1.65 to $3.12. During the same period, public relief spending exceeded private charity by three to one. But, again, localized relief could not endure widespread economic upheaval. Katz reports that by the early stages of the Great Depression New York City was able to provide relief to only one-quarter of its unemployed. During the first quarter of 1931, New York cities with more than thirty thousand residents spent more than $15 million on relief, compared to only $4 million in the first quarter of 1929. By 1933, as relief expenditures rose and property tax bases disappeared, almost one thousand local governments defaulted on their debts. Private

t h e mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

69

charities faced a similar decline, as one-third of the private social-service agencies in New York City had closed by 1929.55 The relationship between legal doctrine and the orthodox economic theory of local redistribution reflects this checkered history. A political ideology that restricted the scope of local authority in the late nineteenth century translated into a series of constitutional and judicially imposed limitations on redistributive activity. The failure of publicly funded improvement projects left municipalities without the wherewithal to pay debts that had been incurred for development purposes. The resulting fiscal crises produced widespread acceptance of the propositions that exercise of municipal authority required state legislative approval and that any approval that was forthcoming should be narrowly construed. Whether municipal defaults resulted from political expedience or fiscal incapacity may be a matter of some debate,56 but the legal reaction was clear: doctrines such as public purpose limitations, prohibitions on lending of credit, and restrictions on governmental debt promised to avoid the interplay of public and private enterprises. Judge John Dillon’s famous, or notorious, “Rule,” which required prior state enactment and parsimonious construction of statutes that confer municipal authority,57 was an explicit reaction to prior municipal largesse in favor of private promoters of railroads, canals, and other private enterprises that subsequently failed. Dillon’s Rule was a principle of statutory construction rather than a legislative mandate, and thus it did not itself preclude localities from engaging in any specified activities. Instead, it directed courts to limit the scope of those grants of municipal authority that the legislature provided. Even though the Rule evolved from the failures generated by redistribution of municipal resources to entrepreneurial activity, the doctrine of limited local autonomy was also employed to forestall assistance to the relatively poor.58 Miserly interpretation of municipal autonomy implied that, at the very least, local redistributive efforts had to receive state approval, presumably on the theory that state legislators were more responsible arbiters of municipal fiscal propriety.59 State courts subsequently employed other doctrines to limit local redistribution. While some courts invalidated local rent controls or minimum wage ordinances on the grounds that localities lacked authority to create the programs,60 others pronounced that the local program, even if authorized, conflicted with state statutes, or that the state had preempted

70

t he mean in g an d s c o pe o f l o c a l r e d i s trib ut io n

the field into which the locality had injected itself.61 The home rule movement that emerged in the late nineteenth and early twentieth centuries, and which permitted municipalities to engage in activities without the state’s prior approval in areas of local concern, might have been thought to permit greater latitude to municipal redistributive efforts. But even redistributive ordinances enacted by home rule municipalities have been invalidated as courts have discovered conflict with state statutes or have identified extraterritorial effects that question whether the challenged program was sufficiently “local” to fall within the scope of the home rule grant.62 Courts similarly have used constitutional restrictions on borrowing or expending tax revenues to invalidate municipal expenditures that would finance housing for the poor or slum eradication.63 To a significant extent, however, these judicial restrictions on local autonomy appear to have dissipated, at least with respect to redistribution to the wealthy in the name of local development. More-recent decisions consistently approve subsidies not only to the local poor, but to private enterprises that promise broad-based local benefits, or even to middleclass individuals whose very residence in the city is seen as consistent with the promotion of local welfare.64 In the past several decades, courts have become more tolerant of local subsidies for economic development, to the point of effectively eviscerating constitutional prohibitions on debt or lending of credit to private enterprises in order to subsidize commercial and industrial enterprises.65 Recent opinions reveal semiautomatic acceptance of legislative determinations that proposed programs will redound to the public’s benefit or do not implicate assets of the locality in the event the project fails. Implicit, and sometimes explicit, in these opinions is a view that results that emerge from the political process trump any judicially determined metric of the proper range of municipal activity.66 At the federal level, the Supreme Court’s recent refusal to interject judicial review into local decisions to use eminent domain for economic development, at least where the locality appears to have proceeded according to a “comprehensive plan” that indicates a working political process,67 reveals similar deference to political markets to determine the appropriate scope of economic subsidies. Local redistribution, therefore, appears to be gaining in legal acceptance at the very time that economic theory would seem to counsel against it. Indeed, given the reasonable assumption that individuals who pay local

t he mean i n g a n d s c o pe o f l o c a l r edis t rib ut io n

71

redistributive taxes have significant mobility and that more widespread legal approval of public subsidies for private firms offers mobile capital more locational choices, the admonitions of the orthodox theory seem increasingly relevant. In light of the risk of losing individuals and firms, the expansive view of municipal functions does little to explain the economic puzzle: why would localities engage in redistribution? That answer, I suggest in the next two chapters, is significantly tied to the objectives we have for local governments and the political processes that have evolved to implement those objectives.

ch ap t e r f o u r

Why Do Local Governments Redistribute?

BENIGN AND MALIGN REDISTRIBUTION

the orthodox theory denies both the practical availability and the desirability of local redistribution. Efforts to redistribute allegedly induce exit by individuals and firms who are necessary to the financial health of the locality and attract a disproportionate share of residents who will receive net benefits from redistributive programs. So why do localities redistribute wealth? I suggested in chapter 1 that there are both benign and malign answers to that question. In this chapter, I expand on that distinction, discuss some of the explanations that have been offered for this phenomenon, and elaborate on their credibility. I have also indicated that to the extent that decisions to redistribute wealth are made through a process in which advocates and opponents of redistribution are represented, redistributive policies presumptively reflect the interests of residents, including interests that the orthodox theory improperly ignores. A local political process that is dominated by either the beneficiaries or the payers of redistribution, by contrast, serves as evidence of something gone awry that permits divergence between the interests of local residents and the level of local redistribution. Redistributive expenditures that reflect such a distortion of local political markets cannot easily be reconciled with the objectives of local autonomy. As a result, we might be skeptical that the presence of redistributive expenditures reveals flaws in 72

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

73

the orthodox theory rather than in the political process. Local expenditures that emerge from defects in the political process might be better described as examples of malign redistribution in that they fail to serve any reasonable conception of municipal interest, as opposed to the personal objectives of their proponents. It is important at the outset, however, to emphasize some caveats to this oversimplified dichotomy of local redistribution. First, I do not use the term “benign redistribution” to signify only those programs that reflect the aggregate preferences of individual municipal residents. I indicated in chapter 2 that much of the Tieboutian justification for local government is rooted in preference satisfaction. Given the tendency of rational individuals to satisfy their own preferences, allowing individuals to migrate among jurisdictions that facilitate that objective seems unobjectionable in light of the alternatives of restricting free movement, as long as the constraints on exercising invidious preferences or those that adversely affect others are sufficiently strong and opportunities for mobility are sufficiently robust. The imperfections of preference satisfaction to which I alluded in chapter 2 limit but do not eradicate the utility of that objective as a standard for sorting for local public goods. Different jurisdictions do offer different bundles of local public goods and services, and it is difficult to attribute the subsequent distribution of individuals as inherently invidious. Those preferences are not necessarily stable. Governments and private entities may seek to alter the preferences of individuals—to make them more environmentally conscious, more frugal, more accepting of nontraditional forms or subjects of education. Nor, as I indicated in chapter 1, does the binary nature of voting permit residents’ preferences to be reflected in the electoral process. Even if preference aggregation were possible, there are reasons tied to temporal externalities and distributive justice that would cause us to constrain choices based solely on preferences; for example, local residents, especially nonhomeowners who do not have assets into which the future value of current investments can be capitalized, may have too limited of a time horizon to incur short-term costs that return long-term benefits. Even where constituent preferences are relatively clear, local officials could choose not to pursue them out of a benign concern that doing so is inconsistent with constituents’ own self-interest. Local officials may

74

wh y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

conclude that constituents possess insufficient information to evaluate the benefits of a proposed policy and that their opposition to it would dissipate if they had fuller information. I include such conduct within the realm of benign rather than malign redistribution, notwithstanding the deviation from preferences. One may argue that theories of democratic governance require more fealty to constituent preferences than a paternalistic view toward insufficiently informed preferences suggests. But that is a very different matter from the political process defects that I suggest characterize “malign” redistribution. It is difficult to castigate local officials who, in good faith, attempt to provide a service that they actually believe will confer optimal benefits on the locality. Local officials, for instance, who believe that economic development will genuinely improve residents’ well-being in ways that justify subsidies for developers, or that redistribution to the poor is a civic obligation despite its cost, act in a manner consistent with a reasonable conception of municipal interest, and hence benignly, even if a majority of constituents opposes the proposal. Well-meaning officials who fail to convince their constituents to alter preferences to reflect their own interests may fall victim to electoral redress, but that does not condemn either the political risks they incurred or the political process that precipitated their defeat. Paul Peterson similarly rejects any effort to define municipal interest in terms of the summation of individual interests or the pursuit of optimum size hypothesized by Tiebout.1 But Peterson then defines policies as serving municipal interest when they “maintain or enhance the economic position, social prestige, or political power of the city, taken as a whole.”2 I reject that definition as well, insofar as it tends to identify municipal interests with successful programs and to identify unsuccessful programs as inconsistent with municipal interests. As with his definition of “redistribution,” one difficulty with Peterson’s analysis is that it calls for a conclusion that is extremely difficult to calculate. Will a municipally financed art show attract more tourist dollars? Or will money spent by attendees only displace entertainment spending that would have been made at other venues in the city? Peterson suggests that investments in winning the high school basketball tournament will serve city interests because the school system will gain prestige. But will that investment have been contrary to the city’s interest if, contrary to expectations, the high school team loses in the finals?

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

75

Even if we could calculate ex post whether an investment had paid off, why would that matter? Investments are necessarily probabilistic, and it seems peculiar to gauge their propriety from an ex post rather than an ex ante perspective. If, as a matter of institutional design, we are attempting to determine which level of government is best positioned to engage in redistribution and whether local officials are likely to advance the interests of their constituents, then the inquiry must be ex ante, not ex post. All we can ask of the purest political process is that it generate investments that have a substantial likelihood of success. Redistributive programs that have low probabilities of advancing local interests are likely to coincide with programs that emerged from a flawed political process, but it is the latter feature that condemns them, not the fact that a risk of project failure materialized. Nor do I identify either benign redistribution or municipal interest as identical to programs that are predicted to return economic benefits to the locality in excess of their tax price, although benign redistribution may frequently have that characteristic. The increased tax rates necessary to subsidize redistributive programs may in fact reduce local property values and therefore have negative revenue effects.3 If local residents knowingly exceed efficient levels of taxation in order to fund the redistributive services they prefer, it is difficult to contend that they have acted in a way that requires some legal corrective, as is implied by designating the inefficient program as malign. Indeed, to the extent that redistribution is motivated by altruism, it will necessarily require expenditures in excess of monetized returns. S O C I A L LY B E N E F I C I A L D O M I N A N T I N T E R E S T G R O U P S

Although my benign/malign dichotomy depends significantly on evaluation of the political process by which a redistributive program is enacted, political processes fraught with rent-seeking conduct by dominant interest groups do not necessarily result in programs that offer negative returns to the municipality. By definition, publicly interested legislation has the characteristics of a public good. That is, once enacted, potential beneficiaries cannot be excluded from obtaining its benefits, and one person’s enjoyment of the rights granted by the legislation typically does not preclude others from simultaneous enjoyment of similar rights. As a result, the median member of the public has little incentive to invest resources

76

wh y d o lo cal g o v e r n m e n ts r e d i s tr i bute?

in securing its passage. She can obtain the benefits of the legislation without incurring any of the costs necessary to secure its passage simply by sitting back and allowing others to do the work of lobbying, accumulating data, and presenting findings to legislators, or, in less sympathetic terms, influencing legislators with promises of campaign contributions or more-tawdry inducements. The expectation of free riding on the efforts of others often leads to the prediction that political systems will systematically underproduce public goods, since each potential beneficiary will await costly efforts by others to lobby for them. But contrary to expectations, many public goods tend to be supplied in an amount at least as great as is optimal, and sometimes in excess of that amount. We don’t tend to think, for instance, that the quintessential good of national defense is undersupplied. The reason for the anomalous result is that those who are willing to avoid free riding will tend to be groups that are not “median” members of the public but instead constitute those who will obtain an idiosyncratically high benefit from the legislation—high enough to warrant incurring the costs related to enactment. But those are the very characteristics of an interest group that is likely to be successful in obtaining a supraoptimal level of benefits from the legislation. The median citizen will desire national defense but would prefer not to expend personal resources in obtaining it. Thus, even those enactments that would serve the public interest will likely require advocacy by parties who will receive a disproportionate benefit and therefore are willing to expend the personal resources necessary to procure passage. We might think of the excessive benefit that the activist subgroup attains as a reward for its willingness to act as something like a private attorney general in creating the public good that would otherwise be underproduced. Defense contractors may be willing to lobby for national defense and thus to serve as surrogates for more-passive citizens who desire defense but wish to free ride on the efforts of others to procure it. But the “price” that defense contractors will extract constitutes an amount of national defense in excess of that desired by the average citizen. Absent those contractors, however, pervasive free riding would likely generate less national defense than the median voter prefers. We might rationally prefer too much national defense to too little of it, even if we would desire just the right amount. The problem of oversupply is analogous to the situation modeled by Thomas Romer and Howard Rosenthal in which officials who have a monopoly in proposing

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

77

laws can induce those charged with enacting laws to provide more than a socially optimal amount on a project because the enactors are even less willing to accept the alternative of forgoing the project.4 At the local level, think in terms of developers who wish to construct commercial buildings that may have positive effects on municipal employment. Those who might find better employment as a result of such enterprises cannot readily identify themselves and, even if they could, are unlikely to pool resources to procure legislation that might support commercial development, such as tax breaks for developers. Developers themselves, however, are direct and identifiable beneficiaries of such plans. They have the motive and are likely to have the resources necessary to coalesce in favor of reducing barriers to development. To the extent that their interests align with the interests of residents generally in attracting economic growth, developers serve as proxies for those beneficiaries who are less able to organize and less willing to contribute to the development effort. But developers are less likely to be attentive to the local costs of development, including the elimination of low-value residential housing or community-based commercial enterprises (the proverbial “mom-and-pop” stores). Given that the benefits for employees, employers, and consumers could outweigh the losses imposed by development, it is plausible that the efforts of developers could produce net municipal benefits. But “overdevelopment” may be the price that developers extract in order to produce what is essentially a public good insofar as economic development positively affects the locality generally. So, as in the national defense example, we are faced with a difficult choice. In the absence of organized opposition, we are unlikely to achieve an optimal level of development. In the absence of prodevelopment interests, the locality may suffer more than it does as a consequence of too much development. If that is the case, we can as easily categorize developers as producers of public benefits who effectively charge for their services as we can categorize them as venal violators of the public interest. Narrow judicial construction or invalidation of legislation that was perceived as evincing the influence of a dominant interest group, therefore, could reduce the benefits of providing that service and hence cause underprovision of the publicly desired good. Living wage campaigns could also exemplify this phenomenon. Let us assume for the moment that these campaigns actually do increase

78

wh y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

public welfare in the localities that enact them, although they may impose net costs on individual employers and reduce employment among the working poor. How would a locality come to enact this hypothetically beneficial program? Certainly not through the efforts of employers who might organize with relative ease but who, by hypothesis, bear net costs. Indeed, as the examples of Santa Fe and Santa Monica reveal, employers might have a sufficient stake to form a cohesive interest group against a proposed living-wage ordinance, so that any interest group advocating it would not qualify as dominant. For the sake of my example, therefore, I will assume that, for some reason, employers fail to organize or to participate in the political debate, and I will inquire into whether a dominant group’s victory necessarily translates into malign redistribution. Those who will benefit most immediately from an increase in the minimum wage, low-income workers, typically do not form substantial interest groups. Especially where the low-wage workforce comprises immigrants, some of whom may be undocumented or ineligible to vote, there is little reason to believe that group members will overcome the costs of organizing into a cohesive political force and lobbying for an optimal wage. The very conditions that suggest that low-wage workers are vulnerable to exploitation in the workforce should have even greater purchase in the political context. In the labor market, unskilled workers may receive some wage protection if demand for their services is sufficiently great that employers are price takers. But there is little reason to believe that unskilled labor receives any analogous protection in the political market. Finally, self-interested other residents of the locality would have little incentive to organize for the living wage ordinance, notwithstanding its hypothetical net benefits, because their per capita benefit would not be sufficient to warrant their participation. If my hypothetically beneficial program is to be enacted locally, therefore, it must occur either through altruism or through the intervention of some group that does have political power and that also stands to gain indirectly from enactment of the legislation. Unions, for instance, may have significant access to the local political decision-making process and may also believe that increased wages for the lowest-skilled wage earners will push up wages for their own more-skilled members. Unions, therefore, may be willing to advocate adoption of living wage ordinances. In doing so, again in my hypothetical example, the union may be acting

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

79

solely out of the interest of its members; nevertheless, its concerted activity overcomes a collective action problem that otherwise would have foreclosed adoption of a locally beneficial program. The union’s presence as a dominant interest group in the decision to enact the ordinance, therefore, probably should not be used to classify the ordinance as an example of malign redistribution, since it was necessary to obtain legislation that was, by definition, beneficial to the locality. The possibility that rent-seeking behavior can be consistent with municipal interest obviously complicates any effort to determine how to defuse the negative effects of dominant interest groups and to define the relationship between interest groups and the desirable scope of local redistribution. Even the most narrowly interested expenditure will typically be articulated in terms of the returns that will ultimately redound to the public; no legislator proclaims that he or she supports a proposal in order to pay off supporters. One can even redescribe what might seem to be the definitive act of municipal self-degradation, the purposeful inducement for productive political opponents to exit the jurisdiction, in positive terms. Glaeser and Shleifer explain a variety of redistributive policies in the Boston of Mayor James Michael Curley as an intentional effort to induce relatively wealthy, established residents, whom one might believe essential to continued productivity in the city, to emigrate in order to entrench a political base that benefited from the redistributive programs. But it is unlikely that Curley would have explained his strategy as an effort to cement a political monopoly; instead, his supporters would have depicted a positive program to empower and develop solidarity among an underserved political minority. At best, then, I can provide a theoretical construct for what my concept of “malign” redistribution is meant to embody. I reserve the term for local policies that cannot credibly be reconciled with constituents’ expected interests at the time they are implemented. Instead, malign programs tend to be a consequence of the more personal objectives of local officials, such as securing electoral support from influential groups or increasing chances for post–public service employment. Unlike “benign” programs that may fail to produce anticipated benefits despite the best intentions of publicly interested officials, malign ones are not proposed to foster any reasonable conception of constituents’ welfare. It is in that sense that malign programs have an affinity to, but are not coterminous

80

w hy d o lo cal g o v e r n m e n ts r e d i s tr i bute?

with, current understandings of “corruption.” Local officials who act malignly are not necessarily obtaining explicit bribes or otherwise violating the law. They are, however, violating an obligation to do their best for their constituents. The best analogy may be to Machiavelli’s use of “corruption” to mean degradation in the quality of government, regardless of its source, so that the public sphere becomes a place solely for the promotion of individual interests.5 Constraints on such behavior do not necessarily come from laws directed at specific acts, in the way that antibribery statutes identify a specific payment. Instead, constraints on malign behavior come from the creation of a process the effect of which is to ensure robust political conversation prior to enacting a policy in ways that neutralize the purely personal incentives of the public’s fiduciaries. But even if this theoretical construct is coherent, it remains to be seen whether the distinctions between benign and malign redistribution are sufficiently recognizable to constitute the basis for critiquing municipal policy or invalidating specific programs. Thus, one of the challenges of the next several chapters is to determine whether it is possible to put the theoretical distinction between benign and malign redistribution into operation in a manner that has practical implications. W H Y D O L O C A L I T I E S R E D I S T R I B U T E W E A LT H ?

Malign redistribution is easy enough to understand—a group with interests that deviate from those of the general populace desires to have municipal wealth transferred to it and enjoys the capacity to organize and to offer political support to local officials, notwithstanding that the redistributive effort fails to serve any plausible municipal interest. It is perhaps particularly comprehensible in the context of redistribution to the wealthy, where private parties who benefit from government largesse have the resources and the incentives to share the gains with political officials in order to procure expenditures or mandates that might not be forthcoming if municipal interest alone dictated results. Alternatively, malign redistribution could be supply based rather than demand based. Local officials could redistribute wealth to a particular group in order to consolidate political power, perhaps with the very intention of inducing prospective opponents to exit rather than to use a more costly voice option of organizing political opposition. Benign local redistribution is more puzzling in light of the orthodox theory and the risk that redistribution will prompt the emigration of net

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

81

payers. Presumably localities would engage in benign redistribution only when doing so promotes local benefits that cannot be replicated through redistribution by more-centralized levels of government. Activities that satisfy that criterion are usually identified with redistribution to the relatively wealthy, such as the use of local revenues to attract industrial development for which multiple localities are competing. Centralized jurisdictions have less incentive to subsidize local development, since doing so may mean shifting resources from one part of the general jurisdiction to another, without any necessary net social gain. Indeed, it is somewhat puzzling that broad-based national subsidies for local economic development with few spillover effects, such as a tax exemption on interest for bonds issued to support local economic development, persist. Localities that commit their own resources to these projects have presumably calculated that winning the interjurisdictional competition for firms will attract tax revenues in excess of the subsidies necessary to induce taxpayers and thus provide a basis for increasing the welfare of all residents. Numerous studies either contest these claims or contend that benefits garnered by the attracting locality are offset by social losses to other localities from which relocating firms emigrate. But the evidence about the net effects of subsidies for local economic development is at least sufficiently ambiguous to provide support for local officials inclined to compete for development programs through subsidies to the wealthy. The orthodox theory is less receptive to benign local redistribution for the poor, in the form of either direct expenditures, abatements, or regulatory mandates such as living wage ordinances. It is tempting to attribute redistributive programs to pure altruism—local residents may be more willing to share their wealth than rational self-interest (defined to exclude a selfish preference for altruism) would predict. But that explanation, standing alone, provides little support for local redistribution. Individuals could fulfill their desire to behave altruistically by paying redistributive taxes to a more centralized government or through private giving. Doing so would serve the desire to assist the relatively poor and simultaneously avoid the negative consequences that the orthodox theory attributes to local redistribution. Some additional factor must explain why redistribution occurs locally. There may be a basis for benign redistribution that does not emanate from local discretion and thus does not depend on unique local benefits.

82

wh y d o lo cal g o v e r n m e n ts r e d i s tr i bute ?

In our federal system, states typically exercise plenary power over their political subdivisions, at least with respect to those issues that have effects beyond local boundaries. It is plausible, therefore, that local redistribution is the result of a state-imposed obligation rather than a local decision. As I indicated in the previous chapter, much of the history of local redistribution for the poor is explained by state mandates that localities care for impoverished residents. States might impose these obligations in order to require all localities to share the redistributive burden and thus avoid the exit that might occur if only some localities undertook redistributive spending, or in order to take advantage of local administrative expertise with respect to the delivery of welfare services. Alternatively, and less benignly, state officials might wish to impose the funding obligation on localities in order to shift responsibility for concomitant tax increases to local officials and thus avoid electoral redress from disgruntled voters. Some state laws continue to impose unfunded or incompletely funded mandates on localities to provide services to the poor.6 Connecticut requires the town of an indigent person’s residence to provide support.7 New York state law requires local governments to pay 10 to 25 percent of Medicaid costs for their residents,8 and the state constitution mandates that the state and city governments provide for the “aid, care and support of the needy.”9 California requires each county to “relieve and support all incompetent, poor, indigent persons . . . lawfully resident therein, when such persons are not supported and relieved by their relatives or friends, by their own means, or by state hospitals or other state or private institutions.”10 States and federal governments may also impose on localities unfunded mandates that disproportionately benefit the poor. Think, for instance, of state lead-paint abatement requirements or No Child Left Behind requirements that compel local expenditures more likely to benefit low-income children. More-centralized levels of government may also indirectly influence the degree of local redistribution. Paul Kantor has suggested that federal policies control the extent of local spending because changes in federal spending parallel changes in local social welfare spending. For instance, movement from categorical to block grants could reduce local redistribution to the poor because block grants could be used for development rather than social welfare.11 Michael Craw hypothesized that if Kantor

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

83

was correct about the federal role, then jurisdictions receiving intergovernmental revenues earmarked for social services would spend more on social services than jurisdictions that did not receive intergovernmental revenues, and that increases in social service intergovernmental revenues would at least commensurately increase the receiving locality’s social welfare expenditures. Craw does find evidence that federal and state intergovernmental grants to localities are reflected in local social welfare spending levels and do not substitute for those levels. Craw thus concludes that “local governments have little autonomy in setting social welfare spending levels.”12 Given Craw’s findings, one might conclude that local preference is less relevant than intergovernmental policies in explaining variations in local redistributive expenditures. But it is not clear how much local redistribution can be explained by looking at more-centralized policies. Again, the centralized policies that Craw examines do not account for redistribution that takes the form of regulatory mandates. Indeed, one would anticipate that to the extent local governments have less flexibility to increase revenues than morecentralized taxing authorities (due to tax limits, balanced budget requirements, absence of authority to impose certain taxes, etc.), they are more likely to effectuate redistributive preferences through regulation than through direct expenditures. Moreover, even if intergovernmental transfers explain some local redistributive spending, there still remains significant redistribution from local revenues that is not reimbursed from centralized programs. It appears, therefore, that—contrary to both intuition and the orthodox theory of fiscal federalism—substantial impetus and funding for these programs arise at the local level. The difficulty, then, is explaining how local redistribution could generate unique local benefits unavailable through centralized subsidies. Most explanations for this phenomenon proceed from Mark Pauly’s observation almost forty years ago that there may be a spatial dimension to redistribution.13 Pauly’s theory rests on two assumptions. First, it assumes that one individual can obtain utility from the welfare of another, that is, that personal utilities are interdependent.14 Thus, the spatial theory rejects the orthodox story insofar as it treats local redistributive payments as purely a cost item, which local residents would prefer someone else to incur. Spatial theory instead assumes that at least some potential subsidizers enjoy benefits from redistributive services for others. Second,

84

wh y d o lo cal g o v e r n m e n ts r e d i s tr i bute?

the theory assumes that the utility A obtains from an increase in B’s welfare increases as the geographical distance between A and B decreases. One is assumed to care more about the well-being of neighbors, even unknown neighbors, than about the welfare of distant strangers. To the extent that is true, some local role will be necessary to achieve a socially optimal level of redistribution. Reliance on centralized provision alone will undersupply redistribution relative to the social demand for it. The combination of spatial theory and the sorting function of local government provides a plausible explanation for local redistribution. Again, sorting along lines of preference allows for a preference for redistribution. Some individuals may value caretaking or a sense of well-being among community members in the same way that other residents have preferences related to schools, proximity to workplace, public parks, or other public goods. In essence, local redistribution to the poor can be a solution to a market failure problem, just as local provision of street paving responds to the inability of landowners to enter into a multilateral contract for paving the street that all agree should be paved. There is little reason to believe that those who share preferences for redistribution would be any less capable of sorting into a locality than those who share preferences for a more tangible bundle of local public goods. The possibility that such communities exist is by no means fanciful. The orthodox theory, for instance, implies that there will be little variance among states and localities with respect to redistributive policies. Any deviation from the average, under that theory, invites migration. But there appears to be substantial variance along a variety of measures. The findings of Sharp and Maynard-Moody that I referred to in chapter 2, concerning wide variance in the funding of redistributive services, are supported by other studies. At state levels, Paul Peterson found that redistributive expenditures among the states over a twenty-five-year period reflected a relatively constant coefficient of variation of .28 to .37 (with an outlier during a period of uneven state implementation of federal aid programs). These figures indicated that about one-third of the states expended about onethird more or less on redistributive services than did the average state.15 Howard Chernick and Paul Sturm find that from the perspective of effective tax rates, the most progressive states had progressivity ratios almost three times as high as the most regressive states.16 Anecdotal observations support similar findings. Some localities signal antipathy toward

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

85

local redistribution to the poor by adopting large-lot zoning or subsidies for activities primarily engaged in by the relatively wealthy. Other localities, however, court a redistributive image, even though they contain a significant population with the means to live elsewhere. Cities such as San Francisco, Santa Monica, Cambridge, and Berkeley herald substantial welfare programs, either for target groups, such as the homeless, or for the relatively poor generally.17 Spatial theory also suggests a mechanism that induces residents to make local contributions that they would not make to more-centralized governments. Payers may enjoy personal psychic benefits from living in an area where their contributions are applied, and thus they may be more willing to make those contributions because they can see tangible evidence of their contributions, or simply because they have an affinity to the locality that justifies a higher contribution level. This may result either from achieving greater internal satisfaction from giving to a group with which the donor identifies (the “warm glow” effect18) or from achieving greater prestige within a group within which the donor is concerned about his or her reputation. We are not surprised, for instance, when alumni donate to educational institutions that they attended and may be perplexed when they donate to needier institutions with which they had no previous affiliation. Similarly, a payer of redistributive taxes may get more benefit from knowing that the municipal treasury funds a local battered women’s shelter than from knowing that a nationally funded one was erected in another jurisdiction, even though the number of beneficiaries is greater in the latter. Alternatively, a mix of personal and altruistic motives may favor local redistribution because it permits subsidizers to publicize their selflessness to those more likely to recognize them. Just as philanthropists signal their status by placing their names on buildings, professorships, or foundations, altruists can signal their munificence by living in a locality known for redistributive programs or by engaging in philanthropic activity at a local level, where they can readily be identified with their contributions. An alternative explanation that also mixes altruism and self-interest involves an affirmative desire to construct a sense of community among those who live in the same geographic area. Individuals may prefer to assist those who live in the immediate vicinity because they feel a sense of camaraderie with their neighbors, notwithstanding that they may have

86

w h y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

different socioeconomic status. Local subsidizers, on this account, may fully appreciate that local redistribution will impose excess costs in the delivery of public goods, will cause exit by some who would prefer not to subsidize those less well off, will make labor markets more inefficient by inducing firms to move from productive locations—and nevertheless conclude that all of these costs are worth incurring in order to construct a community whose members subordinate the efficient delivery of local public goods to an ethos of connectedness to the worst-off members of the community. Indeed, for some, that is what it would mean to have a community; mere cooperation in an enterprise that made all better off might be less conducive to the creation of community than sacrifice that allows parties to transcend self-interest and exhibit “virtue.”19 Philanthropists, for instance, may make contributions locally rather than nationally in order to enhance civic pride as much as to display their participation in a communitarian culture. Simply paying national taxes would dilute the communal benefits of contributions and would send a less amplified signal of one’s charitable tendencies. Experimental literature concerning redistribution is consistent with the community-building explanation. Esteban Klor and Moses Shayo find that for a nontrivial number of subjects in their experiments, preferences for redistribution are linked to social identity. They find that members of a group are willing to sacrifice a personal economic benefit to support the group as a whole as long as the personal cost is not too high. One consequence of these findings is that the definition of the relevant group can influence the amount of redistribution. If those who identify socially with others define the relevant group as coresidents of their municipalities, rather than coresidents of the nation or members a religious or ethnic group, then they should be willing to contribute locally to a greater degree than they would redistribute to national residents.20 The social identity story also supports an alternative, less congenial account of how local redistribution could improve on centralized redistribution. Social identity is often linked to measures that, at least in some contexts, we consider more invidious than mere geographical location but that might coincide with residence. For instance, there is some potentially disturbing evidence that racial differences counteract tendencies to provide assistance to others. At least that is one interpretation of findings that income redistribution decreases in racially heterogeneous areas and that

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

87

survey respondents are more willing to support redistribution to members of their own racial group.21 A related and broader literature suggests that ethnic diversity is correlated with a low level of public goods provision.22 To the extent that localities sort on ethnic grounds, therefore, it is plausible that local redistribution will exceed contributions that would have been made to the same population through a more centralized system, but we might be reluctant to endorse the motivations if we believed that animosity toward other groups was doing more work than a desire to build cohesion between the giving and receiving group. The “warm glow” of giving may similarly be generated by mixed motives, some of which are too self-indulgent to be admirable. While the impure altruism that is associated with internal benefits is typically attributed to a feeling of selfsatisfaction about assisting others, that same self-satisfaction may less generously be related to a desire to confirm relatively high socioeconomic status.23 But the argument here is not that local redistribution is necessarily motivated by the highest of human instincts. Rather, the argument is only that localities provide opportunities for individuals to redistribute in ways that cannot readily be replicated at more-centralized levels of government, and that localities have incentives to utilize those advantages to increase redistributive budgets. Finally, redistributive programs may be profitable for a locality that receives partial reimbursement from centralized governments to offset the costs of participation. Return, for instance, to New York State’s requirement that local governments pay 25 percent of Medicaid costs for acute care services and 10 percent of the cost of long-term care services for Medicaid beneficiaries. In the 2002–2003 fiscal year, New York City spent approximately $4 billion of nonreimbursed funds on these obligations, which amounted to 13 percent of the city’s local tax revenues.24 These costs are not voluntarily incurred by the city and thus may initially be viewed as inefficient unfunded mandates. Nevertheless, the program may actually return localized benefits that justify the local contribution. We could recharacterize the 25 percent limitation as a program in which New York City obtains matching funds from the state or federal government on a 3-to-1 basis for any local expenditures on Medicaid. Assuming that the recipient localities benefit from the economic activity generated by the matching funds, the required local expenditure may ultimately create net benefits for the localities that make them.

88

why d o lo cal g o v e r n m e n ts r e d i s tr i bute?

These explanations for local redistribution, however, face significant limitations. The “community building” explanation makes most sense when members have long-standing ties to the community. The mobility of those who pay and those who receive subsidies, combined with legal constraints on the denial of benefits to newcomers, means that those who wish to subsidize the local poor will not be able to exclude new members of the community in order to reward those with long-standing communal bonds.25 While proximity to the local poor may generate sympathy that could increase redistribution, those effects could be offset by countervailing tendencies on the part of others to create romantic visions of the distant poor.26 Daily interactions with the visible, proximate poor, especially in the form of aggressive panhandling or discomfiting sights of the homeless, may be as likely to produce negative reactions toward the local poor as compassionate ones. These reactions overtly take the form of calls for more regulation of public spaces, at least in large cities with heterogeneous populations.27 But the same sentiment may simultaneously diminish support for redistribution to those viewed as presenting a public nuisance, and in favor of the distant poor. The latter may be viewed as relatively deserving, existing only in sentimental images in media coverage and in organized solicitations for assistance. Given that repeated interactions with the local poor may generate either sympathy or animosity, it is not clear that one reaction prevails over the other. Of course, the fact that different people will have different reactions to the local and distant poor does reinforce the view that local redistribution may allow for sorting among redistributive preferences. If those who are more sympathetic toward the local poor gravitate to a particular jurisdiction, then that jurisdiction will more likely provide local redistributive aid than neighboring jurisdictions that attract residents who are more sympathetic to the distant poor. The “matching funds” explanation similarly supports only a limited type of redistribution,that is, redistribution to those local programs that might be a prerequisite to obtaining federal or state funds. It tells us little about why localities would generate redistributive programs on their own initiative. Alternatively, local residents may enjoy tangible rather than merely psychic benefits from local redistribution that cannot readily be duplicated by centralized redistribution. Some of these benefits simply involve the avoidance of disamenities that are often associated with the poor.

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

89

Local subsidies, for instance, may reduce the number of homeless people sleeping on sidewalks in one’s neighborhood. These perceived disamenities may have powerful effects even for those who support redistribution on other grounds. The recent shift in San Francisco from cash grants to in-kind assistance that provides the homeless with shelter and support services may be illustrative. Support for the change may have been attributable to a desire to remove the homeless from public sight as much as to provide shelter.28 Obviously, one significant disamenity involves crime or violence in the community. Accordingly, one may explain local redistribution as a form of “bribe” to the local poor to maintain a certain level of social peace that the relatively wealthy may otherwise believe will be threatened. Local residents who fear social disorder from the poor may be more willing to support local redistribution in the belief that their own personal welfare and property are more at risk from the geographically proximate poor. If the cost of the perceived “bribe” is less than the expected value of personal loss, and if that same “bribe” cannot be obtained at lower cost from centrally funded redistribution, then local residents may prefer to purchase social order. This motivation is consistent with the “social disturbance” thesis for local redistribution, promulgated several decades ago by Frances Fox Piven and Richard Cloward.29 They contended that fear of social disruption was the best explanation for the increase in social welfare spending that followed urban riots in the 1960s. On this theory, redistribution is not motivated by sympathy for the relatively poor or by altruism but by fear that denial of funding will threaten civic harmony. Attempts to verify this thesis have generated mixed results. The more recent studies, which attempt to correct some methodological shortcomings of earlier analyses, support the proposition that social disturbance has at least some explanatory effect for redistribution, though other characteristics, such as racial composition of the jurisdiction, also have significant effects.30 While these motivations seem plausible, they should skew the types of localized redistribution that find support. They explain support for highly targeted forms of redistribution, such as those directed at reducing violence or removing panhandlers or the homeless from streets, rather than broad redistributive programs that would increase services for the relatively poor throughout the community. The desire for social order has less explanatory force for municipal senior centers than for after-school

90

wh y d o lo cal g o v e r n m e n ts r e d i s tr i bute?

programs, for example, because the elderly poor pose less of a threat to property owners than do roaming youths. Social disturbance theory provides mixed explanations for living wage campaigns, since employees may improve their financial standing, but there is some risk that higher minimum wages will increase the ranks of the unemployed.31 Moreover, the “bribe” story seems incomplete because the relatively wealthy who fear crime could receive more-personalized benefits by investing in private protection, supporting additional public expenditures for police, or exercising their exit option. Investment in private protection would avoid multiple threats, not simply crime by poorer residents, and the prevalence of economically segregated neighborhoods would facilitate collective action (e.g., walled neighborhoods, private security patrols) that would reduce individual costs. Indeed, the possibility that some residents could privately obtain a higher level of local public goods while avoiding contributions to increased levels of citywide services appears to motivate much of the opposition to municipal subunits, such as business improvement districts, or to privatization of public services. REDISTRIBUTION AND LOCAL ECONOMIC INTEREST

There is, however, a stronger claim that one can make about the willingness of local residents to support redistributive programs for the relatively poor. These programs can potentially be construed as investments in the local economy rather than as pure cost items. Thus, they can be justified on the same basis as redistribution to the relatively wealthy, which is traditionally defended as a means of enhancing a locality’s position in interlocal competition for development. Initially, the notion that redistribution to the poor will increase local competitiveness seems anomalous. After all, the orthodox theory predicts that local redistribution at best constitutes a pure wealth transfer and at worst induces exit by net subsidizers and attracts more individuals who require subsidy. It would turn that theory on its head to propose that local redistribution instead produces positive returns. Nevertheless, the literature of local economic growth suggests just such a possibility. Peterson implied as much when he classified certain programs as developmental, notwithstanding their necessary redistributive elements. For Peterson, programs such as locally funded job training programs or free medical care for unskilled workers needed to staff manufacturing industries fall

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

91

within the developmental rather than the redistributive functions of local government.32 Recent literature provides a more complex picture about the net positive consequences of local redistribution. This more complicated story not only supports the provision of services to those who might otherwise drag on local growth but also justifies measures affirmatively to attract those who will require services. Much of this literature depends on a positive relationship between city development and socioeconomic diversity. This literature is somewhat in tension with the work I mentioned above that suggests redistribution and public goods are likely to be provided at lower levels in ethnically heterogeneous jurisdictions than in ethnically homogeneous ones. That research suggests that ethnic diversity inhibits interpersonal cooperation, which one would imagine is a prerequisite to increased productivity. The more optimistic literature indicates that the value of human capital increases in ethnically diverse localities in ways that would seem dependent on intergroup cooperation. Much of the literature either explicitly or implicitly links productive cultural diversity to the influx of immigrants into a locality. Significant research concludes that more culturally diverse cities have higher wages, and that the causal arrow indicates that immigration-related diversity creates the attendant higher productivity, rather than that higher wages attract immigrants. It is plausible, however, that immigrants may require at least short-term redistributive assistance. On that assumption, a locality that seeks an advantage in interjurisdictional competition by attracting and maintaining a diverse population should be willing to make redistributive expenditures that induce immigration and that would not be replicated by centralized redistribution. Alberto Alesina and Eliana La Ferrara examine the tensions within these competing literatures and cautiously conclude that “a diverse ethnic mix . . . brings about variety in abilities, experiences, and cultures that may be productive and may lead to innovation and creativity.”33 They argue that the complementary skills of diverse groups may increase the productivity of a private good within a jurisdiction. But the same diversity may reduce the ability to agree on public goods and public policies, since different groups will have different preferences and diversity may engender more conflict and less communication. Some trade-off occurs between these forces and determines the net effect of diversity on the

92

wh y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

local economy. The authors find, however, that the beneficial aspects of diversity are likely to dominate in jurisdictions at more-advanced levels of development, perhaps because in that environment the production process is sufficiently mature to take advantage of complementary skills. That characteristic is more likely to be found in American cities of the type with which I am currently concerned than in the African villages with which Alesina and La Ferrara make their comparison. Some authors imply that ethnic diversity is less likely to create conflict in jurisdictions that are democratic or that otherwise employ trust-generating institutions because the groups that might otherwise be disruptive are represented in normal political and economic processes.34 This idea is consistent with the claim that diverse, advanced jurisdictions realize net benefits. Studies that have directly measured the relationship between diversity and economic performance of American jurisdictions suggest that it is rather complicated. Alesina and La Ferrara look at fractionalization, by which they mean the likelihood that two randomly selected residents belong to the same group. They conclude that high fractionalization along ethnic lines (primarily meaning race) has a negative effect on population growth in initially poorer counties and a less negative or positive effect in initially richer counties. But cultural diversity measured in terms of language, lifestyle, and attitudes is positively correlated with productivity and a high level of public goods. Work by Ottaviano and Peri has similar implications. They found that persons born in the United States living in cities with higher proportions of foreign-born (“culturally diverse”) residents pay higher rents and are paid higher wages than those living in more homogeneous cities.35 Wages of white workers are higher in cities with a higher percentage of residents who speak a language other than English at home. They conclude that diversity has positive effects on production, which is reflected in housing markets and labor markets, rather than simply that immigrants are attracted to cities of high productivity. Glaeser and Shapiro’s preliminary evaluation of the 2000 census indicated that cities that attracted immigrants grew faster than cities that did not.36 They found that, as a group, medium-sized cities (the second hundred largest cities) grew faster in the 1990s than the largest cities. Vey and Forman found that growth was largely dependent on attracting Asian and Hispanic residents. The Asian population in medium-sized cities grew 58 percent, while the Hispanic population grew 67 percent. At the

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

93

same time, more than two-thirds of these cities lost non-Hispanic whites.37 While it is conceivable that much of the Asian and Hispanic influx arrived from other cities rather than through immigration, and thus Asians and Hispanics might have already achieved a level of success that obviated the need for redistributive services in their new jurisdictions, the large transition suggests that new immigration was also a factor.38 Indeed, recent analyses of the 2000 census reveal that the foreign-born population of the United States grew by 57.4 percent in the 1990s and that immigrants reside in a broader range of gateway municipalities than had previously been the case, with new gateway cities experiencing the most rapid growth of foreign- and native-born populations. Moreover, it appears that these immigrants will require some redistributive assistance. Many of the recent arrivals, especially those in new gateways, are poorer than the average local resident and have low English proficiency.39 These findings are not definitive. Glaeser and Shapiro determined that immigrant rates and growth rates were not perfectly correlated; cities with moderate immigrant growth grew faster than cities with high immigrant growth. Additionally, one possible explanation for the relationship is that the Census Bureau has simply improved its counting of immigrants. Third, it may be that growing cities attract immigrants rather than that immigrants create growth.40 Nevertheless, the fact that cities deliberately undertake to attract immigrants suggests both that local officials conclude that the causal chain runs in the opposite direction and that they are willing to invest in redistributive programs as a means to local economic development. These conclusions are buttressed by findings that cities with heavy concentrations of immigrants outperformed cities with few immigrants over a recent fifteen-year period, where performance was measured by a variety of fiscal and economic variables.41 Other studies of immigration suggest mildly positive financial consequences for cities that attract high number of immigrants, or at least an avoidance of significant negative financial consequences. David Card concludes that immigration produces a small positive effect on the average level of native wages, though those wages are distributed somewhat more unequally in the same cities, and high immigration produces a small negative effect on the relative wages of low-skilled native workers.42 Immigration, he finds, also increases rents in housing markets. Card warns that these effects may be offset by outflows from the city of natives

94

w h y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

who prefer to live in less socioeconomically or ethnically diverse neighborhoods and school districts and by political effects that “diminish[ ] the smooth functioning of local political systems.”43 Those consequences are consistent with the orthodox theory that suggests local redistribution will be rare. Nevertheless, Card ends on an optimistic note. “If anything . . . the evidence suggests the opposite: cities with more immigrants tend to have higher native wages, higher rents, and higher income per capita.” At the very least, the possibility that redistribution could attract immigrants who confer positive effects on the locality helps explain why local officials would be willing to reject the orthodoxy, at least up to a point when net outflows of relatively wealthy natives made redistribution counterproductive. The causal relationship predicted by this literature would be all the more convincing if we could point to a mechanism that explains how diversity translates into local economic growth. John Quigley attributes the relationship to five effects generated by socioeconomic heterogeneity.44 First, different groups have different knowledge, and intergroup knowledge spillovers may facilitate growth by increasing the variety of options that firms can deploy to increase productivity for both the firm and the locality within which the firm operates. This explanation appears to be consistent with the proposition that diversified cities are more likely to generate innovation.45 The mechanism by which this development occurs can be directly related to socioeconomic diversity. Prospective entrepreneurs and employees from different backgrounds presumably have different bases of knowledge, and the interaction among them increases the store of knowledge available to the firm as a whole. Just as interactions among personnel from different firms may generate an exchange of ideas that redound to the benefit of each of the firms,46 so may interactions among individuals from different socioeconomic groups generate beneficial exchanges within the firm. Second, local economies thrive on the capacity to realize economies of scale by supporting amenities that are susceptible to multiple uses. Think, for instance, of activities as varied as sports stadiums or transportation systems. These activities can be efficiently utilized only in an area that contains a population whose members differ in their use of the amenity and thus make its availability financially plausible.47

w h y d o l o c a l g o v e r n m e n ts redis t rib ut e?

