Information and Exclusion 9780300171693

Citation preview

INFORMATION AND EXCLUSION

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

information and Lior Jacob Strahilevitz exclusion

New Haven & London

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

Copyright © 2011 by Lior Jacob Strahilevitz. 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 Galliard and Copperplate 33 types by IDS Infotech, Ltd. Printed in the United States of America. Library of Congress Control Number: 2011922304 ISBN 978-0-300-12304-3 (alk. paper) 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 iris and eli

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

CONTENTS

Acknowledgments  ix Introduction  1 PART I: THE MECHANICS OF EXCLUSION   1 Three Mechanisms for Exclusion  11   2 The Bouncer’s Right  28   3 Exclusionary Vibes  42   4 Exclusionary Amenities  55 PART II: A THEORY OF INFORMATION AND EXCLUSION   5 Asymmetric Information and Exclusion  75   6 Regulating the Choice Among Exclusion Strategies  93 PART III: DIVERSITY-PROMOTING STRATEGIES FOR THE TWENTY-FIRST CENTURY   7 Bundled Amenities to Reduce Discrimination  113   8 Information Is a Variable, Not a Constant  127

vii

viii

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

CONTENTS

  9 Carrots, Sticks, Curtains, and Searchlights  157 10 Winners and Losers  173 Conclusion  193 Notes  199 Index  249

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

ACKNOWLEDGMENTS

I began working on the project that would become this book in 2004. The property theory that forms the core of Parts I and II was conceived over the course of a few weeks spent thinking and using colleagues as sounding boards. Several years passed before I started thinking productively about the effects that new technologies might have on racial and other forms of discrimination, and this research formed the basis for Part III of this book. Part III is about information privacy and exclusion at its core, but the three parts together try to advance the argument that property theory has much to say about privacy, and that privacy theory can teach us much about privacy. The book you see before you is therefore substantially different from the manuscript I thought I was starting on in 2004, and hopefully the better for it. The book draws on several previously published law review articles: Exclusionary Amenities in Residential Communities, 92 Virginia Law Review 437 (2006) (chapters four and seven); Information Asymmetries and the Rights to Exclude, 104 Michigan Law Review 1835 (2006) (chapters one, two, three, five, and six); Privacy versus Antidiscrimination, 75 University of Chicago Law Review 363 (2008) (chapter eight); and Reputation Nation: Law in an Era of Ubiquitous Personal Information, 102 Northwestern University Law Review 1667 (2008) (chapters eight, nine, and ten). There are countless people who deserve my hearty thanks for their assistance and guidance. Scholars who provided me with useful commentary on earlier drafts include Noah Zatz, Tal Zarsky, Albert Yoon, Tim Wu, David ix

x

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

ACKNOWLEDGMENTS

Weisbach, Amy Wax, Cass Sunstein, Kathy Strandburg, Geof Stone, Henry Smith, Peter Siegelman, Paul Schwartz, Alan Schwartz, Peter Schuck, Mike Schill, Max Schanzenbach, Allen Sanderson, Adam Samaha, Daria Roithmayr, David Reiss, Eric Posner, Ariel Porat, Randy Picker, Eduardo Peñalver, Jide Nzelibe, Jonathan Nash, John Nagle, Hiroshi Motomura, Beth Milnikel, Tom Miles, Tracey Meares, Miranda McGowen, David McGowan, Richard McAdams, Yoram Margalioth, Anup Malani, Doug Lichtman, Saul Levmore, Ron Lee, Sarah Lawsky, Doug Kysar, Eugene Kontorovich, Jerry Kang, Christine Jolls, Uri Itkin, Todd Henderson, Scott Hemphill, Bernard Harcourt, Henry Hansmann, Jeff Gordon, Eric Goldman, Nevin Gewertz, Heather Gerken, Nicole Garnett, Oscar Gandy, Owen Fiss, Lee Fennell, Richard Epstein, Liz Emens, Adam Cox, Howard Beales, Will Baude, Shyam Balganesh, Ian Ayres, Ronen Avraham, Amitai Aviram, Theo Angelis, and Michael Abramowicz. I also benefited greatly from presenting early drafts of chapters at the University of Chicago, Yale, UCLA, the University of Virginia, the University of San Diego, Notre Dame, Northwestern, New York University, George Washington University, Fordham, Indiana University–Indianapolis, UC Berkeley, and the annual meetings of the American Law and Economics Association and the Midwestern Law and Economics Association. Michael O’Malley and Jack Borrebach provided helpful guidance and editing acumen. A special thanks is owed to my terrific research assistants who have worked on this project: Daphne Hsu, Levi Giovanetto, Patrick Dunn, and Smitha Nagaraja. Further appreciation is owed to the entities that have provided generous research support: Visa USA, the Morton C. Seeley Fund, the John M. Olin Fund, and the Milton and Miriam Handler Foundation. I am immensely indebted to the late Nelson Polsby, who turned me into an academic. I also want to thank my avid golfer in-laws, who moved to a residential golf community while I was writing this book, and who took my arguments seriously but never personally. Further thanks are owed to my own parents, whose home looks onto a forest, not a fairway. My most enthusiastic appreciation is owed to Joanna.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INTRODUCTION

Perhaps because of their proximity to major research universities, campus fraternities and sororities are among the most studied and analyzed residential communities in the United States. Scholars interested in understanding the “Greek system” can draw on a wealth of serious research spanning a period of decades. We can find some real gems in these archives of fraternity research. For example, during the 1950s, sociologists Gene Levine and Leila Sussman studied the process by which selective fraternities chose which students to admit. Fraternity members were frank about who they were looking to admit and how a student rushing the fraternity could make a favorable impression on them: [We seek] the man who can express himself maturely, who has an interest in things outside of books; the one who has a neat appearance. [We are looking for] likeable guys, ones who know how to talk and who aren’t loud. Men who’ll fit in here . . . one who is the all-American-boy type but with brains. You can tell right away by their faces or the way they shake hands. A cold, clammy handshake as versus a warm, friendly one. Or by the way they are dressed. You can see right through them in five minutes. First, I’m concerned with how they are dressed. It’s not a matter of Ivy League clothes, but looking neat and well-groomed. The

1

2

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INTRODUCTION

next thing is a hearty handshake with a smile to go with it. A cold, clammy handshake and a tense face don’t go with me.1 The stakes in decisions over who to admit to a fraternity were not trivial. Existing fraternity members would have to live in close proximity to their successful pledges for up to three years. If the fraternity admitted sociable and popular applicants, the brothers would see their status rise, would find their experiences enriched, and could expect a more promising group of applicants in future years. If the fraternity admitted duds, the members would have to live with the consequences of their mistakes. With that much at stake, one would hope that fraternity members would rely on more than a handshake or an outfit or a five-minute conversation to help them separate the wheat from the chaff. But the fraternity members no doubt thought that one could learn a lot about a student from each of those readily observable characteristics. A weak handshake might signal low self-esteem. A shabby outfit could indicate a shortage of the financial wherewithal necessary to graduate. A stumbling five-minute chat might be a good indicator of low social intelligence. Assessing self-esteem, financial resources, and social intelligence directly would be costly, time consuming, and intrusive. As shortcuts, the handshake, outfit, and conversation heuristics seem to have worked well enough. Fast forward almost six decades. Fraternities and sororities persist at many colleges and universities, as do selective admissions policies at the more desirable ones. But some things have changed. Fraternity members no longer need to rely as much on proxies like handshakes or clothing to evaluate applicants, though they surely continue to do so. In an era of social networking websites, virtually every freshman student seeking admission will have made available a detailed account of his or her tastes in music, reading materials, and hobbies, along with an informative map of his friends, acquaintances, and intimates. By substituting a five-minute perusal of a Facebook page for a five-minute conversation, the fraternity member can do a better job of evaluating those attributes he deems relevant for admission purposes. It is not surprising that fraternity and sorority members have turned to these social networking websites in recent years to evaluate pledges.2 We can now abstract away from the world of community service projects, keg parties, and sweatshirts sporting Greek letters. The trend I just

Introduction

highlighted on college campuses is indicative of a dynamic that arises whenever people seek to establish exclusive communities, clubs, or workplaces. In each case, someone wants to ensure that the right sorts of people gain admission, and the wrong sorts of people are turned away. This book asks a central question, but one that scholars have largely ignored: What predicts which mechanisms individuals will use in order to promote exclusivity? The short answer to that fundamental question is that the choice among strategies for achieving exclusivity will be driven largely by information economics. In the story above, the availability of a new and informative resource (Facebook) for evaluating fraternity pledges permits members to reduce their reliance on crude and often misleading evaluative proxies like handshakes, clothing, and short conversations. What’s more, once we see how access to information enables or constrains private actors’ exclusion strategies, we begin to realize the wealth of strategies at the state’s disposal if it wants to curtail problematic exclusion or promote desirable exclusion. In the pages that follow, we will explore a host of settings in which information affects the mechanisms that a real estate developer, landlord, business owner, website proprietor, philanthropist, or city planner employs in order to achieve the desired level of exclusivity. Our journey will take us through British poorhouses in the 1830s, Manhattan apartment buildings in the 1960s, suburban residential golf communities in the 1990s, and present-day virtual communities. We will see similar themes playing out across these eras: the use of crude proxies for desirable and undesirable characteristics, efforts to refine those proxies so as to achieve greater precision, the displacement of crude proxies with more refined metrics as information economics makes greater precision possible, and, in some cases, social and economic dislocation as once-reliable bases for sorting individuals become less trustworthy. We will use our fuller understanding of the relationship between information and exclusion to help us tackle a number of contemporary puzzles about the world in which we live: Why do homeowners associations frequently permit murderers and arsonists, but not sex offenders, to live within their gated walls? Why might doctors be more willing to prescribe narcotic pain medication to Caucasians than minorities? What might explain the astonishing appeal of homes built surrounding golf courses during the 1980s and 1990s and the near collapse of residential

3

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

4

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INTRODUCTION

golf construction today? How come blue-collar African American males face better job prospects at firms that routinely conduct criminal background checks than at those that do not? Read on. Chapter one begins with a brief primer on exclusion. It introduces in an accessible way two important developments in legal scholarship. First, it explains the centrality of what are called in rem exclusion rights—rights that are good against the entire world, not just against parties to a contract. Second, it explains the relationship between exclusion and governance as competing strategies for regulating access to a collective resource. Along the way, the chapter explains what makes exclusion so enticing and important in some contexts, yet so dangerous and damaging in other settings. Chapter two introduces the most familiar form of exclusion, what I call the “bouncer’s right.” This sort of exclusion is backed by trespass law, and is enforced by a device for regulating access—perhaps the bouncer at a nightclub, a gate with a combination lock, a rule governing who is admitted to a sorority or a covenant delineating who is permitted to buy a home in a gated community. When scholars of property refer to the right to exclude, it is this bouncer’s exclusion that they typically have in mind. Much of the law that tries to clamp down on discriminatory exclusion has bouncer’s exclusion as its target. But I will argue that bouncer’s exclusion is hardly where all of the action has been, when it comes to exclusion, and an overemphasis on that single mechanism of exclusion can blind us to the effects of exclusion achieved via other means. Chapter three introduces the concept of “exclusionary vibes,” where language, aesthetics, and marketing choices facilitate exclusion—not by physically restraining people from accessing a collective resource, but by encouraging people to exclude themselves from that resource. Exclusionary vibes are at their most effective in those instances where people prefer to be sorted with like-minded individuals. In such instances, exclusionary vibes can perform a valuable role. In other instances, however, the types of sorting that consumers might want individually may not be optimal for society. Those are the domains in which the regulation of exclusionary vibes is warranted. While housing discrimination laws do regulate some uses of exclusionary vibes, the use of communicative tools to create homogeneity is by and large unregulated by law.

Introduction

Chapter four introduces another tool of exclusion, dubbed the “exclusionary amenity.” An exclusionary amenity bundles access to a resource with the use of a polarizing and costly communal good. If some groups are predictably willing to pay for the communal good, and others are not, then the decision to bundle the amenity with the collective resource may engender significant homogeneity. The chapter discusses at some length the evidence suggesting that during the 1980s and 1990s some golf course communities may have facilitated residential racial segregation. It also discusses the current development of Ave Maria Township in Florida, which has used the exclusionary amenities strategy to create an overwhelmingly conservative Catholic municipality. Interestingly, exclusionary amenities are essentially a blind spot where the law is concerned—their use is unregulated, which means they can be used to circumvent antidiscrimination protections. Chapter five proposes a theory to explains why the entity controlling a resource might favor one exclusionary strategy over another. It suggests that the most important, and heretofore unrecognized, factor that governs when a resource owner opts for one exclusion strategy or another is asymmetric information concerning the attributes of potential resource users. If the person controlling access to the resource can easily and accurately determine whether a prospective user possesses the desirable attributes that the owner is hoping to screen for, then bouncer’s exclusion will be the most suitable strategy. If, on the other hand, it would be very difficult for a bouncer to discover which prospective entrants should be admitted and which should be excluded, it will usually be optimal to shift the decision concerning exclusion to the prospective entrants themselves, via exclusionary vibes or exclusionary amenities. The chapter concludes by pointing out how legal considerations, social meaning, and other factors might also influence a resource owner’s choice among exclusion strategies. Chapter six examines what is at stake for society when a resource owner chooses whether to exclude and via what mechanism. Not surprisingly given how much is at stake, exclusion itself is often controversial. Interestingly, as this chapter will suggest, some forms of exclusion are much more visible than others, so an owner can often achieve a result via subtle strategies that never could have been achieved had the mechanism and results of his strategy been more transparent.

5

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

6

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INTRODUCTION

Chapter seven returns to some of the themes introduced in chapter four and examines what a legal framework that took exclusionary amenity strategies seriously and sought to regulate their more pernicious effects might look like. It explains that requiring developers to embed “inclusionary amenities” in communities can often offset some of the polarizing effects achieved via the use of exclusionary vibes or amenities. It then suggests that there is a role for law to play in requiring that inclusionary amenities be bundled into new residential developments, or that the absence of inclusionary amenities in new developments be viewed skeptically by the state. Chapter eight observes that in the last decade or so, industrialized societies have begun to embrace technologies that will facilitate much more finely grained exclusion than would have been possible since the onset of industrialization and urbanization. This development, called the “reputation revolution,” has the potential to fundamentally change the nature of the game when it comes to exclusion. More precisely, the chapter shows how this revolution has often caused resource owners to shift back toward bouncer’s exclusion, precisely because bouncers can now gather and analyze information about prospective entrants so much more easily than before. Notably, however, the move toward greater precision sometimes has very laudable effects. For example, empirical research suggests that when employers have more accurate information about which job seekers have criminal records and which ones do not, their hiring decisions will be less racially discriminatory than those of employers who lack access to reliable criminal history information. Using this body of research as a model, the chapter shows how information policy—whereby the state makes transparent information about individuals that would otherwise be obscure (searchlight strategies), or obscuring information that would otherwise be readily apparent (curtain strategies)—can be an essential tool of twenty-first-century antidiscrimination strategies. Chapter nine discusses at a higher level of generality the suggestion that information policy be used as a tool for influencing the propensity of private actors to discriminate. It discusses the circumstances under which moving away from traditional legal interventions like carrots and sticks, and toward interventions like searchlights and curtains, is appropriate. It turns out that the effectiveness of searchlight and curtain strategies will be

Introduction

driven by considerations such as the ratio of animus-based discrimination to statistical discrimination, the reliability of the reputation information to which decision makers might have access, and the ability of prospective entrants to create misleading profiles of themselves. Chapter ten tackles the issue of carrots and sticks versus searchlights and curtains from a distributive framework. Specifically, who wins and who loses when information policy is used to constrain or promote exclusion by private actors? The answers to that question are both interesting and important. Focusing on a case study like the United States’ evolving treatment of sex offenders who have completed their terms of incarceration provides a useful glimpse of the sorts of issues that will arise with greater frequency in the coming years. The chapter then steps back to consider briefly normative questions about what it might feel like to live in a world where exclusionary and inclusionary judgments are made on the basis of increasingly fine-grained and accurate information about individuals, rather than blunt and imprecise information about visible groups.

7

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

PART

I

THE MECHANICS OF EXCLUSION

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

1 THREE MECHANISMS FOR EXCLUSION

When discussions of “rights” appear in the newspapers or blogosphere, the authors are often referring to individuals’ rights that correlate with distinct obligations on the part of governments. An individual’s right to free speech correlates with a governmental duty not to, say, throw a dissident in jail on the basis of his unpopular expressions. An individual’s due process rights prevent the government from depriving an individual of liberty without first following a set of fair procedures. These sorts of rights—long dubbed negative liberties—prevent the government from taking particular actions against individuals. In some democracies, especially European ones, individuals are guaranteed positive rights—such as a right to shelter, health care, or education—and the government has affirmative duties to vindicate those rights. Analysts of all stripes have spilled a great deal of ink concerning political rights that vindicate affirmative and negative liberties. Another set of rights figures somewhat less prominently in the popular discourse but perhaps influences individuals’ everyday lives to a more significant degree than political rights. There are contract rights and property rights. When Americans arrive at work each weekday, it is a series of contract rights that governs their relationship with their employer. When Americans return from work at the end of the day, it is a series of property rights that protects the privacy of their homes—indeed, property rights were necessary to keep their homes free of undesired interlopers while the owners were at work. Contract rights facilitate the worker’s commute to 11

12

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

work (probably in a vehicle purchased by the worker via a contract with a car dealership), and property rights govern the entertainment websites that the worker visits during a coffee break. Contract and property rights are, in short, omnipresent in people’s everyday affairs. The basic distinction between contract and property rights is straightforward. A contract right is a right that can be enforced only against parties to a bargain. Lawyers call such rights “in personam.” An owner can enforce a property right, also called an “in rem” right, against any member of society. A couple of illustrations will be helpful. Suppose I contract with a handyman to paint my house at 9:00 a.m. on Monday. When nine o’clock rolls around, I see no sign of the handyman, so I call him on his cell phone. He says that after agreeing to paint my house, he got a call from James, who offered him $200 more than I was paying to paint his similarly sized house at 9:00 a.m. on Monday, so he decided to do that more lucrative job instead. The handyman seems to have breached our contract, and I may be able to recover damages against him in court. But I have no contract rights against James. My rights in this instance are in personam, and they only run against individuals with whom I have entered into contractual arrangements. Now suppose that I wake up the next morning and find James splashing in my swimming pool. I don’t know James from Adam. I never gave him permission to use my pool, but neither did I tell him he couldn’t dive in whenever he pleased. No matter, my right to exclude James from my pool is an in rem right—it is good against the whole world, and certainly good against James in this instance. I can sue James for trespass, and will certainly win, even though James and I never entered into a contract whereby he promised me not to swim in my pool. This right to exclude James is a quintessential property right. Indeed, the United States Supreme Court has characterized the right to exclude as the “hallmark of a protected property interest,”1 and property scholars routinely describe the right as the core, or the essential element, of ownership.2 We need not think too hard to imagine a world without rights to exclude—the virtual world in Second Life essentially has no prohibitions on trespassing. As a result, some mayhem ensues: vandals destroy the virtual campaign headquarters set up by presidential candidate John Edwards, smearing his posters with excrement or covering them with blackface, and openings of virtual stores in Second Life are disrupted by virtually armed paramilitary forces.3

THREE MECHANISMS FOR EXCLUSION

The right to exclude protects more than backyard swimming holes. It is the essential protection for a pharmaceutical company seeking to bring a new drug to the market, a Manhattan nightclub owner trying to establish a haven for the trendy and the beautiful, and a baseball stadium owner who wants to make sure that only paying customers get to see the action on the field firsthand. My right to keep James out of my swimming pool, backed by trespass law, is indeed a vital sort of exclusion. Yet it is not the whole ballgame. Individuals can be excluded effectively via other mechanisms as well. Namely, I can use my protected political rights to keep James off my property. My right to display a “No Trespassing” sign depends as much on First Amendment rights of expression as it does on property rights. And an individual who exercises power over a community, such as a real estate developer, a landlord, or the proprietor of a nightclub, can use contract law to control which sorts of individuals wind up gaining admission to a community, and which sorts of individuals wind up on the outside looking in. In my view, it is helpful to think about exclusion more broadly, so as to encompass those rights that are not themselves founded on trespass law but can nevertheless substitute for in rem exclusion rights. Exclusion, in these terms, includes a property owner’s efforts to exclude prospective entrants from a resource, as well as the entrants’ decisions to exclude themselves from the owner’s resource. Properly understood, the right to exclude encompasses three component rights. The first right enables the resource owner to exclude outsiders using trespass law. The owner can exclude everyone from the property, as the government may do with an environmentally fragile wilderness area; admit only a few stylish individuals to the club while keeping most wouldbe entrants shivering behind a red velvet rope; or admit the overwhelming majority of entrants but exclude a small segment of undesirable prospective customers (“No shirt, no shoes, no service”). This right to admit prospective entrants selectively is known as the “bouncer’s right,” and it will be our focus in chapter two. A second exclusionary right relies not on trespass law, but on language and the predictable social dynamics that it can create. This right is really a free speech right, one that permits a resource owner to convey messages about who is welcome or unwelcome on the property. Will different

13

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

14

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

prospective freshmen show up if you change the name of your university from “The University of the South” to “Sewanee”? You had better believe it, and a concern about student demographics explained that recent rebranding campaign by a well-regarded liberal arts school.4 We will refer to intentional use of language in order to prompt undesired entrants to exclude themselves as “exclusionary vibes.” Chapter three will discuss them in detail. Finally, in some instances where a resource owner wants undesired prospective entrants to sort themselves out, an exclusionary vibe may be ineffective, perhaps because the law sometimes prohibits the use of exclusionary vibes (as we will see later), or because the stakes are high enough to prompt members of undesired groups to ignore the exclusionary vibe. In such circumstances, a resource owner may embed an “exclusionary amenity” in the community. An exclusionary amenity is a costly resource that desired entrants will benefit from but undesired residents will not. Because of that simple dynamic, a resource owner can convince undesired prospective entrants who would be willing to ignore an exclusionary vibe that joining the community, and paying for an amenity that will not benefit them, will be too costly. Chapter four will offer a number of instances where exclusionary amenities seem to have been employed by real estate developers, town planners, and other social engineers. It should be immediately apparent that a savvy resource owner can substitute one exclusionary strategy for another. When the law restricts the discretion available to an owner to exclude people directly, as it often does, that owner may turn to exclusionary vibe strategies, which may be scrutinized less closely by government officials, or which might be favored by some measure of constitutional protection. The potential substitutability of trespass-based and non-trespass-based exclusion rights raises an important question that chapters five and six will answer: How does a resource owner choose which exclusion strategy to adopt? One of this book’s core insights is that information costs are often the primary factor guiding a resource owner’s decision about which exclusionary strategy to exercise in a particular context. More precisely, when a prospective entrant has private information about her preferences and behaviors that the resource owner cannot obtain at a low cost, the resource owner essentially will delegate the exclusion function to the prospective entrant, using either exclusionary

THREE MECHANISMS FOR EXCLUSION

vibes or an exclusionary amenities strategy. When, by contrast, the resource owner has or can easily obtain information about prospective entrants’ relevant attributes, the resource owner generally will prefer a bouncer’s right strategy.

The Demand for Exclusivity Baseball legend Yogi Berra famously said about Ruggeri’s, a St. Louis restaurant, that “nobody goes there anymore, it’s too crowded.” Although interpreting Berraisms has always been a risky endeavor, the quote makes sense if we interpret Berra to be saying something along the lines of “I don’t go there anymore; it’s too crowded.” Indeed, that is a sentiment that most readers will have identified with at some point in their lives. Popularity has a downside, and one such downside is that the original customers who gave the place some of its charm may be driven elsewhere by crowds of new patrons. It was the exclusivity of Ruggeri’s years ago, manifested in the lack of crowds and the presence of the right kind of patrons, that made the restaurant worth going to at the outset. Similar sorts of dynamics still play out—even where there are no physical constraints on crowd capacity. In 2007, corporate management at the ubiquitous clothing chain the Gap confronted a deeply disturbing downward trend in their sales. The cause? Gap was trying to sell clothes to too many customers. In their effort to appeal to teenagers, middle-aged shoppers, and the elderly, they wound up exciting nobody.5 Fashion industry analysts were seemingly unified in their diagnosis of what ailed the Gap: “If you stand for everything in fashion today, you stand for nothing,” said Paul R. Charron, the former chief executive of Liz Claiborne. “Brands like the Gap . . . have a special challenge to be relevant in a period when focus and exclusivity are so important.” “The definition of a specialty store is focus,” said Howard Davidowitz, chairman of Davidowitz & Associates, a national retail consulting firm and investment bank. Gap does not have that focus, Mr. Davidowitz said. And in trying to meet the needs of infants, teenagers, and even the elderly, its designers play it safe, season

15

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

16

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

after season. “The merchandise is booooooring,” Mr. Davidowitz added. “Too basic.” “They have to pick out a demographic and go after it with a maniacal focus, to the exclusion of anyone else,” said Bob Buchanan, an analyst at A.G. Edwards & Sons. “If there is one thing you cannot be in the middle of the mall anymore, it is all things to all people,” he added. “And that is what Gap has been trying to do.”6 As an exemplar of the sort of a store whose strategies should be a model for the Gap, industry insiders identified Abercrombie & Fitch, a store that is “openly hostile to what it considers the wrong customer—typically anyone over 30—warding them off with booming music, dark shades on the front windows and teenage employees standing out front. Gap has veered to the other extreme, putting out a welcome mat to nearly everyone, with well-lighted, sparsely decorated stores and ageless fashions.”7 A deep global recession followed this fashion commentary in short order, and to readers perusing this text not long after its publication date, the sentiments expressed might seem like nostalgic reminders of a different, gilded age. Abercrombie has fallen on hard times too—with sales down substantially in the recession as the company was criticized for failing to offer discounts to suddenly value-conscious consumers.8 Perhaps going forward the fashion industry will reconsider its niche-oriented design and marketing choices. But I doubt it. The same desire for exclusivity that dominates the fashion world characterizes the markets for many consumer goods and services: hair stylists, art, and of course real estate, which will occupy much of our attention in the pages that follow. In common parlance, the condominium and the cooperative are treated interchangeably. A typical home buyer in a densely populated community will go out looking to buy an apartment, not a condo or co-op per se. As a practical matter, the would-be buyer who decides to make an offer on a unit in a cooperative building will have more work ahead of him than his counterpart who tries to buy a condominium unit. Whereas the condominium buyer needs to make only one person—the seller—happy, the cooperative unit buyer must convince the cooperative’s board of directors that he should be allowed to buy into the building. What explains this difference?

THREE MECHANISMS FOR EXCLUSION

When a condominium purchaser buys a condominium he is purchasing a combination of kinds of ownership rights. He will own a particular apartment in fee simple absolute—the same sort of ownership arrangement that characterizes most single-family homes. He will also obtain rights to use common spaces in the condominium—the elevator, the lobby, the swimming pool, and so on—that are shared with all his fellow condominium residents. A cooperative owner buys something different. By law, a cooperative is a nonprofit corporation comprised of its shareholders. In a co-op building containing twenty identical apartments, each owner will own a 5 percent share of the building as a whole. Technically, every fractional owner will have the same rights to each apartment that he has to the elevator or lobby. As a practical matter, however, rules arise in each cooperative building giving every owner possessory rights of one (and only one) apartment. That said, because of the nature of cooperative ownership, owners are on the hook for their fellow owners’ defaults. More precisely, if one of the twenty owners stops making mortgage payments, the remaining nineteen will have to pay that owner’s share or risk defaulting on the mortgage for the building as a whole. This greater financial interdependence among cooperative owners has affected the laws governing their decision making. More precisely, the law has given cooperative owners more discretion in deciding who to approve as an owner than condominium owners are allowed.9 The other significant differences between condominiums and cooperatives stem largely from the divergent ownership structures: cooperative owners are typically heavily involved in screening would-be owners, they have more detailed rules governing owners’ use of the building, and they spend more time than their condominium counterparts on internal governance as a consequence.10 In an era where apartment owners are sensitive to the risks that defaulting neighbors might impose on them and generally resistant to spending precious free time in condominium board meetings, it is little wonder that the condominium structure is preferred by most. Developers of new apartment buildings overwhelmingly prefer to adopt the condominium structure over the cooperative one, and some cooperative buildings have endured the lengthy, costly, and contentious process of converting themselves to condominiums. Still, these transition costs are formidable, so a lot of cooperative buildings have retained their structure despite its inefficiencies.11

17

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

18

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

This discussion brings us to recent research by Michael Schill, Ioan Voicu, and Jonathan Miller.12 Consistent with the foregoing analysis, apartments in condominiums generally attract a premium over similar apartments in housing cooperatives. Controlling for the many variables that differentiate housing units, Schill and his coauthors found that, as a general matter, a condominium apartment commands an 8.8 percent premium over a similarly situated cooperative.13 This finding was consistent with the expectations of Manhattan real estate agents. Strikingly, however, Schill and his coauthors identified a group of apartments in which the ordinary patterns were reversed. For these apartments, the cooperative form actually conferred a very substantial premium—approximately 25 percent—on owners.14 The distinguishing characteristic of cooperative units that command a premium is that they bar financing as part of the purchase of a unit. These units, in short, are in buildings where the owners can afford to buy homes without any need for a mortgage. Prohibitions on mortgage financing arise in both condominium and cooperative buildings, but it is the cooperative apartment buildings that command a hefty premium as the domain of Manhattan’s economic elites. Let us be quite clear about what this data means. Wealthy owners of Manhattan cooperative apartments seem willing to pay a hefty premium, sacrifice substantial leisure time, and forgo a great deal of financial privacy at the time of purchase, all for the benefits of exclusivity and having a much greater say in who their neighbors are. For money-is-no-object types, the leisure-time premium paid by cooperative owners may be even more substantial than the economic premium. Cooperatives’ authority to exclude has been exercised to keep the likes of Madonna and Richard Nixon out of prestigious New York buildings,15 but there is also some evidence suggesting that it has been used to exclude members of historically marginalized groups.16 The New York courts have begun policing decisions to exclude members of protected groups from cooperative apartments closely in recent years.17 This data suggests something else that is equally important. Before the advent of antidiscrimination laws and doctrines, restrictions on free alienation could keep undesired prospective residents out of certain communities. But once the state began enforcing antidiscrimination

THREE MECHANISMS FOR EXCLUSION

laws, people who wished to exclude these undesirables had to do so on the basis of proxies. Wealth and income often provide important proxies, and suburbs in particular managed to maintain substantial exclusivity by restricting neighborhoods to single-family homes built on large lots.18 The Manhattan cooperatives, however, show that price will sometimes be an inadequate exclusionary proxy. People may want to exclude “new money” or “old money” or members of a particular political party from their communities, and they will seek out some mechanism for doing so. This helps explain the cooperative premium at the high end. In recent decades, income and wealth have become less accurate proxies for race and characteristics that have often formed the basis for exclusion. Once wealth and income become less useful proxies, people interested in screening their neighbors may have to turn to other characteristics. On the basis of the Schill et al. research and similar studies,19 it seems appropriate to assume a market demand for exclusion in the residential setting, particularly at the highest income levels. Some other studies suggest that, as incomes rise, the demand for racially homogeneous neighborhoods actually increases.20 Residential exclusion, in that sense, may be something of a luxury good. But why are individuals willing to pay such a high price? What do exclusion and exclusivity get them?

Exclusion’s Appeal We can begin by dismissing the happy but unlikely possibility that everyone’s associations can be ordered by mutual consent without engendering conflict. As Lee Fennell has noted, “where people in a bounded metropolitan area form geographic subgroups, it is not possible to extend rights of exclusion to each of those subgroups simultaneously, unless people happen to have perfectly congruent grouping preferences. Because every household must end up somewhere, and because adjacency to other households is an inevitable spatial fact in a metropolitan area, each household’s associational choice affects the associational outcomes of others.”21 The same zero-sum aspect of exclusion often emerges outside of the land use context too—consider workplaces, colleges, country clubs, and even friendship circles as environments where exclusivity leaves some people in residual associations of which they would rather not be a part.

19

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

20

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

Some Americans genuinely would not want to be a part of any club that would have them as members. As a result of exclusion’s not fully consensual nature, there is often unseemliness associated with exclusionary practices and outcomes. Being a member of a country club with exclusionary admission policies is itself controversial—politicians sometimes have to resign from these clubs when seeking higher office. Getting one’s child into an exclusive primary school or one’s self into an exclusive boutique medical practice is often accompanied by some guilt about the people who were thereby excluded. Individuals who were able to obtain scarce military deferments in wartime often expressed angst about those who were sent to war in their place. Some college freshmen find exclusionary fraternities and sororities repulsive and do not rush even though they would have little difficulty gaining admission. When I was a third-year law student at Yale, dozens of the first-year students tried, unsuccessfully, to convince the law review board that membership in that organization should be open to any law student who wanted to join. Membership in the institution had long been determined on the basis of a writing and editing competition. Students who had chosen to enroll in one of the nation’s most selective law schools evidently wanted to take a principled stand against exclusivity.22 As uncomfortable as it may feel to be the beneficiary of an exclusionary policy, those on the outside looking in surely feel worse. The Supreme Court in Brown v. Board of Education recognized that the exclusion of African American students from public schools attended by whites would demoralize and stigmatize them.23 Related concerns about exclusion in particular contexts would open up a series of battlegrounds in the civil rights movements of the twentieth century: lunch counters, universities, the armed forces, fraternal lodges, and country clubs. In many of these cases, exclusion from institutions meant exclusion from powerful social networks and the opportunities they presented. Beyond these considerations, however, exclusion from spaces and organizations can be deeply depressing and hurtful. Social psychology experiments have found that being excluded from social situations causes individuals to direct aggression not only against those who excluded them, but also against innocent third parties as well.24 Other laboratory studies reveal that individuals who are excluded from groups are less likely to contribute

THREE MECHANISMS FOR EXCLUSION

to charities or assist strangers in need as a result, are more likely to make foolish, self-defeating choices, and have a hard time solving problems demanding intellect in the aftermath of their exclusion.25 Experimental evidence suggests that exclusion from even fleeting interactions can cause the excluded party to feel significantly less happy and regard her life as less meaningful.26 Other research has suggested that experiencing social exclusion causes individuals to feel less empathy for others, while increasing aggressive and antisocial behaviors.27 Being excluded against one’s will tends to make a person disempowered, marginalized, selfish, masochistic, powerless, and short-sighted. Notwithstanding the discomfort that exclusionary policies sometimes create for both the excluder and the excluded, exclusion has obvious appeal. At some level, exclusion is so central to the functioning of a market economy that we could not seriously contemplate its abolition. The President’s Council of Economic Advisors could not have particularly useful policymaking discussions if any member of the public were permitted to participate in the discourse whenever he or she pleased. A great chef wouldn’t be able to concoct his most innovative and scrumptious dishes if he were forced to charge patrons a price that even the poorest citizen could afford to pay. And the Boston Red Sox wouldn’t have much of a starting rotation if any fan was able to demand an opportunity to pitch to Alex Rodriguez with the playoffs on the line. Nobody relishes exclusion for its own sake, but we recognize that an economic discussion is likely to be improved by the exclusion of nonexperts from technical meetings. Most of us agree that twelve-course meals prepared by Charlie Trotter are not a basic human right, so we tolerate his use of the market system to allocate scarce space in his dining room. And all of us can recognize that it is far more entertaining to see Alex Rodriguez face Josh Beckett than Josh Bolton, so we tolerate Major League Baseball’s player selection and promotion process, even though it disappoints the overwhelming majority of individuals who would love to play professional baseball at the highest levels. The rest of us can try to hit Josh Beckett’s fastball via video games. Exclusion might serve any number of important human interests. In the examples above, excluding the disadvantaged or unskilled seemed to allow the quality of the enterprise to improve dramatically. In other settings, exclusion fosters other important values: it might protect privacy, promote

21

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

22

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

innovation, facilitate social ties, confer status on those who are included, or enhance public safety and private security. The personal privacy interest is particularly important in the context of residential property. Home purchasers care a great deal about shielding what goes on in the exterior of their homes, or on the surface of their backyards, from nosy neighbors or passersby. They will invest heavily in purchasing extra land as a buffer zone around their property, will seek out landscaping and drapery that effectively shield their land from prying eyes, and may opt for soundproof walls and windows both to keep exterior noises out and interior noises in. The proof here is in the pudding: Multiple Listing Service entries for particular properties commonly emphasize that a particular property provides its owners with exceptional privacy. Exercising control over private information becomes much easier if one can exclude third parties from one’s property. Someone stationed in my living room, or even my front yard, could learn a lot more about my sleep habits, political views, interactions with family members, and consumption choices than could someone forced to respect my property lines. Corporations will value privacy for a different set of reasons. Unlimited access to a corporation’s manufacturing facility will enable competitors or would-be competitors to learn about each other’s innovations. Trade secrets law provides an alternative to patent law for firms that develop innovative approaches to problems and wish to prevent their competitors from misappropriating those innovations. It is also a common form of protecting resources that are valuable and require a great deal of effort to assemble, but are not protectable under patent or copyright law. The classic example is a closely guarded customer list. In cases concerning both innovations and “sweat of the brow” products, the courts have required that a party seeking to establish a trade secret misappropriation claim demonstrate that it has taken reasonable precautions to prevent its trade secret from falling into the hands of a rival.28 A critical determination is often the extent to which the firm has secured its premises against the possibility that an outsider might enter and discover the trade secret in question. Trade secrecy protections promote productive economic activity. If firms cannot be sure that they can benefit from their innovations or sweat of the brow, then they will be discouraged from investing in the creation of those works in the first place. The right to exclude and trade secrets law

THREE MECHANISMS FOR EXCLUSION

thus work in conjunction to encourage research and development and prevent commercial rivals from reaping where they have not sown. The argument for letting the military exclude outsiders from military bases is similar—protecting the secrecy of military technologies and installations confers national security advantages. The ability to exclude outsiders from land may be the most effective means of protecting the owners’ property interests in moveable resources that are stored on that land. A landowner who wishes to prevent people from stealing his livestock can do so much more cost-effectively by fencing off his land and preventing unauthorized entry than by letting anyone come onto his property and trying to catch the thief by monitoring the activities of every entrant.29 And the same dynamic plays out in the context of an ordinary suburban homeowner. A locked front door, a security system, a guard dog—these are all common means of protecting one’s home, one’s possessions, and one’s family from criminal wrongdoing. The right to exclude trespassers from land thereby indirectly strengthens the right to prevent trespasses to moveable property stored on that land.30 Setting aside the question of whether fences and alarms make individuals safer, the evidence seems to be quite strong that such tools of exclusion make property owners feel safer. In America’s suburbs, residents of gated communities report that they feel more secure inside the walls of their communities than they do elsewhere, and they view their communities as safer than neighboring communities, even when there is no difference in the comparative crime rates.31 The ability to exclude might also promote social connectedness. The right to exclude is what makes inclusion meaningful. Imagine a world where I had a legal right to walk into your dining room at any moment and demand to be served. In addition to imposing burdens on you, this practice would render meaningless the gesture that a dinner invitation generally entails. Yet it is often through such gestures that friendships, professional collaborations, or romantic relationships are initiated, developed, and preserved. By giving you the right to exclude, the law ensures that your decision to include a particular person can generate good will. A similar sort of dynamic can play out in business relations too—consumers will often get annoyed when their interactions with their lawyers, doctors, or real estate agents are interrupted so the service provider can take another

23

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

24

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

customer’s call. By ignoring the call from another, the service provider signals that he values his relationship with a particular client enough to forsake all others (at least temporarily). Because inclusion cannot exist without exclusion, exclusion furthers an important role in promoting status, which in turn facilitates effort and innovation. Many undergraduates want to be admitted to the very best graduate or professional schools; they recognize that competition for scarce slots will be fierce, and so they work extremely hard to get good grades, develop strong mentoring relationships with college professors, and devote themselves to extracurricular projects. The same is true in athletics: it is because the National Basketball Association will exclude so many talented basketball players that each aspirant puts in thousands of hours on the practice court in an effort to maximize his innate talent for the game. For homeowners and even renters living in gated communities, such a residential arrangement makes them feel that they have achieved middle-class respectability.32 Circumstances may also arise where the owner of a resource exercises the exclusion right as a means of protecting a would-be entrant. For example, the proprietor of a theme park may impose height or age restrictions on patrons wishing to ride a roller coaster. Airlines may refuse to let an eightmonths-pregnant woman fly on a commercial flight. A landowner may fence off a construction site to stop the neighborhood kids from using it as a playground. The resource owner in question in each of these examples faces a mix of motives—the owner may have altruistic reasons to avoid injuring third parties, the owner may fear that accommodating one user with special needs may inconvenience other users (for example, the other passengers on a flight if a pregnant passenger goes into labor, forcing an emergency landing), or the owner may fear that she will be sued if she fails to prohibit third parties from using a dangerous resource. In these various scenarios, exercising the right to exclude will likely be the optimal strategy, given the inadequacy of alternatives. The airline cannot extract from the pregnant passenger a promise that she will not go into labor during the flight, and even if the owner posts signs warning kids not to play in a construction zone, she may face “attractive nuisance” liability under tort law if one of the children is injured on his property. When the government wants to test a new weapons system, it will exclude outsiders from a large swath of federal land for their own protection.

THREE MECHANISMS FOR EXCLUSION

The advantages of exclusion are therefore significant and widespread. It would be an exercise of foolhardy idealism to try to outlaw exclusion wherever we find it. But in some domains, the undesirable effects of exclusion outweigh its benefits. How do we know which domain is which? Social psychologists Dominic Abrams, Michael Hogg, and José Marques provide us with a reasonable starting point: Even when relationships are interdependent exclusion may in some cases be consensual. For example, boys and girls in elementary school seem quite happy to self-segregate for purposes of play and social activities. Doctors and nurses, academic teachers and college cleaners may be quite happy to dine in different workplace canteens or common rooms. Men and women seem reasonably content with the idea that they should use different rest rooms in public buildings. Likewise, people may choose to exit relationships consensually. . . . Whether consensual or not, the outcomes may vary in terms of their desirability for building a more stable or progressive society, and should not be assumed to be unproblematic, either conceptually or practically. The types of exclusion that are more likely to concern researchers and policy makers are those involving conflict or threat. In these situations exclusion is likely to be contested and the dynamics are likely to involve a struggle over who is excluded and on what basis.33 Their intuition—that the forms of exclusion that generate conflict are most likely to be the ones that warrant state intervention—has genuine appeal. At the same time, this book will discuss some forms of exclusivity that have generated little controversy, but ought to be far more controversial than they are because of their effects on social welfare. Indeed, sometimes the forms of exclusion that are nearly invisible to outsiders turn out to be the most interesting cases of all.

Governance—Exclusion’s Alternative As all the previous examples suggest, the benefits of letting an individual owner exercise the right to exclude potential entrants can be substantial. But there are always alternatives to exclusionary policies.

25

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

26

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

Rather than excluding economists from the Council of Economic Advisors meeting, we could just enact rules about the contents of what participants could say at the meeting. If all statements must concern technical economic matters, then it is extremely unlikely that someone with no economic sophistication would have anything to say. Rather than fencing the neighborhood children out of a construction site, we could let them play, but require them to be accompanied by a parent who would watch them, or require them to wear helmets, elbow pads, and shin guards. In regulating a particular resource, there are essentially two strategies available: you can regulate who gets to use the resource or how people use the resource. According to a terrific series of papers written by Henry Smith, property regimes mediate a choice between two strategies for controlling a resource: governance and exclusion.34 Exclusion, in Smith’s framework, refers to “a low-cost, but low-precision, method that relies on rough informational variables like boundaries to define legal entitlements.”35 When the law grants an owner a right to exclude, it delegates authority over that resource to an owner or group of owners, who can decide whether to fence it off from the outside world or allow users to come and go as they please. This delegation to the resource owner can be advantageous for a host of reasons: it assigns the gatekeeper right to the party with the greatest incentive to exercise it in a way that maximizes the value of the property; it reduces the need to coordinate among multiple stakeholders in order to make decisions about how to use the resource; and it may maximize consumer options if different owners of fungible resources adopt varying exclusion strategies for optimizing the value of their property. If exclusion lies at one end of the spectrum for controlling access to a resource, what strategy is at the other? Smith contrasts the exclusion strategy with governance rules: At the pole opposite of exclusion along the organizational dimension are what I am calling governance rules. These rules . . . pick out uses and users in more detail, imposing a more intense informational burden on a smaller audience of duty holders. For example, village herdsmen may have rights to graze animals that are circumscribed as to number of animals, time of grazing, and

THREE MECHANISMS FOR EXCLUSION

so on. . . . [A] wide range of rules, from contractual provisions, to norms of proper use, to nuisance law and public environmental regulation can be seen as reflecting the governance strategy; compared to basic trespass and property law, all these governance rules require the specification of proper activities.36 Governance, in short, manages a resource, not through blunt access restrictions, but through sets of rules or standards that regulate the conduct of those who do have access. While governance might involve governmental decision making about uses via zoning laws or land use regulations, exclusion involves a delegation of authority to the owners of property, with the government acting only to enforce trespass laws. This distinction is important for my purposes, as will become clear below.37 How should society regulate the use of a scarce resource: through an exclusion regime or a governance regime? Smith argues that the choice hinges on the tradeoffs between exclusion’s simplicity and governance’s precision: “The exclusion strategy bunches together a lot of uses and does not inquire into details; it lacks the benefits of precision in terms of maximizing the value of individual uses, say from specialization by different actors in different uses of the same asset. At the same time, the exclusion strategy also avoids the costs of precision. By contrast, governance captures the benefits of precision but at a higher cost. Governance deals directly with problems that are left to the owner to handle under exclusion.”38 Governance, in this model, is the appropriate response to resource controversies in which a great deal is at stake. In such circumstances, the likely gains from tailoring a set of rules or norms to the resource in question exceed the costs of creating a governance regime, communicating the rules of that governance regime to stakeholders, and resolving any disputes about whether the rules have been followed.39 When there is little at stake, relying on blunt exclusion strategies spares society the substantial information costs associated with governance and still generates resource allocations that, while perhaps not optimal, are good enough.

27

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

2 THE BOUNCER’S RIGHT

The traditional account of the property right to exclude emphasizes a solitary, isolated individual who excludes everyone from his land.1 This is a right that seems useful primarily to hermits or to conservationists wishing to preserve land in a wilderness state. To the rest of us, it is a right of little value. Indeed, it is almost impossible to locate a reported judicial opinion involving a permanent invocation of the right to exclude everyone from land that has positive economic value but little environmental value. The closest case, Brown v. Burdett,2 involves a testator’s wishes that her home be bricked and boarded up “with good long nails” for twenty years following the testator’s death, a will provision that the court invalidated on public policy grounds.3 Ordinary people find the idea of a totally isolated life repellent.4 Rather, the prospect of hosting friends, neighbors, relatives, and service providers on one’s property for visits of varying durations is a large part of what makes land ownership valuable. As James Penner puts it, the “right to property is like a gate, not a wall. The right to property permits the owner not only to make solitary use of his property, by excluding others, but also permits him to make a social use of his property, by selectively excluding others. . . . The property system . . . is not there to ensure that people live like hermits.”5 Once we move away from extreme and economically unproductive exercises of the right to exclude, we arrive quickly at rights that take on enormous economic importance.6 As soon as an owner wishes to allow potential entrants onto his property at certain times of day, or 28

THE BOUNCER’S RIGHT

admit some parties while refusing entry to others, or establish some criteria that will govern entry onto the land, he is exercising the sort of discretion that makes the right to exclude a right worth exercising. The bouncer’s right, then, is the landowner’s right to discriminate among various parties, permitting some to enter or use the land while keeping others off the property entirely. Like the bouncer at a nightclub, the owner must exercise discretion as to who can utilize the resource, and the criteria for exclusion need not be transparent. There are commercial and noncommercial variations on the bouncer’s right, but they are analytically similar. A business owner will value the right to admit some customers and vendors but not others, whereas a homeowner will care about his right to invite friends and family into his home while excluding foes and strangers. The bouncer’s right is vital because it can replace the narrowly tailored but costly governance regimes used to solve common resource problems. Setting aside “fugitive resources,” such as air, that do not lend themselves to exclusion-oriented strategies, exclusion invariably emerges as an alternative to governance for regulating the way in which a collective resource is exploited. A landowner can admit all comers and then control their use of a common resource through governance (e.g., via zoning laws or environmental regulations), or the landowner can be selective about who gets to use the resource, weeding out undesirable uses of the resource by refusing to admit those deemed likely to engage in the undesirable uses. Recall that the choice for a real estate developer is between controlling what his residents do and controlling who his residents are. The more effective the real estate developer’s selection mechanism, the less need there will be for costly governance mechanisms. The bouncer’s right is not, and has never been, absolute. Legal thinkers have sometimes referred in passing to absolute exclusion rights—most notably, William Blackstone’s Commentaries refer to “that sole and despotic dominion which one man claims and exercises over the external things of the world, in total exclusion of the right of any other individual in the universe.”7 But Blackstone proceeded to enumerate many restrictions on that supposed right of total exclusion, and Carol Rose has convincingly demonstrated that when read in the context of the rest of Blackstone’s work about property, this famous sentence is “at most a cartoon or trope.” Rose notes that property “regimes always consist of some individual rights,

29

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

30

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

mixed with some rights shared with nearby associates or neighbors, mixed with still more rights shared with a larger community, all held in relatively stable but nevertheless changing and subtly renegotiated relationships.”8 The clearest manifestation of the contingent nature of rights to exclude arises in the form of antidiscrimination law.

The Fair Housing Act In 1968 Congress enacted the Fair Housing Act (FHA), which makes it unlawful “to discriminate against any person in the terms, conditions, or privileges of sale or rental of a dwelling . . . because of race, color, religion, sex, familial status, or national origin.”9 The law goes into substantial detail about how this broad language is to be effectuated, and the most obvious thrust of the law is to limit a property owner’s discretion to pick and choose renters or buyers. For example, the law prohibits a refusal to rent or sell to an individual on the basis of membership in a protected group.10 It is less widely recognized that several provisions in the FHA actually limit the sorts of things that individuals can say about property. Many of these restrictions proscribe property owners from making statements concerning their own real estate. For example, an owner cannot make misrepresentations as to the availability of housing on the basis of group membership.11 The classic case the law tries to prevent is one where the owner places an advertisement for an apartment for rent, but then announces (disingenuously) that the apartment has been rented when an African American family shows up to express interest in renting the unit. The act strongly restricts the speech of nonowners as well. Real estate agents cannot steer clients toward blocks or neighborhoods where members of their demographic group predominate and away from areas where members of their demographic group are in the minority. Nor can real estate agents, or anyone else, engage in blockbusting activities, whereby someone tries to induce panic sales on the basis of perceived changes in the demographics of a neighborhood.12 And if an owner runs an advertisement in a newspaper that “indicates any preference, limitation, or discrimination” about a renter or buyer, both the owner and the newspaper are liable.13 As interpreted by the courts, even subtle messages can engender liability: a

THE BOUNCER’S RIGHT

pattern of using exclusively Caucasian models in housing advertisements might unlawfully convince African American renters to look elsewhere.14 Finally, according to a 2008 judicial decision, a website that facilitates renters’ search for potential roommates on the basis of race is violating the FHA.15 The law, in short, substantially limits the property owners’ ability to admit renters selectively, to transfer property to purchasers selectively, and to communicate any message indicating an effort, willingness, or intention to do so. Real estate agents and newspaper publishers are substantially constrained in what they can say about other people’s property. Limitations on the right to exclude are widespread and ever-present.

Common Law Limitations on Exclusion Rights There is a libertarian account of nondiscrimination principles that regards them as a startling abrogation of exclusionary rights. Criticizing a court case that precluded a shopping mall owner from excluding pamphleteers who were advocating on behalf of political causes with which the owner appears to have disagreed, Richard Epstein argues that “private property gives the right to exclude others without the need for any justification. Indeed, it is the ability to act at will and without need for justification within some domain that is the essence of freedom . . . of property.”16 Yet even to a principled libertarian like Epstein, substantial limitations on exclusion rights are necessary in order for a private property system to work well. Necessity presents the classic case. The law has long permitted a third party to trespass on another’s property where doing so is necessary to protect the third party’s personal safety. This limitation on the right to exclude is embraced by libertarians like Epstein,17 because the harm to the third party if absolute exclusion rights are enforced substantially exceeds the harm to the property owner if those exclusion rights are temporarily abrogated.18 In such cases, it makes sense to require the third party to compensate the property owner after the fact, but the property owner’s (small) loss, not the third party’s (large) gain should be the basis for calculating compensation. Indeed, the intuition is that everyone is better off living in a world where they know that necessity will excuse the trespass,

31

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

32

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

provided compensation is paid after the fact. The alternative rule would encourage people to take excessive precautions against the possibility that they might wind up facing risks to life and limb. For similar reasons, the government may enter private property in many cases despite the landowners’ lack of consent. Indeed, the government may even destroy private property, in cases involving public necessity, without having to compensate the owner. Firemen can enter onto one landowner’s property in order to fight a blaze on a neighboring parcel. Policemen can pursue a fleeing suspect onto private property without fear of liability. In wartime soldiers can take actions that result in the destruction of buildings if doing so is reasonably related to their military objectives of subduing rioters.19 In contrast to the private necessity cases, the law has not required the state to pay any compensation to the burdened landowner after the fact. The rationale seems to be that the law should not unduly complicate the decision making of firefighters, police officers, or soldiers when quick reactions to avert serious threats are required.20 Though necessity is the classic case limiting the right to exclude, it is hardly the only one. The common law prohibits landowners from constructing spite fences on their own property. A spite fence is a structure whose only purpose is to block a neighboring owner’s view or access to sunlight. The line between a spite fence and a permissible privacy-enhancing fence will be blurry in some cases, but the law’s basic intuition is as follows: purely antisocial uses of the right to exclude, designed to vex or extract money from a neighbor, are socially wasteful and produce no benefits to society. The law therefore deems them verboten. As the common law has evolved, further restrictions on exclusion rights have arisen. The 1971 New Jersey Supreme Court case of State v. Shack 21 has become a canonical opinion. That case holds that a landowner cannot exclude lawyers or doctors who would like to meet with migrant farm workers housed on his property. The court reasoned that allowing the landowner to exercise complete control over who could visit the migrant workers on his property put him in a position of acting like a feudal lord. In the background, the court wrestled with concerns that if the farm workers were denied access to doctors and lawyers, then the state could not be made aware of public health threats or violations of the law regarding, say, workers’ safety or minimum wage. To the court, society’s interest in

THE BOUNCER’S RIGHT

seeing that the migrants had relatively unfettered access to the professionals who wished to see them trumped the landowner’s interest in controlling access to his property. Of course, the big question mark surrounding State v. Shack is just how far its rule can go. Courts in other jurisdictions have not hesitated about extending its protection to migrant farm workers who desire access to family planning services22 or charitable assistance from nonprofit organizations.23 But a significant source of uncertainty exists surrounding the question of whether a migrant farm worker can sign away his rights to receive visitors—perhaps in exchange for higher pay or better working conditions. State v. Shack itself suggests that they will not be permitted to enter into these kinds of bargains with their employers.24 Yet in another context where the law deals with a vulnerable population, we enable tenants to waive the rights to receive visitors via contract. For example, an individual entering an in-patient facility for drug treatment is often required to waive the right to receive visitors for a rather lengthy period of time as a condition for entry. As a result, even close family members may be excluded. Yet permitting these sorts of arrangements is uncontroversial, perhaps because the loss of the associated right is deemed therapeutically useful to the addicts themselves. By contrast, the Shack court seemed to view the exclusion of lawyers and health care professionals as something that would undermine the migrant tenants’ interest. What of third parties’ interests in not being excluded? Do these issues figure into the calculus? The court in Shack suggested that the press might have rights to access a farmer’s property to report on the possibility of mistreatment directed against his tenant workers.25 This aspect of the ruling is unusual—private property owners typically do have rights to exclude muckraking journalists, even when the exposés they write would benefit readers and consumers.26 As a general matter, the law is not terribly sympathetic to the interests of third parties in not being excluded, with civil rights law and necessity defenses to trespass forming important limitations. This is true even when the stakes are quite high. Think of an invitationonly gathering where lots of business gets done—the Davos, Switzerland, World Economic Forum, for example. There are lots of ambitious entrepreneurs or civic leaders who would very much like to participate in those meetings so that they could secure face time with the movers and shakers

33

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

34

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

of international politics and commerce. Davos’s exclusivity helps ensure that those at the periphery remain so and those with the power, connections, and wealth hang onto it. By exercising dominion over its property, the organizers at Davos exercise some control over the lives of those they include and those they exclude. Yet virtually no one thinks that Davos must open itself to everybody. Rather, people recognize instinctively that if Davos were opened up to everyone, then the movers and shakers would stop showing up. Nobody goes there anymore. It’s too crowded.

Sex Offender Residency Restrictions During recent years, American society has focused significant attention on the problems surrounding sex offenders who have served their time and been released into the community. These concerns have manifested themselves in the enactment of Megan’s Laws at the state and federal level, which require released sex offenders to register their home addresses with the state.27 Many states publicize the addresses of sex offenders on websites, and through other means, often publish offenders’ photographs and other relevant information. More recently, some states have enacted legislation that provides for twenty-four-hour electronic monitoring of high-risk sex offenders who have completed their terms and been released into the community.28 In the wake of the new laws directed at sex offenders, homeowners associations and other common-interest communities have recorded covenants prohibiting the sale of homes to registered sex offenders.29 In Mulligan v. Panther Valley Property Owners Ass’n, a resident of one association challenged its prohibition on the sale of her property to Tier 3 sex offenders.30 Tier 3 sex offenders are those who, the state has determined, pose a high risk of recidivism.31 The court refused to invalidate the relevant covenants, finding that because there were only eighty Tier 3 registrants living in New Jersey at the time of the litigation, the restriction did not unreasonably restrain Mulligan’s ability to sell her home.32 The Mulligan court had a harder time determining whether the restrictions were contrary to public policy. The court seemed to sympathize with the desire of residents to protect themselves from sexual predators, but expressed worries about what would happen if most homeowners associations in New Jersey mimicked

THE BOUNCER’S RIGHT

Panther Valley’s policies. The court held that the restrictions did not violate public policy at present, but noted that it would revisit the issue if sex offenders had too few residential options in the future.33 Bouncer’s exclusion seems like an intuitive strategy for a residential community whose residents want to reduce their susceptibility to sex crimes. Although there is a scholarly debate about the level of recidivism for sex offenses and the frequency with which sex offenders prey on strangers,34 the presence of convicted offenders seems to raise serious alarm among neighbors, such that targeting sex offenders for exclusion may be a rational response for some homeowners associations.35 So let us walk in the shoes of the Panther Valley Property Owners Association and review the various strategies that it could use to protect itself against these sorts of crimes. Panther Valley’s decision to record at the county recorder’s office the covenant restraining residence by sex offenders precluded it from making exceptions to its exclusion policy on a case-by-case basis, as any owner of a home in Panther Valley would have had legal standing to enforce the covenant if another owner violated it.36 Hence Panther Valley’s choice of strategies is a clear instance of bouncer’s exclusion. The association merely took away the bouncer’s discretion, and forced her to apply a straightforward bright-line rule. Panther Valley could have opted for a governance regime instead of its bouncer’s restriction on Tier 3 registrants residing in the community. Indeed, a number of possible governance systems spring to mind in this instance, some of them ordinary and some quite unorthodox. For example, the community could invest in traditional law enforcement strategies, like hiring constables to deter and detect sexual offenses. With the consent of its residents it could install ubiquitous video-surveillance cameras in the interiors of homes as well as in public spaces. More provocatively, the association could require that all those who enter the community provide a bond, whose proceeds shall be forfeited to the victim if the entrant was subsequently convicted of a sex offense. As should be readily apparent from these examples, governance strategies seem poorly suited to addressing the danger of sex offenses in Panther Valley. The various governance regimes impose substantial privacy costs, economic costs, or administrative burdens on the community’s residents. Among the strategies we have discussed so far, excluding the likely sex offenders from Panther Valley, while allowing

35

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

36

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

everyone else who can afford to purchase a home to do so, seems to be the cost-effective approach to this problem, especially given the fact that a very small number of prospective Panther Valley residents pose any risk of becoming sex offenders. The lesson here is that bouncer’s exclusion strategies can provide more precise and far more efficient solutions to problems than governance. This idea should surprise no one. We have long understood that the most cost-effective way to prevent a store’s night watchman from robbing the owner blind is not necessarily to watch the watchman’s every move or to provide him with a compensation package that allies his interest with that of the owner’s, but to try to hire an honest watchman.37

The Greek System Suppose a group of college students wishes to maximize some subjective variable within a communal residence. That variable might be physical attractiveness, athletic prowess, intelligence, or sociability. Governance will not work particularly well here, absent some means of exclusion (beyond the university’s admissions requirements). It is not as if the group can mandate plastic surgery or psychotherapy for its members. If the community has no control over membership composition, it therefore cannot expect its residents to be any more attractive, athletic, smart, or fun than the average student at its institution. Because of the desire to shape the character of the community, some students devote substantial resources to obtaining information about those who wish to join their communities, and offer admission only to those applicants who convince the majority of current residents that they would make good housemates. This is an apt description of the pledge and rush processes in fraternities and sororities, the example with which this book began.38 Although both the Panther Valley residents and the Greeks seek to exclude large segments of the population from their respective communities, the Greeks’ method of exclusion raises different considerations. The Panther Valley criterion for exclusion is relatively objective and transparent. (Tier 3 sex offenders are excluded, and anyone else who can afford to purchase a home may do so.) The Greeks’ criteria for admission, by contrast, hinge on relatively subjective factors like affability and

THE BOUNCER’S RIGHT

fashion sense,39 and the members of the community may be reluctant to reveal fully the bases for admission, perhaps out of fear that those seeking admission will misrepresent themselves strategically.40 This distinction has important consequences for both the existing members of a community and those who seek entry.41 Applying objective, readily observable admissions criteria requires few resources. Thus, the residents of Panther Valley need not gather very much information in order to determine whether a would-be owner or occupant will be admitted. By contrast, the members of fraternities and sororities must devote substantial energy to collecting, verifying, synthesizing, and evaluating information about applicants. As discussed at the very beginning of this book, fraternity and sorority members historically opted for timesaving methodologies like snap judgments, visual cues, and the evidently popular handshake heuristic. To the extent that the individual members apply different criteria or obfuscate the criteria, fraternity members must devote resources to aggregating preferences or reconciling conflicting preferences. The difference, in short, between Panther Valley and Pi Beta is the quality of the information that the decision makers have at their disposal. Fraternity members historically had a much harder time discovering the attributes of applicants and might have worried that if they publicized their admissions criteria, unsociable members could gain admission by practicing their handshakes or investing in a few nice outfits. To prevent this from happening, fraternity members kept their admissions criteria a secret. The same holds true in the Facebook era—fraternity members won’t publicize what sorts of profiles they’re looking for. The underlying assumption is that in the absence of reliable information about what type of Facebook profile will appeal to the existing fraternity members, applicants will describe themselves truthfully. Panther Valley, thanks to Megan’s Law, can determine who is a sex offender at a low cost, so it publicizes its criteria in unambiguous terms. The choice of objective or subjective bases for admission has consequences for the applicants as well. Given the clarity of Panther Valley’s restrictions on entry, it is unlikely that many ineligible residents will attempt to join the community. Real estate agents presumably will inform would-be buyers whom they suspect of having shady pasts of the restrictions before they even look at any properties in Panther Valley, and they might transmit

37

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

38

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

the same information to seemingly squeaky clean purchasers as a selling point. The Greeks’ more subjective entry requirements may do a poorer job of dissuading applicants whose chances of admission are remote. As a result, the Greeks’ exclusionary tactics may raise applicant expectations in a manner that ultimately disappoints those who are not admitted to their fraternity or sorority of choice. To avoid this problem, the Greeks try to send hints to applicants who seek admission but are unlikely to gain entry that they should look elsewhere.42

Immigration Policy Exclusion at the macro-level closely resembles its micro-level counterpart. Nation-states control access to a valuable and exclusive resource: their own soil. Little wonder, then, that we see nations pursuing many of the same exclusionary strategies that landowners employ. Bouncer’s exclusion is commonplace. Throughout American history, policy makers have been using national origin proxies to try to shape the nation’s workforce and polity. Most obviously, national origin is a proxy for race, so Americans who cared about the racial composition of the United States could use immigration policy as a tool for achieving their desired ends. Employers might use proxies in much the same way, seeking out immigrants from particular nations because that nation’s inhabitants are believed to possess particular characteristics.43 Adam Cox and Eric Posner have analyzed immigration design decisions as a response to asymmetric information in the international migration context.44 They suggested that the immigrant has private information about her own attributes, preferences, and intentions that United States customs and immigration officials lack. The government thus faces a choice between two sorts of strategies: it can invest in gathering more information about prospective entrants before they are admitted to the United States or it can admit entrants freely, gather information about them once they are here, and then deport those whose actions in the United States do not make positive contributions to society.45 Where the information gathering and processing costs of assessing individuals’ likely contributions to society based on their behavior abroad are high, we can expect to see the government rely more on ex post deportation. Where those costs are low, we can expect to see the government rely on ex ante exclusion at the border.

THE BOUNCER’S RIGHT

In the developed world, the rise of consumer behavior databases and law enforcement databases has enabled countries to evaluate some visa and asylum seekers in a fine-grained and individuated way, rather than relying on crude proxies like nation of origin. That said, the protective approach of European Union lawmakers toward information privacy interests has helped apply the brakes to this development somewhat.46 Of course, while immigration to the United States from developed nations like Canada, Japan, Germany, and Australia is not negligible, it pales in comparison to migration from developing nations in which accurate consumer information databases do not exist.47 It is precisely the absence of reliable information about individuals from developing nations that causes immigration officials to rely on proxies like national origin. In response to the events of September 11, 2001, the United States substantially curtailed the number of visas issued to residents of predominantly Islamic nations and made it more difficult for foreigners to study in the American higher education system.48 (One of the terrorists involved in the September 11 attacks was in the United States on a student visa).49 There were some exceptions to this policy of national-origin-based exclusion. In-demand specialists like doctors, nurses, engineers, and academics had a somewhat easier time obtaining visas,50 in part because their skill sets were in particular demand, but also in part because their degrees in those fields suggested that it was unlikely that they would present a threat to American national security.51 Wealthy foreigners with at least $1 million to invest in new American businesses also have a much easier time gaining visas.52 Although this program was justified as an engine of job creation in the United States,53 it might, alternatively, be characterized as another immigration proxy: foreigners with disposable investment capital and a commitment to investing it in U.S.-based startups might be particularly likely to be desirable migrants.54 Notice what is going on here. Proxies are everywhere in immigration decisions: national origin, education, career path, age, gender, English fluency, family ties in the United States, wealth. All of these attributes wind up determining who is admitted to the United States and who is not. We rely on these proxies to indicate who might pose a security risk, who might fill a technically or physically demanding job, who might wind up on the welfare rolls, who might wind up starting a business, who is likely to

39

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

40

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

commit ordinary crimes, who might have a hard time finding employment in the United States, who is likely to be assimilated, who is likely to return home when a visa expires, and who is likely to overstay a visa. It would, of course, be far simpler if immigration authorities could rely on accurate information about individuals rather than using these group-based proxies, but there are serious impediments to such an approach. With individually based information scarce in the developing world, there may be little to augment reliance on crude proxies like national origin or English fluency. This dichotomy suggests something counterintuitive, which is that the United States ought not to embrace a one-size-fits-all methodology for evaluating visa applicants. Rather, it should rely more heavily on individualdossier-based bouncer’s exclusion when assessing immigrants from those countries with well-developed reputational infrastructures and more heavily on group-based bouncer’s exclusion or post-entry surveillance and deportation for visa seekers from nations where the reputational infrastructure is nonexistent or of unreliable quality. A Japanese immigrant can be assessed rather reliably prior to entry and then probably can be ignored; a Mongolian immigrant might need to be judged via proxies or admitted provisionally and subjected to greater monitoring of employment status and involvement in the criminal justice system by immigration officials. Tailoring our immigration strategies to the varied information environments that exist around the globe would entail a radical reworking of the present approach. And, in some ways, an approach that looked to the information environment would be inequitable. Why should a Japanese citizen have an easier time obtaining a visa than a Mongolian? It is generally a matter of luck where someone happened to be born. But these sorts of distinctions are already rampant in immigration law—it is already far easier to get a visa to the United States as an Irish migrant than as a Venezuelan. If the State Department is going to pick and choose based on immigrants’ national origins, we should prefer that they do so on the basis of a principled criteria, rather than, say, political clout. Before we leave the interesting topic of immigration behind, one further thought is worth flagging. A great deal of popular discourse about immigration policy touches on policies that the Immigration and Customs Enforcement agency (ICE) uses to control access to the border. The provision of welfare benefits, health care, education, and employment

THE BOUNCER’S RIGHT

opportunities to undocumented aliens has created great controversy in the United States. There are many who fear that generous welfare benefits or lax employment status verification policies will act as a magnet to undocumented aliens from other nations. By being stingy with welfare or emergency room services, immigration opponents argue, we can discourage aliens from trying to cross the borders illegally. As a descriptive matter, these concerns certainly have some validity. Generous benefits will attract the notice of would-be migrants, because information travels easily between domestic immigrants and potential immigrants in the home country. Thus welfare or employment policy can be used as an important lever to attract or repel would-be immigrants. Indeed, even less tangible aspects of an immigrant’s existence, such as a community’s insistence on English-only instruction or rallies against immigrants, might become very effective mechanisms for signaling to immigrants that they should look elsewhere.55 It turns out that these effective forms of exclusion have little to do with the bouncer’s right. Financial inducements or tough talk might replace the activities of ICE. But they are different in kind. It is not just a person with a formal legal right who can control where people go and which sorts of people can enter. One of New York City’s most famous bouncers was Dorothy Dean, a tiny Manhattan socialite who shamed people not fabulous enough to belong at the nightclub Max’s Kansas City into leaving. She had no formal right to exclude anybody, but through her sheer chutzpah and the tacit consent of other patrons, the riffraff were effectively kept from mingling with nightclub regulars like Andy Warhol.56 In the immigration context, the high-profile community organizer working for English-only instruction in public schools aspires to be a modern-day Dorothy Dean. Groups of such individuals can be very effective in influencing the choices of others to enter or not enter particular communities. To focus on immigration policy while ignoring everything that falls outside of ICE’s purview is to really miss what is driving the relevant debates. We need a new vocabulary for understanding all these dynamics of exclusion, and the following chapter will begin to develop one.

41

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

3 EXCLUSIONARY VIBES

When I was an undergraduate at a large, state-subsidized university, I got the sense that there was a fair bit of homogeneity within each of the many fraternity and sorority houses on campus,1 and I attributed this homogeneity to the rush and pledge processes. But then I moved into cooperative student housing and noticed a similar level of homogeneity within particular houses, which was initially puzzling, since any student could move into a campus cooperative.2 The co-ops did not exercise the bouncer’s right at all (except to exclude non-student residents), and yet each house seemed to have a distinct personality, not unlike the fraternities and sororities that I occasionally visited. Governance and socialization seemed like incomplete explanations for this homogeneity.3 If my impressions were correct, this homogeneity in the campus cooperatives raised interesting questions about what was substituting for the Greek system’s bouncers. We can generalize this question more broadly. In a big city like New York or Los Angeles, the most exclusive nightclubs have bouncers while the most exclusive restaurants do not. Indeed, many of the snazziest restaurants in each city even allocate reservations through opentable.com, which enables an individual to secure the right to dine at the establishment without having to interact with any human being. Provided restaurant patrons dress in accordance with the establishment’s dress code, behave appropriately, and can pay for the meal, there will be no basis for the restaurant to exclude them. 42

EXCLUSIONARY VIBES

Not so at the club. A nightclub bouncer might elect not to admit a patron purely because the bouncer does not like the look of him, notwithstanding the patron’s ability to pay for drinks, willingness to queue for a spot in the club, and compliance with the norms governing appropriate attire for that kind of club. Both the restaurant and nightclub are exclusive, but the mechanisms for exclusion vary. In the restaurant, the exclusivity will be established on the basis of price (if the restaurant takes reservations and reservations are easily obtainable), or willingness to incur nonmonetary costs (if reservations are very hard to come by or if a popular restaurant refuses to take reservations, seating patrons on a first-come, first-served basis). In the nightclub, the bouncer decides who gets admitted and who should be removed for improper conduct. The divergence of practices among restaurants and nightclubs suggests that the decision to maintain exclusivity can be put into practice through alternatives to the bouncer’s right. There are costs that arise whenever a resource owner wishes to use the bouncer approach. For example, the bouncer must be trained and paid. The bouncer may make mistakes and admit too many of the wrong people or too few of the right people. Those targeted for exclusion might bribe the bouncer to let them enter. Excluded individuals may resort to shoving or even more serious violence if they believe the resource owner is excluding them improperly. And people seeking entry may misrepresent themselves as a way of fooling the bouncer into letting them enter a space to which they would not be admitted absent the deception. When a resource owner finds both governance and bouncer’s exclusion cost-prohibitive, she will often employ the “exclusionary vibes” strategy.

How Vibes Work An exclusionary vibes approach involves the landowner’s communication to potential entrants about the character of the community’s inhabitants. Such communication tells potential entrants that certain people may not feel welcome if they enter the community in question, because they will not share certain affinities with existing or future residents. Although the landowner invokes no legal right to exclude anyone from the property in question, an exclusionary vibe may still be effective at excluding a targeted population thanks to two mechanisms. First, a prospective entrant may

43

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

44

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

view the exclusionary vibe as an effective mechanism for creating a focal point around which people can organize their affairs.4 Focal points are markers around which individuals can coordinate. As Thomas Schelling explained, a focal point must be conspicuous and unique5—tell an undergraduate student that he is to meet up with an unknown confederate in New York City on a particular day, but provide the student with no additional information, and there is a good chance the student and confederate will both find their way to the clock at Grand Central Station at noon—there is only one such clock, and the building’s architecture, central location, and function as a transit hub render it highly conspicuous. Focal points are used to great effect in residential development too—the name, location, or marketing of a new development might facilitate prospective residents’ coordination around it. A variation on this focal point effect arises if the prospective entrant assumes that the exclusionary vibe will create a community population that is likely to embrace bouncer’s exclusion at a later date as a means of removing the entrant from the community. Second, the potential entrant may assume, incorrectly, that the exclusionary vibe is backed by a bouncer’s right to exclude those who are not made to feel welcome by the exclusionary vibe.6 I will elaborate on both of these mechanisms in detail, using a hypothetical community. Suppose that a condo developer sees a market niche for residential communities targeted toward extroverted individuals. To that end, the developer advertises the new condominium as Social Butterfly Place. This advertising should suffice to make the condominium attractive to social butterflies and their families, but unattractive to more introverted individuals, even if the developer does not invest in any amenities that are designed to appeal to the extroverted. How come? Here we see the dynamics working together. Extroverted individuals probably will value proximity to fellow extroverts, so that they can easily find outgoing partners for conversation and joint social activities. Introverts may feel left out or marginalized living in the building, and this marginalization may impose real social and psychological costs on them. Because they anticipate incurring some of these costs if they move into Social Butterfly Place, many introverts will opt for a residence in some other building, all else being equal. This phenomenon illustrates the possibility for exclusionary vibes to serve as focal points.

EXCLUSIONARY VIBES

Savvier introverted prospective condominium purchasers may be deterred from moving into Social Butterfly Place as well. These potential entrants would understand that the developer could do nothing to stop them from purchasing a home in the building, but would recognize the effectiveness of the focal point strategy at establishing a homogeneously extroverted population. Even if one of these introverts did not care whether she felt left out of her neighbor’s social interactions, she would rightly worry about the prospects that her extroverted neighbors might in the future decide to use the bouncer’s right to expel introverts if they concluded that there were too many introverts in their midst, or adopt, by majority vote, governance rules that made life pleasant for extroverts and unpleasant for introverts, such as mandatory weekly condominium association meetings or lax nighttime noise regulations for hallway conversations and parties within units. Finally, some would-be condominium purchasers will see the sign “Social Butterfly Place” and erroneously assume that only extroverts are permitted to reside there. In other words, they may misread the exclusionary vibe as indicative of a developer’s intent and authority to exercise a trespass-based right to exclude them. If they were to ask the developer whether introverts may reside in the tower, the developer would say that all are welcome, but many people are embarrassed to ask questions of that sort or are ignorant of their legal rights. Hence an exclusionary vibe may act as an effective bluff that prevents some potential entrants who are targeted for exclusion from moving into a community. As these examples indicate, the simple act of naming a new development “Social Butterfly Place” could prove effective at excluding the introverted from residence in the development. Exclusionary vibes can function as a substitute for, or a complement to, the bouncer’s right. It should be equally clear that every exclusionary message is implicitly inclusionary with respect to those people who would prefer to live in a community that is devoid of those people who are targeted for exclusion.

Exclusionary Vibes in Practice In the real world, real estate developers sometimes do market their residences as paradise for extroverts.7 The exclusionary vibes strategy is prevalent where other groups or attributes are targeted for exclusion or

45

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

46

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

inclusion as well. Condominium buildings adopt names like Cotton Hope Plantation8 and Sholom House.9 And individual cooperative houses near my old university campus described the character of their communities in great detail on the Internet.10 We need not strain our minds too much in order to see the power of exclusionary vibes. Imagine, for example, the sales center for a mixed-income planned development in a large southern city. The sales center looks identical to any other sales center, with one difference: a large confederate battle flag flies on the flagpole out front. The mere presence of this flag would produce a first generation of homeowners who are overwhelmingly white. College Campuses The year is 2005 and a small, well-regarded liberal arts college in Sewanee, Tennessee, is confronting a problem of too few applications and too little diversity in the student body. More precisely, it accepts two-thirds of the students who apply for admission, and less than 9 percent of its students are African American, Latino, or Asian American.11 The college hires an out-of-state marketing firm to figure out how it might make itself more attractive to a broader swath of the nation’s college applicants. The firm’s conclusion: change your name. The University of the South thus became Sewanee. The Chicago-based marketing firm suggested to school administrators that the name “University of the South” brought to mind “negative connotations for students around the country.” University president Joel Cunningham noted, “We are in the South, and we benefit from the literary tradition, the warmth, the friendliness. Do we have to recognize that there are those who might have other connotations and that we have to draw them in to better understand the place? Yes.” Although “The University of South” remains part of the school’s official name, the font size for “Sewanee” has been made very large in the university logo and the font size for “The University of the South” quite small. The school’s website address is now www.sewanee.com. Soon after the change it was of course too early to tell how successful the rebranding effort had been. In an academic cycle following the change, admissions rates were down slightly, to 63.6 percent, though the racial makeup of the school had not shifted significantly. The school’s list of incoming freshman for 2008–09,

EXCLUSIONARY VIBES

posted to its website in July of 2008, included 392 students, of whom 78 were from non-southern states. Another 7 hailed from outside the United States. In all, then, 21.7 percent of the school’s class of 2012 consisted of non-southerners. Without knowing how this figure compares to previous classes or to comparable southern liberal arts colleges, and without a sense of other measures that the school took to change its applicant pool, it is hard to know what to make of this figure. Perhaps the initial data suggests that even a rather aggressive rebranding campaign will not make an institution’s reputation turn on a dime. That said, evidence from around the country suggests that campus name changes may be associated with increases in the size and quality of the applicant pool. After Beaver College became Arcadia University in 2001, applications doubled and the average SAT score of enrolled students increased by 60 points.12 The College of New Jersey, formerly Trenton State, saw similarly impressive increases in the credentials of its incoming students.13 Western Maryland College (now McDaniel College), California State University, Hayward (now California State University, East Bay)— all posed the question, “What’s in a name?” and provided the same answer: “A lot.” Why should a name make a difference? The professors, campus facilities, alumni, and staff at the colleges remained the same after these rebranding efforts. But the people behind these name changes believed that students would change. In switching its emphasis from the University of the South to Sewanee, one established college hoped to draw more students of color and more students from above the Mason-Dixon line. In becoming Arcadia University, Beaver College sought to leave behind a history as a women’s college and a legacy of sophomoric jokes about its name. Trenton State wanted to become a player at the state level, not just locally. And Western Maryland College, which turns out to be in Central Maryland, not Western Maryland, wanted to stop confusing potential applicants.14 (California University of Pennsylvania should take note. So should the Los Angeles Angels of Anaheim, for that matter.) By changing its name, each school tried to create a new focal point around which applicants could sort themselves. A new Sewanee student would be someone attracted by the new name, and someone who wanted to be around other students who were also attracted by the same name. A new College of New Jersey

47

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

48

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

student would be enamored with her school’s state-sized ambitions and its association with a wealthy state, not a struggling city. No doubt, she wanted roommates and classmates with a similar orientation. For a university, a name change is a relatively inexpensive way of altering the student body’s composition. It requires some transition costs, requiring the university to educate alumni, vendors, prospective employers, and the like. In some cases, like Sewanee, alumni who preferred the old name raise hackles. But a name change is cheap in it does not require a school to do anything different in its admissions, say. They can maintain the same SAT and grade-point cutoffs, and they need not read the admissions essays any differently or more closely than before. Indeed, if a small school is interested in getting greater representation of northerners on its campus, a rebranding strategy will often prove superior to a policy of giving admissions preference to the northerners who do apply. After all, if the school establishes a focal point that northerners will find unattractive, then the students who apply from northern states and still get rejected are likely an unimpressive lot. Critically, the rebranding will aim to secure more applications from students in the north, and this in turn will permit the school to admit more northerners without lowering academic standards. Of course, name changes are only a part of the exclusionary vibe strategies that a university employs. The university where I teach, in its effort to attract serious-minded students, embraces as an unofficial slogan the mantra “University of Chicago: Where Fun Comes to Die.” And the University of Wisconsin, in an overzealous effort to present itself in promotional brochures as a racially integrated campus to prospective students, was caught Photoshopping an African American student into a picture of allwhite Wisconsin students at a football game.15 The Legal Regulation of Residential Advertising Before breaking ground on a new residential community, a real estate developer will spend a great deal of time thinking about the types of residents he would like to attract, given prevailing market conditions and the nature of the property in question. Once he has made that determination, the developer will create a marketing strategy designed to attract those types of people and, simultaneously, convince those types of people whom he does not want to attract to take their housing dollars elsewhere.

EXCLUSIONARY VIBES

Today’s media outlets enable a great deal of message tailoring, with particular messages intended to reach particular recipients. Real estate developers market directly to consumers, mostly through print and Internet-based advertisements, but occasionally via television or radio.16 Developers also market to consumers through real estate agents, and the means of communicating with agents are quite varied. A developer’s choice of advertising media may have substantial effects in skewing the audience of potential buyers who hear about a development. The choice to advertise locally, regionally, or nationally may skew the audience too. The choice to advertise in mostly English-language publications or non-English publications will make a big difference, as will the choice of program for radio or television advertisements. Internet microtargeting has enabled an unprecedented degree of precision and efficiency in the delivery of ads and tailored content. All these various strategies for marketing are reasonably well understood and regulated to varying degrees.17 Advertising can communicate exclusionary vibes through more subtle means, as well. For example, print advertisements or billboards may make use of models enjoying their idyllic residential surroundings. Advertisements or billboards depicting exclusively Caucasian models will tend to attract Caucasians, whereas those depicting exclusively (or largely) African American models attract African Americans.18 Even more subtly, developers may use particular color schemes or themes in advertising, sales centers, or model homes to “signal” consumers about the characteristics of people they would like to attract.19 Advertising, in short, can be an effective means for attracting certain types of consumers and dissuading other types of consumers from purchasing in a new development. In some instances, seemingly innocuous actions steeped in history convey a powerful message to white and black audiences alike. For most of the twentieth century, a siren atop the Villa Grove, Illinois, water tower rang out at six o’clock every evening.20 The siren’s purpose was to provide a warning to African Americans that they were required to leave Villa Grove for the night if they had not already done so. When sociologist James Loewen visited Villa Grove to investigate, all twelve of the town’s residents that he interviewed independently confirmed the significance of the siren’s sounding. Remarkably, the town continued to sound the siren every evening until 1999, when the practice stopped. Loewen writes, “I

49

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

50

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

had hoped it stopped the practice because residents became ashamed of why it was first put in place, no longer cared to explain its origin to their children or guests, and had reconsidered their [African Americans must leave by] sundown policy. No, I learned, it stopped owing to complaints about the noise from residents living near the water tower.”21 In the most recent census, Villa Grove counted just 8 African Americans among its 2,553 residents. The nearby community of Champaign, by contrast, is 15.6 percent African American.22 Aesthetics After a developer has sold off the lots in a new development, there will no longer be any reason for him to maintain an advertising presence with respect to the new development. This vacuum presents a potential problem for residents who expect to live in the development for a long time and hope that a neighborhood’s character will be maintained beyond the first generation of buyers. What can a developer do to give these residents some peace of mind? A highly effective exclusionary advertising campaign can have secondgeneration and third-generation consequences. Upon searching for housing, second-generation buyers may attempt to discern some information about the composition of the existing community, and, if there is a real shortage of people with whom they feel affinity, they may elect to purchase elsewhere.23 Individuals may also be more likely to move into neighborhoods where they already have friends. But predicting these second-generation decisions is going to be made more difficult by the law’s restrictions on the types of information that second-generation buyers can obtain about their neighbors. Real estate agents can get into serious trouble if they answer questions about a neighborhood’s racial or religious composition, for example, and alternative sources of information—such as census data or repeated strolls around the neighborhood—may be imprecise. Moreover, although coordinated purchasing by minority groups is rare, it can happen, and there are historical examples of such purchases snowballing into rapid neighborhood transformation.24 In such circumstances, developers and first-generation residents may seek more permanent exclusionary devices. For example, a developer may arrange for the construction of homes in his development that exclusively

EXCLUSIONARY VIBES

adopt particular architectural styles. Different architectural styles predictably appeal to divergent groups of people. A subdivision of Tudor-style homes may attract stodgy families, whereas a subdivision constructed exclusively of homes in the California modern style may attract edgier first- and secondgeneration residents. Given the expense of transitioning from one type of architectural style to another, these first-generation decisions might lock into place particular homeowner homogeneities for many generations.25 Having said that, what seems edgy today often will become stodgy within a decade. Brazilian Beaches The beautiful beaches of Rio de Janeiro are one of Brazil’s national treasures. Sunbathing and socializing on those beaches is an essential Rio experience, and the beaches are called Brazil’s “most democratic spaces” by the natives. The beaches are also deeply segregated through exclusionary vibes. Rio’s elite beaches are divided by twelve postos (posts), each approximately half a mile apart. Each posto attracts a certain demographic and proves inhospitable to sunbathers who do not fit the mold.26 Ipanema Beach’s Posto 9 is the most exclusive beach; it is popular among entertainers and left-wing intellectuals. A rainbow-colored flag flies near Posto 9, demarcating an area of the beach popular with homosexuals. Posto 7 at Ipanema is where the surfers congregate. Upper-middle-class moms prefer the area between Postos 11 and 12 in Leblon beach—a tradition dating back to the installation of a diaper changing area and small playground near the sidewalk adjoining the beach.27 As a formal legal matter, each grain of sand on a Rio beach is public property. No legal entitlements block gay men from bathing on Leblon beach or families with children from building sand castles at Posto 9. But social norms and traditions are strong, and it is understandable that beachgoers who are wearing little clothing, are potentially vulnerable to thieves, and are susceptible to involuntary interactions with neighboring sunbathers would prefer to pitch their towels and umbrellas near like-minded folks. For marginalized groups, like the farofeiros—dark-skinned, working-class families from distant suburbs—staying near each other provides protection against harassment by snobs or racists from central Rio.28 Exclusionary vibes are strong on the beaches of Rio. It turns out they are present on American beaches as well.

51

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

52

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

“No Trespassing” Signs and Public Trust Lands In California and other states that recognize a robust public trust doctrine, members of the general public have statutory, common law, or even constitutional rights to access those portions of the beach that fall below the high tide line.29 In many cases, these rights are supplemented by dry sand access rights that state or local governments have negotiated on the public’s behalf as a condition of granting a private homeowner a building permit. Notwithstanding these public easements, lots with unspoiled beachfront views remain tremendously desirable for California homeowners. Many of these homeowners have done their best to create the misimpression that the wet sand behind their homes is closed to the public.30 To that end they have posted “no trespassing” signs on the beach. Even though members of the public would not have been trespassing had they ignored the signs, the signs have served their purpose, preventing many members of the public from using the wet sand portions of the beach behind privately owned homes. As a result, the California Coastal Commission has had to fight back, issuing violation notices against these owners and mandating the removal of these signs. That the commission felt compelled to act in these circumstances underscores the point of this discussion, which is that exclusionary vibes can be quite effective exclusionary devices, even when those vibes merely amount to bluffing by the owner.31 A “beware of dog” sign might work almost as well as a fence, even if there is no dog. Massively Multiplayer Online Games (MMOs) Software developers seeking to create interactive video games understand that attracting the right sort of gamers can make the difference between a profitable and an unprofitable business plan. To that end, game developers spend a great deal of time thinking about how games ought to be marketed, who the core audience is, and what sorts of messages will attract them. Many of the most popular MMOs are aimed squarely at young men, but even within this group, there will be different audiences. Some teenaged boys on vacation from school will have the capacity to play games for tens of hours in a row, whereas older teens or young adults may have outside responsibilities preventing that level of commitment. Game

EXCLUSIONARY VIBES

designers often try to design a game so as to appeal to either casual or hard-core gamers, reasoning that casual gamers will encounter frustration in a virtual world where the other players have too much time on their hands and can acquire skills and treasure at a pace that the casual gamer cannot match. Hard-core gamers similarly might prefer a virtual environment populated by fellow hard-core types, as a universe with too many casual gamers may present an insufficient challenge. Designers thus assemble packages of game features and marketing strategies designed to reach the desired audiences. “Games like EverQuest and Star Wars Galaxies are designed specifically to reward people with limitless amounts of free time: teenagers, adults with no jobs or kids, and people who don’t eat.” Games like World of Warcraft and City of Heroes, by contrast, are pitched at “a busy adult [who] can play without signing his own divorce papers.”32 The hard-core games require players to team up with fellow players in guilds if they hope to advance, but the casual games provide ample rewards to individualistic players who do not want the hassle of coordinating schedules with teammates. World of Warcraft even gives players rewards for not playing all the time—its “rest bonus” allows players who have logged off for a few days to advance through levels at a faster clip. And City of Heroes consciously markets itself toward those who find World of Warcraft and other Dungeons and Dragons–inspired games “too nerdy.” It replaces wizards and orcs with superheroes, “admittedly, still a bit geeky, but it’s nowhere near the hail-fellow-well-met medievalist vibe of World of Warcraft.”33 Exclusionary vibes, in short, are an important part of the virtual world precisely because game players are not indifferent about the characteristics of their fellow gamers in multiplayer environments. As more human activity moves online in the years ahead, we can expect to see the architects of the virtual world employing exclusionary vibe strategies with some regularity. With individuals’ online profiles becoming increasingly detailed, some virtual environments may find it worth their while to experiment with fine-grained bouncer’s exclusion too.

The Inadequacies of Exclusionary Vibes Exclusionary vibes may be ineffective at creating homogeneity if too many people whom the landowner would prefer to exclude are

53

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

54

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

oblivious to the signal, are poor at self-assessing, or have contrarian instincts. Exclusionary vibes will also be a poor device for constraining entry by those who are willing to endure ostracism and other penalties upon joining a community because the gains from entry are sufficiently large. For example, in the context of common-interest communities like condominium associations, existing members have strong incentives to ensure that new entrants will stay current on their mortgages and continue to pay monthly assessments. Properties in foreclosure and pre-foreclosure often fall behind on monthly assessments, with the result being that the solvent condominium owners have to pay higher assessments or suffer service cutbacks when some of their neighbors are delinquent. Condominium owners and managers therefore have incentives to screen out those most likely to default on mortgage and assessment obligations. But prospective purchasers who know that their own credit history is shaky have an incentive to buy into the units where their fellow owners are the best credit risks. After all, a condominium owner barely making his payments will have difficulty contributing to the shortfall generated if his neighbors are delinquent in paying assessments. In such environments, exclusionary vibes may be helpful, but insufficient, as a strategy for discouraging those with risky credit histories from buying into a common-interest community. Alternatively, exclusionary vibes may be too controversial if they are noticed and denounced by third parties who object to the content of the exclusionary message, asserting that such a message implies second-class citizenship for community members who are targeted for exclusion. In such instances, a landowner may seek an exclusion strategy that is both more effective and less in-your-face than an exclusionary vibe. Exclusionary amenity strategies present an attractive alternative.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

4 EXCLUSIONARY AMENITIES

Because we lack an adequate vocabulary to discuss the phenomenon I describe in this chapter, we must begin by defining some terms. A “club good” is classically defined as a resource for which “the optimal sharing group is more than one person or family but smaller than an infinitely large number.”1 Think of a library, or a security guard, or a swimming pool, not a refrigerator, or a toothbrush, or a calzone. “Exclusionary amenity” is a term I will coin here and define as a type of club good that is paid for by all members of a collective body, at least in part because willingness to pay for the club good functions as an effective proxy for other desired membership characteristics. In the residential setting, exclusionary amenities function to engender homogeneity among neighborhood residents with respect to any particular characteristic and prevent the neighborhood’s population from reflecting the heterogeneity that exists in the larger community. Demand for exclusivity helps fuel demand for an exclusionary amenity, along with inherent demand for the club good itself. Although not all club goods entail social interactions among fellow users, exclusionary amenities often do, for reasons that will be explained shortly. To function as an effective sorting device, an exclusionary amenity must be both relatively expensive and relatively visible. If the club good in question is too cheap, then the decision to join a particular community might not be affected substantially by its presence. A “cheap” club good may engender homogeneity through the operation of focal points, but it will not have any sorting effects. If, on the other hand, the club 55

56

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

good is relatively expensive, such that an undesired residential purchaser will conceptualize it as a high differential tax without any associated benefit, then it may convince the undesired purchaser to buy a home in a community that does not provide the club good in question. Similarly, a club good that is invisible or that does not predictably attract purchasers with particular characteristics will not operate as an effective focal point.2 Homogeneity will result from sorting and focal point mechanisms acting in concert. To consumers about to make the most important investments of their lifetimes, the synergy between sorting and focal points may prove critical, and this may explain the preference for an expensive club good over a cheap focal point alone. To the extent that focal point messages are misinterpreted, see their meaning change over time, or reach an audience without particularly widespread preferences for homogeneity, the presence of an expensive sorting device will be a critical guarantee that a homogeneous population will arise in the first instance and be maintained through multiple generations of buyers.3 Exclusionary amenities are rarely employed in circumstances where more straightforward mechanisms for exclusion are legally permissible and normatively uncontroversial. For example, as we saw in chapter two, residential communities in the United States are permitted by law to discriminate against convicted sex offenders who present high risks of recidivism.4 The legality of overt discrimination renders it unnecessary for a community to invest in exclusionary amenities that would be attractive to those who are not sex offenders but unattractive to sex offenders.5 Instead, communities use covenants or even local ordinances to exclude sex offenders.6 Similarly, when a developer seeks to fill a market niche by creating a common interest community devoted to housing members of a politically disfavored group, employing exclusionary amenities would be overkill. The cheaper alternative of a focal point alone should suffice to establish residential homogeneity within the common interest community. Thus, the Palms of Manasota, the nation’s first retirement community for homosexuals, need not invest in exclusionary amenities to keep heterosexual retirees from residing there.7 Anti-gay sentiment and heterosexual seniors’ discomfort with homosexuals are sufficiently powerful to prevent integration.

EXCLUSIONARY AMENITIES

When club members or real estate developers have a preference for excluding members of protected classes, the number of available options shrinks. For example, African Americans and members of all other racial groups are protected by various laws designed to combat discrimination in the housing sector. Such laws reach not only refusals to sell or lease, but also the ability of landlords or sellers to advertise in a racially discriminatory manner.8 As we saw in previous chapters, this body of law substantially constrains a developer’s choice of human models in housing advertisements by imposing liability on landlords whose advertisements feature exclusively Caucasian models.9 Indeed, in some respects, housing advertising is more tightly regulated than the sale or leasing of housing. For example, the FHA permits “mom and pop” landlords to refuse to lease certain apartments to tenants on the basis of race, but bars those same landlords from advertising their discriminatory preferences with respect to said apartment.10 Deprived of “efficient”11 tools of discrimination, such as racist refusals to deal or advertisements, those with a preference for discrimination may explore less precise strategies that the law does not proscribe.12 Exclusionary amenities may become a viable option under such circumstances.

Comparing Private Goods Self-sorting occurs in many contexts. Developers might distort the population of a new housing development by providing larger-than-average kitchens (attracting gourmets) or minuscule kitchens (attracting those who do not cook). That said, there are two critical differences between self-selection through these private goods and self-selection through club goods. The first distinction is sociological. Club goods often involve social interactions among the members who are entitled to use them.13 Private goods, by contrast, typically involve more limited social interactions. In a neighborhood comprised entirely of quiet shut-ins living in single-family homes, homeowners probably will not care that much about the characteristics of their neighbors. As interactions among neighbors increase, we can expect that homeowners will care more about the characteristics of their neighbors. Club goods often become a locus of social activity within common interest communities, offering additional dimensions in which

57

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

58

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

interactions can occur. For that reason, one might expect that people will pay a greater premium for desirable neighbors in a community offering many club goods than they would for desirable neighbors in a community offering no club goods.14 One reason why racial segregation is a public policy problem stems from the connection between residential propinquity and the composition of individuals’ social networks.15 Residential segregation helps explain the segregated nature of social interactions in public schools, political gatherings, and some workplaces.16 It may explain a significant portion of the economic, safety, public health, and educational gaps between African Americans and Caucasians in contemporary American society.17 Neighborhood residential racial segregation is also associated with declines in generalized trust, an economic resource that drives people’s willingness to cooperate economically or socially with strangers.18 More troubling still, residential segregation is strongly associated with adherence to negative racial stereotypes, and selection effects only explain part of the heightened animosity toward minorities in overwhelmingly white neighborhoods.19 In short, if there were no social interactions among neighbors, then it would be hard to get upset about residential segregation. Residential segregation is a public policy concern precisely because we know that interactions among neighbors are often frequent and take on substantial political and economic importance. The second distinction is economic. Private goods are excludable. Hence, where the law sees no variation in kitchen sizing, it might examine the costs and benefits of permitting variance, 20 and perhaps mandate variance if the cost-benefit calculus suggests that an invidious motive is at work. Club goods present more difficult issues. With those goods, there may be a very good reason for requiring that each individual contribute toward the good in question. In the absence of such a mandate, residents who value the good could have strong incentives to try to free ride on their neighbors’ contributions. The strength of this justification for mandatory membership can provide excellent cover for bad acts.21 Thus the legal system usually will have a great deal of difficulty discerning which club goods are motivated by a well-meaning desire to solve a collective action problem and which are motivated by more nefarious objectives.

EXCLUSIONARY AMENITIES

Exclusionary Amenities in Action To date, the discussion has been rather abstract. Are there realworld instances of developers using exclusionary amenity strategies? An example from the Washington, D.C., suburbs suggests an affirmative answer. At the very least, this example shows that developers are conscious of the ways in which the presence or absence of communal amenities can deter certain groups of undesired residents from joining a new common interest community, and that targeted consumers understand those messages.22 Falls Church, Virginia, like many suburban communities, has had trouble keeping its tax burden low while maintaining high-quality public schools for its residents.23 One way of satisfying both objectives involves trying to limit the development of new housing that is attractive to families with children. To that end, the Falls Church government permitted Waterford Development to build Broadway, an 80-unit condominium, but gave the developer a financial incentive to ensure that no more than eight school children moved into the complex.24 For the ninth child living in Broadway, and every additional child beyond nine, the developer would have to pay Falls Church $15,000.25 The developer agreed to pay such fees for the first five years of the development’s life.26 A Washington Post article described the Broadway developer’s response: The president of Waterford Development, Jan A. Zachariasse, said he was happy to accommodate the city to win approval of the building, which is under construction on Route 7 at the center of the city. Coming in under the eight-child ceiling was easy, he said, because a building’s demographics can be shaped simply by choosing the right amenities. The Broadway, for example, has a cozy library and a clubroom with a billiard table and bar. It does not have a playroom. . . . Once the deal was signed, “I then could steer the project in a certain direction to maximize or minimize the number of children,” Zachariasse said. “You didn’t have to be a brain surgeon to decide which way to go.”27

59

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

60

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

The developer provided a library and bar, but failed to provide a playroom, making the condominium more attractive to childless residents and less attractive to families. A real estate agent who sold units in the development noted that families with many children never even inquired about living in the Broadway.28 It should not be particularly surprising that developers understand how to use exclusionary amenities. The more surprising aspect of this story is Zachariasse’s willingness to discuss his actions and motivations so candidly with a Washington Post correspondent. Zachariasse later regretted his candor, no doubt, when the Department of Housing and Urban Development launched an investigation into Falls Church and Waterford Development for violating the FHA by intentionally discriminating against families with children.29 The investigation ultimately resulted in a settlement, whereby Falls Church agreed to alter the way in which it collects school impact fees from developers, and the developers agreed to devote $120,000 toward a fair housing partnership that would provide training for the developer’s employees to avoid further discrimination against families with children.30 Following this investigation and settlement, one expects that developers will be more tight-lipped when discussing the motivations behind their provision of amenities in residential developments. This raises a serious problem. How are agencies charged with enforcing antidiscrimination laws to ensure that the laws are not thwarted through exclusionary amenities strategies once developers learn from Zachariasse’s mistake and instead offer pretextual but plausible explanations for the provision of exclusionary amenities? There are two reasonable responses to this question. One possible, and perhaps appropriate, response is to do nothing. That will sometimes be the optimal approach in light of the danger that the cure for exclusionary amenities will be even worse than the disease. Doing nothing is an unsatisfying approach, however, in those instances where developers undermine antidiscrimination laws that reflect important normative commitments. A second possible response is to try to identify club goods that seem particularly susceptible to exclusionary strategies and devote careful scrutiny to developers’ use of those goods. In the section that follows, I will identify a few trends in the residential golf course industry and raise

EXCLUSIONARY AMENITIES

the possibility that residential golf courses sometimes have functioned as exclusionary amenities, with African Americans as the undesired group targeted for exclusion.

Golf and Race in the United States During the 1990s, one could predict with a high degree of accuracy a person’s race upon learning that he or she played golf. Among warm-weather leisure activities attracting 25 million or more participants, golf stood out as the most racially segregated. From 1994 to 1995, 27.7 million Caucasian Americans played golf—approximately 16.9 percent of all Caucasians aged fifteen and older. 31 By contrast, only 900,000 African Americans played golf during that timeframe, representing just 4.2 percent of the African American population.32 After adjusting the size of these groups to reflect the general population of the United States, we see that 93.4 percent of all golfers were Caucasian, 3.1 percent were African American, and 3.4 percent classified themselves as “other,” a group that includes Hispanics and Asian Americans.33 African American golfers played fewer rounds of golf than Caucasian golfers did, which skewed the participation data even further.34 The data suggests that, during the 1990s, golf was a substantially better proxy for race than income and a somewhat better proxy than household wealth.35 That differential is critical. After all, if income provided a better proxy for race than golf participation did, those interested in residential racial homogeneity could use large lot sizes or occupancy restrictions to exclude African Americans. This strategy—referred to in the literature as “exclusionary zoning”—is well documented and widely practiced. But once substantial numbers of African American families achieve higher incomes and higher wealth, exclusionary zoning strategies lose their effectiveness. Notably, during the 1980s and 1990s, the United States saw a substantial exodus of African Americans into the suburbs.36 Given the illegality of alternative discrimination strategies, construction of an expensive, racially polarizing amenity may have provided the next most effective strategy for keeping these upwardly mobile African Americans out of particular communities. Golfing facilities constituted an especially attractive exclusionary amenity for developers during the 1990s because it was difficult to find any activity

61

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

62

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

in which participation was as racially polarized as golf. First, other landbased, warm-weather sports were far more racially integrated. For example, African Americans represented 13.6 percent of joggers, 8.2 percent of bicyclists, 15.5 percent of baseball players, 19.1 percent of basketball players, 8.3 percent of soccer players, and 12.6 percent of volleyball players.37 Even tennis, stereotypically a leisure activity with low levels of African American participation, attracted a rather integrated playing population. Fully 8.2 percent of tennis participants were African American, and participation rates were not starkly different among the races.38 Second, sports that exhibited the same level of racial segregation as golf tended to be either extreme, aquatic, or snow-based, most of which are far less popular than golf. The only warm-weather sport with a greater percentage of Caucasian participants was water skiing, which attracted approximately half as many participants as golf did, 94.4 percent of them Caucasian. Motorboating attracted more participants than golf, but was slightly less segregated, with 92.5 percent of participants identifying as Caucasian and 3.3 percent of participants identifying as African American. Rock climbing exhibited a similar skew but drew only 7.5 million participants in 1994–95.39 Similarly, 94.9 percent of cross-country skiers were Caucasian, but the sport drew less than 7 million participants.40 Third, the nature of golf renders it a more attractive exclusionary good. Golf courses are quite expensive to develop and maintain (unlike, for example, rock climbing walls);41 they can be built in virtually any climate or locale (unlike cross-country skiing courses or marinas); they can be enjoyed by virtually any age demographic (again, unlike rock climbing walls); and they do not generate substantial noise pollution (unlike marinas that house motorboats).42 Finally, golf was historically associated with racial exclusion and played at country clubs that had discriminatory membership policies.43 As a result, golf has an “image as ‘a white man’s game.’”44 To the extent that communities wished to employ racially discriminatory selection mechanisms using exclusionary amenities, golf presented the best opportunities.45 Given the racial dynamics of golfing in the United States, a residential development built around a high-quality, mandatory membership golf course would have attracted two types of residents: avid golfers (who would be overwhelmingly white), and people with a preference for living among avid golfers

EXCLUSIONARY AMENITIES

or other non-golfers attracted to such communities. It is therefore worth investigating the exclusionary amenities hypothesis by examining statistics on golf-course-related residential developments. A residential golf course is a golf course surrounded by residential properties—single-family homes, townhouses, or condominiums. During the 1990s, golf participation intensified,46 and the United States saw a rapid increase in the number of residential golf course developments.47 By 2000, 40 percent of current golf course construction was residential, and the growth rate of residential golf courses far outpaced the growth rate for real estate developments in general.48 In Florida, which has more golf courses than any other state, as many as 54 percent of golf courses were residential.49 It would be inappropriate to assert at this juncture that the exclusionary amenity phenomenon I have identified is largely responsible for this boom in residential golf courses. Alternative explanations cannot and should not be discounted. However, an investigation into the growth of residential golf communities reveals several intriguing data points, all of which are consistent with the hypothesis that exclusionary amenities strategies were responsible for some of the changes and growth in the residential golf course market. The first intriguing data point concerns the shifting mix of mandatory golf course memberships and optional memberships offered to residents of residential golf communities. Early residential golf course developments followed a particular financing model: Those who purchased residences in the development were obligated to purchase “equity memberships” in the adjoining golf course.50 In this arrangement, all homeowners would pay for the development and maintenance of the course, regardless of their utilization of it. In the mid- to late-1990s, however, the market shifted somewhat, with developers increasingly embracing semi-private golf course developments, where membership is optional among homeowners and members of the public can play for a daily use fee.51 Two groups of golf courses did not shift away from equity memberships: high-end courses played by the very wealthy, and courses located in areas with large African American populations, such as Broward and Miami-Dade counties.52 For wealthy homeowners, mandatory golf course membership might have functioned in the same way that the cooperative structure functioned in Manhattan. Wealthy people can afford to pay a premium

63

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

64

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

for the perceived benefits of exclusionary policies and are happy to do so. Instead of paying more for apartments and association governance via the cooperative corporate form, these Floridians might have been opting for a luxury amenity that effectively excluded those who were unwilling to pay substantial amounts for a world-class golf facility. To complete the story, consider the second intriguing data point: many purchasers who buy into residential golf courses do not play golf. This phenomenon of non-golfer households in residential golf communities— including those with mandatory membership policies—has been widely noted in golf industry periodicals.53 To be sure, not all of these people are overt racists or segregationists.54 Indeed, it is likely that many of these non-golfing residential golf course dwellers are willing to pay a premium because they enjoy the open space or low densities offered within golf course developments.55 That said, real estate appraisal research suggests that golf course views provide only one-third as much of an increase in real estate values as views of a creek or marsh.56 Artificial lakes and waterways are cheaper to build and maintain than golf courses and add similar value,57 although they are less of a mainstay of new real estate developments than golf courses.58 Rather surprisingly, proximity to a golf course appears to add less to residential property values than it does to commercial, industrial, institutional, or agricultural properties.59 In short, golf courses qua golf courses add less value to nearby or adjacent residences than one might expect. A desire for open space plainly was not driving all the demand for residential golf courses among non-golfers. Is there any evidence for more insidious explanations? The marketing data appears to suggest that many non-golfer residents of residential golf courses find the homogeneous nature of these communities’ populations appealing. D. Robert DeChaine has conducted the only systematic study of the ways in which residential golf communities market themselves.60 DeChaine noted “recurring themes emphasized in the persuasive sales appeals for golf community property. These themes included focus on the ‘purity’ of the community; the privacy and exclusivity of community membership; the safety, security, and serenity of a lifestyle removed from the maddening crowds; the prestige of the golf course as a community focal point; and the sense of freedom afforded by spacious property and surroundings, among others.”61 Marketing materials

EXCLUSIONARY AMENITIES

certainly discussed the quality of the golf courses at length,62 but DeChaine appeared to notice as much, if not more, emphasis on the exclusivity of golf courses behind gates, membership rules that limited outsiders’ access to the property, and the homogeneity of the community’s residents.63 Advertisements for mandatory membership golf communities sometimes provide not-so-subtle exclusionary messages. For example, Harbour Ridge, a residential golf community in Stuart, Florida, described its community in the following manner: Harbour Ridge Yacht & Country Club is a warm and friendly community of 695 families. Every resident at Harbour Ridge is a member of the Club, thus ensuring universal interest in the care and integrity of the community and the club. Members come from every section of the United States, Germany, England, France and many other countries. They bring with them the traditions of some of their nations’, and the world’s, great golf clubs. Members embrace traditional values and are known to jealously guard their privacy and comfort. Harbour Ridge’s advertisement seemed evocative enough to send clear messages to prospective purchasers about the nature of the community.64 Other residential golf communities opt for an even less subtle approach, selecting names like “Magnolia Greens Golf Plantation” or “Sea Trail Plantation.”65 In some ways, this focus on exclusivity in marketing materials should not be surprising. Even if non-golfers were to constitute a small minority of members within mandatory membership residential golf communities, one would expect to see developers working hard to try to attract them. After all, in some sense the golfers within mandatory membership communities free ride off the contributions by non-golfers for course upkeep. Someone who loved playing golf but did not have strong preferences for residential homogeneity or heterogeneity might rationally prefer to live in a community where non-golf-playing mandatory members subsidized his golfing. Easy access to tee times and little waiting on the course would both be attractive amenities to such golfers. Optional membership residential golf communities, by contrast, should not have been expected to market themselves to non-golfers with a preference

65

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

66

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

for homogeneity. After all, an optional membership residential community faces a “tragedy of the commons” if too many non-golfers join it. The tragedy arises when many people try to take advantage of the views and open space provided by a golf course, but only those residents who are members of the course pay for its upkeep. A residential community can solve this problem only by shifting toward some form of mandatory membership or by permitting nonresidents to use its course, which potentially raises privacy, safety, or traffic concerns for residents.66 There is one final piece of the puzzle. In order for this story to work, it must be the case that Caucasian non-golfers seeking racial homogeneity understood the demographics of golf participation. Ideally, we would be able to access data from the 1990s about white non-golfers’ perceptions of who plays golf. Not surprisingly, however, no one ever thought to ask such a question. That said, James Loewen’s fascinating book on residential exclusion in the United States notes that, at least in the context of retirement communities, both whites and blacks understood the connection between mandatory membership golf communities and residential racial homogeneity.67 Even if there were no intentional discrimination associated with the bundling of golf with residences, the popularity of bundling residential developments with participation in a costly activity that exhibits dramatic racial skews should be disconcerting to those who worry about the effects of residential segregation. The foregoing evidence suggests that, during the 1990s, residential golf communities could have functioned as exclusionary amenities, and may have facilitated substantial residential segregation if housing consumers were acting upon widespread preferences among whites for residential racial homogeneity. Namely, such communities would have attracted three types of residents: (1) whites who wanted racial homogeneity; (2) golfers who did not care about racial homogeneity but were overwhelmingly white; and (3) whites who did not care about racial homogeneity so much as a form of cultural homogeneity. This latter group would be happy to live with “assimilationist” African Americans—precisely those African Americans who would make a conscious decision to live in overwhelmingly white neighborhoods and participate in a sporting activity that has historically been closed to blacks.68 These sorting and focal point mechanisms would have been reinforced by the behavior of middle- or

EXCLUSIONARY AMENITIES

upper-income African Americans who did not want to pay for a costly resource that they were unlikely to use, did not want to be the “token” family in an overwhelmingly white environment, or did not want to live in neighborhoods where they would encounter hostility or social snubs from their neighbors. An exclusionary amenities strategy could enable all these effects to operate in unison.

Other Examples of Exclusionary Amenities Before ending this part of the discussion, it is worth noting the possibility that exclusionary amenities might be used as part of a less obnoxious strategy for promoting residential homogeneity. Racial exclusion is, for very good reasons, regarded as more problematic than other forms of residential sorting. Communities sometimes employ exclusionary amenities strategies, however, to achieve innocuous, or perhaps even beneficial, objectives. Exclusionary Religious Goods Suppose the existence of a religious minority scattered within a large metropolitan area. Suppose further that members of this religious minority value homogeneity in matters of faith and behavior, and that they feel a critical mass of believers in a confined geographic space is necessary for the religious community to thrive.69 In such a setting, one might expect to see the community try to limit the entrance of nonbelievers into the community. For example, a homeowners association might record covenants barring property sales to people who are not members of the religious community in question. Alas, as we shall see in chapter six, such restraints on alienation will be invalidated by courts as contrary to public policy.70 Reliance on exclusionary amenities may provide an alternative strategy. In such a scheme, the community would place a large religious temple at the center of the community and require all homeowners within the association to share the expenses and burdens of the church’s upkeep. This temple could function as an exclusionary amenity if some of the community’s members did not plan to attend it, but only wanted to live among churchgoers. Because an exclusionary amenity merely taxes incoming property owners

67

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

68

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

who do not share the faith, without restraining alienation to them outright, such a financing scheme arguably would not violate public policy.71 After all, covenants and equitable servitudes restricting religious institutions from common interest communities have long been deemed enforceable, based on pro-contract and state neutrality rationales that logically could be extended to cover mandates that homeowners subsidize resident religious institutions.72 The scenario I just outlined is not a hypothetical one. Developers in Collier County, Florida, recently used the exclusionary amenities strategy to create Ave Maria Township, a place some are calling “America’s first gated Catholic community.”73 Because marketing the for-profit development exclusively to Catholics is illegal, developers have tied the development to Ave Maria University, a Catholic institution of higher learning established by the founder of Domino’s Pizza, Tom Monaghan.74 Besides noting the development’s proximity to the new university and its many resources, Monaghan describes a “stunning church in the center of town” and private chapels “within walking distance of each home,” envisioning “an extremely Catholic” population.75 Ave Maria fundraising letters noted that at “the intersection of the town and University will be the focal point of both—a beautiful Catholic Church.”76 Another solicitation from the university said that “when we announced that we would be building a town for Catholics like you, we received 3,000 requests for information.”77 The developers anticipated that the development will be “primarily Catholic,” especially at the outset, but simultaneously stressed that they were “not going to discriminate or market to Catholics.”78 In clarifying remarks, a representative of the developer insisted that Ave Maria “was not going to be a Catholic town. It was to be open to all.”79 The desire among Ave Maria’s planners to create Catholic homogeneity without physically turning anyone away were apparent from early on in the life of the project. A 2002 letter of understanding executed between Ave Maria University and the Barron Collier Companies, which agreed to develop the town, provided that the town, “in its character, ambiance, restricted covenants, zoning, etc. shall allow no public activities which are offensive to traditional Christian values or which might represent a scandal to Catholic and Christian sensibilities. Thus, no topless bars, abortion clinics, ‘adult’ bookstores or the like shall be permitted. . . . The partnership shall donate

EXCLUSIONARY AMENITIES

land for public and private schools (and in the case of Catholic schools, land for churches that will help support such schools) in such size and location as may be appropriate for forecasted demand for such schools.”80 Note the combination of strategies used to substitute for the unlawful bouncer’s exclusion of non-Catholics. Exclusionary vibes that will only appeal to Catholics, a lack of adult-oriented amenities that non-Catholics might be willing to patronize openly, and a donation of land for religious purposes commensurate with the (likely high) demand for parochial education among parents. If the character of Ave Maria is indeed overwhelmingly Catholic, it will underscore my argument in this book. One can create a homogeneous residential development without any targeted marketing or overt discrimination. Of course, it may take a while to evaluate Monaghan’s prospects. Ave Maria Town has been battered by the recent real estate bust, which has hit Florida particularly hard. Only 500 of its projected 20,000 residents have materialized, prices on new homes have been slashed up to $100,000, and workers at the residential sales center eye prospective buyers suspiciously.81 It is an expensive proposition, of course, to construct a golf course or religious university at the center of a residential development. So why would someone seeking to achieve residential homogeneity go to all that trouble? Precisely because an exclusionary amenities strategy may work better than exclusionary vibes alone. After all, an exclusionary amenity may be as effective in establishing a focal point as an exclusionary vibe, allowing people with similar preferences or attributes to find each other and live as neighbors. And the exclusionary amenity will provide added punch: a tax that falls most heavily on people who lack those similar preferences or attributes. As it happens, Ave Maria Township residents will subsidize the adjacent university by virtue of their developers having given the university its land for free.82 Because these costs must be passed along to the town’s new residents, homeowners in Ave Maria Township are paying a markup compared to nearby homeowners whose association is not affiliated with an institution of higher learning. A devout, traditionalist Catholic homeowner might be happy to pay this extra charge, perhaps because he plans to make use of the theological books in the university’s library and values

69

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

70

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

THE MECHANICS OF EXCLUSION

proximity to it, or because he wants to live near the sorts of neighbors who would value proximity to such a library. But a non-Catholic Ave Maria homeowner who did not particularly want to live in an overwhelmingly orthodox Catholic neighborhood would get nothing of value in exchange for his higher monthly assessment: he would not use the library himself, and would not particularly care about whether his neighbors used the library or not. If there are otherwise similar neighborhoods surrounding Ave Maria, we should expect to see Ave Maria Township take on an overwhelmingly Catholic character and other neighborhoods take on a relatively non-Catholic character. The result of this plan would be religious residential segregation, achieved with no overt discrimination and an advertising campaign that need not include blatant exclusionary vibes. The differential tax on non-Catholic homeowners in Ave Maria will serve the same focal points purpose as the exclusionary vibe and will further exclude prospective entrants who might have been impervious or oblivious to exclusionary vibes. Furthermore, the presence of the university and the capitalized real estate markup will directly affect the purchasing decisions of several generations of owners. It is important to flag the possible connection to racial segregation here. Religious institutions are quite racially segregated in general, and many congregations are racially homogeneous.83 Because members of a religious community typically value proximity to their place of worship, the presence of a church or temple may, independently, promote racial sorting in the surrounding neighborhood. A developer interested in promoting racial homogeneity in his new development might therefore sell a large plot of land within the development to a segregated congregation on quite favorable terms, and then raise the price of the surrounding homes to recoup this subsidy. In the Ave Maria example, the town didn’t need to structure itself as a club good to achieve homogeneity. Even if Ave Maria lets everyone use its library or parochial school, proximate homeowners will find these amenities more convenient. At the same time, with so many residents using the parochial school, there may be little demand for public school, requiring Ave Maria parents who prefer a secular education to bus their children to an inconveniently distant town. In this way, public goods may also constitute a type of exclusionary amenity. Local public goods, which

EXCLUSIONARY AMENITIES

confer greater utility on proximate citizens, will function in an analogous way to club goods in a homeowners association. Local taxes will simply replace association assessments as a sorting mechanism. As the following example suggests, even non-local public goods can function as exclusionary amenities. Exclusionary Public Goods Although this chapter focuses on exclusionary amenities in residential communities, we should not be surprised to observe the same phenomenon in virtual communities as well. Indeed, participants in various virtual worlds have developed alternative languages with their own grammars and conventions, many of which prove befuddling to the uninitiated.84 Although some of these languages appropriate internal messaging abbreviations that help shorten the length of typed communications, one prominent online language—l33t (“leet”)—is properly understood as facilitating encryption, not communication. As a result, l33t is more cumbersome to use than ordinary American English.85 Efficiency considerations do not explain the proliferation of l33t—using English would be easier for most of the inhabitants of these online communities. Imposing these barriers to entry may maximize welfare for these communities by making participation in certain online communities vexing for a naïve newcomer, referred to as a “n00b” (newbie) by the computer savvy. One purpose of these languages is to marginalize newbies and exclude the virtual riff-raff.86 Newbies can of course learn l33t eventually, but this process will take time, and that lag will encourage the greenest entrants into virtual worlds to spend more time observing and less time typing during their initial forays. L33t thus can function as a means of discouraging those who are non-savvy, impatient, or unwilling to incur substantial languagelearning costs from joining Internet-based subcultures. In the next chapter I will consider in a more systematic way the benefits and costs associated with the various exclusion strategies. In the course of examining why a resource owner might opt for one exclusion strategy over another, I also will explore what the imperfect substitutability of these various exclusionary strategies teaches us about appropriate legal responses.

71

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

PART

II A THEORY OF INFORMATION AND EXCLUSION

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

5 ASYMMETRIC INFORMATION AND EXCLUSION

So far we have established that the various exclusion strategies can substitute for one another, that they are imperfect substitutes, and that the law sometimes recognizes and sometimes ignores this substitutability. Given all of that, it is worth situating ourselves in the shoes of a resource owner so that we can evaluate the tradeoffs among the various exclusion strategies and decide which strategy is optimal in a particular context. Most owners face a choice among the bouncer’s right, exclusionary vibes, exclusionary amenities, or governance. What factors drive the decision among these options? I will discuss them below, beginning with the most important unrecognized consideration—asymmetric information. Information asymmetries arise when the parties involved in a transaction or negotiation have disparate access to pertinent facts. For example, suppose that a California farmer and a local grocer are negotiating over the price to be paid for the farmer’s tomatoes. The grocer has just learned that unseasonably cold temperatures in Florida have destroyed much of that state’s tomato crops. As a result, the grocer envisions a shortage of tomatoes and an accompanying rise in their prices. The grocer knows that tomatoes are likely to fetch $1.50 a pound on the wholesale market next week, but the farmer believes that the recent prices of $1.00 a pound are in effect. The grocer takes advantage of his superior information, offers the farmer just $1.00 a pound, and turns a significant profit thanks to his informational advantage over the farmer. 75

76

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

The same dynamic can of course occur in reverse. A farmer may know that his supposedly organic tomatoes were in fact raised using conventional fertilizer, but he represents them to the grocer as being certified organic anyway, and the grocer pays a premium for organic tomatoes, much of which will be passed along to consumers. Now the farmer has made a tidy profit. Between the two asymmetries, the second is more troubling. The farmer could have learned about the Florida freeze using publicly available information and raised his prices accordingly. If he realizes that the grocer has taken advantage of superior knowledge, the farmer will have an incentive to do more research the next time before selling off his crop. In the second situation, though, there is little that the grocer can do to discover that the tomatoes aren’t really organic. A grocer will generally rely on the good work of organic certification bodies, but even these entities can be fooled, and they will only conduct sporadic audits of the farmer’s growing practices. Little wonder, then, that the law regards the farmer’s misrepresentations as unlawful fraud but the grocer’s savvy negotiating tactics as fair game. Having said that, both the farmer and the grocer may have strong incentives to behave “honorably” in both situations because dishonest conduct, if discovered, can severely damage their reputation, which will hamper their ability to do business on favorable terms in the future. In the domain of residential exclusion, asymmetric information will loom large. Now the parties will often have private information, not about tomatoes or other commodities, but about themselves, and many of the most interesting dynamics in the law will attempt to regulate how that information gets exploited and revealed.