95

Third, heterogeneity increases productivity by permitting more-varied outputs for similar inputs. As multiple uses for the same inputs expand, unit costs of obtaining them within the local area decrease. Quigley’s examples make the point: The same food can be deployed by different ethnic groups to create different cuisines; identical cloths can be used for different forms of dress; the same facility can be used to accommodate different cultural activities.48 Next Quigley speculates that diverse cities may be more productive because the large labor pool they can attract reduces the cost of matching labor and skills. Some support for this proposition can be inferred from the fact that the firms that bear much of the cost of redistributive programs frequently advocate their adoption.49 It is plausible that these firms are acting strategically—portraying themselves publicly as good citizens while simultaneously using access to decision makers to lobby against enactment. But in the absence of some evidence of such a pessimistic story, one might more readily presume that these firms predict that redistribution will translate into a workforce that is larger, healthier, and better educated and thus will increase the firms’ productivity in an amount that offsets any additional tax burden.50 Finally, diverse cities may be better able to achieve stability notwithstanding fluctuations in the economy because some firms and consumers may be thriving when others are not.51 These reductions in variability are likely to increase with the diversity of economic activity, which itself may be contingent on diversity in the population. These productive consequences of diversity require that the locality be able to attract residents who promote multiple uses of city resources or who provide the labor and consumption that allow realization of the benefits of diversity. Thus, a locality may attempt to attract low-wage individuals who, in the first instance, will need assistance in locating and assimilating into the existing environment. Recognition of the benefits of socioeconomic diversity and the desirability of a low-income population explains why cities that face declining populations, including New York, Boston, Pittsburgh, Philadelphia, and Louisville, have initiated programs to attract immigrants who can reduce labor shortages and forestall the degradation of housing.52 I do not mean to rest the entire case for local redistribution on the objective of attracting productive immigrants. Localities may find that aid

96

why d o lo cal g o v e r n m e n ts r e d i s tr i bute?

to nonimmigrants produces local benefits as well, whether by avoiding the deterioration of housing that might be abandoned if the population declines or by reducing unemployment through subsidies for job training or small business loans that would not receive centralized funding. Glaeser and Shapiro found that poverty rates and city growth rates were strongly negatively correlated both in the 1980s and in the 1990s, which suggested that it was the abundance of disadvantaged residents rather than temporary labor market disruptions that retarded city growth.53 Conversely, city growth (outside of California) was directly correlated with high rates of education. Chernick suggests that cities may face unique issues that warrant municipal intervention either because centralized governments refuse to provide needed assistance, notwithstanding external effects, or because the distinctively local nature of the problem warrants localized redress. For example, Chernick notes that AIDS tends to be concentrated in central cities, and thus service responsibilities linked to the disease may be greater there than in suburban communities.54 Thus, while redistribution to take advantage of potentially productive diversity may have the strongest empirical grounding, the localized benefits that it provides is only suggestive of a variety of reasons why local officials may benignly support other initiatives that initially appear to violate the orthodox theory. Local redistribution programs that are targeted at the needs of immigrants, many of which will overlap with the needs of the relatively poor generally, may, however, offer unique opportunities for cities in ways that the orthodox theory fails to recognize. A G G L O M E R AT I O N E F F E C T S A S A C O N S T R A I N T O N E X I T

The previous discussion offers various benign explanations for the widespread presence of local redistribution. Some of these rationales suggest that we will get more, and more highly tailored, redistribution at the local level than at more-centralized levels because potential subsidizers will enjoy localized benefits from local redistribution that would not be replicated by centralized redistribution.55 To the extent that redistribution overlaps with or is connected to local development policies, centralized redistribution would be less likely to take into account efforts by localities to seek competitive advantages from redistributive programs, and might even inhibit local redistribution that was perceived as causing wasteful competition. Thus, the puzzle of benign local redistribution may not be a puzzle

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

97

at all—it may be a rational response to local needs that could not be satisfied by more-centralized levels of government. Unfortunately, this optimistic story threatens to collapse once we reintroduce the premises of the orthodox theory of local redistribution. Redistribution may be valuable for the locality as a whole. But, apart from the altruistic explanations, that does not mean that it is valuable for any individual resident or firm of the locality who is a net subsidizer. If potential subsidizers can obtain the benefits of local redistribution while avoiding personal contributions by relocating outside the redistributive jurisdiction, then why would any but the most altruistic remain? A suburban resident who had access to the city, or whose suburban environment enjoyed positive spillovers from a vibrant city economy, could still take advantage of the central city for employment, shopping, quality of life, or culture without directly contributing to redistributive programs. Suburban firms could take advantage of the urban labor pool and transportation networks that facilitate the inflow of labor and of a lower rate of nonbenefit taxation. Suburbs could bid for firms by adopting ordinances that essentially serve as precommitment devices not to tax for redistributive purposes. Large-lot zoning ordinances may fit that pattern, as may local tax limitations that are more restrictive than state tax limits. Alternatively, localities may reject tax limitation overrides or adopt expenditure limitations in order to signal potential commercial and industrial residents that they will not be exploited after making location-specific investments. Localities that desire to redistribute, but that fear loss of residents and tax base, appear to have few weapons against the strategic response of less redistributive neighbors. The former may attempt to impose part of the redistributive burden through taxes or exactions that are largely paid by nonresidents, such as commuter taxes, surcharges on hotel rooms and parking and meal taxes. But the extraterritorial effect of those efforts typically means that state approval will be required for their imposition. If representatives from the nonredistributive jurisdictions can outvote those from the redistributive jurisdictions, state approval is not likely to be forthcoming unless the benefits will demonstrably be regional rather than simply local. It is plausible, therefore, that potential subsidizers will be able to frustrate even benign local redistributive programs by locating in areas that eschew redistribution while simultaneously enjoying the amenities of urban localities that prefer to redistribute.

98

w h y d o lo cal go v e r n m e n ts r e d i s tr i bute ?

It is not enough, then, to stipulate that localities may benefit from redistribution. Benign explanations for successful redistribution also require a mechanism by which cities can either discourage subsidizers from exiting or attract new residents who obtain sufficient benefits from city residence to offset their personal redistributive burden. The most plausible mechanism for accomplishing that goal is the exploitation of some situational monopoly that cities possess. One source of such a monopoly is immobile capital. Firms that have invested heavily in nontransferable assets and individual residents who have made location-specific investments in jobs and friends face serious obstacles to emigrating. But municipal officials may avoid exploiting those investments for fear of deterring potential new firms and residents from migrating to the locality. Other local characteristics, however, may raise the costs of emigration sufficiently to permit implementation of benign redistributive programs without fear of exit. Localities’ ability to constrain firms largely depends either on geographical advantages or on agglomeration economies—benefits realized by proximity to other firms within the industry or related to the industry—that cannot readily be duplicated in smaller jurisdictions.56 Geographical benefits, such as proximity to a river or a necessary source material, obviously are not easily substitutable. But those benefits have dissipated with the decline of the manufacturing sector. They may at least initially have been replaced with the benefits that cities could offer as transportation centers. Firms may desire access to transportation networks not only to ship products to distant purchasers but also to be accessible to prospective suppliers or purchasers.57 Hence, highways, airports, and railroads can take the place of mineral deposits or other resources that motivated manufacturers to locate in particular geographic areas.58 But the dramatic reduction in transportation costs that constrained the mobility of capital in an earlier period may have substantially dissipated the transportation benefits of location, and thus caused some disintegration of what Paul Kantor describes as the “magnetic capability” that cities could once exploit in order to redistribute wealth without fear of generating exit.59 Nevertheless, location decisions by firms may hinge on criteria other than monopoly over a locality’s proximity to physical assets or transportation networks. Instead, less tangible but equally effective attractions may tie firms to a locality from which they cannot easily exit. The phenomenon of

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

99

industrial clustering within a small geographic area suggests that agglomeration economies may evolve from interactions with other firms in the same industry or related industries. Firms benefit from locating near other firms in the same business so that they can exchange ideas about issues of common interest. According to Michael Porter, “[c]luster thinking suggests that companies have a tangible and important stake in the business environments where they are located in ways that go far beyond taxes, electricity costs, and wage rates.” Firms can also realize benefits from locating near the professionals with whom they consult, such as their lawyers, bankers, and accountants. Those professionals, in turn, may develop specialized expertise by constant interactions with clients in the same industry, and that specialization loops back to attract more firms that benefit from proximity to their advisors. Trade associations may develop to facilitate the flow of information among related firms and to engage in lobbying for the benefit of members.60 These agglomeration economies constrain the location decisions of firms in the same way that proximity to a natural resource or more costly transportation networks did in an earlier era. There is at least some evidence, however, that agglomeration benefits dissipate rapidly beyond short geographical distances.61 Thus, those who wish to take advantage of these benefits cannot readily migrate far from the cluster that generates them; instead, they must stay in a relatively concentrated geographic area. It is unlikely that suburban areas will be able to accommodate all the related firms that wish to take advantage of these economies. Moreover, once a cluster has formed serendipitously or intentionally in a particular location, no individual member has an incentive to depart except in the unlikely event of a simultaneous movement by a large number of other members. The benefits of geographic proximity may outweigh any benefits that a firm would receive by being in a different location away from the cluster. These agglomeration benefits are characteristic of relatively large cities—the very localities that I have suggested are most likely to implement local redistributive programs. At the same time, the capacity of cities to accommodate clusters further supports local redistributive efforts. As Porter argues, government policies can enhance the efficiency of clusters. Porter appears to support the orthodox theory against local redistribution. He writes, “Ideally, all government policies that inflict costs on firms without any compensating, long-

100

w h y d o lo cal g o v e r n m e n ts r e d i s tr i bute?

term competitive or social value should be minimized or eliminated.”62 But the caveat leaves enormous room for local redistribution. Porter finds that the productivity benefits from clustering flow from characteristics such as access to a larger-sized specialized employee pool and complementarities among different functions within an industry and among different industries. These characteristics are reminiscent of those that Quigley attributes to urban diversity, and arguably they support similar levels of local redistribution. These constraints on the mobility of human and financial assets reduce the risk of free riding by those who would otherwise emigrate to less redistributive suburban locations even though they benefit from redistributive cities.63 Constraints on mobility are frequently seen as a negative characteristic of local policy, insofar as they permit municipalities to impose higher taxes or other exploitative policies on capital that is unable to exit. But a more positive way to think about capital immobility is to attribute the possibility of benign local redistribution to its presence, since capital that might otherwise exit and leave the funding of admittedly benign redistribution to others will not have that option. Before adopting this more optimistic explanation and justification for local redistribution, however, consider three points. First, immobile firms are likely to recognize that they are locked into a location once they make their initial investment, and thus that they will be susceptible to the imposition of redistributive taxes. Even if firms ultimately benefit from local redistributive policies that generate agglomeration effects, they would rationally prefer to enjoy the benefits of those policies while avoiding their costs. As a result, immobile firms may seek ex ante compensation in the form of tax abatements or low-cost financing before agreeing to locate in a particular jurisdiction. These concessions offset subsequent redistributive payments and, by reducing local tax revenues, may limit the locality’s future capacity to redistribute income.64 At the same time, the possibility of ex ante compensation answers claims that immobile firms are being exploited when redistributive payments are demanded from them. Moreover, fear that firms will either eschew a redistributive locality or demand ex ante compensation before locating there should serve as a check on the locality’s ability to exploit immobile firms.65 Second, while geography and agglomeration economies may constrain exit by firms, these local characteristics have less impact on individuals.

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

1 01

Relatively mobile individuals who prefer not to bear redistributive burdens could still migrate to suburban locations, pay only benefit taxes to their jurisdiction of residence, and continue to receive significant financial and cultural benefits from the central city. Suburban and exurban residents may desire a culturally and financially productive central city, but that does not mean they will pay significant redistributive taxes for projects they believe fail to generate suburban benefits. Central cities may be able to impose some charges on commuters in the name of defraying the costs that nonresidents impose on the city, and those taxes may discourage emigration by reducing its benefits. But central cities that are surrounded by less redistributive suburbs will have difficulty imposing redistributive taxes on those who are more mobile than altruistic. Third, and most important for present purposes, even if constraints on exit make benign redistribution more plausible, it does not follow that the relationship between agglomeration and redistribution confers an unqualified benefit. Any constraints on mobility reduce an important check on the capacity of local officials to frustrate residents’ preferences. Mobility is necessary to create the competition for residents and tax base that constrains officials who, but for fear of emigration or loss of desirable immigrants, might commit public funds to private ends. More to the point, the same situational monopoly that permits a locality to impose redistributive exactions for benign reasons without fear that dissenters and free riders will exit also reduces the ameliorative effects that exit provides against malign redistribution. As David Schleicher has astutely observed, there is a tension between the Tiebout model and agglomeration economies, notwithstanding that each is used to explain the provision of municipal services.66 The former indicates that municipal services cannot deviate significantly from residential preferences; the latter indicates that they can because mobility is necessarily constrained. Just as firms cannot obtain agglomeration benefits without paying benign local redistributive taxes, immobile residents are unable to avoid redistributive taxes imposed for objectives that serve much narrower interests. A firm that enjoys higher productivity because of its proximity to networks of competitors, suppliers, trained employees, and customers is unlikely to exit, even though it understands that its tax payments are applied to malign objectives, as long as the costs of being exploited are less than the agglomeration benefits the firm receives from remaining within the

102

w h y d o lo cal go v e r n m e n ts r e d i s tr i but e?

locality. Of course, the firm may exercise its voice option and publicly oppose programs that it perceives as malign, or it may work to unseat the officials behind these programs. Exit and voice are complementary, and the inability to use one heightens the incentives to use the other. But a firm may be reluctant to employ the voice option where the costs of opposition are high or where opposition may generate a reputation for animosity toward redistributive programs, a reputation that the firm might prefer to avoid in order to cultivate its status as a good citizen in the community. The very agglomeration economies on which localities depend to be able to implement benign redistribution, therefore, also enable local officials to impose malign programs. I next turn to a more thorough exploration of that possibility. M A L I G N E X P L A N AT I O N S F O R L O C A L R E D I S T R I B U T I O N

In an ideal world, we would retain local redistribution that is benign and invalidate its malign forms. But the various rationales for local redistribution, combined with the conundrum in which the same program is susceptible both to malign and to benign explanations, reveal the difficulty in classifying any given proposal either as an effort to enhance local welfare or as a sop to narrower political interests at the expense of municipal welfare. Is the proposed living wage ordinance a signal of an enlightened community’s sympathy for low-wage workers? Does it reveal a preference for avoiding labor strife or offset what might be higher tax payments for welfare services? Or is it a concerted effort by local unions to increase wages for more-skilled workers or to enhance their ranks in ways that may ultimately reduce local employment and productivity? Is the publicly financed stadium a source of civic pride that generates additional fiscal and communal benefits, even if it is not financially self-sustaining? Or is it merely a substitute for alternative forms of entertainment, constituting a subsidy for a vocal and politically effective minority? One might conclude that resolution of these difficult issues necessarily lies outside the domain of legal doctrine and solely within the political marketplace. Malign local redistribution will have effects primarily within the locality. Thus, all the reasons alluded to in chapter 1 for embracing broad interpretations of local autonomy apply, one might claim, to permit local redistribution as enacted by local officials. The proper remedy for any official defalcations arguably lies either at the ballot box or in the

w h y d o l o c a l go v e r n m e n ts redis t rib ut e?

1 03

market for residence. On this theory, in the absence of significant extraterritorial effects, some combination of voice and exit displaces any role for legal regulation about the scope of local activity. Attempts to interfere with local decisions, the argument may continue, either result in erroneous decisions about local preferences or dilute the benefits of local autonomy by imposing the will of more-centralized entities on municipalities. On reflection, however, neither the ballot box nor the market for residence is likely to provide sufficient controls on malign redistribution. To say that resolution of the issue of redistribution should be left for political processes is to deny the very predicate on which malign redistribution is based—political market failures due to the presence of a dominant interest group that monopolizes the political debate. And to say that the market for residence can resolve the issue is to deny the unfortunate consequence of agglomeration economies—the very constraints that make imposition of benign redistributive taxes possible without fear that payers will exit also allows exploitation of those who have to pay taxes for malign redistribution. Unless we are willing to accept a certain degree of malign redistribution as a necessary cost of implementing benign redistribution, the imperfections of political and residential markets imply that morecentralized entities must be involved in the review or approval of local redistributive projects. But, concerns about local autonomy aside, that claim raises more difficult issues about whether alternative institutions can do a better job of ferreting out the malign from the benign. In theory, ad hoc adjudication could allow parties who cannot command a majority within the political market, but who have idiosyncratic interests sufficient to warrant active opposition to a perceived malign program, to demonstrate that the challenged program fails to serve a plausible municipal interest. One can look at the multiple cases in which employers challenged living wage ordinances through that lens. Unlike political challenges, judicial intervention does not require convincing a majority of representatives to oppose the program. Instead it requires convincing a court that a legal standard for the exercise of municipal authority has been violated. If those standards are interpreted in a manner consistent with allowing benign and invalidating malign forms of redistribution, then legal challenges may, in theory at least, address deficiencies in the allocation of municipal authority that are left by imperfect markets for

104

w hy d o lo cal go v e r n m e n ts r e d i s tr i but e?

residence (exit) and politics (voice). This is not to say, of course, that judicial intervention will be perfect; indeed, the balance of this book explores those very imperfections. It is only to say that legal doctrine may play more of a role in defining the propriety of local redistribution than writings from economics or politics alone convey. No legal doctrine explicitly recites that “localities may implement benign programs but not malign ones.” Instead, those who have opposed local redistributive programs have sought to fit their challenge within one of the established cubbyholes that allocate governmental authority among the various jurisdictions in our federal system. Most frequently, these doctrines involve the legal scope of municipal authority, the reach of home rule, and the resolution of alleged conflicts between local ordinances and state statutes, though doctrines as far-flung as the Takings Clause of the federal constitution have also been impressed into service.67 It is tempting, therefore, to argue that these challenges should be decided by reference to the standard factors that dominate debates about the scope of municipal autonomy: the extent to which local decisions generate negative externalities, the need for statewide uniformity, and the fit between the proposed regulation and traditional local functions. Those factors do not directly address distinctions between benign and malign redistribution. They do, however, share an inherent ambiguity that demands principled clarification if decisions about the scope of municipal autonomy are to have any predictive power or bear any relationship to optimal allocation of government authority in a multilevel system. If we think of local autonomy as a means of allocating to local governments those decisions that can be decided at decentralized levels through a process in which all affected parties are effectively represented, then there is no inconsistency between clarifying ambiguous legal standards for local autonomy and investigating the presence or absence of dominant interests. If, for instance, a statewide legislative forum is more likely to accommodate all those interested in a rent control ordinance, then perhaps the decision about rent control is best made at the state level. If, however, landlords were expected to dominate the debate at the statewide level, while landlords and tenants were expected to be represented in local debates about the issue, then perhaps any ambiguity about whether rent control lay within the domain of home rule should be resolved in favor of the local prerogative. The difficult issue that remains is whether courts can legitimately and accurately reverse

w h y d o l o c a l g o v e r n m e n ts r edis t rib ut e?

1 05

engineer local decision-making processes to discern the presence or absence of dominant groups. T H E R O L E O F S TAT E C O N S T I T U T I O N A L L AW

The structure of state constitutional law, from which the doctrines that allocate authority among levels of government are derived, potentially invites a substantial degree of judicial intervention. In the next chapter, I contend that state constitutional doctrines that define the scope of municipal autonomy are best explained as preventatives against interest groups that seek favorable treatment from the state or local legislature. Doctrines that pervade state constitutions, such as public purpose clauses, prohibitions on special legislation or special commissions, expenditure limitations, prohibitions on unfunded mandates, single-subject requirements—which by and large have no federal constitutional counterpart—arguably create a presumption that malign redistributive legislation is a serious threat to which the judiciary is authorized to attend.68 The very presence of these provisions and the doctrines to which they give rise arguably invites judicial inquiry into legislative processes in ways that would raise significant skepticism if practiced by federal courts. But the structural arguments about institutional legitimacy that arguably justify judicial restraint at the federal level may have less force when considering the propriety of state judicial intervention into local decision-making processes. If these provisions signal a constitutional concern about the inordinate influence of special interests in legislative processes, then one might more readily consider it part of a legitimate state judicial function to address the role of interest groups in the promulgation of legislation. In particular, if we believed that courts could fruitfully and accurately make the distinction, we might applaud judicial efforts to invalidate malign local redistribution while upholding its benign counterpart. But any such judicial mission necessarily confronts the perception that courts lack legitimacy and institutional capacity to perform the task of reverse engineering legislation to discover the process by which it was produced. In the next chapter, I explore the implications of state constitutional structure and the issue of institutional competence that it necessarily raises.

ch ap t e r f i v e

Local Political Markets and State Constitutional Constraints

i have suggested in the preceding chapter that local officials have both publicly interested and self-interested motivations to redistribute wealth, and that agglomeration economies provide the opportunity to act on either of those motivations without fear of significant exit by those who pay redistributive taxes. The same situational monopoly that makes benign redistribution possible, therefore, simultaneously facilitates malign redistribution. In theory, we would prefer that localities enact benign redistribution and avoid the malign. Achieving that result, wholly apart from the operational difficulty of distinguishing between the benign and the malign, poses theoretical difficulties of its own. Any effort to invalidate malign redistribution necessarily implies some constraints on the capacity of local officials to determine the proper scope of municipal activity, or local autonomy. Presumably those constraints would come from constitutional or statutory limitations on municipal authority, enforced through judicial invalidation of specific municipally initiated projects. Interference with municipal initiatives necessarily confronts arguments discussed in chapter 1 about the desirability of broadly construing municipal authority in order to achieve the objectives of decentralized government. That is, participatory or democratic objectives, efficient signaling and sorting, and matching local benefits and burdens all require 1 06

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 07

significant autonomy, even at the cost of allowing localities to engage in ill-advised activities, and even where political processes work imperfectly. But in the absence of clear indicia of municipal autonomy, centralized officials, who presumably have incentives to expand the scope of their own jurisdiction, arguably would overdetermine the adverse effects of local action and would severely restrict the scope of municipal authority.1 To the extent that we believe that localities should be able to make autonomous decisions about what goods and services to provide, external checks on municipal competence are anathema, at least when the exercise of one locality’s autonomy does not create such significant external effects as to interfere with autonomous decisions by neighboring jurisdictions. But autonomy does not mean freedom to make decisions by any process; it entails a quality of decision making consistent with the democratizing or preference-registering objectives for which local governments are formed. Where redistributive decisions are likely to emerge out of a process that disserves those objectives, deference to local politics in the name of local autonomy is more difficult to justify. If we are searching for some institution better positioned than state executives and legislators to strike the appropriate balance between the scope of state and local jurisdictions, then perhaps the solution lies in judicial checks on local redistribution. Judicial intervention in this context does not mean substantive review of the local decision. Instead, judicial competence might be limited to consideration of the process through which challenged redistributive programs were approved. If the primary concern is with the disproportionate influence of dominant interest groups that induces local officials to deviate from the interests of their constituents, then ex post review of the political process plausibly provides a better corrective than constrained exit or state processes standing alone. Courts would not necessarily be better positioned than local officials to perform the difficult calculus by which the propriety of redistributive programs is evaluated. Whether a locality can profitably undertake a redistributive program is a function of preferences, the local economic environment, alternative uses for the same funds, and the extent to which agglomeration economies can restrain adverse effects on those who might exit. Even in an environment of legislative fact-finding and the give and take of political debate, evaluation of those factors is likely to be imperfect. In the winner-take-all context of litigation

108

p o lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

characterized by the limited authority of courts to engage in independent fact-finding, substantive analysis and the compromise indicative of political bargaining are near impossible. But if each of the objectives of local autonomy is contingent on something like a working political process in which residents’ preferences and interests are considered and debated, then perhaps courts can more suitably play a role in policing the environment in which local redistributive decisions are made. Judicial intervention to negate malign redistribution, however, presumably would not take the form of a roving commission that evaluates the underlying political process of any proposed program of local redistribution. Rather, it would take the form of reviewing those programs under the doctrinal strictures that define the scope of local decision making generally. Local governments, as creatures of the state, must find some affirmative doctrinal predicate in either state constitutional or statutory law for initiating local programs. Other constitutional or statutory doctrines specifically restrict the scope of local autonomy. As a result, the relevant form of judicial constraint would call for broad or narrow interpretation of the relevant constitutional and statutory grants and limitations, depending on whether the challenged program appears to have emerged from a well-operating political process. Calls for judicial intervention in this context are reminiscent of debates concerning the proper role of courts when they interpret ambiguities in statutes that arguably were procured by dominant interest groups as a result of a “deal” with legislators rather than through the merits of their position.2 Some commentators admonish courts to resolve statutory ambiguities against rent-seeking interest groups in order to avoid legislation that serves narrow rather than public interests.3 Proponents of judicial intervention to narrowly construe the terms of the deal have relied on structural arguments about the nature of our federal constitutional system rather than on the language of specific constitutional clauses.4 For other students of federal constitutional law, however, assigning the state judiciary the role of monitoring the local political market seems peculiar. It implicitly compares a realistic view of a legislative process ridden with negative-sum logrolling to a myopic, highly sanitized view of the judicial process as inherently pure and publicly interested. Einer Elhauge makes the most compelling case for judicial abstention from efforts to infer interest group dominance. He contends that judicial decision making is

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 09

itself fraught with interest group pressures. Moreover, he demonstrates, any judicial effort to detect rent seeking is necessarily frustrated by the difficulties of aggregating preferences, cycling, and institutional limitations on judicial inquiry.5 Given that some amount of interest group activity will be visible in any political system, it is difficult to assess whether those interests have nefariously monopolized a biased process or simply participated in the bargaining inherent in a well-operating one. On this theory, courts that evaluate results to make the distinction between benign and malign processes run the risk of becoming involved in discredited substantive due-process review and simply substituting judicial interpretations of public welfare for legislative ones. These debates about judicial interpretation of statutes in light of interest group influence are obviously relevant to, but not necessarily determinative of, the argument for judicial review of municipal authority. Virtually the entire literature that assesses judicial interpretation comprises debates about the appropriate scope of federal court intervention into federal legislation. The debate centers on issues of institutional design, and thus it emphasizes, on the one hand, the relative competence of legislative and judicial bodies to make the relevant inquiries and, on the other hand, the susceptibility of decisions by each institution both to private influence and to correction by democratic processes. The characteristics of the federal institutions that drive these debates, however, map only imperfectly onto their state and local counterparts. The constitutional and institutional arrangements among the legislative, executive, and judicial branches in the states vary from those that exist in the federal context. Indeed, those arrangements vary among the states. Given the institutional differences, it is by no means clear that the role of the state judiciary in determining the constitutional and statutory authority of municipalities should be the same as the role of federal courts in interpreting federal statutes. Few would say that the political process should be devoid of any policing mechanism. Instead, the relevant question is what is the appropriate form of that mechanism: structural constraints such as term limits and open government requirements? the market for political competition enhanced by organized political parties? judicial intervention to invalidate certain practices proscribed by statute? legal doctrines that limit the scope of permissible legislative activity? The desirability of any of these constraints must be evaluated on the basis of “as compared to

110

p o lit ical mark e ts a n d c o n s ti tuti o n a l co ns t raint s

what?” The comparative desirability of judicial policing of the political process at the state level, therefore, could deviate significantly from its desirability at the federal level if the alternatives were also different. Consider first the doctrinal differences that govern the relationship between courts and legislators. The relevant doctrines do not explicitly assign roles to the state judiciary that differ from those assigned to the federal judiciary. Rather, the distinctions are more subtle. A variety of common clauses in state constitutions dictate that states and their political subdivisions exercise legislative authority within parameters that are inherently vague but that are suggestive of more formal restrictions than are suffered by their federal counterpart. These state constitutional clauses include prohibitions on special legislation and the legislative creation of special commissions, requirements that legislation address only a “single subject,” debt limitations, tax limitations, executive authority for lineitem vetoes, requirements that legislation serve a “public purpose,” and prohibitions on unfunded mandates. Most of these constraints apply with equal force to local as well as to state decisions. Localities are no more able to enact legislation for “private” purposes than is the state. Localities frequently face the same “single-subject” requirements that constrain state legislatures.6 These doctrines, moreover, typically have no explicit federal counterpart, presumably for institutional and historical reasons. Most of these clauses are products either of nineteenth-century reactions to state involvement with private firms in capital projects that failed, often leaving citizens rather than shareholders responsible for debts incurred to finance the underlying transaction, or of concerns about state political processes that allowed subgroups within cities to reverse municipal decisions. The legal response was the amendment of state constitutions to add restrictions on the scope of state legislation. Newer states that modeled their constitutions on those of existing states subsequently adopted these provisions as well. No such efforts occurred at the federal level. The more difficult process of amending the federal constitution may have accounted for the continuing omission of similar clauses from that document. Moreover, the federal government had not been involved in the financial failures that generated many of these clauses, so federal constitutional constraints may have been deemed less necessary. Finally, the different nature of federal and state constitutions may have led reformers to deem federal constitutional constraints superfluous. The federal

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

111

constitution is generally viewed as a document of grant, while state constitutions are typically viewed as documents of limitation. As a consequence, states are permitted to engage in any activity not denied to them by the constitution or ceded to the federal government, while the federal government can engage in only those activities for which a textual predicate can be found. Although more-contemporary interpretation of clauses such as the Commerce Clause provides for a very broad grant of federal authority, at the time when reformers were first considering the inclusion of constraints on governmental action, the perceived limited power of the federal government may have rendered federal constitutional restrictions unnecessary, while explicit restrictions on the states were deemed essential to deny power that had been abused. Given the inherent ambiguity in these state constitutional provisions, they necessarily require some interpretation in order to define the permissible scope of governmental activity in questionable cases. But the appropriate source of interpretation is itself ambiguous. These clauses could be considered self-enforcing political questions as to which state and local legislatures serve as the primary source of clarification. That is, any project funded by a local legislature could be said to satisfy a “public purpose” by virtue of the simple fact that the legislature has approved it, thus evidencing a public interest in the subject matter. Or a court could abandon the dubious search for the essential characteristics of a “single subject” and accept the reality that the “subject” of a legislative enactment can be defined either with great specificity (the Uniform Commercial Code may be considered to deal with multiple subjects because it governs remedies for the sale of goods as well as the means of perfecting security interests) or at a higher level of generality (the Uniform Commercial Code may be considered to deal with a single subject because all of its provisions relate to commercial transactions). That is, the resolution of any ambiguity concerning the scope of legislative activity need not entail judicial interpretation. One distinctive feature common to state constitutions, however, argues for a substantial judicial role in the interpretation of these clauses. The constraints impose limits on the legislature itself. They are largely reactions to perceived legislative conferral of benefits on a small group of constituents, notwithstanding that the legislation was costly to constituents at large. Prohibitions on special legislation, public purpose

112

p o lit ical mark e ts a n d c o n s ti tuti o n a l c o ns t raint s

requirements, debt limitations, tax limitations, and indeed the panoply of restrictions on governmental activity appear to be directed less at some substantive theory about the scope of governmental competence than at the threat that legislation characterized by particular features was likely to have arisen from a process dominated by those with disproportionate access to legislators and capable of directing that process toward malign redistribution from which that group would benefit. It would be perverse, therefore, to propose that the legislature itself should adjudicate the appropriate scope of these clauses. Perhaps the most explicit recognition that an alternative interpretation would be tantamount to allowing the fox to guard the henhouse arose in Illinois. After the state judiciary abdicated to the legislature the role of deciding whether proposed legislation was invalid as “special” in contravention of a state constitutional restriction, the state constitution was amended to provide explicitly that the question of whether the prohibition had been violated was “a matter for judicial determination.”7 In short, these constraints on state and local action evolved primarily from concerns about legislative redistribution, though typically to the wealthy rather than the poor. This objective is more obvious in some cases than others. The history of the public purpose doctrine, for instance, has been written in terms of trying to devise meaningful tests of what distinguishes public from private purposes. Since virtually any governmental expenditure will confer disproportionate benefit on some private parties (the firm that receives a local subsidy on the grounds that it will be able to employ more local residents still enjoys a competitive advantage over a firm that does not), some mechanism is necessary to determine how much private benefit is permissible before the public purpose of the expenditure is eviscerated. But these clauses were essentially unknown prior to the nineteenth-century grants of state and local largesse to promoters of railroads and other promised local improvements, the failure of which caused default by and fiscal distress to those governments that had issued debt to pay for the promised projects. Other common clauses impose less obvious but perhaps more effective constraints on redistribution to favored groups. Consider, for example, debt limitations. As in the case of public purpose clauses, they were enacted in reaction to financial failures of private promoters of railroads and other capital projects that had received subsidies from state and

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

113

local governments.8 But rather than explicitly prohibiting governmental loans to private entities (a role played by clauses that expressly prevent states and localities from lending credit to corporations), these provisions restrict the amount of debt that a state or city can have outstanding at any given time. One might conclude, therefore, they are intended solely to prevent fiscal distress, regardless of the use of debt. The effect of the limitations (if any), however, is to create competition among sponsors of capital projects for a share of a limited pie. The optimistic side of this story is that public officials who might otherwise lend public funds to support private benefits will be unable to do so because the zero-sum quality of the limitations means that serving narrow interests will require abandonment of other projects preferred by a majority of constituents. Voting to subsidize a privately operated industrial plant may be acceptable when it does not preclude the construction of public roads, but it might be more contestable when it does.9 Control through the political process, however, is unlikely to generate an optimal level of state or local debt. The reason lies in the temporal externalities implicit in the issuance of debt. Officials can obtain current benefits by issuing debt for capital projects, regardless of their long-term viability, while deferring costs for projects that turn out to be ill-advised to future generations. As a result, local officials who can garner political favor by funding projects that will prove costly only when those same officials are out of office have incentives to overextend the local treasury. Local residents may also fail to discern the consequences of debt, unless future debt service is fully capitalized into physical assets through property taxes. Even then, residents face the standard collective-action problem. Any increases in property taxes due to ill-advised debt is likely to be miniscule in the individual case although substantial in the aggregate. As a result, the incentive of any resident to monitor against fiscal overextension is limited. The few who receive more individually concentrated benefits from the project financed with debt, however, have substantial incentives to lobby for its adoption. Debt limitations help to offset that imbalance. Not all obligations of municipalities create the risk of political distortion, however. Localities incur obligations in a variety of forms, ranging from the commitment to pay the salaries of municipal employees on a regular basis to the funding of a project through bonds to be repaid over a

114

po lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

period of decades. Given the function of debt limitations, they reasonably apply only to those obligations that risk municipal payment for projects from which insufficient benefit is received. But someone must define that category. If debt limitations are effectively to constrain the impulses of local officials, the role of defining their scope would arguably have to lie with some external entity, and, again, the likely mechanism would be judicial determination of whether a particular form of obligation constituted debt as that term is used in constitutional restrictions. Importantly, not only do the state constitutional clauses with which I am concerned have the effect of limiting the conferral of private benefits, but they appear to have been enacted for this very purpose. Take, for instance, single-subject requirements. The commonly expressed concern that underlies these restrictions is the risk that cobbling together a series of measures, none of which could individually command a majority of the legislature, into a single piece of legislation will allow all of them to pass since the supporters of each will accept the unrelated measures in order to obtain passage of the one they prefer.10 Presumably, the coalition necessary to assemble disparate bills would consist of various minority interests that can enact their favored legislation only by combining with similarly situated groups. Any interest group that could command a majority for its proposal would presumably do so without incurring the costs related to logrolling with other groups. Thus, one might infer from the very existence of a bill that contains multiple subjects that it is necessarily a result of logrolls among minority positions. Of course, the single-subject requirement does not preclude all such logrolls; they may be effected by promises to trade votes on different bills. But the fact that all measures are combined in a single bill, all parts of which are voted on simultaneously, reduces the cost of monitoring and enforcing the logroll among its multiple participants. Aggregation of disparate provisions, each of which is favored by a minority interest, may be even more important in initiatives, since the sponsors of unrelated proposals will have little opportunity to monitor compliance with explicit logrolls that require separate votes among the electorate. As a result, interests that seek to enact their agenda through initiatives are more likely to accumulate a majority by seeking simultaneous voting. For instance, in 1988 the California electorate approved Proposition 105, the “People’s Right to Know Act.” An appellate court