Private Information A central claim in this book is that when potential resource users possess private information about their own attributes and intentions, and when it is costly for a landowner to obtain or verify this information, the landowner will be more likely to employ a non-trespass-based exclusion strategy, such as exclusionary vibes or exclusionary amenities. When, by contrast, information asymmetries are negligible, the landowner will be more likely to exercise the bouncer’s right, a trespass-based exclusion

ASYMMETRIC INFORMATION AND EXCLUSION

strategy.1 This central claim is based on a dynamic whereby sorting among desirable and undesirable third-party entrants to the landowner’s property will be costly to the landowner, sometimes costly enough to warrant delegating that sorting process to the potential entrants themselves.2 A few illustrations will be helpful in explaining the centrality of information costs in a resource owner’s choice among multiple exclusionary strategies. Recall our hypothetical real estate developer who determines that in a world of declining social capital and increasing atomization, pent-up demand exists for a subdivision whose residents are enormously outgoing and social. To make things more concrete, we will assume that if the developer can promise potential purchasers, credibly, that enthusiastic social entrepreneurs will be overrepresented in the new development, homeowners will be willing to pay a 5 percent premium for new homes. How might our developer go about capturing that premium? Suppose, in other words, the existence of an exclusionary premium (like the one found by Schill and his coauthors) and an entrepreneur intent on capturing it. The developer might opt to exercise the bouncer’s right. He could, for example, interview all people who wish to purchase homes in his new development, and only permit those homeowners who seem particularly friendly, gregarious, and warm to purchase units. Conducting these interviews will be time consuming for the developer and for the applicants. Of course, the interviews might fail to sort potential purchasers effectively if the developer exercises poor judgment by mistaking physically attractive or eloquent applicants for extroverts, or is overly responsive to his idiosyncratic notions of what makes someone a social butterfly. Because of these concerns, the developer might ask each potential purchaser for personal references who can then be interviewed, or perhaps a list of social, charitable, or professional organizations with which each potential purchaser is involved. Again, however, relying on these proxies for sociability will be costly and imprecise. References may have incentives to be overly rosy in their assessments, applicants may mischaracterize their involvement in civic society in ways that are difficult for the developer to discover, and extroverted potential purchasers may regard such a rigorous background investigation as unduly intrusive of their personal privacy. The problem with a bouncer’s right approach to this problem, then, is that each potential purchaser possesses private information about his own propensity for

77

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

78

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

sociability, and it may be inordinately costly for our developer to discover that private information.3 What if the developer opts for an exclusionary vibe instead? He might, for example, name his development in a manner that conveys its status as a mecca for social butterflies. He could advertise using testimonials from purchasers who talked about how they bought homes in this particular community because the neighbors they met were so outgoing and actively involved in community affairs. He might follow the lead of some past developers by enticing a few particularly well-known social entrepreneurs to purchase homes in the new community and then invest resources in publicizing these high-profile purchases.4 Now the developer can avoid having to judge whether a particular homeowner is indeed a social butterfly. Rather, he can rely on a focal points strategy, and assume that introverted people will be deterred from purchasing homes in the new community by the fear of feeling left out among their neighbors. The developer no longer needs to spend time or resources discovering prospective purchasers’ private information because they will be making the decision about who joins the community and who does not. In this setting, what game theorists would dub a pure coordination game,5 self-sorting replaces developer sorting. Of course, it may be that exclusionary vibes will not be up to the task. For example, the developer’s ads might be too vague about the development’s focal points, the necessary ratio of extroverts may be too difficult to achieve in light of current market conditions, or too many introverted prospective purchasers may be oblivious to the messages. Alternatively, there might be too much pressure from introverted prospective purchasers who want to buy homes in an extroverted development because they anticipate that they can free ride off the lower crime or higher property value appreciation that could result from their neighbors’ contact with one another. To that end, the developer may feel the need to combine a focal points strategy with some monetary inducement that would help sort out the introverted and extroverted—an inducement that would prevent many free riders from moving into the community but would not deter genuine socialites from buying a home. For example, the developer may devote large swaths of land within the development to playgrounds, dog runs, swimming pools, and clubhouses, then mandate that the costs of creating

ASYMMETRIC INFORMATION AND EXCLUSION

and maintaining these social spaces be assessed against all homeowners in the community. At the margins, these amenities would discourage the introverted from purchasing a home within the development because they will face the prospect of paying hefty monthly assessments for costly amenities that they will never use. Because willingness to pay for social amenities may function as a proxy for sociability, the exclusionary amenities strategy seems like a promising tack for the developer to pursue. And this strategy, like the exclusionary vibes approach, permits the developer to avoid most of the information costs that make the bouncer’s right so inefficient in this context. The developer merely needs to choose some basis for exclusion, and self-sorting will take care of the rest, since potential entrants will have incentives to sort themselves in accordance with the selected criteria. When private information is easily discovered by a landowner, the costs of exercising the bouncer’s right are far lower. Recall chapter two’s discussion of sex-offender-free subdivisions. Robust demand for sex-offender-free developments has coincided with the widespread availability of data about sex offenders, circulated easily as a result of the Megan’s Laws enacted by all fifty states.6 Because of these laws, an individual’s status as a sex offender can be discovered by a residential developer at very low cost.7 The value of excluding sex offenders from a residential development appears to be high and the costs for a developer of sorting among sex offenders and non-sex offenders is now low, we should predict that the bouncer’s right strategy will be prevalent here, as opposed to exclusionary vibes or exclusionary amenities. And indeed, that is evidently the dominant strategy that real estate developers now use. The nonfungible and communal nature of real property renders private information decisive in shaping owners’ strategies and makes exclusion a particularly intriguing strategy. Compare relatively fungible resources, like automobiles or designer clothing. In both cases, potential buyers have private information that is relevant to their purchasing decision. In the former case, car dealers try to obtain this private information to engage in price discrimination, and in the latter case, clothing boutiques elect not to invest resources in trying to discover a buyer’s private willingness to pay, selling garments at preset prices. So the process by which private information is extracted or ignored by sellers is rather straightforward, constrained largely by transaction costs and the possibility of arbitrage.

79

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

80

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

Someone excluded from a fungible resource for which there is a competitive market typically would lack standing to complain about being excluded by any particular seller in court, since the availability of perfect substitutes would prevent him from suffering an injury in fact. In the real property context, by contrast, perfect market substitutes for particular homes and communities do not exist. Every location is unique. Hence exclusion might engender real social harms that cannot be solved by market competition and arbitrage. Moreover, an individual’s enjoyment of a home may be heavily dependent on the identity of his neighbors, whereas someone’s enjoyment of a car or dress is only slightly dependent on the identities of particular individuals who have purchased the same cars or suits. The possibility of substituting various exclusion strategies for governance, which is practically a non sequitur in the context of nonsocial goods such as cars or garments, takes on much greater importance in the real property context. My hypothesis, that exclusionary strategies are chosen based on the presence or absence of asymmetric information, finds support in the economics literature on “tagging.” The tagging literature has focused on the problems associated with government wealth-distribution policies, which require the state to distinguish between those deserving of welfare and those who try to obtain welfare payments even though they are capable of obtaining gainful employment. In their study of nineteenth-century poor laws, Timothy Besley, Stephen Coate, and Timothy Guinnane conclude that new informational burdens brought on by rapid economic change explained the English government’s 1834 mandate that only citizens who resided in government workhouses would be eligible for welfare.8 Although building and staffing these workhouses entailed far higher per-pauper expenditures than the prior regime of locally dispensed monetary assistance, the English government nevertheless adopted the workhouse mandate because of its effectiveness at sorting among the deserving and undeserving poor.9 The sorting device was effective because life in the workhouse was so unpleasant that no one with other options would be willing to endure it.10 During earlier epochs, English workers lived and died in a single county. This stability allowed the state to rely on local vestries to dole out welfare, since those officials knew who was destitute because of economic circumstance and who was a loafer or a drunk. But in the early nineteenth

ASYMMETRIC INFORMATION AND EXCLUSION

century, as “society became increasingly anonymous and market relations supplanted personal relations,” this old system broke down.11 Once the government could no longer rely on the bouncers (vestries) to sort the deserving from the undeserving, it had to turn to something akin to an exclusionary amenity.12 The same kinds of information shocks that would prompt the government to rethink its sorting strategies in the welfare context should cause a private resource owner to reevaluate its exclusion strategies in the property context. Before leaving the topic of private information, it is worth mentioning that private information is a two-way street. There may be, on occasion, a few instances in which the landowner has private information about the prospective entrants that the entrants themselves do not possess. Creditworthiness scores are an example of such private information. Many landlords check these credit scores before leasing an apartment to a new tenant,13 and these credit reports may reflect information to which the prospective tenant herself is not privy.14 In such settings, the resource owner’s strategy is obvious: he has the relevant information and the right incentives to use it in a way that maximizes his profit. The landlord will use the bouncer’s right to exclude those prospective tenants whose creditworthiness is deemed too problematic. That said, there may be cases in which potential entrants are poor at self-assessing. For instance, we might imagine the delusional introvert who believes she is an extrovert, or who unrealistically expects to be extroverted if she can just maneuver herself into the right social setting. My analysis so far has been premised on the notion that individuals are almost always better at assessing themselves than third parties are at assessing them, but readers who do not share that supposition can simply supplement my model by adding self-assessment as an additional variable. All else being equal, when self-assessment skills are high, resource owners will be more likely to employ non-trespass-based exclusion, but in instances when self-assessment skills are poor, bouncer’s exclusion will become more common. As a general matter, prospective entrants’ private information will be a concern more often than a resource owner’s private information. After all, vagueness on the part of the resource owner may discourage applicants from seeking entry into the relevant community because they fear being

81

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

82

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

dissatisfied with their fellow resource users or because they are concerned that they will expend resources in an ultimately futile effort to gain admission.15 In that respect, then, resource owners will often have an economic incentive to disclose publicly at least some otherwise private information about the desired mix of entrants. For example, fraternities and sororities do brand themselves with niche personalities so as to attract like-minded rushees, and many Greek houses try to give hints to those rushees who are unlikely to be admitted as pledges, so that they will instead spend their time rushing houses to which they seem better suited.16 To summarize, when potential entrants have private information that the landowner cannot easily discover, it is likely that the landowner will employ a non-trespass-based exclusion strategy. When, by contrast, access to information is roughly symmetrical, the landowner likely will use trespassbased exclusion. In those rare settings in which the resource owner has information about the would-be entrants that the entrants themselves lack, the owner probably will rely on bouncer’s exclusion. Although private information will often be the most important factor in determining which exclusion strategy a resource owner should pursue, it will not be the only relevant factor. In the discussion that follows, I explore factors that may prove decisive in particular contexts.

The Nature of the Game On a conventional economic account, it is not private information that drives the resource owner’s choice of strategies, but rather the nature of the game. Game theorists describe many human interactions as coordination games or conflict games. If all the extroverts in a community wanted to live in one subdivision and all the introverts wanted to live in another subdivision, we would describe the situation as a coordination game. It would not matter whether the extroverts wound up living in subdivision A or B, so long as they were surrounded by similar neighbors. Everyone’s preferences can be satisfied, provided there is some mechanism (a focal point) for telling individuals which subdivision they should go to. If, on the other hand, the extroverts wanted to exclude the introverts but the introverts did not want to be excluded, we would describe the interaction as a conflict game.

ASYMMETRIC INFORMATION AND EXCLUSION

In a pure coordination game, we should expect to see reliance on the communication strategies that economists dub cheap talk.17 Cheap talk tells people whether they should go to subdivision A or B. If, on the other hand, the game is structured as a conflict game, we should expect to see the resource owner relying on the bouncer’s right, because cheap talk is thought to be relatively ineffective when potential users of the resource have conflicting and competitive interests vis-à-vis each other.18 The extroverts will listen to cheap talk telling them to go to subdivision A, but the introverts who want to be surrounded by extroverts will also go to that same subdivision. Thus, one conventional economic account suggests that the nature of the game is the decisive consideration. The central problem with the conventional account is its unrealistic assumptions, at least in the real estate setting. The cheap talk model introduced by Joseph Farrell and others suggests that in a game with some conflict, cheap talk will be ignored by rational actors. But some recent research suggests that even in conflict situations the effects of cheap talk can be substantial.19 If cheap talk influences some potential purchasers even a little bit, then rational actors should pay a great deal of attention to it. I argued before that some people will confuse an exclusionary vibe for an indication that the resource owner intends to exercise the bouncer’s right to preclude undesired types from entering. If a few people make this mistake in an association that is to be governed by majority rule, the development will have the propensity to tip in the direction of desired types. After all, rational actors will assume that once the development has been populated, the majority (of desired types) may gang up on the minority (of undesired types) through onerous governance rules or exclusionary amenities that tax the undesired without offering them commensurate benefits.20 The key point is that residential communities are self-governing once populated, and homogeneous owners can be expected to implement policies that favor the (homogeneous) majority’s preferences. Hence, residential exclusionary vibes are not exactly the sort of cheap talk referenced in the economics literature, and in a subdivision that has not embraced supermajority voting rules, they may well prove quite powerful in a game setting characterized by conflict. Similarly, it is not obvious that exclusionary vibes will always be the exclusion mechanism of choice in a pure coordination game. Even in

83

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

84

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

strong sellers’ markets for real estate, advertising that tries to disseminate exclusionary vibes will be expensive. In some settings, the costs of disseminating effective exclusionary vibes may be higher than the costs associated with exercising the bouncer’s right effectively. The choice among exclusion strategies will depend on these costs and the associated benefits. This analysis shows why it is appropriate to think of resource owners’ actions and potential entrants’ actions in game theoretic terms as a repeatplay game. At time one, private information will drive the resource owner’s choice of strategies, but this choice of strategies also will affect the potential for the entrants to take corrective action at time two or thereafter, and this possibility of subsequent shifting exclusion strategies will alter potential entrants’ incentives at time one. When a resource owner selects an exclusion strategy, she is simultaneously stacking the deck with respect to future decisions that the entrants will make about exclusion and governance. Put another way, an owner influences future governance by controlling present exclusion. But what if potential entrants understand this dynamic too well, and try to engage in coordinated action that undermines the resource owners’ objectives? In the real estate context, blockbusting is the most prominent sort of aggressive collective action. Blockbusting occurs when real estate agents arrange the sale of a few homes in an ethnically homogeneous neighborhood as a means of triggering panic selling and prompting rapid neighborhood turnover, thereby obtaining many commissions.21 Buyers may find participation in endgame blockbusting advantageous because it may allow them to purchase more attractive housing stock than they would otherwise be able to afford. The bouncer’s right provides the best protection against these forms of collective action. For example, a cooperative apartment board typically exercises the bouncer’s right with respect to new purchasers22 and for that reason excels at maintaining the community’s character. Exclusionary amenities offer the next most robust protection against collective action of this sort. Existing homeowners in an exclusionary-amenity community may find themselves saddled with substantial debt to pay off the amenity’s construction, and replacing the amenity with something else after it has been constructed may prove exceedingly costly or contentious if a decent number of original owners remain.23 Hence, an exclusionary amenity

ASYMMETRIC INFORMATION AND EXCLUSION

strategy deters opportunistic collective action by raising the costs associated with neighborhood demographic transition. Exclusionary vibes, of course, provide a poor defense against collective action. Recall that an exclusionary vibes strategy relies entirely on the establishment of focal points among those permitted to enter a property. But a group of like-minded individuals acting in concert could avoid the social or psychological costs associated with entering a property where one is not wanted. And if enough people ignore the exclusionary vibe, then members of the preexisting population will feel like outsiders and may decide to move elsewhere.24

Law There will be circumstances in which a developer’s choice of exclusion strategies is dictated not by information asymmetries or the nature of the game, but by legal regimes that favor some forms of exclusion over others. Sometimes the law merely discourages a particular exclusion strategy. For example, consider the current legal landscape governing bouncer’s exclusion and exclusionary vibes. Courts conceptualize the former as a property right and the latter as an expressive right. Accordingly, government infringements of the bouncer’s right are analyzed under the takings clause of the federal Constitution, whereas government infringements of the right to emit exclusionary vibes are analyzed under the First Amendment’s commercial speech doctrines. Prior to the Supreme Court’s 1976 decision in Virginia State Board of Pharmacy,25 commercial speech was not protected by the First Amendment at all. Since that decision, however, the Court has gradually expanded commercial speech protections,26 and it seems plausible that the Court may eventually scrutinize restrictions on commercial speech with the same framework that it currently applies to core political speech.27 This ratcheting up of commercial speech protection has coincided with a ratcheting down of takings protections,28 such that the takings clause, applied properly, seems to have become a less severe check on the government regulation of exclusion than the First Amendment. In this environment, we should expect to see the government take advantage of this asymmetric judicial scrutiny by overregulating bouncer’s exclusion and underregulating exclusionary vibes.

85

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

86

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

In other instances, government prohibitions dictate a resource owner’s choice among exclusionary strategies. The clearest example of this is my suggestion that real estate developers may have used golf courses during the 1990s as part of an exclusionary amenities strategy to promote Caucasian residential homogeneity. This behavior probably did not stem from the presence of private information. In most cases, a prospective entrant’s race is discernible during a face-to-face meeting. But of course the law prohibits discriminatory refusals to sell, as well as most race-based exclusionary vibes in the housing sector. Yet the law leaves exclusionary amenities largely unregulated. Hence exclusionary amenities are the only lawful racially exclusionary strategies open to real estate developers. Note, however, that if what real estate developers are trying to achieve isn’t so much racial homogeneity as cultural homogeneity (“acting white” or being “WASPy”), then the golf course strategy may be explained with reference to information asymmetries. Cultural orientation may be difficult to discern through short-lived interpersonal interactions. Suppose that Dan is a member of a minority racial group who feels more comfortable among members of a majority racial group. Suppose further that a developer seeks to set up a residential community for members of the majority racial group and could do so without violating the law. If the developer adopts a bouncer’s right approach to regulating access, then it is likely that Dan will be excluded from the development, since his race will probably be discerned easily by the bouncer.29 If, on the other hand, the developer selects an exclusionary vibes approach or an exclusionary amenities approach, then Dan may well wind up entering the community, since he values the prospect of living in a community that consists overwhelmingly of majority racial group members and he may be willing to pay extra for that “privilege.” All else being equal, we can expect to find greater heterogeneity in those communities that employ non-trespass-based exclusion strategies, at least in those instances when the criteria for exclusion can be applied efficiently by a bouncer (because of a lack of private information). On the other hand, it may be that what homogeneity-loving prospective purchasers prefer is not a community that consists entirely of members of a majority racial group. Rather, they might prefer a community that consists entirely of members who accept the cultural preferences and practices commonly associated with majority racial group membership.30 These

ASYMMETRIC INFORMATION AND EXCLUSION

citizens would prefer to live with Dan over a majority-group member who nevertheless embraces minority-group values. Indeed, Dan’s presence may help them feel better about their preferences for neighbors of a homo­ geneous cultural orientation (“some of my best friends are black/Jewish/ gay/etc.”). Thus, the critical question concerns what type of homogeneity is demanded by the market. The answer to that question boils down to the private information considerations identified at the outset of this chapter. If there is market demand for homogeneity across a dimension characterized by private information (such as cultural affinity), the landowner is likely to employ a non-trespass-based exclusion strategy. If the market demands homogeneity across a dimension characterized by easily discoverable information (such as race, typically) then the landowner is likely to employ a trespass-based strategy, assuming the law permits it.

Social Meaning It will not always be obvious to a landowner what the best exclusionary strategy is in a particular setting. The mix of private and public information available to a landowner can create ambiguity. Ties and nearties will arise. How should they be broken? Under these circumstances, a landowner should pay substantial attention to the ways in which differing exclusion strategies might carry with them divergent social meanings.31 Social meanings differ substantially with respect to various exclusion strategies. Bouncers police popular dance clubs, letting the famous, beautiful, and well-dressed people enter, and keeping out the less famous, less beautiful, less well-dressed clientele. This conduct by nightclub bouncers is relatively uncontroversial in the United States.32 But society does not seem to tolerate bouncers at restaurants, so restaurants seeking to establish an exclusive atmosphere must rely on blunt exclusion strategies (for example, high prices or Byzantine systems for allocating reservations); governance rules (“no smoking,” “no shirt, no shoes, no service,” or “jacket required”); or exclusionary vibes (viral advertising, unusual background music, or décor choices). Why is bouncer’s exclusion tolerable in one context, but not in another? The answer plausibly lies in the richness and nature of social interactions among those who enter the facility. That is to say, patrons can expect

87

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

88

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

to interact substantially with strangers in the dance club, but not in the restaurant. Recognizing this, prevalent social norms tolerate more obvious forms of exclusion. Moreover, because the parties’ objectives when interacting with strangers in a dance club will often be romantic, and because exclusionary interactions based on snap aesthetic judgments are de rigueur in romantic settings, what would be offensive in the restaurant setting seems perfectly natural at the entrance to the club.33 But even this analysis is unsatisfying. Bouncers at singles bars are typically just there to ensure that all patrons are twenty-one or older. Although singles bars could become more popular by screening out “creepy” patrons, they typically will not turn away any sober new adult patrons at the door. Instead, they behave in a manner similar to American immigration officials—letting lots of people in and then expelling those who behave badly. As this analysis suggests, our social norms are often mysterious and sometimes irrational. When race, religion, gender, or some other suspect classification is used as the basis for exclusion, the social meaning associated with blatant exclusionary vibes or well-publicized exercises of the bouncer’s right can be deeply problematic. Members of the public may be willing to tolerate subtle exclusion far more than obvious exclusion. Hence, in his study of exclusionary zoning,34 political scientist James Clingermayer found that because arguments about excluding the poor and minorities from middle-class and affluent neighborhoods are politically unpopular, advocates of such separation invoke nonexclusionary rationales for exclusionary policies.“Therefore, justifications that have exclusionary impacts, whatever the intent behind them, are generally couched in terms of neighborhood protection, defense of property values, good planning principles, enhancing environmental quality, promoting historical preservation, and so on. Sometimes, the justification can even be couched in terms of a need for more land devoted to industrial purposes. . . . On [some] occasions, political actors may refer to many values other than the ones that are most dear to their hearts. In doing so, they shift the terms of the debate and manipulate the agenda in such a way as to lead to their more preferred result.”35 There is every reason to think that some of the same considerations play out in the context of exclusionary vibes as in the exclusionary zoning context. To the extent that landowners seeking to establish or preserve homogeneity in their surroundings exercise the bouncer’s right, they will

ASYMMETRIC INFORMATION AND EXCLUSION

point to some criteria less controversial for exclusion than race or religion. Exclusionary vibes often will be more subtle than usual when the basis for exclusion is controversial. And exclusionary amenities may be marketed using code words or proxies for race that are understood by both blacks and whites but avoid the most confrontational aspects of overt racial appeals. As a general matter, then, the controversial and divisive social meaning of exclusionary vibes and visible bouncer’s right activities in the racial segregation domain will help explain the choice of some developers to opt for exclusionary amenity strategies. Even when the bouncer’s right or exclusionary vibes are employed, concerns about social meaning and popular backlashes will modify the nature of the exclusion.36 For example, a landowner excluding some people on the basis of race or religion may seek out an effective proxy or leave his criteria exceptionally vague. As a general matter, the more controversy generated by an exclusionary criterion, the more subtle a landowner will be about the existence and content of that criterion.37 This account of social meaning is, of course, descriptive rather than normative. From a normative perspective, it is not entirely clear why we should differentiate between, say, bouncer’s exclusion and exclusionary amenities, if we suppose that the two strategies are equally effective at excluding prospective entrants. Moral equivalency seems particularly appropriate when the landowners intentionally choose an amenity because of its anticipated exclusionary effects.

Mistakes While concerns about controversial social meanings may tempt a landowner to employ less transparent criteria for exercising the bouncer’s right, or a less obviously exclusionary vibe, there is a countervailing concern. The greater the vagueness, the higher the likelihood that potential entrants targeted for inclusion or exclusion will make errors. These mistakes will arise when parties targeted for inclusion or exclusion fail to realize it. Such errors impose costs on the landowner. Vagueness in the exclusionary criteria behind bouncer’s exclusion will be costly for the landowner in that it will force her to spend substantial time and resources turning away those who do not meet the criteria. On one hand, this turning away of

89

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

90

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

people targeted for exclusion may be valuable for the resource owner. A virtue of the bouncer’s right as an exclusionary strategy is that it provides the landowner with accurate information about the lost economic opportunities associated with exclusion. More specifically, the landowner no longer need rely on guesswork about the gross private costs associated with turning away particular customers who can afford to pay the applicable entrance fee. As a consequence of exercising the bouncer’s right, the landowner necessarily learns more about the customers he is turning away. Such information is not easily discernible if the landowner chooses an exclusionary vibes or exclusionary amenities strategy instead. Bouncer’s exclusion thus has an advantage of providing the landowner with some useful information, and reliance on the bouncer’s right may be particularly attractive under circumstances characterized by substantial uncertainty or boundedly rational landowners. Of course, this additional information may be of marginal utility, since exercising the bouncer’s right does not accurately inform the landowner of the net costs associated with exclusion. Though the landowner knows how many customers she is turning away, she is far from certain how many of her existing customers she would lose if she stopped being so selective about entry. Having said that, an experienced resource owner operating in a stable environment may find the constant rejection of entrants to be costly and annoying. Moreover, if the owner repeatedly turns away prospective entrants without identifying any applicable criteria for entry, she runs the risk that those excluded will become frustrated by having expended resources in a futile effort and will take the law into their own hands as a result. Dampening or obscuring the message contained in an exclusionary vibe imposes a different set of costs on the landowner. Recall that through exclusionary vibes, the owner expects to command a premium by virtue of his successful effort to establish a focal point around which similar people can organize themselves. But the less obvious the exclusionary message, the less confident prospective purchasers or entrants will be in the success of the landowners’ effort to establish a focal point. Moreover, some prospective purchasers or entrants may not realize that they are targeted for inclusion by the landowners’ message, and this will depress the size of the market, thereby further reducing the premium that the landowner

ASYMMETRIC INFORMATION AND EXCLUSION

can expect to command.38 Indeed, as a general matter, we should expect to find a higher percentage of “tone-deaf” entrants in communities whose access is regulated by exclusionary vibes than in those in which a bouncer restricts entry. Because some prospective purchasers or entrants will fail to understand the content of an exclusionary vibe, there may be a substantial portion of the population targeted for exclusion that nevertheless purchases access to the landowners’ community. Most of these people will realize that they failed to perceive an exclusionary message with the benefit of hindsight, and their lack of fit with their neighbors may trigger either buyer’s remorse—with rapid and costly turnover in homes as they decide to move to a community where people are more like them—or efforts to undermine the community’s homogeneity from within—which may trigger the exodus of those homeowners who purchased after correctly perceiving the content of the exclusionary vibe. The same sort of mistakes can occur in the exclusionary amenities setting. Here the problems arise if a potential purchaser or renter underestimates the expenses associated with a costly, polarizing amenity. For example, in the rental environment, a prospective tenant may not know how much of his monthly rent payment is used to subsidize an adjacent amenity. The renter can probably rely on educated guesses here, however, and use the amenity’s apparent quality as a proxy for its monthly costs. In a homeowners association or condominium association, mistake costs will be lower, both because a prospective purchaser will have access to a budget specifying association expenses by line item, and because the prospective purchaser who plans to live in a residence for many years will probably invest more time than the transient renter in discerning where his assessment dollars will be going. But, as the subprime mortgage meltdown of the first part of the millennium has shown us, some home buyers are surprisingly inattentive to critical terms and conditions relating to their purchases of new homes. None of this is meant to suggest that relying on the bouncer’s right will avoid the problem of mistaken admission or exclusion. In an episode that gained local notoriety, a bouncer at Hannah’s, a bar in Washington State’s capital city of Olympia, refused to admit the state’s 61-year-old governor, Christine Gregoire.39 Gregoire told reporters that “the man checking identifications at the door said she couldn’t get in without ID, even when others pointed out that she’s Washington’s governor.” The governor therefore

91

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

92

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

had to go home. According to the bar’s owner, “his 23-year-old part-time bouncer needs more training.”40 Ordinarily, a remotely competent bouncer can be expected to avoid these sorts of mistakes, because even a 61-year-old who looks good for her age will not be mistaken for a 20-year-old. Age is a reliably observable characteristic, give or take a decade, and competent bouncers will do a better job of excluding those too young to drink from a bar than the patrons themselves will if the bar relies on the honor system.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

6 REGULATING THE CHOICE AMONG EXCLUSION STRATEGIES

The previous chapter analyzed the important variables that affect a resource owner’s decision to choose among the three primary exclusion strategies. With a few exceptions, we can expect that resource owners will act to serve their individual interests well in restricting access to their property. Of course, resource owners’ decisions about exclusion necessarily affect the interests of third parties and may implicate broader societal values as well. When a resource owner’s exclusion of prospective entrants generates substantial harms of this sort, intervention by the state may be warranted. In this chapter, I will explore some of the social considerations that ought to guide the law’s response to various exclusion strategies. I will also point to various ways in which the law, consciously or not, alters the incentives for individual resource owners to engage in particular forms of exclusion. Many of the examples in this section refer to “externalities,” activities that an individual engages in that affect third parties. These effects can be of small or large magnitude. If you drove to work this morning, you produced small negative externalities by polluting the air a little bit and contributing a small amount to roadway congestion. If you killed a pedestrian on the way into work this morning, you created massive externalities—wiping out a life, and perhaps making a family destitute, turning a spouse into a widow, depriving a firm of a valuable employee, and taking away a devoted friend from scores of people. Even if the law punishes you for creating these harms—charging you a gasoline tax to discourage driving, 93

94

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

or prosecuting you for reckless driving if you killed a pedestrian— these efforts are unlikely to force you to shoulder all of the social costs associated with your behavior. Externalities can be positive, too. If you plant flowers outside your home that are visible from the street, you are conferring positive externalities on your neighbors and passersby. If you compose a terrific song, you are enriching the lives of others, and it will be difficult for you to capture all of the benefits that you confer. Surely there is someone, somewhere, reading this book who values it more than the amount they paid for it. If so, then the production of this book created positive externalities. (To be sure, since the opposite is surely true too, I am responsible for negative externalities as well.) Externalities are the classic justification for government intervention. Since people do not “internalize” all the negative costs associated with their morning commute, the government may need to tax gasoline or subsidize mass transit so as to more closely align the individual’s interest with society’s interest. It turns out that the choice among exclusion strategies implicates significant externalities, so it is to those sorts of externalities that we now turn our attention.

Symbolic Externalities As noted earlier, certain forms of residential exclusion are highly controversial. Exclusionary actions by a landowner might upset prospective purchasers or tenants who would like to enter the landowner’s property but are deterred or prevented from doing so. Such actions might also impose genuinely felt harms on people who have no interest in entering the landowner’s property but object to the exclusionary device all the same. These sentiments might arise among bystanders because they feel special kinship with the disappointed group of prospective entrants, or because they resent the exclusionary device as a matter of principle. As an exclusionary device becomes increasingly obvious, we might expect that these bystanders will be more offended by its use. This analysis suggests that exclusionary vibes may prove particularly problematic. After all, exclusionary vibes necessitate some form of advertising visible to third parties as a means of enticing compatible people to enter the property in question. While savvy landowners will work hard to target their advertisements to a

REGULATING THE CHOICE AMONG STRATEGIES

receptive audience by selecting the media outlets thought to provide the most sympathetic eyeballs, there will inevitably be some disclosure of an exclusionary message to people who will be offended by such a message.1 From a social-welfare perspective, bouncer’s exclusion and exclusionary amenities will be better strategies when large numbers of people object to the exclusion that is occurring. It might be difficult for bystanders to learn that a landowner is exercising the bouncer’s right in a controversial or problematic manner, unless the landowner for some reason publicizes this fact, a celebrity is involved, or large numbers of people are admitted and rejected in a visible setting. Similarly, one of the advantages of exclusionary amenities is that they can sort people in subtle ways. That subtlety keeps exclusionary amenities off the radar screens of many bystanders.

Misperception Externalities I have already pointed to the possibility that exclusionary vibes might be misunderstood by prospective entrants. Not all of this misperception amounts to an externality, however, since the landowner will suffer economic repercussions if too few people understand his exclusionary message. One set of externalities consists of the costs imposed on prospective purchasers or tenants who are forced to relocate or experience regret after they learn that the property differs from their expectations. A second set of externalities arises when potential entrants reach mistaken conclusions based on the messages implicit in a landowner’s choice of exclusion strategies. Misperception externalities of the first sort will be reduced if the resource owner opts for bouncer’s exclusion or exclusionary amenities instead. In the bouncer’s exclusion case, the prospective purchaser’s misperception is not particularly relevant to the resource owner since the bouncer will prevent the mistaken party from entering the property in any event. In the exclusionary amenities setting, misperception externalities might arise if an individual finds a particular amenity attractive but finds the residential homogeneity that is created by the amenity surprising and unattractive. As a general matter, these forms of misperception will be rare if people who value costly social amenities enough to buy homes that include them have prior exposure to the demographics of that amenity’s user base.

95

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

96

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

Misperception externalities of the second sort also will be most prominent in the exclusionary vibes setting. Indeed, two separate circuit courts have explained the FHA’s peculiar treatment of “Mrs. Murphy” homeowners with reference to these sorts of misperception externalities. Under the so-called Mrs. Murphy exception, owner-occupied housing containing four units or fewer is not subject to the federal FHA’s antidiscrimination provisions regarding residential leases. Congress evidently thought that just as the law permits people to be racist about who they want to marry or invite over for dinner, it ought to let a small-time landlord act on racist impulses with respect to who would live under her roof. Congress, however, did not exempt Mrs. Murphy from liability under the FHA if Mrs. Murphy ran an advertisement stating that she would only rent to Caucasians. So the law lets her discriminate, but it does not let her flaunt her discriminatory preferences.2 To rationalize this counterintuitive result, the Fourth Circuit Court of Appeals noted in 1972 the negative externalities that might result from racist advertising by Mrs. Murphy landlords: “In combating racial discrimination in housing, Congress is not limited to prohibiting only discriminatory refusals to sell or rent. Widespread appearance of discriminatory advertisements in public or private media may reasonably be thought to have a harmful effect on the general aims of the Act: seeing large numbers of ‘white only’ advertisements in one part of a city may deter nonwhites from venturing to seek homes there, even if other dwellings in the same area must be sold or rented on a nondiscriminatory basis.”3 The court noted that Congress might have worried that potential renters would assume, falsely, that if one apartment in a Mrs. Murphy building was advertised as being “for whites only” many other apartments in larger buildings (which are subject to FHA restrictions on the bouncer’s right) were also available only to whites.4 The federal Court of Appeals for the D.C. Circuit embraced somewhat different logic in 1990, noting that permitting Mrs. Murphy landlords to publish discriminatory advertisements “created a public impression that segregation in housing is legal, thus facilitating discrimination by . . . other property owners and requiring a consequent increase in [civil rights] organizations’ educational programs on the illegality of housing discrimination.”5 Both these courts correctly pointed out the possibility that members

REGULATING THE CHOICE AMONG STRATEGIES

of the public might draw erroneous conclusions that undermine public policy objectives if Mrs. Murphy landlords were permitted to advertise their ability to discriminate against tenants without violating the FHA. In short, both judicial panels held that there are public policy justifications for a congressional policy whereby landlords are allowed to exercise the bouncer’s right, but not to disseminate exclusionary vibes. Because of these externalities and the structure of the FHA, a Mrs. Murphy landlord who advertises her racial preferences may be liable even if her own unit is either racially integrated or no more segregated than it would have been in the absence of such an advertisement. Given the Supreme Court’s more recent hostility to various restrictions on commercial speech,6 it is not clear whether these earlier circuit court decisions remain good law. But what is most interesting about the decisions is the recognition by both courts that exclusionary vibes could be powerful and overbroad influencers of decisions as important to individuals as the question of where to live.

Liberty Externalities Free-market societies pride themselves on offering their citizens a wide array of choices, and the variety of homes available in most industrialized societies reflects the diversity of preferences along that dimension. Still, there are constraints on citizen mobility, and many people face very high social or economic costs if they are forced to relocate to another community. Within a given metropolitan area, there may be few, if any, close substitutes for particular residential communities. Under such circumstances, a resource owner’s decision to exclude potential entrants may impose substantial restrictions on the liberty of those excluded. Bouncer’s exclusion regimes that are based on immutable characteristics will be particularly problematic in that respect, since excluded prospective entrants will not be able to gain admission to the property in question by altering their behaviors or preferences. Compared to bouncer’s exclusion, exclusionary amenities and exclusionary vibes score well on this front. Someone whom the resource owner has targeted for exclusion can circumvent the resource owner’s wishes, either by absorbing the costs of being an outlier, or by paying for a costly amenity that he does not value highly.

97

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

98

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

There is a particularly troubling form of the liberty externality, best dubbed the “dumping grounds” problem. The concern here is that there will be some types of people who are so universally loathed and economically disadvantaged that they are denied effective choice among subdivisions and come to be concentrated in the few communities that are willing to accept them or unable to keep them out. In twenty-first-century America, sex offenders seem to represent the least desirable neighbors of all. Indeed, they have been singled out for harsher post-release restrictions than murderers. If this particular dumping grounds problem worsens in the future, courts may crack down on exclusion rights. Recall that after refusing to enjoin a homeowners association ban on sales to sex offenders, the court in Mulligan v. Panther Valley Property Owners Association warned that it might well change its mind in the future if too many homeowners associations followed Panther Valley’s lead, such that “large segments of the State could entirely close their doors to such individuals, confining them to a narrow corridor.”7 Courts considering limitations on other forms of exclusion by landowners have embraced similar reasoning and voiced analogous misgivings about unduly constraining the choices of prospective entrants subject to exclusion.8

Mechanisms for Selective Regulation of Exclusion Rights In symbolic externalities, misperception externalities, and liberty externalities, I have identified three possible considerations that may warrant government restrictions on certain exclusion rights, even in those instances when one suspects the resource owner’s decision to exclude maximizes his private welfare. When the government does elect to intervene, there are essentially two kinds of strategies available. First, the government might attempt direct regulation of the various rights to exclude. We have already seen various selective efforts along these lines: restrictions on the exclusionary vibe but not the exclusionary amenity or bouncer’s exclusion in the Mrs. Murphy context; restrictions on bouncer’s exclusion and exclusionary vibes but not exclusionary amenities in most other fair-housing-related contexts; and administrative crackdowns on exclusionary vibe “bluffs” in the beachfront public-trust context. These legal strategies are straightforward enough. The government permits activities

REGULATING THE CHOICE AMONG STRATEGIES

that it is willing to tolerate and prohibits those forms of exclusion that it deems harmful to social welfare. Less obvious are the ways in which government information policy can alter the incentives for private parties to adopt various exclusion strategies. Yet these information policies may be just as powerful as governmental prohibitions in shaping the incentives of private parties to adopt particular exclusion strategies. As I shall argue in chapters eight and nine, many governmental policies—such as privacy tort laws, as well as government publication outlets like Megan’s Law and the Freedom of Information Act, and government subsidies for the Internet and other information technologies—affect parties’ incentives to exclude. Because the presence or absence of private information is so critical in determining what exclusion strategy a party adopts, government policies that affect the costs of obtaining private information may spark dramatic shifts in the ways resource owners order their affairs. More provocatively, when concerns about negative externalities associated with exclusion are prominent, prohibitions on worrisome forms of exclusion are not the only way to go. Rather, the government might more effectively discourage undesirable forms of exclusion by tweaking information policy.

Applications The foregoing analysis provides an analytical framework that explains why landowners choose one exclusion strategy over another, and why the state sometimes regulates some exclusion strategies but not others. So let us apply this framework to a few real-world cases. Sex Offenders Sensitivity to government information policy helps us understand the recent popularity of homeowners associations’ restrictions on sex offender residency. Prior to the enactment of Megan’s Law, it was difficult for a landlord or real estate developer to learn whether a prospective resident was a convicted sex offender. New Jersey initiated the Megan’s Law trend in 1994, and every state eventually followed suit, prompted in part by congressional legislation that gave the states very strong incentives to adopt their own versions of Megan’s Law.9

99

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

100

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

The key part of various Megan’s Laws is the Internet registry provision. Almost every state now publishes a list of its sex offenders on the Internet.10 Many of these Internet registries do not restrict access in any meaningful way and do not charge for access.11 More recently, efforts have been undertaken to nationalize the database in a way that will allow for the easy tracking of sex offenders who move interstate.12 These days, then, someone’s status as a convicted sex offender is hardly private, in that a landlord or developer can obtain this information about a potential entrant almost costlessly.13 My account of exclusion suggests that as the government reduces the extent to which sex offender status is private information, those interested in excluding sex offenders will rely increasingly on bouncer’s exclusion. This hypothesis is borne out by recent history. Indeed, prior to the enactment of Megan’s Law, sex offender exclusions were practically unknown in common-interest communities, but they appear to have proliferated in recent years. Previously, common-interest communities tried to keep criminals out by holding meetings designed to raise awareness about the threats posed by sex offenders and then publicizing those meetings to the community at large, or by using “neighborhood watch” signs and meetings as a means of signaling to sex offenders and other criminals that they were not welcome.14 Megan’s Law, then, may have caused homeowners associations to substitute bouncer’s exclusion for exclusionary vibes. What’s particularly revealing about the proliferation of sex offender residency restrictions is the relationship between homeowners associations’ restrictions on sex offenders and those same associations’ lack of restrictions on potential purchasers who have committed even more serious crimes (such as murder) or crimes more likely to target proximate strangers (such as burglary and automobile theft). Perhaps the most important explanation for this disparity is the relative dearth of readily available Megan’s Law–style lists for murderers. If this information were available for free, we would expect to see significant numbers of homeowners associations prohibiting the sale of units to murderers, burglars, and car thieves as well.15 As it happens, a few states have begun to make information about all serious crimes freely available on the Internet. We can therefore develop a testable hypothesis: homeowners associations in these communities will

REGULATING THE CHOICE AMONG STRATEGIES

begin enacting covenants prohibiting residency by convicted felons of any kind, at least to the extent that they are permitted to do this by law. I do not mean to imply that state governments enacted Megan’s Law because they wanted to encourage a shift to bouncer’s exclusion in the sex offender context. I actually suspect that exclusion regimes were not on the minds of the relevant elected officials at the time. But this analysis of Megan’s Law underscores the possibility that government information policy might function as an alternative mechanism for influencing the extent to which private parties pursue particular exclusion strategies. When there are public policy rationales for favoring bouncer’s exclusion over exclusionary vibes, but restricting exclusionary vibes is problematic in other dimensions—perhaps because of First Amendment limitations—the government can accomplish its goal by collecting and publishing the applicable private information. Alternatively, the government might subsidize private enterprise to disseminate previously private information more widely. We will discuss this strategy at length in chapters eight, nine, and ten. Racial Discrimination in Rental Housing Since the mid-1960s, the federal government has committed substantial resources to prevent landlords from engaging in racist behavior that promotes racial residential segregation. These efforts, combined with fair-housing enforcement efforts by the states, have probably contributed to declining racial segregation over the past two decades, although other economic and demographic factors no doubt have played a part.16 In the realm of racial discrimination, it appears that the law treats exclusionary vibes with great hostility, treats bouncer’s exclusion with substantial hostility, and leaves exclusionary amenity strategies largely unregulated. As we saw earlier in this chapter, case law interpreting the federal FHA and its Mrs. Murphy exception in particular has recognized communicative exclusion as a form of behavior that should be regulated, regardless of whether that communication is successful in achieving residential segregation. Legal discussions of exclusionary amenity–style arguments have been almost nonexistent. Still, there is some indication that the courts will treat exclusionary amenities and exclusionary vibes differently under the FHA. In the only published case that implicitly addresses a plaintiff’s exclusionary amenity style of argument, the Second Circuit Court of Appeals held that

101

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

102

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

in order to recover under a housing-discrimination-by-amenity theory, a plaintiff asserting housing discrimination must show not only the potential for such an amenity to promote housing segregation, but also that the presence of an amenity actually did produce a segregated residential environment.17 Hence, it was not enough for Orthodox Jewish students at Yale to show that Yale’s exclusively coeducational dormitories, coupled with a rule requiring undergraduates to secure on-campus housing, forced them to choose between following their religious beliefs and paying for a dormitory room that they would never use.18 Surveying this terrain, discriminatory exclusionary vibes alone violate the FHA, regardless of their effectiveness in promoting segregation, but exclusionary amenities do not.19 In some respects, the Second Circuit’s lax approach to exclusionary amenities is puzzling because the same appellate court has taken a very hard line on exclusionary vibes under the FHA.20 But perhaps the court’s differentiation between the two exclusion strategies under the FHA stemmed from a hunch that the negative externalities associated with Yale’s exclusionary amenities were insubstantial. The law’s permissive attitude toward exclusionary amenities probably reflects ignorance about a strategy that has only recently been discussed in the legal literature. The relationship between bouncer’s exclusion and exclusionary vibes might indicate a predominant concern with the symbolic harms associated with racial discrimination, whereby discriminatory statements and messages are conceptualized as a kind of hate speech, but discriminatory outcomes are tolerated if they are shrouded behind sufficiently polite messages that obscure a discriminatory objective. Although exclusionary vibes typically engender more misperception and symbolic externalities than bouncer’s exclusion, the greater liberty externalities associated with bouncer’s exclusion seem especially troublesome. Here, I suspect that the law’s hands-off approach to bouncer’s exclusion in certain contexts is only partially tied to the symbolic harms that are magnified by exclusionary vibes. A fuller explanation for the FHA’s hierarchy no doubt stems from the fact that discriminatory exclusionary vibes are much easier to detect than discriminatory bouncer’s exclusion, particularly in those instances when a landlord rents out only a few units and random statistical variation conceivably explains segregation within the landlord’s building.21 In short, it may well be that the variable costs of enforcement

REGULATING THE CHOICE AMONG STRATEGIES

efforts by the state explain the law’s attitude toward bouncer’s exclusion in the racial segregation context. In the context of racial discrimination, the law has relied heavily on selective prohibitions on the exercise of certain rights to exclude. By and large, the law has not used information policy to shift the incentives for private parties to adopt particular exclusion strategies. The explanation for this observation is fairly straightforward, in that most apartment rentals involve face-to-face interactions between landlords and prospective tenants, and through these interactions, landlords can easily discover the racial affiliation of most potential tenants.22 At present, it would be difficult for government policy makers to render a prospective tenant’s racial status private information. In future decades, as Internet listings for real estate rentals and sales become increasingly sophisticated and communications technologies advance, information policy may provide an attractive avenue for altering landlords’ and developers’ incentives. Religious Exclusion in Housing At the end of chapter four we discussed the use of exclusionary amenities by religious communities. We shall now see that the whole panoply of exclusionary strategies can be employed to promote theological homogeneity: bouncer’s exclusion, exclusionary amenities, and exclusionary vibes. Taormina Theosophical Community, Inc. v. Silver 23 typifies the effort to replace governance with bouncer’s exclusion in a residential community. The Theosophical Society is not technically a religion.24 Rather, it is an association with “three basic objectives: to work towards the universal brotherhood of man; to study and to compare religions, sciences, and philosophies, and . . . to explore the psychical powers latent in man.”25 There were approximately six thousand members of the Theosophical Society living in the United States at the time of the lawsuit.26 In 1967, a nonprofit corporation purchased two tracts of land in California for the purpose of founding a retirement community for Theosophists.27 To that end, the corporation recorded covenants limiting ownership and occupancy of land within the Taormina community to persons aged fifty or over who had been members of the Theosophical Society for more than three years.28 The covenants gave Taormina a right of first refusal to prevent transfers of property within the community

103

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

104

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

that would violate the occupancy or ownership limitations. Note the coupling of a bouncer’s right with reliance on readily verifiable information (age and membership in an affiliated association).29 The Silvers bought land in the community with actual knowledge of these covenants.30 As Robert Silver was under fifty years of age and Esther Silver had not been a Theosophist for the requisite three years, Taormina invoked its right of first refusal. Because Theosophy was not a religion, the trial court held that the covenants did not directly violate any housing discrimination laws.31 The appellate court, however, found that the restrictions on religion did violate the purpose of the antidiscrimination laws, thereby warranting that court’s refusal to enforce the occupancy restriction on equitable grounds.32 The more interesting part of the opinion dealt with the ownership restrictions. The appellate court invalidated Taormina’s as an unreasonable restraint on alienation: In determining whether a restraint is unreasonable, the court must balance the justification for the restriction against the quantum of the restraint. . . . On balance, we find that the covenant restricting ownership of land in Taormina to Theosophists older than 50 is unreasonable. The quantum of restraint in this case is very great. Southern California is a highly desirable place to live and people from all over the country seek to buy property here. In contrast to a vast potential market, the number of Theosophists in the United States is exceptionally small. The purpose of the restriction is to insure that those who settle in Taormina are sincere in their commitment to Theosophy. Taormina was expressly developed as a retirement community for people who share interests in the study of comparative religions and the latent powers of men. While the gathering together of likeminded people may be a laudable goal, such purpose is not sufficient to sustain the heavy burden of alienability.33 Thus, held the court, the exclusion of so many people from the pool of potential purchasers of the homes in Taormina rendered the restraint on alienation unpalatable, however laudable the motivations of the

REGULATING THE CHOICE AMONG STRATEGIES

community’s founders. Had the excluded group been smaller, presumably the restriction would have passed muster. A disturbing implication of analyzing the discrimination issue through the lens of unreasonable restraints on alienation is that religious pluralities and majorities may get more deference from the state in their efforts to exclude outsiders than minorities. But it is exclusion exercised by large religious groups that may be most troubling, both because such exclusion may ensure the majority’s grip on political and economic power and because the history of places like Northern Ireland reveals that deep segregation between large religious groups can promote sectarian violence.34 In economic terms, the members of the Taormina Theosophical Society were trying to benefit from the network effects associated with bringing together people who share interests and values. When everyone at a local café, post office, or general store shares an affiliation with a particular group, the opportunities for conversations of mutual interest are enhanced. There may be substantial gains from limiting membership in a club to law school professors (the American Association of Law Schools), Boston Red Sox fans (the Sons of Sam Horn), or people with high IQs but middling accomplishments (Mensa). The imposition of such a membership criterion facilitates conversation about specialized or advanced topics that would be inaccessible to most members of the general population. Being surrounded by people who think like you do may have certain psychological benefits, as well. An interesting 2010 study by Matthew Brashears presents evidence that for individuals who have a strong religious worldview, the more time they spend with others who adhere to that worldview, the happier and less isolated they feel.35 Brashears hypothesizes that such segregation of social networks might shield individuals from information or stimuli that are likely to make them unhappy. Of course, those are not the only possible benefits of deferring to the covenants at issue in Taormina. Maybe membership in the Taormina Theosophical Society correlates strongly with other attributes that are desirable in a residential setting. Perhaps Theosophists are perceived to be more honest, more neighborly, more engaged, or less boisterous than ordinary Americans. If so, selecting for Theosophist membership might permit a community to devote fewer resources to the kinds of governance regimes necessary to settle disagreements, prevent litigation, encourage

105

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

106

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

civic participation, or regulate noise. Taormina’s ownership and occupancy restrictions thus might engender first-order benefits, like communities of interest, as well as second-order benefits, like better neighborly relations. To be sure, however, there will be costs associated with the Theosophists’ residential homogeneity.36 In evaluating Brashears’s work we should recall that what makes people happy and what is good for them are not necessarily the same thing. Making oneself happy by limiting one’s own exposure to different ways of thinking may well be an economically self-defeating strategy. The costs associated with religious homogeneity, though, are surely lower than the costs associated with residential racial segregation in the American context.37 Once Taormina is barred from enforcing its covenants via bouncer’s exclusion, it must opt for alternative strategies. Might governance mechanisms—regulating what community members do instead of who they are—allow the community to capture the aforementioned benefits? It would appear not. Taormina could enact rules mandating that each resident receive five hours per week of instruction in comparative religion, science, philosophy, and ESP. Fines could be instituted for members who skipped these classes. Moreover, residents who failed to discuss religion, science, philosophy, or psychic phenomena in an intelligent manner when invited to do so by a fellow resident in a public space might face monetary fines. Numerous community rules could be enacted to encourage neighborliness, civic participation, and quiet coexistence, including various neighborhood easements, voting inducements, and noise restrictions. Of course the community would need to set up a comprehensive governance apparatus to determine whether residents were violating these rules and to resolve the inevitable disputes about whether Mr. Silver’s contributions to his neighborhood’s conversations about Eastern philosophy were sufficiently cogent to ward off a hefty fine. And the rules would no doubt need to be fine-tuned to prevent people from circumventing their communal obligations. The limitations on liberty necessary to effectuate such a community would be quite substantial. The reader, no doubt, sees where this is going. It is impossible to design a governance regime that will generate the same level of first-order and second-order benefits as the bouncer’s exclusion regime that Taormina sought to put into place. Any governance regime will be done in by the

REGULATING THE CHOICE AMONG STRATEGIES

enormous costs and complexity associated with designing, interpreting, applying, and enforcing pro-Theosophical rules.38 In this instance, and no doubt many others, exclusion turns out to be more precise and far more efficient than governance. Of course, that does not mean it should be permitted. In light of the legal precedents, what are the options left open to religious orders seeking neighborhood homogeneity, particularly those that, unlike the Theosophists, are classified as religions under applicable laws? They might rely on exclusionary vibes or exclusionary amenity strategies. But exclusionary vibes must be rather subtle, lest an advertising campaign violate the FHA. And the subtler they are, the less homogeneity they will engender. So, to avoid the problems associated with misperception externalities, religious organizations are beginning to turn to an exclusionary amenity strategy, whereby all members of the community must subsidize the creation of a house of worship or religious university. All else being equal, then, efforts to engender religious homogeneity may result in the diversion of more resources to religion-oriented club goods. Hence a ban on religious discrimination in residential sales or advertising promotes the creation of more churches and religious institutions. The case of the Theosophists and their restrictive covenants likely illustrates the challenges faced by a religious-minded minority seeking to achieve homogeneity as a means of creating a critical mass. A subsequent case, Bloch v. Frischholz, suggests that the law struggles with the question of what impediments should stand in the path of religious majorities that take steps that exclude religious minorities.39 In Bloch, a condominium association enacted a rule limiting the placement of “mats, boots, shoes, carts, or objects of any sort . . . outside Unit entrance doors.”40 The rule was in place for three years before the association began citing it as a justification for removing Lynne Bloch’s mezuzah from the exterior of her doorframe. A mezuzah is a small casing (6 inches long or less, typically) containing the text of a Hebrew prayer rolled up inside and attached to the doorframe of an observant Jew’s home. The commandment that Jews display a mezuzah on the doorposts of their homes is articulated in Deuteronomy 6:4–9. The association repeatedly removed the mezuzot that Bloch placed outside her home, in one instance doing so while she and the other members of her family were mourning at her husband’s funeral.41

107

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

108

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

Judaism, unlike Theosophy, is obviously a religion whose members are protected under the FHA. The Blochs therefore sued, alleging that the association’s actions violated the FHA. Chief Judge Frank Easterbrook, writing for a divided panel of the Seventh Circuit Court of Appeals, affirmed the lower court’s ruling against the Blochs. The court suggested that while the association could not enact a policy targeting mezuzot for disparate treatment, it could enact a rule that precluded the display of objects in hallways more generally. To the court majority, the Blochs were requesting that their religion receive special relief from a generally applicable condominium rule, and the FHA does not impose such accommodation requirements on condominium associations.42 Judge Diane Wood wrote a forceful dissent. She argued that the issue of accommodation was a red herring—a rule prohibiting the permanent placement of religious objects on doorframes would be “irrelevant to practitioners of Christianity, Islam, Buddhism, Hinduism, or any other religion, but it is a duty” for Jews.43 The effects of the neutral regulation would impede the free exercise rights of one religious minority, and them alone. Judge Wood wrote that the restriction “operates exactly as a redlining rule does with respect to the ability of the owner to sell to observant Jews. No such person could buy a unit at Shoreline Towers. The Association might as well hang a sign outside saying, ‘ No observant Jews allowed.’” 44 To Judge Wood, the association rule would function like an exclusionary vibe. Although we might more easily characterize the enforced rule as a form of bouncer’s exclusion employing a proxy for non-Jewishness, there are indeed respects in which the vibe framework is the appropriate one—a religious Jew seeking a tour of the building might want to see if her fellow congregants reside in the building. Observant Jews might be struck by the absence of any mezuzot at Shoreline Towers and conclude that they would not be welcome there. Alternatively, knowing the content and application of the association rule could communicate that Jews were unwelcome even to a prospective owner who had never set foot in the building. Judge Wood concluded that the Blochs had produced enough evidence to warrant a jury trial on the questions of whether the association had intentionally discriminated against them on the basis of their religion and whether, by prohibiting mezuzot, the association had as a practical matter evicted the Blochs from their homes.

REGULATING THE CHOICE AMONG STRATEGIES

The conflicting opinions attracted a great deal of attention from the media and elected officials. Reacting to the Bloch case, both Chicago and Illinois enacted laws prohibiting condominium associations from banning the display of religiously mandated objects on homeowners’ doors.45 But such laws would not have been applied retroactively to protect the Blochs. As a result, the Bloch controversy was still a live one, prompting the appellate court to write another chapter in 2009. This time the entire Seventh Circuit Court of Appeals agreed to hear the case and wrote a unanimous new opinion reversing Easterbrook’s opinion. Judges Easterbrook and William Bauer, who had originally ruled in favor of the condominium association, reversed their votes. They did so after re-argument and debate among the judges.46 Now all the judges on the court embraced the idea that the condominium association’s rule, as enforced, could constitute unlawful housing discrimination if it constituted discrimination “in the provision of services or facilities” by the association or if the Blochs could prove that the association “coerced, intimidated, threatened or interfered with the Blochs’ . . . right to inhabit their condo units because of their race or religion.”47 The Blochs could prevail on remand to the lower court if they were able to show that “the Association reinterpreted the Hallway Rules to apply to mezuzot ‘because of’ and not merely ‘in spite of’ the Blochs’ religion. In other words, the evidence must indicate that the Association was not simply indifferent when it reinterpreted the Hallway Rules; the evidence must show that the Association reinterpreted the Rules with Jews in mind.”48 The court found it telling that the Association had ordered its maintenance employee to remove the Blochs’ mezuzah while the Blochs were mourning at the family patriarch’s funeral but that the Association did not remove the coat rack and table that were left in the hallway by the Blochs to accommodate the garments of mourners who would be visiting the Bloch home after the service.49 There is a decent chance the Blochs will be able to prevail in the further legal proceedings, although they do face an uphill battle, given what the courts will require them to show. Bloch and follow-on legislation raise some difficult issues. What if a condominium building employs architecture that effectively limits the ability of homeowners to install mezuzot? Does the law, in other words, preclude the construction of condominium buildings with all-glass entryways whose doors are flush to the wall? Should we really go so far as to

109

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

110

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

A THEORY OF INFORMATION AND EXCLUSION

mandate the construction of doorframes so as to accommodate a religious minority? And while mezuzot are discreet, what about larger display items that may impose on neighbors’ aesthetic tastes to a more significant degree? To make their arguments easier, the Blochs might have argued, as the plaintiffs did in Taormina, that even if the enforced rule did not violate the FHA itself, that law suggests the existence of a public policy commitment to religiously heterogeneous housing units, such that the courts should refuse to enforce the association’s rule under equitable principles. This argument would be grounded in state property law, not federal antidiscrimination law. Setting aside these strategic considerations, however, juxtaposing the respective burdens faced by the plaintiffs in Taormina and Bloch is disturbing. The exclusion of a religious minority from a majority building may well be more troubling than the exclusion of a religious majority from a minority community. Majority status can bring with it access to the most powerful social networks that help the economy, government, and society run.

PART

III DIVERSITY-PROMOTING STRATEGIES FOR THE TWENTY-FIRST CENTURY

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

7 BUNDLED AMENITIES TO REDUCE DISCRIMINATION

We have seen that an exclusionary amenities strategy is neither good nor evil. Rather, it might further good or evil purposes, depending on the particular setting in which it is employed. Normative considerations might cause us to view the exclusionary amenities strategy unfavorably if used by Caucasians to exclude African Americans from an affluent neighborhood, but favorably if used by members of a religious minority that risks losing its identity to establish a critical mass of believers in a particular physical space.1 This discussion of exclusionary amenities raises an additional implication: Inclusionary amenities logically must also exist. The presence of such goods would spark residential heterogeneity, and the absence of such amenities in a community should function in the same way as the presence of exclusionary amenities.

Examples of Inclusionary Amenities An inclusionary amenity is a heterogeneity-promoting club good that does not, by itself, provide enough welfare to the existing residents of a particular community to explain its presence. The inclusionary amenity does, however, make the community attractive to residents who would not otherwise choose to live there. Inclusionary amenities are likely to arise in settings where the members of a community believe that they share undesirable homogeneities, and that the community will be better off if a more heterogeneous resident pool is integrated into the community. 113

114

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

Inclusionary amenities will be adopted, in short, to make the composition of a neighborhood or development better reflect the heterogeneity that exists in the wider surrounding community. For example, student residential buildings on college campuses occasionally acquire reputations as not being academically rigorous, and sometimes these reputations are well deserved. At some point, members of a community may decide that an “animal house” reputation imposes substantial costs on the members, such as diminished access to employment networks, lower status on campus, or unwelcome scrutiny from university administrators. To that end, the members may decide to devote a large amount of scarce public space to a “study room,” and renovate the study room to make it look tranquil and attractive. Although the current residents, and those in the subsequent few years, may use the study room only infrequently, this pattern will change. In time, as successive groups of incoming residents come and go, the presence of the study room might cause more studious students to self-select into the house, and some of these newcomers will eventually start using the amenity. Initially, the study room functions as an inclusionary amenity, but eventually it is transformed into an ordinary club good that provides residents with significant utility quite apart from its inclusionary tendencies. Inclusionary amenity status thus may be a temporary condition. If the strategy succeeds, a collective resource that provides little value to the current generation will eventually provide great value to subsequent generations. In other instances, however, an amenity may be permanently inclusionary—some residents may individually prefer that the study room be converted into a bar, but their desire for a bar is trumped by their interest in attracting studious residents. Anecdotal evidence suggests that inclusionary amenities of this nature are common. Some condominium buildings provide gyms that are under­ utilized by the members, but the space is not converted to higher-utility uses because of a concern that the absence of a gym would send the wrong message to certain kinds of buyers. Similarly, some condominiums maintain party rooms and other social spaces that go underutilized by their introverted residents. The idea here is that incoming buyers may value sociability within a condo’s corridors, but reliable information about sociability is hard to come by for many potential purchasers. The party room may provide a reassuring message to such potential purchasers and,

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

over time, may become a more efficiently utilized amenity through the operation of selection effects.2 Perhaps the most prominent example of the inclusionary amenities approach is the retrofitting of various residential buildings to permit access by the disabled. Prior to the enactment of the Americans with Disabilities Act (ADA) and similar state laws mandating reasonable accommodations for the disabled, numerous building owners voluntarily embraced ramps, elevators, and other accommodations designed to make their buildings more hospitable to handicapped individuals.3 Such voluntary steps were designed to undercut the segregation of the disabled and to permit disabled Americans to interact freely with their able-bodied peers. Indeed, these voluntary steps helped mainstream the disabled, which in turn galvanized them as a political interest group that lobbied for the enactment of the ADA. Inclusionary amenities of other sorts are prevalent in college and university settings. Most provocatively, there is a sense in which affirmative action policies function as inclusionary amenities, as opposed to mere inclusionary devices. That is, it is likely that many sought-after Caucasian college students want to attend a university that has a racially diverse student body. Racial preferences in admission therefore may be designed to attract not only members of minority groups, but also to attract these heterogeneityseeking Caucasians. Were a university to abolish race-based affirmative action, this might not only increase the percentage of Caucasians in the student body, but it might also skew the attributes of those Caucasians in the student body by attracting Caucasians who prefer racial homogeneity or do not care much about racial diversity, while turning off potential applicants who value racial heterogeneity. Similarly, college athletic programs may function as inclusionary amenities. Outstanding academic universities with strong Division I-A sports programs, like Michigan, Stanford, and Duke, may use their college athletic programs to ensure that a wide range of applicants seek admission at their schools. In the absence of high-profile athletic teams, a research university may struggle to attract the proverbial “well-rounded” students who value more than just academic intensity in a learning environment. Local governments use inclusionary public goods to compete for heterogeneous residents as well. In recent years, communities with

115

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

116

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

declining economic bases, like Peoria, Memphis, and Fresno, have begun investing significant resources in the creation of “artist colonies” and other efforts to attract young members of the creative class.4 This effort, inspired in large part by Richard Florida’s influential book, The Rise of the Creative Class,5 is designed to boost economic growth by attracting the young, energetic, and well-educated art and culture lovers who are sought by major employers. Communities across the United States are investing in public goods and club goods that are not terribly appealing to the existing residents. The movement toward magnet schools in urban public school districts reflects a similar dynamic. In many cities, white flight has rendered the population of urban school districts, and the cities themselves, heavily African American and Latino.6 This widespread exercise of the exit option by middle-class whites has imposed real costs on lowerincome populations that lack the resources to exit urban school districts.7 Several cities have tried to counter this trend by investing heavily in selective magnet schools as a means of attracting middle-class parents back to public school systems.8 In communities where the magnet schools rely on aptitude tests or grades to help assign coveted slots to students, the existing public school population may derive little direct benefit from these schools—few children from poor neighborhoods have the credentials to be admitted to selective magnets. Support for these schools may still exist in poorer parts of the city, however, on the theory that attracting middle-class white parents back to the school district will, in the long run, result in an expansion of resources available to all the district’s schools. While such a dynamic is playing out, a magnet school will function as an inclusionary public good. As some of these examples suggest, people concerned about various forms of residential homogeneity should perhaps support the inclusionary amenities strategy. Though they appear to be vastly outnumbered by those Americans who prefer homogeneous subdivisions, a constituency of Americans who want to live in neighborhoods that exhibit genuine racial and economic diversity exists. Consider the young, upwardly mobile, urban pioneers who have been occupying the loft apartments within earshot of Los Angeles’s Skid Row.9 Bernard Harcourt describes the situation:

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

The advocates for the homeless and the non-profit SRO-operators, naturally, want to buy as much real estate on the Row as possible in order to increase the housing stock for low-income tenants and to maintain Skid Row as Skid Row. Oddly, the high-end real estate developers may share this desire to retain the Skid Row flavor. For it is precisely that Skid Row flavor that gives the neighborhood its edginess, that makes it feel like Manhattan—at least, like the old, 1970s, edgy, lower-Manhattan that attracted urban pioneers in the early days of Soho and TriBeCa, of the Lower East Side and the East Village. It is precisely that juxtaposition of high-end lofts and homeless beggars that gives L.A.’s Skid Row a trendy, urban, edgy, noir flavor that is so marketable. To be sure, if the developers succeed in gentrifying parts of Skid Row, at some point the neighborhood will go upscale mainstream. . . . But the point is, right now, the two opposing forces—the real estate developers and the homeless advocates—may not be entirely in head-to-head conflict.10 The preference for heterogeneity that Harcourt describes seems to be an extreme outlier view, at present. But it is interesting to note that for a small subset of home purchasers, the presence of nearby poverty and blight might actually be a selling point. The prevalence of this unusual preference might help solidify the heterogeneous character of a neighborhood subject to conflicting economic pressures like gentrification and decay. Two important points about inclusionary amenities are worth making before proceeding further. First, although it is likely that exclusionary amenities are more common than inclusionary amenities, examples of the latter may be more readily accessible. One likely explanation for this phenomenon is that people are generally quite willing to talk about inclusionary motivations, but reluctant to discuss exclusionary strategies in polite society. Thus, when developers create exclusionary amenities, they usually choose not to discuss their true motivations out of fear of violating antidiscrimination laws or generating controversy. Indeed, their marketing strategies may be aimed at potential customers who, thanks to unconscious racism, prefer racial homogeneity, but would be reluctant to admit that preference to third parties or even to themselves.11 By contrast, inclusionary amenities designed to increase heterogeneity within a residential setting

117

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

118

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

are generally thought laudable, and they may even require substantial publicity if they are to be effective. For instance, if Peoria wants to create an artists’ colony, it cannot simply draw on artists who live in Peoria’s suburbs. Rather, it will need a regional, or perhaps even national, campaign in order to achieve the critical mass of artists who will alter the nature of the community.12 That said, inclusionary amenities often will not be costeffective because of the legality of inclusionary advertisements, which will function as a reasonably close substitute for inclusionary amenities. This situation contrasts sharply with the legal regime governing exclusionary advertisements, which helps drive the use of exclusionary amenities. When should one expect to find inclusionary amenities, then? Perhaps only in those instances where “talk is cheap” and a more expensive investment in inclusion is necessary to attract a heterogeneous audience to a homogeneous community. Second, the determination of what constitutes an exclusionary or inclusionary amenity will be highly context-dependent. It is possible to imagine circumstances under which a particular amenity might exclude in some contexts and include in others. For instance, if citizens in a predominantly poor African American neighborhood decided to replace a dilapidated public housing project with a high-quality golf course surrounded by stylish bungalows, the residential golf course would function as an inclusionary amenity—a resource designed to desegregate a heavily segregated neighborhood—and make its population more reflective of the racial and economic diversity that exists in the United States more generally. Indeed, residents have pursued a similar strategy at the Franklin Park Golf Club in Boston, a racially mixed golf club described as “a large oasis of peace and racial harmony within a generally hostile environment.”13

Inclusionary Amenity Voids Just as inclusionary amenities can be used to attract diverse residents to homogeneous neighborhoods, communities can maintain their homogeneity through the conscious choice to avoid inclusionary amenities or public goods. A desire to avoid offering inclusionary amenities might cause community residents to forego the provision of the communal resources that they would otherwise prefer.14

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

For instance, many communities desire access to public transportation hubs. Even if such hubs are shunned by commuters, who increasingly prefer to drive to work alone, they provide enormous value to those not yet old enough to drive, those too old to drive, and those unable to afford or use motor vehicles of their own. People who drive to work every day may also garner substantial benefits from having bus or subway routes nearby, for example, by freeing up scarce freeway space or making it easier for babysitters, housecleaners, or other vehicularly challenged service providers to reach their homes.15 Perhaps most importantly, proximity to efficient light rail and subway lines generally increases property values.16 Yet many communities are nearly devoid of efficient public transportation, our old friends at Ave Maria among them. The following excerpt is from the university’s website Frequently Asked Questions posting: how far is ave maria university from the beach ?

Ave Maria University is located approximately 30 miles from the beach. are freshmen allowed to have cars on campus ?

No, the university has limited parking spaces available and priority is given to upperclass students. Exemptions may be made and all requests should be made through either the Security Department or the Office of Student Life. i s t h e r e p u b l i c t r a n s p o r tat i o n ava i l a b l e f o r t h o s e without cars ?

While the local public transportation company, Collier Area Transit (C.A.T.), services locations in Naples, there are currently no stops located in close proximity to Ave Maria University’s location in eastern Collier County.17 The beach is nearby (hurray!), but you will have no way of getting there (darn!), at least not as a freshman. Not to worry, though, because car-less people living near the beach won’t be able to visit you either. Part of the resistance to public transportation may stem from concerns about the extent to which such transportation amounts to an inclusionary public good. For example, in the process of planning the Washington, D.C., subway, citizens in various relatively affluent areas opposed the establishment

119

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

120

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

of subway stations because of concerns that inner-city denizens would ride the subways into their neighborhoods.18 Affluent neighborhoods in other parts of the country have done likewise, foregoing otherwise desirable investments in valuable amenities like well-maintained public roads, parks, and even street signs because of fears that such amenities would attract undesirables.19 Exclusionary zoning would be adequate to keep the poor from living in these communities, but an exclusionary dearth of public goods is necessary to keep them out entirely.20 In other affluent neighborhoods, such as the Hamptons, where the incursion of African Americans is viewed as unlikely but the perceived threat posed by Latino immigrants is substantial, homeowners are happy to support the construction of basketball courts on public land, but they fight hard against the creation of soccer fields.21 In some communities, the desire to exclude is sufficiently powerful to overcome the added value associated with transit and recreational improvements. It should surprise no one that similar dynamics arise in the developing world too. In post-Apartheid South Africa, mainly white suburbanites have blocked new public bus routes linking their neighborhoods to the overwhelmingly black townships, angrily insisting that opening such routes would threaten their property values.22

Antidiscrimination Law This chapter has argued that when the law bars discriminatory restraints on alienation, entry, and advertising, communities whose residents prefer particular kinds of homogeneity may use strategies based either on exclusionary amenities or a lack of inclusionary amenities. This raises the question of what is worse: the medicine or the disease? If the exclusionary amenities strategy produces worse societal outcomes than overt discrimination, savvy policy makers could conceivably contemplate doing away with antidiscrimination laws altogether. Exclusionary amenities present a form of discrimination less “efficient” than overt discrimination. However, this inefficiency yields some social benefits, as well. The social costs are the deadweight losses that result from the expenditure of scarce societal resources on exclusionary amenities. For example, society may have built too many residential golf courses during the 1990s, resulting in wasteful land use policies and, in this instance, substantial environmental damage.23 Repealing antidiscrimination laws might well

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

eliminate this excess demand for golf course construction. However, much of this social waste is funded by those seeking residential homogeneity. To that extent, permitting exclusionary amenities as a lawful alternative to overt discrimination might function as an excise tax on residential homogeneity. It would be a wonderful coincidence if the costs imposed by this tax were equivalent to the social costs of the resulting residential homogeneity. In other words, what if the premium paid by in-groups for exclusivity precisely matches the harm to out-groups stemming from their exclusion? Sadly, the probability that harms and benefits match up so well is exceedingly low.24 Moreover, the fact that golf course designers or church builders are making profits by building exclusionary amenities hardly helps those who are targeted for exclusion. Little of the “exclusion tax” finds its way into the pockets of marginalized community members. That said, in a society that values residential heterogeneity as a general matter, “taxing” exclusion in this way may help ensure that people who choose to engage in this form of exclusion have rather strong preferences for doing so. Gary Becker, Richard Epstein, and others have argued that the market adequately punishes people who refuse to deal with African American customers by depriving them of a large number of potential customers.25 The presence of strong and broad consumer demand for segregated environments will, by the same token, reward developers who cater to that demand. In a world where large numbers of Caucasians are willing to pay a premium for neighborhoods that exhibit rather substantial racial homogeneity,26 the waste associated with the provision of exclusionary amenities may provide the only significant penalty suffered by an entrepreneur who satisfies these discriminatory preferences. Exclusionary amenities strategies necessarily create a second kind of inefficiency: they will be less precise than overt discrimination. Tiger Woods is not the only affluent African American golfer. Consequently, a homeowners association that tries to use golf as a proxy for race may not achieve complete racial homogeneity. That might be beneficial in several respects. First, exposure to some racial heterogeneity, albeit a limited amount, may result in preference changes within the majority population that would not occur in a world of complete homogeneity. Evidence shows that both Caucasians and African Americans possess fewer interracial prejudices and will be more willing to integrate following more sustained

121

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

122

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

interracial interactions.27 Second, exclusionary amenities strategies diminish the liberty of members of the excluded group less than overt discrimination does. An African American non-golfer can join a mandatory membership residential golf community—he will just have to pay a premium to do so. As a result, we might expect that he will resent the exclusionary device less than a flat prohibition on entry. Finally, preferences for the good in question may change over time. In recent years, African Americans have taken up golf in increasing numbers.28 If this trend continues, then golf courses will no longer function effectively as exclusionary amenities, and Caucasians interested in racial homogeneity will have to resort to other sorting devices. For all these reasons, then, the “inefficiencies” associated with exclusionary amenity strategies may enhance social welfare. It is probably preferable if communities are forced to substitute exclusionary amenities for overt exclusionary admission criteria or restraints on sales. That said, it is still worth considering whether society would be better off trying to restrict exclusionary amenities strategies or leaving them unregulated, as the law currently does.

Administrative Concerns Let us focus on the use of exclusionary amenities to achieve objectionable ends, such as the exclusion of African Americans from overwhelmingly Caucasian neighborhoods. Should the law proscribe the creation of club goods that deter African Americans from joining a particular community? Not necessarily. In a world where courts are prone to error, and evidence of discriminatory intent is difficult to gather, policing the provision of exclusionary amenities will often prove quite difficult. After all, there is substantial demand for residential golf courses, and a desire for racial homogeneity is not the only plausible explanation for a mandatory membership structure. Mandatory membership may be designed to combat free riding by those who benefit from a golf course’s views and open space but do not contribute to its upkeep. Moreover, mandatory membership might be designed as a device by which residents can precommit to high levels of social interactions among neighborhood residents. 29 Finally, there may be alternative reasons, quite apart from racial bias, to explain

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

why golfers want to live among fellow golfers. For example, doing so may reduce the search costs associated with obtaining useful golf tips or make it easier for golfers to assemble a foursome. As a result, it is appropriate to proscribe exclusionary amenity strategies only where the club good in question would not have been provided but for the desire to achieve a type of residential homogeneity that violates public policy interests. If a developer can show that consumer demand for an amenity is sufficient to explain its procurement, and that preferences for resident homogeneity do not drive that demand for the amenity, antidiscrimination law should not interfere with the developer’s choices about what amenities to offer. Where the law does attempt to defeat exclusionary amenities strategies, some governmental approaches will be superior to others. Given the risk of false positives, it seems wise to police the financing mechanisms for club goods before policing the actual provision of those club goods themselves. There is not necessarily anything objectionable about mandatory membership in golf communities that charge all residents for the positive externalities that the golf course confers on them. This structure requires charging golfers within a residential golf development for open space, views, and golf, and charging non-golfers for open space and views. In fact, many residential golf communities provide such two-tiered membership structures,30 and even those that do not are likely to implicitly charge non-golfers by capitalizing the extra value of a view into the original purchase price of a home. Accordingly, where strong evidence suggests that the provision of exclusionary amenities promotes residential segregation, the appropriate solution is not to ban residential golf communities. Rather, the remedy should be to invalidate mandatory membership schemes for golf playing, which is racially skewed, as opposed to enjoyment of golf course views, which is less likely to create polarized racial preferences. The approach discussed here is patterned on the unbundling strategy that is well integrated into antitrust law.31 Most famously, the courts contemplated requiring Microsoft to unbundle its web browser from its operating system in order to prevent the software giant from leveraging its dominant position in the latter market into a dominant position in the former.32 If consumers were required to select operating systems and web browsers separately or faced no price pressure to select an operating system plus a web browser rather than selecting each product individually, then concerns

123

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

124

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

about anticompetitive behavior would fade away. Translating this antitrust remedy into an antidiscrimination remedy is straightforward. Developers could still build homes next to golf courses, but they could not mandate that these homeowners purchase costly memberships to those courses or otherwise force purchasers to bear the capitalized costs of golf course land acquisition and upkeep. The unbundling strategy essentially eliminates the opportunity to use exclusionary amenities as a discriminatory tax that falls hardest on members of undesired groups.33 The prospect of inclusionary amenities brings to mind another remedial possibility. Rather than require unbundling, the law might mandate bundling of a different sort. Where there are substantial concerns about the use of exclusionary amenities to promote homogeneity, the law might demand the coupling of exclusionary club goods with inclusionary amenities. For example, if a developer wants to put a rock climbing wall in a new development, the law might also require him to build a basketball court next door. Such a coupling scheme might produce a world with too few residential golf courses, but it would also promote the construction of more basketball courts, which are probably undersupplied by the market for the reasons given earlier. Even if such a mandate results in a basketball court glut, the positive externalities associated with interracial relationships established on the court seem to make basketball an activity worth subsidizing, particularly in suburban residential communities. That said, unbundling is probably a more precise tool than this form of superbundling, in the sense that it would be difficult to calibrate the optimal level of extra bundling to offset the adverse consequences of an exclusionary amenities approach. Mandated bundling would, however, be preferable in instances where the relevant decision makers felt that there were too few social interactions in the community, and that, in the absence of collective amenities that prompted face-to-face interactions, relations among diverse members of a residential community would suffer too much. This chapter has argued that divergent preferences for amenities and activities among members of different racial groups are not innocuous, because those divergences create an opening for developers interested in promoting residential segregation. Perhaps the most promising strategy for combating the use of exclusionary amenities is to try to alter the preferences of the group that is targeted for exclusionary treatment. Tiger Woods’s

BUNDLED AMENITIES TO REDUCE DISCRIMINATION

success on the PGA Tour has coincided with a staggering increase in the percentage of African Americans who identify themselves as avid golf fans.34 Furthermore, the Tiger Woods Foundation has sought to provide golfing opportunities to minority and disadvantaged youth.35 Although some aspects of golf participation remain racially skewed,36 these demographic developments might render residential golf courses ineffective race-oriented exclusionary club goods at present and in the years ahead.37 Group disparities in preferences for club goods are socially constructed. As such, they may be amenable to concerted efforts by government or private groups to homogenize preferences as a means of thwarting insidious exclusionary amenities strategies.38 In this form of social engineering, the government would try to defeat racial residential segregation through its spending powers rather than through traditional antidiscrimination laws. Such expenditures would conceivably generate more bang for the government’s buck than sending the Justice Department after residential golf course developers. At first blush, it is hard to imagine the government trying to alter domestic preferences in this way, but it is hardly so different from efforts to encourage civic-minded behavior like water conservation or phoning the police at the sight of a suspicious package. As a general matter, we would ordinarily think that it is less intrusive for the state to try to convince people to take up a new hobby than it is for it to criminalize commercial speech.

Promoting Exclusionary Strategies Given society’s interest in promoting diversity among communities, as well as diversity within communities, there are arguably instances in which the law should promote the use of exclusionary amenities. Consider the efforts by deaf Americans to establish a community made up largely of sign language speakers in Laurent, South Dakota.39 There are strong welfarist arguments for such a residential arrangement, given the network effects and economies of scale associated with bringing speakers of this language together in one place. There are sensible political representation arguments as well, and Laurent organizers were particularly enticed by the prospect of electing representatives who will be forceful advocates for their interests.40 Deaf candidates for public office currently face long odds

125

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

126

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

of getting elected in communities where the deaf are a small minority. Although Canada and several European nations have elected deaf parliamentarians, there has evidently never been a deaf member of Congress. At present, few nondeaf people will want to live in a community where sign language is the lingua franca. Laurent’s developer declared bankruptcy in 2007, but if such a community were ever to become economically successful, one can imagine that those who are not fluent in sign language would move there in search of economic opportunities. To curtail such behavior, the community may find it worthwhile to invest in exclusionary amenities. Where a religious, linguistic, or other minority community genuinely requires some measure of critical mass to thrive, it may be appropriate for the state to subsidize the creation of exclusionary amenities or, failing that, at least to remain neutral. In such an instance, neutrality would mean permitting the enforcement of covenants and equitable servitudes designed to support the creation and maintenance of these kinds of club goods. The law is a sufficiently precise instrument to differentiate between relatively innocuous uses of exclusionary amenities by the deaf and more worrisome strategies designed to exclude marginalized racial minority groups from affluent neighborhoods.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

8 INFORMATION IS A VARIABLE, NOT A CONSTANT

My information-based theory of exclusion posits that those who control access to a resource will decide among exclusion strategies, and about whether to exclude or not, based in large part on how easily they can obtain information about the attributes of potential resource users. It is tempting to assume that the presence or absence of pertinent information is determined by the private sector, and that there is nothing the state can do to affect it. But that view is mistaken. In many contexts, the state does use policy levers to make what was previously private information public. In a few important contexts, the state can even transform public information into private information, although this is more difficult. Once we recognize the variable nature of private information, it becomes apparent that there is much the state can do to alter the ability of private actors to exclude. In light of recent technological advances, the state’s ability to offset information asymmetries may now be at its historic high-water mark.

The Reputation Revolution One of the most significant developments in the industrialized world during the Past decade has been the increased availability of information about individuals. Personal information that was once obscure can be revealed almost instantaneously via a Google search. The flea market transaction, in which a consumer had to hope that a vendor of goods was 127

128

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

trustworthy, has been largely displaced by the eBay auction where a prospective bidder can review information about hundreds of the seller’s prior transactions in an effort to ensure that the seller is trustworthy, prompt, courteous, and the like. Anxious parents thinking about purchasing a home in an unfamiliar neighborhood can acquire information about all the registered sex offenders living nearby, complete with comprehensive information about their crimes, via a few keystrokes. Sizing up a potential blind date to get a sense of his peer group and family background—a task that was once arduous and blatant—has now become easy and discreet thanks to Facebook and similar social networking sites. A car seller can comfortably show a consumer a new car and let her drive it home, having paid for the vehicle with little or no money down, thanks to the comprehensive and nearly instantaneous credit checks that now take place on the dealer’s lot. And we can even assess the credibility of people we have no expectation of ever meeting—the amateur movie critic on Netflix, book reviewer on Amazon, or commentator on Slashdot. I can look up their respective histories of movie, book, and article reviews to see how seriously I should take their recent reviews of Inglourious Basterds, A Thousand Splendid Suns, or the latest iPod. A 2010 research paper by Zhongmin Wang found that Yelp—which requires reviewers to select pseudonyms, flags the most prolific reviewers as “elite” members, permits users to rate each others’ reviews, and encourages social interactions among reviewers—obtains higher-quality and less extreme reviews from more prolific reviewers than competing review sites like Citysearch and Yahoo Local, which do not engage in these practices and rely instead on purely anonymous reviewing.1 Reputation tracking technologies log customers’ preferences and quirks too. For example, Open Table, the popular online reservations system for restaurants, tracks dining patrons’ tendencies: if I repeatedly show up late for reservations, Open Table will alert restaurateurs so that they know to expect me fifteen minutes after my stated reservation time.2 In short, the anonymity and pseudonymity that once characterized our interactions with strangers is fading. Let us refer to this change as the “reputation revolution.” In the years ahead, it seems likely that existing imperfections in the reputation market will dissipate.3 For example, it is presently difficult for

INFORMATION IS A VARIABLE, NOT A CONSTANT

an individual to translate a strong offline reputation into a strong online reputation. eBay does not let users take their existing brick-and-mortar reputations into the online auction world. Thus, a well-established merchant in a small town with lots of satisfied customers providing her repeat business gets zero feedback points upon opening a new eBay account, just like everyone else. Because a strong feedback score permits a seller to obtain higher prices for her goods,4 the new entrant on eBay thus may have to sell goods with a discounted reserve price in order to build up the positive reputation that will allow her to compete effectively with established online sellers. This is a significant market inefficiency, and firms like Spokeo and Pipl are making progress toward solving the “problem” by linking information about individuals from varied disparate databases. Next-generation meta-ranking sites have the capacity to make reputation more readily transferable from one online forum to another and between online and offline marketplaces. Just as each American has a credit score that banks, lenders, landlords, and others can access before determining whether and on what terms to do business with us, we should anticipate the emergence of “social credit scores,” where individuals’ personality traits are assessed and translated into a profile that tries to predict an individual’s suitability for various social and economic opportunities. Cruder versions of such profiling already exist in the online advertising realm.5 As technologies improve, the reputation revolution has the potential to alter more than just considered judgments with obvious economic and social consequences, such as the decision to purchase a car, buy a home, go on a date, or acquire a pair of Nikes. In these settings we would hope that the decision maker will conduct some research before electing a course of action. In the near future, it is plausible that information about individuals will seep into interactions where it is presently unavailable. Such interactions require split-second decisions, but technologies being developed for the mass market will nevertheless enable us to take advantage of the reputation revolution in making those decisions. A concrete example is useful here. It is late at night, and an unaccompanied adult is walking home. There are no other pedestrians or moving vehicles immediately visible, but a twenty-four-hour pharmacy stands across the street. The adult suddenly sees a group of five male teenagers turn the

129

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

130

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

corner. The teens are now walking directly toward the adult. Will the pedestrian cross the street, perhaps entering the pharmacy for a moment? The answer probably will depend on a series of proxies: What race are the teenagers? How are they dressed? How are they interacting with one another? In a split second, and with potentially high stakes, the first question may prove decisive in guiding the adult’s response to this situation. The adult may well cross the street out of caution and then feel ashamed after receiving quizzical or bemused looks from the probably harmless teenagers. Given presently available technology, we have a hard time understanding the application of a strategy to give the pedestrian more relevant information so he can rely less on race or wardrobe information. But wearable computers that are already being tested and rolled out permit precisely those types of calculations. In the past few years, our old-school cell phones have seen the addition of camera, text messaging, email, Global Positioning System, video chatting, gaming, and mp3 player functionalities. They have begun to integrate impressive social networking and data storage capabilities while becoming platforms for tens of thousands of applications.6 Computer scientists have now spent a decade researching, implementing, and writing about “wearable communities,” which employ ubiquitous computing resources to help provide individuals with information about the people and products around them.7 These technologies have numerous functionalities, but among the most promising is an application of social networking sites into real space. By 2011, a number of GPS-enabled social networking applications were already allowing smartphone users to broadcast their physical location to friends and see where their friends were in real time, including Google’s Latitude, Loopt, and Pelago.8 If everyone is carrying around a wearable computer that can talk to everyone else’s wearable computer, then at a moment’s notice I might be able to discern whether anyone else in a café has seen the movie I just saw.9 While waiting in line at the grocery store or box office, I might be alerted to the presence of a friend-of-a-friend just behind me. Given existing technologies like Latitude, it is a short leap to a world in which having told my wearable computer that I am looking for an electrician, I might learn that there is an available electrician in aisle six at Home Depot,

INFORMATION IS A VARIABLE, NOT A CONSTANT

that a trusted colleague employed the electrician recently and rated his performance nine out of ten, and that the electrician has experience doing the sorts of lighting rewiring jobs that I need done. These sorts of technologies can help us navigate the everyday snap judgments about how to interact with strangers on a mostly deserted block. My computer might tell me that the five youths approaching late at night are honors students or dropouts with long rap sheets.10 And if my computer can tell me that, a police officer’s wearable computer can tell him just as much or more, diminishing the likelihood of racial profiling and tragic mistakes.11 In the years ahead, my computer could tell me whether the individual who just cut me off on the freeway has a history of violence and vehicular collisions, or is a surgeon likely racing to the hospital. Summary statistics, much like eBay’s reputation scores or Slashdot’s karma points, could flash on my screen to tell me that the teens are trusted by people I trust but the driver is someone with a violent temper who is best avoided. Of course, if it is late at night and I am being approached by teenagers with suspicious intentions, do I really want to be pulling out a portable computer and scanning an eBay profile? If a taxi driver has to decide whether to pick up a pedestrian late at night, won’t the driver always prefer statistical discrimination to spending ten seconds evaluating the pedestrian’s reputation score?12 These are fair questions, but they are ones with good answers—the problems raised by these examples are being addressed in e-commerce. Automated negotiation protocols allow users who own devices that can communicate with other users’ devices to set the parameters for such communication. For example, these protocols have been implemented in digital rights management, p3p (platform for privacy preferences) systems, and wearable communities. In such environments, automated negotiation lets users specify what levels of trust are necessary before their digital media device will download content from another device or website. They permit users to identify in advance what privacy-protection characteristics a website offers and only show users items for sale from websites that meet the specified criteria. Or they allow someone participating in a wearable community to request notification whenever someone else approaches whose wearable computer indicates specified attributes (unmarried, fellow Hoosier, Vampire Weekend fan, and so on). Our hypothetical taxi driver might program his

131

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

132

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

computer to notify him whenever a tourist, university student, nurse, or airline pilot seeking a cab is in the vicinity, but he might also assign lower priority to taxi requests from those who are currently unemployed and have the computer provide an audible warning beep whenever a proximate convicted felon tries to hail a cab. Alternatively, the taxi driver might identify fifty cab drivers whom he trusts and pick up any passenger who had received positive feedback from any of those cab drivers.13 Under the law, of course, a taxi company is a common carrier, with an obligation to provide service to everyone.14 But we should not kid ourselves. Cab drivers routinely avoid picking up African Americans, often avoid serving African American neighborhoods entirely, and resist efforts to enforce antidiscrimination laws.15 With infinite law enforcement resources, we could stop cab drivers from discriminating. But in our highly imperfect world, we might pursue “lesser of two evils” strategies. As we will see shortly, information asymmetries currently cause taxi drivers to sort on the basis of race and gender when picking up rides, penalizing both the African American male gang member and the African American male clergyman. A technology that allows sorting between gang members and clergymen, from this perspective, makes the world a better place than the one in which we live, where the only decision-making factors available to the cab driver are the information revealed by quick visual inspection (race, gender, dress, and so forth). To be sure, there are technological and consumer preference impediments to the implementation of these technologies. Networking capabilities will have to improve somewhat, kinks will have to be worked out of the existing wearable computing software, and consumers will have to decide that the benefits of these technologies warrant the associated loss of privacy. In some cases, technology firms may design around technological bottlenecks or legal impediments. For example, while those with criminal records might elect not to carry around smartphones so as to avoid broadcasting their status, automated photograph recognition software could allow those whose mug shots appear in public databases or even those whose images have been tagged by others on Facebook to be identified instantly.16 No significant hypothesis in this book depends on the successful implementation of these technologies. But the preceding discussion is designed to open the reader’s eyes to the dramatic nature of the reputation revolution

INFORMATION IS A VARIABLE, NOT A CONSTANT

that has already occurred in the past decade and the not fanciful possibility that equally dramatic changes lie ahead in the next ten years.

The Conventional Wisdom About Consumer Information and Discrimination The rise of social networking websites, consumer information databases, blogs, and online feedback systems have not made scholars who write about privacy law happy. In a host of books and articles, these scholars have bemoaned the disappearance of anonymity, the loss of autonomy, and the dangers of shame sanctions that have accompanied these trends.17 One influential and important part of that critique began in 1993 with Oscar Gandy’s book The Panoptic Sort,18 and the issue remains vibrant and visible today thanks to work by David Lyon and other surveillance studies scholars.19 According to Gandy and Lyon, the growth of technology-aided surveillance, consumer information databases, and other mechanisms for accessing increased information about individuals facilitates pernicious forms of discrimination—what Gandy calls the “panoptic sort” and Lyon calls “social sorting.” It prevents governments from treating citizens alike and similarly prevents firms from treating consumers as an undifferentiated mass. As a result, people are categorized, grouped, divided, and treated in disturbingly differential ways.20 I would submit that this research by Gandy and Lyon oversimplifies the relationship between discrimination and the availability of personal information.21 The increased availability of information about individuals will prompt some decision makers to shift from not sorting to sorting, as Gandy and Lyon argue. But it will prompt other decision makers to shift from sorting via problematic group-based stereotypes to less problematic, more individuated judgments. The key questions, as I explain below, are the magnitudes and consequences of the two relevant changes. Before the developed world made heavy investments in its “reputational infrastructure,”22 citizens and consumers were not treated equally. Then, as now, governments sorted citizens and firms sorted consumers. The difference between our present age and prior epochs was not the temptation to sort, but the basis for sorting. Lacking comprehensive consumer information databases, criminal history databases, Google searches, Facebook

133

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

134

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

profiles, family background, and the like, institutions interested in sorting used easily available criteria like race, gender, and age as their criteria. When Gandy, Lyon, and other scholars writing in the surveillance-studies tradition advocate greater privacy protections as a mechanism for decreasing discrimination, they (unwittingly) propose policies that will shift sorting techniques away from less problematic criteria, like purchasing patterns, social affiliations, criminal histories, insolvency records, and Internet browsing behavior, back toward the older and more readily observable sorting standbys—race, gender, and age. Often, the choice isn’t between sorting and not sorting; the economic and social gains from sorting are simply too great, while banning sorting in many contexts will be simultaneously costly and not terribly effective. The real choice is between sorting on the basis of troubling criteria and sorting on the basis of positively obnoxious criteria. The remainder of this chapter takes stock of the enormous changes that industrialized democracies have seen in the past couple of decades and asks, in a systematic and sustained way, what the law should do to respond to those changes. It explores how the widespread availability of information about individuals has already transformed the market for rental apartments and then analyzes the law’s initial responses to this revolution in the context of employment discrimination, jury selection, medical treatment decisions, immigration law, consumer protection law, and the law of defamation. In all these settings, courts and policy makers are beginning to confront the new informational environment, but they have not fully thought through its implications. The discussion illustrates that context matters, such that the law’s optimal response in one racial discrimination context—hiring—should differ from its response in another racial discrimination context—jury selection. More information is not always better. Nor is it always worse. The discussion that follows begins to develop principles that should help us determine when the state should facilitate the proliferation of personal information.

Landlord-Tenant Law Recall the discussion of nineteenth-century English workhouses in chapter five. Urbanization delivered shock therapy to the welfare delivery

INFORMATION IS A VARIABLE, NOT A CONSTANT

system. Dispensers of charity no longer had good information about those seeking help. As a result, charities moved away from grants to individual benefit seekers, which required the charities to have a lot of information about individuals, toward a bundled system in which charity was combined with a living arrangement that the undeserving poor and nonpoor would find unpalatable. A sudden decline in pertinent information altered the appropriate mechanisms for delivering charity. Approximately a century and a half later, a similar dynamic would play out in housing, but in the opposite direction. This time a sudden increase in pertinent information about individuals would prompt a shift away from proxy-based discrimination. During the 1970s and 1980s, it was not unusual for landlords in New York City to refuse to rent apartments to lawyers.23 At first blush, this seems like an odd trend. Lawyers may be loathed by the public, but they typically bring home a nice paycheck. A New York City landlord who refused to rent to lawyers would be depriving himself of many well-heeled prospective tenants. The explanation for this seemingly irrational landlord conduct was a story about the use of proxies. As one landlord explained to a New York court, his refusal to rent an apartment to a qualified attorney applicant was based on his preference for “a person who was likely to be less informed and more passive” rather than someone “attuned to her legal rights.”24 That court noted that lawyers were not a protected class under fair housing laws and thus ruled in the landlord’s favor.25 It took nine years, but New York City eventually prohibited discrimination in the housing market on the basis of profession, at the urging of lawyers who had similar troubles finding rental units in the city.26 We shall overcome someday. The 1986 enactment of New York City’s profession-based fair housing protections did not prompt the city’s landlords to wave the white flag of surrender with respect to tenant screening. Landlords still wanted to screen out those tenants who seemed likely to invoke their rights under New York’s landlord-tenant laws. Some landlords responded by continuing to avoid renting to lawyers but offering instead some pretext—new or old.27 But other landlords stopped relying on profession as a proxy and started relying on involvement in prior litigation as a proxy for litigiousness. Information brokers began data-mining state and municipal court records, hoping to identify tenants who had been involved in landlord-tenant litigation of

135

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

136

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

any sort.28 Those landlords essentially blacklisted tenants who had been involved in such litigation.29 In such a world, even tenants who had won suits against their landlords faced a difficult time obtaining housing. As the founder of a tenant screening company told the New York Times, “It is the policy of 99 percent of our customers in New York to flat out reject anybody with a landlord-tenant record, no matter what the reason is and no matter what the outcome is, because if their dispute has escalated to going to court, an owner will view them as a pain.”30 In New York, at least, even sex offenders seem to have an easier time finding a place to live than tenants who had previously litigated against their landlords.31 To be sure, market conditions always affect the balance of power between landlords and tenants. There is a lot of demand for rental housing in New York and a finite supply. Were the situation reversed, the pressure to exclude potentially litigious tenants would decline. Yet that familiar dynamic represents only part of the story. Legal environments matter as well. On this account, tenant screening services are particularly popular in New York in part because landlord-tenant regulations in that city slant so heavily in the direction of tenants.32 It is hard to get rid of deadbeat tenants. Faced with high eviction and litigation costs, landlords devote more resources to trying to screen out prospective tenants who pose heightened risks of future legal entanglements. This screening presents a social problem because society has an interest in ensuring that the landlord-tenant laws are enforced, and the common law system requires the help of plaintiffs who generate appellate cases, the resolution of which will clarify the law.33 Landlord-tenant law, as it exists in ordinances and case law, bears little resemblance to the set of rules that govern everyday relations between landlords and tenants. The reputational consequences of involvement in litigation are so severe that a rational tenant should often elect not to seek enforcement of the substantive entitlements provided by formal law. In this environment, it is fair to wonder whether tenant’s rights clinics, a staple of legal education at many schools, should even exist in their current form. A legal aid attorney ought to advise an indigent client that involvement in landlord-tenant litigation is likely to be counterproductive in the long run. Pro-tenant regulations, in short, might not make tenants better off, though they do seem to improve the lot of tenant screening firms.

INFORMATION IS A VARIABLE, NOT A CONSTANT

This dynamic explains why some of the landlord-tenant reforms of the 1960s and 1970s, which were supported by well-meaning tenants’ rights advocates, may have ultimately backfired. One such reform is the prohibition on self-help evictions by landlords and tenants. At common law, landlords were able to evict tenants who had violated the terms of their lease by locking them out, provided that the landlords did not use excessive or unreasonable force.34 Beginning in the 1960s, tenants began legal reform efforts, arguing that legal process should be the exclusive means of ousting a tenant in possession.35 This reform movement was premised on the view that self-help evictions tend to spark violence between the landlord and tenant, and that unless checked by the courts, some landlords would evict tenants who had a legal right to remain on the premises.36 Although some states still permit self-help by landlords, their ranks have been shrinking, and even the jurisdictions that permit landlords to use reasonable force in self-help evictions have defined “reasonable force” so narrowly that self-help has become a “theoretical but not a practical alternative.”37 The movement to prohibit self-help evictions by landlords has long had a few critics, with the most prominent ones suggesting that landlords would pass the high costs of judicially evicting deadbeat tenants onto the tenants who paid their bills on time.38 Some passing on of these costs undoubtedly occurs, but the reputation revolution suggests a deeper criticism of the prohibitions on landlord self-help. Eviction via self-help typically creates no public records. Courts are not involved in a self-help eviction, and a landlord has little economic incentive to report such a dispossession to a credit bureau or any other information broker.39 Evictions via summary proceedings, on the other hand, necessarily generate public records, and it is those public records that will prove so damaging to tenants the next time they try to rent an apartment. From the perspective of facilitating tenant rehabilitation and second chances, a law prohibiting self-help by landlords will prove counterproductive. Many tenants who have trouble making rent payments will fail to appreciate the reputational repercussions of involvement in summary proceedings. For these tenants, the law’s prohibition on self-help can be a particularly raw deal. This is a point overlooked by defenders of the prohibition on self-help.40 Tenants’ rights advocates who appreciate the ways in which earlier reforms have produced unintended consequences are not powerless to

137

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

138

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

address this situation. If society believes that second chances are important in the landlord-tenant context, then it might require that information about involvement in landlord-tenant litigation be purged from consumers’ credit reports after a relatively brief period of time. Currently, the Fair Credit Reporting Act prohibits credit reporting agencies from disclosing information about someone’s involvement in landlord-tenant litigation more than seven years after its conclusion.41 There is an exception, however, that permits the disclosure of such information if it concerns a credit transaction of more than $150,000, so even very old litigation might adversely affect someone’s ability to use mortgage financing to purchase a home.42 In any event, a seven-year cloud on one’s suitability as a tenant will still impose substantial harms on tenants who become involved in litigation. During the 1990s, California tried to address this broader concern legislatively, prohibiting credit reporting agencies from including information about a tenant’s involvement in landlord-tenant litigation where the tenant was the prevailing party. Alas, this legislation was invalidated on First Amendment grounds.43 In short, legislators cannot easily prevent landlords from receiving information about tenants’ prior involvement in litigation. They can try to ban landlords from acting on that information, but the enforcement of such prohibitions, like the enforcement of other antidiscrimination provisions that regulate decision makers’ behavior, will be spotty, expensive, and prone to false positives. Let us survey landlord-tenant law in the twenty-first century. Tenant background checks have gotten so cheap that for many tenants, involvement in litigation of any sort will place meaningful constraints on their future ability to obtain rental housing. These tenants’ apartment search costs will be very high because landlords do not advertise their unwillingness to rent to people who have previously been involved in litigation. There are, in other words, no exclusionary vibes to which blacklisted tenants might respond. Assuming nontrivial vacancy levels, some landlords will be willing to rent to those who have been evicted in summary proceedings or sought to vindicate their legal rights, but only after extracting higher rents and security deposits. In such a world, the only tenants who should be willing to defend their rights in court are those who have previously litigated and those with the ability to purchase housing. It seems like a hopeless world for a tenants’ rights advocate. But such an advocate should not lose all hope.

INFORMATION IS A VARIABLE, NOT A CONSTANT

There are two sides to every reputational coin, and that is equally true of this particular doubloon. Just as a substantial market has developed for tenant screening services, the market has also produced clearinghouses for information about landlords’ reputations. In a prior era, these were confined mainly to university housing offices, where students could swap stories about the good, bad, and ugly landlords.44 Today, in markets where prospective tenants sometimes hire real estate agents to assist in their searches, agents who depend on repeat business and positive word of mouth will have an incentive to learn which landlords behave inappropriately. Not surprisingly, the Internet has given rise to far more sophisticated resources for tenants. The best developed among them is ApartmentRatings.com, a website that contains tens of thousands of landlord ratings written by current and former tenants. This and some similar websites provide a wealth of new information, and their existence gives tenants some recourse in dealing with recalcitrant or bullying landlords.45 To the extent that websites like these are used by prospective tenants, landlords should fear developing a reputation for unfair or overly aggressive behavior. The best check on landlord misbehavior is probably not the threat of a lawsuit by the tenant, but the threat of a series of complaints by aggrieved and eloquent tenants. At present, an information asymmetry exists. ApartmentRatings.com is not as extensive or reliable as websites that monitor hotels and restaurants, like TripAdvisor.com. Impressionistic evidence suggests that it is underutilized and that the website devotes less attention to eliminating bogus or suspect ratings than TripAdvisor.com does. But as landlord-tenant law recedes into the background, we can expect that the demand for services like ApartmentRatings.com will expand, and this added demand will induce the site to provide a better service or be driven out by competitors offering a superior product. Summarizing the landlord-tenant market, then, one can see themes playing out that will become familiar refrains as we proceed through the remainder of this book. Where information costs are high, landlords use rough proxies, like occupation, to sort out litigious or otherwise undesirable prospective tenants. As reputational information becomes far more widely available through commercial data brokers, landlords reduce their reliance on these proxies and begin blacklisting those prospective tenants with previous involvement in litigation. As a result of this transformation,

139

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

140

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

landlord-tenant reforms designed to help marginal tenants and constrain landlords may have had the opposite effect, creating far more extensive public records of litigation that have tarred some tenants with undesirable status. In the short term, the result is tenant litigants being caught unaware by the reputational implications of trying to vindicate their rights through legal process. In the long run, the effect will be a substantial deterrent to litigation in the landlord-tenant arena. This de-emphasis on litigation may be tolerable, however, if existing resources designed to track and monitor landlords’ reputations become more reliable and widely used.

Antidiscrimination Law There are two basic forms of discrimination: animus-based discrimination and statistical discrimination. Animus-based discrimination occurs when an individual treats members of a group differently because of (conscious or unconscious) antipathy toward that group. Statistical discrimination occurs when an individual treats members of a group differently because he believes that group membership correlates with some attribute that is both relevant and more difficult to observe than group membership.46 One engaged in statistical discrimination would not harbor any ill will toward members of the group against which he is discriminating beyond the belief that membership in that group correlates with some undesirable characteristic. To continue with our first example, a landlord who refuses to rent to lawyers because he fears litigious tenants and thinks lawyers are more likely to be litigious is engaging in statistical discrimination. A landlord who will not rent to lawyers because he hates lawyers is an animus-based discriminator. Statistical discrimination is often a likely explanation for instances in which decision makers who themselves belong to a minority group nevertheless make decisions that favor majority group members over minority group members. For example, newspaper accounts suggest that nannies generally prefer to work for Caucasian clients over African American clients. African American clients were widely perceived as being too demanding, too likely to live in unsafe neighborhoods, or unable to pay as much as white couples.47 These stereotypes were shared by nannies of all races, including African Americans and Caribbeans.48 Similar phenomena have

INFORMATION IS A VARIABLE, NOT A CONSTANT

been observed in explaining the difficulties that African American professionals have hailing cabs successfully—African American riders are viewed as less safe passengers and poorer tippers.49 Waiters similarly perceive African Americans as poor tippers, an expectation that becomes a self-fulfilling prophecy if African Americans receive less attentive service as a result.50 Even doctors seem to rely on race-based statistical discrimination in diagnosing various ailments.51 There will be some contexts in which animus-based discrimination predominates and others in which statistical discrimination predominates. In 2011, overt racial animus persists but is probably waning as a result of generational replacement. A younger generation of Americans has embraced Martin Luther King as an icon, elected Barack Obama president, and feels embarrassed by slavery, Jim Crow, and massive resistance to integration. To them, racial animus seems distasteful and passé. That said, implicit bias, an unconscious form of animus-based discrimination, is alive and well.52 Painting with broad brushstrokes, it appears that while the effects of animus-based discrimination are still widely felt, implicit bias and statistical discrimination are more prevalent today than they were in prior eras and overt animus-based discrimination is less prevalent.53 To illustrate how statistical discrimination plays out in contemporary society, suppose a person charged with hiring a sales clerk wants to avoid employing someone with a criminal background because she feels that prior criminal convictions are a good proxy for future criminality. Assuming the decision maker lacks reliable access to information about applicants’ criminal records, she might choose to hire a Caucasian female over an equally qualified African American male based on the relatively high percentage of African American males and the relatively low percentage of Caucasian females who are involved in the criminal justice system.54 This decisionmaking process will impose a distasteful form of collective punishment on African American males who have had no run-ins with the law, penalizing them for crimes that others have committed. Because many decision makers may use the same decision-making criteria, a law-abiding African American male may face repeated rejection and economic marginalization.55 For these reasons, antidiscrimination law prohibits the use of these race or gender proxies even where race or gender might correlate with some relevant qualification.56

141

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

142

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

Policing statistical discrimination through traditional antidiscrimination measures has proven difficult: many victims of statistical discrimination never bring suit, many nonvictims bring unmeritorious suits that prompt defendants to settle so as to avoid the costs of litigation, and enforcement of the laws by the Justice Department and state attorneys general has been sporadic.57 Concerned about the courts being flooded with frivolous claims, judges have imposed substantial burdens on plaintiffs seeking to enforce antidiscrimination laws, often hamstringing pretrial discovery, interpreting statutes of limitations aggressively, or hastening resolution of claims on summary judgment.58 A fascinating research paper published in the Journal of Law and Economics by Harry Holzer, Stephen Raphael, and Michael Stoll illustrates the prevalence of statistical discrimination and the failure of antidiscrimination laws to curtail it.59 The paper began by noting that 28 percent of African American males, 16 percent of Hispanic males, and 4 percent of white males would be incarcerated at some point in their lives and that the median prison sentence was less than two years.60 As a result, Holzer and his coauthors observed that a sizable minority of the male labor pool in the United States consists of people with criminal records.61 The study then surveyed employers about their most recent hire for a position that did not require a college degree. The authors collected demographic information about each firm’s most recent hire and information about the firm’s willingness to hire employees with criminal records generally.62 The findings of the study suggested that statistical discrimination against African American males was widespread and that employers were using race as a proxy for involvement in the criminal justice system. Employers who conducted criminal background checks on applicants were 8.4 percent more likely to hire African Americans than employers who did not. Consistent with the statistical discrimination hypothesis, the effect was highly significant for employers who expressed an unwillingness to hire ex-offenders (10.7 percent greater likelihood) and only marginally significant for employers who stated their willingness to hire ex-offenders (4.8 percent greater likelihood). The effects for African American males were far greater than the effects for African American females, which is consistent with the statistical discrimination hypothesis and harder to square with a racial animus hypothesis. Further, the study found the same

INFORMATION IS A VARIABLE, NOT A CONSTANT

effects even after controlling for differences in the racial composition of the applicant pool. The study also found evidence that surveyed employers who did not conduct criminal background checks used other proxies for criminal convictions as well, including spotty work histories and periods of unemployment lasting over a year, and that those employers who expressed an unwillingness to hire ex-offenders were significantly less likely to hire members of these stigmatized groups. Surveying their results, the study authors reached the following conclusion about the effects of statistical discrimination on African American job applicants: The empirical estimates indicate that employers who perform criminal background checks are more likely to hire black applicants than employers that do not. . . . This positive net effect indicates that the adverse consequence of employer-initiated background checks on the likelihood of hiring African Americans is more than offset by the positive effect of eliminating statistical discrimination. . . . In addition, we find that the positive effect of criminal background checks on the likelihood that an employer hires a black applicant is larger among firms that are unwilling to hire ex-offenders. This pattern is consistent with the proposition that employers with a particularly strong aversion to ex-offenders may be more likely to overestimate the relationship between criminality and race and hence hire too few African Americans as a result. . . . The results of this study suggest that curtailing access to criminal history records may actually harm more people than it helps and aggravate racial differences in labor market outcomes.63 Subsequent research by Keith Finlay reached broadly similar conclusions on the basis of a different methodology.64 Finlay looked at variation among the states in the accessibility of criminal history information. He found that in states with highly accessible criminal history databases, ex-offenders faced worse employment prospects but members of high-offending-rate groups faced marginally better employment prospects. The implications of these studies and of similar studies on statistical discrimination in the employment market65 are chilling, but they should

143

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

144

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

not be surprising. Employers in a variety of sectors admit their heavy reliance on proxies, especially in sorting between “good” and “bad” African American workers.66 Many employers wish to avoid hiring ex-offenders because they consider them untrustworthy, because they are worried about vicarious liability if the ex-offenders commit crimes or torts on the job, or for other reasons.67 Employers who expend resources on criminal background checks will be able to sort effectively among those African Americans who have had run-ins with law enforcement and those who have not, but other employers will rely on race as a proxy for criminality, imposing a disturbing sanction on law-abiding African American males.68 Given the deleterious consequences of this predictable behavior, it is worth examining the possible avenues, other than ex post litigation, for the state to prevent statistical discrimination. One way to protect African Americans and other disadvantaged groups would be to make them appear indistinguishable from whites. Indeed, some efforts to reform antidiscrimination law have suggested that statistical discrimination can be mitigated if the relevant decision makers are deprived of information about a candidate’s race, religion, or gender.69 With less information, decision makers presumably will focus more on the black and white of a job applicant’s resume and less on the black or white of the applicant’s skin. The drawbacks to this strategy are obvious. For example, to deprive employers of information about the race of a job applicant, the law would have to prevent them from meeting face-to-face. But such restrictions would deprive both the employer and the applicant of very valuable information that has nothing to do with race. In an era of ubiquitous personal information, we should consider approaching the statistical discrimination problem from the opposite direction: using the government to help provide decision makers with more complete information about each applicant. Thus, readily discernible facts like race or gender will not be overemphasized, and more relevant, less observable facts, like an applicant’s past job performance or social capital, will loom larger. For instance, government subsidies could promote the creation of employment evaluation clearinghouses where individuals’ employment reviews from all prior jobs would be aggregated in a single source that would be accessible to human resources personnel. The government might further improve the quality of the data in these clearinghouses by providing tax subsidies to encourage the collection

INFORMATION IS A VARIABLE, NOT A CONSTANT

of “360-degree” feedback within firms and promote the availability of such feedback between them. Such feedback aggregates an employee’s performance evaluations from subordinates and peers, as well as supervisors. Because it draws on information from many people and because the ranks of junior employees tend to be more diverse than upper management, reliance on 360-degree feedback is likely to decrease the weight associated with any particular evaluation and minimize the likelihood that race or gender dynamics will taint the accuracy of the employee evaluations.70 When firms lack reliable information about the performance of a job applicant at a previous employer, they are more likely to rely on old boys’ networks, their impressions of applicants’ interviewing skills, and other highly imperfect mechanisms. To be sure, when firms know that their own employment evaluations will be available to their competitors, they will have an incentive to encourage their employees to be less than fully candid in their evaluations. The strategy of making such evaluations more widely available will therefore be most beneficial when the evaluations come from large, bureaucratized entities, where there are too many employees for senior management to easily determine which employees should be promoted or transferred to other divisions. Pushing this sort of strategy further, the state could publish information about all individuals’ involvement (or lack thereof) in the criminal justice or bankruptcy systems; it might publish military records that document individuals’ performance and conduct while in the service; or it might verify and vouch for applicants’ educational credentials. All of these steps would help shift an employer’s gaze away from race, gender, and age. On this theory, a major factor driving unlawful discrimination on the basis of race, ethnic status, gender, or religion is a lack of verifiable information about the individual seeking a job, home, or service. By making the publication of criminal histories tortious71 or making it more costly for the media to obtain aggregated criminal history information that is already in the government’s hands,72 information privacy protections become the enemy of antidiscrimination law. The tradeoff makes privacy law and institutional arrangements that obscure information about individuals’ reputations far more problematic than courts and theorists presently suppose. From the perspective of reducing statistical discrimination on the basis of race, the problem with Megan’s Law registries is not that they reveal

145

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

146

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

too much, but that they reveal too little. Such online resources should be expanded to include all criminal convictions of adults, as Colorado, Florida, and a handful of other states have done. Indeed, much of the information in the federal government’s National Crime Information Center’s database could be brought online and made available to the general public free of charge.

The Jury Pool The same sorts of dynamics that confront decision makers in the hiring, leasing, and sales contexts play out within the criminal justice system as well. That said, the jury selection process shows the dangers of extrapolating arguments that arise in the context of necessities—like employment or shelter—to contexts where individual resentment of excessive data collection and dissemination may discourage socially productive behavior. The relatively few high-stakes legal disputes that go to trial are often won or lost during the voir dire process, when attorneys seek to seat the jurors deemed most sympathetic to the sorts of arguments they will make and strike those deemed least sympathetic. Here, as in other contexts, the relevant decision makers appear to rely heavily on characteristics that they can discern at a relatively low cost—race, gender, age, and national origin.73 They also rely on additional information that the jurors themselves provide, but there are significant problems with taking that information at face value. First, prospective jurors often knowingly provide inaccurate information.74 Second, jurors may be poor at self-assessing. For example, they may overestimate their ability to be impartial in light of relevant life experiences. Third, there are lots of questions that attorneys would love to know the answers to but that they do not dare ask jurors. In some cases, lines of juror questioning are placed off limits by the law, as is usually the case with prospective jurors’ voting records, medical conditions, or reading habits.75 In other instances, attorneys fear antagonizing prospective jurors by asking them questions deemed overly intrusive. Attorneys have imperfect information about prospective jurors, and what attorneys do not know about jurors may prove decisive in deliberations. Yet attorney folklore about the correlation of jurors’ observable characteristics with their decision-making tendencies is often unreliable

INFORMATION IS A VARIABLE, NOT A CONSTANT

and unscientific.76 In recent years, a large industry of jury consultants has sprung up, hoping to make jury selection less of an art and more of a science.77 Jury consultants use information from jury questionnaires and voir dire to predict prospective jurors’ behavior. Thus, jury consultants may try to read body language, analyze handwriting, suggest revealing but not intrusive voir dire questions, or provide lawyers with information about how members of various demographic groups behave as jurors.78 To a jury consultant, the observable and trivial detail is often a proxy for the unobservable and critical characteristic. A large part of what modern jury consultants now provide to their clients is even more fine-grained. Jury consultants increasingly run background checks on the various prospective jurors in the pool, pulling credit reports, employing search engines, looking for rap sheets, and examining property tax records.79 In high-stakes cases, jury consultants work with private investigators who photograph prospective jurors’ homes and vehicles, searching for any pertinent information like a political yard sign or a religious bumper sticker.80 Litigants who cannot afford a jury consultant may turn to juryselection software like JuryQuest, which identifies jurors who should be stricken on the basis of demographics and a large database of prior juror survey responses and verdicts.81 What’s more, jury trial surveillance is beginning to operate in both directions. Empanelled jurors are Googling the attorneys who are making arguments before them with increasing regularity, and trial lawyers are trying to make their web profiles as appealing as possible to these curious jurors.82 With jurors using social networking tools to learn about criminal defendants, lawyers, and their fellow jurors, search engines to conduct their own research about the allegations in a case, and Twitter to disclose information about deliberations, trial court judges are having to rule on novel motions asserting that these actions merit mistrials or dismissals.83 In what may be the beginning of an emerging trend, courts are examining the backgrounds of jurors on their own initiative rather than relying on the parties to do the digging. Interest in this practice is particularly strong in Illinois, where the high-profile corruption conviction of former Governor George Ryan was nearly reversed on appeal thanks to the dismissal of two jurors eight days into the deliberations.84 The jurors were dismissed after Chicago Tribune reporters discovered that they had each concealed arrest

147

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

148

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

records during the jury selection process. To help prevent a repetition of these problems, the chief judge of the Northern District of Illinois has instructed court personnel to conduct criminal background checks on all prospective jurors in high-profile cases.85 If the Illinois experience proves successful, then one can imagine its duplication and expansion in other jurisdictions. What should we make of the federal district court’s reforms? Although criminal and other background checks are increasingly relied upon by litigators in high-stakes trials, not all parties employ them, and the George Ryan trial suggests that prosecutors sometimes fail to do their homework on prospective jurors. In the case of criminal records, the state is in the best position to aggregate the information and use it to remove citizens with felony convictions from the jury pool ex ante. Alternatively, the state can provide the litigants with this information as a matter of course, so that they can do with it what they want. Indeed, the Ryan trial may be somewhat of an outlier in light of the many resources devoted to the former governor’s defense by a large Chicago law firm.86 In the typical criminal case, prosecutors will have better access to criminal history databases than public or court-appointed defense counsel, and requiring the courts to collect and disseminate this information will remove troubling inequalities from the criminal justice system. Thus, on the whole, getting the courts more involved in collecting this information seems like a good thing. That said, we cannot evaluate the Northern District’s reforms without some recourse to first principles. We must decide whether the jury system is an effort to increase fairness in trials or merely another domain for lawyers to seek strategic and tactical advantages over their adversaries. The Supreme Court has developed numerous doctrines to police excesses in jury selection, most famously the Batson limitations on the use of peremptory challenges to strike racial minorities from juries.87 Under Batson, prosecutors violate the equal protection clause when they strike jurors from a panel solely on the basis of race. What is Batson’s objective? Is it an attempt to make race a nonfactor in the selection of jurors? Is it about ensuring that empanelled juries resemble jury pools, which in turn reflect the demographics of the citizenry as a whole? Does it apply to demographic factors besides race?88 Is it about ensuring that criminal defendants receive trials that are deemed legitimate and fair?