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

115

subsequently invalidated the initiative for violations of the single-subject requirement.11 Under the “single” rubric of promoting public disclosure, the initiative included provisions requiring (1) that businesses advertising a household toxic product provide a warning that the product should not be placed in the trash or poured down the drain; (2) that public advertisements for insurance policies supplementing Medicare disclose that the offerer of the policy is or is not associated with a business which sells insurance if the name of the organization resembles a governmental agency, nonprofit or charitable institution, or seniors organization; (3) that all long-term health-care facilities’ admission contracts and advertisements contain a notice of the availability of additional information about the particular facility from the toll-free telephone number of the State Ombudsman’s Office; (4) that an advertisement favoring or opposing any statewide initiative or referendum contain a statement identifying the major funding source; and (5) that any corporation selling stock or securities in California disclose within its prospectus whether it is doing business with South Africa or any person or group located there and that a copy of such notice be filed with the secretary of state. In the context of either normal legislation or initiatives, however, the signal sent by the constitutional restriction is the same: a coalition of groups that represent only minority interests to enact legislation that contains no provision that a majority would approve is not the kind of interaction that one anticipates when multiple interest groups on different sides of the same issue bargain. The latter interaction represents the presence of political debate and a well-operating political market. Under those circumstances, little check on the process seems necessary. But the mutual trading of benefits among minority interests represents something closer to what Richard Stewart has aptly designated “Madison’s Nightmare” than the Madisonian Nirvana that Federalist #10 suggests is the consequence of factional competition.12 Again, if special legislation is problematic, a solution must be found outside of the political process itself. Those who engage in logrolls have no incentive to limit their promulgation. To the contrary, even for legislators who recognize the adverse effects of the practice the logrolling game looks like a classic prisoners’ dilemma in which each player adopts a strategy that each would avoid if all could enforce the same constraints on others. If all legislators were willing to forgo trading, decision making

116

p o lit ical marke ts a n d c o n s ti tuti o n a l c o ns t raint s

would avoid the downward spiraling race to the bottom that occurs with mutual trading. Thus, not trading would constitute a form of cooperation, while trading would constitute defection. Yet absent a cooperative moratorium on trading, the best strategy for each decision maker would be to defect from the cooperative solution while all other localities cooperate, thus obtaining a subsidy for the defector’s favored project. In contrast, the worst position would be to be the lone cooperator, whose constituents would find themselves heavily subsidizing multiple special programs elsewhere without obtaining comparable programs providing local net benefits. Of course, not all logrolls are negative-sum. A substantial literature suggests that logrolling may be a means by which legislators can register their intensity of preference over policy choices. Logrolls may actually enhance representation of divergent political views, the very goal that the countervailing forces of competing interest groups seeks to achieve. For instance, where local representatives are elected from districts, logrolls may be essential to enact legislation that provides net benefits to the locality if those benefits are concentrated in only a few districts. In those circumstances, representatives from the other districts have little incentive to favor the proposed project over other projects that will benefit their own constituents, even if they bear only minimal costs related to the proposal. If the representatives from districts that receive no benefit from the proposal outnumber those from districts that would benefit, the proposal is unlikely to pass, notwithstanding that the concentrated benefits of enactment exceed the project’s total costs. Nevertheless, the presence of negative-sum logrolling arguably explains findings that localities with representatives elected from districts spend more per capita than localities with representatives elected at large. Baqir’s study of city budgeting found that doubling the size of a city council generates a 20 percent increase in city spending per resident.13 Langbein, Crewson, and Brasher also find that, at least within their sample of cities with a council-manager form of government and a weak mayor, per capita expenditures are positively related to the number of elected members of the city council.14 Robert Inman similarly concludes from a case study of Philadelphia that increasing the number of local legislators elected from districts increases the per resident or per legislator budget.15 These relationships have been formalized into the Law of 1/n, which

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

117

holds that there is an inverse relationship between project efficiency and legislature size. The underlying intuition is that each district will pay a smaller share of a project’s costs as the number of districts increases, so that representatives from benefiting districts have greater willingness to fund projects from the common pool of the municipal treasury as they internalize a project’s benefits and share the costs.16 Even recent analysis that questions the scope of the Law of 1/n concludes that it has substantial force with respect to total government expenditures where there are few central government subsidies and with respect to project size where there is some sufficient congestion and small deadweight costs of taxation.17 For much the same reason, Haughwout and Inman propose that localities that are primarily concerned with optimal redistribution policies should replace ward-based city politics with at-large politics and require cities to elect an at-large mayor.18 I don’t want to reach that conclusion yet, since it is plausible that coalition building returns net benefits that justify the increased government expenditures. For the moment, I want only to suggest that the ubiquity of logrolling and its ambiguous consequences complicates the judicial role in extrapolating from the presence of dominant interest groups in a particular piece of legislation. Courts that face a challenge to one provision cannot readily reverse engineer the legislative process to determine whether the apparently malign terms were offset by other terms in the same legislation, or were traded for terms in unrelated legislation that benefited other groups. Indeed, even if courts could identify the separate components of the legislative deal, there would remain the difficult issue of deciding whether the existence of the trade signified the positive-sum logrolling story of a well-functioning political market that reflected the optimal allocation of public goods among contesting groups and incorporated intensity of preference rather than mere aggregation, or the negative-sum logrolling story in which various groups enlarge the size of government by trading projects, none of which would be pursued if the beneficiaries had to bear the entire cost. If courts subsequently invalidate one part of the deal in the erroneous belief that it was motivated by interest group domination, then the compromise that made the legislation possible will be disrupted. While that may leave public-regarding legislation intact, the threat of invalidation will frustrate subsequent efforts to form public-regarding coalitions. If logrolling tends to take a negative-sum form, especially where it is accompanied by

118

p o lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

dominant interest groups, constraints such as single-subject requirements become more comprehensible. Prohibitions on special legislation are a bit more complicated. Although there are variations among the states, these provisions generally preclude the state legislature from conferring benefits or imposing burdens on one locality without doing the same for similarly situated localities. The explanation for the appearance of these provisions in the late nineteenth century has frequently been stated in terms of a developing preference for local autonomy. The provisions allegedly corrected the perceived evil of state legislative interference with the local prerogative either by preventing a locality from undertaking a desired activity or by imposing on a locality an obligation it preferred not to take on itself. In either case, state intervention presumably occurred because representatives from other jurisdictions wanted to control the targeted jurisdiction. Frequently, the problem was perceived as rural efforts to impose burdens on corrupt urban officials. But Jon Teaford and Robert Ireland have demonstrated that local legislation in the nineteenth century was not necessarily a means by which the legislature could control localities. Rather, other forces were at work. Teaford and Ireland contend that special legislation was sought by representatives of the affected municipalities themselves in order to obtain some unique authority. As Teaford concludes, “though state legislatures were potentially masters of the city, throughout much of the United States they were actually compliant servants of urban spokesmen.”19 The state legislature tended to defer to the local delegation with respect to special legislation that affected only a particular city, and the legislation typically arose from within that delegation. Indeed, one perceived danger of special legislation was that delegates from other jurisdictions would be ignorant of or agnostic to its content. But Teaford’s observation that local legislation was usually sponsored by the local delegation does not mean that special legislation was enacted to serve the interests of the locality. It is unclear from Teaford’s account whether the “urban spokesmen” to whom the legislature deferred necessarily represented local interests or the interests of some subgroup within the locality. To some extent, Teaford suggests, special legislation appears to have been a means of resolving contested conceptions of the “local interest.” State legislative action, on this account, was orchestrated

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

119

by interests that represented only a minority view in the local political process, but that had sufficient influence in the state legislature to override the local majority. Reform groups sought state legislative intervention when they could not have their way in local elections. Business interests sought similar intervention when they failed to achieve their agenda with local officials. In what appears to be an odd manifestation of the contest but may represent an underlying battle over land use, Teaford recounts that Philadelphia representatives successfully sponsored a law that resulted in the transfer of the Philadelphia city hall, notwithstanding objections from the Philadelphia city council. Pursuit of special legislation, in short, could be the recourse of groups that failed to obtain a majority in the local electorate. If special legislation was being enacted at the behest of the target city’s delegation, but only because those legislators were attentive to minority interests that city officials had rejected, then special legislation “served the forces of localism”20 only if those appealing to the state delegation were more representative of local preferences than were the municipal officials whose decisions the state overrode. Of course, if the interest group comprised reformers who were appealing to the local delegation to untie the knots of a political machine that failed to serve the locality’s long-term interests, then special legislation might be seen as a purifier of the political process. There do appear to have been some examples in which reformers proceeded along these lines to shift power away from ostensibly corrupt officials to state bodies that were perceived as less vulnerable to nefarious influence. But Ireland suggests that special legislation was more often used as a mechanism to transform “individual legislators [into] . . . little dictators” with plenary authority over the cities they represented.21 Special legislation was not necessarily introduced to implement policies favored by the city’s officials or by reformers, but it could instead emerge in response to the entreaties of constituents with particular requests that the city could not or would not (and perhaps should not) address. Ireland concludes that the process was essentially undemocratic because city residents received no notification of pending proposals until the state legislature enacted them. Several commentators alleged that the laws had been procured by bribery or other misconduct, steps that presumably would have been superfluous if the policies actually advanced municipal interests. Special charters for privileged businesses

12 0

p o lit ical mar ke ts a n d c o n s ti tuti o n a l c o ns t raint s

constituted a significant amount of the special legislation that local representatives sponsored. Other legislation granted exemptions from taxes or other regulations to specific businesses. City officials themselves might refer proposals, but more to circumvent normal political processes than to promote municipal welfare. Special legislation, for instance, could be used to authorize taxation without voter approval that might be difficult to secure. As Ireland concludes: Additional research indicates that special laws relating to municipalities did not always benefit the interests of a majority of the residents of municipalities. Thus the governing authorities of Chicago, without giving advance notice to city residents or discussing the matters in any form of open meeting, secured tax increases, amendments to the city charter, and salary increases by special legislation. Had the authorities been required to make their constituents aware of these matters and to bring them before the city council, and to discuss them in open meeting, it is doubtful that they would have been approved. Special legislation allowed city leaders to defeat or at least evade democracy in other ways.22 The insertion of special legislation prohibitions into state constitutions may have effectively negated this circumvention of local politics. The effect of the prohibition, however, was not to eliminate interest group influence. Instead, it altered the forum within which efforts to exercise influence could occur. After adoption of the constitutional prohibition, interest groups that were unable to prevail at the local legislature could less readily seek a second bite of the apple by appealing to a more hospitable state legislature, unless they could demonstrate that the problem they were addressing was so endemic to a class of localities within which their locality fit to justify enactment of “general” rather than “special” legislation. Presumably, they would be able to do so only with the consent of the other affected localities—and that would serve as a strong signal that the issue reflected the interest of the public at large rather than that of a narrow, local group. But this result did not diminish the influence of groups that could dominate at the local level. Indeed, it may have enhanced their power by reducing the influence of groups that were incapable of exercising influence at the local level but that could dominate

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 21

at the state level. Reformers who depended on state legislative sympathy, for instance, would have more difficulty circumventing a local political machine even if they were more representative of local preferences. One who favors local autonomy can paint an optimistic view of these prohibitions. They arguably entail a conviction that, pace Madison, interest group competition is more likely to exist and to generate publicly interested decisions at the local level than at the state level. Thus, diluting the ability of the state to trump local decisions will increase local decisions that reflect local preferences. There is some reason to endorse that conclusion. The higher transportation costs that groups suffer in coalescing at the state capital and higher costs of lobbying a more numerous state legislature might mean that fewer groups form at that level than at the local level. As a result, those who successfully coalesce at the state level face less competition from adversaries. Alternatively, one may have a more agnostic view that, depending on the issue, there might be more competition at either the state or the local level, but it is unclear ex ante at which level better decisions would dominate. Thus, one might be indifferent as to whether the final approval of local action came from the locality’s municipal representatives or their state representatives. That being the case, ending the discussion at the local level would at least reduce decision-making costs, since one less decision would be required. But even if we are unsure where dominant interest groups are most likely to arise, the prohibition itself does appear to have emerged from a desire to constrain dominant interest groups and to encourage decision making at the level of government most likely to host competition among different views on an issue of municipal interest. State review of local activity through special legislation privileged those groups that could organize at the state level; the special legislation prohibition privileged those that could organize at the local level. The ultimate enactment of special legislation prohibitions can be interpreted as comprising a presumption that decisions by municipal representatives would be more representative of local preferences, and thus subsequent appeals to state representatives should be foreclosed. It seems clear that there will be instances in which that assumption is incorrect and the state will be a superior host of interest group competition. Indeed, where courts can conclude that the state provides such a forum, the underlying objective of special legislation prohibitions seems to augur for defining the resulting state legislation as “general” rather than “special.”

12 2

p o lit ical mark e ts a n d c o n s ti tuti o n a l co ns t raint s

These common constitutional clauses, therefore, have a unifying theme. They appear to have been directed specifically at diluting the influence of dominant interest groups. I am not suggesting that these clauses are necessarily successful in restraining malign redistribution. It is clear that they are not. Studies of gubernatorial line-item veto power, for instance, systematically conclude that item-veto authority does not substantially affect total spending or budget outcomes.23 State constitutional debt limitations arguably have been more successful in providing opportunities for bond counsel and investment bankers to create novel financing schemes that circumvent the limitations than in reducing overall state or local debt. I am suggesting, however, that these clauses imply a more substantial role for judicial monitoring of the local political process than is commonly considered appropriate in analyzing the relationship between federal courts and Congress. The pervasive presence of constitutional clauses that reflect a desire to reduce the influence of dominant interest groups on state and local legislators supports the general proposition that judicial interpretation and enforcement of the clauses is appropriate, since it would be anomalous to confer on the legislature the authority to define the scope of constitutional constraints that are intended to limit its authority. Thus, the implication of these constitutional clauses is that they signal a more substantial role for judicial inquiry into the process by which local (and state) legislation has been enacted than is associated with judicial review of the political process in the federal realm. This does not mean that courts are authorized to invalidate what they perceive as malign redistributive legislation on substantive grounds. But it does mean that where the form of legislation renders it subject to a challenge on the basis of one of the clauses that states have incorporated in order to restrain the influence of dominant interests, courts have the authority (I will later get to the question of whether they also have the capacity) to inquire into the process by which the legislation was enacted. Admittedly, given the potential difficulties of distinguishing situations of normal bargaining among interest groups from those dominated by a single interest group, it would be easy enough for a court that disagreed substantively with a redistributive program to invalidate it on process grounds. But to suggest that nebulous constitutional clauses permit judicial inquiry into political processes is not to admit that the clauses are infinitely malleable.

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 23

The possibility that a court could conflate the substantive and process basis for invalidating legislation does not transform the distinction into a case of sophistry in designating the appropriate scope of judicial intervention. Presumably legislation that takes the forms that are vulnerable to challenge on the basis of one of these clauses has a higher probability of reflecting the influence of dominant interests than legislation that falls outside these forms. While we might believe that allowing judicial intervention to reverse engineer any piece of legislation might be prone to too much error, the probability of finding some defect in the political process may warrant intervention in these cases. The multiplicity of clauses that are directed at constraining the effects of dominant interest groups, the history of these clauses, and the necessity that some institution interpret their scope create a plausible case that state constitutions invite a significant degree of judicial evaluation of the legislative process. The doctrinal support that these constitutional clauses lend to judicial monitoring of local and state political processes reinforces the second—institutional—basis for distinguishing between the relationships of the courts to the legislatures at the federal and state levels. Consider the recent literature that argues for more intrusive state judicial intervention in the creation and protection of individual rights than is frequently thought appropriate for federal courts that interpret the federal constitution.24 James Gardner, for example, discovers broader state judicial authority in the different powers allocated to state judiciaries and to the federal judiciary: a less rigid separation of powers among governmental branches, which permits judicial intervention into political questions; the absence of a need for a case or controversy for state courts to render opinions; and the more relaxed standards for standing or ripeness.25 Some state courts may face less criticism for interfering with “political” branches because numerous states elect at least some of their judges, although the forms of election, such as retention elections or nonpartisan elections, often make political competition for judges less contentious than elections for legislators.26 Notwithstanding increasing criticisms that judicial elections are becoming tainted with a political brush, the fact that some judges have some electoral accountability at least dilutes the claim prevalent in debates about the relationship between federal courts and Congress that the judiciary is obligated to defer to an elected legislature.

12 4

p o lit ical marke ts a n d c o n s ti tuti o n a l c o ns t raint s

Helen Hershkoff’s analysis of state legislatures similarly suggests that the relationships among state institutions mean that state courts may be particularly efficacious in considering the propriety of local redistribution.27 According to Hershkoff, the relatively amateur nature of state legislatures means that they are not characterized by the institutional safeguards—such as full-time membership or maintenance of a precise public record—that are present at the federal level and that can substitute for judicial review. Local legislatures similarly tend to be composed of part-time members who meet relatively infrequently and whose published proceedings bear little similarity to the Congressional Record or other indicia of accountability that give rise to claims at the federal level for judicial deference.28 At the same time, the relative ease with which state constitutions can be amended provides a more readily available corrective to misguided judicial activism, and thus arguably justifies more risk taking by courts and a more generous grant of judicial authority. Gardner and Hershkoff focus on the limitations of state legislatures to explain or encourage more intervention by state judiciaries in the protection of rights. The same institutional rationale, predicated on the relative lack of professionalism for local and state legislators, also increases the likelihood that state legislatures are more susceptible than their federal counterpart to the entreaties of rent-seeking institutions that can distort legislative action away from constituents’ interests.29 These distinctions between federal and subfederal institutional structures do not simply reflect technical differences between otherwise identical governmental entities. Rather, they reflect characteristics that determine the proper scope of each institution’s involvement in the law-making process. Legislatures deserve deference from courts because they have comparative advantages in matters relevant to democratic governance such as fact-finding, opportunities for deliberation, susceptibility to monitoring, and accountability to constituents. The negative implication is that the case for judicial deference wanes directly with the absence of those characteristics in legislative processes. Where there are multiple and strong institutional safeguards—such as committees, bicameral legislatures, competition between political parties, professional legislative staffs, public records of deliberations by full-time legislators—to check legislative favoritism, the need for external checks on legislative processes, such as intrusive judicial review, diminishes. To the extent that monitoring

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 25

of federal institutions against interest group dominance benefits from nonjudicial constraints, judicial intervention becomes superfluous at best and mischievous at worst. But the relative absence of such institutions at the state level perhaps makes judicial intervention the best alternative. As safeguards against dominant groups decline in number and intensity, the traditional argument that courts should refrain from intervention in the political process loses much of its appeal. The more casual processes of local lawmaking suggest an environment in which legislative history is less frequently available, ordinances are drafted by those interested in their enactment and with more-limited input from experts, and little in the way of commissioned empirical data supports policy proscriptions. Local legislators rarely fill their offices as more than a part-time obligation, and if there is any relationship between their legislative and private vocations, there is some fear that the former will cultivate the latter, as when attorneys with professional interests in local commerce serve on local legislatures that decide the appropriate scope of development. I am not arguing that the institutional differences between federal and state or local governments are necessarily sufficient to rebut the weighty arguments against full-fledged judicial review to detect rentseeking legislation. State courts also have limited fact-finding ability, cannot initiate litigation on their own, and cannot readily reverse engineer legislation to determine whether there were trade-offs in nominally publicly interested legislation for what ostensibly seems an interest-group bargain. State judges have no more insight than their federal counterparts into the motivations of individual legislators. One is reminded of Justice Scalia’s admonition that “discerning the subjective motivation of [a legislative body] is, to be honest, almost always an impossible task. The number of possible motivations, to begin with, is not binary, or indeed even finite. . . . To look for the sole purpose of even a single legislator is probably to look for something that does not exist.”30 Whatever truth lies in that proposition applies to local as well as federal legislative bodies. Thus, I am sympathetic to the recent cautious response of the United States Court of Appeals for the Second Circuit to an invitation to consider the process by which New York City took private property by eminent domain. The owners of the condemned property argued that government officials approved the project on the “pretext” of providing a public use in order to confer a private benefit on the developer. Although the

12 6

p o lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

plaintiffs apparently conceded that the project bore a rational relationship to well-established public uses—a conclusion that would typically foreclose further challenge to the propriety of the condemnation—they sought to reveal improper motive by deposing the officials, examining their e-mails, confidential communications, and other pre-decisional documents, and questioning the project’s budgetary assumptions. The court rejected the request, declaring that nothing in the jurisprudence of eminent domain required federal courts to scrutinize the mechanics of a taking in order to gauge “the purity of the motives of the various government officials who approved it.” Any judicial effort to inquire into the motivations of public officials would be fraught with “conceptual and practical difficulties” (and, in the case of federal courts’ examination of state and local decisions, with state-sovereignty and separation-of-power concerns).31 Indeed, there is some reason to be more suspicious of state court intervention in local legislative processes than of federal court review of congressional processes. True, the fact that state court judges may be elected suggests that their voice is as democratic as that of legislators. But that same fact also means that state judges are similarly susceptible to interest group pressures. Moreover, the fact that federal institutions have few local counterparts does not mean that local decision making contains no effective substitutes other than judicial review. The greater capacity of disaffected residents to exit and the greater proximity of residents to their officials can at least occasionally provide substantial antidotes to local rent-seeking that are less available at more-centralized levels of government. For all that I have suggested that collective action problems move local decision making from the ideal, those same problems may be even more prominent when the relevant officials are in a distant capital and are responsive to a larger constituency. Nevertheless, I am suggesting that insofar as the case against judicial intervention rests on considerations of institutional competence, there is at least a theoretical basis for contending that the institutional characteristics of state and local governments differ from those of their federal analogue. What would it mean, then, for courts to reverse engineer local legislation? Certainly it does not entail deconstruction of legislative history, which I have suggested is likely to be sparse or nonexistent in local settings, to determine whether representation by countervailing groups

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 27

was sufficient to guarantee the emergence of publicly interested legislation. Nor could it be based on outcomes alone, since legislation that favors concentrated groups and imposes burdens on diffuse groups serves equally as a hallmark of desirable redistributive legislation and undesirable legislation influenced by dominant interest groups. Instead, the limitations on judicial inquiry mean that review of legislative processes would have to entail a set of presumptions concerning the characteristics both of publicly interested legislation and of enactments that were so one-sided as to warrant invalidation. Those characteristics would serve as proxies for public and private interest, so that their presence could allocate the burden of proof for those attempting to demonstrate the purity or defective nature of the political process by which they were enacted. In the following chapter, I examine whether potential proxies can aid the judicial analysis. But before addressing that issue, it is useful to consider some of the rare situations in which the Supreme Court has implied that it could use municipal political process failures to pass on the validity of local legislation, and thus to imagine what judicial review of political process might entail. In these cases, at least some members of the Court have suggested that even without review of the type that Justice Scalia finds so problematic, a less intrusive consideration of local politics could detect characteristics of malign redistribution sufficient to overcome presumptions of deference to local legislatures. In City of Richmond v. J.A. Croson Co.,32 the Court confronted the issue of when race-based measures could be used to ameliorate the effects of past discrimination. The Court had previously upheld a congressional program of set-asides for minority contractors in federal construction grants. The city of Richmond had adopted its own Minority Business Utilization Plan, described as remedial and enacted to promote participation by minority businesses in public projects. The city’s program required certain contractors to whom the city awarded construction contracts to subcontract at least 30 percent of the dollar amount of the contract to businesses owned by minorities. Although there was no direct evidence of prior race discrimination on the part of the city in letting contracts or any evidence that the city’s prime contractors had discriminated against minority-owned subcontractors, proponents of the measure submitted a study that showed only 0.67 percent of the city’s prime construction contracts had been awarded to minority businesses during a recent five-year period, notwithstanding that Richmond’s population was 50 percent black.

12 8

p o lit ical marke ts a n d c o n s ti tuti o n a l c o ns t raint s

The Court invalidated the Richmond program. In distinguishing the case from the federal set-aside program that had been approved, the Court nominally focused on the constitutional mandate that Congress enjoyed to enforce the dictates of the Fourteenth Amendment’s Equal Protection Clause. Nothing in the Constitution conferred similar latitude on states or their political subdivisions. Indeed, the Amendment’s creation of congressional enforcement power was enacted out of distrust of race-based state legislative classifications. That doctrinal distinction may be sufficient to justify the different decisions. But, even so, a plurality of the Court concluded that it had to review a race-based program such as Richmond’s under a strict scrutiny standard and validate it only if local legislators had acted to remediate past or existing discrimination. In reaching that conclusion, the plurality indicated that without deep inquiry into the justification for race-based measures, courts would be unable to distinguish remedial or benign programs from those that were based on “illegitimate notions of racial inferiority or simple racial politics.” The same test had to be applied whether the challenged measure was intended to assist a group that was a historical target of discrimination, such as the Richmond program, or whether the measure benefited a traditional majority, as in standard cases of discrimination. Cases in which the racial group in power conferred benefits on itself were more vulnerable to judicial inquiry than those in which the racial group in power disadvantaged itself because only the former bore the characteristics of raw racial politics that the Equal Protection Clause was intended to eliminate. But the principal of racial politics applied regardless of which racial group was in power: In this case, blacks constitute approximately 50% of the population of the city of Richmond. Five of the nine seats on the city council are held by blacks. The concern that a political majority will more easily act to the disadvantage of a minority based on unwarranted assumptions or incomplete facts would seem to militate for, not against, the application of heightened judicial scrutiny in this case.33 Thus, at least for a plurality of the Court, it was the threat of redistribution fostered by dominant interest groups, rather than the identity of the beneficiaries of the program, that warranted strict judicial scrutiny.

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 29

That threat could be inferred from the basic political characteristics of the local decision-making process. The presumption created by those characteristics might have been overcome by empirical data that revealed actual discrimination in the Richmond construction industry that was in need of remediation. But the city had, in the plurality’s view, offered no such evidence. It is notable that the threat of dominant interest groups in Croson comes from a perception that a majority had exploited a minority rather than from the organizational advantages that collective action theory suggests minority interest groups will enjoy and that I have suggested creates much of the threat to local interests. That may be a peculiar feature of the racial basis of the proposed program, since the fear of purely racial politics implies such a small amount of room for interracial bargaining that would either forestall raw majoritarianism or permit minority group capture of the local legislature. But what is striking about Croson for current purposes is the willingness of at least some members of the Court to address perceived flaws in the local political process as the basis for invalidating local redistributive programs. An appreciation for the importance of political process is also implicit in the Court’s controversial decision authorizing localities to use eminent domain power for local economic development in Kelo v. City of New London.34 For all the uproar and backlash that the decision has generated, Kelo is in many respects a very conservative opinion that fits neatly within the tradition of balancing the need for flexibility in urban planning with political process protections. The majority found that the public use requirement necessary to the exercise of eminent domain could be satisfied where a taking for economic development “would be executed pursuant to a ‘carefully considered’ development plan.” Moreover, the Court adopted a broad interpretation of “public use,” reflecting deference to legislative judgments about the proper use of public expenditures and the proper interaction between government and business. Thus, the Court rejected any monolithic metric for economic development such as a “blight” requirement, at least as a federal constitutional issue. Instead, at least where the locality was proceeding pursuant to a “carefully considered” development plan, the Court was of the view that the judiciary should defer to the judgment that emerged from those legislative deliberations. At the same time, the Court did find it necessary to note that nothing

13 0

po lit ical marke ts a n d c o n s ti tuti o n a l c o ns t raint s

in the Kelo case indicated that the city’s development plan was adopted “to benefit a particular class of identifiable individuals.” The language used in Kelo is heavy with negative predicates. The majority decision does not grant blanket permission to local governments to use eminent domain for economic development whenever officials so desire. Instead, much of the language of the decision implies that given the political process by which New London decided to take the relevant properties, the Court could not detect any failure that courts could correct better than the political process itself. None of the indicia of either a majority ganging up or a minority ganging up that underlie our concerns about the exercise of eminent domain appeared to exist. That interpretation does not compel judicial restraint in all circumstances. It only establishes the need to articulate those conditions under which judicial intervention is warranted. Consider the language that the majority employs over and again to justify its deference to local decision making: “The takings before us . . . would be executed pursuant to a ‘carefully considered’ economic plan”; “[t]he disposition of this case . . . turns on the question [of ] whether the City’s development plan serves a ‘public purpose’ ”; “[t]he City has carefully formulated an economic development plan that it believes will provide appreciable benefits to the community”; “[g]iven the comprehensive character of the plan, the thorough deliberation that preceded its adoption,” the entire plan will be judged for its satisfaction of the public purpose requirement. The Court’s systematic emphasis on the presence of what it variously refers to as a “plan” or a “comprehensive plan” or a “carefully considered development plan” does not appear accidental. There is an analogy here to comprehensive versus spot zoning and the underlying rationale for the distinction is similar. The existence of a comprehensive plan, as opposed to a land-use plan that affects a single parcel or a small number of parcels, has two implications. First, it entails a process that involves significant engagement by multiple actors in public hearings concerning what should be constructed where and by whom. One would anticipate that any such process will attract competing views—not only between property owners and developers but also among developers, contractors, and planners. As a result, the capacity of a small rent-seeking group to impose its will on a complacent majority or an underrepresented minority is diminished.

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 31

Second, the presence of a comprehensive plan implies that the interests of multiple current landowners are at risk. Comprehensiveness and planning entail simultaneous consideration of multiple parcels. The importance of that characteristic is derived from the political concerns about eminent domain that I expressed at the outset. If eminent domain creates a risk of ganging up on a discrete minority of landowners who, by virtue of their small numbers, have little political power, then those concerns should diminish as the number of landowners increases. There is little need for the affected landowners to reach anything close to a majority to have effective political voice. Recall that most residents either will be indifferent to the proposed comprehensive plan or, more likely, will be insufficiently affected by it to warrant the personal costs of becoming involved to support or oppose it. But those directly affected, those whose personal landholdings are at risk, have sufficient incentive to become involved that even moderate numbers of them can swamp the political process by which a final determination is made. Thus, the implicit conditions established by the Court for judicial abstention ensure both a forum in which opposition to a proposed condemnation can be articulated—the hearings process that is implicated in the promulgation of a comprehensive plan—and an environment in which those who do oppose the plan have sufficient political weight that their voices can effectively be heard. This, of course, does not ensure victory for the vocal minority. It does not prohibit local officials from ignoring their loudly and intensely expressed opposition. But unless we believe that economic development is intrinsically or systematically inconsistent with the interests of the majority, it is not clear why a decision that considered but rejected minority views reflects anything other than an inevitable outcome of democratic processes. The negative implication of all this is that no similar presumption of procedural propriety can be entertained when eminent domain is exercised against a single parcel or a small number of parcels and then transferred to a particular private party. Just as spot zoning raises concerns that the person who got the exception had the fix in, in ways that are less probable when a locality adopts a comprehensive zoning plan, so too are the conditions for deference relaxed where the takings decision implicates so few parcels that one reasonably fears a heightened risk of abuse. Indeed, Justice Stevens invited just such an exception when he wrote,

13 2

p o lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

“a one-to-one transfer of property, executed outside the confines of an integrated development plan, is not presented in this case.” Hanging over those words is the unspoken parenthetical “And when it is presented, I will vote against it.” The Supreme Court’s willingness to countermand the Richmond City Council, and its suggestion of similar willingness to intervene where political processes have characteristics of taint, is reminiscent of one other source of evidence that the institutional relationship between state courts and state legislatures differs from that between their federal counterparts: State courts appear to have a certain comfort level playing precisely the role that I have suggested could be defended from a doctrinal and institutional perspective. I am not trying to infer an “ought” from an “is.” I am not even suggesting that state courts necessarily have a good track record in identifying instances where dominant interest groups have or have not caused malign redistribution. (I take up the issue of judicial accuracy in chapter 6.) Instead, I suggest that to the extent that state courts have been playing this role without strong objection from those adversely affected, the very practice may lend a particular legitimacy to the effort and indicates that injection of the judiciary into the lawmaking process has not caused a noticeable adverse effect on democratic governance at the local level. State courts have demonstrated ambivalence on the issue of discerning motive in order to evaluate the validity of legislative action. Certainly a variety of opinions demonstrates the reluctance to inquire into legislative processes that characterizes Justice Scalia’s distrust of judicial competence in this area. When New York City contended that the state legislature had improperly amended state tax law to prevent the imposition of a commuter tax on New York State residents (but not on non-New York State residents) in order to influence an election in a state senate district rather than to achieve any substantial state interest in New York City affairs, the New York Court of Appeals declined to consider the allegation: “The vortex of the legislative process is often the intersection of politics and policy. Speculation on the ‘political’ motivation of the Legislature as a judicial construct for statutory analysis in an appropriate area of legislative activity would be a slippery and dangerous slope. ‘If the Legislature might constitutionally pass such an act, if the act be clothed with all the requisite forms of law, a court sitting as a court of law cannot inquire into the motives by which law was produced.’ ”35

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 33

Other courts appear to have taken great steps to avoid any intrusive role in the interpretation of the clauses that I have argued serve as a hook for more aggressive judicial intervention in defining the scope of local redistribution. As I have suggested above, recent decisions concerning debt limitations reveal almost total abdication of any role in defining the limits of constitutionally restricted “debt.” A series of cases decided by the Alaska Supreme Court is representative of the general unwillingness of courts to get into the sticky business of distinguishing between those proposals that contain only a “single” subject and those that are constitutionally flawed by virtue of containing multiple subjects. Criminal sanctions provided for in a bill authorizing mortgage aid to earthquake victims were “fairly incidental to the general subject” of the bill and thus could be included within it.36 A law that provided for flood control could also provide for boat harbor projects on the theory that both were part of a “cooperative water resources development program.”37 One legislative act could authorize bond issues both for correctional facilities and for the public safety administration.38 Legislation addressing both state and municipal taxation “relate[d] directly to state taxation.”39 And the Uniform Land Sales Practices Act satisfied the constitutional mandate on the theory that “every section of [the act] in some respect concerns land . . . [even though] it is just as clear that many of its provisions have nothing else in common.”40 But other courts have felt more comfortable evaluating legislation of dubious validity through the lens of legislative motivation, especially when the alleged driver for legislative activity has been characterized as something akin to a dominant interest group. An Illinois appellate court has explained invalidation of legislation under a special legislation clause as appropriate where the statutory classifications “have an artificially narrow focus and which appear to be designed primarily to confer a benefit on a particular private group without a reasonable basis.”41 Other courts have recognized the role of special interests in procuring legislation that might be contrary to public interests, and the injection of constitutional constraints on legislation motivated by private benefits. Although those courts do not necessarily proceed to inquire into the motivations of the legislature when evaluating legislation, opinions that assail the role of special interests and invalidate legislation challenged under constitutional clauses identified as constraints on those interests implies that the

13 4

po lit ical mark e ts a n d c o n s ti tuti o n a l co ns t raint s

court has found a link between the legislation and inappropriate legislative favoritism.42 This is not necessarily to say that courts have an easy time detecting the presence or calculating the importance of special interest advocacy. Indeed, judicial efforts to evaluate the role of special interest groups may occasionally provide support for those who despair of the futility of the effort. Take, for example, the debate within the Kentucky Supreme Court when it decided whether a classification that taxed unmined coal at a rate drastically reduced from the rate that applied to other forms of real property violated a constitutional requirement that all property within a classification be taxed uniformly. The majority concluded that the distinction between unmined coal and other forms of property was unsustainable. In response to the argument that the constitutional constraints, rooted in a desire to avoid the conferral of special tax benefits on a favored group, should not prevent the legislature from “extending a benevolent hand to . . . others who honor their debt to the Commonwealth in ways worthy of legislative consideration,” the majority concluded: With special interests and pressure groups seemingly better organized, better financed and more powerful than ever before, we can hardly take judicial notice that the General Assembly should be free from constitutional restraint to manipulate the tax structure to extend “a benevolent hand” to those “who honor their debt to the Commonwealth in ways worthy of legislative consideration.” The proponents of the special tax treatment for unmined coal . . . have yet to provide any hard information in the record or in briefs as to how this statute benefits anyone but the coal owners. . . . Even if we were inclined to accept as fact that the people of Kentucky no longer need protection from the “corrupt and polluting grasp of wealthy corporations” (or the power of today’s special interest groups), we are not free to strike the constitutional mandate. We have no power to ignore the plain meaning of the Constitution when we believe it expedient to do so.43 Of course, the propriety of a proposed classification within the constitutional mandate of uniformity is anything but “plain.” Any proffered classification requires interpretation to determine whether it is sufficiently

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 35

rational to permit differential taxation. What does seem “plain,” however, was the court’s willingness to consider the likelihood that the challenged classification was enacted solely at the behest of the beneficiaries of the lower tax rate for unmined coal, and that no countervailing interest group had a sufficient stake in the outcome to generate debate. But a dissenting judge inferred just the opposite conclusion from the same facts. He suggested that it was the attack on the classification, rather than the classification itself, that was motivated by special interests. To the dissenting justice, foreign corporations that owned large tracts of coal had obtained the low tax rate through a legitimate political process that domestic interests were now attempting to undo: “The engine that has driven this lawsuit from the beginning to the majority opinion is that out-of-state corporations own large tracts of coal and pay a minimal tax on such property. I am sure that if it was possible to tax the land-holding corporations at a high rate and leave alone the thousands of relatively small-holding coal owners, we would not have the hue and cry over unmined coal taxes.”44 Indeed, some state courts take a view of their role vis-à-vis the political process in a way that reflects the standard federal court deference to congressional statutes. The conflict between more and less active investigations into political process recently played out within a single case that involved efforts to condemn an allegedly blighted area of New York City for development by Columbia University. In Kaur v. New York State Urban Development Corp.,45 a panel of the state’s Appellate Division invalidated the proposal, in part on the grounds that the relationship between governmental entities that supported the condemnation and the beneficiary significantly infected the substantive judgment about the propriety of the project. The decision was all the more remarkable because it followed by a matter of days an opinion of the state’s highest court that endorsed judicial deference toward local blight decisions.46 Nevertheless, the Appellate Division noted that the city’s Economic Development Corporation had worked with Columbia to develop a seventeenacre area for educational and public uses. Prior to this effort, there had been no allegation that the affected area suffered blight. Indeed, the court suggested that any blight in the area had materialized only in properties that Columbia had previously purchased and then allowed to deteriorate. The state agency required to approve a finding of blight employed consultants who had assisted Columbia and who used a methodology that

13 6

p o lit ical mark e ts a n d c o n s ti tuti o n a l c o ns t raint s

the court described as biased in Columbia’s favor. A subsequent consultant, presumably appointed to remove the taint of Columbia’s relationship with the first, apparently accepted and amplified the methodology, the findings of which the court described as “idiocy.” The agency then adopted the consultant’s report, although it failed to consider some of the very factors that the agency’s counsel asked the consultant to evaluate. The decision to condemn the subject property was apparently made only with the Columbia project in mind, rather than being made as a predicate to deciding how best to utilize the condemned property, and Columbia— rather than the city—underwrote the costs of studying and planning for the project. The record relied on by the court revealed that “every document constituting the plan was drafted by the preselected private beneficiary’s attorneys and consultants and architects.” The strong language in which the opinion described the methodology employed for defining blight, ranging from “preposterous” to “egregious,” suggests a court bent on disapproving the finding of blight. A few gratuitous attacks on Columbia University in the opinion reinforce that interpretation. Nevertheless, the court’s focus on perceived capture of the political process allowed it to reach that result in a manner that suggested that it was the agencies’ decision, not the court’s, that had been predetermined. The Court of Appeals quickly reversed the Appellate Division.47 The court began by reasserting the limited role of the judiciary in reviewing findings of blight. At most, there was a reasonable difference of opinion as to the condition of the subject property. As a result, the de novo review of the administrative record by the Appellate Division was improper. Any proper review of that record, however, refuted the claims that the Appellate Division accepted to the effect that the relevant agencies or contractors acted pretextually or in bad faith. Where the Appellate Division found no blight in the relevant area prior to Columbia’s ownership, the Court of Appeals found evidence in the record of longstanding deterioration. This combination of record evidence and an obligation not to look behind that record led the Court of Appeals to dismiss the Appellate Division’s implication that the blight determination had been driven by a contaminated process. The claim that I am making, therefore, is not that state courts reviewing the political process will perfectly distinguish between benign and malign redistributive legislation. Rather it is that when courts apply

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 37

constitutional constraints that are ambiguous in their scope but that in their origins and purpose seek to constrain the capacity of dominant interest groups to distort local preferences, it lies within the judicial role to consider the presence of dominant interest groups in evaluating the validity of legislation subject to those constraints. That appears to be a more common phenomenon of explicit judicial analysis than an examination of political motivations generally. For instance, in deciding whether local governments have the authority to enact ordinances that favor one identifiable subgroup at the expense of another—the very definition of a redistributive ordinance—at least some courts appear to be intuiting to results that constrain the ability of a dominant subgroup of residents to extract wealth from those perceived as less capable of effective resistance. Courts, for example, have narrowly construed local authority to restrict landlords’ ability to convert their property to condominiums or to control rental prices.48 In construing statutes that authorize municipalities to exact fees for services, courts have limited the ability of localities to impose on developers exactions for infrastructure that would benefit the entire locality as a condition of obtaining building permits.49 I will argue later that the perceived victims may enjoy more protection from exploitation than one might initially think; but that argument goes to the propriety rather than the justification or existence of judicial review of local redistribution. Indeed, ultimately the pervasive inclusion within state constitutions of clauses that explicitly or implicitly sanction investigation into rentseeking legislation can justify judicial measures to reduce the distorting effects of dominant interest groups even when other clauses alone are implicated. If specific clauses signal a general role for state courts to monitor the political process in order to neutralize dominant interest groups, then they may license similar analysis of other nebulous constitutional clauses that define the scope of local autonomy. Granted, even multiple clauses within the same state constitution that have similar implications would not necessarily entail that a preference for judicial intervention pervades the entire document. State constitutions tend to be less cohesive than the federal constitution; state constitutions are more frequently amended and thus contain provisions written at very different times, thus diluting the implication that they are unified documents with a consistent message.