INFORMATION IS A VARIABLE, NOT A CONSTANT

Do Batson rights protect prosecutors and civil litigants as well?89 Or does Batson protect prospective jurors’ rights against mistreatment on account of race? The answers to these questions are important because they can help us evaluate the growing reliance on external sources of information about prospective jurors. In post-Batson cases, the Supreme Court has implied that all these interests are furthered by Batson.90 At the same time, by grounding its analysis in the equal protection clause and focusing on the harm to prospective jurors, the Court has suggested that a constitutional violation may have occurred even if the unlawful use of peremptory challenges did not determine the ultimate jury verdict.91 If making race a nonfactor in jury selection is Batson’s primary objective—as a mechanism for protecting the rights of either litigants or prospective jurors—then Batson and its progeny have not achieved their goal.92 Batson can be evaded by the lawyer who strikes some prospective jurors on the basis of race, so long as not “too many” prospective jurors are stricken on that basis and prosecutors articulate consistent rationales to explain their treatment of whites and blacks.93 Alternatively, the lawyer might successfully articulate a nonracial rationale, such as a prospective juror’s demeanor,94 for striking a disproportionate number of racial group members. Coming up with such a pretext is easy—recent experimental evidence suggests that college students and attorneys alike instinctively identify nonracial rationales for decisions that were driven by a prospective juror’s race.95 To make race a nonfactor, much stronger medicine than Batson would be required. Namely, attorneys should not be able to see prospective jurors during voir dire; rather, they would be limited to examining jurors behind an opaque screen (as in The Dating Game or an orchestral audition). Moreover, attorneys should be denied access to prospective jurors’ names and addresses, which often indicate racial background with substantial reliability and enable snooping attorneys to obtain a wealth of information about the jurors from third parties. The court might have to disguise prospective jurors’ voices and colloquialisms as well. This would be a plausible way of conducting jury selection, and it may be optimal as a means of curtailing troubling exclusionary practices, albeit at significant cost.96 At the other extreme, the law might address the problem of racially exclusionary uses of peremptory strikes by providing more information,

149

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

150

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

not less. If one adopts this view, the Northern District policy is merely a baby step in the right direction. The government could report to the parties jurors’ credit scores, military service records, bankruptcy filings, and involvement in prior litigation. It could review mental health records in the state’s possession to screen out those who might be unfit for service. It could scour public records and conduct LexisNexis searches to provide the parties with any relevant information. If the government did all these things, essentially providing dossiers on all prospective jurors, one might expect to see less discrimination on the basis of race, national origin, religion, gender, and other immutable characteristics. Indeed, a regime of full and symmetrical disclosure of juror profile information to the litigants conceivably could do more to combat the improper use of race as a proxy than Batson ever has. Alas, there are important complications here, suggesting that such a regime may be more appealing in the employment context than in the jury context. One would like to see careful data that controls for all the associations between race and other juror characteristics. Unfortunately, too little of the jury behavior research has done this successfully. The best available empirical evidence suggests that race is itself a strong predictor of whether a juror will vote to impose the death penalty in a capital case, with whites twice as likely as blacks to vote for the death sentence initially.97 Socioeconomic status, by contrast, has no predictive effect on initial votes by jurors. A capable defense lawyer thus might not be using race as a proxy for anything other than a juror’s propensity to send his client to death row. Indeed, black and white jurors appear to behave differently in a number of respects, such as their certainty that a defendant is guilty, their perceptions of the defendant’s remorse, and their assessments of the defendant’s future dangerousness.98 These behavioral tendencies evidently alter outcomes in capital jury sentences, with counties that have large African American populations imposing the death penalty at lower rates than counties with smaller African American populations.99 Some less careful survey evidence suggests that race is an equally important predictive factor in civil trials, swamping the effects of income, gender, and political inclinations.100 In civil cases, African Americans appear to be markedly more sympathetic to plaintiffs than Caucasians at both the liability and damages phases.101

INFORMATION IS A VARIABLE, NOT A CONSTANT

In the jury context, race may be an unusually valuable proxy for propensity to convict or impose liability on a defendant. If African American and Caucasian jurors approach legal controversies in fundamentally different ways because of their radically divergent interactions with police officers,102 or large employers, banks, insurers, and other institutional defendants, providing litigants with more information about all prospective jurors may do little to stem the reliance on race in voir dire by prosecutors and defense counsel. But if that is the case, it raises the question of whether the Batson doctrine is even a sensible one, or whether Batson ought to be replaced with affirmative rules mandating that each empanelled jury roughly reflect the racial diversity of the jury pool. Interestingly, the strategy of providing jurors with more information about the litigants can alleviate racial bias in some settings. Contemporary experimental evidence suggests that making the race of an African American criminal defendant salient reduces the propensity of whites to convict or impose sentences in a racially disparate manner.103 The experiments exposed mock jurors to testimony concerning a vehicular homicide, and presented the same evidence as to the African American defendant’s culpability. In the experimental condition, the mock jurors were exposed to testimony suggesting that after the accident, a crowd gathered and yelled racial epithets at the driver. White jurors who heard this testimony were much less likely to convict, and more likely to view the defendant’s case on the merits as a strong one. It is possible that jurors were more lenient because they became more sympathetic to the defendant, but a body of research suggests that alerting white jurors to the racist attitudes of some other whites may cause those jurors to reflect more carefully on their own propensity to convict and compensate for the presence of racial bias.104 A different consideration might temper our embrace of pervasive reputation information in the voir dire context. Jury duty is already viewed as an unappetizing prospect for many Americans. Jurors can be, and often are, identified by the courts while a trial is proceeding, subjecting jurors to potential intimidation and harassment.105 Such a loss of privacy can compromise the quality and integrity of jury deliberations. As Judge Richard Posner writes, “Most people dread jury duty—partly because of privacy concerns.”106 The further loss of privacy associated

151

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

152

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

with comprehensive government background checks could prompt stiff resistance and exacerbate juror absenteeism. To be sure, a savvy juror appreciates the possibility that litigants will gather information about her already, but Americans often feel more dread about governmental possession of private information than they do upon realizing that such information has been obtained by nongovernmental actors.107 Even more important, the loss of privacy associated with submitting a job application or trying to find housing will not be sufficient to deter reasonable people from the enterprise. Employment and housing are necessities, so we can safely assume that the behavioral distortions associated with pervasive reputation screening will be minimal. Not so for jury duty. Jury duty is compulsory, but the government cannot enforce its compulsory nature on the citizenry without incurring substantial costs and generating significant resentment. It may be the case that it is necessary for the government to forgo gathering a great deal of pertinent information about prospective jurors to avoid diminishing the quality of the pool of willing jurors.108

Medical Stereotyping Employers’ use of statistical discrimination is troubling enough, as it endangers people’s livelihoods. When doctors and other health professionals use the same strategies for coping with incomplete information, it risks patients’ lives. Yet the literature on health disparities suggests that doctors behave in much the same way that employers, trial attorneys, and other decision makers do, with wrenching consequences for African American patients. Why would doctors use statistical discrimination in treating patients? One possible explanation not well explored in the medical literature is that statistical discrimination may be a strategy for staying out of prison. In recent years, state and federal law enforcement officials have become more aggressive about tracking the prescription of opioids by pursuing physicians who prescribe medication that is subsequently diverted into the black market.109 Surveys of physicians show that many are very concerned about prescribing narcotics to patients suffering from pain because of a fear that such prescriptions will trigger drug enforcement

INFORMATION IS A VARIABLE, NOT A CONSTANT

agency investigations,110 with a sizable minority of physicians admitting that they “give patients a limited supply of pain medications to avoid being investigated” and a larger minority expressing concern that “prescribing narcotics for patients with chronic pain is likely to trigger a drug enforcement agency investigation.”111 Although one would hope that prosecutorial discretion will prevent prosecutions of doctors who prescribed in good faith subsequently diverted narcotics, this fear of law enforcement investigations is not misplaced. Surveys of prosecutors suggest that doctors face a genuine risk of being charged with serious crimes, even if their decisions to prescribe narcotics were medically defensible.112 Additionally, many jurisdictions track physicians’ prescriptions of narcotics, forwarding copies of all filled prescriptions to law enforcement personnel.113 Successful prosecutions of physicians who were duped by their patients into dispensing narcotics have garnered substantial media attention.114 When a doctor risks incarceration if her patients divert prescribed medication to the black market or abuse their medication, it seems likely that she will try to sort between those patients who are risky and those who are less risky. Race is a plausible proxy here, and there is evidence suggesting that doctors use it in prescribing narcotics. A survey of 397 patients being treated for chronic nonmalignant pain and their primary care physicians found that, although African American patients reported experiencing more pain than Caucasian patients,115 Caucasians were significantly more likely to be treated with stronger and longer-acting opioids. Even after the researchers controlled for socioeconomic and other factors, Caucasians were more than twice as likely as African Americans to be given opioids. Notably, there was no correlation between race and the use of non-opioid pain drugs. What explained the racial disparity? Not differential access to insurance or other payment options, and not divergent preferences among African American and Caucasian patients. The study authors concluded that the “more pronounced racial differences for strong and long-acting opioids suggest that systematic mistrust, bias, or stereotyping phenomena could be in play.”116 As with statistical discrimination in employment, there is a large literature to suggest that doctors treat patients differently on the basis of race in a variety of settings. Although fear of prosecution is a plausible explanation

153

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

154

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

for statistical discrimination in the opioid prescription context, African American patients appear to receive worse treatment in a variety of settings, even after controlling for socioeconomic factors.117 This inferior treatment can be explained by physicians’ attitudes. A well-designed study by Michelle van Ryn and Jane Burke found that physicians view African Americans as more likely to abuse medication and less likely to comply with doctors’ orders.118 These differences do not go unnoticed by African American patients.119 Statistical discrimination on the basis of socioeconomic status also has been observed in some settings. For example, nephrologists are more likely to refer a child of higher socioeconomic status than one of lower socioeconomic status to a transplant doctor, with the likely explanation being the belief that wealthier parents are more likely to ensure compliance with a rigorous postoperative recovery regime.120 It would be tempting to say that these disparities could be eliminated by imposing liability on physicians who engage in statistical discrimination or by embracing efforts to reeducate physicians to act in a colorblind manner. But scholars of health disparities understand the inadequacies and dangers of such approaches.121 Just as litigation and employer diversity training are incomplete remedies for discriminatory hiring practices,122 they would be incomplete remedies in the prescription context. Ana Balsa, Thomas McGuire, and Lisa Meredith sensibly note that doctors’ “reliance on ‘priors’ related to age, gender, or race, when low-cost reliable tests are available, is difficult to justify.”123 In the treatment of pain, however, where existing diagnostic constraints often force doctors to rely on a patient’s own statements and visible acts to discern the extent of his suffering, there are no reliable low-cost tests. But there are information-based strategies that can reduce the appeal of statistical discrimination. When doctors encounter patients who they believe are abusing prescription medication, diverting medication to the black market, overstating symptoms, or failing to comply with protocols for taking medications or recuperating, they routinely include this information in the patient’s medical records.124 If the contents of these medical records were easily transferred from one physician to other physicians likely to encounter a particular patient, then doctors would not need to engage in statistical discrimination on the basis of race and other problematic proxies. The statistical discrimination problem, in short, helps make a compelling case

INFORMATION IS A VARIABLE, NOT A CONSTANT

for the computerization of medical records, a process that is proceeding slowly in many parts of the country.125 A large literature documents the costs of continued reliance on antiquated medical recordkeeping. 126 Diagnosis becomes more difficult. Patient choices of doctors are effectively constrained. Tests and other procedures may be duplicated unnecessarily. I would argue that in addition to all these problems, our failure to bring medical recordkeeping into the twenty-first century may be contributing to distasteful discrimination and that African American patients are enduring unnecessary pain and inappropriate treatment as a consequence.

The Reputation Revolution and Non-TrespassBased Exclusion If my argument in this chapter is right, then the dramatic increase in previously private information about individuals has made bouncer’s exclusion much easier than it was previously. Under the model that lies at the core of this book, the result of this change will be twofold. First, as I have already argued, bouncer’s exclusion will become more fine-grained and more precise. As a general matter, this is a positive development because exclusion will relate increasingly to the behavior of individuals, not groups. Second, as bouncer’s exclusion becomes more efficient, there will be a shift away from exclusionary vibes and exclusionary amenities. This section will evaluate the effects of that latter change. Let us put together the lessons from chapters five and eight. The existence of asymmetric information might cause a resource owner to delegate the task of excluding those deemed undesirable from the owner’s bouncer to the prospective entrants themselves. At the same time, the reputation revolution has radically diminished information asymmetries. Therefore, we can predict that resource owners will, going forward, rely more heavily on bouncer’s exclusion, which permits them to control access to a resource themselves, and less heavily on exclusionary vibes and exclusionary amenities. Application of the theory so far provides a heartening lesson: bouncer’s exclusion, which is what lawmakers and legal scholars focused on in the twentieth century even though a lot of the action was in vibes and amenities, is becoming the main mechanism for exclusion once again in the twenty-first.

155

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

156

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

That straightforward assessment is largely correct, although there are some complexities that muddy the waters a bit. First, the reputation revolution is a two-way street, and it is likely that new information technologies will assist in the collection and analysis of data about decisions to exclude. Controversial exercises of the bouncer’s right can therefore be made more transparent to the constituencies that might object, and those affected may agitate for voluntary inclusion or regulatory interventions. For example, if the behavior of cab drivers is meticulously tracked, it may be feasible to discern which of them are engaged in unlawful discrimination. Second, and more critically, there are reasons to think that the reputation revolution will also lower the costs of exclusionary vibes and exclusionary amenities. The more information resource owners have about prospective resource users, the better they can calibrate their communicative or bundling strategies. Data mining and increasingly sophisticated behavior profiling can facilitate far more accurate amenity tailoring than was previously possible. By profiling those who do use a resource that is access-restricted via vibes or amenities, resource owners will be able to infer who is not using the resource as a result of those strategies. The upshot, then, is that it will be easier to test and perfect exclusionary vibe and exclusionary amenity strategies as a result of the reputation revolution. Although the increased ubiquity of personal information and the ease with which such information can be analyzed will, at the margins, shift resource owners toward bouncer’s exclusion, the effect will not be so dramatic as to wipe out non-trespass-based exclusion. The persistence of differential legal regulation of competing exclusionary strategies will drive other owners toward exclusionary vibes and exclusionary amenities as well.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

9 CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

The preceding chapters suggest that the widespread availability of information about individuals and firms ought to alter the way we think about law and public policy in a variety of domains. In some settings, such as the employment context, disclosing previously private information about individuals may prove to be a desirable government intervention. In other settings, such as the juror selection process, there is a stronger argument for maintaining the privacy of information about individuals. This chapter provides a new typology of government information policies and draws some general lessons from our tour through various legal subjects. We are conditioned to think about using law to create private incentives in terms of two well-known tools: “carrots” and “sticks.” The carrot approach rewards desirable private behavior (perhaps through tax incentives, subsidies, or positive recognition), and the stick approach punishes undesirable behavior (through criminal fines, civil liability, or the condemnatory use of the bully pulpit). Upon reflection, however, carrots and sticks are not all there is. We are also familiar with “curtains.” Government can try to make potentially observable characteristics obscure so as to make it more difficult for private decision makers to act on the basis of observable characteristics. This is what legislators do when they attempt to restrict genetic discrimination, and it is what progressive reformers attempted to do when they created nonpartisan elections, thereby preventing voters from electing judges or mayors on the basis of party affiliations. Various 157

158

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

doctrines limiting the sorts of information that attorneys can collect about prospective jurors follow the same tack. These curtain strategies have been advocated in the antidiscrimination context as well. Indeed, some efforts to reform antidiscrimination law have suggested that statistical discrimination can be mitigated if the relevant decision makers are deprived of information about a job applicant’s race, religion, or gender. With less information, decision makers presumably will focus more on the applicant’s qualifications and less on the applicant’s skin pigment. “Searchlights” are less familiar than curtains, but they provide a fourth policy alternative, and an important one. The state can make private discrimination on the basis of illegitimate or misleading observable characteristics less appealing by making legitimate or informative characteristics more easily observable. This is what the government does when it mandates the placement of visible health inspection signs rating the hygiene of each Los Angeles restaurant outside that establishment’s front door, a policy that has significantly reduced hospitalizations from food-related illnesses, increased revenues for restaurants sporting high hygiene grades, and reduced revenues for restaurants that must advertise their barely passing grades.1 In 2010, New York City one-upped Los Angeles by publishing its own public health inspection data on all restaurants and rendering the raw data easily importable into online restaurant review sites like Zagat, Yelp, or TripAdvisor.2 The searchlight strategy is hardly limited to food cleanliness. The government pursues a searchlight strategy when it publishes information about the identities of those with criminal histories.3 This is the basic strategy behind mandatory S.E.C. disclosures in corporate law. 4 By giving investors access to a wealth of information, the government helps them rely less on rumors, sales pitches, and intuitions. It animates California’s widely emulated Data Security Breach notification statute, which requires firms to inform consumers whenever the firms’ actions have exposed the consumers’ personally identifiable information to the public.5 And this is the same intuition driving the implementation of the federal government’s Toxic Release Inventory, which requires polluters to disclose publicly every year their emissions of specified toxic pollutants.6 Both the data security and toxic release regulatory regimes enable

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

affected parties to learn about harmful activities so that they may take corrective measures. As important, the public nature of such mandated disclosures allows other third parties like insurers, investors, bond holders, and others to use transparent information about past slip-ups to weigh future risks associated with a firm. We can find informationforcing rules in a variety of legal domains, and it is probably no coincidence that these strategies have been employed more frequently in the Internet era. Note that the searchlight framework described here eschews any actomission distinction. Just as we can equate the absence of a carrot with a stick, at least in the presence of many other carrots, we can understand the absence of a government searchlight as equivalent to a curtain. In some settings, the law will reduce the observability of individual attributes through affirmative acts (like privacy tort protections); in other settings, through omissions (like failures to make data about public servants accessible online).7 An information asymmetry might result just as easily from an affirmative government act as from its failure to act. To the extent that there are relevant differences between these acts and omissions, they would stem from their differing social meanings. This searchlight strategy will not always be the optimal one, just as carrots, sticks, and curtains may fail us at times. But it is axiomatic that we can achieve more with four tools than we can with three. The remainder of this chapter will identify the settings in which searchlight strategies are well suited or poorly suited to addressing social ills.

Animus-Based Discriminators and Pretext Where statistical discrimination is more prevalent than animusbased discrimination, policy makers should employ searchlight strategies. Where animus-based discrimination is more prevalent, curtain strategies will be appropriate. The reason why is rather straightforward: it is easy for animus-based discriminators to identify a pretextual reason for rejecting an applicant when they have lots of information about the applicant, and harder for discriminators to point to a legitimate nondiscriminatory basis for an adverse decision when they lack information about the candidate. For example, suppose that an employer dislikes African Americans

159

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

160

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

and refuses to hire a well-qualified African American applicant on the basis of the applicant’s race. This would constitute animus-based discrimi­ nation. If the employer has access to information about the applicant’s credit history, social relations, academic record, prior employment evaluations, and the like, it will be relatively easy for the employer to claim falsely that information contained in those resources explains the decision not to hire the applicant. No applicant is perfect, after all. By contrast, if all the employer knows about the applicant is his skin color, it will be impossible for the employer to identify a neutral justification for the refusal to hire. In short, the searchlight strategy can backfire when it provides an animus-based discriminator with additional information. For this reason, information about decision makers’ true motives is crucial, and it is important that public opinion scholars explore the connection between groups’ perceived behavioral tendencies and the actual behavior of group members. Without a reliable understanding of the relationship between those two variables, policy makers cannot determine which contexts lend themselves to particular information-based antidiscrimination strategies. There may be settings in which not knowing the ratio of animus-based discrimination to statistical discrimination paralyzes government efforts to decide between curtains and searchlights. For example, in the jury selection context, it is worth asking if parties’ motivations matter when they seek to strike jurors of a particular race. If Batson is designed to address an injury to prospective jurors, it would seem as though the prosecutor who strikes African Americans from the panel because he hates African Americans is more dangerous than the prosecutor who does so because he believes it is more likely to result in the conviction of an African American defendant.8 But while there are indications in the case law that Batson is designed to protect jurors,9 the doctrine is invoked by litigants, and these litigants have no incentive to make the inflammatory argument that the other side is motivated by something other than a desire to maximize their odds of winning at trial. In any event, the analysis here suggests an inadequacy in Batson’s effective treatment of all litigants as statistical discriminators. The searchlight strategy will only reduce reliance on the prospective jurors’ race if the litigant whose actions are challenged is engaged in rational, statistical discrimination.

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

False Information Accurate information is a necessary part of any effort to combat statistical discrimination via government information policy. Some of the information discussed herein—military records, records of criminal convictions, bankruptcy records, immigration and naturalization documents, and the like—will not pose insurmountable accuracy problems. To the extent that errors occur, they will often revolve around cases of mistaken identity. Where multiple people share the same name, and one of them has committed a crime or filed for bankruptcy, the entire group may be penalized.10 These sorts of problems can be ameliorated through the use of supplemental identifiers, such as birthdates, birthplaces, and partial Social Security numbers. A more serious challenge arises in the context of identity theft. If a bad actor successfully hijacks the identity of a good actor and then uses the good actor’s identity to defraud unsuspecting consumers, significant damage can result. The consumers will be misled. The good actor’s reputation will be trashed unfairly. And confidence in the reputation system as a whole will be eroded. But here is the rub: reliance on searchlight strategies might facilitate identity theft because private information is often used to authenticate people’s identities. The logical response is to discourage reliance on personal history information for authentication purposes. A regime that widely publicizes consumers’ birthdates is a regime in which any bank or credit card company would be foolish to use birthdates as a basis for authentication. The costs of transitioning away from biographical information toward password-based authentication and biometrics will not be excessively burdensome. Nor will they be negligible. Because identity theft represents such a significant threat to reputation-reliant dispute avoidance and resolution strategies, minor tweaks to business practices need to be made. The most daunting false information problems arise in the context of data that is not contained in existing public records, but rather would be the product of new government searchlight initiatives. Nobody believes that the inaccurate feedback problem can be solved entirely in reputation-tracking environments. For example, if the government tries to improve the efficiency of the labor market by subsidizing the collection of

161

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

162

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

360-degree feedback and making that feedback transportable across firms, then it will have to deal with deliberately or unintentionally false feedback that employee A provides about employee B. Critically, if an employer relies heavily on coworker evaluations in deciding not to hire a seemingly qualified applicant, but those evaluations are themselves the product of coworker animus or implicit bias, then the government strategy might be counterproductive.11 There are strong reasons to believe that the problems can be ameliorated significantly through the use of algorithms designed to locate outlier data points, along with spot-checks and audits relying on objective verification.12 Given enough information, these algorithms can identify instances where an individual provides suspiciously high levels of negative feedback toward members of any racial or gender group. Once the animus or implicit bias is identified, the system can then adjust the weight assigned to that individual’s feedback, substantially offsetting the harm done. Let us begin with a disturbing but inevitable hypothetical. Suppose a racist employee persistently gives terrible performance evaluations to African American and Asian American coworkers, and good or neutral performance evaluations to Caucasian coworkers. Designers of online feedback systems have encountered this sort of problem, and they have developed algorithms to detect feedback that is unusually skewed. Essentially, the system discounts outlier scores—instances in which a buyer gives negative feedback on an overwhelmingly well-rated merchant or vice versa.13 There is a cost to eliminating these outliers, in that a good merchant sometimes behaves badly, just as a good driver sometimes makes mistakes on the roadways.14 That said, in an environment like eBay, where most users are behaving honestly, algorithms designed to weed out likely false reports improve system performance.15 As long as reputation systems elicit a lot of user feedback and racists provide a minority of the evaluations, isolating and ignoring problematic feedback is reasonably straightforward. In the next chapter we will consider what happens when very large numbers of users are providing low-quality feedback. The picture is not so pretty. It is worth emphasizing that online reputation-tracking technologies are still in their infancy, and dramatic improvements to existing systems for identifying false feedback can be expected in the years ahead. For example, by gaining more information about feedback creators and recipients, the

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

system could police racist or otherwise problematic feedback quite effectively. For example, if a 360-degree feedback system knows the race of various employees, it can be programmed to discount or even ignore the ratings of white employees who routinely assign suspiciously high levels of negative feedback to African American employees. In other words, so long as we are willing to seed a reputation system with information about characteristics that might form the basis for inaccurate feedback, we can develop algorithms to address deliberate inaccuracy. We cannot solve the problem of intentionally racist feedback entirely, but we can mitigate it substantially. The thornier problems arise when people do not realize that they are being racist in the way they evaluate their peers, but implicit bias is causing them to give skewed marks. Even in these difficult settings characterized by implicit bias, modern evaluation systems have an important role to play. Contemporary psychological research has examined the ability of decision makers to act defensively to control implicit bias and stereotypes.16 Yet one cannot compensate for her own implicit biases without being made aware of them and having some reason to want to control them.17 Here again, recent experimental evidence is encouraging. It suggests that discerning a disturbing pattern in one’s own feedback enables a person who has some motivation to ameliorate her own discrimination to take effective corrective measures.18 Feedback systems can both lower the costs of identifying such patterns and provide reviewers with a stronger motivation to provide unbiased feedback if they understand that feedback exhibiting a significant disparate impact on discrete groups will be accorded less weight. Eventually, well-designed systems may even enable individuals to reduce their own biases at the subconscious level, permitting them to take anti-stereotyping measures without incurring much in the way of added cognitive load.19 Hence, tracking feedback can educate not only the recipients of such feedback, but its providers as well. The way in which reputational information is packaged for decision makers will also be critical. Cognitive psychologists are investigating ways in which information can be presented in a manner that minimizes the likelihood of stereotype-based responses.20 Frames and cues matter. A well-designed searchlight strategy should employ cognitive psychologists to help ensure that information transmitted to decision makers enables them to analyze and act upon that information in a rational way.

163

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

164

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

To reiterate a point made earlier, more information is not always superior to less information, but more information is usually better if it can help decision makers identify their own biases, suppress those biases, or learn to make decisions in a way that overcomes those biases. It is sometimes tempting to use the imperfections in feedback systems as a basis for rejecting their use. But the appropriate contrast is not between a world of perfectly accurate information about individuals and the sometimes flawed information that an eBay-style reputation tracking mechanism can generate. There are daunting false feedback equivalents in the real world. A poor man buys one expensive suit to appear prosperous. A debt-saddled person drives a very nice car or eats at nice restaurants to appear well off. A middle-aged person has plastic surgery to appear young. A person of average intelligence wears geek-chic glasses to appear smarter. An individual fakes an accent to appear worldly. A would-be adulterer at a pick-up joint removes a wedding ring to appear unmarried. An American traveler knows that her countrymen have sometimes behaved badly abroad, so she sews a Canadian flag to her backpack. Reputation tracking systems like eBay’s, Slashdot’s, and TripAdvisor’s introduce some false feedback. However, because most feedback providers are sincere, and because algorithms can help the owners or users of these sites weigh more heavily the feedback provided by reviewers who have proven their reliability, the signal-to-noise ratios are often quite high. There will be contexts in which disseminating feedback information from consumers may not be appropriate. Some aspects of the doctor-patient relationship seem to fall into this category. Patients are quite capable of assessing physicians’ bedside manner, their ability to communicate, their promptness, and the like. But in many cases, patients will do a poor job assessing physicians’ diagnostic skills, their surgical techniques, or the accuracy of their prognoses (at least in the short run). There will be selection effects that make such assessments more difficult to evaluate, too, although feedback can be normalized statistically based on the vagaries of a physician’s patient population. More troublingly, if patients are rating physicians and physicians are rating patients publicly, there may be a strong incentive for physicians and patients to trade unduly favorable feedback with each other. This “Pollyanna effect” has played out on eBay, with the result that, until a 2008 change to its feedback policy, eBay exhibited

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

significant grade inflation in its feedback ratings.21 Given the strong interests identified above in having physicians provide accurate—albeit perhaps unflattering—assessments of their patients in their medical records, the appropriate regime might (a) make physician assessments of patients available only to other physicians in the absence of a court order, or (b) impose a time lag, whereby patients could only assess physician characteristics after some period of time. Either of these interventions will decrease the likelihood that patients will provide strategic feedback instead of sincere feedback. At most, patient evaluations should be one of many factors that consumers use to pick their physicians. The more technical the skill set offered by a medical specialist, the less weight evaluative data should receive.

Overinvesting in Reputation Proxies are necessarily imperfect, and those imperfections can encourage manipulation. Take the case of apples. Consumers usually cannot taste apples at the grocery store, but taste may be a decisive consideration in determining whether they will be satisfied with their purchases. Because the redness of an apple is a proxy for its deliciousness, consumers may sensibly rely on its appearance as an indicator of its taste.22 If they do so, problems may arise. More precisely, apple producers might start investing in growing techniques that make apples redder but not tastier, degrading the quality of the redness signal. In a world where consumers care only about the taste of apples, and not their appearance, these expenditures to make apples redder but not tastier will represent wasteful expenditures. Since increasing redness always comes at some cost, producers will either raise prices or decrease investments in the tastiness of their apples. Along the way, the reliability of the redness signal will diminish, leaving consumers increasingly adrift as they try to figure out which apples to buy. What is true of red apples is also true of reputations. A strong reputation merely correlates with desirable attributes; it is not a perfect proxy for those attributes. As a result, there is a lingering danger that increased reliance on individuals’ reputations for sorting purposes will prompt individuals or firms to overinvest in actions that will improve their reputations.23 For example, a professor might pander to students by providing them with

165

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

166

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

free baked goods on the last day of class, which not coincidentally is the same day that the students will fill out teaching evaluations. Or worse, the professor may try to entertain students throughout the semester at the expense of teaching them. Alternatively, a hotel might provide monetary incentives for its customers to provide favorable reviews, rather than lowering room rates, making capital expenditures that will improve its amenities, or devoting more money to salary so that better-skilled workers will seek employment there. Though they may boost feedback ratings,24 these sorts of wasteful activities also have the potential to degrade the quality of a reputation-rating resource. For the latter reason, ratings websites like TripAdvisor have devoted substantial resources to trying to punish firms that employ these tactics.25 The teaching evaluation context is easier to monitor. While some students will reward a teacher who gives away cookies on the last day of class with stronger evaluations, other students are likely to resent the pandering and note their disapproval of the tactic on their anonymous student evaluation forms. If the professor’s colleagues find out about pandering, then they will discount the significance of the professor’s generally positive evaluations and apply negative peer pressure in the interest of preventing student evaluations from becoming a noisy indicator of teaching quality. In short, there is an effective whistle-blowing mechanism that will deter excessive investments in reputation, as long as there are a sufficiently large number of raters, some heterogeneity in attitudes regarding expenditures that enhance reputation but not service quality, and reluctance on the part of feedback providers to lodge false accusations of pandering. It is even possible that the overreliance on particular proxies might contribute to their deterioration as a signal of quality. Just as the redness of apples might eventually come not to correlate at all with their taste, it is conceivable that so many people are using the same proxy to evaluate individuals’ future behavior that an initially effective proxy ceases working. At that point, the people or firms that can best figure out how to develop a newer, better proxy will have an advantage over the rest of the pack. This is the role for information intermediaries. Such “reputation professionals” will always be able to profit by noticing the presence of strategic behavior before everybody else does. As with picking stocks, this is not an activity in which part-timers are likely to have the upper hand.

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

More generally, it is useful to examine the ex ante effects of ubiquitous personal information. There are obvious upsides and downsides to such a regime. If we are dealing with everyday interactions among people, then it seems likely that the ex ante effects will be quite positive.26 That might be one takeaway point from Bob Ellickson’s extended case study of Shasta County, California, where a well-functioning gossip network facilitated the formation and enforcement of a set of social norms that made the community as a whole much better off.27 It is also a fair takeaway from my own research into motorist behavior, which concludes that the practical anonymity of drivers vis-à-vis one another encourages antisocial driving.28 In other contexts, though, the ex ante effects of reputation monitoring will be undesirable. For example, having every move watched and profiled might discourage socially beneficial forms of identity experimentation.29 Broadly speaking, the use of ubiquitous personal information and feedback will be most desirable when majoritarian norms are particularly unproblematic. As a corollary, we should recoil at the thought of using negative peer evaluations to punish political dissidents. Just because everyone thinks that Joe’s rantings on a soap box are nonsensical does not mean it is desirable for the state to silence him. When political dissent is at issue, society has long recognized the value in letting unpopular or unfashionable arguments be voiced. More broadly, majoritarian sentiment may be too quick to condemn intellectual, political, or artistic innovation. As a result, insecure geniuses whose ideas might have ultimately prevailed if protected by anonymity will be too discouraged by the high costs of nonconformism. For that reason, using citizen evaluations to award, say, National Endowment for the Arts grants could inappropriately reward those artists whose work is not artistically excellent but coincides with the aesthetic preferences of the median voter. Peer evaluations from fellow artists have their problems, but they are surely a superior means of dispensing scarce resources.

Kings in Disguise King Abdullah of Jordan is known for donning various disguises and mingling with his subjects to get a better sense of what life is like for ordinary Jordanians.30 The incognito king has waited in lines at government tax offices, observed traffic regulation from behind the wheel of a

167

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

168

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

taxi, and posed as a television journalist to get a sense of life in Jordan’s free trade zone. Media accounts of the King’s disguised exploits are usually laudatory, suggesting that the experiences enable the monarch to avoid getting an unduly rosy account of life in his country.31 While the King of Jordan is not walking around the average American city, it is possible that disguised reputations help keep our government officials in line as well. Take instances of police brutality in major American cities. Scandals about excessive force often emerge because the police roughed up a prominent minister or elected official who is a member of a minority group.32 By seeing the way in which a prominent, powerful, and law-abiding member of that group is treated, the public may learn about how their less prominent and powerful but nevertheless law-abiding peers are routinely treated. Similarly, to some readers, the story of a Paris Hermès saleswoman refusing to admit Oprah Winfrey and her entourage to the store in the last few minutes before closing time was newsworthy because it suggested the persistence of racism in French society.33 Henry Louis Gates’s arrest by the Cambridge Police on his own porch, having been suspected by a passerby of breaking in, is another contemporary, high-profile iteration of this dynamic.34 But in a world with ubiquitous reputation information, events like these would not occur. It is therefore worth asking whether making reputation information more broadly available would cause the mistreatment of the powerless to get swept under the rug. Should we tolerate some extra police brutality toward the powerful, in the hopes that the resulting outrage will raise public consternation and ultimately reduce the general incidence of police brutality? That seems like a perilous strategy.35 The possibility that an African American motorist might be a well-known minister or city councilman will constrain the behavior of police officers somewhat, and all members of the community therefore will be protected from brutality to some extent. But facilitating sorting creates significant advantages that offset this loss of protection. A police officer pulling over an African American motorist in an economically depressed community will most definitely feel a constraint on his behavior if he learns that the motorist is the nephew of a well-known minister, a receptionist in a prominent city councilman’s office, or the nephew of another police officer. Not every middle-aged African American man will turn out to be a famous Harvard scholar, but a lot of people are connected to people with

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

clout. Officers might actually be surprised to learn how many people can be effective whistle blowers. There is a tradeoff here, but it is probably one society ought to be willing to make. If officers approach motorists or pedestrians in a depressed area thinking that anyone who isn’t a celebrity is part of an undifferentiated mass, then the King Abdullah effect may do little to keep them honest. And if, as seems plausible, police officers usually underestimate the likelihood that an individual is connected to someone with clout, then less anonymity for citizens probably will be a net positive.

When Can Reputation Systems Outperform Courts? In many instances, the government has the best access to information that decision makers will want to use. Criminal records, bankruptcy records, military service records, immigration and naturalization records, academic records from public schools or state-run universities, and records regarding membership in licensed professions are obvious examples.36 Private firms may be better than the government at aggregating this information and publishing it in a format that will be useful to producers or consumers.37 In other instances, valuable information will be generated by private parties, and the government might face political or agency constraints that prevent it from generating equally accurate information. For example, there is little reason to urge the government’s involvement in the generation of consumer credit scores. Although Experian and the other credit scoring agencies sometimes make mistakes, their incentives are properly aligned, and they are insulated from interest group pressure regarding the formula used for credit scores. Similarly, the government should not get into the business of running social networking sites or developing auction websites as a mechanism for combating statistical discrimination. There are market actors with substantial comparative advantages over the government, and they are already doing a fine job of making new information available to the public. In these settings, the government’s role should be confined to facilitating the adoption of uniform standards—for example, through subsidies—so that information can be aggregated easily from among a number of different social networking websites, and reputations can be transported from one auction site to another.

169

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

170

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

In perhaps the majority of instances, the most appropriate role that the state can play in facilitating the development of a robust reputational market is to get out of the way. Take an illustrative example: In July of 2007, Avvo­­­­­­­­­­­­­­­­­ ­ ­­ .com, a website that rates lawyers, was launched. The website aspired to do for attorneys what Zagat did for restaurants: provide consumers with information that they could use to find a suitable lawyer and collect evaluations of lawyers from fellow attorneys and clients.38 Alas, within ten days of its launch, Avvo was sued in a class action lawsuit by attorneys alleging that Avvo had violated Washington State’s Consumer Protection Act by disseminating unfair and deceptive information about lawyers who were rated by the site.39 More precisely, the complaint faulted Avvo’s website for being subjective and unreliable; providing questionably low numerical ratings to U.S. Supreme Court justices, law school deans, and other highly regarded lawyers; using a nontransparent methodology for developing lawyer ratings; and providing incomplete information.40 The suit’s lead plaintiff, Browne, was an attorney who claimed to have lost two clients as a result of a low Avvo rating, a rating that was tied to a state bar disciplinary proceeding against him which had resulted in an admonition.41 Avvo moved for dismissal in short order, arguing that its services were protected by the First Amendment and that the plaintiffs had failed to state a claim under the state’s Consumer Protection Act.42 The district court appropriately granted the motion for dismissal on both grounds.43 As the court surely understood, if Avvo were liable for its conduct, then it seems likely that Zagat may be liable to Pizza Hut if unfavorable restaurant reviews result in a poor rating, U.S. News and World Report may be liable to Texas Wesleyan University Law School for placing that school near the bottom of its rankings, and eBay may be liable to vendors who cannot make sales because they have poor feedback ratings. Under these circumstances, there are superior alternatives to litigation. Namely, individuals like Browne who believe that false information has been disseminated about them ought to have a right of reply—an ability to explain why they believe they have received inappropriate ratings from a website or a complaining consumer.44 In workplace settings, the available empirical evidence suggests that providing subordinates with a right to respond to feedback from managers enhances the effectiveness of feedback systems.45 This right of reply is something that is already built into eBay’s

CARROTS, STICKS, CURTAINS, AND SEARCHLIGHTS

and TripAdvisor’s feedback systems. eBay users and TripAdvisor vendors typically employ the right when they believe they have unfairly received negative feedback. An attorney like Browne could make use of his right of reply to note that other lawyer-rating services, like Martindale Hubbell, rate him highly. He could assert that Avvo’s methodology for calculating lawyer ratings is flawed, using the example of Ruth Bader Ginsburg’s implausibly mediocre rating as a case in point. He could, in short, make many of the arguments that his lawyers made in his complaint to alert consumers to the deficiencies of the Avvo rating and entice potential clients back into the fold.46 Just as Congress has enacted section 230 of the Communications Decency Act47 to avoid chilling Internet discussion, it or the courts should provide services like Avvo with immunity against tort suits stemming from unflattering ratings, so long as the defendant offers the poorly rated individual or firm a right of reply similar to eBay’s. Such a rule would permit a vendor to point out possible biases that formed the basis for an unfair rating. Reputation ratings systems are an alternative to formal adjudication and criminal enforcement. In many cases, these systems will provide the most efficient mechanism for resolving and deterring disputes. Bringing the complex, slow, and costly legal system into run-of-the-mill disputes over whether Tavern on the Green’s Zagat décor rating should be a 24 or a 27, or whether Browne’s attorney score should be a 5.1 or a 6.2, endangers an important and dynamic aspect of the nation’s private reputational infrastructure. The result of legal liability for reputation trackers like Avvo would have been fewer ratings, less information for consumers, and greater reliance on more problematic bases for deciding upon an attorney, like race, limited word-of-mouth data, law school attended, media visibility, claims in attorney advertisements, and the like. Of course, removing the possibility of liability in cases where inaccurate feedback is reported on a ratings website creates the potential for the quality of the published feedback to suffer, just as eliminating defamation liability might cause the quality of newspaper reporting to decline.48 That said, a decline in quality would by no means be inevitable. The question in both cases is whether market forces provide adequate incentives to keep the information on reputation rating sites generally accurate. With respect to this question, it is appropriate to give a provisional affirmative answer. Although there is always the potential for websites providing biased product

171

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

172

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

reviews to receive heavy traffic, it has not happened. Consumer Reports, published by a nonprofit that accepts no advertising from the merchants whose products it reviews, easily outsells its ad-supported competitors.49 Search engines that auction off the top responses to search queries do not have nearly as many users as those that place more popular or more widely linked web pages at the top of their search results. U.S. News and World Report’s rankings of graduate schools and undergraduate institutions—flawed though they may be—do not face serious competition from publications that sell off top rankings to high-bidding universities.50 To the extent that there are variations in the extent to which websites are objective, websites that “rate the raters” can point out those websites that consumers should view with suspicion. There are, in short, rather robust mechanisms for promoting accuracy on reputation rating sites, and it seems far from clear that legal liability for potentially erroneous statements would enhance overall welfare. A final paradox is worth noting. There may be some instances in which the increased use of rating sites will actually lead to more litigation or new regulatory initiatives. One of the positive features of these systems is they have a tendency to educate populations about problems that perturb many but are unlikely to be voiced. The litigation process, by contrast, tends to bring to the forefront problems that affect a few people a great deal. Three important groups may respond aggressively when they discover that a number of people are complaining about the same pathology via a feedback site. Journalists might seize on the story and report it more widely. Regulators and elected officials, responding to both the feedback itself and the media coverage of it, may press for interventions that will please consumer constituencies. Finally, class action attorneys, the legal actors who specialize in cases involving small harms to large numbers of people, might start filing suits, particularly if they see the opportunity to recover substantial fees. It seems likely that, on balance, feedback sites will reduce the need for costly litigation and regulatory interventions, but by creating a focal point where people can gather and gripe, the sites might facilitate legal interventions that would not otherwise occur. And that might not be such a bad thing after all.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

10 WINNERS AND LOSERS

The reputation revolution that industrialized societies are currently undergoing is neither all good nor all bad. In some cases, the legal challenges posed by new technological capabilities are wrenching, and in a few instances those challenges are so severe that they warrant restricting the use of the technologies in question. Moreover, because the reputation revolution seems poised to create a world that resembles the small towns of yore far more than they resemble the urban and suburban environments in which most of us now live, readers should at least feel uneasy about the process by which we might achieve heightened trust, reduced fraud, and decreased statistical discrimination. This chapter examines the normative questions that should be on the minds of policy makers when they consider whether to use carrots, sticks, curtains, or searchlights in tackling a policy problem. It argues that determining who will bear the costs and benefits of a policy intervention ought to be a primary consideration, along with the obviously relevant magnitude of those costs and benefits. Finally, it tentatively explores the question of what it might be like to live in a society where individual reputation is increasingly transparent.

Some Thoughts About the Desirability of an Information Strategy My primary objective here is to identify the ways in which technological and social developments will alter the foundational assumptions 173

174

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

upon which the law is based, and then to examine how the law might respond to those developments. I do so for several reasons, mostly having to do with my desire to engage as many readers as possible and to help readers with conflicting normative priors understand the stakes at issue in particular public policy domains. I am not a moral philosopher, let alone a good one, so I see little reason why the reader should care about my views on the propriety of various forms of discrimination. At the same time, apparent normative judgments inevitably manifest themselves throughout this book. For example, I take the position here that some forms of discrimination (such as statistical discrimination on the basis of race) are particularly undesirable and other forms of discrimination (such as statistical discrimination manifested as an unwillingness to hire ex-offenders, based on the supposition that an ex-offender is probably less trustworthy than an individual with no criminal record) are tolerable. Although I do, in fact, believe that discrimination on the basis of race is worse than discrimination on the basis of criminal history, it is not those priors that guide my analysis here. Rather, my analysis is based on the premise that policy makers in all jurisdictions have decided that racial discrimination is unlawful and lawmakers in most jurisdictions have concluded that criminal history discrimination is not. I take these judgments as a given, noting that nearly a dozen states have limited or even prohibited the reliance on prior convictions as a basis for denying employment but that other jurisdictions have prohibited ex-felons from becoming state government employees.1 Once we recognize the choice between discrimination on the basis of race and discrimination on the basis of criminal history, it is difficult to imagine anyone favoring the former over the latter. In our world of imperfect enforcement of antidiscrimination laws, treating all forms of discrimination as equally problematic ensures social welfare losses. One useful way of getting at this legal hierarchy issue is to consider two attributes that are presently correlated and that plausibly will prompt less consensus among readers than the race versus criminal history distinction. What is worse: discrimination on the basis of sexual orientation or discrimination on the basis of HIV status? HIV-positive individuals are protected against discriminatory treatment by the Americans with Disabilities Act,2 yet some employers may still prefer to keep HIV-positive individuals out of their workplaces, for reasons rational (concerns about rising group

WINNERS AND LOSERS

health insurance premiums) or irrational (stubborn concerns about the possibility of HIV transmission via casual contact). In the United States, HIV historically has been disproportionately prevalent among homosexual men, with male-to-male sexual contact remaining the predominant method by which Americans contracted the disease.3 On reasonable assumptions, homosexual men are ultimately sixteen times as likely as heterosexual men to contract the HIV virus.4 Given the substantial stigma associated with HIV and the relatively high costs of providing health insurance for HIV-positive employees, it is plausible that homosexual men are victimized by statistical discrimination designed to keep HIV-positive individuals out of the workplace.5 One possible strategy for combating this potential statistical discrimination would be to publicize the HIV status of every American. That would be a bad idea. As demonstrated by the Americans with Disabilities Act and a host of common law decisions treating HIV status as a “private fact” whose disclosure is highly offensive to a reasonable person,6 Americans have decided that HIV status itself ought to be a protected classification. By inference, decisions classifying individuals on the basis of HIV status may be nearly as bad as decisions classifying them on the basis of race. Federal law presently gives greater antidiscrimination protections to HIV-positive individuals as a class than it does to homosexuals as a class. But suppose a reader believes that the law has it wrong. Perhaps in this reader’s view, discrimination on the basis of sexual orientation is worse than discrimination against those with HIV. There are plausible rationales that can be invoked to support this view. Homosexuality might be less likely to be caused by an individual’s voluntary choices, sexual orientation may be a more central part of people’s identities, homosexuals may be more likely to be victimized by violence and animosity, or antidiscrimination protections for homosexuality may be less likely to come into conflict with public health interests. As long as the two forms of discrimination are not equally offensive and the other considerations discussed in previous chapters are satisfied, it would be appropriate, under this framework, to publish individuals’ HIV status as a means of alleviating the statistical discrimination that HIV-negative, openly gay men currently endure. So in a world where the law’s present hierarchy of antidiscrimination interests is flipped, publishing individuals’ HIV status might make sense.

175

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

176

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

Things would get more complicated if the law decided that all discrimination is equally bad. Formally, the law holds that discrimination on the basis of race is unlawful regardless of whether the victims are white males or African American females. When one looks more closely at the allocation of government and private resources, as well as popular attitudes, a more nuanced structure emerges. Discrimination on the basis of race against African Americans is regarded by many as worse than discrimination on the basis of race against Caucasians. African Americans and Native Americans are given particularly strong protection because of the prevalence and intensity of the historic disadvantages they have faced. Discrimination on the basis of religion is bad, but it has not exercised American lawmakers or the public to the same extent as discrimination on the basis of race. To equate two forms of discrimination is, as this chapter shows, to remove a potentially useful antidiscrimination arrow from the law’s quiver.

Incidence The discussion so far has been premised on the idea that it is desirable to help, for example, law-abiding African American males at the expense of African American males with criminal records. To defend that proposition, it is worth exploring the counterarguments—why someone might want to make it more difficult for private decision makers to sort among those with criminal records and those without criminal records. There are several possible justifications for this approach. First, we might expect that if all criminal records are widely publicized, those with criminal records will be harmed more than those without criminal records will be helped. Accordingly, facilitating effective private sorting will make the group of job applicants as a whole worse off. Second, we might believe that there are powerful policy justifications for preventing sorting, perhaps because we believe that criminals who have served their time deserve a clean slate. Third, we might believe that the criminal justice system is essentially corrupt, in which case facilitating sorting merely enhances the unjust penalties meted out by an arbitrary government apparatus. I will consider each of these arguments in turn. First, the available empirical evidence suggests that African American males are more likely to be hired by firms that conduct criminal background checks than by similarly situated firms that do not. Recall that Holzer and

WINNERS AND LOSERS

his coauthors found that those responsible for hiring appeared to overestimate the propensity of African American males to have criminal records and hired too few African Americans as a consequence.7 Publicizing accurate information about individuals’ involvement in the criminal justice system should only adversely affect a group’s available opportunities to the extent that decision makers (a) underestimate the prevalence of criminal records among members of a particular group or (b) are effectively prevented from engaging in statistical discrimination to sort out those with criminal records. The Holzer study undercuts both claims in the race-criminal history context. It is plausible that decision makers in employment settings are more concerned about erroneously hiring someone with a criminal background than they are about not hiring someone without a criminal background.8 Because of the information environment in which corporations operate, it may be easier for a firm to recognize the harms associated with a bad hire than the harms resulting from the failure to make a good hire. If decision makers are indeed more concerned by the prospect of making a bad hire than failing to make a good hire, but are not particularly risk averse about the prospect of disparate impact antidiscrimination liability, then publicizing information about criminal records may make the group that was previously the target of statistical discrimination better off as a whole. To the extent that we are concerned about the welfare of a group that is victimized by statistical discrimination, we should limit antisorting strategies to those settings in which employers are engaging in statistical discrimination while at the same time underestimating the correlation between a group classification and an undesirable characteristic. Second, we might think that private sorting harms vital social interests and justify keeping criminal histories obscure for that reason. For example, we might believe on policy grounds that the availability of employment opportunities for ex-convicts will discourage recidivism. Alternatively, we might have an abstract ideological commitment to the proposition that “everyone deserves a second chance” or, more narrowly, that “people who have served their time have repaid their debts to society and should be able to start off with a clean slate.” These sorts of arguments sometimes find their way into the information privacy case law9 and the pertinent academic literature.10 In this case, the appropriate question to ask is: What is the

177

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

178

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

optimal strategy for preventing these negative externalities? The sensible way to answer is by drawing on the tools of optimal redistribution analysis.11 Privacy is one way to promote second chances, but it is rarely the best way to do so. A better way to facilitate the hiring of ex-convicts who deserve a second chance is through direct subsidies to employers who hire them. Such programs have been implemented, with the discontinued federal Targeted Jobs Tax Credit providing one example12 and the current federal tax code’s Work Opportunity Credit providing another.13 Because it is inexpensive for the government to identify ex-convicts, the government can efficiently ensure that only genuine ex-convicts benefit from subsidized second chances.14 Moreover, since the program is funded out of general tax revenues, the costs of promoting second chances is borne by taxpayers as a whole. Compare that tax regime to the status quo. We try to facilitate the hiring of ex-convicts by raising private decision makers’ costs of sorting between ex-convicts and those with no criminal records. As a result, many employers use statistical discrimination tools to penalize those without criminal records, and the ex-convicts who do get hired are likely to be members of groups whose baseline offending rates are low: white males and females of all races. If the law continues on the path of promoting second chances by making it hard to determine who has slipped up in the past, only some of the beneficiaries will be ex-convicts, and the costs of this program will fall heavily on a group that includes other ex-convicts and innocent people who share demographic characteristics with ex-convicts. From an optimal redistribution perspective, there is little reason to prefer our present approach to a tax credit. If reintegrating ex-offenders into the nation’s economy and social fabric is welfare-enhancing, and there are good reasons to believe that it is, then direct subsidies or insurance programs, not information privacy protections, are the most appropriate lever for facilitating that reintegration. There is also an important role for private organizations to play in tackling this social ill. For instance, the San Francisco–based Delancey Street Foundation is a successful nonprofit whose mission is to reintegrate ex-felons into society15 by putting them to work in restaurants, moving companies, landscaping companies, and a host of other ventures.16 Delancey Street publicizes the nature of its workforce, and some customers patronize the foundation’s businesses because they are attracted to its social objectives, whereas others just seem to like the food or service provided.17 Its workforce

WINNERS AND LOSERS

is “not just polite, but energetically polite, and there is a complete lack of foul language.”18 The foundation perhaps benefits from the precariousness of its reputation in light of the attributes commonly associated with ex-convicts. It recognizes that it has more to lose than competitors do from unhappy customers, and it responds to consumer complaints accordingly. Its reputational vulnerability becomes a strength, a sort of bonding mechanism with its customers. Transparency for criminal histories need not create a caste system. But it might require ex-felons to bond themselves to entities like Delancey that use screening mechanisms, bonding mechanisms, training programs, and other strategies for regaining the trust of ordinary citizens and consumers. Delancey Street represents the classy end of ex-offender reentry programs, but the model has been duplicated by relatively crass entrepreneurs as well. Felony Franks is a hot dog stand operated by Jim Andrews in a rough neighborhood on Chicago’s west side. Andrews employs only ex-offenders and sells items like the “misdemeanor wiener” and the “cell mate dog.” Andrews argues that he is helping ex-offenders reintegrate themselves into the labor force and hopes to expand into dozens of neighborhoods. Local Chicago politicians seem supportive of the goal but resentful of a branding campaign that arguably glorifies crime. The local alderman has repeatedly denied Andrews a permit to construct a sign for his hot dog stand, but Andrews wonders about the hypocrisy of a city government that is all too happy to let Hooters advertise as blatantly as it wants.19 The final justification for obscuring information about criminal offenses is connected to disturbing inequalities within the criminal justice system. More precisely, if the criminal justice system is systematically biased against African Americans, Latinos, and men, then a system whereby the government publicizes the crimes of African Americans, Latinos, and men will worsen existing inequalities. This argument comes the closest to providing a compelling reason for suppressing criminal history information about individuals. If criminal punishments are indeed meted out arbitrarily to members of minority groups, then a Rawlsian focused on the lot of the least-well-off members of society should reject the proposal that I have advanced.20 Someone who subscribes to a welfarist or utilitarian framework should not. To a Rawlsian, some form of reputational affirmative action would be necessary to render a searchlight strategy desirable in the context of the employment of those with criminal records.21

179

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

180

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

If we accept a weaker version of the “arbitrary criminal justice system” thesis, then a welfarist certainly need not reject the searchlight approach to criminal history information. More precisely, even in a nation whose criminal justice system discriminates systematically against African American males, there may be important, merit-related differences between those African American males who have criminal records and those who do not. Some of the former will be innocent, but far more of the latter will be innocent. Indeed, because the criminal justice system is biased against African American males, those African American males who nevertheless avoid run-ins with the law should be particularly desirable in the market for jobs where trustworthiness is important and the applicant pool contains a large number of untrustworthy job seekers.22 Why shouldn’t we help decision makers identify these particularly desirable individuals with greater ease?

Social Meaning It is possible to critique the optimal redistribution analysis put forth above while staying within a welfarist framework. The analysis would proceed as follows: Outright prohibitions on discriminatory conduct are preferable to subsidies for nondiscriminatory conduct because the former will instill or strengthen antidiscrimination norms and the latter will not. By this logic, the implementation of prohibitive discrimination policies will eventually change the preferences of the populace, making the discriminatory impulse rarer in the long run.23 A policy that permits discrimination but subsidizes employment of the group discriminated against, on the other hand, might signal tacit approval of the discriminatory conduct at issue, perhaps making discrimination more socially acceptable—and hence more prevalent—in the long run. The story has intuitive appeal, but there are equally plausible stories we can tell that will point us in the opposite direction. The question of whether tax subsidies for firms that employ ex-convicts are desirable is an empirical one on which there is conflicting evidence. A widely cited 1985 study by Gary Burtless suggests that welfare-recipient job seekers whose employment was subsidized fared poorly in the labor market. The presence of the vouchers signaled to employers that the applicants were on welfare, and employers otherwise would have had difficulty discerning who was a welfare recipient.24 As a result of this research and important subsequent scholarship, some scholars have taken a

WINNERS AND LOSERS

dim view of targeted tax credits.25 Yet more recent work suggests that the program Burtless studied was designed so as to maximize the welfare stigma26 and that well-designed tax incentives to employ disadvantaged workers can be effective at increasing their employment.27 And perhaps more to the point, Burtless and other contributors to this field have not explored the troubling possibility that in the absence of a tax subsidy program to help employers sort between the disadvantaged and the nondisadvantaged, employers will instead sort applicants based on their race. A final and related question concerns the social meaning of government information policy. When the government chooses to publish information about attribute x but not attribute y, the populace may understand this policy as an indication that attribute x is relevant or salient but attribute y is not.28 This dynamic is part of a plausible critique of Megan’s Laws, which disclose information about sex offenders’ criminal histories and whereabouts without disclosing the same information about murderers who have completed their sentences, causing communities to overreact to the presence of some ex-offenders and underreact to the presence of others. The theoretically attractive solution to this problem is to release as much information as possible, relying on private actors to distill this information into a format that consumers can readily use. But if individuals respond irrationally to such information or use it in ways that further their self-interest at the expense of collective welfare, then dissemination should be carefully structured. If disclosure cannot be structured so as to ameliorate these dynamics, publication may prove undesirable altogether.29

Price Discrimination The greater availability of consumers’ personal information enables forms of price discrimination that would not have been feasible prior to the reputation revolution. Price discrimination can take multiple forms, but the classic example is of “a firm charging multiple prices for the same good where the difference in price is not attributable to a corresponding difference in cost.”30 Firms price discriminate to extract more money from consumers who are willing to pay more than the market clearing price. To take a short-lived but notorious example, Amazon.com began using information from Internet cookies to engage in dynamic pricing. Loyal

181

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

182

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

customers who frequently bought from Amazon were charged higher prices and customers who were directed to Amazon from comparison-shopping websites were charged less.31 Amazon dropped the practice in the face of significant customer complaints.32 Amazon’s reliance on proxies like repeat purchases or use of bargain-hunting websites necessarily entailed lumping together groups of consumers—a practice that the economics literature refers to as third-degree price discrimination.33 Amazon could extract much more consumer surplus if it was able to charge prices that perfectly reflected each consumer’s willingness to pay for a product—what economists call first-degree price discrimination.34 Put another way, in the absence of perfect information about every individual, price-discriminating firms are required to statistically discriminate in their pricing policies. The widespread availability of personal information about individuals’ behaviors, preferences, and reputations enables firms to shift toward behavior that more closely approximates first-degree price discrimination.35 Provided that selling firms have some market power, can limit arbitrage, and are marketing to consumers possessing varied price elasticities of demand, price discrimination will enable those firms to capture what would otherwise be consumer surplus under a fixed pricing model.36 Price discrimination is prohibited by law only where it adversely affects consumer welfare.37 This is as it should be. The effects of price discrimination, unlike the effects of racial discrimination, are ambiguous. Perfect firstdegree price discrimination will enable a firm to increase output, improving social welfare, but at the cost of a diversion of surplus from consumers to producers.38 In other words, perfect price discrimination will permit firms to offer useful products that would have never seen the light of day if all purchasers had been charged the same amount. So the pie gets bigger, but consumers may get smaller slices of it. The welfare effects of third-degree price discrimination are also context-dependent.39 Even the distributional consequences of price discrimination are indeterminate. While price discrimination necessarily shifts surplus away from consumers, it also enables poor consumers who would otherwise be unable to afford a product the opportunity to obtain it at a reduced price. For that reason, price discrimination often entails a progressive redistribution of resources.40 In short, the desirability of price discrimination is ultimately an empirical question with varied answers in different contexts. Sometimes, the existing empirical work

WINNERS AND LOSERS

suggests that the benefits of price discrimination outweigh the harms.41 We simply do not know the welfare or distributional consequences of facilitating price discrimination in e-commerce generally, in landlord-tenant markets, or in most of the other settings discussed herein. But we do know that price discrimination considerations should be an essential part of the calculus in determining when searchlights or curtains are desirable. That is because searchlights facilitate price discrimination and curtains constrain it.