13 8

p o lit ical mark e ts a n d c o n s ti tuti o n a l co ns t raint s

Nevertheless, failure to extrapolate from specific clauses could also lead to anomalous results. It would, for example, be peculiar to suggest that the influence of dominant interest groups should be relevant in the interpretation of whether legislation contains a single subject but not in the interpretation of a clause that equally implicated the role of groups in municipal legislation. Grants of home rule and doctrines of state preemption of municipal ordinances, for example, involve the same ambiguities as the scope of special legislation or the public purpose doctrine. Constitutional provisions for home rule and state preemption of local ordinances do not create or expand governmental authority. Instead, they allocate authority between state and local levels. Depending on their language and construction, home rule grants can both permit localities to initiate legislation without prior state authorization and immunize localities from state intervention, at least with respect to those activities that fall within the malleable parameters of “municipal affairs.” The common case in which state statutes trump local ordinances that either conflict with or are preempted by pervasive state legislation similarly allocates regulatory authority between different levels of government. As in the case of explicit but vague restrictions on the scope and form of legislation imposed by the constitutional clauses that I have discussed above, doctrines of home rule, conflict, and preemption require some further elucidation. As I noted above, state governments and local governments have incentives to exceed their appropriate jurisdiction, so the judicial function is to strike some balance between the risk of excessive uniformity (where state law inappropriately trumps local variations) and the risk of imposing excessive external costs (where localities are allowed to regulate affairs that require more standardization). Some very basic principles inform the appropriate allocation of authority between state and local governments. The state is more likely the appropriate jurisdiction to decide issues that have consequences in several localities, as the discussion of externalities in chapter 1 suggests. Conversely, the sorting, efficiency, or democracy-enhancing functions of local government suggest deference to local autonomy where externalities are minimal or are likely to be considered by local officials. Most cases in which localities assert authority to act and in which that authority is contested, however, involve something in between these extremes. As a result, no simple metric resolves these issues of the optimal allocation

p o lit ical ma r ke ts a n d c o n s ti tuti o n a l co ns t raint s

1 39

of government authority. In recognition of the flexibility of the doctrinal standards, many courts have entertained presumptions about the proper resolution of claimed preemption. But the nature of that presumption is itself a matter of debate. Some courts conclude that preemption should be restricted to cases where the legislature has expressly stated that intent, or at least a presumption against preemptive intent should arise in such cases.50 Other courts, however, appear to entertain the contrary presumption and find preemption unless the legislature has expressly rejected it. Forests have died so that law professors could search, thus far unsuccessfully, for consistent principles that would reconcile the various decisions of courts in these areas.51 The pervasive state constitutional signals of a desire to neutralize dominant interest groups point to a plausible strategy for resolving this allocation of authority. In contested cases, courts may allocate authority to that level of government in which dominant interest groups are least likely to emerge. That is, proper exercise of home rule may be more readily found where the locality appears to be the superior forum for robust debate among competing interest groups, while home rule may be more restricted where the state appears to occupy that position. Initially, this principle might be viewed as inhospitable to home rule. Standard theories, based on Madison’s principles of Federalist #10, suggest that this metric would systematically disfavor local decision making. Decentralized governments, on that theory, are more likely to be characterized by factional dominance than centralized ones. As a result, judicial allocations of governmental authority will systematically deny local autonomy where there is a colorable claim that home rule powers have been exceeded or that the state has engaged in conflicting or preemptory regulation of an area that the locality seeks to legislate. But recall from chapter 1 that the essentialist Madisonian conception is debatable. Some groups will have organizational advantages at the local level, some at more-centralized levels. On any given issue, countervailing groups may be more prevalent at one level than another. But since organizational advantages are contingent on a variety of factors, such as transportation costs, the emergence of an entrepreneur, group size, and diffusion of beneficiaries and payers, there is less reason than Madison suggests to believe that the centralized forum will systematically be the one in which multiple interests are represented and have significant

140

p o lit ical marke ts a n d c o n s ti tuti o n a l co ns t raint s

influence in the legislative debate. It may be more plausible for a court to identify the presence of organizational advantages in the individual case before it. Programs that appear to reflect a malign redistribution of benefits to dominant interests rather than a benign redistribution in order to facilitate economic growth might fall outside the scope of local authority, while programs that can more readily be justified as advancing local economic development because multiple adverse groups participated in the process would fall within the range of the local initiative or survive a preemption challenge. The living wage ordinance cases illustrate these various judicial strategies. Those cases frequently concern whether the municipality had the authority to enact the challenged ordinance, that is, whether wage regulation constitutes a “municipal affair” within the scope of home rule and, if so, whether the ordinance was nevertheless preempted by state law.52 Neither of these issues generates an obvious response. Some localities may have living costs significantly higher than those of others elsewhere in the state, so that local variations on a statewide minimum wage could be considered appropriately within the scope of “municipal affairs.” But, as the Louisiana Supreme Court maintained, it is also plausible that a high wage rate in one locality would disrupt regional labor markets sufficiently to classify the issue as a matter of statewide interest and thus beyond the scope of home rule. Even if one grants that localities initially possess home rule authority, it may be displaced if there is direct conflict with state law, that is, if the locality permits what the state prohibits or prohibits what the state law permits.53 When the state creates a minimum wage law that requires an employer to pay a wage rate of $x, is it simply creating a floor that municipalities may exceed if they believe that local circumstances justify it? If so, a municipal ordinance that requires a wage rate of $x+y does not permit something that the state has prohibited. Or does the state law, by permitting employers to pay a wage rate of $x, implicitly prohibit municipalities from requiring higher wages? There is no inexorable response to these inquiries, which may suggest that courts are ill-situated to improve on legislative judgments about their resolution. The cases rely on different rationales and reach different conclusions. But behind the disagreement may well lie different conceptions of what motivated the localities to adopt minimum wage ordinances and whether local governments are likely to serve as an appropriate forum

p o lit ical mar ke ts a n d c o n s ti tuti o n a l co ns t raint s

141

in which to debate the issues. Those courts that uphold the ordinances view them as publicly interested efforts to subsidize some local residents at the expense of a broad set of employers who are able to pass the costs to the local population at large. Those courts that reject the ordinances may instead see them as the result of effective but inappropriate capture by groups interested in securing higher wages. The most explicit invocation of this concern comes not from a state court but from a dissent in the Ninth Circuit decision that upheld the Berkeley, California, living wage ordinance. In concluding that the ordinance impaired the obligation of contract that the city had previously struck with the developers of a marina area that was affected by the ordinance, Judge Bybee argued that the limited scope of the ordinance posed the danger that “ ‘the legislature acted less out of concern for the general good than for special interests.’ ”54 As I discuss in the next chapter, there are reasons to share Judge Bybee’s concerns, especially where redistributive ordinances apply only to a small portion of the locality that enacts them. But, even if the legitimacy of state court intervention into the legislative process is less controversial than court intervention in the federal context, addressing those concerns assumes a level of institutional capacity on the part of courts to make the relevant inquiries about an ordinance’s evolution. I turn next to the propriety of that assumption.

ch ap t e r s i x

Proxies for Distinguishing Benign and Malign Redistribution

what does it mean for courts to reverse engineer local redistributive legislation in order to neutralize the distorting effects of dominant interest groups? I certainly cannot claim that review of the legislative history from which local action emerged can provide the needed evidence of interest group influence. Since interest group pressure at the local level to a large extent occurs outside the public processes of government decision making, through lobbying and promises of campaign efforts on behalf of receptive local officials, perusal of public records would not necessarily reveal the presence of dominant interests. Records of debates or committee hearings to determine the sponsorship of bills or witness lists would be useful. Nevertheless, since one of my arguments for allowing greater judicial intervention into the local legislative process than is allowed at the federal level is the absence from the local process of many of the public safeguards that pervade the federal process, I obviously cannot rely on judicial examination of what I have claimed exists only in inadequate form. Moreover, even if courts could detect the presence of a dominant group in the legislative process, perhaps by comparing the presence of winners in hearings with the absence of losers, that would not necessarily translate into malign redistribution, since, as I have suggested above, the interests of those present could significantly overlap with those of less organized constituents. 1 42

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

143

The difficulty of distinguishing between malign and benign redistribution could be eased if the former had distinctive hallmarks that tended not to be present in the latter. If that were the case, then decision makers could at least entertain a presumption about the nature of any particular redistributive proposal. The presence of such hallmarks could serve as a presumptive proxy for malign redistribution, so that courts interested in neutralizing dominant interest groups might be more likely to conclude that proposals bearing those characteristics fall outside the scope of local authority. At the least, courts could impose on supporters of ordinances with those hallmarks the burden of demonstrating that the proposals emerged from a process that included groups adversely affected by their enactment or that the proposals met some plausible definition of municipal interest. The absence of those hallmarks, in contrast, could serve as a presumptive proxy for a proposal that was consistent with local preferences. The very imposition of a burden to demonstrate emergence from a robust political process would presumably have the beneficial by-product of inducing advocates of a proposal to include more perspectives in the debate so that they would be better positioned to survive a subsequent challenge to the ordinance. If inducing a more comprehensive process has the positive effect of generating more-balanced legislation, then the imposition of such a burden may align local legislation with residents’ interests both at the front end (by generating legislation more consistent with those interests) and at the back end (by invalidating legislation less consistent with those interests). Any search for viable proxies begins with the nature of redistributive legislation. The ostensible desire of potential interest groups to motivate members to assist in obtaining a unique benefit, while deterring the formation of effective opposition, suggests that those who seek malign redistribution will want to structure a proposal in a manner that concentrates benefits and diffuses costs. Concentrating benefits increases the per capita rewards for any individual member and thus increases the probability that it is worth the while of a member to contribute to the collective effort. Concentration of benefits also facilitates monitoring members in order to avoid free riding. Diffusion of costs, on the other hand, reduces the willingness of potential opponents to resist the proposal, since the per capita costs of organizing may exceed the costs that they will suffer should the proposal be enacted. As a result, one might be tempted to infer from any

144

d is t in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

ordinance that confers benefits on a relatively small group and spreads costs broadly that malign forces are at work. Any presumption based on the concentration of benefits and diffusion of costs fails, however, because that phenomenon is inherently indicative of most redistribution, not just the malign. There is some redistributive governmental action that imposes burdens on a few in order to confer benefits on the many; eminent domain is perhaps the best example. But the very nature of redistribution, and certainly of redistribution in favor of the relatively poor, typically entails the creation of social benefits by taking small amounts from large numbers of people in order to increase the welfare of a few. Because that simple formula is present in most redistributive legislation, reflexively equating such distributions with rent-seeking behavior would condemn all such programs, the benign along with the malign. Jerry Mashaw has launched a broader attack on efforts to identify the propriety of governmental programs through simple comparison of who suffers costs and enjoys benefits. Mashaw has argued that classification of the costs and benefits of legislation may fail to track any underlying asocial intent. Different provisions within legislation involving multiple subjects may entail different distributions of costs and benefits that may change over the life of the legislation and that do not necessarily make oversight by the courts an appropriate curative even if the distribution signifies the successful intervention of rent-seeking special interests.1 Only if they could identify additional characteristics that systematically correlate with malign local redistributive programs would courts have a basis for using those characteristics to construe the scope of local authority to enact the program. But sponsors of malign redistribution do not necessarily only seek to obtain sufficient funding for the programs they advocate. They also have incentives to deter organized opposition, primarily by obfuscating the costs of the program, the imposition of those costs, or the personal benefits they receive from the program. Unlike the formulaic concentration of benefits and diffusion of costs, these additional efforts may be more susceptible to proxies that can be used by courts to detect malign redistribution. Proponents of such programs, for instance, might seek not only to keep individual subsidy costs low but also to shift those costs to particular nonbeneficiaries who, by virtue of small numbers or political powerlessness, are least likely to have an effective voice before the local decision maker. Alternatively, malign legislation may be implemented

d is t in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

145

through budgetary and expenditure devices that are relatively difficult to detect or quantify. While courts cannot easily reverse engineer local decisions to determine whether legislation was benign or malign in its effects, it is worth inquiring whether legislation that emerges from competitive or dominated political processes would tend to divide along characteristics on which courts could focus to draw conclusions about the benign or malign nature of challenged redistribution. R E D I S T R I B U T I O N T O T H E W E A LT H Y A N D T H E P O O R

We can begin with the mundane observation that politically influential groups have greater access to the local decision-making process, and wealth can translate into political influence through political contributions and opportunities for post–public service employment. The relatively wealthy may also have more opportunities to identify and monitor each other, through trade associations (where the members are firms that share an interest in redistributive legislation) or professional associations (where the members are individuals who share the interest), and thus to maintain a cohesive group. We might presume, therefore, that successful rent seeking is more likely to be present when the challenged redistributive proposal favors the relatively wealthy, but that the opposite conclusion should apply where redistribution favors the relatively poor. It might be appropriate, therefore, to begin by asking who benefits from the proposed subsidy. Mashaw’s concern about shifting benefits over time notwithstanding, at the very least it seems plausible to identify the initial sponsors of the local proposal and its immediate beneficiaries. It is perhaps on the theory that the relatively wealthy had undue access to the local decision-making process that courts, until recently, exhibited antipathy toward public financing for commercial enterprises and projects, such as stadiums for professional sports teams, that appeared to confer benefits primarily on owners of firms rather than on relatively needy residents.2 Industrial projects that promised increased employment might seem consistent with local development; but publicly funded subsidies for commercial projects, especially those that competed with existing enterprises, raised greater skepticism. This is not to say that the relatively poor cannot organize. Organizations such as unions or churches can serve as effective surrogates for the poor or focal points around which coalitions that include the poor can be

146

d is t in g u is h in g be n i gn a n d m a l i gn r e dis t rib ut io n

constructed. But as the debates about living wage ordinances in Chicago, Santa Fe, and Santa Monica reveal, even when the poor can organize to lobby for redistributive legislation, the relatively wealthy who bear the redistributive burden presumably have the capacity to create an opposing interest group. Those examples suggest that it will be more difficult for the poor to dominate the political process. Instead, the living wage debates look like exemplars involving countervailing groups that presumably give rise to desirable legislation. Thus, even if little can be inferred from redistribution to the relatively wealthy, redistributive legislation that favors the poor should arguably carry a presumption of public interestedness. Any simple dichotomy of the politically powerful wealthy and the politically disenfranchised poor, however, fails. It is unlikely that local redistribution to the wealthy results in malign redistribution with sufficient frequency that it alone signifies a political process failure, and local redistribution to the poor does not necessarily signify a working political market. Even when benefits are concentrated and costs are diffuse, redistribution to the wealthy may have positive effects that would be impeded by judicial invalidation based on the economic status of the immediate beneficiaries. Local subsidies to employers, for instance, are typically justified on the principle that they will increase local employment opportunities or attract other businesses within the network of the subsidized firm. While the literature is ambiguous about the social effects of interlocal competition for firms, any locality that can plausibly increase its tax base through employer subsidies can resist the claim that it is acting simply to confer benefits on a favored firm at the expense of other residents. Perhaps counterintuitively, the case against a presumption of malign redistribution seems strongest in those instances, such as subsidies for commercial enterprises, in which the subsidized firm will simply expand the existing commercial base within the locality rather than add a new dimension that could boost the local economy. I suggested above that courts have exhibited hostility toward such programs, perhaps on the theory that the low number of jobs or taxes associated with commercial development is more indicative of favoritism for the subsidized firm than of substantial municipal benefits. But the very presence of competitors within the jurisdiction means that a proposal to subsidize a new commercial enterprise is likely to generate opposition.3 Resistance to subsidizing a new Wal-Mart or other chain store may be offered in the name of maintaining high-wage jobs but will likely

d ist in g u is h i n g be n i gn a n d m a l i gn r edis t rib ut io n

147

be largely supported by the proverbial “mom-and-pop” stores that are likely to suffer from the lower prices that big-box stores offer.4 Moreover, if one believes that agglomeration economies make local redistribution possible without generating exit, then one would expect to see localities attempting to attract enterprises that are similar to those that already exist and that, therefore, exploit the economies present in having firms in associated industries located within a small geographic area. A new industrial plant, in contrast, may face few competitors among current residents. It may, therefore, be better able to extract subsidies from the locality without generating the opposition that firms who would face a competitive disadvantage might raise. That same plant, however, may actually be more likely to generate economic benefits to the locality, so that the absence of organized opposition does not necessarily denote malign redistribution. Even in the absence of opposition from competitors, moreover, relatively wealthy beneficiaries are likely to face scrutiny from other quarters and thus be less able to exploit their political advantage. As I discuss below, local media may have difficulty scrutinizing the details of the municipal budget. But large capital projects of the type that relatively wealthy groups are likely to seek will be more amenable to media attention out of a fascination with sheer size or because those items have more obvious implications for alternative uses of the same municipal funds. Since partisan politics tend to play a lesser role in municipal elections, potential opponents of incumbents are likely to exploit issues that have salience regardless of political ideology, such as financial impropriety. Aspirants for political office, therefore, have incentives to monitor for issues that will attract voter attention; perceived “giveaways” to the wealthy from the public treasury are more likely to satisfy that objective than detailed analysis of fiscal health. The net result is that there is reason to believe that projects that malignly benefit the relatively wealthy will not necessarily go unchallenged, notwithstanding the diffuse costs that they impose on the municipal residents. One might think that redistribution toward the relatively poor could serve as a more fruitful proxy, this time for benign rather than for malign proposals. The variables that facilitate collective action on behalf of the wealthy (repeat play with local officials, ability to contribute to political campaigns, tendency to participate in politics) seem, at least initially, to be absent in the case of the poor. While Glaeser and Kahn state that the poor “are better endowed with votes than they are with income,”5 low turnout

148

d is t in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

rates for the poor imply that political officials who seek to maximize their chances for reelection are less likely to be sympathetic to programs that will assist the poor.6 The lack of organizational resources that the poor possess to capture the local legislature suggests that redistributive programs that favor them must command broad support. Moreover, should some entrepreneur be able to organize for malign redistribution, the relative organizational capacity of the wealthy who pay the subsidies should ensure that their opposition receives a full hearing. Thus, one might presume that redistributive programs that systematically assist the relatively poor, such as living wage ordinances, have been preceded by democratic debate and reflect local preferences. This presumption, however, rests on too thin a description of the conditions that underlie collective action. First, even if the relatively poor constitute a small percentage of the voting electorate, they may still compose an effective voting bloc if they commonly vote their economic interests. Since redistributive policies will provide discrete benefits whose costs are diffused through the remainder of the population, local officials may be willing to offer such programs in the belief that they will be sufficiently salient to the recipients to translate into significant votes and will not generate offsetting opposition from nonbeneficiaries who bear only small per capita costs. Politicians themselves, therefore, may supply redistributive benefits to the poor in an effort to create a voting bloc that ensures their own incumbency. At the extreme, this strategy may entail sharing the spoils of corruption with the relatively poor, in order to induce the wholly predictable exit by the relatively wealthy and the inevitable deterioration of municipal services.7 Second, even if the poor cannot readily coalesce, they may have well-organized and politically powerful surrogates, such as churches or unions.8 Representation of the poor through surrogates complicates the effort to determine the nature of the redistributive effects. Frequently, surrogates will themselves enjoy some benefits as a consequence of their advocacy for the poor. Bureaucrats in agencies that serve the poor may enjoy larger budgets from which they obtain benefits if they can expand their projects;9 skilled unions may be able to push their own wages higher if they successfully advocate higher minimum wages for the unskilled.10 When surrogates act on behalf of the poor, therefore, it can be difficult to disaggregate those efforts that altruistically, and benignly, increase the welfare of the poor in a manner that creates net local benefits and those

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

149

efforts that enhance the welfare of the sponsoring group but that reduce the welfare of the locality as a whole. The evidence on whether surrogates for the poor increase welfare expenditures to efficient levels or cause more malign overexpenditures is mixed. Consider, for instance, the possibility that teacher unionization could cause public school expenditures that are inefficient in the sense that they are not valued by residents. One would expect that land values would decline in localities that spend either more or less than residents prefer. Proceeding from that intuition, Lisa Barrow and Cecilia Elena Rouse find that additional expenditures on public school education are correlated with decreases in property values in large school districts, in those without significant competition, and in areas in which residents are poor and less educated.11 Nevertheless, they find no evidence that unionized school districts are more likely to spend beyond the optimal amount on education. Other studies, however, are less sanguine on this point. Andrew Haughwout and Robert Inman find that cities with strong municipal unions have significantly depressed land values (by approximately 12 percent) relative to cities without such unions.12 Caroline Hoxby finds that teachers’ unions increase spending on education, but that those additional inputs are actually correlated with reduced productivity and a negative effect on student performance.13 Robert Ellickson similarly suggests that the “poverty services industry” in New Haven exercises significant power in the distribution of local public funds and often does so in a manner that directs the funds to projects that serve a small subset of the community (the organizers of the funded program) rather than the broader public.14 Community leaders may obtain funds for their favored antipoverty projects in return for political support, notwithstanding the absence of any correlation between those projects and the preferences of city residents generally.15 These political effects are likely to be exacerbated in cities that utilize wards or districts for purposes of electing representatives to a city council. Ward politics encourage a pattern of logrolling in which each representative seeks to generate benefits for his or her constituents, even though the municipal benefits from any of these expenditures is worth less than its cost. By trading favored projects with other representatives, each can obtain citywide funding for projects that return highly localized benefits and thus represent to his or her constituents that those benefits were obtained at a cost below what the beneficiaries were required to pay.

150

d ist in g u is h in g be n i gn a n d m a l i gn r e d i s t rib ut io n

As I argued in the previous chapter, logrolling of this sort is likely to have significant effects on the municipal budget. But for current purposes, the lesson is slightly different. All spending that results from logrolling can be defined as redistributive to the extent that it causes residents from one part of the municipality to subsidize residents in other districts. Much of the resulting legislation is likely to fall within the malign category.16 To the extent that residents in the relatively poor districts are able to participate in the logrolling process, there is little reason to believe that the relatively poor districts will fail to obtain a share of the “excess” spending. One might contend that such districts are lacking in essential services, and therefore that logrolls involving these districts are more likely to involve the extension of those services to poorer areas, but I am not aware of any empirical support for that claim. These examples suggest that the relatively poor may be better represented in the decision-making process than one might initially believe. But they do not tell us much about the nature of redistributive proposals that favor the poor. Even where surrogates support redistributive expenditures that benefit the poor (as well as themselves), one would imagine that the relatively wealthy have the resources and incentive to organize against such proposals. As a result, the presence of organized interest groups in favor of the poor does not necessarily translate into dominance of the political process, even where that group extracts benefits from the challenged ordinance. Indeed, even if the benefits captured by the surrogate group constitute an overinvestment by the locality, it may be that the overexpenditures are preferable to underexpenditures on the part of the poor, and the involvement of the surrogate facilitates a better approximation of the optimal level of redistribution. Thus, the overexpenditures the surrogate obtains for its own purposes constitute the type of “payment” for serving as a private attorney general that I indicated previously might be acceptable in order to avoid underinvestment. SALIENCE AND OPACITY: ON- AND OFF-BUDGET EXPENDITURES

Local legislators who enact redistributive programs possess a vast toolkit from which to select the means of funding. They could, for instance, enact a broad-based tax or excise and dedicate the proceeds to the program’s beneficiaries. Funding a homeless shelter out of municipal property taxes would fall into this category. Alternatively, they could impose

d is t in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

151

a highly targeted user fee on some subset of constituents and exact full funding from that group. For instance, parking meter collections could be pledged to fund construction of a municipally owned golf course. Or, legislators could avoid taxes and instead impose regulations that require a subgroup to fund the redistributive benefits. Living wage ordinances that require private employers to pay higher-than-market wages or rent control ordinances fall into this category. One way in which these various funding mechanisms differ is in their transparency or salience to those who incur the related burdens. Salience matters because more-obscure funding mechanisms are more costly to calculate and to monitor. Expenditures from taxes are relatively salient, insofar as they require an explicit legislative allocation in a stated amount and affect large numbers of individuals, some of whom may have incentives to question whether the expenditure is appropriate. Regulations may be salient to their initial target, such as employers who must pay a living wage. But to the extent that the regulated party attempts to pass the costs on to customers and clients, regulations will be less salient insofar as those ultimate payers have difficulty calculating either the full or the personal costs and thus have little incentive to explore their propriety. It is plausible, therefore, that funding mechanisms could serve as a proxy for the relative propriety of a redistributive program. Proponents of a benign redistributive program and local legislators who implement it should expect that constituents who pay the subsidies would acquiesce either because they will receive net benefits from the program or they implicitly agree that the subsidies are worth paying. Thus, legislators should tend to make the costs of benign redistribution transparent because transparency signals fidelity to constituents’ preferences and it poses no threat to the legislators. Conversely, consistent with the literature on fiscal illusion, one would predict that proponents of malign redistributive programs and local legislators who deviate from constituents’ interests will raise monitoring costs by obfuscating the expenditures. Beneficiaries of malign redistributive legislation similarly have incentives to advocate transfers through less salient devices. Thus, perhaps programs that are funded in less transparent ways are more likely to represent examples of malign redistribution. One might initially conclude that expenditures made from general municipal revenues provide little practical transparency, notwithstanding

152

d ist in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

that they require an explicit appropriation. No individual municipal resident is likely to monitor even a published budget because the effect of any particular expenditure is too trivial to warrant examination let alone active opposition. Budgets of large cities approach the impenetrable, even when they are easily accessible on the Internet. New York City may be an outlier on this score, as on so many urban matters. But, as originally adopted, the city’s Expense, Revenue, and Contract Budget for Fiscal Year 2009 runs 707 pages, while the Capital Budget for Fiscal Year 2009 runs an additional 544 pages. A separate document contained more than 3,900 pages of supporting schedules.17 None of those documents contains anything close to a line-item schedule that would permit a constituent to discover the cost of a particular expenditure. The city of Cleveland’s 2008 budget consists of a more manageable, if less informative, document of 508 pages.18 Notwithstanding the obstacles to perusing balance sheets and expenditure estimates, some local actors, such as newspapers and political opponents of incumbents, have idiosyncratically high incentives to monitor officials and thus may overcome the collective action problem inherent in deciphering, analyzing, and publicizing municipal expenditures.19 Moreover, funds paid out of a fixed municipal budget create a zero-sum game, so that competitors for public funds are likely to monitor expenditures, especially to discover grants to other firms and organizations, in order to ensure that a sufficient balance remains for their own projects. As in the case of redistribution to the wealthy, potential competitors of incumbent office holders have incentives to scour the municipal budget to detect and publicize financial malfeasance that may provide an edge with the electorate. Common legal doctrines also facilitate constituent monitoring by increasing the transparency of the municipal budget process. Statutory requirements of clarity and specificity in appropriations, and of competitive bidding, have the effect of reducing search costs for citizens who want to examine municipal expenditures to determine the beneficiaries of governmental largesse.20 The result is that local legislators can anticipate that direct expenditures may be publicized, and thus they have some motivation to ensure that those expenditures align with constituents’ interests. Off-budget expenditures, in contrast, are less susceptible to constituent scrutiny, both because information is more difficult to obtain and because, for most residents, the incentives to obtain it are diluted. Think, for instance, of tax abatements to businesses that the municipality

d ist in g u is h i n g be n i gn a n d m a l i gn r edis t rib ut io n

153

is attempting to attract. The amount of the abatement, or its percentage of otherwise applicable property taxes, is likely to be a product of an individual negotiation that is not readily accessible to other residents. Similar difficulties of opacity arose in the plan to permit the New York Yankees to construct a new Yankee Stadium on city-owned land with tax-exempt bonds financed through an agency of New York City. The bonds were payable from payments in lieu of taxes, or PILOTs, to be remitted by the team to the city. In effect, the city abandoned any discretion it otherwise would have been able to exercise over use of the funds represented by the PILOTs. In order to qualify for the federal tax-exemption on bond interest, however, the PILOTs could not exceed property taxes that would have been assessed on the land had it been privately held. In an interesting twist, it was to the advantage of the Yankees to have a high valuation on the land, so that they could justify making PILOT payments sufficient to finance the full $900 million cost of the stadium. When asked to provide an assessment figure that could be used for the Internal Revenue Service in an effort to obtain tax-exempt status for the stadium bonds, the New York City Department of Finance offered a figure of $204 million. The department, however, had assessed the same property at $21 million for purposes of a state requirement to create alternative parkland to replace the land dedicated to the stadium. Was the city properly using different methods for different purposes in a good faith effort to provide accurate assessments in each case? Or were city officials engaged in efforts to assuage the demands of an important interest group by manipulating assessment figures depending on their use? Even congressional hearings on the issue do not appear to have reached definitive conclusions.21 Given the difficulty of obtaining relevant information about off-budget expenditures, it is unclear who would incur the costs of discovering it. The costs of abatement programs are spread to all municipal residents through a reduction in municipal revenues. Unlike direct expenditures, these programs have no zero-sum quality to them; the grant of an abatement to one entity does not directly preclude the grant of another tax abatement to another entity. As a result, tax abatements create less incentive for any other resident to monitor or complain publicly, at least where the recipient does not have a local competitor who would be disadvantaged by the grant of the abatement and who thus has reason to oppose it (though the competitor may instead respond by demanding its own abatement).

154

d is t in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

The result is that legislatures face few inducements to ensure that offbudget expenditures are cost effective or would be approved by residents. Redistributive legislation funded by off-budget expenditures, therefore, might systematically be presumed to reflect malign motives. Nevertheless, there is reason to be cautious about equating on-budget expenditures with publicly interested legislation and off-budget expenditures with malign exploitation. Those equations rest precariously on the assumption that legislators balance the costs and benefits of on-budget expenditures but not of off-budget expenditures. As Daryl Levinson has suggested, local officials who appropriate the local budget are spending the public’s money, not their own. There is little reason to believe that officials who expend the public’s money have incentives to internalize the costs of their activities in the manner that residual owners decide how to invest the funds of their own firm.22 The constraint of maintaining broadbased political support by operating an efficient budget can be offset by the desire to maintain the political support of particular groups by allocating funds in a manner consistent with their more limited interests. In addition, some on-budget expenditures may have nonverifiable malign redistributive effects because they are substitutes for nonredistributive or benign redistributive expenditures. Think, for instance, of public employment. In theory, municipal employers would hire those individuals who are most qualified for the particular positions. In practice, however, public positions may be filled, or even created, for redistributive purposes. Some of this redistribution may constitute “a disguised way of channeling resources from middle class voters to disadvantaged citizens when an explicit tax-transfer scheme would not find political support.”23 While redistribution through employment may be inefficient, insofar as unqualified employees do not perform at a level that warrants the wage they are paid, it may also be benign insofar as it circumvents opposition to more direct efforts at desirable redistribution. Thus, Alesina, Baqir, and Easterly find that cities that suffer more income inequality utilize a greater amount of public employment, perhaps to reduce inequalities that would otherwise frustrate urban growth.24 But inefficient employment may also be used to redistribute revenues to politically favored groups that are essential to the political success of the official who authorizes the expenditures.25 These inefficient and arguably malign transfers can occur notwithstanding that they are part of a fully disclosed program, in

d is t in g u is h i n g be n i gn a n d m a l i gn r edis t rib ut io n

155

large part because their inefficiency cannot readily be detected, even if the expenditure itself can be. The equation of off-budget expenditures with malign redistribution is similarly suspect. Off-budget expenditures may be perfectly appropriate financial tools for achieving local preferences where legal constraints preclude on-budget expenditures for reasons that are unrelated to efficiency. Tax abatements may encourage prospective employers to migrate to or remain in the locality. More-direct appropriations might violate constitutional constraints on expenditures to private parties, even though those restrictions reflect neither an ability to make the associated payments nor a need to attract additional economic actors. As a result, subsidies structured in a manner that circumvents these constitutional restrictions may reflect the obsolescence of the latter more than a prophylactic against rent-seeking redistribution. Mandates imposed on private employers, such as living wage obligations, are necessarily “off-budget” in the sense that they do not directly implicate municipal expenditures, although those mandates effectively constitute subsidies to beneficiaries. As a result, any efforts to identify the transparent with the benign, and thus to create any judicial presumption about such expenditures, is unlikely to be successful. The standard public choice story of regulatory mandates is consistent with the argument that they are utilized to obscure subsidies to favored interests by circumventing the budgetary process. Mandated expenditures may increase costs, but because governmental expenditures are not implicated, those costs are more difficult for payers to quantify; political actors can use mandates to ingratiate themselves to groups that benefit, and simultaneously to avoid blame from the ultimate end users who pay higher prices as a consequence. After all, the higher prices are reflected in the goods or services provided by the private parties subject to the mandate rather than in public taxes or fees. Think, for instance, of New York City’s Local Law 11 of 1998. The law was enacted after several incidents involving the separation of bricks from building façades and at least one subsequent death. The city council, in its infinite wisdom, enacted inspection requirements for all buildings in excess of six stories in height at intervals of no less than every five years and the repair of any unsafe conditions. The new law certainly conferred a benefit on the architects and engineers authorized to conduct investigations,

156

dist in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

bricklayers who would make repairs, and providers of scaffolding that have dotted the city’s landscape to permit examination of the 12,000 commercial and residential buildings that are subject to the mandated procedures. It is less clear whether the costs of compliance, perhaps compounded by injuries sustained in erecting and working on scaffolding, and that building owners presumably seek to incorporate into rents, will offset the public benefits. But while most analysts can easily recall their favorite example of a mandate that appears to impose excess costs on private providers, even these regulatory mechanisms do not necessarily denote malign redistributive efforts. Given the administrative burdens attached to regulatory programs, including the direct costs of administration and the political risk that expenditures may be used in a manner inconsistent with the program’s stated objectives, it is not clear that regulation is necessarily less efficient than subsidies effected by payments to beneficiaries from tax revenues. Whether regulatory mandates make sense depends on a variety of factors, such as whether government taxation and provision of the redistributive service will be as effective as private provision, and the probability that regulated entities will take socially undesirable evasive steps to avoid regulatory costs. A municipality, for instance, may rationally conclude that a private developer that wishes to market a development would do a better job of constructing affordable housing than the locality itself.26 But that same municipality may also rationally conclude that regulation will deter private developers from entering the municipality at all. Mandates may also be effective ways of solving collective action problems among private actors, each of whom could incur costs that would produce benefits to itself and others similarly situated, but none of whom would desire to incur those same costs unless the others contributed. Assume, for instance, that all business owners in an area would benefit from additional security that would deter crime. No single business owner would be willing to incur the costs of the new security because he would receive the same benefit if other business owners incurred the costs. But each business owner might be willing to share in the costs if she was confident that her neighbors would also contribute. It is plausible that business owners could contract for contributions, but the transactions costs related to obtaining mutual consent, collecting fees, and solving holdout problems could significantly reduce the effectiveness of a private

d ist in g u is h i n g be n i gn a n d m a l i gn r edis t rib ut io n

157

solution. If the city was to mandate the payment of fees, however, then presumably all participants would be willing to contribute their share, secure in the knowledge that other business owners would be obligated to make their “contributions” as well. Given that the benefits of the additional security are limited to one area of the city, it might not be appropriate to impose the financial burden on the city as a whole. Thus, the role of the locality may be limited to providing the assurance of mutual cooperation that business owners need in order to make expenditures that they prefer but might not otherwise make. Under these circumstances, mandates may be superior to broad-based taxes and budgetary allocations and do not necessarily indicate interest group capture to redistribute wealth from less powerful constituents. R E G U L AT O R Y M A N D AT E S A N D I N T E R E S T G R O U P S I Z E

The prior section demonstrated that the problem of opacity increases when a program of local redistribution is implemented through regulations that require private outlays rather than through expenditure or forgoing of governmental revenues. Since government does not make the redistributive payments when it imposes regulations on private actors, the legislature has less incentive to compare a proposal’s costs and benefits, and less fear of retribution from constituents who will blame lawmakers for higher taxes. In this sense, redistributive programs such as living wage ordinances mimic requirements that employers or schools make “reasonable accommodations” under the Americans with Disabilities Act,27 or legislation that requires factories to install certain pollution control equipment. Indeed, some residents may be willing to support mandates that impose the payment obligations on a discrete group of payers, rather than make them through the tax system, because such mandates allow satisfaction of redistributive preferences without requiring any personal financial outlay. Residents who do not employ low-wage workers, for instance, might be more supportive of a living wage ordinance that assured covered workers of higher pay but imposed the payment obligation on a small subset of employers rather than on public subsidies. The risk of either raw majoritarianism or interest group capture, in short, is exacerbated where the off-budget expenditure has few implications for the vast majority.28 Certainly those who are regulated and who must therefore bear the related costs have opportunities and incentives to object. The question, therefore,

158

dist in g u is h in g be n i gn a n d m a l i gn r e d i s t rib ut io n

is whether, relative to the beneficiaries of the regulation, those who are regulated can coalesce with sufficient ease and influence to support a conclusion that the redistributive decision emerged from a political process that fostered benign results. Perhaps the presence of a redistributive mandate that places the compliance burden on a group with a low likelihood of being able to compete for legislative attention is itself a proxy for malign redistribution. Some years ago, Justice Scalia seized on just these concerns to justify his conclusion that a rent control law effected an uncompensated taking if it permitted a hearing officer to take into account “hardship to a tenant” when determining whether to approve a proposed rent increase. Here is Scalia, opining that redistribution through regulation allows interest groups to obtain benefits outside the scrutiny that taxpayers, and perhaps applicants competing for the same municipal funds, would apply to explicit appropriations of tax dollars: The politically attractive feature of regulation is not that it permits wealth transfers to be achieved that could not be achieved otherwise; but rather that it permits them to be achieved “off budget,” with relative invisibility and thus relative immunity from normal democratic processes. San Jose might, for example, have accomplished something like the result here by simply raising the real estate tax upon rental properties and using the additional revenues thus acquired to pay part of the rents of “hardship” tenants. It seems to me doubtful, however, whether the citizens of San Jose would allow funds in the municipal treasury, from wherever derived, to be distributed to a family of four with income as high as $32,400 a year—the generous maximum necessary to qualify automatically as a “hardship” tenant under the rental ordinance. The voters might well see other, more pressing, social priorities. And of course what $32,400-a-year renters can acquire through spurious “regulation,” other groups can acquire as well. Once the door is opened it is not unreasonable to expect price regulations requiring private businesses to give special discounts to senior citizens (no matter how affluent), or to students, the handicapped, or war veterans. Subsidies for these groups may

d is t in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

159

well be a good idea, but because of the operation of the Takings Clause our governmental system has required them to be applied, in general, through the process of taxing and spending, where both economic effects and competing priorities are more evident.29 Saul Levmore had earlier made a similar point in a different context. In attempting to distinguish those state subsidies and preferences for domestic firms that violated the dormant commerce clause from those that did not, Levmore noted that courts tended to allow interventions into the market that the state government funded directly, but they tended to reject interventions that were contracted for outside the scrutiny of the political process that characterizes explicit appropriations.30 Levmore explained this phenomenon as a manifestation of a judicial preference for “conscious funding,” a concept that is reminiscent of Scalia’s description of the “intelligent democratic process” that he attributed to tax collection and explicit expenditure.31 A legislature that appropriates funds for a subsidy, Levmore suggested, presumably has balanced the costs that its action would impose on the treasury against anticipated benefits and against alternative uses of the same fixed amount of dollars. But no such inference applies, Levmore implied, where the legislature has conferred a preference, such as a preference to hire in-state workers, the costs of which are borne by private parties. Hence, one might equate the opacity of regulatory costs with a presumption of malign legislation. Certainly, this opacity suggests a cautious reaction to those mandated expenditures that could, in theory at least, have been implemented through explicit subsidies funded through broad-based taxation. Justice Scalia, for instance, implied that he would not have had the same objection if the local legislature had offered the tenants hardship assistance funded from the city treasury. But here again the issue of benign or malign legislation is more complicated. Put to the side Julie Roin’s very plausible claims that administrative convenience, a sense of fairness, or the presence of compensating sources of accountability complicate the public choice arguments that Levmore and Scalia employ against regulatory redistribution.32 The more problematic objection for Justice Scalia’s analysis emerges from the very principles of collective action theory on which he relies. Scalia’s concern about “relative invisibility and thus relative immunity” imputes a greater

160

dis t in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

transparency to the process of taxation and spending than to the process of regulation. His claim is that the beneficiaries of redistribution effected through regulation will be able to obtain benefits without public scrutiny or the comparison of alternative expenditures that inheres in allocating a fixed budget. Expenditures from a fixed budget, after all, have a zero-sum quality that is absent from regulations, which can simply accumulate. Groups that seek redistributive transfers through the regulatory process, in Scalia’s view, would likely face the collective opposition of the majority if they sought equivalent benefits through a budgetary allocation. But collective action theory also cautions skepticism of Scalia’s dichotomy between regulation and budget making. His analysis rests precariously on the assumption that legislators balance the costs and benefits of on-budget expenditures but not of off-budget expenditures. He assumes that constituents will ignore regulatory decisions that impose costs only on private parties (for instance, landlords of rent-controlled units), but they will be sufficiently attentive to the opportunity costs of public funding decisions that they will scrutinize proposed redistributive programs funded from the local treasury.33 But think about the structure of the local budgetary process that Scalia finds so democratically transparent. Redistribution through the municipal tax system involves the allocation of small per capita tax payments from large numbers of residents to a relatively few beneficiaries. The resulting characteristic of any redistributive program is that it will impose diffuse costs throughout the municipality and confer concentrated benefits. This is the standard situation in which collective action theory tells us the burdened majority has the least incentive to monitor expenditures. Since, for any net payer, the per capita payment from the treasury for the redistributive program is likely to be trumped by the costs of monitoring the program, and since each net payer has incentives to free ride off the efforts of others to provide the public good of monitoring, oversight of specific expenditures is likely to be undersupplied. Thus, it is unclear that Scalia’s preferred process for conferring subsidies will be any more inclusive of competing views than the regulatory one he disdains. Instead, we are faced with a difficult anomaly. On the one hand, supporting a redistributive program through a broad-based tax reflects widespread acquiescence by virtue of residents’ willingness to contribute and hence is a hallmark of benign redistribution. On the other hand, that same characteristic implies the difficulty of

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 61

generating an effective interest group to which it is worthwhile to object, a hallmark of malign redistribution. Consider next the “spurious” benefits that Justice Scalia suggests can more readily be attained by groups that employ the regulatory process. If we focus solely on the beneficiaries, then Scalia’s implicit contention that per capita benefits may be sufficient to justify individual participation in the regulatory process seems reasonable enough, though not necessarily more so than if the same beneficiaries pursued transfer payments through the budget. But once we look at those who bear the costs, matters change. Some regulation will impose costs on diffuse groups whose members cannot readily identify each other or who do not bear sufficient costs to justify collective opposition. Roin suggests, for instance, that minimum wage legislation imposes costs on those who will not receive jobs they otherwise would obtain and on consumers who pay more for goods or services. Neither of these groups, however, is likely to be directly represented in the debate about the legislation. Nonetheless, the fact that the regulation does not require robust interest group competition does not mean that interested parties will not object. The same wage regulation is likely to be at least nominally applied to a discrete group—employers—so that the costs of the program, as well as the benefits, are concentrated.34 Similarly, in Justice Scalia’s rent control case, a small number of residents may qualify as hardship tenants, but a small number of landlords will bear the burden of stabilizing the rents of the beneficiaries. As a result, those who bear the costs of the program may also qualify as the small, privileged group that standard theory suggests is able to overcome obstacles to collective action. Since they will have multiple reasons to address land use regulations considered by the city legislature, landlords also have incentives to coalesce into an effective group that can amortize the costs of organization over several issues. Thus, even if the population at large is unaffected by, and hence indifferent to, the regulation, the costs will be highly salient to a group that has an intense interest in not bearing the redistributive burden for the entire locality. One might initially believe, therefore, that neither the beneficiaries nor the subsidizers of regulatory mandates will be able to dominate legislative debate. This sounds like the Madisonian Nirvana in which legislatures proffer publicly interested proposals after hearing from all interested parties rather than Scalia’s prediction of minority tyranny.