The “Am I Hot or Not?” Society In October of 2000, James Hong launched a strange new website called AmIHotorNot.com.42 In the site’s first month of operation, more than 20,000 individuals submitted photographs (of themselves, typically) to the site so that other users could rate their physical attractiveness on a scale of 1 to 10.43 Although many Americans scratched their heads about the website’s success, it was a viral hit with teenagers and young adults, and it remains online today, although it has morphed into some odd combination of a beauty rating site and an Internet dating site.44 In the intervening years, cultural trends have suggested at least a limited embrace of the “rate me” mentality exemplified by AmIHotorNot. Millions of Americans have begun blogging about their personal thoughts, sharing their homemade videos on YouTube, and dreaming about competing on American Idol, which is in the midst of a strong run as the most popular television show in the United States.45 Not long before a scandal involving a referee’s alleged involvement in game-fixing made headlines, NBA Commissioner David Stern bragged that professional basketball referees were the “most ranked, rated, reviewed, statistically analyzed and mentored group of employees of any company in any place in the world.”46 Those of us in the academy can relate to these referees, as we have come to expect that our teaching will be evaluated by the students at the end of each course and perhaps via websites like RateMyProfessors.com. Such treatment is not reserved for service providers, referees, and professors. The website getunvarnished.com allows any person to write up a Yelp-style assessment of any other person’s reputation for all the world to see.47 Negative reviews cannot be deleted by their users or anyone else, though other users can evaluate each review’s veracity, causing the

183

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

184

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

reviewer’s reliability score to rise or fall. One need not be a member of getunvarnished.com to be a subject of others’ reviews and at press time the website did not appear to have a mechanism, short of a defamation suit against the company, by which people could opt out of being reviewed on the site. In 2008 the Northwestern University Law Review published an article of mine called Reputation Nation. The article began to sketch out what the implications would be of life in a society with ever-present smartphones, omnipresent peer ratings, and ever-diminishing privacy.48 My thesis in that work—the idea that making individualized information widely available would curtail some pernicious forms of discrimination—is echoed in this book. In 2010, the talented novelist Gary Shteyngart described a fictional (but not that fictional) AmIHotorNot society with great wit and in rich detail. Super Sad True Love Story, Shteyngart’s novel, offers a dystopian vision to supplement my more optimistic take.49 Shteyngart’s America of the not-too-distant future is one in which individuals wear small äppäräti around their necks. These äppäräti are sophisticated mobile communications devices that stream constant updates about news, media, and shopping to their wearers, while simultaneously broadcasting information about their wearer to the äppäräti belonging to those nearby. When a person walks into a bar, her physical attractiveness and personality ratings are compared to that of all the other patrons, so that she can be ranked against the others in the room. The same is true of her credit score, bank account information, consumer and romantic predilections, and her medical information.50 There are “credit poles” scattered throughout the city that reveal the creditworthiness of everyone who walks by them.51 When a man gets on a plane without an äppärät, the passengers around him are unnerved. It is not long before soldiers show up to drag the anonymous passenger off the plane.52 Super Sad True Love Story is indeed super sad, and it tells an engaging love story. The pertinent question for our purposes, though, is whether the book rings true. Let us try to make progress on that question by quoting a few particularly interesting passages from his text. In the first passage, the protagonist (Lenny Abramov) enters a bar and describes his first experience with “FACing” (short for “Forming a Community”), which is billed as “a way to judge people. And let them judge you”:53

WINNERS AND LOSERS

Streams of data were now fighting for time and space around us. The pretty girl I had just FACed was projecting my male hotness as 120 out of 800, personality 450, and something called sustainabilit ¥ at 630. The other girls were sending me similar figures. . . . Vishnu worked my äppärät until some rankings came up. He helped me navigate the data. “Out of seven males in the Community,” he said, gesturing around the bar, “Noah’s the third hottest, I’m the fourth hottest, and Lenny’s the seventh.” “You mean I’m the ugliest guy here?” I ran my fingers through the remnants of my hair. “But you’ve got a decent personality,” Vishnu comforted me, “and you’re second in the whole bar in terms of sustainabilit ¥.”54 This formulation of life in a privacy-free world suggests that superficiality has triumphed over substance. Beauty is widely regarded as skin deep, but people are making snap judgments about each other’s personality and financial wherewithal too after glancing at summary data. Snap judgments and quick glances seem to be all that anyone is capable of in Super Sad True Love Story’s “community.” In a haunting passage toward the end of the novel, Lenny tries to show his young lover, Eunice Park, that literature can stir the soul by reading to her from The Unbearable Lightness of Being. The college-educated Eunice has a hard time following Milan Kundera’s text, and confesses to Lenny that she “never really learned how to read texts, just to scan them for info.”55 If intelligent, educated people like Eunice have lost the ability to be reflective and American society generally has become irredeemably crude and crass, then subjective assessments of reputation will only create an echo chamber that reinforces existing pathologies. Social scientists doing research in teacher performance evaluations have occasionally reached similar conclusions. For example, Elizabeth Davison and Jammie Price, professors at Appalachian State University, performed a content analysis on the RateMyProfessors ratings written by students at their school and chastised both their “anti-intellectual tone” and the students’ tendency to provide good marks to instructors whose classes were deemed easy and poor marks to instructors whose courses were perceived to be hard.56 I am unfamiliar with the educational environment at Appalachian State University, so I do not know whether anti-intellectualism is reflective

185

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

186

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

of campus life, nor do I know whether the teachers being evaluated critically there are teaching in a way that ultimately will benefit their students. Taking Davison and Price’s claim at face value, my own experiences suggest that there is variation among the schools on this point. As deputy dean at an unabashedly intellectual professional school, reading all of the teaching evaluations filled out for every course is part of my job description. Among the thousands of evaluations I read last year, I can recall only a handful that could be characterized as even remotely anti-intellectual.57 Bouncers and exclusionary vibes have shaped our pool of raters. Although Davison and Price reached the sweeping conclusion on the basis of their study that “student evaluations are not good data,”58 there is nothing remotely antiintellectual about teaching evaluations if the students doing the ratings are themselves intellectuals. Some crowds are wiser than others. One of Shteyngart’s subtle and subversive moves is to show us that while the judgments Lenny encounters in the bar are distressing, the judgments people encounter when they are unplugged may be even worse. Eunice is a Korean American who is addressed by poor public housing residents as “Chinita” and “China Honey.”59 This sort of crude ethnic stereotype is the sort of thing uttered by people who have no access to äppäräti. Shteyngart, through Lenny’s narration, shows later in the novel how the äppäräti can at least fix this sort of mistake: While we were lunching at one of the louder places on 32nd Street, Eunice saw a man eating by himself and sipping a Coca-Cola. “It’s so sad,” Eunice said, “to see a Korean man without a wife or girlfriend to tell him not to drink that junk.” She lifted up her cup of barley tea as if to show him a healthier alternative. “I don’t think he’s Korean,” I said to Eunice. “My äppärät says he’s from Shanghai.” “Oh,” she said, losing interest as soon as her bloodlines to the solitary Asian Coca-Cola drinker were cut.60 Facts have some ability to correct erroneous subjective judgments, but low stakes, low interest, or low empathy might prompt people with access to pertinent information to ignore facts that are staring them in the face. Perhaps the most bracing aspect of Super Sad True Love Story is the extent to which individuals have come to depend psychologically on being rated

WINNERS AND LOSERS

by peers and strangers. As part of the novel’s climax, much of the nation’s communications network shuts down as the American economy and government simultaneously collapse. People begin taking their own lives, and it is the collapse of the communications network, not the economic or political collapse, that prompts the tragedies. Lenny writes: Dear Diary, My äppärät isn’t connecting. I can’t connect. It’s been almost a month since my last diary entry. I am so sorry. But I can’t connect in any meaningful way to anyone, even to you, diary. Four young people committed suicide in our building complexes, and two of them wrote suicide notes about how they couldn’t see a future without their äppäräti. One wrote, quite eloquently, about how he “reached out to life,” but found there only “walls and thoughts and faces,” which weren’t enough. He needed to be ranked, to know his place in this world. And that may sound ridiculous, but I can understand him.61 This statement encapsulates much of the technological pessimism that runs through Shteyngart’s satire. Like much of the rest of the book it rings true with some of our lived experiences. Young adults who graduate from college or professional school and enter the workplace sometimes confess to friends that they miss being graded on the work they hand in. They are told by employers that they are doing well, but yearn to know “How well?” and “Am I doing better than Jones?” The impulse of teenagers who post pictures of themselves on AmIHotorNot.com and then eagerly await the praise or condemnation of the masses is hard to understand unless we suspect that Oscar Wilde had it basically right. The only thing worse than being rated poorly is not being rated at all. It seems distressing to think that people’s desire to be rated by their peers would create such dependence. Yet upon reflection, we might come to the conclusion that a desire to be thought well of by others is what motivates so much of the social cooperation necessary to make the world a pleasant place. To take just one example of many, we might wonder why most large gifts to charity are not made anonymously. After all, the announcement of such gifts necessarily reduces the donor’s financial privacy and opens her up to more solicitations, and the anonymous donor presumably feels at least as

187

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

188

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

much of a warm glow from giving as the named donor does. The Almighty does not need a donor to be named in order to reward good deeds in the afterlife. The almost inevitable conclusion one reaches about philanthropy is that most people shy away from anonymous large gifts because they want the high status that comes along with generous public giving. Shteyngart’s text makes FACing and the rest of his variant on the reputation revolution seem at the very least disheartening. The suicides described in his text are a part of what makes his story super sad. But we should be somewhat cautious about relying on isolated anecdotes, and this is especially true of fictitious anecdotes. We should instead ask about the effects of this observation on real people. There is a slowly growing empirical literature that addresses this important question. Employees in many industries face constant evaluation by coworkers, customers, and supervisors. By way of example, sociologists have explored the effects of being a call center employee whose actions are constantly under surveillance. Insofar as business-related calls were being monitored and recorded, they did not present significant concerns to the employees being monitored, though the employees expressed significant resistance to the idea of their employers monitoring private calls from friends and relatives.62 Other research on call centers finds that when employers articulated training and career development rationales for electronic monitoring, employees felt much higher levels of job satisfaction than when the rationale was perceived in terms of deterring misconduct.63 Similarly, when researchers at Stanford began studying the effects of publishing feedback generated by speed-dating encounters between opposite-sex experiment subjects, they found some surprising results.64 The researchers matched up pairs of people and asked them to provide feedback about their fellow research subject’s behavior in the speed-dating exercise. In some variations, the feedback was kept private, and in other circumstances, each partner’s feedback about the other partner was publicized. Moreover, in some variations, the feedback included substantive comments, whereas in other variations it consisted entirely of numerical scoring. Participants were asked to assess the extent to which they felt monitored, conformist, and self-conscious in the various experimental set-ups. They reported that having numerical feedback about them made public, but without any basis for the feedback, felt most invasive and conformist, and made them most self-conscious. Feedback that was shared with both parties was preferred,

WINNERS AND LOSERS

and numerical feedback accompanied by substantive explanations for the scores was viewed as far more acceptable than numerical feedback alone.65 If this result is broadly generalizable—and other research in very different contexts suggests at least parts of it may be66—then it suggests that ordinary people may be willing to embrace ubiquitous feedback systems, provided they are sufficiently transparent and universal. Feedback disclosure should be targeted toward those with some legitimate interest in the information at issue. People are most likely to feel that feedback systems are fair and useful when they are used as a teaching tool. The opportunity to receive and react to constructive criticism (or praise) seems to generate goodwill among the subjects of pervasive monitoring. Designing feedback systems that encourage raters to provide this sort of feedback and discourage unhelpful sniping is a critical design challenge. The broader question of whether such rating is good for people’s psychological well-being is more difficult to answer. Individuals who grow up in small towns where their reputations are easily tracked by neighbors commonly describe some aspects of their existence liberating and others as stifling. It is easy to trust your neighbors when you know them well. It may be hard to grow up in such a small town when one’s own values or behaviors are contrary to those of the community. No one is surprised by the tendency of cities to become magnets for homosexuals from small towns.67 The tradeoff between feeling safe and feeling alienated from one’s surroundings is not an easy one. There is, of course, an important difference between the AmIHotorNot submitters, small-town residents, call center employees, speed daters, referees, and professors whose activities are evaluated frequently and those whose activities are rarely, if ever, subject to evaluations. The difference is consent. For the reasons identified above, the option of emigrating from the reputation nation will be more of a theoretical possibility than a practical one. If an individual chooses to opt out of using reputation-tracking technologies, then many people understandably will assume the worst about that individual, even if it does not result in that individual getting dragged off an airplane by the army.68 If enough people have opted in to this system, then opting out will become effectively impossible. The legal system may push firms to insist on the use of these sorts of technologies, even where they are unpopular among employees and

189

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

190

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

customers. If firm A elects not to conduct invasive background checks on its employees but firm B decides to leave no stone unturned, A may be subjected to heightened liability risks. What if an employee is rejected by B on the basis of a background check and then secures employment with A? Suppose that the employee then injures a customer in a way that might have been predicted had firm A conducted the extensive background checks that are standard operating procedure at firm B. The customer sues firm A for failing to take reasonable steps to protect him against harm. Now the deck is stacked toward A mimicking B, even if the employees and customers collectively would prefer for the firm to maintain its existing policies. This is a problem. One solution is for the law to immunize firm A against this sort of vicarious liability. That solution is hardly a satisfying one. It imposes all the costs of firm A’s nonintrusive policies on a customer who derived little if any benefit from those policies and suffered the losses disproportionately. Imposing liability on A looks better comparatively, since such liability would spread the pain of compensating the customer among A’s employees, shareholders, and managers. Alternatively, to the extent that society is genuinely interested in letting firms and individuals make autonomous decisions about how fully they should embrace the reputation revolution, it might consider a social insurance approach. The harm to the customer will be paid for by everyone in society, including the customers and employees of A, B, and every other firm. Such a step will help ensure that the move to embrace the reputation revolution is genuinely majoritarian, though by removing firms’ incentives to consider the costs imposed on the injured customer, it arguably stacks the deck unduly in favor of privacy. Is there a good reason to stack the deck in that respect? Privacy theorists have long argued that protecting privacy is essential so that individuals can relax,69 experiment with different personalities to figure out who they truly are,70 or develop the insights that will make them more productive citizens.71 True enough, the private sphere of the home will remain a respite largely free of rating, and there will be market demand for zones of privacy where everyone will agree to suspend the use of rating technologies; it is easy to imagine the proprietors of the Las Vegas strip going this route, though it seems likely that they would first establish minimum reputational requirements for entry into the reputation-free zone. The question, though, is whether those zones of privacy are sufficient to prevent the societal harms

WINNERS AND LOSERS

to which privacy theorists have pointed. In line with the peculiarities of Americans’ attitudes toward privacy generally,72 and the well-recognized dangers of surveillance in one-party states, part of the answer may well hinge on the extent to which the state can be prevented from utilizing widely available personal information to identify, intimidate, or otherwise disfavor members and supporters of the political opposition. The key point here is that privacy prevails in some portions of our lives for very good reasons and in other aspects of our lives largely because of path dependence. Privacy is necessary where it furthers vital instrumental goals and where no substitutes can further those interests as effectively. Privacy in the bedroom lets people love each other and facilitates intimacy. Privacy at the ballot box ensures that citizens in a democracy are voting for the candidates they actually favor, not the candidates whose vengeance they fear. Privacy for medical records can encourage individuals to seek out diagnostic testing and treatment, although the law sometimes favors disclosure over privacy where privacy might place the lives of third parties at risk. Super Sad True Love Story is at its most disconcerting where privacy in bedrooms and at ballot boxes is eviscerated, regardless of what the people whose information is being disclosed want. It is not that everyone needs privacy; some people in contemporary society readily disclose information about their sex lives, political behavior, and medical conditions to perfect strangers. But without some control over whether sensitive information about them is widely known, the capacity of many people to live fulfilling lives is substantially compromised. In other contexts, though, privacy should become obsolete. Until recently, practical obscurity created a wide zone of privacy on urban freeways. But new technologies linking smart phones with driver information databases may soon permit someone to report the license plates of dangerous drivers so that other drivers, law enforcement, and insurers will be able to see the information and respond accordingly. We may have come to expect practical anonymity on the roadways, but there is no evidence that such anonymity makes people’s lives better on the whole, and a lot of evidence to suggest that roadway anonymity causes numerous accidents and widespread stress.73 Driving in small towns where everybody knows everybody else is safer, more pleasant, and hardly less free. The transition to a world with no anonymous driving may be jarring, but it will leave us better off.

191

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

192

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

DIVERSITY-PROMOTING STRATEGIES

A critical question, then, is what parts of our society are like bedrooms and what parts are like roadways. Increasingly, contemporary neighborhoods, marketplaces, and selective workplaces and schools are coming to resemble roadways, where the sort of practical obscurity that prior generations had gotten used to is beginning to break down. This is a technological development that the law could try to resist, and understanding more about what is at stake will help policy makers understand whether such intervention is appropriate. As the previous chapter makes clear, the decline of privacy will not necessarily result in the decline of exclusion, though it will certainly modify the strategies by which exclusivity is achieved. A related critical question is whether privacy can be compartmentalized, such that its disappearance on the roadways or in the marketplace would not threaten its vibrancy in bedrooms or at ballot boxes. It is tempting to think that privacy must be defended in all settings so that it survives where it matters most. Yet a competing story may resonate more with the facts. After all, European law and culture have long provided governments with awesome powers to gain information about their citizens but tied the hands of businesses that wish to collect consumer information for marketing purposes. In the United States the reverse is true: the law and culture have limited what the state can do while barely restraining industry efforts to conduct surveillance of online activities or engage in data mining.74 The absence of privacy protections in one domain has not brought about their disappearance from the other, although it has perhaps exerted some pressure toward convergence.75 What explains the stubborn divergence between the laws of public and private surveillance on both sides of the Atlantic? Why has neither the European nor American legal system converged towards a uniformly high- or low-privacy equilibrium? Perhaps the lesson here is that privacy protections are reasonably resistant to technological threats where those protections are grounded in sound arguments that resonate with dominant cultural beliefs. If so, it is plausible that by defending privacy to the hilt in contexts where it is actually detrimental, advocates weaken the persuasiveness of privacy arguments in those settings where a great deal is at stake. The more liberally privacy is invoked, the less force it has as an objection.

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

CONCLUSION

Robert Moses, New York’s nearly omnipotent builder of infrastructure, understood exclusion in all its shapes and forms. Jones Beach was one of Moses’s most impressive legacies—a pristine six-and-a-half-mile beach on Long Island that became one of the area’s most popular weekend destinations. Moses decided that keeping the beach beautiful required keeping African American bathers from spending time there. So he vetoed the Long Island Railroad’s proposal to construct a branch leading to the beach. And then he insisted that newly constructed parkways linking the beach to population centers be covered with bridges that were too low to let buses use the parkways. He required that buses entering the beach obtain permits, and then made it difficult for buses chartered by African American groups to get those permits. He bucked tradition by imposing parking fees at New York state parks, reserving the highest fees for Jones Beach, and he refused to lift them when his boss, then-governor Franklin Delano Roosevelt, repeatedly requested that he do so. Through all these mechanisms, Moses acted as a bouncer on the beach, with the objective of keeping Jones Beach white.1 But he did not rely on bouncer’s exclusion alone. So he tried subtler strategies as well. For reasons that will remain a historical mystery, Moses was convinced that African American bathers would not swim in pools whose water was kept cold. He therefore instructed park workers to keep the pools at Jones Beach icy cold. And he tried to signal to different populations where they should go via his choice of lifeguards—a strategy 193

194

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

CONCLUSION

he called “flagging.” African American lifeguards employed by the state parks commission were dispatched to relatively inaccessible and poorly developed beaches. African Americans who somehow managed to make it to Jones Beach or other desirable beaches would not see any black lifeguards there.2 With respect to park facilities located within Manhattan, public transportation restrictions and infrastructure limitations were not going to be sufficient to allow Moses to achieve his racist ends. But relying on exclusionary vibes and exclusionary disamenities proved more successful. As Robert Caro tells it: Whether it was the temperature or the flagging—or the glowering looks flung at Negroes by the Park Department attendants and lifeguards—one could go to the pool on the hottest summer days, when the slums of Negro and Spanish Harlem a few blocks away sweltered in the heat, and not see a single non-Caucasian face. Negroes who lived only half a mile away, Puerto Ricans who lived three blocks away, would travel instead to Colonial Park, three miles away—even though many of them could not afford the bus fare for their families and had to walk all the way. The fact that they didn’t use their neighborhood pool—and the explanation for this fact—was never once mentioned by any newspaper or public speaker.3 We are talking about public parks, which by law are open to anyone. Yet subtle and not-so-subtle government exclusionary policies achieved the complete racial segregation of their users. Robert Moses understood that exclusion is about so much more than trespass law. But as Caro also points out, the sorts of exclusionary strategies he employed, and the effects those strategies created, were simply not a part of the public discourse in the first part of the twentieth century. Still, it is surprising that by the first part of the twenty-first century, scholars have still not developed a comprehensive account of how non-trespassbased exclusion works, and what, if anything, the state can do to regulate it. This book has tried to contribute a systemic discussion of the topic that the newspapers all ignored in Moses’s era, drawing on important developments in the law and social sciences in subsequent decades.

CONCLUSION

Trespass law, and the fences, bouncers, and police officers who enforce it, turns out to be an important, but nonessential, ingredient of exclusion. Those seeking to exclude have a number of other arrows in their quiver. Proxy-based exclusion, by which an owner excludes on the basis of visible or lawful criterion A as a means of excluding those with correlated characteristic B, is one such tool. Exclusionary vibes, like the hiring of only white lifeguards or the naming of a subdivision, is another. Exclusionary amenities, which embed a costly and polarizing amenity, like a golf course or a religious institution, in a community, is another still. Society’s blind spot toward non-trespass-based exclusion has consequences. The unwillingness of anyone to discuss Robert Moses’s segregation strategies in the 1930s and 1940s meant that no one ever did anything about the results of those policies. It is, after all, not the use or non-use of trespass law that concerns us, but the existence of racially segregated pools and beaches. If the law wants to attack segregation head-on, it can only do so by recognizing the multiplicity of exclusionary strategies and designing approaches to combat those strategies in each of their forms when they are used to perpetuate odious forms of discrimination. After identifying a plurality of exclusion strategies, it is natural to ask what factors cause a landowner to choose one over another. It turns out that the presence or absence of private information drives many landowner decisions about what exclusion strategy to adopt. When prospective entrants possess private information about their own preferences, behaviors, and intentions, and the landowner cannot discover this private information at a low cost, we can expect to see the landowner employ non-trespass-based exclusion strategies. When there is little private information involved, or private information can be discovered by the landowner at a low cost, we can expect to see the landowner employing trespass-based exclusion rights. When intolerable segregation or its prospect prompts the state to regulate a landowner’s choice among exclusion strategies, there are two strategies the government might pursue. First, the government can engage in “hard shove” strategies, whereby the government prohibits some forms of exclusion and permits other forms of exclusion.4 This is the strategy the government has adopted in the housing discrimination arena. Alternatively, the government can opt for gentle nudges—strategies that alter the incentives of landowners by making various types of exclusion more or less attractive.

195

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

196

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

CONCLUSION

Megan’s Law is the most prominent and far-reaching example of the gentle nudge approach, although many aspects of information-privacy law affect landowner incentives in much the same way. What unifies these strategies is the government’s ability to regulate exclusion indirectly, through its control over information access policy. By rendering private information public or public information private, the state can alter, sometimes radically, the mix of exclusion strategies that landowners employ. As the last part of this book suggests, we are in the midst of a reputation revolution that is changing the fundamentals of exclusion. If a person who controls access to a resource formerly only had data about attributes A and B, he now can quickly analyze A through Z. This change has both happy and sad consequences—to the extent that decision makers were using race as a proxy for other characteristics, we will usually prefer that these decision makers admit or exclude entrants based on the relevant characteristics themselves instead of race. In other contexts, though, more information may indeed result in more exclusion, which can be problematic if the resource people are seeking to access is essential to equality of opportunity or other meritocratic aims. We cannot look inside Robert Moses’s soul to see what he would have done with all the modern tools of exclusion. Maybe he would have used technology to identify African Americans who he could “trust” at “his” precious beach. Or maybe he would have sliced and diced the data to exclude all African Americans in even more insidious ways. An important lesson from this book’s analysis is that information policy is an underutilized and undertheorized tool for the state to influence the behavior of private parties. There will be settings in which the government can reduce the prevalence of unlawful discrimination by publicizing previously private information about individuals. For example, the best available empirical evidence suggests that publicizing criminal history information could reduce racial discrimination in the employment of blue-collar and service workers. Pushing the point further, we can wonder whether a similar strategy might reduce the prevalence of statistical discrimination in the prescription of narcotics. Another vital lesson is that optimal public policy design must take account of the availability of a private market for reputation information. In regulating the landlord-tenant market, pro-tenant reformers failed to

CONCLUSION

anticipate the reputational repercussions of insisting on summary proceedings as the sole avenue for evicting tenants. In the immigration context, a comparative analysis of reputational infrastructure in the developed and developing world demonstrates the possible appeal of a bifurcated immigration policy—one that focuses on pre-entry screening for residents of developed nations and post-entry surveillance for residents of nations with poor reputation infrastructures. The tools available in a twenty-first-century economy will allow a modern-day Robert Moses to exclude with precision that the historical Robert Moses could have never imagined. But in some ways it is the imprecise exclusion that is most troubling of all. In other respects, the insights and technologies that we now have access to can help us root out and address subtle forms of improper discrimination that would have gone unnoticed or ignored in earlier eras. The aspiration of this book has been to shine the light on exclusionary practices that have long been shadowed. Exclusion is sometimes quite valuable, it is oftentimes inevitable, and it has always been misunderstood.

197

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

This page intentionally left blank

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES

Introduction 1. Gene Norman Levine & Leila A. Sussmann, Social Class and Sociability in Fraternity Pledging, 65 Am. J. Soc. 391, 395–96 (1960). 2. Barbara Allen, Saving Face, Tulsa World, January 14, 2007, at D4.

1.  Three Mechanisms for Exclusion 1. Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 673 (1999). This language arguably elevated the status of the right to exclude beyond the Court’s earlier characterization of it as “one of the most essential sticks in the bundle of rights that are commonly characterized as property.” Kaiser Aetna v. United States, 444 U.S. 164, 176 (1979). 2. J. E. Penner, The Idea of Property in Law 71 (1997); Felix S. Cohen, Dialogue on Private Property, 9 Rutgers L. Rev. 357, 374 (1954); Thomas W. Merrill, Property and the Right to Exclude, 77 Neb. L. Rev. 730, 740–52 (1998). 3. Allison Fass, Sex, Pranks, and Reality: Second Life’s Virtual Web World Can Be a Weird, Chancy Place for Real-Life Brands, Forbes, July 2, 2007. 4. Alan Finder, In Desire to Grow, Colleges in South Battle with Roots, N.Y. Times, Nov. 30, 2005, at A1. 5. Michael Barbaro & Hillary Chura, Appealing to All Has Resulted in Sales to Too Few, N.Y. Times, Jan. 1, 2007, at C1. 6. Id. 7. Id. 8. Andrea Chang, Abercrombie Tries to Keep Its Shirt On as Sales Slacken, L.A. Times, Aug. 15, 2009, at 1. 9. Bluvias v. Winfield Mut. Housing Corp., 540 A.2d 1324, 1328 (N.J. Super. 1988).

199

200

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 17–22

10. Michael Schill et al., The Condominium v. Cooperative Puzzle: An Empirical Analysis of Housing in New York City, 36 J. Legal Stud. 275, 284 (2007). 11. Id. at 314–16. Schill et al. identify substantial transaction costs and adverse tax consequences associated with transitioning a cooperative building into a condominium. During the last three decades, the percentage of common-interest communities that have used the cooperative form has plummeted. Evan McKenzie, Common-Interest Housing in the Communities of Tomorrow, 14 Housing Pol’y Debate 203, 207 tbl.2 (2003). 12. Schill et al., supra note 10. 13. Id. at 312. 14. Id. at 309 tbl. 6. 15. Id. at 284 n.9. 16. Rosemarie Maldonado & Robert D. Rose, The Application of Civil Rights Laws to Housing Cooperatives: Are Co-ops Bastions of Discriminatory Exclusion or Self-Selecting Models of Community-Based Living?, 23 Fordham Urb. L.J. 1245, 1245–46 (1996); Sabrina Malpeli, Comment, Cracking Down on Cooperative Board Decisions That Reject Applicants Based on Race: Broome v. Biondi, 73 St. John’s L. Rev. 313 (1999). 17. Maldonado & Rose, supra note 16, at 1245–46. 18. Richard Thompson Ford, The Boundaries of Race: Political Geography in Legal Analysis, 107 Harv. L. Rev. 1841, 1894–1906 (1994); J. Peter Byrne, Are Suburbs Unconstitutional?, 85 Geo. L.J. 2265, 2265–72 (1997) (book review). 19. Casey J. Dawkins, Recent Evidence on the Continuing Causes of Black-White Residential Segregation, 26 J. Urb. Aff. 379 (2004) and sources cited therein. 20. See, e.g., Patrick Bayer et al., An Equilibrium Model of Sorting in an Urban Housing Market: The Causes and Consequences of Residential Segregation 65–66 (Yale Univ. Econ. Growth Ctr., Center Discussion Paper No. 860, 2003). This trend is evidently more pronounced for Caucasians than for African Americans. Richard D. Alba et al., How Segregated Are Middle-Class African Americans?, 47 Soc. Probs. 543, 556 (2000); Dawkins, supra note 19, at 382–83. 21. Lee Anne Fennell, The Unbounded Home: Property Values Beyond Property Lines 163 (2009). 22. The arguments in favor of open admissions to the law review were largely deontological in nature. Some advocates of open admissions argued publicly that all Yale law students would be equally capable law review members, but it seems implausible that anyone actually believed that. 23. 347 U.S. 483, 494 (1954). 24. Jean M. Twenge & Roy F. Baumeister, Social Exclusion Increases Aggression and Self-Defeating Behavior While Reducing Intelligent Thought and Prosocial Behavior, in The Social Psychology of Inclusion and Exclusion 28, 30 (Dominic Abrams et al. eds., 2005). 25. Id. at 32–38. 26. Tyler F. Stillman et al., Alone and Without Purpose: Life Loses Meaning Following Social Exclusion, 45 J. Exper. Soc. Psych. 686, 688–89 (2009). 27. Id. at 692. 28. Rockwell Graphic Systems v. DEV Industries, 925 F.2d 174 (7th Cir. 1991).

NOTES TO PAGES 23–28

29. Robert C. Ellickson, Property in Land, 102 Yale L.J. 1315, 1327–28 (1993). 30. Fennell, supra note 21, at 12–13. 31. Thomas W. Sanchez et al., Security Versus Status?: A First Look at the Census’s Gated Community Data, 24 J. Planning Educ. & Research 281, 283–84 (2005). 32. Id. at 290; Cynthia L. Pickett & Marilynn B. Brewer, The Role of Exclusion in Maintaining Ingroup Inclusion, in The Social Psychology of Inclusion and Exclusion 89, 96–97 (Dominic Abrams et al. eds., 2005). 33. Dominic Abrams et al., A Social Psychological Framework for Understanding Social Inclusion and Exclusion, in The Social Psychology of Inclusion and Exclusion 1, 20 (Dominic Abrams et al. eds., 2005). 34. See, e.g., Henry E. Smith, Exclusion Versus Governance: Two Strategies for Delineating Property Rights, 31 J. Legal Stud. 453 (2002) [hereinafter Exclusion Versus Governance]; Henry E. Smith, Intellectual Property as Property: Delineating Entitlements in Information, 116 Yale L.J. 1742 (2007). 35. Henry E. Smith, Exclusion and Property Rules in the Law of Nuisance, 90 Va. L. Rev. 965, 981 (2004). 36. Exclusion Versus Governance, supra note 34, at 455. 37. More complex governmental efforts to facilitate exclusion by private parties, such as exclusionary-zoning regimes, are appropriately understood as governance mechanisms, notwithstanding the presence of the word “exclusionary” in the term. This is appropriate, as a government-driven process like exclusionary zoning is analytically different in kind from the decentralized, individual-rights-oriented exclusion remedies I discuss in detail below. For discussions of exclusionary and inclusionary zoning, see Peter H. Schuck, Diversity in America: Keeping Government at a Safe Distance 203–27 (2003); David J. Barron, Reclaiming Home Rule, 116 Harv. L. Rev. 2255, 2357–61 (2003); James C. Clingermayer, Heresthetics and Happenstance: Intentional and Unintentional Exclusionary Impacts of the Zoning Decision-making Process, 41 Urb. Stud. 377 (2004); and Robert C. Ellickson, The Irony of “Inclusionary” Zoning, 54 S. Cal. L. Rev. 1167 (1981). 38. Henry E. Smith, Property and Property Rules, 79 N.Y.U. L. Rev. 1719, 1756 (2004). 39. A. Smith’s analysis of the information costs associated with governance proceeds from a welfarist perspective. In other words, he would not be satisfied if governance merely shifted the information costs associated with resource management from the state to an individual owner. Rather, he argues, exclusion shifts the decision over a resource’s use to the individual who usually can obtain information about that resource most efficiently—its owner. Smith, supra note 35, at 985.

2.  The Bouncer’s Right 1. Eduardo M. Peñalver, Property as Entrance, 91 Va. L. Rev. 1889, 1898–1900 (2005). 2. [1882] 21 Ch.D. 667. 3. Id. at 669. I have argued elsewhere for greater symmetry in the law’s treatment of purportedly antisocial impulses by living owners and dead ones. Lior Jacob Strahilevitz, The Right to Destroy, 114 Yale L.J. 781, 839–52 (2005).

201

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

202

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 28–34

4. Stephen R. Munzer, Property as Social Relations, in New Essays in the Legal and Political Theory of Property 36, 40 (Stephen R. Munzer ed., 2001); Eduardo M. Peñalver, Is Land Special? The Unjustified Preference for Landownership in Regulatory Takings Law, 31 Ecology L.Q. 227, 257–60 (2004). 5. J. E. Penner, The Idea of Property in Law 74–75 (1997). 6. Id. at 75 (“The important feature of property is the individual’s determination of the disposition of a thing, not any requirement that he use it on his own.”). 7. 2 William Blackstone, Commentaries on the Laws of England *2 (1765–1769). 8. Carol M. Rose, Canons of Property Talk, or, Blackstone’s Anxiety, 108 Yale L. J. 601, 631 (1998). 9. 42 U.S.C. § 3604(b). 10. 42 U.S.C. § 3604(a). 11. 42 U.S.C. § 3604(d). 12. 42 U.S.C. § 3604(e). 13. 42 U.S.C. § 3604(c). 14. Ragin v. New York Times, 923 F.2d 995 (2d Cir. 1991). 15. Fair Housing Council of San Fernando Valley v. Roommates.com, 521 F.3d 1157 (9th Cir. 2008) (en banc). 16. Richard A. Epstein, Takings, Private Property and the Power of Eminent Domain 66 (1985). 17. Richard A. Epstein, Takings, Exclusivity, and Speech: The Legacy of Pruneyard v. Robins, 64 U. Chi. L. Rev. 21, 22 (1997). 18. Richard A. Epstein, The Ubiquity of the Benefit Principle, 67 S. Cal. L. Rev. 1369, 1383–84 (1994). 19. National Board of Y.M.C.A. v. United States, 395 U.S. 85, 93 (1969). 20. Lior J. Strahilevitz, When the Taking Itself Is Just Compensation, 107 Yale L.J. 1975, 1979–80 (1998). 21. 277 A.2d 369 (N.J. 1971). 22. State v. DeCoster, 653 A.2d 891, 894 & n.3 (Me. 1995). 23. Flogueras v. Hassle, 331 F. Supp. 615, 623–24 (W.D. Mich. 1971). 24. 277 A.2d at 374–75. 25. Id. 26. Food Lion Inc. v. Capital Cities / ABC, 194 F.3d 505 (4th Cir. 1999). 27. 42 U.S.C. § 14071 (2000); N.J. Stat. Ann. § 2C:7 (West 2004). Congress acted after many states had enacted their own versions of Megan’s Law, and this congressional action “effectively mandate[d]” that the remaining states follow suit. Daniel M. Filler, Silence and the Racial Dimension of Megan’s Law, 89 Iowa L. Rev. 1535, 1546 (2004). 28. See, e.g., Jessica Lunsford Act, H.B. 1877, 107th Reg. Sess. (Fla. 2005); States Track Sex Offenders by GPS, Wired, July 30, 2005, available at http://www.wired .com/news/technology/0,1282,68372,00.html. 29. Mulligan v. Panther Valley Prop. Owners Ass’n, 766 A.2d 1186, 1192 (N.J. Super. Ct. App. Div. 2001); Henry Gottlieb, Fighting a Local Ban on Sex Criminals, Nat. L.J.,

NOTES TO PAGES 34–37

May 17, 1999, at A7; Robert Hanley, Condominiums Consider Barring Some Paroled Sex Offenders, N.Y. Times, May 4, 1998, at B1. 30. Mulligan, 766 A.2d at 1192. 31. Id. at 1189. 32. Id. at 1192. 33. Id. at 1193–94. 34. For a quick overview of the conflicting studies of recidivism rates, see Peter T. Wendel, The Case Against Plea Bargaining Child Sexual Abuse Charges: “Déjà Vu All Over Again,” 64 Mo. L. Rev. 317, 331 n.46 (1999). For a more comprehensive analysis of the literature, see John M. Fabian, Kansas v. Hendricks, Crane, and Beyond: “Mental Abnormality,” and “Sexual Dangerousness”: Volitional vs. Emotional Abnormality and the Debate Between Community Safety and Civil Liberties, 29 Wm. Mitchell L. Rev. 1367, 1426–33 (2003). 35. Alternatively, these restrictions may be insufficiently exclusionary in that sex offenders are excluded from residing in the community, but not excluded from entering the community. The calculus here involves a straightforward cost-benefit analysis whereby a community would weigh the cost of doing background checks whenever someone enters a community as a visitor against the anticipated decrease in sex offenses associated with such a policy. 36. In New Jersey, as in other states, any property owner whose land is benefited by a covenant has standing to sue to enforce it. Syrian Antiochian Orthodox Archdiocese of N.Y. v. Palisades Assocs., 264 A.2d 257 (N.J. Super. Ct. Ch. Div. 1970). 37. Edgar Kiser, Comparing Varieties of Agency Theory in Economics, Political Science, and Sociology: An Illustration from State Policy Implementation, 17 Soc. Theory 146, 149–50 (1999). The Ottoman Turks’ selection of eunuchs to guard harems was an extreme (and extremely effective) use of selection effects to solve a principal-agent problem. Richard Sherr, Gugliemo Gonzaga and the Castrati, 33 Renaissance Q. 33, 47 (1980). 38. The rush process at fraternities and sororities has been studied from both sociological and economic perspectives. For interesting accounts, see Monica Biernat et al., Selective Self-Stereotyping, 71 J. Personality & Soc. Psychol. 1194 (1996); Gene Norman Levine & Leila A. Sussmann, Social Class and Sociability in Fraternity Pledging, 65 Am. J. Soc. 391 (1960); Susan Mongell & Alvin E. Roth, Sorority Rush as a Two-Sided Matching Mechanism, 81 Am. Econ. Rev. 441 (1991); Guillermo De Los Reyes & Paul Rich, Housing Students: Fraternities and Residential Colleges, 585 Annals Am. Acad. Pol. & Soc. Sci. 118 (2003). 39. Levine & Sussman, supra note 38, at 395–96. 40. See generally Mongell & Roth, supra note 38, at 460 (finding strategic behavior by pledges seeking admission to sororities). 41. Another potential advantage of vague criteria and bouncer’s discretion is that it makes it more difficult to prevail in a lawsuit against the bouncer for improper exclusion. Some scholars have advocated legal rules requiring the articulation of justifications for why particular people were excluded from property. See, e.g., Cynthia L. Estlund, Labor, Property, and Sovereignty After Lechmere, 46 Stan. L. Rev. 305, 308–09 (1994). The problem with

203

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

204

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 38–42

this approach is that articulation of a justification invites costly second-guessing and legal intervention. Giving the owner the right to exclude unreasonably hardly guarantees that he will behave unreasonably. He has his reputation at stake, after all. 42. Id. 43. See, e.g., Leticia M. Saucedo, The Employer Preference for the Subservient Worker and the Making of the Brown Collar Workplace, 67 Ohio St. L.J. 961, 970–71 (2006) (arguing that some employers prefer to hire undocumented immigrants from Latin America because they are more likely to be subservient). 44. Adam B. Cox & Eric A. Posner, The Second-Order Structure of Immigration Law, 59 Stan. L. Rev. 809, 824–27 (2007). 45. Id. at 835–40. 46. Joel R. Reidenberg, E-Commerce and Trans-Atlantic Privacy, 38 Hous. L. Rev. 717 (2001). 47. U.S. Census Bureau, The Foreign-Born Population: 2000, at 5 tbl.2 (2003), available at http://www.census.gov/prod/2003pubs/c2kbr-34.pdf. 48. Samuel G. Freedman, Grad School’s International Glow Is Dimmed by Security Concerns, N.Y. Times, Oct. 27, 2004, at B10. 49. Josh Meyer, Al Qaeda Linked to Plot to Use Student Visas, L.A. Times, Jan. 23, 2007, at A13. 50. See, e.g., Brandon Ortiz, Foreign Aid for Texas: Visa Waiver Program Is Just What the Doctor Ordered to Meet Medical Need, San Antonio Express-News, June 27, 2004, at 1A. 51. The 2007 attacks on the Glasgow airport, apparently perpetrated by several foreign-born physicians working at United Kingdom hospitals, show the pitfalls of using occupation as a proxy for the security threat posed. Janet Stobart & Sebastian Rotella, British Manhunt Widens: Three Physicians Are Among Eight Suspects Now Held in Bomb Plots, L.A. Times, July 3, 2007, at A1. This episode may have been an instance of the terrorists recognizing the profile and finding someone who didn’t fit the profile to carry out the attacks. This danger was anticipated in Bernard E. Harcourt, Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age 227–36 (2007). 52. 8 U.S.C. § 1153(b)(5)(A) (2006). 53. Beth MacDonald, Note, The Immigrant Investor Program: Proposed Solutions to Particular Problems, 31 Law & Pol’y Int’l Bus. 403, 406–10 (2000). 54. But see Cox & Posner, supra note 44, at 855 (suggesting that auctioning off visas might be a poor mechanism for facilitating immigration by upwardly mobile migrants). 55. Alex Kotlowitz, Our Town, N.Y. Times Magazine 30, Aug. 5, 2007. 56. John Seabrook, The Back Room, New Yorker, Aug. 30, 2010.

3.  Exclusionary Vibes 1. For more on homogeneity within the Greek system, see Ernest T. Pascarella et al., Influences on Students’ Openness to Diversity and Challenge in the First Year of College, 67 J. Higher Educ. 174, 190 (1996).

NOTES TO PAGES 42–46

2. That is not to say that the Greek system in general and the cooperative system in general attracted the same kinds of students. The student populations were quite different, and campus stereotypes evidently were at least somewhat grounded in reality: cooperative residents as a group appeared to be more liberal and less well-off, more likely to dress like slackers, more politically correct, more pierced, and less interested in the university’s athletic teams. 3. Nor is it the case that the co-ops were homogeneous because co-op members were the “leftovers” of the Greeks’ selection process. Approximately 2,500 students presently live in Berkeley’s fraternities and sororities, and the campus cooperatives have approximately 1,300 members. Patrick Hoge & Steve Rubenstein, Cal Bans Alcohol at Campus Fraternities, S.F. Chron., May 10, 2005, at B1; Univ. Students’ Coop. Ass’n, History of the USCA, http://www.usca.org/home/history.php. The vast majority of Berkeley’s 20,000 undergraduates sought housing in neither the Greek system nor the cooperatives. 4. There is an extensive literature on the use of “cheap talk” (i.e., low-cost communication) to establish focal points in coordination games. See, e.g., Joseph Farrell, Cheap Talk, Coordination, and Entry, 18 RAND J. Econ. 34, 39 (1987); Joseph Farrell & Matthew Rabin, Cheap Talk, 10 J. Econ. Persp. 103, 116–17 (1996). Elsewhere, Richard McAdams has pointed to the value of communication to establish focal points in not only pure coordination games, like the meeting place problem, but also in games that involve a mix of coordination and conflict strategies. Richard H. McAdams, A Focal Point Theory of Expressive Law, 86 Va. L. Rev. 1649, 1677–89, 1714–22 (2000). For an invocation of the focal points idea in the property case law, see United States v. Hunter, 459 F.2d 205, 213–14 (4th Cir. 1972). For a discussion of sorting in the context of homeowners associations, see Clayton P. Gillette, Courts, Covenants, and Communities, 61 U. Chi. L. Rev. 1375, 1394–1400 (1994). 5. Thomas C. Schelling, The Strategy of Conflict 57–58 (1980). 6. Spann v. Colonial Vill., Inc., 899 F.2d 24, 30 (D.C. Cir. 1990). 7. See, e.g., Kevin Weaks, New Displays, Old Charm at Whittaker’s New Town, St. Louis Post-Dispatch, Aug. 15, 2005, available at http://www.stltoday.com/stltoday/ realestate/stories.nsf/buildingahome/story/045852EAA8E1C16D862570570051A3 3C?OpenDocument. 8. Davenport v. Cotton Hope Plantation Horizontal Prop. Regime, 508 S.E.2d 565 (S.C. 1998). 9. Aquarian Found., Inc. v. Sholom House, Inc., 448 So. 2d 1166 (Fla. Dist. Ct. App. 1984). Property names act as effective sorting devices in other contexts as well. See, e.g., Stephan Weiler, A Park by Any Other Name: National Park Designation as a Natural Experiment in Signaling 11 (Fed. Reserve Bank of Kan. City, Research Working Paper 05–09, 2005) (finding that when national monuments are redesignated as national parks, they experience significant increases in attendance, drawing largely on “a constant marginal set of visitors interested in National Park designations in themselves”). 10. The cooperatives have several “theme houses,” such as the vegetarian theme house, the African American theme house, and the Gay-Lesbian-Bisexual theme house. Univ.

205

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

206

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 46–49

Students’ Coop. Ass’n, The Co-ops at USCA, http://www.usca.org/understand/ thecoops/. Any student may live in these buildings, but the theme certainly affects the applicant pool. Other houses without themes nevertheless intentionally present differentiated personalities on the Internet. Compare Univ. Students’ Coop. Ass’n, Ridge House at USCA, http://www.usca.org/understand/thecoops/rid.php (“On the spectrum of studious to social, Ridge House would be on the more studious end. [For example, the study room gets the best afternoon light and you can probably find someone studying in there at any given hour.] We don’t host any huge co-op notable parties . . . and you know what?—at Ridge, we’re all pretty fine with that. . . . Pretty much we’re all kind of normal-looking, with a pretty even spread of majors, and artsy/ sportsy/nerdy interests.”), with Univ. Students’ Coop. Ass’n, Andres Castro Arms at USCA, http://www.usca.org/understand/thecoops/aca.php (“We will forever host the BEST co-op party ever! . . . The Infamous Disco Party. We will always road trip the hardest; from Tahoe to Joshua Tree—snowboarding to rock climbing—we do it all. . . . Overstuffed sofas so soft that you can melt into them. A game room stocked with Double Dragon, foosball, and a pool table. All the Oreos and Nutella you can eat. And of course . . . a hot tub for those cold winter nights. Yes . . . Castro has more luxury features than a Coupe DeVille. If it ain’t high livin’, we ain’t havin’ it.”). Residents of each house also gave tours to prospective members on request, and these tours tended to emphasize many of the same themes mentioned in the Internet house profiles. At studious Ridge House, students who seemed on the surface to be good fits for the house typically received more enthusiastic tours than prospective students who seemed like poor matches. 11. Alan Finder, In Desire to Grow, Colleges in South Battle with Roots, N.Y. Times, Nov. 30, 2005, at A1. 12. Alan Finder, To Woo Students, Colleges Choose Names that Sell, N.Y. Times, Aug. 11, 2005, at A1. 13. Id. 14. Id. 15. Joe Saltzman, A Picture Is Worth a Thousand Lies, USA Today Magazine, Nov. 11, 2007, at 55. 16. See, e.g., Lucia Moses, Grabbing the Rebound, Editor & Publisher, Jan. 1, 2004, at 1; Plaza Developers Take to the Airways, Real Estate Wkly., Apr. 20, 2005, at 39; What You Can Do If Your Home Isn’t Selling, Kansas City Star, Sept. 19, 2004, at K12. 17. See, e.g., Reginald Leamon Robinson, The Racial Limits of the Fair Housing Act: The Intersection of Dominant White Images, the Violence of Neighborhood Purity, and the Master Narrative of Black Inferiority, 37 Wm. & Mary L. Rev. 69 (1995); Debra Alligood, Comment, When the Medium Becomes the Message: A Proposal for Principal Media Liability for the Publication of Racially Exclusionary Real Estate Advertisements, 40 UCLA L. Rev. 199 (1992). 18. Tyus v. Urb. Search Mgmt., 102 F.3d 256, 262–64 (7th Cir. 1996); Ross D. Petty et al., Regulating Target Marketing and Other Race-Based Advertising Practices, 8 Mich. J. Race & L. 335, 349–52, 373–77 (2003).

NOTES TO PAGES 49–52

19. On advertising of this nature, see Reginald Leamon Robinson, White Cultural Matrix and the Language of Nonverbal Advertising in Housing Segregation: Toward an Aggregate Theory of Liability, 25 Cap. U. L. Rev. 101, 204 (1996). 20. James W. Loewen, Sundown Towns: A Hidden Dimension of American Racism 203–04, 384 (2005). 21. Id. at 384. 22. Id. at 388. 23. Camille Zubrinsky Charles, Processes of Racial Residential Segregation, in Urban Inequality: Evidence from Four Cities 217, 259 tbl.4.6 (Alice O’Connor et al. eds., 2001) (noting that 11 percent of whites responded in a survey that they wanted to live in neighborhoods that were 100 percent white, and that 2 1/2 percent of black respondents said they wanted to live in all-black neighborhoods); id. at 230–31 (finding that African Americans are unlikely to move into neighborhoods that are believed to contain a large percentage of residents who do not want African American neighbors); see also Michael O. Emerson et al., Does Race Matter in Residential Segregation? Exploring the Preferences of White Americans, 66 Am. Soc. Rev. 922, 927–32 (2001) (finding that the presence of Asian Americans and Latinos had little effect on whites’ willingness to move into a neighborhood once crime, public school quality, and anticipated appreciation of real estate were controlled, but that the presence of African Americans had a very substantial effect on whites’ willingness to move into the neighborhood, even after controlling for these variables). 24. See, e.g., Arnold R. Hirsch, Making the Second Ghetto: Race and Housing in Chicago, 1940–1960, at 31–36 (1998). 25. Costly governance mechanisms, like architectural review boards, will be unnecessary for preserving uniformity, except in those instances where real estate is pricey enough or the existing housing stock is sufficiently deteriorated to encourage “knockdowns” (i.e., the demolition of existing residences, to be replaced by modern, often larger homes). 26. Larry Rohter, Drawing Lines Across the Sand, N.Y. Times, Feb. 6, 2007, at A1. 27. Id. 28. Id. 29. See, e.g., Marks v. Whitney, 491 P.2d 374 (Cal. 1971); Jessica A. Duncan, Coastal Justice: The Case for Public Access, 11 Hastings W.-Nw. J. Envtl. L. & Pol’y 55, 59 (2004). In some states, such as New Jersey, these public trust rights also create ancillary rights to use the dry sand as well. Matthews v. Bay Head Improvement Ass’n, 471 A.2d 355 (N.J. 1984). 30. Duncan, supra note 29, at 64–65. 31. Rick Brooks has suggested that even after the highly publicized Shelley v. Kraemer decision rendered racially restrictive covenants unenforceable, real estate attorneys continued to make use of those covenants on the grounds that some people would still assume that they were legally enforceable. Richard R. W. Brooks, Covenants and Conventions 37 & n.103 (Nw. Univ. L. & Econ. Research Paper No. 02–8, Sept. 2002). The prevalence of sundown signs in racist communities throughout the United States

207

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

208

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 53–57

is another example of bluffing. Such signs were placed at the town border and typically used the language: “Nigger, Don’t Let The Sun Go Down On You In This County.” Loewen, supra note 20, at 195. James Loewen’s historical research reveals 184 towns that displayed such signs, only 7 of which were south of the Mason-Dixon line. Id. Such edicts obviously were unenforceable under the Equal Protection Clause, but the presence of the signs undoubtedly emboldened private citizens to use violence and intimidation to keep blacks out of town after sundown. Id. at 182, 270–79. 32. Clive Thompson, An Elf’s Progress: Finally, Online Role-Playing Games That Won’t Destroy Your Life, Slate, March 7, 2005, available at http://www.slate.com/ toolbar.aspx?action=print&id=2114354. 33. Id.

4.  Exclusionary Amenities 1. James M. Buchanan, An Economic Theory of Clubs, 32 Economica 1, 2 (1965). 2. A large body of real estate law mandates that sellers disclose various attributes of their property to potential purchasers. As the analysis above suggests, various forms of mandatory disclosure may have the unintended consequence of promoting residential homogeneity. 3. The implicit assumption here is that preferences for certain types of common amenities are more stable over time than linguistic signals, which are the cheapest tools in a focal point strategy but which might see their meanings change radically, thanks to linguistic reclamation, government actions, or other behavioral shifts. 4. People v. Leroy, 828 N.E.2d 769, 776–84 (Ill. App. Ct. 2005); Mulligan v. Panther Valley Prop. Owners Ass’n, 766 A.2d 1186, 1192–94 (N.J. Super. Ct. App. Div. 2001). 5. It is not difficult to imagine a club good that might provide a good proxy for sex offender status. Community members might make extremely heavy investments in school child-abuse-awareness programs or domestic violence police as a way of discouraging dangerous sex offenders from settling in a particular community. 6. Betsy Blaney, Safe at Home: Lubbock Company Creating Sex Offender–Free Subdivision, Ft. Worth Star-Telegram, June 7, 2005, at B5; Stephanie Simon, Ex-Cons Exiled to Outskirts, L.A. Times, Dec. 5, 2002, at A1. 7. Debra Rosenberg, A Place of Their Own, Newsweek, Jan. 15, 2001, at 54. 8. Ross D. Petty et al., Regulating Target Marketing and Other Race-Based Advertising Practices, 8 Mich. J. Race & L. 335, 373–77 (2003); Robert G. Schwemm, Discriminatory Housing Statements and § 3604 (c): A New Look at the Fair Housing Act’s Most Intriguing Provision, 29 Fordham Urb. L.J. 187, 191 (2001). 9. Ragin v. N.Y. Times, 923 F.2d 995 (2d Cir. 1991). 10. 42 U.S.C. § 3603(b)(2); see also Petty et al., supra note 8, at 376. The Seventh Circuit has held that a nineteenth-century federal statute, 42 U.S.C. § 1982, bars racial discrimination by Mrs. Murphy landlords. Morris v. Cizek, 503 F.2d 1303, 1304 (7th Cir. 1974).

NOTES TO PAGES 57–58

11. There may be a few senses in which exclusionary amenities strategies are more efficient than overt discrimination in admission or advertising. First, adopting the exclusionary amenities strategy may be less “in your face,” or confrontational, than excluding members of undesired groups, and excluders may value this opportunity. James C. Clingermayer, Heresthetics and Happenstance: Intentional and Unintentional Exclusionary Impacts of the Zoning Decision-making Process, 41 Urb. Stud. 382–83 (2004) (noting that exclusionary zoning proponents rarely discuss racial segregation in public, even where segregationist sentiments are motivating them, because such language “is generally not considered socially acceptable or politically correct” and may invite a lawsuit); Allan C. De Serpa, A Theory of Discriminatory Clubs, 24 Scot. J. Pol. Econ. 33, 34, 39 (1977) (“[P]eople are apt to be reluctant to admit, face to face, that the characteristics of others are repulsive to them. As a consequence, the exclusion of individuals exhibiting certain characteristics evolves as a second best solution.”). Second, club members may want to attract members of disfavored groups who actually loathe other members of their disfavored groups, because the presence of such “self-hating” group members solidifies negative stereotypes about the excluded group or provides cover against discrimination suits. To maximize this preference, overt discrimination will be ineffective, but exclusionary amenities may be highly effective. 12. Formally, the federal FHA and Fair Housing Act Amendments (FHAA) recognize disparate impact claims. Lapid-Laurel v. Zoning Bd. of Adjustment of the Twp. of Scotch Plains, 284 F.3d 442, 466–67 (3d Cir. 2002); Gamble v. City of Escondido, 104 F.3d 300, 304–07 (9th Cir. 1997). That said, FHA and FHAA claims are almost always brought against local governments, as opposed to individual developers, perhaps because it is so easy for a developer to rebut a prima facie case of disparate impact by pointing to a “legitimate, nondiscriminatory reason for its action,” such as consumer demand, unconnected to exclusionary motives, for the club good in question. Gamble, 104 F.3d at 305; see also Lapid-Laurel, 284 F.3d at 467. 13. Marilyn Gardner, An Empty Nest—Now What? Once the Kids Move Out, Couples Start to Ask Themselves What They Want in Life and How Much Space They Need, Christian Sci. Monitor, Apr. 21, 2004, at 15; see also Ronald T. Mitchelson & Michael T. Lazaro, The Face of the Game: African Americans’ Spatial Accessibility to Golf, 44 Se. Geographer 48, 70 (2004) (“The golf course can be a wonderful landscape of intense social and environmental interaction.”). 14. The social nature of many club goods also allows prospective purchasers to obtain information about neighborhood composition at a low cost. By contrast, in a neighborhood with neither common spaces nor front porches, it may be difficult for a prospective purchaser to discover the characteristics of the neighborhood’s residents. 15. Lee Sigelman et al., Making Contact? Black-White Social Interaction in an Urban Setting, 101 Am. J. Soc. 1306, 1324–26 (1996). 16. Charles T. Clotfelter, Spatial Rearrangement and the Tiebout Hypothesis: The Case of School Desegregation, 42 S. Econ. J. 263, 268 (1975) (noting that whites’ opposition to residential integration increases when they believe that residential integration will result in the desegregation of local public schools); Yannis M. Ioannides & Linda

209

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

210

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 58–59

Datcher Loury, Job Information Networks, Neighborhood Effects, and Inequality, 42 J. Econ. Literature 1056, 1071–82 (2004); Wilfred G. Marston & Thomas L. Van Valey, The Role of Residential Segregation in the Assimilation Process, 441 Annals Am. Acad. Pol. & Soc. Sci. 13, 16–17 (1979). But cf. Timothy Bledsoe et al., Residential Context and Racial Solidarity Among African Americans, 39 Am. J. Pol. Sci. 434, 451–53 (1995) (finding that residence in integrated neighborhoods and increased social contact with whites may decrease social solidarity among African Americans). 17. Oscar H. Gandy, Jr., Coming to Terms with Chance: Engaging Rational Discrimination and Cumulative Disadvantage 93–102 (2009). 18. Melissa J. Marschall & Dietlind Stolie, Race and the City: Neighborhood Context and the Development of Generalized Trust, 26 Pol. Behav. 125, 139–44 (2004). 19. J. Eric Oliver & Janelle Wong, Intergroup Prejudice in Mutliethnic Settings, 47 Am. J. Pol. Sci. 567, 577–80 (2003). 20. Variance in this context means a development with both large and small kitchens. 21. This explains why plaintiffs asserting disparate impact claims under the FHA would face an uphill battle if they attacked a private developer’s use of exclusionary amenities. 22. This book focuses on developers’ uses of exclusionary amenities, as opposed to decisions by populated common-interest communities to add exclusionary amenities. Barzel and Sass provide an illuminating explanation for why one might expect to see developers making decisions about common amenities, instead of leaving this decision to residents. Yoram Barzel & Tim R. Sass, The Allocation of Resources by Voting, 105 Q. J. Econ. 745, 764–65 (1990). They argue that creating expensive common amenities in a preexisting community will generate substantial controversy, especially where residents will derive differential utility from these amenities. Complex voting procedures will be needed to resolve these disputes, particularly in common-interest communities that have homes of different sizes and values. Id. at 765–70. My account is consistent with Barzel and Sass’s, although it supplements it in important ways. Demand for certain common-interest communities may sort potential residents of a community in many ways, potentially contributing to homogeneities beyond a common desire for the amenity in question. Thus, developers may create common amenities at the outset, not only because creating such amenities would be more difficult down the road, but also because the absence of such an amenity at the outset will cause potential purchasers who would like that amenity to purchase elsewhere instead. Indeed, it may be that the presence of certain common amenities promotes homogeneity across a number of dimensions, and these forms of homogeneity lend themselves to less contentious governance within common-interest communities. 23. John J. Delaney, Addressing the Workforce Housing Crisis in Maryland and Throughout the Nation, 33 U. Balt. L. Rev. 153, 175 (2004). 24. Peter Whoriskey, No Kids? That’s No Problem: Falls Church’s Deal with Builder Highlights Area School Crowding, Wash. Post, May 25, 2003, at A1. 25. Id. 26. Id.

NOTES TO PAGES 59–61

27. Id. 28. Id. (“‘We haven’t had any inquiries from people with lots of kids. It’s kind of like how water seeks its own level. It just happens.’”) (quoting real estate agent Mary Alice Kaplan). In other contexts, housing consumers with a choice of suburbs seem to understand that the choice of common sporting activities entails a choice about the nature of one’s neighbors and social networks. A New York Times series on class in America quoted a homeowner’s description of his Atlanta suburb and the role tennis played in organizing social interactions: “The good thing about it is that it is a very comfortable neighborhood to live in. . . . These are very homogeneous types of groups. You play tennis with them, you have them over to dinner. You go to the same parties.   “. . . When you talk about tennis, guess what? Everybody you play against looks and acts and generally feels like you. It doesn’t give you much of a perspective.” Peter T. Kilborn, The Five-Bedroom, Six-Figure Rootless Life, N.Y. Times, June 1, 2005, at A1. 29. Press Release No. 04–142, U.S. Dep’t of Hous. & Urban Dev., HUD Settles Investigation of Falls Church and Condo Developers (Nov. 19, 2004), http://www.hud.gov/news/release.cfm?content=pr04–142.cfm. 30. Id. 31. R. Jeff Teasley et al., Recreation and Wilderness in the United States 20 (Univ. of Ga. Dep’t of Agric. & Applied Econ., Working Paper No. 97–13, 1997), available at http://www.agecon.uga.edu/~erag/finalreport.htm. 32. Id. 33. Id. Data from a 1997 study showed an even more substantial gap in participation. In that year, 2.7 percent of African Americans played golf, compared with 12.6 percent of Caucasians. Jill Lieber, Golf Finally Reaching Out, USA Today, Aug. 15, 2001, at 1C (quoting statistics from a 1997 study by the National Golf Foundation). For an explanation of the various possible causes of low minority participation in golf, see Paul H. Gobster, Explanations for Minority “Underparticipation” in Outdoor Recreation: A Look at Golf, 16 J. Park & Recreation Admin. 46, 48–49 (1998). 34. Nat’l Golf Found., Minority Golf Participation in the U.S. 6 (2003) (noting that the average golfer played 19.2 rounds during the previous year, whereas the average African American golfer played 13.9 rounds during the previous year). Some caution is in order in interpreting this data, however. African American golf participation increased during the first few years of the millennium, and it may be that an influx of new African American golfers explains the lower intensity of participation. See infra text accompanying note 28. 35. In 1995, 19.6 percent of Caucasians lived in households with annual incomes in excess of $75,000, whereas 8.1 percent of African Americans lived in such households. Bureau of the Census, Money Income in the United States: 1995, at 11–12, in Current Population Reports P60–193 (Sept. 1996), available at http://www.census.gov/

211

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

212

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 61–62

prod/2/pop/p60/p60–193.pdf. Thus, Caucasians were 2.4 times as likely as African Americans to have household incomes above $75,000 per year, but 4 times as likely to play golf. Income inequality between Caucasians and African Americans has been diminishing consistently over time. Bureau of the Census, Measuring 50 Years of Economic Change Using the March Current Population Survey C-7 tbl.C-4 (1998). During the 1980s and 1990s, the racial gap between blacks and whites participating in white-collar jobs declined dramatically. In 1980, 36.6 percent of blacks and 53.9 percent of whites were in white-collar occupations. In 2000, 51.3 percent of blacks and 62.6 percent of whites were in white-collar occupations. Marshall H. Medoff, Revisiting the Economic Hypothesis and Positional Segregation, 32 Rev. Black Pol. Econ. 83, 91 (2004). Wealth is more racially skewed than income in the United States, a result partially due to decreasing marginal consumption as incomes rise, demographic variables, asset allocation decisions, and disproportionate demands for assistance from low-income family members faced by higher-income African Americans. Joseph G. Altonji et al., Black/White Differences in Wealth, 24 Econ. Persp. 38, 38, 48–49 (2000); N. S. Chiteji & Darrick Hamilton, Family Connections and the Black-White Wealth Gap Among Middle-Class Families, 30 Rev. Black Pol. Econ. 9, 21–25 (2002). Wealth differentials, like income differentials, appear to be less dramatic than golfing participation differentials. See, e.g., Sharmila Choudhury, Racial and Ethnic Differences in Wealth and Asset Choices, 64 Soc. Security Bull. 1, 8 tbl.3 (2002) (noting that a white household in the top income quartile had $551,818 in mean net worth, whereas a black household in the top income quartile had $247,555 in mean net worth). Between 1969 and 1995, the percentage of Southern Caucasians in the top three U.S. wealth quintiles stayed constant at 60 percent, while the percentage of Southern African Americans in this group increased from 27.6 percent to 34.6 percent. MDC Inc., Income and Wealth in the South: A State of the South Interim Report 10, chart 10 (1998). Moreover, among high-income, middle-aged college graduates, wealth disparities between Caucasian and African American families disappear. Ronald L. Straight, Survey of Consumer Finances: Asset Accumulation Differences by Race, 29 Rev. Black Pol. Econ. 67, 76–77 (2001). If one adjusts for age, income, education, and employment, interracial differences in wealth tend to disappear. Id. at 80. 36. Bledsoe et al., supra note 16, at 440; Medoff, supra note 35, at 91 (“By 1999, the number of blacks living in a suburb outside a central city was nearly eleven million, or more than 30% of the total black population, as compared to 9% in 1980.”). Note, however, that suburbanization did not end racial segregation. Many African Americans moved into deteriorating inner suburbs that were becoming majority African American. Joe T. Darden & Sameh M. Kamel, Black Residential Segregation in Suburban Detroit: Empirical Testing of the Ecological Theory, 27 Rev. of Black Pol. Econ. 103, 105 (2000). 37. Teasley et al., supra note 31, at 20–21. 38. Id. at 21. The tennis participation rates were as follows: 10.8 percent of Caucasians; 7.8 percent of African Americans; and 12.8 percent of “Others.” Id. 39. Id. at 24–25. 40. Id. at 23.