162

d ist in g u is h in g be n i gn a n d m a l i gn r e d i s t rib ut io n

But here, too, the issue is more complicated than first impressions suggest. Small group size may indeed facilitate communication among and collective opposition by those who bear regulatory costs. The willingness of decision makers to consider the interests of those who do monitor, however, further depends on the capacity of those monitors to translate their interests into political capital. The relevant group to which the expenditure is salient may be too small to have significant political effect. Assume that the proposed subsidy is to be paid by a relatively small number of residents for whom the personal stakes of defeating the proposal are sufficiently great to offset the costs of marshalling opposition. Members of the burdened group have incentives to identify and oppose the proposal, to emphasize competing priorities, and to serve as surrogates for others adversely affected by the proposal but who have neither the information nor the incentive to defend those interests. Under these circumstances, there is little reason to believe that decision makers will be left unaware of the proposal’s costs. But what happens at that point is uncertain. In some cases, those few residents who bear the immediate costs of a program will be able to represent others whose interests are too remote to warrant participation in the public debate. For instance, independent-bookstore owners within a municipality may object to a development grant to attract a Borders or a Barnes and Noble to the area, as may book lovers who fear the demise of locally owned boutique bookstores. While the latter may have insufficient concern to warrant vocal public opposition to the grant, they may take low-cost measures, such as joining a downtown rally or signing petitions, where those activities are subsidized by the independents. In such a case, even the small group that bears the immediate burden of the expenditure may be able to command legislative attention. At times, however, the concentrated group that suffers disproportionate costs from the regulation may have too few members and allies to translate their interests into political influence, notwithstanding their capacity to coalesce.35 If Scalia’s local landlords can organize but lack sufficient numbers to generate political support, then they may be excluded from any effective role in the political process. Small numbers will matter less if a group can form coalitions with prospective allies. But it is difficult to determine whether that “if” has been realized. To continue the landlord example, that group may have difficulty coalescing with prospective allies, such as anonymous prospective tenants who might be willing to

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 63

pay market rents and thus share landlords’ concerns about artificially depressed rents that induce existing tenants to remain in rent-stabilized apartments. Proponents of malign redistribution have incentives to impose the burdens of the redistributive program on just such a small, politically impotent group, while leaving the great majority of residents unaffected. The rosy picture of countervailing interests then founders on the shoals of indifference. The fact that only a small subset of residents will benefit from the program and only a (different) small subset will bear the redistributive costs belies, or at least renders more uncertain, any conclusion about the locality’s collective redistributive intent. It is difficult enough to infer some local preference for redistribution where the proposed ordinance affects a significant percentage of residents but affects each resident only minutely. Where regulations have so few implications for the vast majority of residents that those residents lack any incentive to contemplate the relative costs and benefits of the proposal, however, any inference about local preferences is more difficult to justify.36 One might contend that this very concern for imposing the costs of programs of questionable value on a group too small to offer an effective countervailing force in political debate is reflected in the constitutional incorporation of the Takings Clause: because eminent domain is likely to impose net costs on a very small number of landowners in order to confer benefits on a relatively large population, a just compensation clause is necessary to prevent public officials from running roughshod over the discrete minority, simply to gain the favor of the majority. It is perhaps this sympathy for the imposition of costs on a distinct minority that underlies much of the judicial invalidation of municipal programs that impose costs on groups that seem capable of organizing, but that arguably have insufficient membership to translate organization into influence. For instance, the common rejection of “impact fees” that require developers to pay the costs of municipal improvements that have only tangential relation to the developers’ projects implies a concern about majoritarian ganging up on developers, notwithstanding that few would describe real estate interests as suffering an incapacity to organize.37 One might object that those concentrated costs may be less problematic where benefits are similarly concentrated, since the size of the relevant groups would then suggest that they might have approximately equal

164

d is t in g u is h in g be n i gn a n d m a l i gn r e dis t rib ut io n

political power. If few firms face the mandate of a living wage ordinance, then, by hypothesis, few employees will enjoy the benefits. Moreover, one might suggest, since the immediate beneficiaries will necessarily be lowwage workers, a group that traditionally has little political influence, those firms that bear the cost would presumably be able to hold their own in the political forum against a relatively equally sized group of advocates for increased wages. Thus, the fact that unorganized, low-wage workers are able to prevail over the expressed interests of organized firms may signal a local commitment to redistribution. There are two difficulties with this rosy analysis, however. First, the fact that, organizational capacity aside, only a small subset of residents bears the redistributive costs belies, or at least renders more uncertain, the notion that the locality as a whole has some redistributive intent. It is one thing to infer some local desire to redistribute where residents are willing to bear the relevant costs, but arguably quite another where those costs fall almost exclusively on a subset of residents or nonresidents. Thus, one might reasonably distinguish between ordinances that impose living wage requirements on public entities, or on those who do business with public entities, and those ordinances that require high wage payments only by a subset of private employers. In the case of contracts that involve the city, that higher wage presumably will be reflected in the cost the city incurs when dealing with its employees or contractors. As a result, individual residents will bear the full cost of the ordinance, and its imposition will be somewhat salient, since the additional payments must come directly from the municipal budget. The local legislature presumably will internalize these costs of the mandate, at least to the extent that the local legislature ever internalizes the costs of direct expenditures. The imposition of the same obligation on private employers, however, entails a cost that is not internalized by the local legislature or most of the public. As the number and location of firms that bear the costs narrow, the likelihood that the process of cost internalization serves as an effective check on malign redistribution decreases. Thus, where, as in the cases of Santa Monica and Berkeley, the ordinance applies only to a limited geographical area within the city, an inference that the redistributive program represents local largesse becomes more problematic. The second problem with simply matching a small set of burdened residents with a small set of beneficiaries is that while group size matters,

d is t in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 65

its effects may be misleading. The relatively unorganized group that, of itself, appears to have little political influence may enjoy the support of proxies that can dominate the debate. Here, again, living wage ordinances are instructive. Organized labor has been a proponent of and active participant in local campaigns for living wage ordinances, either out of altruism or in an effort to increase the ranks of union membership. Its intervention, however, makes it difficult to disaggregate altruistic from selfish motives. I noted at the outset that more than a hundred localities have enacted living wage ordinances. But there are dramatic distinctions among these ordinances. It would be surprising if these differences were not influenced at least in part by drafters’ efforts to minimize opposition. Indeed, comparison of the different types of ordinances reflects how they can be structured in a manner that affects the incentives for different parties to become involved in the issue. Some ordinances apply solely to contracts that the city itself enters into and require that those who perform services for the city pay their employees a living wage. Presumably, that wage will be reflected in the price that contractors charge the city. As a result, city taxpayers generally will pay for the requirement through their taxes. The local legislature presumably will internalize these costs of the mandate, though the fact that they are spending constituents’ funds means that internalization is likely to be highly imperfect. While use of a broad-based tax and limited scope of the ordinance may mean that too few residents will bear a sufficient share of the added costs to make expression of any opposition worthwhile, use of a broad-based tax may also suggest that the ordinance reflects local preferences, as evidenced by residents’ willingness to pay for it. That conclusion may be even more appropriate because diffuse costs will not necessarily preclude organized and politically effective opposition to the proposed ordinance. Employers who expect to contract with the city may be concerned that they will not be able to pass the full additional costs to the city and thus will bear some of the costs of the ordinance themselves. If those employers have a sufficiently intense interest in the issue, then they would be expected to subsidize the costs of opposition among others in the community, perhaps by suggesting that higher taxes will be a consequence of the ordinance. Enactment of the ordinance under these circumstances, in which there is a significant likelihood that a politically influential group was able to participate in the

166

d ist in g u is h in g be n i gn a n d m a l i gn r e dis t rib ut io n

debate but the proposal still passed, likely signals that the result was consistent with local preferences. One might respond that the per capita cost that any resident incurs from a given city expenditure is so small that the distinction between costs imposed by the city on itself and costs imposed on private firms is irrelevant for purposes of inferring residents’ intent. Perhaps, but it is not clear which way that argument cuts. If the per capita costs are so small as to discourage residents from contemplating the relative costs and benefits at all, then there is less reason to believe that the ordinance reflects broad-based commitment to any position. Other living wage ordinances affect private employers but cover a sufficient number of employers to imply a greater likelihood that effective opposition will form. The robust debates among employers and other residents in Chicago, Santa Fe, and Santa Monica testify to the organizational capacity of local employers when the proposed ordinance affects a large enough number. Employers would appear well positioned to document the costs and thus to demonstrate that the proposed regulation is likely to impose net burdens on the locality or to contradict local preferences. The fact that they are outnumbered obviously cannot be a measure of political powerlessness sufficient to invalidate the ordinance as the result of a malign process. All that a group can ask is a realistic opportunity to be involved in the legislative bargaining process in a manner that permits effective consideration of its members’ concerns. That might mean that the substance of their concerns cannot be dismissed for reasons that are irrelevant to a socially acceptable determination of the issue, such as standard invidious bases of discrimination. Simply being in a distinct minority, however, does not qualify as such a basis for discrimination. After all, as Bruce Ackerman has noted, “minorities are supposed to lose in a democratic system—even when they want very much to win.”38 The difficulty, then, lies in separating out legitimate strategic structures that increase the likelihood of achieving local preferences from ganging up on a minority in order to allow another group to achieve its own naked preferences. If small size or a highly targeted imposition of regulatory mandates allowed some inference about malign redistribution, then, again, courts might be competent to play the role that I have argued is appropriate from the perspectives of both legal doctrine and institutional structure. Drawing such an inference, however, is more complicated than the simple metric of size and targeting suggest. Take, for example, the debate

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 67

about the proposed Santa Fe living wage ordinance. The fact that employers were able to procure an amendment that applied the proposed ordinance to businesses with more than twenty-five employees, rather than the original ten employees, might be thought to demonstrate that even those small-business employers who had the least political influence among the group were able to influence the deliberations. That suggests that the living wage ordinance reflected a well-working political process, notwithstanding all the hesitation that we might have about the risks of using regulatory mandates rather than direct taxation to accomplish redistributive objectives. But that same fact could be interpreted in a more negative way. By increasing the threshold number of employees necessary to trigger the ordinance, the amendment reduced the number of employers who had a direct interest in the outcome of the debate and thus arguably effectively reduced the incentives of potentially influential opponents to participate in the debate. The subsequent litigation supported by opponents may augur in favor of the first interpretation. If the opponents who remained active after the amendment were able to initiate a litigation campaign to overturn the ordinance once it was enacted, one might infer that they also had the resources to mount a political campaign against its initial passage. Nevertheless, resources aside, the reduced number of covered employees does raise the specter that the remaining opponents simply constituted too small of a group to exercise any political power in the face of what were effectively dominant proponents. This concern similarly emerges from other living wage ordinances that appear to have been drafted with a narrow focus, perhaps to minimize the number, if not the intensity, of active opponents. The contested Hudson County, New Jersey, ordinance required that higher wages be paid only to employees of contractors that supply food service workers, janitorial workers, and unarmed security guards to the county.39 An amendment to the ordinance in Berkeley, California, extended coverage to a limited number of employers that satisfied requirements relating either to number of employees and revenues or to location in certain areas of the city.40 Recall that the controversial Chicago living wage ordinance applied to a mere handful of employers—big-box stores that occupy more than ninety thousand square feet and make more than $1 billion in annual gross revenue.41 It did not apply to all employers, to all chain stores, or to stores with a certain number of employees. Rather, the

168

dist in g u is h in g be n i g n a n d m a l i g n r e d is t rib ut io n

criteria included square footage and the annual gross income of the parent company. Indeed, the targeting was sufficiently tailored that one might suspect that if the ordinance had ultimately gone into force, it would have faced a “special legislation” challenge. It is less clear that these ordinances reflect local preferences in a manner that ensures their benign origins. Opponents who pay the direct regulatory costs may be so small in number that they are swamped by the numbers of advocates. Residents who are not affected by the mandate, either because they do not employ lowwage workers or because they do not contract with the covered employers, may be indifferent as to enactment of the ordinance. Some residents who favor redistribution generally may actively support mandates that impose redistributive expenses borne solely by others. That does not mean, of course, that they would be willing to support the ordinance if they were required to internalize its costs. Given the difficulty of distinguishing the needs of low-wage workers in the covered industries from the needs of those in exempt industries, the narrow range of the mandate’s application begins to look like a strategic effort to minimize opposition by concentrating on a group that will have difficulty exercising significant influence. Of course, it may be hard to make that case where, as in Chicago, the narrow group of covered employers includes international corporations with significant political and financial resources, such as Wal-Mart—especially since Wal-Mart won the war after losing the battle. But Wal-Mart’s failure to participate in the explicit public debate about the issue suggests only that private interests were operating at multiple levels, not that the initial ordinance or the mayor’s veto of it reflected local preferences. Finally, the structure of some living wage ordinances suggests even more strategic efforts to structure the proposal in a manner that distorts debate away from ideal interest group competition. This could be done, for instance, by identifying potential opponents and neutralizing them by providing benefits not shared with other subsidizers who are less capable of organized resistance. For example, recall that the living wage ordinances enacted in Berkeley and initially in Santa Monica contained an exemption for those employees who would otherwise be eligible but who were subject to a collective bargaining agreement that explicitly waived the provisions of the ordinance.42 It is, of course, plausible that the presence of such an exemption reflects a local policy to promote collective bargaining and to ensure that labor has sufficient tools to negotiate effectively with firms.

d ist in g u is h i n g be n i gn a n d m a l i gn r e dis t rib ut io n

1 69

But note the direct effects of the exemption in light of the role that unions played in serving as surrogates for the working poor who were directly covered by the ordinance. The exemption provides unions with a powerful tool for reducing opposition to union organizing. An employer that enters a collective bargaining arrangement enjoys the prospect of paying employees less than would be mandated under the living wage ordinance, although more than the employer would be paying in the absence of a union agreement. The degree of the incentive depends on the disparity between the market wage and the mandated living wage. Assume, for instance, that the market wage for unskilled workers covered by the ordinance is $7.00 per hour. Assume that the living wage ordinance requires a minimum wage of $10.00 per hour. An employer can negotiate a collective bargaining agreement for the covered employees that contains a wage rate of between $7.00 per hour and $10.00 per hour and reduce its costs below those that would be required by the ordinance. The employer is therefore less likely to oppose the union-organizing efforts in its workplace. The prospective benefit to the union of increased membership encourages unions to support living wage ordinances, regardless of whether employees would receive the mandated wage. The result, however, is that what initially looks like a contest between a small number of low-wage beneficiaries and relatively wealthy employers in which the latter may have an organizational edge might actually become one in which the employers are dominated by the more numerous and highly organized proxies of the ordinance’s immediate beneficiaries. Again, the presence of the dominant group does not guarantee a malign outcome. It does, however, complicate efforts to explain the redistributive outcome as an expression of majoritarian municipal preference. Moreover, proponents of benign redistribution might prefer to cabin the costs for administrative reasons rather than to quash political debate. Deliberation may be a social benefit, but the group that assembles even the politically popular and desirable proposal might simply want to get the deed done and treat robust debate as a cost. Targeting may stem from recognition that the targeted group is receiving a unique benefit or imposing a unique burden on the public that warrants imposition of disproportionate cost. A mandate that only those employers situated in particular locations of a locality be subject to a living wage ordinance may be defended on the grounds that the targeted area is sufficiently popular that the employer can earn supracompetitive profits and should be required

170

d is t in g u is h in g be n i gn a n d m a l i gn r e d i s t rib ut io n

to share that benefit with its employees. That rationale assumes that the extraordinary value of the employer’s location has not been factored into the prices it paid to own or rent the land, but if that assumption is true, then there may a logical basis for distinguishing among employers in a locality. A tax on overweight vehicles that was designed to compensate for the extra wear and tear they impose on municipal streets might fall on a small group that could not easily coalesce or exercise much political influence, but it could not be said to impose on them any burden disproportionate to the costs they are imposing on the majority of residents. EXTERNALIZED COSTS

In true beggar-thy-neighbor fashion, proponents of malign redistributive legislation are also likely to structure proposals in a manner that imposes costs on nonresidents to whom local legislators are unaccountable. Local legislators, and their constituents, are likely to be less attentive to, or even to ignore, spillovers of local policies; proponents of malign redistribution, therefore, have incentives to design programs in a manner that imposes the related costs on nonresidents who have limited access to the local political process. From exclusionary zoning, which imposes the costs of poverty on less affluent areas, to the siting of locally undesirable land uses at the boundaries of localities, the concern that local governments will seek to capture the benefits of local programs while externalizing costs is frequently invoked to advocate policy making at more-centralized levels of government.43 If we are skeptical of local ordinances that externalize costs, then perhaps exactions that are disproportionately borne by nonresidents can serve as a presumptive proxy for malign redistribution. Facially neutral local taxes on hotel rooms, parking garages, or access to entertainment fall into this category, as they are likely to fall largely on tourists or other transients to whom the individual costs are too small to warrant monitoring or the registering of complaints. But explicitly redistributive programs may share the same characteristics. The proposed living wage ordinance in Santa Monica, for example, would have affected almost exclusively hotels and restaurants in an area populated by tourists.44 Although the proposed ordinance applied directly to resident businesses, one would imagine that those businesses would have attempted to pass on as much as possible of the increased cost of doing business to nonresident patrons. Indeed, part of the debate about the propriety of the ordinance revolved

d ist in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 71

around the issue of whether hotel operators would be constrained by competition from non-Santa Monica resorts from passing costs on to patrons. Similarly, the Santa Fe living wage ordinance excludes private employers with fewer than twenty-five employees.45 In a city as dominated by tourism as Santa Fe, it is plausible that many of the businesses subject to the ordinance are hotels, restaurants, and museums that cater to nonresidents. But before presuming that any exaction that disproportionately burdens nonresidents smacks of a defective political process or is likely to be malign, it is useful to consider perfectly legitimate reasons to impose taxes or fees on nonresidents. The tourists and commuters on whom these exactions typically fall, after all, consume or benefit from municipal services such as police and fire protection but do not help defray their costs through the standard system of local property taxes. Thus, to the extent that the exactions reflect the pro rata costs of the municipal services from which nonresidents benefit, their imposition does not reflect an effort to exploit those without political voice. But this argument for imposing exactions on nonresidents also defines the limit of their appropriate use. Exactions imposed on nonresidents may be more suspicious where they exceed the per capita value of the services that the payers utilize.46 One might contend that this limitation provides courts with a basis for intervening on a more objective basis. Courts arguably could review the costs of supplying municipal services to nonresidents and determine whether the exactions roughly approximate the per capita expenditure. Exactions that are excessive could be perceived as indicative of malign distribution. Indeed, courts play this function already, insofar as they inquire into whether a particular exaction constitutes a permissible user fee, the imposition of which is intended to defray the expenses of providing a service to a payer, or an impermissible tax, the imposition of which need not be linked to the costs or opportunity costs of the service to which it is attached.47 Similarly, courts traditionally consider whether fees imposed on nonresidents for the municipal provision of extraterritorial services are appropriate in light of the costs related to providing the service. The same type of inquiry could be made, for instance, to determine whether an exaction explicitly imposed on tourists, transients, or commuters was sufficiently related to the costs of the services provided. But the literature on the issue of whether nonresidents contribute their fair share of municipal expenses suggests that a simple comparison

172

dist in g u is h in g be n i gn a n d m a l i gn r e d i s t rib ut io n

of costs and payments is more elusive than the recommendation for judicial intervention implies.48 Nonresidents and tourists who do not pay for the services they utilize through property tax payments may warrant some credit for increasing the economic activity of the locality that offsets the burdens they impose on municipal services. For instance, additional tourism boosts employment and tax revenues, even if the tourists do not make explicit payments for services. If the local financial benefits of nonresident activity offset the additional costs of serving tourists and commuters, then the economic case for additional exactions collapses. In theory, the judiciary could assess the evidence on whether nonresidents consume more municipal services than their economic activity within the locality supports. But given that the judiciary will likely have limited capacity to evaluate the available evidence—an enterprise that bedevils even professional economists who attempt to resolve the debate—it is doubtful that courts would be able to arrive at a reliable calculation about whether exactions were exploitative or whether they simply recovered the costs of providing public services to nonresidents. Judicial inquiry into exactions imposed on nonresidents might be less appropriate if there existed alternative constraints on municipal exploitation of outsiders. Interlocal competition for mobile capital and labor may do much of this work. To the extent that there is interlocal competition for mobile capital and labor, local officials have incentives to consider the consequences of local policies on the ability to attract and retain firms and workers who serve the city’s economic development. Mobile capital that must increase wages to compensate for higher taxes on commuting employees has incentives to exit the city in order to reduce those costs. Employees who must pay the taxes, because employers have passed them on, have incentives to exit for employers in more hospitable jurisdictions. Indeed the subsequent constraint on cities may lead them to be conservative in their estimates of the positive effects of exacting payments from nonresidents. Commuters may recognize that taxes reflect the value of the services that they receive from the locality, but they may still object to paying that value because they believe that they can get similar resources for “free” from jurisdictions more willing to subsidize nonresident services. Recall that, as a result, some observers suggest that interlocal competition will cause a race to the bottom in which localities impose taxes on mobile capital that are, from a social perspective, too low.49 The

d is t in g u is h i n g be n i g n a n d m a l i g n r e dis t rib ut io n

1 73

exit threat might seem limited by any unique benefits or agglomeration economies that localities can exploit. But even with respect to those benefits, nonresidents do not necessarily lack the kind of representation that bespeaks a flawed political process or malign redistribution from nonresidents to residents. Residents whose livelihood depends on nonresidents to shop or work in the locality serve as surrogates for those who are nominally disenfranchised. Resident restaurant or garage owners, for instance, may depend sufficiently on commuter or tourist patronage that they are able to represent the interests of nonresidents. Paris, for example, might be thought to have a situational monopoly that allows it to exploit tourists. But restaurateurs led the recent successful battle to reduce the tax on meals from 19.6 percent to 5.5 percent.50 In other cases, the incidence of the redistributive exaction is more complicated. As discussed above, the burden of some living wage ordinances initially falls on employers who depend on nonresident, tourist patronage. Note that in the battle over the Chicago big-box ordinance, the local suppliers and businesses that would profit from Wal-Mart’s presence argued on behalf of the national chains. As long as these businesses believe that they cannot fully pass the higher wage costs on to nonresidents, they have incentives to oppose the redistributive program and thus to serve as surrogates for those who have no direct access to the decision-making process. Where surrogates effectively represent the interests of the disenfranchised, nonresidency alone provides little reason for inferring that exactions directed at nonresidents necessarily signal malign redistribution.

ch ap t e r s e v e n

The Limits of Judicial Intervention

let us assume that there are cases in which courts reliably conclude that the political process by which a locality enacted a redistributive program was sufficiently dominated by a particular interest group to trigger a presumption of malign consequences. I have suggested to this point that courts might consider that factor when the program is challenged for violating some constraint on local autonomy. In theory, such interventions could reduce distortions of residents’ preferences and thus enhance the values inherent in decentralized decision making. Are there still reasons why we would not want courts to take such motives into account when construing the permissibility of local redistribution? In this chapter I examine three reasons for a cautious approach to judicial intervention, even where we are confident that courts can identify malign local redistributive programs. I have alluded earlier to some reasons that are frequently cited for inviting courts to protect the political process against interest group dominance. As most thoroughly illuminated by Einer Elhauge, courts suffer from their own biases and susceptibilities to interest group pressures, so that allowing them to correct perceived process failures might simply displace one set of distortions from municipal interests with another. These distortions might be greater where judges are elected and thus susceptible to lobbying by potential supporters, though empirical evidence on the political vulnerability of courts is inconclusive.1 1 74

th e l i m i ts o f j ud i c i a l i nt erv ent io n

1 75

Even apart from electoral influences, however, the same organized interest groups that effectively lobby legislators can also sponsor litigation campaigns, while free riding may undermine the willingness of less cohesive groups to subsidize court challenges. Dominant interest groups may matter less for litigation, however, because even a relatively small number of people with an intense interest can initiate judicial process. A small business owner who faces atypical competition from a potential beneficiary of a municipal tax subsidy would likely have little capacity to wage an intensive lobbying campaign that others have no incentive to join. But that same business owner could trigger the full panoply of judicial remedies by individually challenging the subsidy.2 Organized interest groups, however, might have a qualitative advantage. They might be able to afford better legal representation and thus increase their chances of success in litigation. In this chapter, however, I make the conservative assumption that judges are benign and impervious to interest group entreaties. I assume that courts seek to enforce the law and interpret it in a manner consistent with the intentions of legislators or constitutional drafters. Even under these circumstances, I suggest, there are reasons to be skeptical of judicial intervention into local redistributive decisions. First, I expand on an argument to which I have alluded previously concerning the relative ability of interest groups to coalesce at state and local levels. Judicial intervention of the type I have suggested takes the form of narrowly construing constitutional or statutory authority for local action where local decision makers are vulnerable to special pleading. Frequently when courts deny such local authority, they do not conclude that the programs are beyond the competence of government. Rather, they conclude that local governments have not received the appropriate dispensation from the state to implement the challenged redistributive program. Groups that prefer the programs can seek approval from more-centralized levels of government. Thus, judicial limitations on local autonomy are appropriate only if we believe that malign redistribution is more likely to occur at the local level than at a more centralized level. Otherwise, we are simply increasing transactions costs by requiring groups to seek permission both from state and from local levels of government without affecting the end result. In contrast, if decision making at the state level is less vulnerable to interest group dominance, or if simply increasing transactions costs reduces the probability that dominant groups will obtain authority for localities to act

176

t h e limit s o f j ud i c i a l i n te r v e n ti o n

malignly, then a requirement of state approval could reduce local deviations from constituents’ preferences or interests. Second, while negating the authority of a locality to implement a redistributive program may initially look like a desirable outcome, results may be more questionable when placed in a more global context. Institutions that are competent to make decisions about allocation of a particular resource at a micro level may have less capacity to consider the macro implications of their decisions. Courts exemplify the problem. Courts do a pretty good job of resolving discrete disputes by applying principles to limited facts that can be presented in an adversarial proceeding. Courts are less proficient in deciding how the discrete issue before them in any given case fits into a more global analysis of related issues. A court, for instance, may be capable of analyzing the facts of an individual tort case and applying established rules to those facts. Devoid of independent fact-finding ability or the capacity to achieve compromises among multiple constituencies, that same court would have more difficulty assessing whether those rules provide a more efficient or equitable resolution for all tort cases than, for instance, an administrative compensation scheme. Given the narrow institutional competence of judicial decisions, we do not ask courts to address the broader implications of their decisions. If judicially established rules generate adverse social effects, we typically change the rules by legislative or administrative procedures, since these institutions have broader capacity to initiate proceedings, engage in independent fact-finding, and balance competing criteria rather than engage in winner-take-all decision making. In the current context, the implications of discrete decisions flow from the inability of courts to address the role of interest groups in situations that do not give rise to legal cases. As a result, even if courts can discern that interest group influence has skewed local decision making about a particular redistributive program, courts will be unable to calculate the effect that the challenged program has on the full range of redistributive programs. The scope of judicial consideration matters because redistributive programs come in many forms, some of which are highly susceptible to judicial review, and some of which are less so. If programs that are susceptible to review tend to favor one segment of society, while programs that are immune from judicial review tend to benefit another, then judicial

th e l i m i ts o f j ud i c i a l int erv ent io n

1 77

invalidation of programs within the former group may distort the overall mix of redistribution in the society. Third, and related to the second point, judicial intervention has the potential to exacerbate rather than reduce the distorting effects of local redistributive programs. Courts can address malign influences only in redistributive programs that the local legislature enacts and that are subsequently challenged by opponents. Interest groups, however, may exercise influence both to enact programs that they support and to block enactment of programs that they oppose. In the latter case, where the result of interest group intervention is legislative inaction, the constitutional doctrines on which courts may rely to negate legislative action are not triggered. Thus no basis exists for judicial review. Hence, if one group is favored by the capacity to block legislation and has an equal opportunity with other groups to have enacted legislation that it endorses, judicial invalidation of some programs may skew the total mix of redistribution. I next turn to more thorough consideration of these issues. L O C A L A N D S TAT E S U S C E P T I B I L I T Y T O I N T E R E S T G R O U P P O L I T I C S

I have suggested that, in theory, courts could interpret the scope of local autonomy narrowly in order to invalidate malign local redistributive programs when they believed that the local political process was insufficiently representative of multiple perspectives on the problem to produce a program consistent with municipal preferences. The effect of doing so, however, is not necessarily to preclude enactment of the redistributive program. Rather, judicial determinations that localities lack authority to enact a program may induce those who favor the redistributive program to seek explicit authority from the state. That is because judicial denials of local authority do not necessarily prohibit government involvement in the invalidated project. Instead, they require more explicit state approval than exists for local involvement. In effect, judicial decisions about the scope of local authority allocate decision-making responsibility between state and local governments. While state legislatures will still be unable to authorize programs that run afoul of constitutional constraints on the state itself, such as public purpose or special legislation restrictions, explicit legislative grants would at least remove obstacles to local redistribution predicated on the absence of local authority, a narrow scope of home rule, or state preemption.

178

t h e limit s o f j ud i c i a l i n te r v e n ti o n

One could imagine that judicial intervention could exacerbate malign consequences at the state level in either of two situations. First, it is plausible that a court would properly recognize that an interest group dominates at the local level and thus would invalidate the local ordinance as outside the scope of local authority. This could leave the defeated local interest group to seek explicit permission for the program from the state legislature, thus reversing the judicial decree. This might initially seem anomalous, as one would expect that the interest group would have begun with the state legislature if it held a dominant position there, as it presumably would have to possess in order to obtain state approval. But the fact that the group began its efforts at the local level might indicate only that it thought it had a sufficiently high probability of success to initiate the effort at that level rather than to incur the higher organizational costs that the group might encounter at the state level, notwithstanding an even higher probability of success at that level. Alternatively, a group that was competitive at the local level and that unsuccessfully opposed the local redistributive project might seek to have it judicially invalidated. Under these circumstances, a court that perceived the local process as competitive would uphold the result as benign. If the opposition group was dominant at the state level, however, then it might subsequently prevail on the state legislature explicitly to deny localities the authority to engage in such programs. One could interpret explicit legislative prohibitions on local rent control ordinances as reflecting that scenario, though I will argue below that other interpretations are also plausible. If a group is competitive at the local level but dominant at the state level, then judicial invalidation of the local program at best only increases the costs of obtaining the malign result (by inducing the dominant group to use its political power at the state level) and at worst exacerbates matters, because state legislative action means that the prohibition on local activity would extend throughout the state, notwithstanding the contrary preferences of some localities. Of course, it is plausible that the state legislature could also serve as a corrective to judicial misjudgments or misapplications of presumptions. For instance, if a court invalidated a local living wage ordinance on the doctrinal grounds that it was preempted by a state minimum wage law, but motivated by a perception that proponents dominated at the local level, those who favored the ordinance might seek an explicit legislative statement that the

th e l i m i ts o f j ud i c i a l i nt erv ent io n

1 79

state minimum was intended only as a floor, not as a ceiling. If the court was mistaken in its belief about the dominated political process at the local level, and if the state legislature similarly provided a forum for countervailing groups, then one would anticipate that the state legislation that emerged would compensate for judicial error in evaluating local political processes. In all of the above events, the state legislature retains the capacity to reverse judicial findings of insufficient local authority. Thus, judicial invalidation of perceived malign local redistribution on authority grounds makes sense only if more-centralized levels of government are less likely than local governments to engage in malign redistribution. When courts determine whether to allocate decision-making authority to a state or local government, therefore, it is plausible that they could consider which jurisdiction is more likely to be susceptible to participation by multiple and competing interest groups and thus more likely to increase the probability that benign redistribution will emerge. Initially, this might appear to be a false issue. Standard theories of federalism suggest that local governments are more susceptible than morecentralized entities to dominant interest groups. Centralization typically increases the preference heterogeneity of the population; that is, a state is likely to contain a broader range of interests than any city within it. As a result, groups that may have been dominant in a decentralized area are more likely to face groups that have countervailing preferences at the more centralized level of government. Madison’s argument for strong centralized governments in Federalist #10 evolves from just this preference dilution effect and what he perceived as the relative absence of countervailing interest groups—factions—at decentralized levels.3 Indeed, although Madison was speaking of the relative vulnerability of the federal government and states to factions, one might think that his point has even more force when applied to the relative vulnerability of states and cities, given the greater homogeneity of preferences within localities. Thus, one might conclude that a court that desired to allocate decision-making responsibility to the jurisdiction most susceptible to robust debate among countervailing groups would systematically define local autonomy narrowly. But the likelihood of capture is only indirectly related to the preference dilution effects that are characteristic of centralized governments. Although there is little empirical work on the relative vulnerability of different levels of government to capture, some economists have created models

180

t he limit s o f j ud i c i a l i n te r v e n ti o n

that demonstrate, under certain assumptions that are not overly heroic, that local governments can be less prone to capture than more-centralized governments. Pranab Bardhan and Dilip Mookherjee find that susceptibility to capture at different levels of government will vary with factors such as the extent of electoral competition, electoral uncertainty, the value of campaign contributions, intradistrict inequality, and the type of electoral system.4 Michela Redoano presents a model under which the level of centralization affects the level of lobbying, but the effect is ambiguous. That is, there are conditions under which lobbying is more pervasive under centralization and conditions under which lobbying is more pervasive when decentralized governments make policy.5 For instance, if centralization increases the size of the affected group, then lobbying should increase at the centralized level, if lobbying costs are fixed and equally shared among members. Some commentators have drawn support for the relative vulnerability of centralized governments to lobbying from empirical work that demonstrates a negative correlation between decentralization and corruption.6 Consistent with these models, my argument for ambiguity about the relative vulnerability of local governments to dominant lobbies is predicated on the relationship between the conditions for collective action and the organizational structures of state and local governments. Capture will be a function of the capacity of some groups to organize while potential groups with countervailing interests do not. While heterogeneity of preferences decreases the likelihood of capture, it is only one characteristic that affects the capacity of a group to coalesce internally and, once organized, to lobby decision makers in a manner that influences policy. Within the group, the costs of identifying, communicating with, and obtaining contributions from like-minded individuals will determine the capacity to organize. Once organized, the group must have the capacity to attain access to and influence within the decision-making forum. For example, groups that are repeat players before a legislature may be willing to invest more in creating a relationship with representatives than groups that cannot amortize that investment over a series of projects. Developers who expect to have multiple dealings with a city council over issues such as zoning, tax abatements, and building codes are more likely to invest in relationships with council members than any landowner who is adversely affected by one of the developer’s projects. The variety of characteristics that affect both internal organization and access to decision making introduces more uncertainty into

th e l i m i ts o f j ud i c i a l i nt erv ent io n

181

the question of which level of government is more susceptible to capture. For some groups, the conditions for collective action will be satisfied more readily at the local level than at more-centralized levels of government. For other groups, the opposite will be true. The result is that the relative vulnerability of decentralized governments to capture is less obvious than the Madisonian preference for centralized government admits. Think, for example, of the relative costs of attaining access to legislators, and hence of asserting influence, at the state level. The fact that elections for state office typically require campaigning over a larger geographic area, and the fact that they tend to be more partisan than elections for local office, translate into a greater need in statewide elections for substantial campaign contributions. Groups that perceive electoral support as a prerequisite to influence, therefore, will have to provide greater support for state than for local legislators. Once in office, state legislators are more costly to lobby. Even though the state legislature may meet at a centralized location, the transportation costs of organizing at the state capitol may be within the reach of a smaller number of groups. If key representatives have to be lobbied when the legislature is not in session, the costs of lobbying them across the state will be higher. State legislative processes are also more likely than local ones to be characterized by bicameral legislatures and multiple committee hearings, each of which increases lobbying costs. Because concurrent majorities of both houses are necessary to pass legislation, bicameral legislatures privilege the status quo by reducing the number of votes necessary to block new enactments. A majority in the smaller sized chamber can effectively defeat a redistributive proposal favored by a majority of the total of the two chambers. Thus, higher expenditures will be necessary to ensure that the minimum number of opponents is not reached. Committees exercise monopoly power over the agenda that falls within their jurisdiction. Committee members, therefore, can demand resources (time, effort, money) from proponents of legislation in order to advance a particular proposal to the full chamber for a vote. “Demand” in this sense is not necessarily equivalent to a bribe. Even the most publicly interested committee member must be persuaded that one piece of legislation rather than another should pass through the gate that the committee maintains, so that proponents of legislation will have to devote resources to making their proposals salient.

182

t he limit s o f j ud i c i a l i n te r v e n ti o n

Finally, groups that seek to facilitate local redistribution will not necessarily be able to persuade the state to mandate that localities implement a program. A variety of programs may be appropriate for only a few localities within the state, and representatives from jurisdictions that incur few benefits from the proposal, or for which the proposal would be inappropriate, are likely to balk at a state requirement that imposes obligations on their constituents. They may be indifferent to a state law that permits localities the option of adopting the program, since their own districts can ignore the legislation even if other districts exercise the option. But legislation that permits the exercise of local autonomy without requiring local action also raises costs because groups that successfully obtain passage of enabling legislation at the state level will also have to persuade local decision makers to implement the programs that the state has authorized. Groups that sponsor legislation, therefore, are likely to prefer state-imposed mandates over permissible legislation. One might think that the higher costs of obtaining influence at the state level supports the Madisonian preference for centralized decision making because higher costs mean that fewer interest groups will be able to subsidize statewide coalition and lobbying. Constraints on dominance, however, are not a simple function of a small number of interest groups; rather the constraint arises from the presence of countervailing groups. High access costs may reduce the number of groups that are able to coalesce at the state level. But the necessary implication is that any group that can successfully organize is less likely to be met by effective opposition. Groups that have the means to navigate through the maze of state legislative processes, therefore, may be more likely to dominate the debate in the centralized government than at the local level. If wealthy, privileged groups have the capacity to organize both at centralized and at decentralized levels, while poorer groups can organize only at the local level, then one would expect interest group competition at the local level, but not at the state. Fewer interest groups might survive the more costly process of lobbying at the state level, but those that do can survive with a vengeance. If that is the case, then the state legislature may be more inclined explicitly to authorize malign redistributive programs advocated by dominant, wealthier groups that are likely to have unopposed access at the state level. Return, for example, to the debates about rent control that I discussed above. Landlords of multifamily housing units of the type that are subject

th e l i m i ts o f j ud i c i a l i nt erv ent io n

183

to rent control are likely to be able to coalesce relatively easily at either the state or the local level. They are typically repeat players before the local legislature because they deal with multiple legislative issues such as zoning ordinances, variances, building permits, and tax abatements. Thus, should landlords desire to gain access to address the additional issue of rent control, they face relatively low costs given that the decision makers are those with whom the landlords have previously established relationships. Rent control is also likely to impose significant income effects on them, so they have incentives to organize in opposition to it. Tenants who would benefit from rent control are likely to be transient and relatively poor. Nevertheless, if an entrepreneur (a church, an organizer for a tenants’ union) sponsors organization, tenants will enjoy relatively low costs of appearing before the local legislature and have incentives to do so because rent control confers significant benefits on them. Thus, one would expect that both sides of the rent control ordinance will be heard at the local level. At the more centralized level of government, however, one would expect a very different mix. Landlords are likely to have incentives to block rent control at the state level because they may own properties in multiple jurisdictions, and thus they will benefit from statewide prohibitions that preclude the need to battle the issue on a locality-by-locality basis. Landlords will also be able to identify each other across jurisdictions at relatively low cost. Tenants, however, may have more difficulty identifying each other across jurisdictional boundaries and coalescing at the state level, if for no other reason than the prohibitive cost of mounting statewide campaigns for tenants who are, by definition, of relatively low income. As a result, contrary to Madison’s instincts, we would expect to see landlords dominate rent control debates at the state level but to see more interest group competition at the local level. Indeed, consistent with that prediction, rent control ordinances tend to arise at the local level. Of course, different local circumstances might mean that local decision making is appropriate because some localities might need rent control while others do not. But if that fully explained the scope of government intervention into residential rental markets, we would expect to see no state legislation on the issue. Instead, we do see state intervention. That intervention, however, typically takes the form of statewide prohibitions on local rent control, a conclusion consistent with the relatively dominant presence of landlords over tenants at the state level.7

184

t h e limit s o f j ud i c i a l i n te r v e n ti o n

There are, however, some competing characteristics that suggest that dominant groups seeking malign redistribution are more likely to prevail at the local level, while state legislatures would more likely be the focal point of interest group competition. Because party competition tends to be less prevalent at the local level, local autocrats may be more solicitous of narrow interests without triggering criticism from political opponents. Some grassroots movements may be able to coalesce at the local level, while countervailing groups that have significant national or state organizational advantages would have difficulty funding opposition to initiatives that arose simultaneously in multiple localities. The opponents might tactically concede the local battle and seek to overturn the results either in the state legislature, where they may dominate local grassroots efforts that don’t translate into effective statewide organizations, or in litigation, in which they may have a funding advantage. An example here is a recent spate of ordinances that limit the terms of predatory loans, and which have been challenged in litigation underwritten by national organizations for the lending institutions.8 Finally, the availability of gubernatorial vetoes and the capacity of a robust state bureaucracy to frustrate implementation of malign redistribution may dilute the incentives of groups to seek legislation that they might otherwise advocate. It is, therefore, difficult to generalize about whether local or state legislatures are systematically more susceptible to lobbying by dominant interest groups. Instead, different groups may have an organizational advantage at different levels of government on different issues, so that relative vulnerability to capture depends on which groups a particular issue attracts. As a result, it is difficult to conclude that a judicial decision that invalidates local legislation is likely to “stick” or be legislatively reversed, or that any such reversal will be more or less publicly interested than the invalidated local program. While this agnosticism may be appropriate for legislation generally, perhaps more can be said about explicitly redistributive legislation. If we assume that redistribution will most likely have the support of the beneficiaries of the proposed program, then the characteristics that I have attributed to state and local decision making may mean that there is a difference in the forum most vulnerable to proposed redistribution in favor of the wealthy and the poor. Groups that favor redistribution to the relatively wealthy are more likely to have the resources necessary to

th e l i m i ts o f j ud i c i a l int erv ent io n

185

overcome the structural impediments to influence that exist at the state level (committees, bicameralism, coextensive branches of government). Groups that assist the poor may have greater difficulty organizing and purchasing access and influence at more-centralized levels of government simply because the combination of transportation costs and the multiple barriers to enactment require resources beyond their means. I noted above that the relatively poor may be sponsored by other groups, such as labor unions or church-based groups, whose interests coincide with those of the poor. Those groups may have sufficient funds to overcome the obstacles presented by more-centralized governments. But because these groups, by definition, represent a broader constituency, it is less clear that they would dedicate the resources necessary to overcome those obstacles to efforts for which their members are only indirect beneficiaries. While these groups may be willing to invest some resources in programs that will assist the poor at the local level (especially if the union members are themselves highly localized), they may be reluctant to use the significant capital necessary at the state level to support the same effort. The somewhat perverse result is that judicial invalidation of local redistributive efforts to assist the poor may be more likely to “stick” than invalidation of local redistributive efforts to assist the wealthy, since the latter will more readily be able to substitute state decision makers, who are not susceptible to the same grounds for invalidation, for local ones. States, for instance, may be more likely to override judicial invalidation of local subsidies for economic development than judicial invalidation of living wage ordinances. The indeterminacy of whether states or localities are more susceptible to dominant interest groups provides an opportunity for courts to evaluate the local political process in individual cases. But the same phenomenon might make us suspicious of the competence of courts to perform the necessary ad hoc analysis. Essentially, it requires courts to reverse engineer the process by which the locality acted, to determine whether countervailing groups exercised sufficient influence to legitimize the process, and—at least in some cases—to determine whether the state legislature would be a superior forum for accommodating countervailing interests. The problem that courts face in resolving these issues is reminiscent of the discussion concerning the history of constitutional special legislation prohibitions in chapter 5. There I suggested that there were two plausible explanations for the state legislative reversals of local

186

the limit s o f j ud i c i a l i n te r v e n ti o n

policies that preceded the adoption of the constitutional clauses. One explanation was that groups that represented only minority interests at the local level had an organizational advantage at the state level, so that the state’s action distorted municipal preferences. The alternative was that the decision at the local level was itself enacted through influence of a dominant interest group and was countermanded by majority local interests that were better able to organize at the state level. Groups that lose the interest group competition at the local level, but that dominate at the state level, can always enact their agenda simply because policies set at the state level consistently prevail over conflicting local policies, even when the locality is exercising constitutional home rule authority.9 But that does not demonstrate that the state decision is more consistent with either state or local interests; it demonstrates only that legal doctrine privileges states that have plenary authority over their political subdivisions. Einer Elhauge illustrates the difficulty of discerning which explanation accounts for state reversal of local policies by positing that the state could enact a statute repealing city rent control ordinances.10 Consistent with what I suggested above, Elhauge contends that landlords might be successful in securing repeal of the ordinances because landlords are more organized (at the state level) than tenants. But, he argues, that success may represent capture by the landlords, or reversal by the landlords of a local decision initially obtained by more numerous tenants. For Elhauge, this conundrum represents the poverty of interest group analysis predicated on reversing decisions made by those who exert “disproportionate” influence. He thus suggests that whether either group exercised such influence depends on one’s conception of “disproportionate” influence, and that inquiry necessarily entails a normative baseline. Thus, if one’s normative baseline is economic efficiency, then the identification of the group that exercised disproportionate influence depends on one’s view of the efficiency of rent control. If rent control is efficient, then the landlords exercised disproportionate influence. If not, the landlords did not. But one need not attempt to address the problem of disproportionate influence by reference to outcomes based on any normative premise, other than a premise about optimal decision making. That is, if the objective is to ensure municipal autonomy that actually reflects municipal interests, then substantive outcomes matter less than creating a decision-making structure that is most likely to express those interests, including whatever

th e l i m i ts o f j ud i c i a l int erv ent io n

187

normative principles the municipality cares to adopt (subject, of course, to the externalities that restrict all local action). From that perspective, the problem is to determine, for any given issue, which level of government is most likely to provide a structure in which countervailing forces exist so that a more benign decision can emerge. Judicial invalidation of local redistribution makes sense only where more-centralized levels of government are less susceptible than their localities to dominant interest groups that have the capacity to undermine judicial determinations. Elhauge, however, is no more enamored of this possibility than he is of efforts to define disproportionate influence. He acknowledges the possibility that the normative baseline that he finds inherent in the decisionmaking process “need not turn on the substance of outcomes. Imagine, for example, a baseline that accepts any political outcome that would have been reached under circumstances where each person had full information and equal political influence. Such a baseline might seem purely process based.” That seems pretty much to describe the process that I have recommended. The underlying assumption of interlocal competition is that different localities will and should provide different bundles of goods and services. Potential residents will then gravitate to the jurisdiction that offers the bundle most consistent with their interests. In order for that form of sorting to occur, however, potential residents must believe that the bundle that their jurisdiction offers will continue to coincide with residents’ interests. The role of courts, therefore, is to ensure that the mechanism for deciding the content of the bundle allows residents effectively to make their interests known. The content of the bundle (again, aside from external effects) is essentially irrelevant. That is, “any political outcome” is acceptable, assuming that the process reflects the results that would be obtained if decision makers had full information about residents’ interests and residents had influence over the results. Elhauge’s assumption of “equal political influence” seems relatively strong, and it is certainly stronger than what I have assumed in requiring the presence of countervailing forces, each of which has an opportunity to influence policy, but not necessarily as an equal of the others. Nevertheless, Elhauge still rejects the utility of interest group theory as administered by courts to produce an uncontroversial result. He concludes that because interest group theory assumes that each participant acts solely out of self-interest, a process justification is equivalent to applying a

188

the limit s o f j ud i c i a l i n te r v e n ti o n

substantive majoritarian standard in which more people are benefited than are harmed by an action. Such a standard, Elhauge contends, is simply an alternative to other substantive standards, such as utility or wealth maximization, rather than a replacement for them. Moreover, Elhauge contends, majoritarian standards are inadequate insofar as they fail to account for intensity of preferences. As a result, he concludes, concentrating on defects in the political process is inherently tied to a set of political outcomes and only substitutes one set of normative preferences for others. Subsequently, judicial intervention in the name of constraining the disproportionate influence of interest groups would only generate normatively contestable results and “may not reflect the views of the polity at all.”11 These are compelling criticisms and cannot be lightly dismissed. To the extent that judicial intervention is predicated on a finding of disproportionate influence, it necessarily implicates the question, disproportionate to what? An economic response of disproportionate to economic interests in the matter is as contestable as a majoritarian response of disproportionate to numbers of people holding the view expressed or the more difficult to measure issue of disproportionate to aggregate utility, taking intensity of interest into account. Nevertheless, Elhauge’s rejection of process may depend on demands for a level of exactitude in calculating proportions of influence that is unnecessary. In this sense, the criticism he levies invites clarification of an appropriate judicial role rather than total rejection of intervention. Judicial intervention as I have proposed it is intended to enhance, not displace, the political decisions concerning the provision of local public goods that would emerge from well-operating markets for residence. Recall that in the idealized Tiebout world, even those political decisions would be superfluous because individuals who are dissatisfied with the bundle of goods or services in their current jurisdiction would costlessly migrate to another. The unachievable nature of the Tiebout assumptions makes politics necessary. But if politics are to compensate for the absence of perfect mobility, then the political market must function in a manner that adequately allows deliberation among competing views about what comprises the bundle of local public goods for any given jurisdiction. In the face of imperfections in the political marketplace, some regulatory mechanism must deter political market failures from occurring and provide some corrective when they do occur, just as regulatory mechanisms are employed

th e l i m i ts o f j ud i c i a l int erv ent io n

189

in imperfect private markets to constrain market failures. Presumably, the constitutional prohibitions on certain kinds of expenditures and lawmaking provide ex ante deterrents against activity that would distort the results of a well-operating political market. Similarly, one can conceptualize a judicial role as correcting those distortions that materialize notwithstanding constitutional barriers. This does not mean that courts can or should attempt to identify and quantify all distortions, any more than we expect that the Federal Trade Commission or the Antitrust Division of the Department of Justice will correct all deviations from perfect competition in product markets. That level of precision would likely impose excessive costs, including error costs. What one would expect, however, is for courts to make a correction in those situations where there is a high probability that legislative results represent substantial political market failure, notwithstanding the presence of ex ante deterrents. Moreover, “correction” does not mean the imposition of a result that courts believe would emerge from a well-working political process, a remedy that would entail many of the objections that Elhauge raises concerning judicial competence. It means increasing the probability that local decisions about redistribution arise out of political processes that are consistent with the theories of decentralization that permit local decision making in the first instance. Indeed, judicial authority to invalidate perceived malign redistribution based on process defects is likely to increase the level of inclusion of groups that are more difficult to organize. The prospect of subsequent invalidation of redistributive benefits purchased at substantial cost induces those groups that can more easily organize to secure those benefits to include the group of payers that are more difficult to organize in the process. Doing so will facilitate the proponents’ efforts to rebut any claim of political market failure inherent in a determination of malign redistribution. But that same act of including net payers is likely to affect the substance of the redistributive proposal and reduce the distortions from municipal interests. LOCAL REDISTRIBUTION AND LOCAL JUSTICE

An additional objection to judicial invalidation of local redistributive programs emerges from jurisdictional limitations that courts face. At best, courts can allocate an entitlement, risk, or asset by applying legal principles governing the case to the facts before them. But courts, which are limited to fact finding based on evidence presented by parties, cannot easily analyze

190

th e limit s o f j ud i c i a l i n te r v e n ti o n

how application of the legal principles relevant to the instant case fits into a more universal allocation of goods in the society. Thus, we would be surprised to hear a court say, “Under the legal principles applicable to this case, the plaintiff has been negligently injured by the defendant and thus has a right to recover. Nevertheless, there are other situations in which persons in the position of defendant are injured by persons in the position of plaintiff and do not have the opportunity to recover. Thus, to balance the parties, we have decided not to make any award to plaintiff.” In this sense, courts can do local justice but have little capacity to do global justice. The consequence is that judicial intervention may exacerbate rather than cure broad societal misallocations. As applied to the case of local redistribution, the institutional constraints on courts affect the desirability of judicial intervention to constrain malign redistribution in at least three ways. First, courts may not comprehend the manner in which individual decisions implicate the mix of redistributive programs in an implicit bargain among different groups. Second, courts will only review redistributive programs that are enacted. They will be unable to review proposed benign redistribution that is blocked by dominant interest groups. If programs that are enacted systematically favor one group while those that are defeated systematically favor another group, then the availability of judicial review may skew the balance of distribution in the society at large, or global allocation. Third, the very existence of judicial review may be skewed in that some groups may have an advantage in prosecuting challenges to redistributive proposals. Judicial review and the redistributive mix Local redistributive programs constitute parts of a much larger redistributive puzzle. When seen from a broader perspective, the aggregate pieces of that puzzle may satisfy overall social preferences for redistribution, even though individual pieces, viewed in isolation, appear inconsistent with those preferences. Imagine, for instance, a community that consists of the very wealthy and the very poor, the local budget of which includes subsidies for local activities that are disproportionately used by the relatively rich (e.g., a municipal golf course, a municipal park in a wealthy neighborhood, a tax abatement for a building that will contain luxury condominiums), and explicit expenditures for activities that are disproportionately used by the relatively poor (a municipal homeless shelter, grants

th e l i m i ts o f j ud i c i a l int erv ent io n

1 91

to small businesses in low-income areas). This combination of subsidies through tax expenditures and explicit expenditures may represent an implicit bargain within the community about how to spend redistributive dollars in order to ensure that different groups (here, the rich and the poor) receive pro rata shares of the budget.12 Now assume that a court invalidates municipal funding of a golf course on the grounds that the proposed expenditure fails to serve a “public purpose.”13 The consequence may be to overturn the compromise among municipal residents about proper expenditures from the municipal budget. If golf course financing was included in order to create a majority coalition in favor of grants for municipally funded shelters, for instance, then invalidation of the former threatens the willingness of constituents to fund the latter. This is not to say that legitimate redistributive funding for the poor should be held hostage to illegitimate funding for the wealthy, or that courts should ignore explicit violations of municipal authority in order to uphold redistributive programs that may be part of an implicit local bargain. Rather, it is to suggest that municipal budgetary decisions necessarily reflect trade-offs between competing interests. Our willingness to allow courts to examine these processes emanates from a concern about legislative disregard for constituents’ interests. But the complexity of budgetary compromises frustrates efforts to determine whether a particular program does, in fact, reflect constituent preferences in a pluralist jurisdiction. Courts may correctly perceive a single expenditure to be contrary to the interests of a majority of constituents, but they may fail to perceive how that expenditure was necessary to attract the support of a minority interest necessary to create the majority coalition or to maintain stability in a community that assigns different parts of the budget to the discrete interests of different groups. Even if a court were to identify the expenditure as the consequence of a group that dominated on that interest, the issue of whether to classify it as malign is complicated to the extent that it was necessary to achieve majority support on the broader budget. The risk of altering the global mix is exacerbated if some of the redistributive expenditures within the municipal budget are susceptible to judicial review while others are not. Once a locality is authorized to enact a progressive local income tax, for instance, the applicable rates are likely to be within legislative discretion and not subject to judicial review. Explicit

192

t he limit s o f j ud i c i a l i n te r v e n ti o n

local expenditures or tax expenditures for particular projects, however, are more likely to be reviewable under a standard such as “public purpose.” Progressivity will favor the poor, while an explicit application of tax-exempt financing, such as to obtain funds to landscape a private housing development, may favor the wealthy.14 In this example, even if the same level of malign legislation and the same standards of judicial review exist with respect to all redistributive programs (those that benefit the wealthy and those that benefit the poor), judicial intervention will systematically affect the net effects of redistribution on local residents because some forms of redistribution will not be reviewed while others are. If programs that are reviewable systematically benefit a different group than those that are not, again the mix will be affected. More perverse results would obtain if, contrary to my example, the set of programs susceptible to judicial review, and hence to invalidation, tends to benefit the poor, while the second set, which escapes judicial review and potential invalidation, tends to benefit the wealthy. That may well characterize much of local redistribution. Tax expenditures in the form of abatements are likely to be insulated from challenge because the amounts are subject to administrative discretion concerning the underlying value of the real property, and information about the size and basis for the abatement may be difficult to obtain. Subsidies through tax expenditures may also be more likely to be used by the relatively wealthy, both because politically unpopular subsidies are easier to obfuscate this way and because abatements and tax credits will have greater value to those who would otherwise pay significant taxes than to the relatively poor. Explicit expenditures, in contrast, may be more valuable to the relatively poor, and their relative palatability may make them more useful to politicians. But if explicit expenditures are also more vulnerable to challenge than tax expenditures, then we get the perverse result that the relatively wealthy enjoy subsidies less vulnerable to judicial reversal than subsidies to the poor. Balancing malign motives in enacted and defeated programs The potential objection that I have raised to local redistribution arises from defects in the political process. That objection suggests that local interest groups may be able to lobby for enactment of programs that would not be enacted by local officials whose actions coincided perfectly with the interests of their constituents. Those same defects in the political process,

th e l i m i ts o f j ud i c i a l i nt erv ent io n

1 93

however, may also forestall enactment of programs that would serve the local public interest. Assume, for instance, that a prospective employer is considering locating in locality A and that it will create five hundred badly needed jobs for low-income workers in that locality. Locality A is considering a cost-effective tax abatement to attract the prospective employer. Locality A is already the home of firm X, which would have to compete for workers and consumers if the prospective employer locates in the municipality. Firm X, therefore, lobbies against enactment of the proposed tax abatement and is joined by other local residents who do not want to pay additional redistributive taxes. Even if the tax abatement—hypothetically necessary to attract the prospective employer—would return net benefits to the locality, the proposal may be defeated because those who would benefit (prospective employees who cannot easily identify each other and taxpayers who would receive a small individual benefit, albeit one that is large in the aggregate) cannot readily organize to lobby for enactment. Those who oppose it (such as firm X) can more readily organize since they can identify each other and, as current taxpayers and likely repeat players within the municipality, already have significant access to the local legislature. If the efficient abatement is denied, the interest of the public has been no less ignored than in the case where a locality enacts a program to satisfy a private interest. But the failure of the locality to enact the legislation means that there is no action for a court to review. My example is crafted to involve a situation in which the locality fails to enact legislation that would redistribute wealth to the wealthy, a prospective employer, even though that redistribution would, by hypothesis, benefit the locality as a whole. But my intuition is that when special interests defeat proposed local redistribution, they are more likely to do so with respect to programs that directly benefit the poor. The reason for the intuition should be familiar by now. Prospective wealthy employers have access to the decision-making process and, by virtue of repeat play within the locality, are likely to have the capacity to organize effectively to obtain the benefits of redistributive programs. We might be less confident of that result with respect to proposed redistribution on behalf of the poor. While I have suggested that there may be surrogates who help compensate for the relative inability of the poor to organize, those surrogates are likely to be most willing to intercede in support of programs that they have proposed and that also fit within the overall agenda of the

194

t h e limit s o f j ud i c i a l i n te r v e n ti o n

surrogate group, such as in the case of union support for the living wage ordinances. The result is that interest group intervention that defeats desirable redistribution, the consequences of which are not reviewable, may be more likely in the case of proposals to assist the poor than those to assist the wealthy. If courts subsequently review and invalidate malign local redistributive programs that are enacted, then the effect could be a significant skew in the sum total of redistribution within a locality. Redistribution for the wealthy will be more readily implemented than redistribution for the poor. Unless there is also a reason to believe that redistribution in favor of the wealthy will be less commonly enacted in the first instance or more frequently judicially invalidated than redistribution for the poor, the sum of redistributive programs that ultimately go into effect (that is, that are enacted and that avoid judicial invalidation) will, perversely, be skewed in favor of those that assist the wealthy and against those that favor the poor. Benign programs that benefit the poor will be frustrated at the enactment level, while those that benefit the wealthy will not. Malign programs that benefit each group will be frustrated at the judicial review stage. One possible implication of this argument is that courts should be reluctant to invalidate even malign local redistributive legislation that assists the poor in order to maintain a balance with local redistributive programs that assist the wealthy. If, as I have suggested, there are justifications for local redistribution in the first instance, then that objective may be better attained by allowing offsetting deviations from the public interest (that is, by judicial refusal to interfere with local decisions) rather than by invalidating all cases of malign redistribution apparent to the court. Otherwise, judicial invalidation would leave intact a distribution of local resources that failed to satisfy the locality’s overall interest in redistribution. In short, the “wrong” of failure to enact benign redistributive programs as a result of interest group opposition could be offset by the “wrong” of judicial refusal to invalidate malign programs that assist the poor. A very rough but perhaps defensible balance could be struck in which the total of redistributive spending was about right, even if the ultimate beneficiaries were not identical to those who would benefit under ideal conditions. But even if this “two wrongs make a right” conception is defensible in theory, it is less clear that courts have a capacity for analyzing how their decisions in particular cases affect the net redistribution within the

th e l i m i ts o f j ud i c i a l i nt erv ent io n

1 95

locality. If courts cannot analyze how their decisions fit into more global redistribution, perhaps it would be better for them to address individual cases without regard to global consequences. Even skepticism about judicial competence to measure global justice, however, does not allay the concern about the consequences of judicial intervention in local redistribution. Courts cannot avoid affecting the net amount of malign redistribution. Whatever courts do, whether they uphold or invalidate admittedly malign redistributive programs, will inevitably leave intact some maldistribution. If the “two wrongs” argument is accepted, courts may minimize global discrepancies but will preserve inappropriate distributions in individual cases. If courts invalidate all malign programs, they may minimize maldistribution in particular programs, but only at the expense of global distributive justice within the locality. In the absence of a way of comparing the benefits of these outcomes, I ultimately remain agnostic on the best judicial strategy. The concern about interest group capacity to defeat benign redistribution also has implications for the first issue I raised in this chapter concerning whether the state or local legislature is more vulnerable to dominant interest groups. As I indicated above, the organizational structure of state legislatures—including the presence of bicameralism and committee gatekeeping devices—privileges the status quo and allows minority interests to prevail more readily. At the state level, opponents of a benign redistributive proposal can forestall its passage by prevailing in one house of the legislature or by convincing only the committee that has jurisdiction over the matter to refrain from bringing it to the full body for a vote. Proponents must clear multiple hurdles to obtain enactment of the legislation. Given my conclusion above that, for at least some groups, organization and access costs are likely to be higher at the state level than at the local level, wealthier groups or those who desire to maintain the status quo may systematically enjoy an advantage in finding a pressure point in which it is sufficiently dominant to stifle state authorization for local redistributive legislation, notwithstanding its benign character.

ch ap t e r e i g h t

Conclusion

when practice deviates from theory, something is wrong. The problem could be with the practice, in that actors charged with achieving the results that theory predicts act in an inappropriate manner. Or, the problem could be with the theory itself. With respect to the deviation between the orthodox theory of urban finance, which criticizes local redistribution, and practice, which endorses it, both problems appear to be at work. The orthodox theory understates the extent to which local governments can obtain unique benefits or otherwise reflect local preferences by redistributing wealth. The practice, however, may reflect political dominance by groups who do not represent a majority of constituents. The difficult part is figuring out, with respect to any given redistributive effort, which of these motivations is at work. Political bargaining that obfuscates any inference of majority preferences, collective action problems that interfere with registering those preferences, the scope of judicial legitimacy to inquire into political processes, and the limited accuracy of judicial inquiries even when legitimately made all complicate any effort to ensure that local actions coincide with local interests. Nevertheless, some effort in that direction seems appropriate if a plausible strategy can be devised. Theories of local government vary among those based in economics, such as sorting, and those based in democratic theory. Each of these theories, however, is ultimately rooted 1 96

co ncl us io n

1 97

in a belief that local action is appropriate if, but only if, it reflects the interests of a majority of local residents. That is not a sufficient condition for local initiative: neither theory of local government would permit the imposition of significant consequences on nonresidents who were not represented in the decision-making process. But it is a necessary condition. No theory of local government law provides that localities should act in accordance with the interests of a politically dominant minority that is capable of ignoring the majority. The question, then, is how to neutralize the effect of dominant interest groups in municipal governance. I have suggested that judicial intervention is a plausible response, perhaps supported more by the constitutional relationship among government branches in the states than has previously seemed appropriate. But the practical difficulties of a judicial role cannot be understated. Courts cannot easily reverse engineer municipal ordinances in order to detect the presence of dominant interest groups, cannot easily discern malign effects from the presence of conditions that suggest dominant interest groups played a role in enacting ordinances, and cannot predict with confidence the social consequences of invalidating even demonstrable malign redistribution. Nevertheless, when courts are able to define situations in which there is a high probability that redistribution was motivated by something other than concern for municipal welfare, the risk of inaccuracy seems sufficiently small to warrant intervention, especially in light of the institutional advantages that even imperfect courts have over other state and local institutions for constraining special interests. Of course, the need to risk inaccurate judicial intervention could be reduced if we are able to identify municipal processes or review mechanisms that promote benign redistribution and impede malign redistribution. If dominant interest groups are a problem, then one ostensible solution would be to promote the participation of competing interest groups in the local decision-making process. I have suggested that the problem of collective action by those who have only small stakes in a local proposal underlies the ability of a dominant minority to redistribute municipal resources for malign purposes. It is possible that subsidizing the formation of interest groups to represent competing perspectives would facilitate a debate among countervailing groups that is more likely to reflect municipal interests. This might mean inviting community representatives whom

198

c o n clu sio n

the local legislature can identify as affected by proposals to enter the fray, rather than waiting for them to organize and request to be heard. In the area of economic development, for instance, some municipalities have induced developers to negotiate with community representatives in an effort to reach some accommodation about the scope of projects that would displace existing residents and businesses. It is unclear, however, that self-appointed representatives will actually serve as faithful agents of their presumed constituents. Given the standard problem of collective action, it is virtually certain that anyone who would assume the role of a community representative with respect to a discrete issue such as development of a specific project will have an idiosyncratic interest. Perhaps the project threatens that person’s home, business, or tranquillity in a way not shared by the “community” at large. Any settlement between the beneficiary of local redistribution and such representatives may involve resolution of the latter’s discrete difficulty but will not necessarily achieve the Madisonian solution of countervailing interest groups composed of elected representatives debating within a public forum. Alternatives that could facilitate countervailing forces involve reducing search costs that frustrate organization. Cost reduction could alter the calculus of those who avoid collective action because the personal benefits of success do not warrant the personal costs of participation. For instance, cities could be required to provide more notice of redistributive proposals, or even require an “economic impact statement” of specified redistributive proposals that would indicate the incidence as well as the costs of regulatory and direct redistribution. The accounts of living wage battles with which I began indicate that local debates have been characterized by competing claims about the fiscal impact of higher wages on local businesses. But local legislatures rarely explicitly adopt one set of findings to support the votes they ultimately cast. Legislative adoption of one set of claims about fiscal impacts both reduces the need for constituents to decide which of the contested studies is likely to be most influential and provides a readily accessible focal point for those who want to analyze the basis for the ultimate legislative determination. If one can’t raise the bridge by creating more incentives for more representation in local decision making, perhaps one can lower the water by facilitating judicial distinctions between malign and benign

co ncl us io n

1 99

redistribution. I have suggested that a major obstacle to judicial review of local action lies in the paucity of information concerning the local legislative process, especially as compared to the full reports of hearings and debates that occur at the federal level. Increasing the quantity and quality of reporting of debates in legislative chambers would provide courts with significant information about local legislative motivations and provide a means of testing the purported public purposes to be served by redistributive ordinances. This suggestion would do little to mollify those who oppose the use of legislative history to discern legislative intent on any grounds; but those objections are often directed at judicial efforts to discern congressional objectives, and detecting legislative intent in a body that consists of a small number of local representatives is arguably quite different from engaging in the same practice with respect to a bicameral legislature that comprises some 535 members. Economic impact statements of regulatory mandates would similarly provide a means by which courts could consider the effects of proposed redistribution. Indeed, similar information that already exists tends to be dramatically underutilized by courts. In many cities, for many capital projects, especially those to be funded by bonds, feasibility studies will have been created to determine the size and scope of the project’s expected benefits. A feasibility study for a publicly financed stadium, for instance, is likely to contain projections about revenues from operation of the stadium and from related business as well as the underlying assumptions on which those projections are based. Of course, many of these projects have now been in operation for an extensive period of time. Yet, there is little if any work that compares the projections contained in the earlier feasibility study on which approvals were based to the actual performance of the project. One plausible check on the incentives of public officials to overstate benefits and understate costs ex ante would entail a requirement of postcompletion publication of comparisons between actual performance of a subsidized project at various intervals and the anticipated performance as detailed in feasibility studies on which the locality relied in approving the project. Perhaps the greatest problem of these recommendations involves their cost. Even if increasing information about legislative proceedings was useful to reviewing courts, the costs of subsidizing potentially competing interest groups, generating economic impact statements, and creating and

2 00

c o n clu sio n

maintaining records would have to be incurred in all cases, while judicial use of them would be required only in the very small number of cases in which the proposal was challenged. It is not clear, therefore, that the game is worth the candle, even if greater transparency of decision making prompted invalidation of some malign redistribution. Of course, these procedures themselves would have other benefits. They might constrain the enactment of malign legislation in the first instance, or they might have the democratizing effect of providing additional information to citizens, wholly apart from the effect on the quality of legislation. None of these recommendations suggests a panacea to the perpetual difficulty of neutralizing dominant interest groups whose self-interest is at odds with the larger municipal interest. But perhaps a greater understanding of the positive and negative roles of redistribution in the theory of municipal corporations will allow more-nuanced appreciation for those redistributive programs that advance municipal interests and more-subtle impediments for those that do not.

NOTES

CHAPTER ONE. THE SCOPE OF LOCAL COMPETENCE

1. Katrina Vanden Heuvel and Sam Graham-Felsen, “Chicago’s Living Wage,” The Nation, 25 Sept. 2006, (29 June 2009). 2. Monica Davey and Michael Barbaro, “Minimum Wage Ordinance Fails Last Hurdle in Chicago,” New York Times, 14 Sept. 2006, (29 June 2009). 3. Deanna Bellandi, “Chicago Council Passes ‘Living Wage’ Act,” Washington Post, 27 July 2006, (29 June 2009). 4. “Chicago May Rethink ‘Living Wage’ Law,” 16 Aug. 2006,

(29 June 2009). 5. Mayor Daley’s veto letter can be found at (29 June 2009). 6. See Richard H. Sander and E. Douglass Williams, “Santa Monica’s Minimum Wage: Assessing the Living Wage Movement’s New Frontier,” Economic Development Quarterly 19, no. 1 (2005): 25–44. 7. The transcript of the hearing reveals the political, economic, and emotional stakes: Dr. Robert Pollin Dr. Pollin, an economist, thanked Councilor Coss for inviting him to this meeting. He said he has done studies on living wage ordinances around the country, including Los Angeles, New Orleans and Santa Monica. . . . Dr. Pollin 201

2 02

n o t e t o p ag e 5 said the minimum wage in 1968, adjusted for inflation, was $8.15; today, it is $5.15. He stated that productivity in the U.S. economy has gone up by about 70% during this period while there has been a 35% fall in the absolute value of what somebody can buy with the minimum wage. He pointed out that, had the minimum wage gone up penny for penny with the rate of productivity, the minimum wage today would be $14.65. Dr. Pollin spoke to concerns on the so-called law of unintended consequences, i.e., 1) the City is trying to help low wage workers but may end up hurting them by causing unemployment, because businesses would be less willing to hire low wage workers, and 2) that businesses may relocate out of the city. . . . Dr. Pollin said the cost increase to the average firm, according to his own research, would be about 1% to 2% of their total costs. He stated that these costs could be absorbed “through a 1% to 2% price increase or through small improvements in productivity.” David Walther Mr. Walther stated that he is a lawyer in Santa Fe with a small firm, which has 10 employees. He pointed out that “labor is not a commodity, and workers have a right to a living wage. A business does not have the right to continue if it cannot pay a living wage.” David Gold Speaking to the argument against this ordinance that businesses will go elsewhere rather than come to Santa Fe, Mr. Gold said the only businesses that would go elsewhere are those looking for low wages, and those businesses will go to Mexico or Asia, which is a national trend. He said the only businesses that would come here are those attracted to the high quality of life here, and they would be willing to pay more money. Addressing the argument that children will drop out of school to take advantage of the $8.50 an hour wage, Mr. Gold said he has coached basketball at Agua Fria Elementary, and to get kids to stay in school means giving them a good job when they leave. He said having good jobs in Santa Fe is the best incentive for kids to stay in school. Mara Lloyd Ms. Lloyd, a waitress, appeared on behalf of the workers in Santa Fe who could not be present tonight because they are single parents or are working. Ms. Lloyd stated that many restaurant workers are Spanish-speaking and have been excluded from tonight’s proceedings because they do not speak English, and also because of socioeconomic barriers. She stated, “We’re not asking for a handout. We’re asking for a living wage. Please remember the people like me who could not be here tonight.” Ms. Lloyd stated that she works at least 40 hours per week, and pays $250 a month rent, and her paycheck for two weeks does not cover her rent. She said she lives in a trailer with two other people and has a tiny bedroom with just enough room for her bed.

n o t es t o p ag e 5

203

Loretta Martin Ms. Martin, a small business owner in Santa Fe, stated that she has invested all of her savings in this business, and has moved into a smaller house so she can afford to own the business. She stated that many small business owners, including her, “are part of the working poor.” She said she pays her employees more than the proposed living wage because they earn it by showing up on time and providing exceptional value to her customers. She stated that, in the last two years, she has hired people “who couldn’t find the tab key on the keyboard after three days, people who quit showing up after a week, sat down at every opportunity.” She commented, “I would gladly hire someone if I could find someone who had skills and a strong work ethic.” Jill Markstein Ms. Markstein, a co-owner of the Marketplace Natural Grocery, said her store would be at a great disadvantage if this ordinance passed compared to large business. She stated that the Marketplace starts people at $8 and that goes up to over $20. She said they offer health insurance and have an EAP program. She stated that they never lay people off at the top; and when times are tough, they do not lay off their longtime employees, do not cut benefits, and do not cut hours. She stated that her store has not raised its prices in years, and if it has to do so, it will not be able to compete with large businesses that pay a lot less for their product, and are also subsidized by having many stores. Michael Chapman Mr. Chapman said his family has been in business in Santa Fe for 37 years, and their average employee has been with them over 20 years. He stated that he thought it “abominable and ridiculous” for some people to suggest that employees in Santa Fe are “slaves.” Mr. Chapman said the Council has no right or responsibility to come between him and his employees, and to take their bargaining rights with him away. He said, “Shame on you for considering this.” Mr. Chapman stated that he is past chairman of the Santa Fe Business Incubator, and while some people have said this ordinance will not affect startups, there are several new businesses with more than ten employees and this will affect them. Ed Tinsley Mr. Tinsley, president of a restaurant chain that employs approximately 1,000 people, said Santa Fe led the job growth market in New Mexico in 2000, and 39% of those private sector jobs were directly related to the restaurant industry. He stated that this ordinance will have a domino effect, extending to the contractor who builds the restaurant, service providers such as plumbers and electricians, and the tax base. He urged the Council to think globally. He said his restaurant chain will not locate in Santa Fe if the ordinance is passed. 8. New Mexicans for Free Enterprise v. City of Santa Fe, 126 S.W.3d 1149 (N.M. Ct. App. 2005).

2 04

n o t es t o p ag e s 6– 1 9

9. New Orleans Campaign for a Living Wage v. City of New Orleans, 825 So. 2d 1098 (La. 2002). 10. Rachel I. Rosen, “The Rise and Potential Fall of Living Wage Laws: Missouri Hotel and Motel Ass’n v. City of St. Louis,” Journal of Law and Commerce 21, no. 1 (Fall 2001): 131–146. 11. Visiting Homemaker Service of Hudson County v. Board of Chosen Freeholders, 883 A.2d 1074 (N.J. App. Div. 2005). These decisions follow earlier approvals of local minimum wage ordinances. Mayor and City of Baltimore v. Sitnick, 255 A.2d 376 (Md. 1969); City of Atlanta v. Associated Builders and Contractors, 242 S.E.2d 139 (Ga. 1978). 12. RUI One Corp. v. City of Berkeley, 371 F.3d 1137 (9th Cir. 2004). 13. For reviews of literature testing the effects of differential welfare benefits, see Jan K. Brueckner, “Welfare Reform and the Race to the Bottom: Theory and Evidence,” Southern Economic Journal 66, no. 3 (Jan. 2000): 505–525; Robert Moffitt, “Incentive Effects of the U.S. Welfare System: A Review,” Journal of Economic Literature 30, no. 1 (March 1992): 1–61. Brueckner finds mixed evidence that welfare recipients migrate for higher benefits but does not speculate about emigration of the rich from those jurisdictions. For an interesting historical study that suggests that the poor did not migrate from one locality to another nearby one that offered more generous welfare benefits, see Kyle D. Kauffman and L. Lynne Kiesling, “Was There a Nineteenth Century Welfare Magnet in the United States? Preliminary Results from New York City and Brooklyn,” Quarterly Review of Economics and Finance 37 (1997): 439. 14. 2007 Census of Governments: Finances of Municipal and Township Governments, available at . The explanation of “Public Welfare” expenditures indicates that they include “any payments for applicable cash benefits in excess of, or supplementary to, those financed with federal or state participation. General relief, which is wholly financed from state and local sources, makes up most other cash assistance.” Other public welfare spending includes “vendor payments under various public welfare programs such as the federally supported medical care program commonly known as Medicaid; institutional care for the needy; and administration of welfare activities.” Thus, it appears that these figures are net of federal and state contributions. See ibid., ix. 15. Ibid. Total direct expenditures were calculated at $466.4 billion. 16. Jessica Trounstine, Political Monopolies in American Cities (Chicago: University of Chicago Press, 2008), 148–161. 17. Thomas Romer and Howard Rosenthal, “Bureaucrats Versus Voters: On the Political Economy of Resource Allocation by Direct Democracy,” Quarterly Journal of Economics 93, no. 4 (Nov. 1979): 563–587. 18. Edward L. Glaeser and Andrei Shleifer, “The Curley Effect: The Economics of Shaping the Electorate,” Journal of Law, Economics, and Organization 21, no. 1 (2005): 1–19. 19. Cycling involves situations in which there are more than two choices and no choice can beat the other choices in a majority-rule contest. For instance, in a

n o te s to p ag es 24 – 37

205

choice among A, B, and C, A might beat B in a pairwise contest, B might beat C, and C might beat A. 20. Richard B. Stewart, “Madison’s Nightmare,” University of Chicago Law Review 57, no. 2 (Spring 1990): 335–356. 21. See Albert O. Hirschman, The Passions and the Interests (Princeton, N.J.: Princeton University Press, 1977), 20–31. 22. Alexander Hamilton, James Madison, John Jay, The Federalist Papers, ed. Clinton Rossiter (New York: New American Library, 1961), 322; David Hume, “On the Independency of Parliament,” in Hume, Political Essays, ed. Knud Haakonssen (Cambridge: Cambridge University Press, 1994), 24–27. CHAPTER TWO. WHY DOES LOCAL AUTONOMY MATTER?