NOTES TO PAGES 62–63

41. W. J. Florkowski & G. Landry, An Economic Profile of Golf Courses in Georgia: Course and Landscape Maintenance 4 (Ga. Agric. Experiment Stations Research Report No. 681, 2002) (noting that the average maintenance expenditure—not including land acquisition costs and property taxes—for a Georgia golf course was $417,042 per year); J. Richard McElyea et al., Golf’s Real Estate Value, Urb. Land, Feb. 1991, at 14 (noting the cost of constructing an 18-hole golf course to range from $2 to $8 million). 42. Club goods are not the only means of sorting residents. Common-interest communities conceivably could achieve the same ends through direct subsidies for “sorting” activities, as opposed to club goods provision. For example, a homeowners’ association might provide a subsidy of up to $5,000 per household for rock climbing expenses, and tax all homeowners equally to pay for this subsidy. Presumably, African Americans would be as deterred by this approach as they would be by a residential golf community with a $5,000 annual mandatory membership fee. In light of my theory, why are such arrangements not present in the real world? The puzzling absence of these arrangements is probably explained by legal doctrine. Covenants and equitable servitudes that do not “touch and concern” the land do not bind successors in interest under American property law. Affirmative promises to pay money for common amenities located within a development, such as communal golf courses, have long been held to “touch and concern” the land, but affirmative promises to pay money for rock climbing or other activity subsidies presumably would not satisfy the “touch and concern” requirement. Anthony v. Brea Glenbrook Club, 130 Cal. Rptr. 32, 34–35 (Cal. Ct. App. 1976); Streams Sports Club, Ltd. v. Richmond, 457 N.E.2d 1226, 1230–31 (Ill. 1983); Regency Homes Ass’n v. Egermayer, 498 N.W.2d 783, 791–93 (Neb. 1993); Homsey v. Univ. Gardens Racquet Club, 730 S.W.2d 763, 764 (Tex. App. 1987). 43. Calvin H. Sinnette, Forbidden Fairways: African Americans and the Game of Golf 58–60, 121–32 (1998); Mitchelson & Lazaro, supra note 13, at 48–51. 44. James D. Davidson, Social Differentiation and Sports Participation: The Case of Golf, in Social Approaches to Sport 181, 200 (Robert M. Pankin ed., 1982). 45. This view is premised on the idea that golfers are at least somewhat evenly spread across income levels. If, by contrast, all African American golfers were wealthy, then residential golf courses would not provide an effective means of engaging in the exclusionary amenities strategy. The best available data indicates that the household incomes of African American golfers skew slightly higher than those of Caucasian golfers, but the difference is not particularly pronounced. Nat’l Golf Found., supra note 34, at 16. 46. The number of Americans who played one round or more per year declined from 27,800,000 in 1990 to 26,446,000 in 1999. U.S. Census Bureau, Statistical Abstract of the United States: 2001, at 761 tbl.1244 (2001). These Americans played golf more frequently, however, as the total number of golf rounds played increased from 502,000,000 to 564,100,000 during the same period, a 12 percent increase. Id. 47. John L. Crompton, Designing Golf Courses to Optimize Proximate Property Values, 5 Managing Leisure 192, 192–93 (2000). 48. Id. at 193 (“While the real estate industry in the United States as a whole grew at an annual rate of 2–3% in the 1990s, the annual growth rate of developments which

213

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

214

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 63–64

incorporated golf courses approached 10%, making it one of the hottest sectors in real estate.”). Some subsequent evidence suggests that the construction of new residential golf courses has declined. Kevin Allison, Golf Comes out of the Bunker, Fin. Times, Feb. 1, 2005, at 10. This is consistent with the exclusionary amenities hypothesis. See infra text accompanying notes 34–37. 49. John J. Haydu & Alan W. Hodges, Economic Impacts of the Florida Golf Course Industry, Economic Information Report No. 02–4 (Univ. of Fla. Inst. of Food & Agric. Sci., Gainesville, Fla.), June 13, 2002, at 1, 3; see also Lewis M. Goodkin, Out of the Rough?, Florida Trend, Dec. 1998, at 78, 81 (quoting an earlier estimate that 40 percent of Florida’s golf courses are residential). In 1996, approximately one-third of all newly constructed golf courses were residential. Jordan N. Roberts & Darla DomkeDamonte, Utilization of Golf Course Facilities by Residents of Golf Course Communities in Myrtle Beach, 1 Coastal Bus. J. 13, 14 (2002), http://www.coastal.edu/business/ cbj/pdfs/golfcommunities.pdf. 50. Goodkin, supra note 49, at 78. 51. See, e.g., id. (discussing the emergence of this financing design in Florida). 52. Id. at 80. According to the 2000 census, Miami-Dade and Broward have the largest African American populations among Florida counties. Among Florida’s large counties, they rank second and third, respectively, in percentage of African American residents. Duval County’s population is 27.8 percent African American; Broward’s is 20.5 percent African American; and Miami-Dade’s is 20.3 percent African American. Florida as a whole is 14.6 percent African American. U.S. Census Bureau, County and City Data Book: 2000, at 71 tbl.B-2 (2001). 53. See, e.g., McElyea et al., supra note 41, at 16 (“Golf-course-oriented homes appeal to nongolfers as well as to golfers. [Only about one-third of golf-frontage home-buyers in nonretirement projects play golf regularly.]”); Crompton, supra note 47, at 193; Stella M. Chavez, Subdivisions Want Residents to Join the Club, S. Fla. Sun-Sentinel, Feb. 15, 2000, at A1; Goodkin, supra note 49 (quoting a developer’s expectation that “50% of buyers will be golfers”); Nancy Kressler Murphy, Golf Course Communities Sprouting, Mercer Bus. June 1990, at 15 (quoting a New Jersey developer’s statement that “[f]ifty percent of my buyers are golfers, and then 50 percent have never picked up a club and never plan to”). This pattern, of non-golfers buying homes in residential golf communities, persists today. Robert Johnson, Golf Homes Attract Even Those Who Don’t Play, N.Y. Times, May 8, 2005, at 15. A cautionary note is in order. Although the above-cited sources suggest the presence of large numbers of non-golfers in all types of residential golf courses, I have been unable to find data that breaks down the prevalence of non-golfers in mandatory membership developments. 54. But some of them may be. A 2005 New York Times article discussed a county in North Carolina where overwhelmingly white residential golf communities are surrounded by overwhelmingly black unincorporated areas. The townships containing the residential golf communities refuse to incorporate the largely black neighborhoods and, as a result, the latter are left without the most basic municipal services, such as

NOTES TO PAGES 64–66

garbage collection, piped water, and police protection. Shaila Dewan, In County Made Rich by Golf, Some Enclaves Are Left Behind, N.Y. Times, June 7, 2005, at A1. 55. See, e.g., Murphy, supra note 53, at 15. 56. James R. Rinehart & Jeffrey J. Pompe, Estimating the Effect of a View on Undeveloped Property Values, 67 Appraisal J. 57, 60 (Jan. 1999) (“The results show that ocean views add 147% to lot values, location on a creek or marsh adds 115% to lot prices, and golf course location adds 39% to lot values.”). 57. E-mail from Jim Kass, Research Director, National Golf Foundation, to author (Feb. 15, 2005, 09:15:04 CST). 58. Even within residential golf courses, lots with views of water hazards are particularly desirable and command the highest premiums. Gregory L. Cory et al., Golf Course Development in Residential Communities 37 fig.2–12 (2001); Crompton, supra note 47, at 198. 59. Haydu & Hodges, supra note 49, at 23 (“Commercial, agricultural, industrial, institutional, and government land use types all showed an increase in total value associated with golf courses, averaging $10,942 per parcel, and ranging from nearly $20,00 [sic] for residential properties, $70,000 for commercial properties, $114,000 for industrial, to nearly $121,000 for agricultural land.”). This study included not only residential golf communities, but homes near such communities, as well as those within a mile of public courses, country club courses, and semi-private courses. 60. D. Robert DeChaine, From Discourse to Golf Course: The Serious Play of Imagining Community Space, 25 J. of Comm. Inquiry 132 (2001). Sadly, for my purposes, DeChaine did not distinguish between mandatory-membership and optional-membership communities. 61. Id. at 134. 62. Id. at 138–39. 63. Id. at 139–43. DeChaine’s analysis lacked a quantitative dimension, but his article devoted far more space to discussions of exclusivity than discussions of golf quality. I cannot determine whether this reflects a selection bias or a proportional treatment based on the relative prevalence of developer rhetoric. It would also be helpful to know the extent to which residential golf communities stress exclusivity more or less than other gated communities do in their marketing materials. 64. Harbour Ridge’s Internet advertisements, like Jan Zachariasse’s statements to a reporter, might have been sufficiently candid to invite scrutiny from HUD’s attorneys. See generally Ragin v. N.Y. Times, 923 F.2d 995, 998–99 (2d Cir. 1991) (subtle discriminatory modeling in newspaper advertisement). A quick Internet survey suggested that most advertising messages used by bundled-membership communities do not violate FHA guidelines. 65. Mitchelson & Lazaro, supra note 13, at 69 (emphasis added). 66. For discussions of the heated debates that arise when optional-membership golf communities try to solve this tragedy of the commons by mandating membership, see Tal Abbady, No Change for Boca Lago: Mandatory Membership Voted Down, S. Fla. Sun-Sentinel, May 12, 2004, at 8B; Leon Fooksman, Residents Fight Rule on

215

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

216

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 66–68

Joining Golf Club, S. Fla. Sun-Sentinel, Sept. 17, 2003, at 1B; Lee Hoke, Mandatory Memberships? Solution or Band-Aid?, Club Mgmt., Dec. 2004, at 18; Patty Pensa, Country Club Battle Heads to Court: Community Split Over Required Membership, S. Fla. Sun-Sentinel, Dec. 19, 2004, at 3B. On the privacy drawbacks of solving a tragedy of the commons by opening up the golf course to outsiders, see Mary Shanklin, Golf Communities Tee Off, Orlando Sentinel, Nov. 10, 1996, at J1. 67. James W. Loewen, Sundown Towns: A Hidden Dimension of American Racism 392 (2005) (“Today the tradition of retiring to white enclaves continues, often gated and built around private beaches, golf courses, marinas, or all three. They may provide community, because purchase of a house or town house includes use of a clubhouse, restaurant, sports facilities, and other amenities. . . . While not quite racially segregated, these new towns and developments advertise themselves as ‘exclusive’ and are often overwhelmingly white, although race goes unmentioned.”). Loewen notes that African Americans who move into such communities may face social sanctions from fellow African Americans. Id. at 318 (“A resident of an overwhelmingly white neighborhood near a golf club in south Tulsa told me of a black doctor who moved there. He had to move back to north Tulsa, she said, because ‘his [black] patients rose up in protest.’”) (bracketed text in original). 68. See supra note 12. These latter residents might not object to the presence of an African American celebrity, either. At least two highly prestigious golf-oriented country clubs have Michael Jordan as a member, though virtually no other African American members. Marcia Chambers, The Changing Face of Private Clubs, Golf Dig., Aug. 2000, at 93, 100–01. 69. For discussion along these lines, see Eduardo M. Peñalver, Property as Entrance, 91 Va. L. Rev. 1889, 1962–71 (2005). There is some evidence suggesting that religious residential homogeneity may have some beneficial effects on social welfare, though this data analyzes metropolitan-level homogeneity, as opposed to neighborhood-level homogeneity. Christopher G. Ellison et al., Religious Homogeneity and Metropolitan Suicide Rates, 76 Soc. Forces 273, 287 (1997) (finding that religious homogeneity is associated with decreased suicide rates). 70. Taormina Theosophical Cmty., Inc. v. Silver, 190 Cal. Rptr. 38 (Cal. Ct. App. 1983). Under the FHA, a religious organization may discriminate on the basis of religion with respect to housing that the organization owns or controls through a nonprofit. United States v. Columbus Country Club, 915 F.2d 877, 882 (3d Cir. 1990). A forprofit developer would not be able to take advantage of this exemption. Id. at 882–83. 71. Under the Restatement approach, an equitable servitude generally binds successors unless it (1) “is arbitrary, spiteful, or capricious”; (2) “unreasonably burdens a fundamental constitutional right”; (3) “imposes an unreasonable restraint on alienation”; (4) “imposes an unreasonable restraint on trade or competition”; or (5) “is unconscionable.” Restatement (Third) of Prop.: Servitudes § 3.1 (2000). 72. See, e.g., Hall v. Church of the Open Bible, 89 N.W.2d 798, 799–800 (Wis. 1958) (noting that restrictive covenants excluding churches have been universally enforced).

NOTES TO PAGES 68–77

73. Adam Reilly, City of God: Tom Monaghan’s Coming Catholic Utopia, Boston Phoenix, June 17, 2005, at 17. 74. Id. 75. Id. at 17, 19. “Extremely Catholic” is a double entendre here. Some have suggested that Ave Maria seeks to differentiate itself from other Catholic institutions, like Notre Dame, which Ave Maria’s founders regard as unduly progressive. Sharon Tubbs, School of Faith, St. Petersburg Times, Mar. 28, 2004, at 1E. 76. Questioning Ave Maria Town, May 19, 2007, available at http://avewatch .com/2006–2007/files/6ddac18b8f7eace7c84bf53104edd231–74.html. 77. Id. 78. Reilly, supra note 73, at 19. This “focal points” statement from the developer, quoted by a reporter, may well violate laws that bar religious discrimination in advertising. See generally supra notes 27–30 and accompanying text. 79. Questioning Ave Maria Town, May 19, 2007, available at http://avewatch .com/2006–2007/files/6ddac18b8f7eace7c84bf53104edd231–74.html. 80. Id. 81. James Thorner, Testing Their Faith, St. Petersburg Times, May 8, 2009, at 1A. 82. Amy Keller, Leap of Faith, Fla. Trend Magazine 60 (Sep. 1, 2007), available at 2007 WLNR 17904892. 83. Kevin D. Dougherty, How Monochromatic Is Church Membership? Racial-Ethnic Diversity in Religious Community, 64 Soc. Religion 65, 74–77 (2003) (noting substantial racial homogeneity among U.S. congregations, but a great deal of heterogeneity with respect to income and education); Donald R. Kinder & Tali Mendelberg, Cracks in American Apartheid: The Political Impact of Prejudice Among Desegregated Whites, 57 J. Pol. 402, 417 (1995). 84. F. Gregory Lastowka & Dan Hunter, The Laws of the Virtual Worlds, 92 Cal. L. Rev. 1, 26 n.128 (2004). 85. Blake Sherblom-Woodward, Hackers, Gamers and Lamers: The Use of l33t in the Computer Sub-Culture 6–9 (Fall 2002) (unpublished senior thesis, Swarthmore College), Microsoft.com, A Parent’s Primer to Computer Slang, Feb. 4, 2005, http:// www.microsoft.com/athome/security/children/kidtalk.mspx; see also English to Hackerspeak Translator, http://www.cs.utk.edu/~cjohnson/computing/javascript/ round_hackerspeak.php. 86. Sherblom-Woodward, supra note 85, at 14–15.

5.  Asymmetric Information and Exclusion 1. The story underlying this hypothesis draws on a fundamental premise of agency theory. Namely, the landowner is ordinarily a better agent for her own interests than a potential entrant would be. Hence, if the landowner can make decisions about whom to exclude at the same cost as the potential entrants, she will prefer to keep this discretion for herself.

217

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

218

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 77–79

2. My thesis here can be conceptualized as a contribution to the economics literature on mechanism design, which examines the ways in which actors can structure their affairs to induce individuals to reveal private information that they would otherwise prefer to keep secret for strategic reasons. This literature begins with A. Michael Spence’s work on signaling in the employment context, which discussed employees’ strategy of investing in costly but readily observed educational credentials as a means of signaling to employers that they possess other desirable, but less easily observed, characteristics. A. Michael Spence, Market Signaling: Informational Transfer in Hiring and Related Screening Processes 26–28 (1974). 3. The unattractiveness of utilizing the bouncer’s right might discourage the real estate developer enough so that he starts contemplating governance solutions. For example, he may explore whether he can include rules mandating socialization among residents of his community in the covenants, conditions, and restrictions (CC&Rs) that govern his development. But trying to mandate socialization directly through such covenants would be problematic. In the first place, it would be hard to create rules that defined socialization among neighbors with sufficient precision. Enforcing the rules would be cumbersome and costly; residents might resent the rules and consequently honor their letter rather than their spirit. Furthermore, there may be an important difference in the quality of mandatory socialization versus voluntary socialization. It seems likely that people resent the former and enjoy the latter, and that only the latter produces the welfare gains that would prompt prospective purchasers to pay a premium. Hanoch Dagan & Michael A. Heller, The Liberal Commons, 110 Yale L.J. 549, 581 (2001) (finding that voluntary association engenders higher levels of cooperation than involuntary association). 4. Developers and cooperative boards do sometimes publicize purchases of homes by celebrities, civic leaders, or captains of industry. Ralph Gardner Jr., There Goes the Nabe: Up, Up, Up, N.Y. Times, Jan. 16, 2003, at F1. 5. A pure coordination game is one in which there are two or more possible equilibria, none inherently superior to the others, but the players all have incentives to coordinate “on some equilibrium.” Judith Mehta et al., The Nature of Salience: An Experimental Investigation of Pure Coordination Games, 84 Am. Econ. Rev. 658, 658 (1994). Thomas Schelling devised the most famous pure coordination game experiment, in which he instructed Yale graduate students that they were to meet another student (unknown to the subject) in New York City on a particular day, promising a reward if they successfully met up despite their inability to communicate. A majority of the students identified the clock at Grand Central Station and noon on the relevant day as the natural focal points and claimed the prize by meeting their partners there and then. Thomas C. Schelling, The Strategy of Conflict 55 n.1, 56 (1980). 6. Doron Teichman, Sex, Shame, and the Law: An Economic Perspective on Megan’s Laws, 42 Harv. J. on Legis. 355, 415 (2005). 7. It is appropriate to flag the government’s role in making previously private information about prospective purchasers widely available over the Internet. I will argue in

NOTES TO PAGES 80–83

the next part that by regulating the information market, the government can alter the incentives for resource owners to choose one exclusion strategy over another. 8. Timothy Besley et al., Incentives, Information, and Welfare: England’s New Poor Law and the Workhouse Test, in History Matters: Essays on Economic Growth, Technology, and Demographic Change 245 (Timothy W. Guinnane et al. eds., 2004). The analysis by Besley and his coauthors echoes earlier work on tagging as a means of sorting among deserving and undeserving recipients of welfare in situations involving asymmetric information. See, e.g., George A. Akerlof, The Economics of “Tagging” as Applied to the Optimal Income Tax, Welfare Programs, and Manpower Planning, 68 Am. Econ. Rev. 8, 15–17 (1978); Timothy Besley & Stephen Coate, Workfare versus Welfare: Incentive Arguments for Work Requirements in Poverty-Alleviation Programs, 82 Am. Econ. Rev. 249, 259 (1992); Albert L. Nichols & Richard J. Zeckhauser, Targeting Transfers Through Restrictions on Recipients, 72 Am. Econ. Ass’n Papers & Proc. 372, 375–77 (1982). 9. Besley et al., supra note 8, at 254–55. 10. Workhouse residents were denied contact with members of the opposite sex, including family members. They were deprived of tobacco, served terrible food, forced to work in difficult circumstances, and had their waking hours strictly regulated. Id. at 253–54. 11. Id. at 251. 12. Id. at 257 (“Screening is necessary only because obtaining information on the state of the poor required costly and potentially acrimonious and fraudulent investigation. The workhouse test dispensed with all investigation. By accepting or declining the workhouse, the applicant in effect told the guardians whether he or she was needy.”). 13. Brian S. Prestes, Comment, Application of the Equal Credit Opportunity Act to Housing Leases, 67 U. Chi. L. Rev. 865, 866 (2000). 14. Martha F. Davis, Solving Statute of Limitations Problems Under the Fair Credit Reporting Act, 18 Ind. L. Rev. 507, 515 (1985). 15. Resource owners generally do not sell “grab bags” to members of the public, in which the resource owner knows exactly what the purchaser is buying but the purchaser has not a clue. Still, it is often the case that a resource owner has some private information about his product, though in the context of social goods, this private information will not typically concern the resource’s most salient aspects. 16. At the same time, just as making it past a picky nightclub bouncer may confer social status on a patron, the fraternity and sorority members may benefit from not admitting everyone. The interesting question is whether that status benefit is greater when the resource owner adopts a bouncer’s exclusion strategy as opposed to a nontrespass-based strategy. 17. Joseph Farrell, Cheap Talk, Coordination, and Entry, 18 RAND J. Econ. 39 (1987); Joseph Farrell & Matthew Rabin, Cheap Talk, 10 J. Econ. Persp. 116–17 (1996). 18. Farrell & Rabin, supra note 17, at 116–17. 19. Experimental and historical evidence suggests that cheap talk can have significant effects on participants’ behavior in these settings. Richard H. McAdams & Janice Nadler, Testing the Focal Point Theory of Legal Compliance: The Effect of Third-Party Expression

219

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

220

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 83–85

in an Experimental Hawk/Dove Game, 2 J. Empirical Legal Stud. 87, 116 (2005) (“The results here . . . suggest something more surprising, that third-party expression can by itself influence the behavior of individuals outside of a pure coordination game, in a situation involving significant conflict.”); Richard R. W. Brooks, Covenants and Conventions 4 (Nw. Univ. L. & Econ. Research Paper No. 02–8, Sept. 2002) (“[T]he effectiveness of covenants was not exclusively predicated on their legal enforceability. Covenants were also valued signals that served to coordinate the behavior of a variety of private and institutional actors—signals that remained effective despite the legal unenforceability of covenants. Consistent with this claim, the empirical analysis of the article suggests that the impact of racial restrictive covenants endured long after the Supreme Court ruled their enforcement unconstitutional.”). 20. This analysis assumes simple majoritarian voting rules in the relevant homeowners associations. If an association’s covenants provide for supermajority-voting rule requirements in order to change governance rules or buy out existing residents, then exclusionary vibes may prove insufficient for solving this conflict game, and the developer will have to turn to exclusionary amenities or bouncer’s exclusion. 21. Gerald Gamm, Urban Exodus: Why the Jews Left Boston and the Catholics Stayed 41–42 (1999); Brooks, supra note 19, at 21; Dmitri Mehlhorn, A Requiem for Blockbusting: Law, Economics, and Race-Based Real Estate Speculation, 67 Fordham L. Rev. 1145, 1145 (1998). 22. Henry Hansmann, Condominium and Cooperative Housing: Transactional Efficiency, Tax Subsidies, and Tenure Choice, 20 J. Legal Stud. 25, 31 (1991). 23. Yoram Barzel & Tim R. Sass, The Allocation of Resources by Voting, 105 Q.J. Econ. 745, 764–70 (1990). 24. One of the immediately interesting aspects of bouncer’s exclusion is its capacity to be useful in individual-on-individual exclusion settings, as well as individual-on-group and group-on-group settings. So, a bouncer’s right might prove handy when only one individual is being sought (e.g., the search for a prom date), or when a resource owner seeks a large group of resource users (e.g., the bouncer at an expensive nightclub). Upon reflection, however, it appears that non-trespass-based strategies have the same flexibility. A high school student might use exclusionary vibes to ward off undesirable courtiers (e.g., by publicizing her unique fashion sense and musical taste) or via an exclusionary amenities strategy (e.g., “I will only go to the prom with someone who can also find a date for my homely friend.”). By the same token, groups might employ any of these three exclusion strategies—bouncers, vibes, or amenities—to control the composition of their memberships. 25. Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, 425 U.S. 748 (1976). 26. See, e.g., Lorillard Tobacco Co. v. Reilly, 533 U.S. 525 (2001); 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996). 27. Note, Making Sense of Hybrid Speech: A New Model for Commercial Speech and Expressive Conduct, 118 Harv. L. Rev. 2836, 2853–54 (2005). 28. Tahoe-Sierra Pres. Council v. Tahoe Reg’l Planning Agency, 535 U.S. 302 (2002); Penn Cent. Transp. Co. v. City of New York, 438 U.S. 104 (1978).

NOTES TO PAGES 86–88

29. There are, of course, exceptions. It may be difficult to determine the racial status of a mixed-race person, and some bouncers may be unable to differentiate between members of particular racial groups through visual inspection (e.g., between East Asians and Native Americans) or reliance on surnames (e.g., between Latinos and Filipinos). See generally Amanda E. Lewis, Everyday Race-Making, 47 Am. Behav. Sci. 283, 291–94 (2003); E. J. Pérez-Stable et al., Use of Spanish Surnames to Identify Latinos: Comparison to Self-Identification, 18 J. Nat’l Cancer Inst. Monograph 11 (1995); Mary C. Waters, The Everyday Use of Surname to Determine Ethnic Ancestry, 12 Qualitative Soc. 303 (1989). Note that in this context I am referring to race in a biological sense, rather than in a socially defined sense. 30. Alternatively, their preferences for homogeneity may shift over time. They might begin with preferences for racial homogeneity, but their contact with the Dans of the world may convince them that their preference for racial homogeneity was misplaced. For scholarship providing support for this “contact hypothesis” in the racial setting, see Casey J. Dawkins, Recent Evidence on the Continuing Causes of Black-White Residential Segregation, 26 J. Urb. Aff. 379, 389 (2004); Robert D. Tollison, Consumption Sharing and Non-Exclusion Rules, 39 Economica 276, 283 (1972). 31. Classic legal scholarship on social meaning includes Dan M. Kahan, Social Influence, Social Meaning, and Deterrence, 83 Va. L. Rev. 349 (1997); Lawrence Lessig, The Regulation of Social Meaning, 62 U. Chi. L. Rev. 943 (1995); and Cass R. Sunstein, Social Norms and Social Roles, 96 Colum. L. Rev. 903 (1996). 32. See generally Coco Henson Scales, The Hostess Diaries: My Year at a Hot Spot, N.Y. Times, July 11, 2004, § 9, at 1 (describing a bouncer’s interactions with would-be entrants to a trendy club). 33. Clubs are as capable as restaurants of using price as a blunt exclusionary device, but many clubs rely on bouncers anyway. Perhaps the explanation for this is that many resource-poor club patrons can still enter desirable clubs because they either expect other club patrons to buy them drinks, or they do not expect to drink much. Furthermore, rich but ugly patrons pose a greater danger in a nightclub than they do in a restaurant, where interactions among strangers are limited. Sober bouncers serve a “vital” function of preventing the intoxicated beautiful people from succumbing to the advances of the riffraff. 34. Despite its name, exclusionary zoning does not fit into my typology of exclusion. In this book, I deal with instances in which a resource owner is deciding to exclude, as opposed to those instances in which a government is deciding to exclude, albeit sometimes in response to the concerns of multiple resource owners. That said, the exclusionary zoning strategy (mandating large lot sizes, for example) is in some ways similar to the exclusionary amenities strategy (requiring homeowners in a residential subdivision to join a golf club, for example). For a fuller discussion, see supra chapter one, note 37. 35. James C. Clingermayer, Heresthetics and Happenstance: Intentional and Unintentional Exclusionary Impacts of the Zoning Decision-making Process, 41 Urb. Stud. 382–83 (2004).

221

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

222

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 89–97

36. Loewen describes this phenomenon as the “paradox of exclusivity,” whereby overwhelmingly white communities greatly desire racial homogeneity but “develop a motivated blindness to the workings of social structure” and adamantly take issue with any suggestion that racist sentiment pervades the community or motivates its policies. James W. Loewen, Sundown Towns: A Hidden Dimension of American Racism 316–20 (2005). 37. These concerns about backlash are interesting in and of themselves. Consumers who feel some shame about their own preferences may blame real estate developers for responding to those same consumer preferences, as a psychological coping mechanism. See generally Lyn H. Lofland, The Real-Estate Developer as Villain: Notes on a Stigmatized Occupation, in 27 Studies in Symbolic Interaction 85, 98 (Norman K. Denzin ed., 2004) (noting that housing consumers simultaneously demanded suburban homes and vilified developers for providing suburbanized homes for the masses). 38. Some people who plainly belong to the group targeted for inclusion will not respond to an ambiguous message if they are too busy to make deciphering the message worthwhile. 39. Associated Press, Bouncer Turns Gov. Gregoire Away from Olympia Bar, Seattle Times, July 30, 2008. 40. Id.

6.  Regulating the Choice Among Exclusion Strategies 1. On the other hand, it is conceivable that seeing a bouncer unfairly turn away a prospective entrant is a particularly visceral experience, both for the thwarted entrant and bystanders. If so, the intensity of the negative reaction may make up for its relative rarity, rendering the symbolic externalities associated with bouncer’s exclusion substantial. 2. United States v. Hunter, 459 F.2d 205, 213–14 (4th Cir. 1972); 42 U.S.C. § 3603(b)(2) (2000). For a criticism of the persistence of the Mrs. Murphy exception, see James D. Walsh, Reaching Mrs. Murphy: A Call for Repeal of the Mrs. Murphy Exemption to the Fair Housing Act, 34 Harv. C.R.-C.L. L. Rev. 605 (1999). 3. Hunter, 459 F.2d at 214; see also Spann v. Colonial Vill., Inc., 899 F.2d 24, 30 (D.C. Cir. 1990) (Bader Ginsburg, J.) (noting, in the FHA standing context, that discriminatory ads could create “a public impression that segregation in housing is legal, thus facilitating discrimination by . . . other property owners”). 4. United States v. Hunter, 459 F.2d 205, 214 (4th Cir. 1972). 5. Spann v. Colonial Vill., Inc., 899 F.2d 24, 30 (D.C. Cir. 1990); see also Felicia R. Lee, ABC Drops Show After Complaints by Civil Rights Groups, N.Y. Times, June 30, 2005, at C3 (describing how a television network pulled the plug on a reality television program in which neighbors selected which of seven diverse families would live in a new home in a Christian, Republican subdivision, after civil rights groups complained that this program would cause members of the public to believe that racial and other forms of discrimination are permissible). 6. See supra text accompanying chapter five, notes 26–28.

NOTES TO PAGES 98–100

7. 766 A.2d 1186, 1193 (N.J. Super. Ct. App. Div. 2001). It is not only homeowners associations that are restricting the residential choices of sex offenders. State and local governments have ushered in the era of “zoning people” by enacting laws restricting sex offenders from residing, working, or even approaching within 1,000 or 2,000 feet of schools, daycare facilities, and other areas where minors congregate. Robert F. Worth, Exiling Sex Offenders from Town, N.Y. Times, Oct. 3, 2005, at B1. Iowa’s law rendered 77 percent of the state’s housing units off-limits to sex offenders, and most of the remaining 23 percent of the state’s units consisted of rural farmhouses. Nevertheless, the Eighth Circuit upheld the constitutionality of the statute. Doe v. Miller, 405 F.3d 700, 706 n.2, 714–15 (8th Cir. 2005). 8. Perhaps the most influential case on this score is Matthews v. Bay Head Improvement Ass’n, 471 A.2d 355 (N.J. 1984), a staple of property casebooks. In Matthews, the New Jersey Supreme Court held that the owners of land abutting beachfront public trust lands were not required to permit public access across their property as long as sufficient numbers of their neighbors voluntarily permitted the public to cross their property to reach the beach. Id. at 369. The court warned that if too many neighbors stopped permitting the public to access the beach via their lands, the court would mandate that all owners of beachfront property in the area be subject to a public easement. Id. Thus, the court embraced an analog to the dumping grounds argument, permitting exclusion by private landowners only insofar as that exclusion did not unduly limit the options available to potential entrants. 9. For a thorough review of the legislative debates, see Daniel M. Filler, Making the Case for Megan’s Law: A Study in Legislative Rhetoric, 76 Ind. L.J 315, 315, 327 (2001). 10. Dan Eggen, Sex Offender Registry Announced, Ft. Worth Star-Telegram, May 21, 2005, at A11 (forty-eight states as of 2005); Note, Making Outcasts Out of Outlaws: The Unconstitutionality of Sex Offender Registration and Criminal Alien Detention, 117 Harv. L. Rev. 2731, 2731 n.5 (2004) (forty-five states as of 2004). 11. Meghann J. Dugan, Comment, Megan’s Law or Sarah’s Law: A Comparative Analysis of Public Notification Statutes in the United States and England, 23 Loy. L.A. Int’l & Comp. L. Rev. 617, 622–23 (2001) (describing variations among the states with respect to the ease of accessing information about sex offenders); see generally Daniel J. Solove, The Virtues of Knowing Less: Justifying Privacy Protections Against Disclosure, 53 Duke L.J. 967, 1061 (2003) (“Megan’s Law data are beneficial when disclosed for certain purposes, but not necessarily for all purposes. When placed on the Internet for any curious individual around the world to see, Megan’s Law information becomes disconnected from its goals.”). California appears to be something of an outlier in that respect, in that the state permits ready access to its sex offender registry but criminalizes the use of information so obtained in housing, employment, and other contexts. Cal. Penal Code § 290.46(j)(2) (West 1999). 12. Eggen, supra note 10, at A11. 13. On the (close) connection between privacy as a legal concept and obscurity, see Lior Jacob Strahilevitz, A Social Networks Theory of Privacy, 72 U. Chi. L. Rev. 919, 930–31 (2005).

223

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

224

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 100–102

14. See, e.g., Maria Giordano, Lookout Group’s Effort Curbs Crime, New Orleans TimesPicayune, July 4, 1993, at F1; Ignacio Lobos, Mill Creek Rape Causes Fear Beyond City Limits, Seattle Times, May 21, 1991, at A1; Dale Rodebaugh, M.H. Parents to Discuss Kids Being Accosted, San Jose Mercury News, May 3, 1989, at 1B. In one instance, neighbors sent a signal to potential sex offenders by taunting a sex offender who had moved into a group home. Lois M. Takahashi & Michael J. Dear, The Changing Dynamics of Community Opposition to Human Service Facilities, 63 J. Am. Planning Ass’n 79, 79 (1997). 15. In pointing to the potential for information policy to cause resource owners to substitute bouncer’s exclusion for non-trespass-based exclusion strategies, or vice versa, I am not suggesting that government information policies have no effect on aggregate exclusion levels. To the contrary, the various Megan’s Laws plausibly further stigmatized sex offenders, resulting in an increase in the private resources devoted to excluding sex offenders from neighborhoods. See generally Deborah Hellman, The Expressive Dimension of Equal Protection, 85 Minn. L. Rev. 1, 61 n.261 (2000). 16. Abraham Bell & Gideon Parchomovsky, The Integration Game, 100 Colum. L. Rev. 1965, 1975–81 (2000); David M. Cutler et al., The Rise and Decline of the American Ghetto, 107 J. Pol. Econ. 455 (1999); Reynolds Farley & William H. Frey, Changes in the Segregation of Whites from Blacks During the 1980s: Small Steps Toward a More Integrated Society, 59 Am. Soc. Rev. 23, 40–41 (1994). 17. Hack v. President & Fellows of Yale Coll., 237 F.3d 81, 91 (2d Cir. 2000). 18. Id. at 88. Perhaps the court went easy on Yale because it was thinking in terms of antidiscrimination law more generally, where facially neutral classifications with a racially disparate impact are viewed with less skepticism than the sort of express classifications that might be referenced via an exclusionary vibe. 19. The FHA prohibits discriminatory “statements,” not just advertisements. 42 U.S.C. § 3604(c) (2000). As a result, there may be some exclusionary amenities that confer such an obvious signal as to run aground of the FHA. For example, Lee Fennell suggests that, had Yale placed wallpaper in every common hallway reproducing pages from the New Testament, Yale might have fared poorly in the litigation. 20. See, e.g., Ragin v. N.Y. Times Co., 923 F.2d 995 (2d Cir. 1991). 21. As a general matter, violations of exclusionary vibes prohibitions will be easiest to detect. After all, exclusionary vibes must be publicized to outsiders in order for them to work effectively. In the process of advertising to her intended (and unintended) audiences, a resource owner pursuing an exclusionary vibe strategy will be alerting law enforcers to the nature of her conduct as well. Improper exercises of the bouncer’s right would seem to be much harder to detect. In the case of the bouncer’s right, the resource owner makes no representation of whom she wants to attract and may not keep records about whom she has turned away. Discovering a prohibited use of the bouncer’s right therefore involves substantial effort by law enforcers. Problems of proof seem to go hand in hand with problems of detection, too. Indeed, resource owners can often get away with controversial bouncer’s right strategies and tend to get caught only if they are loose-lipped. The fact that exclusionary vibes are so easy to detect strengthens the case for policing them more than other forms of exclusion, but perhaps also suggests that penalties should be higher

NOTES TO PAGES 103–106

when individuals violate prohibitions on bouncer’s exclusion. Gary S. Becker, Crime and Punishment: An Economic Approach, 76 J. Pol. Econ. 169, 189–96 (1968). 22. There will, of course, be cases in which a prospective tenant’s racial status is ambiguous, or the landlord fails to identify the prospective tenant’s racial status accurately. 23. 190 Cal. Rptr. 38 (Ct. App. 1983). 24. Id. at 41. 25. Id. at 43. Readers interested in more background on Theosophy might enjoy reading Theosophical History, http://www.theohistory.org. 26. Taormina, 190 Cal. Rptr. at 43–44 & n.7. 27. Id. at 40. 28. Id. Purchasers who had been members for less than three years would be allowed to purchase or occupy homes in the community with the consent of Taormina’s Board of Trustees. Id. 29. Thinking about the religion issue from a private information perspective is also revealing. An individual’s religious affiliation is almost never evident from visual inspection, and some people may refuse to disclose their religious affiliations or give misleading answers if asked. Moreover, even where religious affiliation is readily observable, the intensity of one’s devotion to one’s faith is not, absent access to the religious leaders of a particular congregation. If a developer seeks to set up a new religious community, as opposed to one that will draw exclusively on the members of an existing congregation, it will be very difficult for him to differentiate among potential residents on the basis of the intensity of their ties to the religion. Hence, costly club goods in the form of an exclusionary amenities strategy might have real appeal quite apart from the law’s stance toward bouncer’s exclusion and exclusionary vibes. 30. Taormina, 190 Cal. Rptr. at 40. 31. Id. at 42–43. There is a narrow exemption in the federal FHA that permits religious organizations and the nonprofits that they supervise to limit the sale or rental of housing on the basis of religion. 42 U.S.C. § 3607(a) (2000); United States v. Columbus Country Club, 915 F.2d 877, 881–83 (3d Cir. 1990). This exemption does not preempt states from creating liability under their own housing discrimination laws. See generally Melissa Fishman Cordish, Comment, A Proposal for the Reconciliation of Free Exercise Rights and Antidiscrimination Law, 43 UCLA L. Rev. 2113, 2133 (1996). 32. Taormina, 190 Cal. Rptr. at 43. 33. Id. at 42–43. 34. Miles Hewstone et al., Intergroup Contact in a Divided Society: Challenging Segregation in Northern Ireland, in The Social Psychology of Inclusion and Exclusion 265, 266–280 (Dominic Abrams et al. eds., 2004). 35. See Matthew E. Brashears, Anomia and the Sacred Canopy: Testing a Network Theory, 32 Social Networks 187 (2010). 36. See, e.g., Gregory S. Alexander, Dilemmas of Group Autonomy: Residential Associations and Community, 75 Cornell L. Rev. 1, 50, 52 (1989) (arguing that exclusion might decrease sympathy for outsiders and that intentional communities typically want to isolate themselves from the outside world); J. Eric Oliver, The Effects of Metropolitan

225

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

226

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 106–115

Economic Segregation on Local Civic Participation, 43 Am. J. Pol. Sci. 186, 198–200 (1999) (finding that residential homogeneity leads to declining political participation); Cass R. Sunstein, Deliberative Trouble? Why Groups Go to Extremes, 110 Yale L.J. 71, 98–102, 108–11 (2000) (arguing that heterogeneous groups are likely to deliberate more effectively than homogeneous groups). 37. Alexander, supra note 36, at 38; Hanoch Dagan & Michael A. Heller, The Liberal Commons, 110 Yale L.J. 549, 571 (2001). 38. Seen in this light, the presence of lengthy, detailed CC&Rs in a homeowners association’s founding documents might be a symptom of weakness and dysfunction in a community. A community that can substitute exclusion for governance probably need not regulate resident conduct heavily through CC&Rs, because selection effects prevent conflicting uses and preferences from arising. Note further that the late date on which CC&Rs are usually disclosed to a buyer (after the seller accepts an offer, but before closing) renders CC&Rs themselves largely irrelevant in prompting sorting by purchasers. Stephanie Stern, Temporal Dynamics of Disclosure: The Example of Residential Real Estate Conveyancing, 2005 Utah L. Rev. 57, 93–94. 39. 533 F.3d 562 (7th Cir. 2008). 40. Id. at 567 (Wood, J., dissenting). 41. Id. 42. Id. at 565. 43. Id. at 572. 44. Id. at 570. 45. Id. at 568. 46. For an excellent analysis of the case and the oral argument, see Martha C. Nussbaum, Deliberation and Insight: Bloch v. Frischholz and the “Chicago School” of Judicial Behavior, 77 U. Chi. L. Rev. 1139 (2010). 47. Bloch v. Frischholz, 587 F.3d 771, 779, 782–83 (7th Cir. 2009) (en banc). 48. Id. at 785. 49. Id. at 786.

7.  Bundled Amenities to Reduce Discrimination 1. It may be more difficult to justify religious residential segregation by members of vibrant, commonly practiced religions, such as Roman Catholicism, though distinct Roman Catholic subpopulations may be able to make colorable “critical mass” arguments. 2. On the connection between condominium amenities and resident selection effects, see Kathy McCormick, Condo Amenities Reflect Changing Needs: Cover the Gamut from Car Wash Bays to 24-Hour Concierge, Nat’l Post, Mar. 3, 2001, at N4. 3. See, e.g., Dick Thornburgh, The Americans with Disabilities Act: What It Means to All Americans, 64 Temp. L. Rev. 375, 376, 383 (1991); Tim Gilmer, A Tale of Two Cities, New Mobility, June 2002, available at http://www.newmobility.com/review_article .cfm?id=555&action=browse (noting that Venice, Florida, tried to make itself accessible to the disabled decades before the enactment of legislation mandating access).

NOTES TO PAGES 116–118

4. See, e.g., Abe Aamidor, Cool Indy, Indianapolis Star, Oct. 3, 2004, at J1; Timothy J. Gibbons, The Cool Factor: Jacksonville Has Much to Do to Attract Young, Creative Workers, Fla. Times-Union, Feb. 16, 2004, at 10; Keith Herbert, Struggling Borough Tries to Get Creative: Norristown Hopes Artists Will Be Drawn by Low Rent and Incentives Such as Tax Breaks, Phila. Inquirer, Aug. 15, 2004, at B1; Elaine Hopkins, Cheap Rent + Good Light = Art: Arts Project Rep Says that Peoria’s Buildings Are the Perfect Places to Foster Creativity, Peoria J. Star, June 12, 2004, at B3; E. J. Schultz, Artists, Writers and Young Professionals See Potential in the Region’s Budding Arts and in Fresno’s Reviving Downtown, as They Try to Remake the City into . . . Creative Fresno, Fresno Bee, Jan. 9, 2005, at D1. 5. Richard Florida, The Rise of the Creative Class: And How It’s Transforming Work, Leisure, Community, and Everyday Life (2002). 6. James E. Ryan, Schools, Race, and Money, 109 Yale L.J. 249, 281–83 & n.152 (1999). 7. Lee Anne Fennell, Beyond Exit and Voice: User Participation in the Production of Local Public Goods, 80 Tex. L. Rev. 1, 25–31 (2001). 8. Robin D. Barnes, Black America and School Choice: Charting a New Course, 106 Yale L.J. 2375, 2402 (1997). Several papers have critiqued the use of magnet schools to diminish white flight. See, e.g., Christine Rossell, The Desegregation Efficiency of Magnet Schools, 38 Urb. Aff. Rev. 697 (2003); Kimberly C. West, Note, A Desegregation Tool That Backfired: Magnet Schools and Classroom Segregation, 103 Yale L.J. 2567 (1994). 9. Bernard E. Harcourt, Policing L.A.’s Skid Row: Crime and Real Estate Redevelopment in Downtown Los Angeles [An Experiment in Real Time], 2005 U. Chi. Legal F. 323, 333 (2005). 10. Id. at 333. 11. See generally Charles R. Lawrence III, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, 39 Stan. L. Rev. 317 (1987). 12. For this reason, it may be appropriate to define exclusionary amenities with reference to people who live within a metropolitan area but are targeted for exclusion from a particular development, and inclusionary amenities with reference to people who live throughout the United States but are targeted for inclusion in a particular development. Residential developers sometimes try to attract residents from distant states or regions, but they rarely worry about excluding residents from distant states or regions. 13. Ronald T. Mitchelson & Michael T. Lazaro, The Face of the Game: African Americans’ Spatial Accessibility to Golf, 44 Se. Geographer 52–53 (2004). 14. There are important connections between my argument here and an argument voiced by Clayton Gillette. He notes that within common-interest communities, certain types of restrictive covenants might be imposed, not because the residents object to the proscribed land uses themselves, but because they object to the types of people who might engage in the proscribed uses. Gillette gives the following example, justifying restrictions on trailer homes: [E]ven where individuals do not have an aversion to certain practices that are prohibited in covenants, such as maintenance of trailer homes, they may believe that there is a correlation between the subject of the covenant and characteristics

227

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

228

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 119–120

that can serve as the basis for a desirable affinity. I may have nothing against trailer homes, other things being equal. That is, I may believe that they are not aesthetically displeasing, and may believe that they offer the best available housing opportunities for a large segment of the population. I may, however, simultaneously seek a relatively noise-free environment, or assurances that I live among others who do not mind a high degree of regimentation, and hence are less likely to be offended when I complain of what to me is excessive noise. A covenant against “unreasonable noise” may be too imprecise to accomplish my objectives. I therefore may prefer a more certain surrogate that reflects the level of comfort to which I aspire. If I believe that the presence of trailers is positively correlated with bothersome levels of noise, a covenant against trailer homes may serve this proxy role. Clayton P. Gillette, Courts, Covenants, and Communities, 61 U. Chi. L. Rev. 1375, 1396 (1994). The essential difference between Gillette’s example and my own is strategic. Gillette focuses on covenants that restrict the use of particular private goods, whereas my examples show how the same objectives can be satisfied through the provision (or lack thereof) of club and public goods. 15. In theory, service providers ought to be able to pass these transportation costs onto homeowners whose homes are not proximate to public transportation. Their ability to do so may be constrained, however, to the extent that demand for these services is elastic. 16. See, e.g., Hong Chen et al., Measuring the Impact of Light Rail Systems on Single Family Home Values: A Hedonic Approach with GIS Application (Portland State Univ. Ctr. for Urban Studies, Discussion Paper No. 97–3, 1997), available at http:// www.upa.pdx.edu/CUS/publications/docs/DP97–3.pdf; Roderick B. Diaz, Impacts of Rail Transit on Property Values (Am. Pub. Transp. Ass’n, Washington D.C.), May 1999, available at http://apta.com/research/info/briefings/documents/diaz.pdf. 17. Residence Life Frequently Asked Questions, available at http://www.avemaria .edu/FAQs/#23. 18. Zachary Moses Schrag, The Washington Metro as Vision and Vehicle, 1955–2001, at 268–71 (2002) (unpublished Ph.D. dissertation, Columbia University). Although it is often asserted that neighbors’ fear of outsiders explains the absence of a subway station in Georgetown, see, e.g., Stephen C. Fehr, Where D.C. Wants Metro to Go Next, Wash. Post, Mar. 23, 1994, at D3; Juan Williams, Georgetown: Separate City, Wash. Post, Dec. 8, 1981, at A21. Schrag concludes that there is only “a kernel of truth” to the Georgetown story, since engineering challenges and economic considerations helped steer the Metro away from Georgetown. Schrag, supra, at 268–69; see also Bob Levey, Metro’s Not Coming to Georgetown—and Nobody’s Crying, Wash. Post, June 30, 1977, at D.C.1 (noting several bases for neighborhood opposition). 19. James W. Loewen, Sundown Towns: A Hidden Dimension of American Racism 254–55 (2005) (“At the behest of the wealthy . . . officials in Nassau County allowed all public roads to fall into disrepair. . . . [R]esidents of . . . a New York City suburb would rather bear the inconvenience of narrow and congested streets on a day-by-day basis than make it easier for the inhabitants of New York City to reach the town. Even

NOTES TO PAGES 120–122

street signs are in short supply in Darien, Connecticut, making it hard to find one’s way around that elite sundown suburb. Darien doesn’t really want a lot of visitors, a resident pointed out, and keeping Darien confusing for strangers might deter criminals—perhaps a veiled reference to African Americans. . . . Sidewalks and bike paths are rare and do not connect to those in other communities inhabited by residents of lower social and racial status. Some white suburbs of San Francisco opted out of the Bay Area Rapid Transit system, fearing it might encourage African Americans to move in. . . . Parks, tennis courts, and playgrounds may be few or located on minor roads where visitors will be unlikely to find them. . . . San Marino, an elite suburb of Los Angeles, closes its parks on weekends to make sure the neighboring Asian and Latin communities are excluded, thus keeping out everyone, even its own residents.”) (internal quotation marks omitted). 20. A similar example arises in Chicago’s Hyde Park community—an increasingly affluent university neighborhood that, quite conspicuously, lacks a movie theater. Hyde Park had a movie theater in the 1990s, but it drew large numbers of African American youths from surrounding Chicago neighborhoods. Eventually, the University of Chicago, which owned the land, elected to close the cinema entirely, notwithstanding complaints from students. Hyde Park’s lack of a cinema and other entertainment amenities prompts many graduate and professional students to live in distant neighborhoods and endure long commutes to the campus. 21. Corey Dolgon, The End of the Hamptons: Scenes from the Class Struggle in America’s Paradise 124–25, 156 (2005). 22. Celia W. Dugger, A Bus System Reopens Rifts in South Africa, N.Y. Times, Feb. 21, 2010, at A1. 23. For discussions of the ecological consequences of golf course development, see James C. Balogh et al., Background and Overview of Environmental Issues, in Golf Course Management & Construction: Environmental Issues 1 (James C. Balogh & William J. Walker eds., 1992); M. K. Brewin, An Annotated Bibliography and Literature Review on the Potential Impacts of Golf Courses on Freshwater Environments 44–129 (1992) (summarizing the existing literature and providing an annotated bibliography); Michael A. Lewis et al., Effects of a Coastal Golf Complex on Water Quality, Periphyton, and Seagrass, 53 Ecotoxicology & Envtl. Safety 154 (2002); Dep’t of Envtl. Res. Mgmt., Environmental Quality Monitoring at Five Municipal Golf Courses in Miami-Dade County (2002), available at http://www.miamidade.gov/derm/land/library/golf_course.pdf. 24. R. H. Coase, The Firm, the Market, and the Law 182–85 (1988). 25. See, e.g., Gary S. Becker, The Economics of Discrimination 41 (2d ed. 1971); Richard A. Epstein, Forbidden Grounds: The Case Against Employment Discrimination Laws 41–42 (1992). 26. Casey J. Dawkins, Recent Evidence on the Continuing Causes of Black-White Residential Segregation, 26 J. Urb. Aff. 387–88 (2004) 27. Id. at 389; see also Marcia Chambers, The Changing Face of Private Clubs, Golf Dig., Aug. 2000; Robert D. Tollison, Consumption Sharing and Non-Exclusion Rules, 39 Economica 283 (1972).

229

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

230

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 122–125

28. Nat’l Golf Found., Minority Golf Participation in the U.S. 4 (2003); April Adamson, Tiger Draws Many to Sport, Phila. Daily News, June 24, 2004, at 30. 29. Members might value social interactions as such, and may therefore want to bind themselves to interact socially with their neighbors. Mandatory membership will reduce each household’s disposable income, thereby limiting their opportunities for social interactions with people from outside the residential golf community. Since people have difficulty ignoring sunk costs, having already paid for a membership at a golf club might cause them to play more golf and attend more golf-course-related events than they otherwise would have. 30. Gregory L. Cory et al., Golf Course Development in Residential Communities 166–73 (2001). 31. See generally Jerry A. Hausman & J. Gregory Sidak, A Consumer-Welfare Approach to the Mandatory Unbundling of Telecommunications Networks, 109 Yale L.J. 417 (1999); Randal C. Picker, The Digital Video Recorder: Unbundling Advertising and Content, 71 U. Chi. L. Rev. 205 (2004). 32. United States v. Microsoft Corp., 253 F.3d 34, 96–97 (D.C. Cir. 2001). 33. To be sure, permitting the construction of exclusionary amenities near homes might still facilitate pernicious forms of segregation through a focal points mechanism. That is to say, in a world with no mandatory membership exclusionary amenities, those with a preference for racial homogeneity would be drawn to residential communities that are located near racially polarizing amenities, and those with a preference for racial heterogeneity might be deterred from moving into these communities. But a central argument of this chapter is that focal points and sorting are particularly powerful when they function together, and the unbundling strategy at least prevents sorting from occurring. Moreover, many new residential developments are surrounded by undeveloped or agricultural land. Robert W. Burchell, Economic and Fiscal Costs (and Benefits) of Sprawl, 29 Urb. Law. 159, 160 (1997). In these communities, developers prevent people from “free-riding” on the homogeneity that results from exclusionary amenities by ensuring that only residents of the common-interest community can live near the exclusionary amenity. 34. In 1996, 10.1 percent of Caucasians and 2.5 percent of African Americans identified themselves as avid fans of professional golf. In 2003, 11.8 percent of Caucasians and 12.0 percent of African Americans identified themselves as avid fans of professional golf. Thus, whereas avid fandom increased by 16.8 percent among Caucasians, it increased by 380 percent among African Americans. The increases among casual fans were not as dramatic. Casual fandom increased by 10.5 percent among Caucasians and 73.2 percent among African Americans. Golf 20/20, Golf 20/20 Vision for the Future: Industry Report for 2003, at 12 (2004). 35. Jill Lieber, Golf Finally Reaching Out, USA Today, Aug. 15, 2001, at 1C. 36. John Paul Newport, PGA Moves to Address a Past Racial Injustice, Wall Street J., Aug. 8–9, 2009, at W4 (“Today, of about 28,000 members and apprentices in the PGA, 145 are African-American.”). 37. Residential golf courses might still function as exclusionary club goods, but they would prompt sorting on the basis of some factor other than race. For example,

NOTES TO PAGES 125–130

men are noticeably more likely than women to participate in golf. Nat’l Golf Found., supra note 28, at 20 (noting that 22 percent of white adult males play golf, versus 6 percent of white adult females, although the discrepancies are less pronounced for racial minorities). Given this disparity, it may be that married couples who purchase homes in residential golf communities are more patriarchal than ordinary married couples, in the sense that the husband plays a dominant role in making important family decisions, like the choice of residential location. 38. Paul H. Gobster, Explanations for Minority “Underparticipation” in Outdoor Recreation: A Look at Golf, 16 J. Park & Recreation Admin. 60–61 (1998). 39. Their goal was to establish a new town “expressly created for people who sign.” Monica Davey, As Town for Deaf Takes Shape, Debate on Isolation Re-emerges, N.Y. Times, Mar. 21, 2005, at A1. Community planners were excited about the prospect of a town in which signing is the language of choice and community services could be geared toward a largely deaf population. 40. Indeed, organizers selected South Dakota as a home for their community in no small measure because of the state’s small population and their anticipated ability to achieve real political representation in short order. Id. Given South Dakota’s climate and its dearth of urban amenities, South Dakota may itself function as an exclusionary public good. Signers were attracted to South Dakota, not because of what it offers, but because of its effectiveness in keeping non-deaf outsiders from outnumbering the deaf population in Laurent.

8.  Information Is a Variable, Not a Constant 1. Zhongmin Wang, Anonymity, Social Image, and the Competition for Volunteers: A Case Study of the Online Market for Reviews, 10 B.E. Journal of Econ. Anal. & Pol’y (Article 44 2010). 2. Katie Hafner, Service at Restaurants Changing Subtly with Online Information, N.Y. Times, June 18, 2007, at C1. 3. See generally Nolan Miller et al., Eliciting Informative Feedback: The PeerPrediction Model, 51 Mgmt. Sci. 1359 (2005) (developing “methods to elicit feedback effectively when independent, objective outcomes are not available”). 4. See, e.g., Paul Resnick et al., The Value of Reputation on eBay: A Controlled Experiment, 9 Experimental Econ. 79, 80–81, 96 (2006). 5. See, e.g., Wayne R. Barnes, Rethinking Spyware: Questioning the Propriety of Contractual Consent to Online Surveillance, 39 U.C. Davis L. Rev. 1545, 1550–57 (2006); Lee Kovarsky, Note, Tolls on the Information Superhighway: Entitlement Defaults for Clickstream Data, 89 Va. L. Rev. 1037, 1042–47 (2003); Brad Stone, MySpace Mining Members’ Data to Tailor Ads Expressly for Them, N.Y. Times, Sept. 18, 2007, at C1. 6. On the latter, see Dana Cuff et al., Urban Sensing: Out of the Woods, Comm. ACM, Mar. 2008, at 24, 27–28. 7. See, e.g., Adam Greenfield, Everyware: The Dawning Age of Ubiquitous Computing (2006); Marije Kanis et al., Toward Wearable Social Networking with

231

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

232

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 130–132

iBand, in CHI ’05 Extended Abstracts on Human Factors in Computing Systems 1521 (Wendy Kellogg et al. eds., 2005), available at http://portal.acm.org/citation .cfm?id=1056956; Gerd Kortuem & Zary Segall, Wearable Communities: Augmenting Social Networks with Wearable Computers, Pervasive Computing, Jan.–Mar. 2003, at 71, available at http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1186728; Steve Mann et al., Sousveillance: Inventing and Using Wearable Computing Devices for Data Collection in Surveillance Environments, 1 Surveillance & Soc’y 331 (2003), http:// www.surveillance-and-society.org/articles1(3)/sousveillance.pdf; Michael Terry et al., Social Net: Using Patterns of Physical Proximity over Time to Infer Shared Interests, in CHI ’02 Extended Abstracts on Human Factors in Computing Systems 816 (Loren Torveen et al. eds., 2002). For a technical overview about various wearable community projects, see Roy L. Ashok & Dharma P. Agrawal, Next-Generation Wearable Networks, Computer, Nov. 2003, at 31. The primary legal treatment of this subject is Jerry Kang & Dana Cuff, Pervasive Computing: Embedding the Public Sphere, 62 Wash. & Lee L. Rev. 93 (2005). 8. LuAnn LaSalle, Where Google Meets Facebook Meets GPS, Toronto Globe & Mail, Feb. 26, 2009, at L3. 9. Kang & Cuff, supra note 7, at 131. 10. Id. at 133–34; see also Mark A. Blythe et al., Little Brother: Could and Should Wearable Computing Technologies Be Applied to Reducing Older People’s Fear of Crime?, 8 Pers. & Ubiquitous Computing 402, 407–12 (2004) (describing the potential for wearable computing technologies to address the information problems that underlie urban crime). 11. The People’s Republic of China is beginning to use wearable computers and identity management systems to help the police keep track of the government’s contacts with its civilian population, as well as the citizenry’s movements through public spaces. Keith Bradsher, China Enacting High-Tech Plan to Track People, N.Y. Times, Aug. 12, 2007, at A1. Less disconcertingly, Maryland is developing the capacity for police officers to conduct criminal background checks via squad car–based terminals, as well as enabling biometric identification. Philip J. Weiser, Communicating During Emergencies: Toward Interoperability and Effective Information Management, 59 Fed. Comm. L.J. 547, 564 (2007). 12. See infra text accompanying note 15 (discussing the issue of discrimination in taxi cab pickups). 13. He might well decide to extend the circle of trust by one degree—trusting favorable feedback from any of his fifty cabbie friends or any of his fifty cabbie friends’ friends—a community of interest that could well include many of the cab drivers in town. 14. Ian Ayres et al., To Insure Prejudice: Racial Disparities in Taxicab Tipping, 114 Yale L.J. 1613, 1624 (2005). 15. Danita L. Davis, Note, Taxi! Why Hailing a New Idea About Public Accommodation Laws May Be Easier Than Hailing a Taxi, 37 Val. U. L. Rev. 929 (2003). 16. Photograph recognition software and other emerging technologies that are poised to reduce personal privacy are discussed in Jeffrey Rosen, The End of Forgetting, N.Y. Times Mag. 30 (July 25, 2010).