1. Robert Dahl and Edward Tufte, Size and Democracy (Palo Alto, Calif.: Stanford University Press, 1973); J. Eric Oliver, Democracy in Suburbia (Princeton, N.J.: Princeton University Press, 2001). 2. We are, apparently, willing to impose some constraints on mobile capital. Witness plant-closing laws and the federal Worker Adjustment and Retraining Notification Act (WARN), which requires sixty days’ advance notification of certain plant closings and mass layoffs. 3. See Michael McConnell, “Federalism: Evaluating the Founders’ Design,” University of Chicago Law Review 54, no. 4 (Fall 1987): 1484–1512. I am assuming that the higher per capita cost that the six hundred will pay when they alone fund the playground does not affect preferences. 4. See Timothy Besley and Anne Case, “Incumbent Behavior: Vote Seeking, Tax Setting and Yardstick Competition,” American Economic Review 85, no. 1 (March 1995): 24–45; Mark Schneider, “Intermunicipal Competition, Budget-Maximizing Bureaucrats, and the Level of Suburban Competition,” American Journal of Political Science 33 (Aug. 1989): 612–628. 5. See Wallace E. Oates, “Fiscal and Regulatory Competition: Theory and Evidence,” Perspektiven der Wirtschaftspolitik 3, no. 4 (2002): 377–390. 6. Wallace E. Oates, “The Effects of Property Taxes and Local Public Spending on Property Values: An Empirical Study of Tax Capitalization and the Tiebout Hypothesis,” Journal of Political Economy 77, no. 6 (Nov.–Dec. 1969): 957–971; Jan Brueckner, “A Test for Allocative Efficiency in the Local Public Sector,” Journal of Public Economics, 19 (Dec. 1982): 311–331; William A. Fischel, The Homevoter Hypothesis (Cambridge, Mass.: Harvard University Press, 2001); Richard Thompson Ford, “Bourgeois Communities: A Review of Gerald Frug’s City Making,” Stanford Law Review 56, no. 1 (Oct. 2003): 231–252, 243–245; Frances Fitzgerald, Cities on a Hill (New York: Simon and Schuster, 1987), 203–216. See Daniel Rubinfeld, “The Economics of the Local Public Sector,” in Handbook of Public Economics, ed. A. Auerbach and M. Feldstein (New York: North Holland, 1987), 2:571–545. On evolutionary processes that cause arbitrarily selected routines to stick, see Armen Alchian, “Uncertainty, Evolution, and Economic Theory,” Journal of Political Economy 58, no. 3 (June 1950): 211–221.

2 06

n o t es t o p ag e s 3 7– 5 1

7. Gerald E. Frug, City Making (Princeton, N.J.: Princeton University Press, 1999), 169. 8. Peter Enrich, “Saving the States from Themselves: Commerce Clause Constraints on State Tax Incentives for Business,” Harvard Law Review 110, no. 2 (Dec. 1996): 377–468; John Douglas Wilson, “Theories of Tax Competition,” National Tax Journal 52, no. 2 (June 1999): 269–304; David E. Wildasin, “Interjurisdictional Capital Mobility: Fiscal Externality and a Corrective Subsidy,” Journal of Urban Economics 25, no. 2 (March 1989): 193–212. 9. Clifford J. Levy, “Seeing Life Outside New Orleans Alters Life Inside It,” New York Times, 20 Nov. 2005, (29 June 2009). 10. Lee Anne Fennell, “Contracting Communities,” University of Illinois Law Review, 2004, no. 4:829–898, 873–877. 11. Ronald Coase, “The Problem of Social Cost,” Journal of Law and Economics 3 (1960): 1–44. 12. Mark Schneider, “Income Homogeneity and the Size of Suburban Government,” Journal of Politics 49, no. 1 (Feb. 1987): 36–53. 13. Elaine B. Sharp and Steven Maynard-Moody, “Theories of the Local Welfare Role,” American Journal of Political Science 35, no. 4 (Nov. 1991): 934–950. 14. Paul E. Peterson, City Limits (Chicago: University of Chicago Press, 1981), 53. 15. Robert C. Ellickson, “Cities and Homeowners Associations,” University of Pennsylvania Law Review 130, no. 6 (June 1982): 1519–1580; Robert C. Ellickson, “New Institutions for Old Neighborhoods,” Duke Law Journal 48, no. 1 (Oct. 1998): 75–110; Richard Briffault, “A Government for Our Time? Business Improvement Districts and Urban Governance,” Columbia Law Review 99, no. 2 (March 1999): 365–477. The caveat in the text is important. There are cases in which those who can benefit from an increased level of provision will simultaneously attempt to reduce public levels of expenditure from which they don’t benefit. But in many of these cases, even those who wish to supplement the baseline level of provision will want that baseline to be high in order to minimize the cost of supplementation. See Clayton P. Gillette, “Opting Out of Public Provision,” Denver University Law Review 73, no. 4 (1996): 1185–1219, 1209–1210. 16. Madison, Federalist #10. 17. Richard Briffault, “Town of Telluride v. San Miguel Valley Corp.: Extraterritoriality and Local Autonomy,” Denver University Law Review 86 (2009): 1311–1328. 18. Town of Telluride v. Lot Thirty-Four Venture, L.L.C., 3 P.3d 30 (Colo. 2000). 19. See, e.g., New Jersey Statutes Annotated § 34:11–56a4, which permits political subdivisions in New Jersey to enter into agreements with vendors and contractors of the subdivision that require wage rates higher than the standards imposed by the state. 20. John Stuart Mill, Considerations on Representative Government, ed. Currin V. Shields (New York: Bobbs Merrill, 1958), 224-226.

n o te s to p ag es 5 3– 5 9

207

CHAPTER THREE. THE MEANING AND SCOPE OF LOCAL REDISTRIBUTION

1. See, e.g., Timothy J. Bartik, Evaluating the Impacts of Local Economic Development Policies on Local Economic Outcomes: What Has Been Done and What is Doable?, Upjohn Institute Staff Working Paper No. 03–89 (Nov. 2002), (30 June 2009); Timothy J. Bartik, Jobs for the Poor: Can Labor Demand Policies Help? (New York: Russell Sage Foundation Publications, 2001), 251. 2. Paul E. Peterson, City Limits (Chicago: University of Chicago Press, 1981), 43. 3. See Aaron S. Yelowitz, Santa Fe’s Living Wage Ordinance and the Labor Market (Employment Policies Institute, Sept. 2005). 4. Nicholas Potter, Measuring the Employment Impacts of the Living Wage Ordinance in Santa Fe, New Mexico (University of New Mexico, Bureau of Business and Economic Research, 30 June 2006), (29 June 2009); Robert Pollin and Jeannette Wicks-Lim, Comments on Aaron Yelowitz, “Santa Fe’s Living Wage Ordinance and the Labor Market,” PERI Working Paper No. 108 (Amherst, Mass.: Political Economy Research Institute, Oct. 2005), (29 June 2009). 5. David Neumark and Scott Adams, “Detecting Effects of Living Wage Laws,” Industrial Relations 42, no. 4 (Oct. 2003): 531–564; David Neumark and Scott Adams, “Do Living Wage Ordinances Reduce Urban Poverty?,” Journal of Human Resources 38, no. 3 (Summer 2003): 490–521; Scott Adams and David Neumark, “Living Wage Effects: New and Improved Evidence,” Economic Development Quarterly 19, no. 1 (2005): 80–102; Scott Adams and David Neumark, “The Effects of Living Wage Laws: Evidence from Failed and Derailed Living Wage Campaigns,” Journal of Urban Economics, 58, no.2 (Sept. 2005): 177–202. 6. Michael Reich, Peter Hall, and Ken Jacobs, “Living Wage Policies at the San Francisco Airport: Impacts on Workers and Businesses,” Industrial Relations 44, no.1 (Jan. 2005): 106–138. 7. Robert Pollin, “Evaluating Living Wage Laws in the United States: Good Intentions and Economic Reality in Conflict?,” Economic Development Quarterly 19, no. 1 (Feb. 2005): 3–24. Richard Sander and E. Douglass Williams calculated that if the Santa Monica living wage ordinance had gone into effect it would have lifted families out of poverty in a miniscule number of cases and would have displaced the lowest-wage workers. They concluded that households in the bottom decile would have received disproportionately smaller benefits from the ordinance than other groups. Richard H. Sander and E. Douglass Williams, “Santa Monica’s Minimum Wage: Assessing the Living Wage Movement’s New Frontier,” Economic Development Quarterly 19, no. 1 (2005): 25–44. 8. Hendrik Hartog, Public Property and Private Power: The Corporation of the City of New York in American Law, 1730–1870 (Ithaca, N.Y.: Cornell University Press, 1983), 147.

2 08

n o t es t o p ag e s 5 9– 64

9. Ibid., 153. 10. Paul Studenski and Herman Edward Krooss, Financial History of the United States (New York: McGraw-Hill, 1951), 134. 11. Oscar Handlin and Mary Flug Handlin, Commonwealth: A Study of the Role of Government in the American Economy, Massachusetts 1774–1861 (Cambridge, Mass.: Harvard University Press, 1969), 53. 12. Charles River Bridge v. Warren Bridge, 11 Peters 420 (1837); Crease v. Babcock, 23 Pickering 334 (Mass. 1839). 13. William P. Quigley, “The Quicksands of the Poor Law: Poor Relief Legislation in a Growing Nation,” Northern Illinois Law Review 18 (1997): 1–98. 14. Ibid.; Michael B. Katz, In the Shadow of the Poorhouse (New York: Basic Books, 1986), 15. 15. Katz, In the Shadow, 12–13. William P. Quigley, “Reluctant Charity: Poor Laws in the Original Thirteen Colonies,” University of Richmond Law Review 31 (1997): 111, 130–132; Quigley, “The Quicksands,” 12–13, 18; William P. Quigley, “Rumblings of Reform: Northern Poor Laws 1820–1860,” Capital University Law Review 26 (1997): 739–774, 760–762. 16. Katz, In the Shadow, 15–23; Benjamin Joseph Klebaner, Public Poor Relief in America, 1790–1860 (New York: Arno Press, 1976), 352. 17. J. V. N. Yates, “Report of the Secretary of State in 1824 on the Relief and Settlement of the Poor,” in 34th Annual Report of the State Board of Charities of the State of New York (Albany, N.Y.: James B. Lyon, 1901), 1063; “Report of the Committee Appointed by the Board of Guardians of the Poor of the City and Districts of Philadelphia to Visit the Cities of Baltimore, New-York, Providence, Boston, and Salem” (Philadelphia: Board of Guardians of the Poor, 1827), reprinted in The Almshouse Experience: Collected Reports (New York: Arno Press, 1971), 7, 16; Robert E. Cray, Paupers and Poor Relief in New York City and Its Rural Environs (Philadelphia: Temple University Press, 1988), 124; Klebaner, Public Poor Relief, 43, 127. 18. Joan E. Marshall, “Shaping Poor Relief for the Sick-Poor in Indiana’s Pioneering Era, Tippecanoe County, Indiana, 1826–1846,” Social Service Review 74, no.4 (Dec. 2000): 560–568. 19. Katz, In the Shadow, 5–22; Quigley, “Rumblings of Reform,” 760–761. 20. Katz, In the Shadow, 22–25. 21. Klebaner, Public Poor Relief, 127. 22. “Report of the Committee,” 16. 23. Yates, “Report of the Secretary of State,” 957. 24. Katz, In the Shadow, 22–24. 25. Klebaner, Public Poor Relief, 99–102. 26. Katz, In the Shadow, 15. 27. Ibid.; Klebaner, Public Poor Relief, 88. 28. Klebaner, Public Poor Relief, 74. 29. Ibid., 96. 30. Quigley, “The Quicksands,” 22–23.

n o te s to p ag es 64 – 69

209

31. Raymond A. Mohl, Poverty in New York: 1783–1825 (New York: Oxford University Press, 1971), 144–145; Eric H. Monkkonen, “Nineteenth-Century Institutions: Dealing with the Urban ‘Underclass,’ ” in The “Underclass” Debate, ed. Michael B. Katz (Princeton, N.J.: Princeton University Press, 1993), 347. 32. Klebaner, Public Poor Relief, 329. 33. Ibid., 339. 34. Katz, In the Shadow, 36. 35. M. J. Heale, “From City Fathers to Social Critics: Humanitarianism and Government in New York, 1790–1860,” Journal of American History, 63 (June 1976): 29–38. 36. Katz, In the Shadow, 36–71. 37. Ibid., 46–51. 38. Frederic Almy, “The Relation Between Private and Public Outdoor Relief,” Charities Review 9, no. 1 (1899): 65–71. 39. Katz, In the Shadow, 37–39. 40. Ibid., 44. 41. Klebaner, Public Poor Relief, 478. 42. Jon C. Teaford, The Unheralded Triumph: City Government in America, 1870–1900 (Baltimore, Md.: Johns Hopkins University Press, 1984), 269. 43. Ibid., 270. 44. Joy J. Jackson, New Orleans in the Gilded Age (Baton Rouge, La.: Louisiana Historical Association and Center for Louisiana Studies, 1969), 188. 45. Teaford, The Unheralded Triumph, 270. 46. Ronald R. Weiner, Lake Effects: A History of Urban Policy Making in Cleveland (Columbus, Ohio: Ohio State University Press, 2005), 80–82. 47. Robin Einhorn, Property Rules: Political Economy in Chicago, 1833–1872 (Chicago: University of Chicago Press, 2001), 19. 48. Stephen Diamond, “The Death and Transfiguration of Benefit Taxation: Special Assessments in Nineteenth-Century America,” Journal of Legal Studies 12, no. 2 (June 1983): 201–240, 206. 49. Ibid., 212–214. 50. Katz, In the Shadow, 147–171. 51. Ibid., 147. 52. Ibid., 139–171. 53. Ibid., 139. 54. Ibid.; Marilyn Thorton Williams, Washing “The Great Unwashed”: Public Baths in Urban America, 1840–1920 (Columbus, Ohio: Ohio State University Press, 1991), 39. 55. Katz, In the Shadow, 208–214. 56. Eric H. Monkkonen, The Local State: Public Money and American Cities (Stanford, Calif.: Stanford University Press, 1995), 57–77. 57. As formulated by its originator, John Dillon, the Rule permitted localities to exercise only those powers expressly granted by the legislature, necessarily implied from express grants, and essential to accomplishment of the declared

2 10

58.

59.

60. 61.

62. 63.

64.

65.

66. 67.

n o t es t o p ag e s 69– 74 purposes of localities. Any doubt was to be resolved against the exercise of a power. See John F. Dillon, Commentaries on the Law of Municipal Corporations (Boston: Little, Brown, 1911), 1:448–451. For example, Lowell v. City of Boston, 111 Mass. 454 (1873) (invalidating bonds proposed to be issued by City of Boston to reconstruct housing destroyed in conflagration). Subsequent constitutional restrictions on the state, however, revealed that state legislators could not be trusted to restrain local redistribution. Constitutional amendments that prohibited the legislature from enacting “special legislation” in favor of one or a few localities became popular in the latter part of the nineteenth century. The legislative enactments at which the prohibition was directed had tended to permit specific localities to dedicate local resources to particular groups that apparently had more political support in the legislature than in city councils. Teaford, The Unheralded Triumph, 84–94, 105–107. See, e.g., City of Miami Beach v. Fleetwood Hotel, Inc., 261 So.2d 801 (Fla. 1972). See, e.g., Wholesale Laundry Board of Trade v. City of New York, 234 N.Y.S.2d 862 (N.Y. App. Div. 1962) (invalidating local minimum wage ordinance). But see McMillen v. Brown, 14 N.Y.S.2d 326 (N.Y. 1964) (upholding local minimum wage ordinance). See, e.g., Greater Boston Real Estate Board v. City of Boston, 705 N.E.2d 256 (Mass. 1999). Some of these opinions indicated antipathy toward redistribution at more-centralized levels of government as well, insofar as they blocked assistance by the state as well as local governments. For example, Opinion of the Justices, 98 N.E. 611 (Mass. 1912) (public funds could not be used to provide homes to wage earners); Salisbury Land & Improvement Co. v. Commonwealth, 102 N.E. 619 (Mass. 1913). See, e.g., Massachusetts Home Mortgage Finance Agency v. New England Merchants National Bank, 382 N.E.2d 1084 (Mass. 1978); Massachusetts Housing Finance Agency v. New England Merchants National Bank, 249 N.E.2d 599 (Mass. 1969). See, e.g., WDW Properties v. City of Sumter, 535 S.E.2d 631 (S.C. 2000). For discussion of this phenomenon, see Robert S. Amdursky and Clayton P. Gillette, Municipal Debt Finance: Theory and Practice (Boston: Little, Brown, 1990), 114–157; Richard Briffault, “The Disfavored Constitution: State Fiscal Limits and State Constitutional Law,” Rutgers Law Journal 34 (Summer 2003): 907–957. See, e.g., In re Oklahoma Dev. Finance Authority, 2004 WL 728854 (Okla.). Kelo v. City of New London, 545 U.S. 469 (2005).

CHAPTER FOUR. WHY DO LOCAL GOVERNMENTS REDISTRIBUTE?

1. Paul E. Peterson, City Limits (Chicago: University of Chicago Press, 1981), 20. 2. Ibid.

n o te s to p ag es 75 – 8 6

21 1

3. Andrew F. Haughwout, Robert P. Inman, Steven Craig, and Thomas Luce, “Local Revenue Hills: Evidence from Four U.S. Cities,” Review of Economics and Statistics 86, no. 2 (May 2004): 570–585. 4. Thomas Romer and Howard Rosenthal, “Bureaucrats Versus Voters: On the Political Economy of Resource Allocation by Direct Democracy,” Quarterly Journal of Economics 93, no. 4 (Nov. 1979): 563–587. 5. See Albert O. Hirschman, The Passions and the Interests (Princeton, N.J.: Princeton University Press, 1977), 40; S. M. Shumer, “Machiavelli: Republican Politics and Its Corruption,” Political Theory 7, no. 1 (Feb. 1979): 5–34. 6. See Kirk J. Stark, “City Welfare: Views from Theory, History, and Practice,” Urban Lawyer 7, no. 3 (Summer 1995): 495–529, 517–520. 7. See Conn. Gen. Stat. Ann. § 17b-116. 8. See Citizens Budget Commission, Confronting the Tradeoffs in Medicare Cost Containment (2004), 11. 9. N.Y. Const. Art. XVII, § 1. 10. Ca. Wel. Inst. §17000 (2004). 11. Paul Kantor, The Dependent City Revisited: The Political Economy of Urban Development and Social Policy (Boulder, Co.: Westview Press, 1995). 12. Michael Craw, “Overcoming City Limits: Vertical and Horizontal Models of Local Redistributive Policy Making,” Social Science Quarterly 87, no. 2 (June 2006): 361–379. 13. See Mark V. Pauly, “Income Redistribution as a Local Public Good,” Journal of Public Economics 2, no. 1 (Feb. 1973): 35–58, 38. 14. Ibid., 57. 15. Paul E. Peterson, The Price of Federalism (Washington, D.C.: Brookings Institution, 1995), 88–89. 16. Howard Chernick and Paul Sturm, “Redistribution at the State and Local Level: Consequences for Economic Growth,” State Tax Notes, 21 April 2005. 17. See, e.g., the services provided by the Cambridge, Massachusetts, Department of Human Services, at ; the Berkeley, California, Municipal Code contains not only a living wage ordinance, Chapter 13.27.010–.100, but also a rent stabilization ordinance, Chapter 13.76.010–.190. 18. See, e.g., William T. Harbaugh, “What Do Donations Buy? A Model of Philanthropy Based on Prestige and Warm Glow,” Journal of Public Economics 67, no. 2 (Feb. 1998): 269–284; James Andreoni, “Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving,” Economic Journal 100, no. 401 (June 1990): 464–477. 19. Michael Sandel, Democracy’s Discontent (Cambridge, Mass.: Belknap Press, 1996), 131–132. 20. Esteban F. Klor and Moses Shayo, Social Identity and Preferences over Redistribution, CEPR Discussion Paper No. 6406 (Centre for Economic Policy Research, July 2007), (30 June 2009).

2 12

n o t es t o p ag e s 8 7– 8 9

21. See Alberto Alesina and Eliana La Ferrara, “Who Trusts Others?,” Journal of Public Economics 85, no. 2 (Aug. 2002): 207–234; Erzo F. P. Luttmer, “Group Loyalty and the Taste for Redistribution,” Journal of Political Economy 109, no. 3 (2001): 500–528; James Habyarimana, Macartan Humphreys, Daniel N. Posner, and Jeremy M. Weinstein, “Why Does Ethnic Diversity Undermine Public Goods Provision?,” American Political Science Review 101, no. 4 (Nov. 2007): 709–725; Edward Miguel and Mary Kay Gugerty, “Ethnic Diversity, Social Sanctions, and Public Goods in Kenya,” Journal of Public Economics 89, nos. 11–12 (Dec. 2005): 2325–2368. 22. Alberto Alesina, Reza Baqir, and William Easterly, “Public Goods and Ethnic Divisions,” Quarterly Journal of Economics 114, no. 4 (Nov. 1999): 1243–1284. 23. See, e.g., Robert H. Frank, Luxury Fever (New York: Free Press, 1999). 24. See Citizens Budget Commission, Fixing New York State’s Fiscal Practices (2003), 6, (30 June 2009). New York City residents pay local taxes approximately 72 percent higher than the national average, and the city’s generous Medicaid payments account for approximately one-quarter of that difference. It is perhaps difficult to explain this significant deviation from the average as an indication of local preference. 25. Saenz v. Roe, 526 U.S. 489 (1999); Shapiro v. Thompson, 394 U.S. 618 (1969). 26. Helen F. Ladd and Fred C. Doolittle, “Which Level of Government Should Assist the Poor?,” National Tax Journal 35 (Sept. 1982): 323–336, 331. 27. See, e.g., Fifth Ave. Presbyterian Church v. City of New York, 293 F.3d 570 (2d Cir. 2002) (city prohibited from removing homeless persons from church property notwithstanding apparent complaints from church neighbors); Robert C. Ellickson, “Controlling Chronic Misconduct in City Spaces: Of Panhandlers, Skid Rows, and Public-Space Zoning,” Yale Law Journal 105 (March 1996): 1165–1248. 28. Of course, some support may also have been attributable to a desire to remove the homeless—or have them remove themselves—from San Francisco if they could no longer receive cash. That certainly has been one effect of the program. The San Francisco Department of Human Services reported in July 2005 that its County Adult Assistance Program caseload declined 78 percent in the first fifteen months of the Care Not Cash program. See (30 June 2009). But even that effect suggests the willingness of the relatively wealthy to support programs that decrease what they perceive as disamenities. 29. Frances Fox Piven and Richard Cloward, Regulating the Poor (New York: Pantheon, 1971). 30. See Richard C. Fording, “The Conditional Effect of Violence as a Political Tactic: Mass Insurgency, Welfare Generosity, and Electoral Context in the American States,” American Journal of Political Science 41, no.1 (Jan. 1997): 1–29; Joe Soss, Sanford F. Schram, Thomas P. Vartanian, and Erin O’Brien, “Setting the Terms of Relief: Explaining State Policy Choices in the Devolution Revolution,” American Journal of Political Science 45, no.2 (April 2001): 378–396; Elaine B.

n o te s to p ag es 90– 93

31.

32. 33. 34.

35.

36.

37.

38.

39. 40. 41.

21 3

Sharp and Steven Maynard-Moody, “Theories of the Local Welfare Role,” American Journal of Political Science 35, no. 4 (Nov. 1991): 934–950. For my purposes, these motivations remain “benign” notwithstanding that they emerge from selfishness rather than altruism. Recall that I classify explanations for redistribution as benign as long as they are consistent with the interests of residents, regardless of what motivates the formation of those interests. Of course, the evidence here is mixed. Compare Scott Adams and David Neumark, “Living Wage Effects: New and Improved Evidence,” Economic Development Quarterly 19, no. 1 (2005): 80–102, and Scott Adams and David Neumark, “The Effects of Living Wage Laws: Evidence from Failed and Derailed Living Wage Campaigns,” Journal of Urban Economics 58, no.2 (Sept. 2005): 177–202, with Michael Reich, Peter Hall, and Ken Jacobs, “Living Wage Policies at the San Francisco Airport: Impacts on Workers and Businesses,” Industrial Relations 44, no. 1 (Jan. 2005): 106–138, and Mark D. Brenner, “The Economic Impact of the Boston Living Wage Ordinance,” Industrial Relations 44, no. 1 (Jan. 2005): 59–83. Peterson, City Limits, 43. Alberto Alesina and Eliana La Ferrara, “Ethnic Diversity and Economic Performance,” Journal of Economic Literature 43, no. 3 (Sept. 2005): 762–800. Paul Collier, “Ethnicity, Politics, and Economic Performance,” Economics and Politics 12, no. 3 (Nov. 2000): 225–245; Paul Collier, Patrick Honohan, and Karl Ove Moene, “Implications of Ethnic Diversity,” Economic Policy 32, no.16 (April 2001): 129–166; William Easterly, “Can Institutions Resolve Ethnic Conflict?,” Economic Development and Cultural Change 49, no. 4 (July 2001): 687–706. Gianmarco I. P. Ottaviano and Giovanni Peri, “The Economic Value of Cultural Diversity: Evidence from U.S. Cities,” Journal of Economic Geography 6, no. 1 (2006): 9–44. See Edward L. Glaeser and Jesse M. Shapiro, City Growth and the 2000 Census: Which Places Grew, and Why, Brookings Institution Survey Series (May 2001). Jennifer S. Vey and Benjamin Forman, Demographic Change in Medium-Sized Cities: Evidence from the 2000 Census, Brookings Institution Center on Urban and Metropolitan Policy Survey Series (July 2002): 1–17, 6 (30 June 2009). Some confirmation of this can be found in Roberto Suro and Audrey Singer, Latino Growth in Metropolitan America: Changing Patterns, New Locations, Brookings Institution Survey Series (July 2002): 7. Audrey Singer, The Rise of New Immigrant Gateways, Brookings Institution Living Cities Census Series (Feb. 2004): 1, 12–15. Glaeser and Shapiro, City Growth, 12–13. Stephen Moore, “Immigration and the Rise and Decline of American Cities” (Hoover Institution, Aug. 1997), (30 June 2009).

2 14

n o t es t o p ag e s 93 – 98

42. David Card, “How Immigration Affects U.S. Cities,” in Making Cities Work: Prospects and Policies for Urban America, ed. Robert P. Inman (Princeton, N.J.: Princeton University Press, 2009), 158–200. See Albert Saiz, “Room in the Kitchen for the Melting Pot: Immigration and Rental Prices,” Review of Economics and Statistics 85, no. 3 (Aug. 2003): 502–521. 43. Card, “How Immigration Affects,” 188. 44. John M. Quigley, “Urban Diversity and Economic Growth,” Journal of Economic Perspectives 12, no. 2 (Spring 1998): 127–138, 132. 45. See Gilles Duranton and Diego Puga, “Nursery Cities: Urban Diversity, Process Innovation, and the Life Cycle of Products,” American Economic Review 91, no. 5 (Dec. 2001): 1454–1477. 46. Annalee Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Cambridge, Mass.: Harvard University Press, 1994). 47. Quigley, “Urban Diversity,” 131. 48. Ibid., 131–132. 49. See, e.g., Michael Cooper, “Business Group Backs Raise in New York Minimum Wage,” New York Times, 15 July 2004, B3, (30 June 2009). 50. Peterson, City Limits, 43. 51. Quigley, “Urban Diversity,” 132. 52. See Eric Schmitt, “To Fill in Gaps, Shrinking Cities Seek a New Wave of Foreigners,” New York Times, 30 May 2001, A1. 53. See Edward L. Glaeser and Jesse M. Shapiro, “Urban Growth in the 1990s: Is City Living Back?,” Journal of Regional Science 43 (2003): 139–165. 54. Howard Chernick and Andrew Reschovsky, Lost in the Balance: How State Policies Affect the Fiscal Health of Cities, Brookings Institution Center on Urban and Metropolitan Policy (March 2001), (30 June 2009). 55. In theory, it would be possible to address some of these local redistributive needs by collecting redistributive taxes at the central level but allocating them in accordance with formulas that consider local needs. The combination of central funding and decentralized implementation would simultaneously discourage exit from the redistributive burden and permit response to idiosyncratic local needs. These effects, however, must be balanced against the costs of centrally identifying and satisfying local needs. One would anticipate that the process of collecting data and evaluating competing applications for funding could be so burdensome and error prone (in part because the process would be susceptible to politicization, especially if the central and decentralized governments were dominated by different political parties) as to offset losses that arise as a result of local emigration to escape the redistributive burden. The central pool would take on many of the features of a commons, since localities would have made payments and would have incentives to overutilize the pool, indifferent to the effects of their use on needier localities. 56. See, e.g., Masahisa Fujita and Jacques-François Thisse, Economics of Agglomeration: Cities, Industrial Location, and Regional Growth (Cambridge:

n o te s to pag es 98 – 100

57.

58.

59. 60.

61.

62. 63.

64.

21 5

Cambridge University Press, 2002); Glenn Ellison and Edward L. Glaeser, “The Geographic Concentration of Industry: Does Natural Advantage Explain Agglomeration,” American Economic Review 89, no. 2 (May 1999): 311–316. See Andrew F. Haughwout, “Public Infrastructure Investments, Productivity and Welfare in Fixed Geographic Areas,” Journal of Public Economics 83, no. 3 (March 2002): 405–428. Edwin Mills reports that “nearly half of the one million jobs located in the city of Chicago are in its CBD [central business district] and 200,000 are located within a mile of the periphery of O’Hare Airport.” Edwin S. Mills, “Earnings Inequality and Central-City Development,” Economic Policy Review 5, no. 3 (Sept. 1999): 133–142, 135. See Edward L. Glaeser, Cities, Agglomeration and Spatial Equilibrium (New York: Oxford University Press, 2008), 7; Kantor, The Dependent City Revisited, 95. Michael Porter, “Location, Competition, and Economic Development: Local Clusters in a Global Economy,” Economic Development Quarterly 14, no. 1 (2000): 15–34, 16. See, e.g., Peter Dreier, John Mollenkopf, and Todd Swanstrom, Place Matters: Metropolitics for the Twenty-First Century (Lawrence, Kans.: University Press of Kansas, 2005), 35 (noting presence of 380 specialized geographical clusters in U.S. economy); Quigley, “Urban Diversity,” 130–132. See, e.g., Stuart S. Rosenthal and William C. Strange, “Geography, Industrial Organization, and Agglomeration,” Review of Economics and Statistics 85, no. 2 (May 2003): 377–393. Porter, “Location, Competition,” 26. They will not necessarily do so. Craw tested for a relationship between counties with what he described as high monopoly power and redistributive expenditures on public welfare, health care, and community development. He found little support for the proposition that jurisdictions with monopoly power made greater social welfare expenditures than jurisdictions with less monopoly power. But Craw was testing only explicit expenditures, not other forms of redistribution such as tax expenditures or regulatory initiatives. Edward L. Glaeser and Matthew E. Kahn, “From John Lindsay to Rudy Giuliani: The Decline of the Local Safety Net?,” Economic Policy Review 5, no. 3 (Sept. 1999): 117–132, 119. The locality, however, may be more willing to grant concessions in order to attract residents if it is able to externalize some of the costs of compensation. For instance, a locality that issues tax-exempt bonds on behalf of a new firm will not suffer tax collection reductions equal to the benefit conferred on the firm, since the tax exemption typically provides a subsidy from the state and federal governments rather than from the locality. The tax exemption on bonds applies to the interest earned by bondholders. That interest is not included in federal income or state income (where the bond is issued by a locality within the state of the bondholders’ residence). Since few cities have income taxes, the subsidy comes from federal and state sources, not from local ones. This may mean that the locality is engaged in socially inefficient behavior, but the ability to externalize costs does, at least, reduce the negative effects that

2 16

65.

66. 67. 68.

n o t es t o p ag e s 1 0 0 – 1 0 8 ex ante compensation would otherwise impose on the ability of localities to attract potential subsidizers of redistributive taxes. See, e.g., James M. Buchanan and Richard A. Musgrave, Public Finance and Public Choice: Two Contrasting Visions of the State (Cambridge, Mass.: MIT Press, 2000), 179; Vicki Been, “ ‘Exit’ as a Constraint on Land Use Exactions: Rethinking the Unconstitutional Conditions Doctrine,” Columbia Law Review 91, no. 3 (April 1991): 473–545. David Schleicher, “The City as a Law and Economic Subject,” University of Illinois Law Review, 2010, no. 5: 1507–1563. See RUI One Corp. v. City of Berkeley, 371 F.3d 1137 (9th Cir. 2004). See, e.g., 19th Street Associates v. State, 79 N.Y.2d 434 (N.Y. 1992). The prohibition on special legislation which is found in several state constitutions does have an analogue in federal constitutional equal protection analysis. Indeed, some of the living wage ordinances that have been limited in scope to particular industries have been challenged on federal equal protection grounds. See, e.g., Visiting Homemaker Service of Hudson County v. Board of Chosen Freeholders, 883 A.2d 1074 (N.J. App. Div. 2005); RUI One Corp. v. City of Berkeley, 371 F.3d 1137 (9th Cir. 2004).

CHAPTER FIVE. LOCAL POLITICAL MARKETS AND STATE CONSTITUTIONAL CONSTRAINTS

1. See Roderick M. Hills, Jr., “Compared to What? Tiebout and the Comparative Merits of Congress and the States in Constitutional Federalism,” in The Tiebout Model at Fifty, ed. William A. Fischel (Cambridge, Mass.: Lincoln, 2006), 239–263. 2. Compare, e.g., Jonathan R. Macey, “Promoting Public-Regarding Legislation Through Statutory Interpretation: An Interest Group Model,” Columbia Law Review 86, no. 2 (March 1986): 223–268, with William N. Eskridge, Jr., “Politics Without Romance: Implications of Public Choice Theory for Statutory Interpretation,” Virginia Law Review 74, no. 2 (March 1988): 275–338; Frank H. Easterbrook, “Statutes’ Domains,” University of Chicago Law Review 50, no. 2 (Spring 1983): 533–552; Jerry L. Mashaw, Greed, Chaos, and Governance (New Haven: Yale University Press, 1997); Richard A. Posner, “Economics, Politics, and the Reading of Statutes and the Constitution,” University of Chicago Law Review 49, no. 2 (Spring 1982): 263–291. 3. See, e.g., Frank H. Easterbrook, “The Supreme Court, 1983 Term—Foreword: The Court and the Economic System,” Harvard Law Review 98, no. 1 (Nov. 1984): 4–60, 15–18; Cass R. Sunstein, “Interpreting Statutes in the Regulatory State,” Harvard Law Review 103, no. 2 (Dec. 1989): 405–508, 471, 486–487. 4. See, e.g., Sunstein, “Interpreting Statutes,” 471 (“Designed to ensure a kind of deliberative democracy, the constitutional system is hostile to measures that impose burdens or grant benefits merely because of the political power of private groups.”). Some federal constitutional clauses, such as the Takings Clause and the Contracts Clause, are perhaps best understood as warding off interest group grabs in the form of raw majoritarianism. See Richard Epstein, “Toward a Revitalization of the Contract Clause,” University of Chicago Law Review 51, no. 3 (Summer 1984): 703–751, 716–717.

n o te s to pa g es 109– 120

21 7

5. Einer Elhauge, Statutory Default Rules: How to Interpret Unclear Legislation (Cambridge, Mass.: Harvard University Press, 2008); Einer R. Elhauge, “Does Interest Group Theory Justify More Intrusive Judicial Review,” Yale Law Journal 101, no. 1 (Oct. 1991): 31–110. 6. Young v. Kitsap County, 2007 WL 365101 (Wash. App.); Drury v. City of Cape Girardeau, 66 S.W.3d 733 (Mo. 2002); DPR, Inc. v. City of Pittsburg, 953 P.2d 231 (Kan. App. 1998); Seminole County v. City of Winter Springs, 935 So.2d 521 (Fla. App. 2006). 7. Ill. Const. Art. IV, § 13; Best v. Taylor Machine Works, 689 N.E.2d 1057, 1070 (Ill. 1997). 8. Robert S. Amdursky and Clayton P. Gillette, Municipal Debt Finance: Theory and Practice (Boston: Little, Brown, 1990), 90–96. 9. There is a more pessimistic story: the limitations may be driven by public officials themselves who can demand greater contributions from the private beneficiaries of public credit because the debt limitations make it a scarcer resource that is worth more to the few private beneficiaries able to receive it. 10. For judicial recognition of this function of single-subject requirements, see, e.g., Arbino v. Johnson & Johnson, 880 N.E.2d 420, 438 (Ohio 2007); City of Fircrest v. Jensen, 143 P.3d 776, 784–785 (Wash. 2006); Californians for an Open Primary v. McPherson, 134 P.3d 299, 339–340 (Cal. 2006). 11. Chemical Specialties Manufacturers Assn., Inc. v. Deukmejian, 278 Cal.Rptr. 128 (Cal. App. 1991). 12. See Richard B. Stewart, “Madison’s Nightmare,” University of Chicago Law Review 57 (Spring 1990): 335–356. 13. Reza Baqir, “Districting and Government Overspending,” Journal of Political Economy 110, no. 6 (Dec. 2002): 1318–1354. 14. Laura I. Langbein, Philip Crewson, and Charles N. Brasher, “Rethinking Ward and At-Large Elections in Cities,” Public Choice 88, nos. 3–4 (Sept. 1996): 275–293. 15. Robert Inman, “How to Have a Fiscal Crisis: Lessons from Philadelphia,” American Economic Review 85, no. 2 (May 1995): 378–383. 16. Barry R.Weingast, Kenneth A. Shepsle, and Christopher Johnsen, “The Political Economy of Benefits and Costs: A Neoclassical Approach to Distributive Politics,” Journal of Political Economy 89, no. 4 (Aug. 1981): 642–664. 17. David M. Primo and James M. Snyder, Jr., “Distributive Politics and the Law of 1/n,” Journal of Politics 70, no. 2 (April 2008): 477–486. 18. Andrew F. Haughwout and Robert P. Inman, “Should Suburbs Help Their Central City?,” Brookings-Wharton Papers on Urban Affairs (2002): 45–94, 79. 19. Jon Teaford, “Special Legislation and the Cities, 1865–1900,” American Journal of Legal History 23, no. 3 (July 1979): 189–212. 20. Ibid., 207. 21. Robert M. Ireland, “The Problem of Local, Private and Special Legislation in the Nineteenth-Century United States,” American Journal of Legal History 46, no. 3 (July 2004): 277–299. 22. Ibid., 283.

2 18

n o t es t o p ag e s 1 22– 1 3 5

23. See Glenn Abney and Thomas P. Lauth, “The Line-Item Veto in the States: An Instrument for Fiscal Restraint or an Instrument for Partisanship?,” Public Administration Review 45, no. 3 (May–June 1985): 372–377; John R. Carter and David Schap, “Line-Item Veto: Where Is Thy Sting?,” Journal of Economic Perspectives 4, no. 2 (Spring 1990): 103–118; Douglas Holtz-Eakin, “The Line Item Veto and Public Sector Budgets: Evidence from the States,” Journal of Public Economics 36, no. 3 (Aug. 1988): 269–292. 24. See, e.g., Helen Hershkoff, “State Courts and the ‘Passive Virtues’: Rethinking the Judicial Function,” Harvard Law Review 114, no. 7 (May 2001): 1833–1942, 1928; Douglas S. Reed, “Popular Constitutionalism: Toward a Theory of State Constitutional Meanings,” Rutgers Law Journal 30, no. 4 (1999): 871–932. 25. James A. Gardner, Interpreting State Constitutions (Chicago: University of Chicago Press, 2005), 154–164. 26. See G. Alan Tarr, Understanding State Constitutions (Princeton, N.J.: Princeton University Press, 1998), 174–175. 27. Hershkoff, “State Courts,” 1922. 28. See the breakdown of state legislature meeting times and salaries at “Fulland Part-Time Legislatures,” National Conference of State Legislatures Web site, (30 June 2009). 29. See Stephen Ansolabehere and James M. Snyder, Jr., Party Control of State Government and the Distribution of Public Expenditures, MIT Department of Economics Working Paper No. 03–28 (Aug. 2003), (30 June 2009); Hershkoff, “State Courts,” 1928. 30. Edwards v. Aguillard, 482 U.S. 578, 636–637 (1987) (Scalia, J., dissenting) (emphasis in original). 31. Goldstein v. Pataki, 516 F.3d 50, 62–63 (2d Cir. 2008). 32. 488 U.S. 469 (1989). 33. 488 U.S. 495–496. 34. 545 U.S. 469 (2005). 35. City of New York v. State of New York, 730 N.E.2d 920, 926 (N.Y. 2000). 36. Suber v. Alaska State Bond Comm., 414 P.2d 546 (Alaska 1966). 37. Gellert v. State, 522 P.2d 1120, 1123 (Alaska 1974). 38. Short v. State, 600 P.2d 20 (Alaska 1979). 39. North Slope Borough v. Sohio Petroleum Corp., 585 P.2d 534, 546 (Alaska 1978). 40. State v. First Nat’l Bank, 660 P.2d 406, 414 (Alaska 1982). 41. Best v. Taylor Mach. Works, 690 N.E.2d 1057, 1070 (Ill. 1007). 42. See, e.g., Board of Trustees of the Judicial Form Retirement System v. Attorney General, 132 S.W.3d 770 (Ky. 2003). 43. Gillis v. Yount, 748 S.W.2d 357, 359–360 (Ky. 1988). 44. Ibid., 370 (Stephenson, J., dissenting). 45. 892 N.Y.S.2d 8 (N.Y. App. Div. 2009). 46. Goldstein v. New York State Urban Development Corp., 921 N.E.2d 164 (2009). That decision involved the same project with which the federal court that decided the Goldstein v. Pataki case, referred to above, had refused to interfere.

n o te s to p ag es 136– 14 7

21 9

47. 2010 WL 2517686 (N.Y.). 48. Town of Telluride v. Lot Thirty-Four Venture, L.L.C., 3 P.3d 30 (Colo. 2000) (invalidating rent control ordinance under home rule grant); Wagner v. Mayor and Municipal Council of City of Newark, 132 A.2d 794 (N.J. 1957) (invalidating rent control under a home rule grant). Municipal rent control in New Jersey was subsequently upheld in Inganamort v. Borough of Fort Lee, 293 A.2d 720 (N.J.Super. 1972). 49. See, e.g., Idaho Bldg. Contractors Ass’n v. City of Coeur d’Alene, 890 P.2d 326 (Idaho 1995); Eastern Diversified Properties, Inc. v. Montgomery County, 570 A.2d 850 (Md. 1990); Wielepski v. Harford County, 635 A.2d 43 (Md. Ct. Spec. App. 1994), vacated on other grounds, 648 A.2d 192 (Md. 1994). 50. See, e.g., People v. Llewellyn, 257 N.W.2d 902 (Mich. 1977), cert denied sub nom. East Detroit v. Llewellyn, 435 U.S. 1008 (1978). 51. See, e.g., Paul Diller, “Intrastate Preemption,” Boston University Law Review 87, no. 5 (Dec. 2007): 1113–1176. 52. See, e.g., New Mexicans for Free Enterprise v. City of Santa Fe, 126 P.3d 1149, 1156 (N.M. Ct. App. 2005). In 2005, New Jersey explicitly permitted political subdivisions to adopt ordinances setting wage rates higher than those established at the state or federal level. See N.J.S.A. § 34:11–56a4. Thus, the court in Visiting Homemaker Service of Hudson County v. Board of Chosen Freeholders, 883 A.2d 1074 (App. Div. 2005), concluded that the issue of state preemption and local authority had been rendered moot while the appeal was pending. 53. This is the standard statement of conflict between state laws and local ordinances. See Miller v. Fabius Township, 114 N.W.2d 205 (Mich. 1962). 54. RUI One Corp. v. City of Berkeley, 371 F.3d 1137, 1172 (9th Cir. 2004) (Bybee, J. dissenting). CHAPTER SIX. PROXIES FOR DISTINGUISHING BENIGN AND MALIGN REDISTRIBUTION

1. Jerry L. Mashaw, Greed, Chaos, and Governance (New Haven: Yale University Press, 1997), 94–95. 2. Recent court decisions have been more receptive to claims that municipal expenditures on behalf of firms satisfy a public purpose. For a review of the history and evidence of that transition in the context of publicly financing sports stadiums, see, e.g., Haney v. Development Authority of Bremen, 519 S.E.2d 665 (Ga. 1999); Peacock v. Shinn, 533 S.E.2d 842 (N.C. Ct. App.), app. dismissed 546 S.E.2d 110 (N.C. 2000); Ragsdale v. City of Memphis, 70 S.W.3d 56 (Tenn. Ct. App. 2001); CLEAN v. State of Washington, 928 P.2d 1054 (Wash. 1996); Citizens for More Important Things v. King County, 932 P.2d 135 (Wash. 1997); Poe v. Hillsborough County, 695 So. 2d 672 (Fla. 1997); Libertarian Party of Wisconsin v. State, 546 N.W.2d 424 (Wis. 1996). 3. See, e.g., WDW Properties v. City of Sumter, 535 S.E.2d 631 (S.C. 2000). 4. See William H. Simon, The Community Economic Development Movement (Durham, N.C.: Duke University Press, 2001), 105–109; Richard Schragger, “The Anti-Chain

220

n o t es t o p ag e s 1 4 7– 1 4 8

Store Movement, Localist Ideology, and the Remnants of the Progressive Constitution, 1920–1940,” Iowa Law Review 90, no. 3 (March 2005): 1011–1094. 5. Edward L. Glaeser and Matthew E. Kahn, “From John Lindsay to Rudy Giuliani: The Decline of the Local Safety Net?,” Economic Policy Review 5, no. 3 (Sept. 1999): 117–132, 119. 6. The U.S. Census Bureau reports the following statistics for the November 2004 election:

TOTAL Less than $10,000 $10,000 to $14,999 $15,000 to $19,999 $20,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $74,999 $75,000 to $99,999 $100,000 to $149,999 $150,000 and over Income not reported

Percent Reported Registered

Percent Reported Voting

67.2 49.5 49 53.9 58.1 62.9 69.8 75.6 79.4 82.2 82.6 53

60.1 36.5 39.1 45.2 49.4 54.3 62.3 68.1 74.1 77.8 78.3 47.6

Source: U.S. Census Bureau, Current Population Survey, November 2004, Voting and Registration of Family Members, by Race, Hispanic Origin and Family Income: November 2004, Table 9, .

222

n o t es t o p ag e s 1 5 4 – 1 67

22. See Daryl Levinson, “Making Government Pay: Markets, Politics, and the Allocation of Constitutional Costs,” University of Chicago Law Review 67, no. 2 (Spring 2000): 345–420, 347–348. 23. Alberto Alesina, Reza Baqir, and William Easterly, “Redistributive Public Employment,” Journal of Urban Economics 48, no. 2 (2000): 219–241, 220. 24. Ibid.; Glaeser and Shleifer, “The Curley Effect.” 25. See, e.g., James A. Robinson and Thierry Verdier, “The Political Economy of Clientelism,” CEPR Discussion Paper No. 3205 (Centre for Economic Policy Research, Feb. 2002), (1 July 2009); Stephen Coate and Stephen Morris, “On the Form of Transfers to Special Interests,” Journal of Political Economy 103, no. 6 (Dec. 1995): 1210–1235. 26. See Mark Kelman, Strategy or Principle? The Choice Between Regulation and Taxation (Ann Arbor, Mich.: University of Michigan Press, 1999), 75–79; Julie A. Roin, “Reconceptualizing Unfunded Mandates and Other Regulations,” Northwestern University Law Review 93, no. 2 (1999): 351–388. 27. 42 U.S.C. 12101–12122B. 28. See John C. Jeffries, “Disaggregating Constitutional Torts,” Yale Law Journal 110, no. 2 (Nov. 2000): 259–292, 268; Roin, “Reconceptualizing Unfunded Mandates,” 353 n6. 29. Pennell v. San Jose, 485 U.S. 1, 22–23 (1988) (Scalia, J., concurring and dissenting). 30. Saul Levmore, “Interstate Exploitation and Judicial Intervention,” Virginia Law Review 69, no. 4 (April 1983): 563–632. 31. 485 U.S. at 23. 32. Roin, “Reconceptualizing Unfunded Mandates.” 33. This is not to say that any process in which all are represented will necessarily pass muster. If all municipal residents (except landlords) approved the rent control ordinance, I take it that Justice Scalia could still object on the grounds that it constituted a raw majoritarian grab that violated any of a variety of legal constraints, ranging from state uniformity of taxation clauses to the Takings Clause. 34. Roin, “Reconceptualizing Unfunded Mandates,” 358. 35. Russell Hardin, Collective Action (Baltimore, Md.: Johns Hopkins University Press/Resources for the Future, 1980), 72–75. 36. Jeffries, “Disaggregating Constitutional Torts,” 268. 37. See e.g., Idaho Building Contractors Ass’n v. Coeur d’Alene, 890 P.2d 326 (Idaho 1995); New Jersey Builders Assn. v. Mayor and Township Committee of Bernards Township, 528 A.2d 555 (N.J. 1987). 38. See Bruce A. Ackerman, “Beyond Carolene Products,” Harvard Law Review 98, no. 4 (Feb. 1985): 713–746, 719. 39. Visiting Homemaker Service of Hudson County v. Board of Chosen Freeholders, 883 A.2d 1074, 1077 (N.J. App. Div. 2005). 40. RUI One Corp. v. City of Berkeley, 371 F.3d 1137, 1145 (9th Cir. 2004).

n o te s to p ag es 167– 174

223

41. Deanna Bellandi, “Chicago Council Passes ‘Living Wage’ Act,” Washington Post, 27 July 2006, (29 June 2009). 42. See Berkeley City Code § 13.27.070(H), available at . 43. See, e.g., Sheryll D. Cashin, “Localism, Self-Interest, and the Tyranny of the Favored Quarter: Addressing the Barriers to the New Regionalism,” Georgetown Law Journal, 88 (July 2000): 1985–2048. 44. See City of Santa Monica, Santa Monica Municipal Election Information, Election 2002, (1 July 2009). 45. New Mexicans for Free Enterprise v. City of Santa Fe, 126 P.3d 1149, 1156 (N.M. Ct. App. 2005). 46. See, e.g., Bold Corp. v. County of Lancaster, 801 A.2d 469 (Pa. 2002); Ace Rent-A-Car, Inc. v. Indianapolis Airport Authority, 612 N.E.2d 1104 (Ind. Ct. App. 1993); City of New York v. State, 730 N.E.2d 920 (N.Y. 2000); Cavanaugh v. Town of Narragansett, 1997 R.I. Super. LEXIS 21 (R.I. Super. Ct. 1997). The very existence of these cases, of course, reveals that sometimes nonresident litigants who object can be found. That same phenomenon suggests that nonresidents may be virtually represented by residents, such as restaurant owners and garage owners, whose business depends on attracting nonresidents. 47. Clayton P. Gillette and Thomas D. Hopkins, “Federal User Fees: A Legal and Economic Analysis,” Boston University Law Review 67, no. 5 (Nov. 1987): 795–874. 48. See, e.g., Howard Chernick and Olesya Tkacheva, “The Commuter Tax and the Fiscal Cost of Commuters in New York City,” State Tax Notes 25 (5 Aug. 2002): 451–456; R. S. Smith, “Are Nonresidents Contributing Their Share to Core City Revenues?,” Land Economics 48, no. 3 (Aug. 1972): 240–247. 49. Peter Enrich, “Saving the States from Themselves: Commerce Clause Constraints on State Tax Incentives for Business,” Harvard Law Review 110, no. 2 (Dec. 1996): 377–468; John Douglas Wilson, “Theories of Tax Competition,” National Tax Journal 52, no. 2 (June 1999): 269–304; David E. Wildasin, “Interjurisdictional Capital Mobility: Fiscal Externality and a Corrective Subsidy,” Journal of Urban Economics 25, no. 2 (March 1989): 193–212. 50. Matthew Saltmarsh, “France to Lower Tax to Entice Diners to Restaurants,” New York Times, 28 April 2009, (1 July 2009). CHAPTER SEVEN. THE LIMITS OF JUDICIAL INTERVENTION

1. See Stephen J. Choi, Mitu G. Gulati, and Eric A. Posner, Professionals or Politicians: The Uncertain Empirical Case for an Elected Rather Than Appointed Judiciary, University of Chicago Law and Economics, Olin Working Paper No. 357; 2nd Annual Conference on Empirical Legal Studies Paper (Aug. 2007), (1 July 2009); Chris W. Bonneau and

224

n o t es t o p ag e s 1 75 – 1 92

Damon M. Cann, “The Effect of Campaign Contributions on Judicial Decisionmaking” (4 Feb. 2009), (1 July 2009). 2. See, e.g., Mazur v. Trinity Area School District, 961 A.2d 96 (Pa. 2008). 3. See, e.g., Michela Redoano, Does Centralization Affect the Number and Size of Lobbies?, CESifo Working Paper No. 1968 (April 2007), (1 July 2009). 4. Pranab Bardhan and Dilip Mookherjee, “Capture and Governance at Local and National Levels,” American Economic Review 90, no. 2 (Papers and Proceedings of the One Hundred Twelfth Annual Meeting of the American Economic Association, May 2000): 135–139. 5. Redoano, Does Centralization Affect. 6. See, e.g., ibid.; Massimo Bordignon, Luca Colombo, and Umberto Galmarini, Fiscal Federalism and Endogenous Lobbies’ Formation, CESifo Working Paper Series No. 1017 (Aug. 2003), (1 July 2009). 7. See, e.g., Colo. Rev. Stat. Ann. §38–12–301 (“Control of rents by counties and municipalities prohibited”); Mass. Gen. L. Ann. Ch. 40P § 5 (“Because rent control is a matter of statewide concern, this chapter shall preempt, supersede or nullify any inconsistent, contrary or conflicting state or local law”); Wis. Stat. Ann. § 66.1015 (“Municipal Rent Control Prohibited”). 8. See, e.g., American Financial Services Association v. City of Oakland, 104 P.3d 813 (Cal. 2005); American Financial Services Association v. City of Cleveland, 858 N.E.2d 776 (Ohio 2006). 9. The rare exceptions exist in a few states that permit local ordinances to trump state statutes in areas vaguely denominated as “municipal affairs.” See, e.g., Calif. Const. art. XI, § 5(a); Colo. Const. art. XX, § 6. 10. Einer Elhauge, Statutory Default Rules (Cambridge, Mass.: Harvard University Press, 2008), 268. 11. Ibid., 274. 12. I ignore for purposes of this example the difficult issue of what it would mean for the rich and the poor to receive “pro rata” shares of the budget. That is, should the budget be allocated according to tax revenues paid (so that the wealthy receive a greater than per capita share of municipal services) or should the budget be allocated according to need (so that the poor receive a greater than per capita share of municipal services). I assume only that the parties have agreed on the allocation that exists. 13. See, e.g., City of East Orange v. Board of Water Comm’rs, 191 A.2d 749 (N.J. Super.), aff’d 194 A.2d 459 (1963). 14. Northern Palm Beach County Water Control Dist. v. State, 604 So.2d 440 (Fla. 1992).

INDEX

Ackerman, Bruce, 166 ACORN, 1 Adams, Scott, 56 Agenda-setting monopolists, 18 Agglomeration effects as constraint on exit, 98–102, 107, 147, 173 AIDS, 96 Alaska Supreme Court and single-subject requirement for legislation, 133 Alesina, Alberto, 91–92 Almshouses, 61–65, 68 Almy, Frederic, 65 Altruism, xi, 11, 64, 75, 78, 81, 85, 87, 89 Alumni donations, 85 Ambiguity: about municipal autonomy, 104; in state constitutions, 111; in statutory interpretation, 108 Appropriations. See Budget and expenditures Asian population growth in U.S. cities, 92–93 Baby Health Stations (NYC), 68 Baltimore: poor relief programs in, 62, 64; public baths in, 68 Baqir, Reza, 116

Bardhan, Prahan, 180 Barrow, Lisa, 149 Battered women’s shelters, 85 Benefit-based payment mechanisms, 20, 54–55, 67 Benign vs. malign redistribution, 72–75, 142–173; balancing malign motives in enacted and defeated programs, 192–195; benign explanations for redistribution, 10–11, 43, 80–82; concentration of benefits and diffusion of costs in malign redistribution, 143–144; constraints on exit favoring benign redistribution, 98, 101; and defective political process, 50; difficulty in distinguishing between, 143, 197; externalized effects of malign redistribution, 170; judicial intervention to determine, 104, 105, 109, 140, 177–178; malign explanations for redistribution, 11–12, 40–41, 80, 102–105; minorities suffering due to malign redistribution, 163–170; and off-budget expenditures, 154–155; and transparency in decision making, 14, 151

225

226

in d ex

Berkeley, California: living wage ordinance in, 6, 7, 141, 164, 168; welfare programs in, 85 Bicameral legislatures, 181, 195 Big-box stores: ordinance for employee living wage (Chicago), 1–3, 7, 20, 146, 166, 167–168, 173; subsidies to attract, 53, 146–147 Boston: immigrant programs of, 95; poor relief programs in, 62, 63, 64; public baths in, 68 Brasher, Charles N., 116 Bribery, x, 80 Briffault, Richard, 48 Brooklyn, New York, poor relief programs in, 64–65 Brueckner, Jan, 37 Budget and expenditures: constitutional constraints on, 27, 105, 110, 112; implicit local bargains on what to fund, 191; internalized vs. externalized costs, 47–50, 170–173; local officials able to target for favored projects, 18; off-budget expenditures, 152–155; overexpenditures on the poor, 150; trade-offs in, 45–46; and transparency, 151–157 Burke County, Georgia, poorhouse funding in, 63 Bybee, Jay, 141 California: poor relief programs in, 82; Proposition 105 (“People’s Right to Know Act”), 114–115 Capture, vulnerability to, 179–181, 184 Card, David, 93–94 Case or controversy, not required for state court jurisdiction, 123 Census of Governments, 10 Census (2000) on immigrant statistics, 92–93 Centralized governments: costs of influencing, 181–182; involvement in review or approval of local redistributions, 103, 175–176; and lobbying, 180; Madison as advocate for, 23–26; reasons to advocate for, 170;

and special interests, 179; and urban finance theory, ix Centralized redistribution, 96 Charitable organizations, 64, 66 Chernick, Howard, 84, 96 Chicago: big-box ordinance for employee living wage in, 1–3, 7, 20, 146, 166, 167–168, 173; public baths in, 68; special legislation in, 120 Chicago Federation of Labor, 3 Children’s Aid Society, 64 Churches as surrogates for the poor, 145–146, 185 Cincinnati, poor relief programs in, 66 City Limits (Peterson), 54 Cleveland: budget transparency of, 152; poor relief programs in, 66 Cloward, Richard, 89 Coase, Ronald, 39 Collective action problems, 15–23, 58; and centralized governments, 126; distortion caused by collective action, 51; and idiosyncratic interests of community representatives, 198; interfering with local political decisions, 196; mandates as way of solving, 156–157; minimizing incentives for single residents to resist political process, 113; and monitoring and transparency of expenditures, 159–160; opponents of incumbents, ability to overcome, 152; redistributive activities blocked by, 55; voice and transmission of signals via collective action, 46–47 Collective bargaining and living wage ordinance, 3, 168 Commerce Clause, 111 Committees, state legislative, 181 Community altruism and community building, 85–86, 88 Commuter taxes, 97, 101, 132, 171–173 Competition: advantages from redistributive programs, 96–97; among interest groups at local level, 23–28, 121; among local governments, 40, 90, 172

index

Competitive bidding, 152 Condominium conversions, 137 Conflict of laws, state vs. local, 5, 26, 69–70. See also Preemption Connecticut, poor relief programs in, 82 “Conscious funding,” 159 Constitutional constraints, 106–141; ineffectiveness of, 122; and judicial intervention, 105, 122–123, 137; and living wage ordinances, 6; and local autonomy, 13, 70, 108; and special interests, 27, 29; state vs. federal, 27, 110–111. See also specific types of constraints (e.g., Special legislation prohibitions, Single-subject requirements) Contractors and living wage ordinances, 3–4, 165, 167 Corruption, 80, 118, 180 Costs. See Budget and expenditures Council-manager form of government, 116 Countervailing forces: cost reductions to facilitate, 198; and high costs of opposition, 182; and logrolling, 116; Madison’s description of, 24–26, 139; in multiple localities, difficulty and cost of organizing, 184; necessary to benign decision making, 187; and state legislatures, 179; subsidizing formation of, 197–198 Courts, state: ability to constrain interest groups, 15, 51–52, 107; accessibility to individuals and small interest groups, 175; biases of, 174; as checks on redistributive programs, 13, 104, 107; constitutional constraints imposed by, 105, 122–123; “correction” of local redistribution programs by, 189; deference to legislatures, 124, 127, 132, 135–136; determining local autonomy, 6–7, 26–27; distorting balance of distribution in society at large, 190, 192; doctrinal differences from federal courts, 110, 123, 132; failure to reflect true views of polity, 188; individual rights, intervention to

227

create and protect, 123; judges, election of, 123, 126, 174; judicial review and redistributive mix, 13, 190–192; limits of judicial intervention, 15, 174–195; mediating between urban finance theory and public choice theory, xi; negative effect in reducing benefits of goods and services, 77; partial invalidation of legislation by, 117–118; and redistribution, xii, 12, 15, 189–190; reversing court decisions through legislation or administrative regulation, 176. See also Reverse engineering of local legislation Craw, Michael, 82–83 Crewson, Philip, 116 Crime, social programs to discourage, 89–90 Curley, Michael, 79 Daley, Richard M., 2–3 Debt limitations on local governments, 27, 110, 112–114, 122, 133 Decentralized governments: and local autonomy, 13–14, 31–32; and special interests, x, 25–26, 139, 179; and urban finance theory, ix Decision-making responsibility: allocation of, 12; development of structure for, 186–187; in localities where some activities are subsidized by residents and nonresidents, 7; reflecting local constituents’ interests, x; subjects appropriate for local decision making, 27–28. See also Local autonomy Defalcation, x, 102–103 Default of local governments, 68–69, 110, 112 Defeated legislation, court’s inability to review, 190, 192–195 Defense contractors, 76 Deference of courts. See Judicial deference Delaware, poor relief programs in, 63 Democracy: and economic interests, 28–30; and local autonomy, 31–32, 45–47; and monitoring of officials, 32

228

in d ex

Democracy and Distrust (Ely), 27 Depression of 1893, 67 Developers: condemnation of blighted area for development by Columbia University, 135–136; impact fees paid by, 163; likelihood of investing in relationships with local officials, 180; tax breaks for, 77 Dexter, Ebenezer Knight, 63 Dillon’s Rule, 69 Diversity. See Ethnic/racial diversity and inclination for local redistribution Economic crisis, government response to, 67–69 Economic development: investments in poor for, 90; legislatures overriding judicial invalidation of initiatives for, 185; local subsidies for, 70, 147; statistics, 53; wealthy receiving subsidies for, 146 Economic impact statements, 198, 199 Economic theories of local government, 15–17, 70. See also Orthodox view of redistribution Economies of scale, 94 Educational expenditures. See School expenditures Einhorn, Robin, 50, 66 Elections: campaign contributions for state office seekers, 181; challengers to incumbents, incentives to find financial malfeasance of office holders, 152; results considered as signal of constituents’ wishes, 17–18; of state judges, 123, 126, 174 Elhauge, Einer, 108–109, 174, 186–189 Ellickson, Robert, 149 Ely, John Hart, 27 Eminent domain, 7, 70, 125–126, 129–132, 144. See also Takings Clause English Poor Laws, 60, 64, 66 Entertainment taxes, 170 “Equal political influence” (Elhauge), 187 Equal Protection Clause, 6, 128 “Equal services” doctrine, 44

Erie County, New York, poor relief programs in, 65 Eskridge, William, 30 Ethnic/racial diversity and inclination for local redistribution, 86–87, 91–92 Exactions on nonresidents, 170–173 Exclusivity, grants of, 60 Executive line-item vetoes, 110, 122 Exit, 96–102; agglomeration effects as constraint on, 98–102, 107, 147, 173; of disaffected residents, 8–9, 126; immobile capital of firms preventing, 98; and redistribution, 8–9 Expenditures. See Budget and expenditures Exploitation of situational monopoly for benign redistribution, 98 Extraterritorial effects of local redistribution, 49–50, 97 Fear of the poor as reason for redistribution, 89–90 Feasibility studies, 199 Federal grants and funding, 58 Federal institutional design, compared to state and local counterparts, 109, 126, 132 Federalism, 179 Federalist: No. 10 (Madison), 23, 115, 139, 179; No. 15 (Madison), 29 Fiscal capacity and redistribution, 42–43 Fischel, William, 34–37 Forman, Benjamin, 92 Forum choice as determining outcome of living wage proposals, 6 Fourteenth Amendment. See Equal Protection Clause Fractionalization, 92 Free riding, 16–17, 21, 22, 33, 76, 100, 101 Gardner, James, 123, 124 Geographic proximity and clustering of firms, 98–99. See also Agglomeration effects as constraint on exit Glaeser, Edward L., 79, 92, 93, 96, 147–148

index

Global mix of redistribution, 191–192, 195 Golf courses, 54, 191 Grassroots Collaborative, 1 Grassroots movements, 184 Great Depression, 68–69 Green Party of Santa Fe, 4 Hall, Peter, 56 Handlin, Oscar and Mary Flug, 59–60 Hartog, Hendrik, 59 Haughwout, Andrew, 117, 149 Hershkoff, Helen, 124 Heterogeneous preferences of population, 32–33, 45, 88, 179, 180 Hirschman, Albert, 46 Hispanic population growth in U.S. cities, 92–93 Home Depot, 2 Homeless, shelters and support services for, 54, 57, 89, 150 Home rule municipalities, 5, 70, 138–140, 186 Homogenous preferences of population, 32–33, 40–41, 43, 88, 179 Hotel and parking taxes, 97, 170 Housing: avoiding deterioration of, 95–96; immigration’s effect on, 93, 95; ordinances, 49. See also Rent control ordinances Hoxby, Caroline, 149 Hudson County, New Jersey, living wage ordinance, 167 Humanitarian relief organizations, 64 Hume, David, 29 Illinois special legislation, interpretation of constitutional provision, 112, 133 Immigrants: as beneficiaries of minimum wage increases, 78; and short-term redistributive assistance, 91; wages and economic performance associated with increased numbers of, 92–93 Immobility, pros and cons of, 98–102. See also Agglomeration effects as constraint on exit

229

Impact fees paid by developers, 163 Implicit local bargains on what to fund, need for courts to recognize, 191 Income taxes, local, 10, 191 Individual rights, state judicial intervention to create and protect, 123 Individuals: courts accessible to, 175; difficulties for single residents to resist political process, 113; incentives to cooperate among, 16; influence on local government, 32; local governments accessible to, 22, 32; mobility of, 100–101. See also Public preferences Indoor relief for the poor, 61 Industrial clustering, 99. See also Agglomeration effects as constraint on exit Inman, Robert, 116, 117, 149 Interest groups, x; competition at local level, 23–28, 121; constitutional clauses aimed at constraining effects of, 123; influence of, as malign redistribution, 14–15, 51–52, 80; judicial oversight of legislative role of, 105, 108–109, 133–134; landlords as, 22, 161, 162, 182–183; litigation campaigns by, 175; Madison’s views on, 23–26; majority vs. minority interests, in Croson case, 127–129, 132; motivations of, not necessarily for public good, 19; size of and regulatory mandates, 157–170; socially beneficial dominant, 75–80; and special legislation, 118–119; state legislatures susceptible to influence of, 124; susceptibility of local and state government to, 121, 177–189. See also Lobbying; Neutralization of special interests Intergovernmental transfers and redistributive spending, 83 Ireland, Robert, 118, 119–120 Jacobs, Ken, 56 James County, Virginia, relief for the poor, 62 Judges, state election of, 123, 126, 174

230

in d ex

Judicial abstention, 131 Judicial deference, 124, 127, 132, 135–136 Kahn, Matthew E., 147–148 Kantor, Paul, 82–83, 98 Katz, Michael B., 64, 65, 67, 68 Kaur v. New York State Urban Development Corp. (2009), 135–136 Kelo v. City of New London (2005), 129–132 Kentucky: poor laws in, 61; tax on unmined coal in, 134–135 King County, New York, poor relief programs in, 65 Klebaner, Benjamin Joseph, 63–64 Klor, Esteban, 86 La Ferrara, Eliana, 91–92 Landlords. See Rent control ordinances Langbein, Laura I., 116 Law of 1/n, 116–117 Lawsuits. See Courts, state; Living wage ordinances Legislative committees, state, 181 Legislatures, state: amateur nature of and susceptibility to special interests, 124–125; costs of lobbying, 181; judicial deference to, 124, 127, 132, 135–136; localized programs inappropriate for state action, 182; reversing court decisions through legislation, 176, 179, 184; size, relationship with project efficiency (Law of 1/n), 116–117 Levinson, Daryl, 154 Levmore, Saul, 159 Liberal humanitarianism, 60 Living wage ordinances, 1–3, 164–169; in conflict with state law, 26; differences in scope of proposals for, 7, 20, 164– 166; as exemplars of local autonomy and decision making, 14; forum choice as determining outcome of proposals for, 5; funding of, 151; invalidated as preempted by state minimum wage law, 178; lawsuits over, xiii, 5, 6–7, 13, 14, 103, 141; as “municipal affair” within scope of home rule, 140; as

off-budget expenditures, 155; and Peterson’s views, 55; as redistributive mechanisms, 10, 77–80, 148; as regulation to achieve redistribution, 42–43, 157; and social disturbance theory, 90; state legislatures permitting, 50; and unemployment rates, 56; and urban finance theory, 7; wealthy opposition to, 146. See also specific cities Lobbying, 21, 175, 180–181, 184 Local autonomy, 31–52; and allocation of authority between state and local governments, 138–139, 177, 179; and decentralized governments, 13–14, 31–32; and externalization of costs, 47–50; and internal political processes, 50–52; and judicial determinations, 6–7, 12–13, 175; and local democracy, 14, 45–47; and local political processes, 12, 26, 47; and redistribution policies, 11, 26, 177; scope of, 31; state statutes conferring and keeping approval authority, 69, 175–176. See also Constitutional constraints Local competence, 1–30 Local democracy. See Democracy Local economic interest, 90–96 Local government officials: mismanagement by, 51; motivations of, 19, 21–22, 48, 106, 148–149 Local legislatures: part-time nature of, 124; paucity of information on process of, 124, 199–200 Local political processes: inadequacy of, 142; and local autonomy, 47; and redistributive programs, 13, 50–52, 108 Local redistribution. See Redistribution “Local Redistribution, Living Wage Ordinances, and Judicial Intervention” (Gillette), xii Logrolling approach to legislation, 108, 114–118, 149–150 Lotteries for poorhouse funding, 63 Louisiana, poor relief programs in, 64 Low, Seth, 64–65

index

Low-income workers: as beneficiaries of minimum wage increases, 78; reasons for cities to attract, 95. See also Living wage ordinances Machiavelli’s use of “corruption,” 80 Madison, James: advocating centralized decision making, x, 23, 179, 181, 182; description of countervailing forces’ role, 24–26, 115, 139, 198; on need for government and government controls, 29 Malign redistribution. See Benign vs. malign redistribution Mandated redistributions: advantages of, 155–157; regulatory mandates and interest group size, 157–170 Market failure problem, local redistribution as answer to, 40, 44, 55, 66, 84, 103, 188–189 Mashaw, Jerry, 144, 145 Massachusetts: history of private-public development in, 59–60; poor relief programs in, 62, 63 Maynard-Moody, Steven, 41–42, 84 Medicaid, 82, 87 Michigan, unemployment rates in, 67 Migration: individuals migrating to communities that provide valued services or to avoid supporting services they do not value, 37, 44, 101, 188; redistribution as cause of, 8–9, 40, 53, 54, 73; as signals to local officials, 34–36, 38, 46. See also Exit Mill, John Stuart, 51 Millers, 60 Milwaukee, poor relief programs in, 66 Minimum wage, 5, 69, 161, 178. See also Living wage ordinances Minority businesses and set-asides, 127–129 Misinterpretation of public’s preferences, 18–19, 21, 74 Mismanagement by local officials, 51 Missouri living wage ordinance litigation, 6

231

Modern urban welfare programs, 68 Mookherjee, Dilip, 180 National Public Welfare League, 68 Neumark, David, 56 Neutralization of special interests: constitutional signals for, 139; legal system’s objective as, 12, 14, 27, 29, 143, 197; Madison on, 23 New England township system and poor relief, 32, 61 New Hampshire poor relief programs, 62 New Jersey living wage ordinance litigation, 6 New London eminent domain case, 129–132 New Mexico living wage ordinance litigation. See Santa Fe, New Mexico, living wage ordinance New Mexico Restaurant Association, 4 New Orleans: living wage ordinance in, 6–7, 13, 14, 49, 140; poor relief programs in, 64, 66; returning residents after Katrina expressing dissatisfaction with services in, 38 New York (state): Medicaid allocations in, 82, 87; poor relief programs in, 62, 63; unemployment rates in, 67 New York City: budget transparency of, 152; commuter tax, state law to prevent, 132; condemnation of blighted area for development by Columbia University, 135–136; eighteenth-century government of, 59; eminent domain case, review rejected by federal Second Circuit, 125–126; immigrant programs of, 95; Local Law 11 (1998) for building inspections, 155–156; Medicaid funding in, 87; poor laws in, 61; poor relief programs in, 63, 64, 65–66, 68; poor tax on real estate in, 62; public baths in, 68; waterworks in, 59; Yankee Stadium, bonds issued to build, 153 New York Times on residents returning to New Orleans after Katrina, 38

232

in d ex

Nonresident fees, 170–173. See also Commuter taxes; Hotel and parking taxes Northwestern University Law Review, xii Notice of redistributive proposals, 198 Oates, Wallace, 37 Ohio poor laws, 61 On- and off-budget expenditures, 150–157 Orthodox view of redistribution: application to the wealthy, 53; and benign local redistribution, 81; and competition of local governments, 97; denial of desirability of local redistribution, 72; failure to recognize unique opportunities, 96, 196; as pure wealth transfer, 90; on rarity of local redistribution and its inconsistency with local welfare, 10–11, 50, 94; relationship to legal doctrine, 69; and relocation of dissatisfied residents, 20–21; sameness among policies as part of, 84; and Tieboutian model, 41 Ottaviano, Gianmarco I. P., 92 Outdoor relief for the poor, 61, 64–65 “Overdevelopment” as price of developers’ subsidies, 77 Overweight vehicles, tax on, 170 Paris restaurant tax, 173 Parking taxes. See Hotel and parking taxes Pauly, Mark, 83 Paving. See Road paving Pennsylvania poor relief programs, 62; and county system, 61 Pepperell, Massachusetts, poor relief program, 63 Peri, Giovanni, 92 Peterson, Paul, 42, 54–55, 74, 84, 90 Philadelphia: immigrant programs of, 95; increase in number of local legislators, effect of, 116; poor relief programs in, 61, 62, 64; special legislation in, 119 Philanthropy, 85–86 PILOTs (payments in lieu of taxes), 153 Pingree, Hazen, 67

Piven, Frances Fox, 89 Plenary power of states, 5 Political party competition, 184 Pollin, Robert, 56 The poor: defeat of redistribution programs intended to help, 193; fear of, 89–90; housing for, 70; local redistribution to, 50, 54, 147–152, 185; public relief programs, 60–69, 82 Poorhouses, 61–65, 68 Poor Laws, 60–62 Poor localities’ approach to redistribution, 43 Populists, 66 Porter, Michael, 99–100 Poverty services industry, 149 Predatory loan ordinances, 184 Preemption, 5, 13, 48, 69–70, 138, 178 Preferences of public. See Public preferences Prison Association, 64 Private-public alliances, 59, 68 Privilege, grants of, 60 Progressives, 68 Property taxes, 10, 32, 35, 68, 113, 171. See also Tax abatements Providence, Rhode Island, poor relief programs in, 63, 64 Public baths, 68 Public choice theory, ix–x, 155 Public goods and services: dissatisfaction with, 43–44; equal provision to all, 44; and ethnic diversity, 87; function of government to provide, 15–17; impossibility of disaggregating, effect of, 39; and publicly interested legislation, 75–76; residents’ willingness to pay for, 34–36, 41, 43, 67 Public preferences: diversity in, 28, 39, 45, 73; homogenous vs. heterogeneous, 32–33, 40–41, 43, 45, 88, 179, 180; and living wage proposals, 78; misinterpretation of, 18–19, 21; municipal services’ need to match, 101, 187; political bargaining obscuring, 196; and transparency of redistributive

index

programs, 151; willingness to pay as indication of, 35–36, 67, 85, 160, 165. See also Interest groups; Sorting theory Public purpose doctrine, 26, 69, 105, 110, 111–112, 177, 192 Quigley, John, 94–95, 100 Racial differences and inclination for local redistribution, 86–87, 91–92 Railroads, 59, 112 Redistribution, xi, 53–71; advantages of mandated redistribution, 155–157; and agglomeration effects, 98–102, 147; defined, 57; disincentives for, implementation in spite of, 11; experimental literature on, 86; explanations for, 10–11, 80–90; and externalization of costs, 47–50, 170–173; and global justice, 189–190; and global mix, 191–192; and history of poor relief, 58–69; and interest group size, 157–170; and internal political processes, 13, 50–52; and local economic interest, 90–96; and local welfare, 10; malign explanations for, 11–12, 40–41, 102–105; municipal employment as, 154; on- and off-budget expenditures, 150–157; self-defeating results from, 9; and socially beneficial dominant interest groups, 75–80; and validity of decisions, 56–57; variations among localities for, 41–42, 53–54; to wealthy, 7, 53–54, 80, 85, 145–147, 184–185. See also Benign vs. malign redistribution Redoano, Michela, 180 Regulatory mandates and interest group size, 157–170 Reich, Michael, 56 Relocation. See Exit; Migration Rent control ordinances: lack of municipal authority to promulgate, 69, 137; landlords incentivized to act by, 22, 182–183; as poor relief, 10; requiring subgroup to fund redistribution of,

233

151; state enacting statute to prohibit, 178, 186; state level debate of, 104; as uncompensated taking, 158–162 Residential preferences. See Migration: as signals to local officials; Public preferences Reverse engineering of local legislation: to determine countervailing groups’ role, 185; and judicial limitations, 105, 123, 125, 126–127, 197; and living wage ordinances, 14; and partial invalidation of legislation, 117; state courts’ authority for, 15 Richmond, City of v. J.A. Croson Co. (1989), 127–129, 132 Ripeness, 123 Road paving, 16, 54, 57 Roin, Julie, 159, 161 Romer, Thomas, 18, 76 Rosenthal, Howard, 18, 76 Rouse, Cecilia Elena, 149 St. Louis living wage ordinance, 6 San Francisco: homeless shelters and support services in, 89; International Airport’s living wage policy, 56; welfare programs in, 85 Santa Fe, New Mexico, living wage ordinance: community debates over, 23, 166, 167; litigation over, 5, 6, 13, 14, 167; scope of, 4–5, 7, 20, 167, 171; and unemployment rates, 56; wealthy opposition to, 146 Santa Monica, California: living wage ordinance in, 3–4, 7, 20, 146, 164, 166, 168, 170–171; welfare programs in, 85 Savannah, Georgia, poorhouse funding in, 63 Scalia, Antonin, 125, 127, 132, 158–162 Schleicher, David, 101 Schneider, Mark, 40 School expenditures, 57–58, 149 Scientific charity, 64 Self-interest of political actors, 29, 187 Separation of powers, 123 Set-asides, 127–129

234

in d ex

Shapiro, Jesse M., 92, 93, 96 Sharp, Elaine B., 41–42, 84 Shayo, Moses, 86 Shleifer, Andrei, 79 Single-subject requirements, 27, 105, 110, 111, 114–115, 118, 133, 138 Slum eradication, 70 Social disturbance theory, 89–90 Social identity and local redistribution, 86–87 Socioeconomic heterogeneity, benefits of, 94–95 Sorting theory, 40–42, 45, 47, 84 Spatial theory of redistribution (Pauly), 83–85 Special assessments, 67 Special commissions, 105, 110 Special interests. See Interest groups; Neutralization of special interests Special legislation prohibitions, 27, 105, 110, 111–112, 118–121, 133, 177 Spillover effects, 57–58, 81, 94, 97, 170 Sports stadiums, 7, 53, 55–56, 57, 94, 145, 153, 199 Standing, 123 State courts. See Courts, state State grants and funding, 58 State legislatures. See Legislatures, state State preemption. See Preemption Statutory ambiguity, 108 Stevens, John Paul, 131–132 Stewart, Richard, 115 Sturm, Paul, 84 Subsidies: arguments for, 53, 59; for economic development, 70, 77, 81; and local autonomy, 11; presumption of legislative balancing of costs and benefits of, 159; protection from market competition as, 60; redistributions as, 7; wealthy more likely to benefit from, 192. See also Redistribution; Tax abatements Suburban firms relocating outside redistributive jurisdiction, 97 Supreme Court, U.S., 127–132; on eminent domain (Kelo), 70, 129–132;

on minority businesses and set-asides (Croson), 127–129, 132 Surrogates: for disenfranchised in debate over nonresident exactions, 173; unions as, for the working poor, 145–146, 149, 169, 193–194 Susceptibility of local and state government to interest groups, 177–189 Takings Clause, 104, 163 Target Corp., 2 Tax abatements, 21, 54, 60, 77, 152–153, 155, 183, 192–193 Tax avoidance by leaving redistributive municipalities, 8 Taxes, local, imposition of, 18, 67, 75, 97, 170, 191. See also specific types of taxes Taxes, state: and level of redistributive services, 84; special tax benefits, Ky. case on unmined coal tax, 134–135 Tax exemptions, 7, 10 Teaford, Jon, 118–119 Telluride, Colorado, housing ordinance, 49 Tennessee county system and poor relief, 61 Tieboutian model, 34–41, 43, 44, 46, 73, 74, 101, 188 Tippecanoe County, Indiana, poor relief program, 62 Tourists, taxes on, 97, 170–172 Town meetings, 32 Trade-offs, 33, 45–46, 91–92, 125, 191 Transparency, 14, 151, 160 Transportation access as inducement for firms to stay in municipalities, 98 Trounstine, Jessica, 18 Unemployment rates, 56, 67, 68, 96 Unfunded mandates, 27, 82, 105 Uniform Commercial Code, 111 Uniform Land Sales Practices Act, 133 Unions: and living wage ordinances, 3–4, 78–79, 165; as surrogates for the working poor, 145–146, 149, 169, 185, 193–194

index

Urban finance theory, ix–xii; and living wage ordinances, 7; and redistributive functions of government, 8–10, 68. See also Orthodox view of redistribution Urban riots of 1960s, increase in social welfare spending after, 89 Utility franchises, 60 Vagueness, 6 Vermont township system and poor relief, 61 Vetoes, gubernatorial line-item, 110, 122 Vey, Jennifer S., 92 Voting: considered as signal of constituents’ wishes, 17–18, 73; to

235

subsidize projects of special interests, 113. See also Elections Wal-Mart, 2, 168, 173 Ward-based city politics: logrolling encouraged by, 149–150; replacing with at-large politics, 117 Waterworks, 59 Wealthier localities’ approach to redistribution, 42 “Welfare magnets,” 50 Women’s shelters, 85 Zoning ordinances, 85, 97, 130, 170, 183