NOTES TO PAGES 133–136

17. See, e.g., Daniel J. Solove, The Digital Person: Technology and Privacy in the Information Age (2004); Daniel J. Solove, The Future of Reputation: Gossip, Rumor, and Privacy on the Internet (2007); Julie E. Cohen, Examined Lives: Information Privacy and the Subject as Object, 52 Stan. L. Rev. 1373 (2000); A. Michael Froomkin, The Death of Privacy?, 52 Stan. L. Rev. 1461 (2000); Oscar H. Gandy, Jr., Quixotics Unite! Engaging the Pragmatists on Rational Discrimination, in Theorizing Surveillance: The Panopticon and Beyond 318 (David Lyon ed., 2006); Stan Karas, Privacy, Identity, Databases, 52 Am. U. L. Rev. 393 (2002); Matthew J. Hodge, Comment, The Fourth Amendment and Privacy Issues on the “New” Internet: Facebook.com and MySpace.com, 31 S. Ill. U. L.J. 95 (2006). 18. Oscar H. Gandy, Jr., The Panoptic Sort: A Political Economy of Personal Information (1993). 19. See, e.g., Surveillance as Social Sorting: Privacy, Risk, and Digital Discrimination (David Lyon ed., 2003); Theorizing Surveillance, supra note 17. 20. Gandy, supra note 18, at 71–93; David Lyon, Surveillance as Social Sorting: Computer Codes and Mobile Bodies, in Surveillance as Social Sorting, supra note 19, at 13, 13–28. 21. Subsequent work by Gandy has focused on the disturbing cumulative effects of sorting and advanced the more compelling argument that reliance on race as a variable in decision-making algorithms that aim to predict future behavior cannot be justified in the absence of a causal story to explain observed correlations between race and behavior. Oscar H. Gandy, Jr., Coming to Terms with Chance: Engaging Rational Discrimination and Cumulative Disadvantage 69–71 (2009); Oscar H. Gandy, Jr., Engaging Rational Discrimination: Exploring Reasons for Placing Regulatory Constraints on Decision Support Systems, 12 Ethics & Information Technology 29, 34–36 (2010). 22. Reputational infrastructure consists of the technologies that enable information about individuals’ actions and reputations to circulate efficiently among members of society. 23. Jesse Dukeminier et al., Property 382 (6th ed. 2006) and sources cited therein. 24. See, e.g., Kramarsky v. Stahl Mgmt., 401 N.Y.S.2d 943, 944–45 (N.Y. Sup. Ct. 1977). 25. Id. at 945. 26. Dukeminier et al., supra note 23, at 382–83. 27. A 2006 New York Times article mentions continuing discrimination in New York City against “lawyers (too litigious) and musicians (too loud).” Teri Karush Rogers, Only the Strongest Survive, N.Y. Times, Nov. 26, 2006, § 11, at 1. 28. Robert C. Ellickson, Order Without Law: How Neighbors Settle Disputes 277 (1991); Mary B. Spector, Tenants’ Rights, Procedural Wrongs: The Summary Eviction and the Need for Reform, 46 Wayne L. Rev. 135, 181–82 (2000). 29. Rogers, supra note 27, at 1. 30. Id. 31. Id. 32. Id. 33. Rudy Kleysteuber, Tenant Screening Thirty Years Later: A Statutory Proposal to Protect Public Records, 116 Yale L.J. 1344, 1363–64 (2007).

233

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

234

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 137–141

34. Dukeminier et al., supra note 23, at 408; Kara B. Schissler, Note, Come and Knock on Our Door: The Fair Debt Collection Practices Act’s Intrusion into New York’s Summary Proceedings Law, 22 Cardozo L. Rev. 315, 328–29 (2000). 35. See, e.g., Duncan Kennedy, The Limited Equity Coop as a Vehicle for Affordable Housing in a Race and Class Divided Society, 46 How. L.J. 85, 104 (2002). 36. See, e.g., Berg v. Wiley, 264 N.W.2d 145, 149–50 (Minn. 1978). 37. Dukeminier et al., supra note 23, at 408. 38. See, e.g., James J. White, The Abolition of Self-Help Repossession: The Poor Pay Even More, 1973 Wis. L. Rev. 503, 522–24 (making this argument in the context of chattel property); Chi. Bd. of Realtors, Inc. v. City of Chicago, 819 F.2d 732, 741 (7th Cir. 1987) (Posner, J.) (suggesting that legal reforms designed to help tenants are often counterproductive for this reason). 39. Indeed, assuming the absence of repeat-player interactions among landlords, the landlord may have an incentive to suppress information about a self-help eviction in order to inflict an undesirable tenant on competitors. 40. See, e.g., Daphna Lewinsohn-Zamir, In Defense of Redistribution Through Private Law, 91 Minn. L. Rev. 326, 381–83 (2006). 41. 15 U.S.C. §§ 1681c(a)(2)–(5) (2006). 42. Id. at §1681c(b). 43. U.D. Registry, Inc. v. State of California, 40 Cal. Rptr. 2d 228, 233 (Cal. Ct. App. 1995), reaff’d, U.D. Registry, Inc. v. State of California, 50 Cal. Rptr. 3d 647, 667 (Cal. Ct. App. 2006). The latter U.D. Registry court nevertheless held that California’s similar “security freeze” legislation was not unconstitutional on its face. One basis for distinguishing the two cases was entirely unconvincing—the idea that preventing identity theft is a compelling government interest but that ensuring that tenants can exercise their rights under the law without fear of subsequent discrimination in the housing market is not a compelling interest. 50 Cal. Rptr. 3d at 666–67. 44. Ellickson, supra note 28, at 277. 45. Similar websites that deal with the purchase of real estate, like Curbed’s New York, San Francisco, and Los Angeles sites, are transforming the market for homes. Brad Reagan, The Dirt on the Neighbors, SmartMoney, Oct. 2007, at 110, 110–13. 46. Stewart Schwab, Is Statistical Discrimination Efficient?, 76 Am. Econ. Rev. 228 (1986); David A. Strauss, The Law and Economics of Racial Discrimination in Employment: The Case for Numerical Standards, 79 Geo. L.J. 1619, 1622–24, 1626–30, 1639–43 (1991). Richard Epstein analyzes statistical discrimination as a predictable response to information asymmetries in the labor market. Richard A. Epstein, Forbidden Grounds: The Case Against Employment Discrimination Laws 32–40 (1992). 47. Jodi Kantor, Nanny Hunt Can Be a “Slap in the Face” for Blacks, N.Y. Times, Dec. 26, 2006, at A1. 48. Id. 49. On taxi tipping, see Ayres et al., supra note 14, at 1648–53. 50. Michael Lynn, Ethnic Differences in Tipping: A Matter of Familiarity with Tipping Norms, 45 Cornell Hotel & Restaurant Admin. Q. 12 (2004).

NOTES TO PAGES 141–142

51. Ana I. Balsa et al., Testing for Statistical Discrimination in Health Care, 40 Health Serv. Res. 227, 247 (2005) (finding evidence consistent with statistical discrimination in diagnosing hypertension and diabetes). 52. My characterization of implicit bias as a form of animus-based discrimination is contested. For a useful overview, see Kristin A. Lane, Jerry Kang & Mahzarin R. Banaji, Implicit Social Cognition and Law, 3 Ann. Rev. L. & Soc. Sci. 427 (2007). 53. Christine Jolls, Antidiscrimination Law’s Effects on Implicit Bias, in 3 NYU Selected Essays on Labor and Employment Law: Behavioral Analyses of Workplace Discrimination 69, 69–70 (Mitu Gulati & Michael J. Yelnosky eds., 2007). For a discussion of the decline of overt discrimination and the rise of statistical discrimination and implicit bias, see generally Rachel F. Moran, Whatever Happened to Racism?, 79 St. John’s L. Rev. 899 (2005). 54. Kathleen Daly & Michael Tonry, Gender, Race, and Sentencing, 22 Crime & Just. 201, 201–03 (1997); see also Becky Pettit & Bruce Western, Mass Imprisonment and the Life Course: Race and Class Inequality in U.S. Incarceration, 69 Am. Soc. Rev. 151, 156 (2004) (noting that African American males face a lifetime risk of incarceration of 28.5 percent versus 4.4 percent for Caucasian males). 55. Strauss, supra note 46, at 1626–29. 56. Ariz. Governing Comm. for Tax Deferred Annuity & Deferred Comp. Plans v. Norris, 463 U.S. 1073, 1085 n.15 (1983) (“For some jobs, however, there may be relevant skills that cannot be identified by testing. Yet Title VII clearly would not permit use of race, national origin, sex, or religion as a proxy for such an employment qualification, regardless of whether a statistical correlation could be established.”); Strauss, supra note 46, at 1623. 57. See, e.g., Robert A. Kearney, The Coming Rise of Disparate Impact Theory, 110 Penn St. L. Rev. 69, 86 (2005); Stephen L. Ross & John Yinger, Uncovering Discrimination: A Comparison of the Methods Used by Scholars and Civil Rights Enforcement Officials, 8 Am. L. & Econ. Rev. 562 (2006); Michael Selmi, Public vs. Private Enforcement of Civil Rights: The Case of Housing and Employment, 45 UCLA L. Rev. 1401 (1998). 58. See, e.g., Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618 (2007) (construing the time limit for filing employment discrimination claims with the EEOC narrowly), overruled by Congress in the Lilly Ledbetter Fair Pay Restoration Act of 2009; Brown v. City of Oneonta, 221 F.3d 329, 338–39 (2d Cir. 2000) (holding that despite a potential disparate impact, law enforcement stopping suspects on the basis of gender and race did not violate the Equal Protection Clause absent evidence of a discriminatory racial animus). 59. Harry J. Holzer et al., Perceived Criminality, Criminal Background Checks, and the Racial Hiring Practices of Employers, 49 J.L. & Econ. 451 (2006). 60. Id. at 451. 61. Id.; see also Devah Pager, The Mark of a Criminal Record, 108 Am. J. Soc. 937, 938 (2003) (noting that approximately 8 percent of the working-aged population of the United States are ex-felons).

235

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

236

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 142–145

62. Holzer et al., supra note 59, at 464. 63. Id. at 473. One of the authors of this study hedged this conclusion somewhat, noting that the desirability of promoting access to criminal records would depend on the amount of time for which a prior conviction would act as a reliable proxy for future conduct and the degree of nuance in employer reactions to prior convictions. Steven Raphael, Should Criminal History Records Be Universally Available?, 5 Criminology & Pub. Pol’y 515, 516–17, 519–20 (2006). 64. Keith Finlay, Effect of Employer Access to Criminal History Data on the Labor Market Outcomes of Ex-Offenders and Non-Offenders, in Studies of Labor Market Intermediation 89 (David H. Autor, ed. 2009) 65. See, e.g., David H. Autor & David Scarborough, Does Job Testing Harm Minority Workers? Evidence from Retail Establishments, 123 Q.J. Econ. 219 (2008); Shawn D. Bushway, Labor Market Effects of Permitting Employer Access to Criminal History Records, 20 J. Contemp. Crim. Just. 276 (2004); Pager, supra note 61, at 937–38. 66. Joleen Kirschenman & Kathryn M. Neckerman, “We’d Love to Hire Them, but . . .”: The Meaning of Race for Employers, in The Urban Underclass 203, 231 (Christopher Jencks & Paul E. Peterson eds., 1991). 67. Holzer et al., supra note 59, at 453. 68. The pathology described in the text is not the only rational but socially undesirable employer reaction to asymmetric information. Faced with unreliable information about criminal history, an employer may lean more heavily on word of mouth from existing social networks in its efforts to hire trustworthy employees. See generally George J. Stigler, An Introduction to Privacy in Economics and Politics, 9 J. Legal Stud. 623, 632–33 (1980) (“The more costly the acquisition of knowledge, the more expensive it becomes to enter into transactions with new parties. We should expect less mobility of laborers, creditors, etc. . . .”). Because the social networks used for job search purposes tend to be racially segregated, see Antoni Calvó-Armengol & Matthew O. Jackson, The Effects of Social Networks on Employment and Inequality, 94 Am. Econ. Rev. 426 (2004), increased reliance on word of mouth by firms that presently employ few minorities will increase the likelihood that minorities will continue to be underrepresented in that workplace in the future. 69. See, e.g., Robert Post, Prejudicial Appearances: The Logic of American Antidiscrimination Law, 88 Cal. L. Rev. 1, 14–16 (2000) (discussing orchestral auditions, in which a screen separates the judges from the auditioning musician). 70. Edward S. Adams, Using Evaluations to Break Down the Male Corporate Hierarchy: A Full Circle Approach, 73 U. Colo. L. Rev. 117, 117–18 (2002). 71. See, e.g., Briscoe v. Reader’s Digest Ass’n, 483 P.2d 34 (Cal. 1971); Melvin v. Reid, 297 P. 91 (Cal. Dist. Ct. App. 1931). 72. See, e.g., U.S. Dep’t of Justice v. Reporters Comm. for Freedom of the Press, 489 U.S. 749 (1989) (holding that criminal rap sheets are exempt from disclosure under FOIA because their dissemination would constitute an unwarranted invasion of personal privacy).

NOTES TO PAGES 146–148

73. Even after the Supreme Court’s opinion in Batson v. Kentucky, 476 U.S. 79 (1986), made it clear that race could not provide a basis for the use of peremptory strikes, some attorneys continued to do so. Upon opposing counsel’s raising of a Batson challenge, these attorneys would admit to having stricken prospective jurors on the basis of race, provide no explanation for their use of peremptory strikes, or claim to be striking jurors on the basis of a purportedly non-race-based criteria, like “membership in the NAACP” or “people likely to be offended by racist jokes contained in the evidence.” Kenneth J. Melilli, Batson in Practice: What We Have Learned About Batson and Peremptory Challenges, 71 Notre Dame L. Rev. 447, 480–83 (1996). 74. Herald P. Fahringer, In the Valley of the Blind: A Primer on Jury Selection in a Criminal Case, 43 Law & Contemp. Probs. 116, 117–18 (1980). 75. Jonathan M. Redgrave & Jason J. Stover, The Information Age, Part II: Juror Investigation on the Internet—Implications for the Trial Lawyer, 2 Sedona Conf. J. 211, 216 (2001). 76. Solomon M. Fulero & Steven D. Penrod, The Myths and Realities of Attorney Jury Selection Folklore and Scientific Jury Selection: What Works?, 17 Ohio N.U. L. Rev. 229, 229–38 (1990); Paul V. Olczak et al., Attorneys’ Lay Psychology and Its Effectiveness in Selecting Jurors: Three Empirical Studies, 6 J. Soc. Behav. & Personality 431, 443–44, 447–48 (1991). 77. Neil J. Kressel & Dorit F. Kressel, Stack and Sway: The New Science of Jury Consulting (2002). 78. Franklin Strier & Donna Shestowsky, Profiling the Profilers: A Study of the Trial Consulting Profession, Its Impact on Trial Justice and What, If Anything, to Do About It, 1999 Wis. L. Rev. 441, 466–67; see also Fahringer, supra note 74, at 127 (“Jurors usually stand out because of their apparent unfamiliarity with the courthouse. Watch what they do and try to remember their faces. Note what they say in the lobby of the courthouse, how they behave, what newspapers they buy; these activities will reveal in a small way what they are like. When the jury panel is brought into the courtroom, study them. Search for clues that may be helpful in making important choices. A Wall Street Journal tucked under an arm, a sexy paperback book protruding from a purse, or a best-seller in a juror’s hands can be meaningful.”). 79. Redgrave & Stover, supra note 75, at 219. 80. Id. at 213. 81. Roger Allan Ford, Modeling the Effects of Peremptory Challenges on Jury Selection and Jury Verdicts, 17 Geo. Mason L. Rev. 377 (2010). 82. Henry Gottlieb, Should You Design Your Firm’s Web Site with Jurors in Mind?, Law.com, Jan. 2, 2007, http://www.law.com/jsp/article.jsp?id=1167386817011. 83. Del Quentin Wilber, Social Networking Among Jurors Is Trying Judges’ Patience, Wash. Post., Jan. 9, 2010. 84. A divided panel of the Seventh Circuit affirmed the conviction. See United States v. Warner, 498 F.3d 666 (7th Cir. 2007). 85. Matt O’Connor, Jury Pools Can Face Probes in Sensitive Trials: Ryan Case Sparks Move to Avoid Bias, Scandal, Chi. Trib., Dec. 11, 2006, § 1, at 1.

237

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

238

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 148–150

86. Greg Hinz & Steven R. Strahler, Saving Client Ryan: A $10-mil. Mission, Crain’s Chi. Bus., Nov. 14, 2005, at 1. 87. Batson v. Kentucky, 476 U.S. 79 (1986). 88. It does. J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127, 146 (1994). 89. They do. Georgia v. McCollum, 505 U.S. 42 (1992) (holding that Batson prohibits the defendant from using race as a basis for peremptory challenges); Edmonson v. Leesville Concrete Co., 500 U.S. 614 (1991) (extending Batson to civil cases). 90. J.E.B., 511 U.S. at 140 (“Discrimination in jury selection, whether based on race or on gender, causes harm to the litigants, the community, and the individual jurors who are wrongfully excluded from participation in the judicial process.”). 91. The Supreme Court, 2004 Term—Leading Cases, 119 Harv. L. Rev. 169, 224 n.45 (2005). 92. Chris F. Denove & Edward J. Imwinkelried, Jury Selection: An Empirical Investigation of Demographic Bias, 19 Am. J. Trial Advoc. 285, 298 (1995); Melilli, supra note 73, at 503. 93. Snyder v. Louisiana, 552 U.S. 472 (2008), suggests that courts will scrutinize lawyers’ reasoning more closely in the coming years. 94. Melilli, supra note 73, at 483. 95. Samuel R. Sommers & Michael I. Norton, Race-Based Judgments, RaceNeutral Justifications: Experimental Examination of Peremptory Use and the Batson Challenge Procedure, 31 Law & Hum. Behav. 261, 269 (2007) (“College students, law students, and attorneys playing the role of a prosecutor trying a case with a Black defendant were more likely to challenge a prospective juror when he was Black as opposed to White. . . . When justifying these judgments, participants rarely cited race as influential, focusing instead on the race-neutral characteristics associated with the Black prospective juror. That is, when Juror #1 was Black, participants tended to justify their judgments by citing his familiarity with police misconduct as their reason for excluding him. When Juror #2 was Black, on the other hand, participants reported his skepticism about statistics to be more important than the police misconduct issue.”). 96. Another alternative would be to eliminate the use of peremptory challenges altogether. Miller-El v. Dretke, 545 U.S. 231, 266–73 (2005) (Breyer, J., concurring) (concluding that the Batson framework cannot stamp out discrimination in jury selection and that peremptory challenges therefore should be abolished). 97. Theodore Eisenberg et al., Forecasting Life and Death: Juror Race, Religion, and Attitude Toward the Death Penalty, 30 J. Legal Stud. 277, 285–87 (2001). Race likely functions as a proxy here, too, for an omitted variable like differential life experiences or interactions with law enforcement. 98. William J Bowers et al., Death Sentencing in Black and White: An Empirical Analysis of the Role of Jurors’ Race and Jury Racial Composition, 3 U. Pa. J. Const. L. 171, 241–44 (2001). 99. Theodore Eisenberg, Death Sentence Rates and County Demographics: An Empirical Study, 90 Cornell L. Rev. 347, 367–68 (2005).

NOTES TO PAGES 150–154

100. Denove & Imwinkelried, supra note 92, at 293–95 (“Race emerges from the data as the single most important factor in predicting juror orientation. The impact of race is so strong that it often outweighs the impact of all other demographic factors combined.”). 101. Id. at 293–95, 305–06, 313. 102. See generally. Fahringer, supra note 74, at 119 (“[B]lacks who have seen the ravages of the drug trade in their community must be considered carefully in a narcotics prosecution. . . .”). 103. Ellen S. Cohn et al., Reducing White Juror Bias: The Role of Race Salience and Racial Attitudes, 39 J. Applied Soc. Psych. 1953, 1966 (2009). 104. See id. and sources cited therein. 105. United States v. Blagojevich, 614 F.3d 287, 293 (7th Cir. 2010) (Posner, J., dissenting from denial of rehearing en banc). 106. Id. 107. See generally James Q. Whitman, The Two Western Cultures of Privacy: Dignity Versus Liberty, 113 Yale L.J. 1151 (2004). 108. This assumes, of course, that it is not only those with “something to hide” who value their privacy rights. See generally Daniel J. Solove, “I’ve Got Nothing to Hide” and Other Misunderstandings of Privacy, 44 San Diego L. Rev. 745 (2007). 109. Dana A. Forgione et al., Diversion of Prescription Drugs to the Black Market: What the States Are Doing to Curb the Tide, J. Health Care Fin., Summer 2001, at 65; Stephen J. Ziegler & Nicholas P. Lovrich, Jr., Pain Relief, Prescription Drugs, and Prosecution: A Four-State Survey of Chief Prosecutors, 31 J.L. Med. & Ethics 75, 78 (2003). 110. Sharon M. Weinstein et al., Physicians’ Attitudes Toward Pain and the Use of Opioid Analgesics: Results of a Survey from the Texas Cancer Pain Initiative, 93 S. Med. J. 479, 481 tbl.3, 482–83 (2000). 111. Id. at 481 tbl.3. 112. Ziegler & Lovrich, supra note 109, at 91. 113. The Supreme Court upheld the constitutionality of this practice in Whalen v. Roe, 429 U.S. 589 (1977). 114. See, e.g., Tina Rosenberg, Doctor or Drug Pusher, N.Y. Times Mag., June 17, 2007, at 48; John Tierney, Juggling Figures, and Justice, in a Doctor’s Trial, N.Y. Times, July 3, 2007, at F1. 115. The researchers used the standard NRS-11 scale for assessing pain intensity. Ian Chen et al., Racial Differences in Opioid Use for Chronic Nonmalignant Pain, 20 J. Gen. Internal Med. 593, 594 (2005). 116. Id. at 597. The researchers note their inability to determine whether African Americans’ pain is being undertreated or Caucasians’ pain is being overtreated. Id. The medical literature suggests, however, that pain undertreatment, not overtreatment, is the cause for concern in the United States. See, e.g., David B. Resnik et al., The Undertreatment of Pain: Scientific, Clinical, Cultural, and Philosophical Factors, 4 Med. Health Care & Phil. 277 (2001). 117. Ana I. Balsa & Thomas G. McGuire, Prejudice, Clinical Uncertainty and Stereotyping as Sources of Health Disparities, 22 J. Health Econ. 89, 103–04 (2003);

239

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

240

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 154–159

Darrell J. Gaskin et al., Exploring Racial and Ethnic Disparities in Prescription Drug Spending and Use Among Medicare Beneficiaries, 4 Am. J. Geriatric Pharmacotherapy 96, 103 (2006). 118. Michelle van Ryn & Jane Burke, The Effect of Patient Race and Socio-Economic Status on Physicians’ Perceptions of Patients, 50 Soc. Sci. & Med. 813, 822–24 (2000). 119. Rachel L. Johnson et al., Racial and Ethnic Differences in Patient Perceptions of Bias and Cultural Competence in Health Care, 19 J. Gen. Intern. Med. 101, 106–08 (2004). 120. Susan L. Furth et al., Effects of Patient Compliance, Parental Education and Race on Nephrologists’ Recommendations for Kidney Transplantation in Children, 3 Am. J. Transplantation 28, 30–32 (2003). 121. See, e.g., Balsa & McGuire, supra note 117, at 112 (“Rule-based policies, such as requiring doctors to act in a race-blind fashion, when race conveys information that can help the doctor decide what to do for the patient, can make matters worse for the patients the policy is designed to help.”). 122. See supra text accompanying notes 57–58 (addressing the shortcomings of litigation as a remedy for discriminatory hiring practices). 123. Balsa et al., supra note 51, at 247–48. 124. See generally David W. Bates et al., A Proposal for Electronic Medical Records in U.S. Primary Care, 10 J. Am. Med. Informatics Ass’n 1, 4 (2003). 125. Thomas Bodenheimer & Kevin Grumbach, Electronic Technology: A Spark to Revitalize Primary Care?, 290 JAMA 259, 260 (2003). 126. See, e.g., Bates et al., supra note 124, at 4; Bodenheimer & Grumbach, supra note 125, at 260–61; Ross D. Fletcher et al., Computerized Medical Records in the Department of Veterans Affairs, 91 Cancer 1603, 1606 (2001).

9.  Carrots, Sticks, Curtains, and Searchlights 1. Archon Fung et al., Full Disclosure: The Perils and Promise of Transparency 82–83 (2007). 2. Diane Cardwell, Restaurants Complain Health Department Web Site Misleads the Public, N.Y. Times, Aug. 3, 2010, at A17. 3. Rose Corrigan, Making Meaning of Megan’s Law, 31 Law & Soc. Inquiry 267, 296–301 (2006). 4. James D. Cox et al., Public and Private Enforcement of the Securities Laws: Have Things Changed Since Enron?, 80 Notre Dame L. Rev. 893, 893–94 (2005). 5. Paul M. Schwartz & Edward J. Janger, Notification of Data Security Breaches, 105 Mich. L. Rev. 913 (2007). 6. Bradley C. Karkkainen, Information as Environmental Regulation: TRI and Performance Benchmarking, Precursor to a New Paradigm?, 89 Geo. L.J. 257 (2001). 7. On the latter, see Jerry Brito, Hack, Mash, & Peer: Crowdsourcing Government Transparency, 9 Colum. Sci. & Tech. L. Rev. 119, 122–29 (2008), http://www.stlr .org/html/volume9/brito.pdf.

NOTES TO PAGES 160–165

8. This assumes, of course, that the prosecutor genuinely believes that the defendant is guilty for reasons having nothing to do with the defendant’s race. 9. See supra text accompanying chapter eight, notes 90–91. 10. This is a recurring problem with the Transportation Security Administration’s no-fly and restricted lists. Leslie Miller, Thousands Wrongly Named to Terrorist List, Chi. Sun-Times, Oct. 7, 2006, at 12; see also Rudy Kleysteuber, Tenant Screening Thirty Years Later: A Statutory Proposal to Protect Public Records, 116 Yale L.J. 1344, 1358–59 (2007) (discussing this problem in the context of tenant information databases). 11. On implicit bias, see, for example, Anthony G. Greenwald & Linda Hamilton Krieger, Implicit Bias: Scientific Foundations, 94 Cal. L. Rev. 945 (2006); Christine Jolls & Cass Sunstein, The Law of Implicit Bias, 94 Cal. L. Rev. 969 (2006); Jerry Kang & Mahzarin R. Banaji, Fair Measures: A Behavioral Realist Revision of “Affirmative Action,” 94 Cal. L. Rev. 1063 (2006). 12. Lior Jacob Strahilevitz, “How’s My Driving?” for Everyone (and Everything?), 81 N.Y. U. L. Rev. 1699, 1733–34 (2006). 13. Chrysanthos Dellarocas, Immunizing Online Reputation Reporting Systems Against Unfair Ratings and Discriminatory Behavior, 2000 Proc. 2nd ACM Conf. on Electronic Com. 150, 153. 14. Id. at 155. 15. Sonja Buchegger & Jean-Yves Le Boudec, The Effect of Rumor Spreading in Reputation Systems for Multiple Ad-Hoc Networks § 7 (March 2003) (unpublished manuscript, available at http://icapeople.epfl.ch/sbuchegg/bucheggerL03A.pdf). 16. See, e.g., Jack Glaser & Eric D. Knowles, Implicit Motivation to Control Prejudice, 44 J. Experimental Soc. Psychol. 164, 167–71 (2008); Laurie A. Rudman et al., “Unlearning” Automatic Biases: The Malleability of Implicit Prejudice and Stereotypes, 81 J. Personality & Soc. Psychol. 856 (2001). 17. See generally Gordon B. Moskowitz et al., Preconscious Control of Stereotype Activation Through Chronic Egalitarian Goals, 77 J. Personality & Soc. Psychol. 167, 181–82 (1999) (discussing various strategies for combating stereotype-based subconscious practices). 18. Margo J. Monteith et al., Putting the Brakes on Prejudice: On the Development and Operation of Cues for Control, 83 J. Personality & Soc. Psychol. 1029, 1045–46 (2002). 19. Moskowitz et al., supra note 17, at 180. 20. Monteith et al., supra note 18, at 1045–46. 21. Strahilevitz, supra note 12, at 1754–55. The Pollyanna effect prompted eBay to alter its rules so as to prevent sellers from leaving negative or neutral feedback about buyers. As eBay explained: “The previous system prevented buyers from leaving honest Feedback since they feared retaliation from the sellers if they left a negative. This made it harder for buyers to distinguish between sellers while making bidding or buying decisions.” eBay, Upcoming Changes to Feedback, http://pages.ebay.com/?services/forum/new .html. Shortly afterwards, individuals who opposed the feedback policy change launched their own non-eBay website, whereby sellers could leave negative feedback about their transaction partners. eBuyer Feedback, http://www.ebuyer-feedback.com/ (“Since May

241

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

242

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 165–169

19th, eBay no longer gives its sellers the ability to leave negative or neutral feedback for their customers. Only positive feedback is allowed, ‘at the seller’s option.’ Even though the majority of buyers are honest this new policy leaves sellers vulnerable to new scams, feedback extortion, unpaid items, etc. with no way to defend themselves or warn others. This is not acceptable. This site gives sellers the ability to leave their eBay buyers positive, negative, or neutral feedback.”). 22. Yoram Barzel, Measurement Cost and the Organization of Markets, 25 J.L. & Econ. 27, 42–46 (1982). 23. This is a concern identified in Richard A. Posner, Law, Pragmatism, and Democracy 284–85 (2003). 24. Robert J. Youmans & Benjamin D. Jee, Fudging the Numbers: Distributing Chocolate Influences Student Evaluations of an Undergraduate Course, 34 Teaching Psychol. 245 (2007). 25. Christopher Elliott, Hotel Reviews Online: In Bed with Hope, Half-Truths and Hype, N.Y. Times, Feb. 7, 2006, at C1. 26. See George J. Stigler, An Introduction to Privacy in Economics and Politics, 9 J. Legal Stud. 627 (1980) 27. Robert C. Ellickson, Order Without Law: How Neighbors Settle Disputes (1991). 28. Strahilevitz, supra note 12, at 1705–08. 29. Julie E. Cohen, Examined Lives: Information Privacy and the Subject as Object, 52 Stan. L. Rev. 1373 (2000). 30. Douglas Jehl, The King and the “Cabby” Inspect Jordan Incognito, N.Y. Times, Aug. 9, 1999, at A1. 31. Id. 32. See, e.g., Mary Mitchell, Would Cop Have Pulled a Gun on a White Minister?, Chi. Sun-Times, July 17, 2005, at 14 (discussing the controversy that erupted when a police officer shouted profanity and pointed a gun at James Meeks, a state senator and prominent Chicago religious leader). 33. Mary Mitchell, In Paris, Not Even Oprah Can Escape Reality of Being Black, Chi. Sun-Times, July 3, 2005, at 14. 34. Don Van Natta Jr. & Abby Goodnough, After Call to Police, 2 Worlds Collide in Unlikely Cambridge Meeting, N.Y. Times, July 27, 2009, at A13. 35. It might be that by making these mistakes by the police more rare, you create an even stronger public backlash when mistakes do happen. 36. In many of these settings, government publication of the pertinent information is quite limited. Take criminal histories as an example. In some states, all registered sex offenders have their address information published, but other states only publish the information of those sex offenders who are deemed to present the highest risk of recidivism or who have committed the most serious crimes. Christina Locke & Bill F. Chamberlin, Safe from Sex Offenders? Legislating Internet Publication of Sex Offender Registries, 39 Urb. Law. 1, 15–16 (2007). Only a handful of state governments publish Megan’s Law–style databases for individuals committed of nonsex crimes (see James B. Jacobs, Mass Incarceration and the Proliferation of Criminal Records, 3 U. St. Thomas

NOTES TO PAGES 169–172

L.J. 387, 399–400 [2006]), and such criminal history information may be immune from Freedom of Information Act requests (see U.S. Dep’t of Justice v. Reporters Comm. for Freedom of the Press, 489 U.S. 749 [1989]). 37. The rise of these private databases containing information culled from government records has made the pursuit of a curtains strategy more difficult. When the government no longer controls the databases, and where the contents of databases are widely mirrored, it is quite difficult for the government to erase or redact previously released information from the public domain effectively. 38. Adam Liptak, On Second Thought, Let’s Just Rate All the Lawyers, N.Y. Times, July 2, 2007, at A9. 39. Complaint, Browne v. Avvo, Inc., 525 F. Supp. 2d 1249 (W.D. Wash. 2007) (No. CV7–920 RSL), available at http://blog.seattlepi.nwsource.com/venture/library/ Avvo_Complaint_FINAL_secured1.pdf. 40. Id. at ¶¶ 46, 70–71. 41. John Cook, Respected Lawyer Wants Rating Site Avvo Closed, Seattle PostIntelligencer, June 12, 2007, at D1. 42. Defendants’ Motion to Dismiss Class Action Complaint Under Fed. R. Civ. P. 12(c), Avvo, 525 F. Supp. 2d 1249 (No. CV7–920 RSL). 43. Avvo, 525 F. Supp. 2d 1249. 44. For a discussion of right of reply statutes in the defamation context, see Jerome A. Barron, The Right of Reply to the Media in the United States—Resistance and Resurgence, 15 Hastings Comm. & Ent. L.J. 1 (1992). 45. Alf Lizzio et al., Managers’ and Subordinates’ Evaluations of Feedback Strategies: The Critical Contribution of Voice, 38 J. Applied Soc. Psychol. 919, 938–40 (2008). 46. A right of reply could exist in other settings, too. Systems that disclose an individual’s criminal conviction might allow the criminal himself to explain his life circumstances leading up to the convictions, or profess his innocence. It is likely that most of these “replies” would fall on deaf ears given the strength of existing biases against ex-offenders. But to the extent that some information intermediaries will be able to profit by helping employers sort between those ex-offenders who pose high recidivism risks and those who do not, the contents of these replies might turn out to provide a useful variable to permit effective sorting. 47. 47 U.S.C. § 230 (2000); see also Zeran v. Am. Online, Inc., 129 F.3d 327, 328 (4th Cir. 1997) (finding that America Online was entitled to absolute immunity against defamation liability under the Communications Decency Act). 48. For an exploration of related questions, see Chris Dent & Andrew T. Kenyon, Defamation Law’s Chilling Effect: A Comparative Content Analysis of Australian and US Newspapers, 9 Media & Arts L. Rev. 89 (2004). 49. See generally Alex Taylor III, No Test Dummies, Fortune, June 11, 2007, at 49 (noting that Consumer Reports has 4.3 million subscribers in print and another 2.7 million online). 50. Jonathan Zittrain, The Future of the Internet—and How to Stop It 206 (2008).

243

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

244

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 174–178

10.  Winners and Losers 1. Elena Saxonhouse, Note, Unequal Protection: Comparing Former Felons’ Challenges to Disenfranchisement and Employment Discrimination, 56 Stan. L. Rev. 1597, 1637 (2004). 2. Bragdon v. Abbott, 524 U.S. 624, 641–42 (1998). 3. Centers for Disease Control and Prevention, U.S. Dep’t of Health & Human Servs., HIV/AIDS Surveillance Report: Cases of HIV Infection and AIDS in the United States and Dependent Areas, 2005, at 10 (2007), available at http://www.cdc.gov/hiv/topics/?surveillance/resources/reports/2005report/ pdf/2005SurveillanceReport.pdf. 4. Bruce Elmslie & Edinaldo Tebaldi, Sexual Orientation and Labor Market Discrimination, 28 J. Lab. Res. 436, 442 (2007). 5. Id. at 442–43 (reviewing the literature). The authors’ regression analysis casts some doubt on the statistical discrimination hypothesis, thanks to a finding of greater discrimination against gay men in blue-collar industries than in service jobs. Id. at 449–50. They wonder—quite provocatively—what would lead blue-collar employers but not white-collar employers to engage in statistical discrimination, and see employerbased animus as a more likely explanation for the diminished prospects of homosexual men in blue-collar jobs. 6. See, e.g., Multimedia WMAZ, Inc. v. Kubach, 443 S.E.2d 491 (Ga. Ct. App. 1994); Hillman v. Columbia County, 474 N.W.2d 913, 922 (Wis. Ct. App. 1991). 7. Harry J. Holzer et al., Perceived Criminality, Criminal Background Checks, and the Racial Hiring Practices of Employers, 49 J.L. & Econ. 451, 473 (2006). 8. David A. Strauss, The Law and Economics of Racial Discrimination in Employment: The Case for Numerical Standards, 79 Geo. L.J. 1619, 1641 (1991). 9. See, e.g., Briscoe v. Reader’s Digest Ass’n, 483 P.2d 34, 41 (Cal. 1971); Melvin v. Reid, 297 P. 91, 93 (Cal. Dist. Ct. App. 1931). 10. See, e.g., James B. Jacobs, Mass Incarceration and the Proliferation of Criminal Records, 3 U. St. Thomas L.J. 387, 406–12 (2006); Megan C. Kurlychek et al., Scarlet Letters and Recidivism: Does an Old Criminal Record Predict Future Offending?, 5 Criminology & Pub. Pol’y 483, 498–99 (2006). 11. For an analysis of the distributive dimensions of antidiscrimination policy, see Mark Kelman, Concepts of Discrimination in “General Ability” Job Testing, 104 Harv. L. Rev. 1157, 1182–98 (1991). 12. George K. Yin et al., Improving the Delivery of Benefits to the Working Poor: Proposals to Reform the Earned Income Tax Credit, 11 Am. J. Tax Pol’y 225, 291–92 (1994). 13. 26 U.S.C. § 51(a), (d)(1)(C) (2000). For an overview of the history of tax credits and other subsidies, see Timothy J. Bartick, Jobs for the Poor: Can Labor Demand Policies Help? 207–16 (2001). 14. It is more difficult to identify those ex-convicts who would have been hired despite the absence of a subsidy, however, which creates some inefficiency in tax programs. Contemporary economic scholarship suggests that the resulting inefficiency is manageable

NOTES TO PAGES 178–181

and that employment subsidies do result in many people obtaining employment who would not otherwise have gotten hired. Aki Kangasharju, Do Wage Subsidies Increase Employment in Subsidized Firms?, 74 Economica 51, 52, 63–64 (2007). 15. Delancey Street Foundation, Our Story, http://www.delanceystreetfoundation .org/?ourstory.php. 16. Delancey Street Foundation, Our Enterprises, http://www.delanceystreet­ foundation.org/?enterprises.php. 17. For a profile, see Andrew Friedman, The Prison That Thinks It’s a Kibbutz, Jerusalem Rep., Jan. 14, 2002, at 32, available at http://www.delanceystreetfoundation.org/ pdf/2002/jeru_011402.pdf. 18. Id. 19. Rummana Hussain, Beef Is on with Felony Franks: Hot Dog Joint’s Play off Prison Theme “Not That Cute,” 2nd Ward Alderman Says, Chi. Sun-Times, April 25, 2009, at 4. 20. John Rawls, A Theory of Justice 100–02 (1971) (discussing the difference principle). 21. The question of whether there ought to be reputational affirmative action is a fascinating one that I do not take up here. Jonathan Zittrain’s fascinating book contains a provocative but brief discussion of reputational bankruptcy. Jonathan Zittrain, The Future of the Internet—and How to Stop It 216–17 (2008). Both reputational affirmative action and reputational bankruptcy have a downside, which is that by decreasing the reliability of the personal information available to people, they encourage reliance on other strategies for sorting, such as statistical discrimination. 22. In practice, it does not always work out that way. African American ex-offenders may face a greater incremental harm from involvement in the criminal justice system than Caucasian ex-offenders. Devah Pager, The Mark of a Criminal Record, 108 Am. J. Soc. 937 (2003). 23. Lawrence Lessig, The Regulation of Social Meaning, 62 U. Chi. L. Rev. 965–67 (1995). 24. Gary Burtless, Are Targeted Wage Subsidies Harmful? Evidence from a Wage Voucher Experiment, 39 Indus. & Lab. Rel. Rev. 105, 112 (1985). 25. Anne L. Alstott, Work vs. Freedom: A Liberal Challenge to Employment Subsidies, 108 Yale L.J. 967, 1033–45 (1999). 26. Bartick, supra note 13, at 220–24. 27. Id. at 226–48; Michael Gerfin et al., Does Subsidised Temporary Employment Get the Unemployed Back to Work? An Econometric Analysis of Two Different Schemes, 12 Labour Econ. 807, 820–23 (2005); Lawrence F. Katz, Wage Subsidies for the Disadvantaged, in Generating Jobs: How to Increase Demand for Less-Skilled Workers 21, 40–42, 46–49 (Richard B. Freeman & Peter Gottschalk eds., 1998). 28. See generally Stephen Morris & Hyun Song Shin, Social Value of Public Information, 92 Am. Econ. Rev. 1521, 1522, 1529–32 (2002) (suggesting that under certain circumstances, revealing public information may be undesirable because it causes people to underweight private information they already possess).

245

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

246

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 181–185

29. Dana Cuff et al., Urban Sensing: Out of the Woods, Comm. ACM, Mar. 2008, at 33. 30. Kathleen Carroll & Dennis Coates, Teaching Price Discrimination: Some Clarification, 66 S. Econ. J. 466, 467–68 (1999). Carroll and Coates also note that price discrimination includes “the practice of a firm charging the same price for all units of the same good when there are cost variations in supply.” Id. at 468. 31. Anita Ramasastry, Web Sites Change Prices Based on Customers’ Habits, CNN .com, June 24, 2005, http://www.cnn.com/2005/LAW/06/24/ramasastry.website .prices/. 32. William W. Fisher III, When Should We Permit Differential Pricing of Information?, 55 UCLA L. Rev. 1, 11–12 (2007). 33. Carroll & Coates, supra note 30, at 469–70. 34. Id. at 468–69. 35. Andrew Odlyzko, Privacy, Economics, and Price Discrimination on the Internet, in Economics of Information Security 187, 189–93, 207–08 (L. Jean Camp & Stephen Lewis eds., 2004). 36. Carroll & Coates, supra note 30, at 470–471. 37. Matthew A. Edwards, Price and Prejudice: The Case Against Consumer Equality in the Information Age, 10 Lewis & Clark L. Rev. 559, 575–83 (2006). 38. Carroll & Coates, supra note 30, at 472. 39. Id.; Fisher, supra note 32, at 22. 40. Fisher, supra note 32, at 25. 41. Id. at 27. 42. Matt Richtel, “Face Time,” with a Twist, N.Y. Times, Oct. 26, 2000, at G7. 43. Id. 44. Hot or Not, http://www.hotornot.com. 45. Melanie McFarland, “‘American Idol’” Juggernaut Is No Better—and No Less Popular, Seattle Post-Intelligencer, Jan. 15, 2008, at C1. 46. Alan Schwarz & William K. Rashbaum, Referee Is the Focus of a Federal Inquiry, N.Y. Times, July 21, 2007, at D1. 47. Jessica Guynn, People-Rating Site a Danger, Critics Warn, L.A. Times, April 2, 2010, at 1. 48. Lior Jacob Strahilevitz, Reputation Nation: Law in an Era of Ubiquitous Personal Information, 102 Northwestern U. L. Rev. 1667 (2008). 49. See Gary Shteyngart, Super Sad True Love Story (2010). 50. Id. at 88–93. 51. Id. at 54. 52. Id. at 34–35, 40–43. 53. Id. at 88. 54. Id. at 90. 55. Id. at 277. 56. Elizabeth Davison & Jammie Price, How Do We Rate? An Evaluation of Online Student Evaluations, 34 Assess & Eval. Higher Educ 51, 52–55 (2009).

NOTES TO PAGES 186–192

57. Theodore Coladarci & Irv Kornfield, RateMyProfessors.com versus Formal In-Class Student Evaluations of Teaching, 12 Prac Assess, Rsrch & Eval 1, 8 (May 2007) concluded that the instructor ratings of University of Maine faculty at RateMyProfessors.com correlated closely with those produced by the official student instructor quality surveys conducted in class at the University of Maine. 58. Davison & Price, supra note 56, at 51, 62. 59. See Shteyngart, supra note 49, at 103. 60. See Id. at 166. 61. See Id. at 270. 62. Employees were more concerned about management unintentionally listening to personal calls. Gloria Lankshear et al., Call Center Employees’ Responses to Electronic Monitoring: Some Research Findings, 15 Work Emp. & Soc’y 595, 601–04 (2001). 63. Deborah L. Wells et al., The Impact of Perceived Purpose of Electronic Performance Monitoring on an Array of Attitudinal Variables, 18 Human Resources Dev. Q. 121, 133 (2007). 64. Erica Ann Robles et al., Being Watched or Being Special: How I Learned to Stop Worrying and Love Being Monitored, Surveilled, and Assessed, in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems 831, 833–35 (Rebecca Grinter et al. eds., 2006). 65. Id. at 837. 66. See, e.g., Steve Mann et al., Sousveillance: Inventing and Using Wearable Computing Devices for Data Collection in Surveillance Environments, 1 Surveillance & Soc’y 346–47 (2003), http://www.surveillance-and-society.org/articles1(3)/sousveillance.pdf. 67. Kath Weston, Get Thee to a Big City: Sexual Imaginary and the Great Gay Migration, 2 GLQ 253 (1995). 68. This account helps explain why even people with poorer than average reputations would continue to opt in to the system. Participation would provide a means for such an individual to separate himself from those with horrendous reputations. Even people whose reputational profiles include horrendous past acts might still have some incentive to improve their reputation in the future, so as to separate themselves from the unredeemable sociopaths at the very bottom of the reputation hierarchy. 69. See, e.g., Alan F. Westin, Privacy and Freedom (1967). 70. See, e.g., Julie E. Cohen, Examined Lives: Information Privacy and the Subject as Object, 52 Stan. L. Rev. 1373 (2000). 71. See, e.g., Hannah Arendt, The Human Condition 38–78 (1958). 72. See generally James Q. Whitman, The Two Western Cultures of Privacy: Dignity Versus Liberty, 113 Yale L.J. (2004). 73. See Lior Jacob Strahilevitz, “How’s My Driving?” for Everyone (and Everything?), 81 N.Y.U. L. Rev. 1699, 1705–08, 1743–44 (2006), and the sources cited therein. 74. Whitman, supra note 72, at 1189–1218. 75. Jon D. Michaels, All the President’s Spies: Private-Public Intelligence Partnerships in the War on Terror, 96 Cal. L. Rev. 901 (2008).

247

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

248

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

NOTES TO PAGES 193–195

Conclusion 1. Robert A. Caro, The Power Broker: Robert Moses and the Fall of New York 318–19 (1974). 2. Id. at 319. 3. Id. at 514. 4. The distinction between “hard shove” regulatory approaches and “gentle nudge” approaches is Dan Kahan’s. Kahan argues that when the law tries to prohibit conduct that is tolerated by prevalent social norms, a backlash against the law often undermines the prohibition’s goals. Counterintuitively, more incrementalist approaches by the government may prove more effective, in that they trigger a gradual shift in social norms. Dan M. Kahan, Gentle Nudges vs. Hard Shoves: Solving the Sticky Norms Problem, 67 U. Chi. L. Rev. 607, 610–18 (2000).

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INDEX

Abdullah, king of Jordan, 167–68 Abercrombie & Fitch, 16 Abrams, Dominic, 25 advertising: residential, 48–50, 57, 84, 94–95, 96, 107; online, 129 aesthetics, as exclusionary vibe, 50–51 affirmative action, 115, 179 altruism, 24 Amazon, 128, 181–82 Americans with Disabilities Act (ADA), 115, 174, 175 AmIHotorNot.com, 183, 187 Andrews, Jim, 179 animus-based discrimination, 140, 141, 159–60, 162 anonymity, 128, 133, 166; of motorists, 167, 191; in charitable giving, 187–88 antidiscrimination law, 120–22, 140–46 antitrust law, 123–24 ApartmentRatings.com, 139 arbitrage, 79, 80 Arcadia University (Beaver College), 47 artist colonies, 116, 118 asymmetric information, 5, 6, 38–39, 75–92, 127, 132, 139, 155, 159; bouncer’s exclusion and, 5, 6, 15, 76; offsetting, 127, 132 athletic programs, 115

attractive nuisance, 24 Ave Maria Township (Fla.), 5, 68–70 Ave Maria University, 68, 119 Avvo.com, 170–71 background checks: in labor market, 4, 142–44, 157, 190; in housing market, 77, 138; in jury selection, 147, 151, 157 Balsa, Ann, 154 Batson v. Kentucky (1986), 148–49, 151, 160 Bauer, William, 109 beaches, 51–52, 98, 193–94 Beaver College (Arcadia University), 47 Becker, Gary, 121 Berra, Yogi, 15 Besley, Timothy, 80 blacklisting, 136, 138, 139 Blackstone, William, 29 Bloch, Lynne, 107–10 Bloch v. Frischholz (2008), 107–10 blockbusting, 30, 84 blogs, 133 bluffing, 45, 98 bouncer’s exclusion, 13, 28–41; examples of, 4; asymmetric information and, 5, 6, 15, 76; alternatives to, 42–43; prospective, 44–45, 83; inefficiency

249

250

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INDEX

(continued) of, 77–78, 79; self-assessment and, 81, 82; collective action impeded by, 84–85; as property right, 85; social norms governing, 87–88; vague, 89–90; advantages of, 90, 95, 155–56; mistaken application of, 91–92; mistaken perceptions of, 95; difficulty of circumventing, 97; regulation of, 98, 101, 102–3, 156; sex offenders and, 100–101; religious homogeneity achieved by, 103–4; rules functioning as, 108 branding, 44–48 Brashears, Matthew, 105, 106 Brazil, 51 Browne, John Henry, 170, 171 Brown v. Board of Education (1954), 20 Brown v. Burdett (1882), 28 Buchanan, Bob, 16 bundled amenities, 113–26, 156 Burke, Jane, 154 Burtless, Gary, 180–81 buyer’s remorse, 90, 95 California, 138, 158 California Coastal Commission, 52 Caro, Robert, 194 cell phones, 130 charitable giving, 187–88 Charron, Paul R., 15 cheap talk, 83 Citysearch, 128 clearinghouses, for information, 139, 144–45 Clingermayer, James, 88 club goods, 56, 60–61, 116, 122–25; defined, 55; private goods vs., 57–58; inclusionary amenities as, 113–14 Coate, Stephen, 80 code words, 89 College of New Jersey (Trenton State College), 47–48 Colorado, 146 commercial speech, 85, 97, 125 common carriers, 132 Communications Decency Act, 171

condominiums, 16–18, 54, 107–9, 114–15 conflict games, 82, 83 Consumer Reports, 172 consumer surplus, 182 contract rights, 11; property rights vs., 12 cooperatives, 16–18, 84 coordination games, 78, 82–84 country clubs, 19–20, 62 covenants, 34–36, 67–68 Cox, Adam, 38 creative class, 116 credit reporting, 54, 81, 128, 137, 138, 147, 169 cultural homogeneity, 86–87 Cunningham, John, 46 curtain strategies, 6–7, 157–58, 159, 160, 183 data mining, 156, 192 Data Security Breach statute, 158 Davidowitz, Howard, 15–16 Davison, Elizabeth, 185–86 deaf communities, 125–26 Dean, Dorothy, 41 death penalty, 150 DeChaine, D. Robert, 64–65 defamation, 171 Delancey Street Foundation, 178–79 digital rights management, 131 disabilities, 115 dress codes, 42 drug abuse, 152–53 due process, 11 “dumping grounds” problem, 98 Easterbrook, Frank, 108, 109 eBay, 128, 129, 131, 162, 164–65, 170–71 economies of scale, 125 education: as right, 11; racial discrimination in, 20; magnet schools, 116 elections, 157, 191 Ellickson, Bob, 167 environmental quality, 120; regulation of, 13, 27, 29 Epstein, Richard, 31, 121

index

equal protection clause, 149 equitable principles, 110 evictions, 136, 137, 197 excessive precaution, 32 exclusion: as quintessential right, 12; strategies of, 13–14, 93–110; trespassbased, 14, 45, 76–82; exclusivity and, 15–19; costs of, 20–21; advantages of, 21–25; governance vs., 25–27, 29, 35–36, 43, 80, 103, 106–7; limitations on, 29–34 exclusionary amenities, 55–67, 78–79, 83, 195; defined, 5; offsetting, 6; exclusionary vibes vs., 14; asymmetric information and, 15, 76; religious homogeneity achieved by, 68–70, 103, 107; collective action impeded by, 84–85; lawfulness of, 87, 101–2; marketing of, 89; mistaken perceptions of, 91; subtlety of, 95; ease of circumventing, 97; bundling of, 113–26; overt discrimination vs., 120–21, 122; benign, 122–23, 125–26; combating, 124–25; shift away from, 155; refinement of, 156 exclusionary vibes, 5, 14–15, 42–54, 87, 138, 195; defined, 4; offsetting, 6; language as, 14, 89; asymmetric information and, 76; vague, 78, 90–91; costs of, 83–84; collective action unimpeded by, 85, legal regulation of, 86, 98, 101, 102; subtlety of, 88–89, 107; externalities of, 94–97; ease of circumventing, 97; religious homogeneity achieved by, 103; rules functioning as, 108; shift away from, 155; refinement of, 156 exclusivity, 15–19 ex-offenders, 142–43, 144, 174, 176, 178–79, 181 Experian, 169 expressive rights, 85 externalities, 93; positive, 94; symbolic, 94–95, 102; misperception, 95–97, 102, 107; liberty, 97–98, 102

Facebook, 38, 128, 132, 133–34 Fair Credit Reporting Act, 138 Fair Housing Act, 30–31, 96, 101, 103, 108 Falls Church, Va., 59–60 false information, 161–65 Farrell, Joseph, 83 Federal Housing Administration, 57 feedback, 129, 132, 133, 145, 161; false, 162–63, 164–65, 171–72; inflation of, 165–66; in speed-dating study, 188–89 fee simple ownership, 17 Felony Franks, 179 Fennell, Lee, 19 Finlay, Keith, 143 First Amendment, 85, 138. See also free speech Florida: Ave Maria Township, 5, 68–70; golf courses in, 63, 65; criminal databases in, 146 Florida, Richard, 116 focal points, 44–45, 55–56, 66–67, 69, 70, 78, 82, 85 Franklin Park Golf Club (Boston), 118 fraternities, 1–2, 20, 36–38, 82 fraud, 76 Freedom of Information Act (FOIA), 99 free riding, 58, 65, 78, 122 free speech, 11, 13–14, 85 fugitive resources, 29 fungibility, 79–80 game theory, 78, 82, 84 Gandy, Oscar, 133–34 Gap (clothing store), 15–16 gated communities, 4, 23, 24 Gates, Henry Louis, 168 genetic discrimination, 157 gentrification, 117 getunvarnished.com, 183–84 Ginsburg, Ruth Bader, 171 Global Positioning System, 130 golf course communities, 3–4, 5, 61–67, 86, 120–21, 122–23 Google, 127, 130, 133

251

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

252

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INDEX

gossip, 167 governance, 45, 64, 83, 84, 87; exclusion vs., 25–27, 29, 35–36, 43, 80, 103, 106–7 Gregoire, Christine, 91–92 Guinnane, Timothy, 80 Hamptons (Long Island), 120 Harbour Ridge Yacht & Country Club, 65 Harcourt, Bernard, 116–17 hate speech, 102 health care, as right, 11 HIV, 174–75 Hogg, Michael, 25 Holzer, Harry, 142, 176–77 homeowners associations, 3, 34, 98, 100–101 homosexuality, 174–75, 189 Hong, James, 183 housing discrimination, 4, 5; legal regulation of, 18–19, 30–31, 48–50, 57, 96, 101–3, 195; on religious grounds, 102, 103–10; against tenants, 134–40. See also race identity theft, 161 Illinois, 147–48 immigration, 38–41, 88, 197 Immigration and Customs Enforcement (ICE), 40–41 inclusion, 23–24 inclusionary amenities, 6, 113–20 information: asymmetric, 5, 6, 38–39, 75–92, 127, 132, 139, 155, 159; costs of, 14–15, 38; private, 76–82, 84 information policy, 99, 101, 103 innovation, 22–23 in personam rights, 12 in rem rights, 4, 12 intercollegiate athletics, 115 Internet, 99, 100 journalists, rights of, 33 Judaism, 102, 107–10 judicial elections, 157 JuryQuest, 147

jury selection, 146–52, 157–58, 160 King, Martin Luther, Jr., 141 landlord-tenant law, 134–40, 196–97 land use regulation, 27 language, as exclusionary vibe, 14, 89 Latitude, 130 Laurent, S.D., 125–26 l33t (online language), 71 Levine, Gene, 1 liberty externalities, 97–98, 102 Loewen, James, 49–50, 66 Loopt, 130 Lyon, David, 133–34 magnet schools, 116 majority rule, 83, 105, 107, 110 Marques, José, 25 Martindale Hubbell, 171 massively multiplayer online games (MMOs), 52–53 McDaniel College (Western Maryland College), 47 McGuire, Thomas, 154 medical stereotyping, 151–55 Megan’s Laws, 34, 37, 79, 99–101, 145–46, 181, 196 Mensa, 105 Meredith, Lisa, 154 mezuzot, 107, 108–10 Microsoft, 123 migrant farm workers, 32–33 military secrets, 23 Miller, Jonathan, 18 misperception externalities, 95–97, 102, 107 misrepresentation, 76 mistaken identity, 161 Monaghan, Tom, 68 Moses, Robert, 193–97 moveable property, 23 “Mrs. Murphy exception,” 96–97, 98, 101 Mulligan v. Panther Valley Property Owners Ass’n (2001), 34–36, 98

index

names, as focal points, 44–48 narcotics, 152–53, 196 National Crime Information Center, 146 national origin, 38–39 necessity, in trespass, 31–32 negative liberties, 11 “neighborhood watch” signs, 100 Netflix, 128 network effects, 105, 125 New York City, 135–36, 193–94 nightclubs, 4, 13, 41, 43, 87 Northern Ireland, 105 “no trespassing” signs, 52 nuisance, 27 Obama, Barack, 141 Open Table, 42, 128 optimal redistribution analysis, 178, 180 Orthodox Judaism, 102 Palms of Manasota, 56 Panoptic Sort, The (Gandy), 133 Panther Valley Property Owners Association, 34–36 patents, 22 peer evaluation, 167 Pelago, 130 Penner, James, 28 peremptory challenges, 148–49 philanthropy, 187–88 Pipl, 129 police brutality, 168 Pollyanna effect, 164–65 poor laws, 80 positive externalities, 94 positive rights, 11 Posner, Eric, 38 Posner, Richard, 151 prescription drugs, 152–53, 196 press, freedom of, 33 pretexts, 88, 135, 149, 159–60. See also proxies Price, Jammie, 185–86 price discrimination, 181–83 privacy, 22–23, 99; technological benefits vs., 132, 133–34; access to

information vs., 145, 159; jury system and, 151, 157; importance of, 190–91; in Europe, 192 private information, 76–82, 84 property rights, 11; contract rights vs., 12 proxies, 130; exclusionary, 3, 18–19, 38–40, 89, 108, 195; in rental housing market, 18–19, 134–35, 139; golf courses as, 61, 86; amenities as, 79; in labor market, 141–44; in jury selection, 151; in narcotics prescriptions, 153; reputations as, 165, 166 p3p systems, 131 public goods, 70; exclusionary, 71; inclusionary, 115–16, 118 public transportation, 119–20 public trust lands, 52 pure coordination games, 78, 83–84 race: residential segregation by, 5, 58; golf course communities and, 5, 86; job discrimination and, 6; income and wealth as proxies for, 18–19; Fair Housing Act and, 30–31, 96, 101, 103; national origin as proxy for, 38–39. See also housing discrimination Raphael, Stephen, 142 RateMyProfessors.com, 183, 185 Rawls, John, 179 real estate development, 3, 5, 13, 14, 29, 46; inclusionary amenities in, 6, 113–20; condominiums favored in, 17; focal points in, 44, 45; advertising and, 48–50, 57, 84, 94–95, 96, 107; families with children and, 59–60; of golf course communities, 61–67, 86 rebranding, 48 religious communities, 67–71 religious discrimination, 102, 103–10 rental housing, 134–40 repeat-play games, 84 reputation effects, 76 reputation revolution, 6, 127–33 residential advertising, 48–50, 57, 84, 94–95, 96, 107

253

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

254

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x

INDEX

restaurants, 42–43, 87; health inspection of, 158 restraints on alienation, 104–5 retirement communities, 56, 66 right of reply, 170–71 Rio de Janeiro, 51 Rise of the Creative Class, The (Florida), 116 Roosevelt, Franklin Delano, 193 Rose, Carol, 29–30 Ryan, George, 147–48 Schelling, Thomas, 44 Schill, Michael, 18 search costs, 123, 138 searchlight strategies, 6–7, 158–60, 161, 163, 179–80, 183 Second Life, 12 Securities and Exchange Commission (SEC), 158 segregation, 5, 58, 61 self-help evictions, 137 self-sorting, 57, 79, 114, 165 Sewanee (University of the South), 46–47 sex offenders, 7, 34–36, 56, 79, 98, 99–101, 128, 181 sexual orientation, 174–75 shaming, 133 Shasta County, Calif., 167 shelter, as right, 11 Shoreline Towers, 108 Shteyngart, Gary, 184–88 sign language, 125–26 Silver, Esther, 104 Silver, Robert, 104 singles bars, 88 Skid Row (Los Angeles), 116–17 Slashdot, 128, 131, 164 Smith, Henry, 26–27 social connectedness, 23–24 social meanings, 87–89 social networking, 128, 130, 133, 147 social sorting, 133 social status, 24 sororities, 1–2, 20, 36–38, 82

sorting, 55–56, 66–67, 133–34; self-, 57, 79, 114, 165 South Africa, 120 speed dating, 188–89 spite fences, 32 Spokeo, 129 sports: ethnicity and, 62, 120, 124–25; intercollegiate, 115 standing to sue, 80 State v. Shack (1971), 32–33 statistical discrimination: animus-based discrimination vs., 140, 160; in labor market, 141–46, 175, 177, 178; in medicine, 152–55, 196; curtain strategies to combat, 158, 159; forms of, 174 stereotypes, 58, 133, 140, 163 Stern, David, 183 Stoll, Michael, 142 Super Sad True Love Story (Shteyngart), 184–88, 191 surveillance, 134, 188 Sussman, Leila, 1 symbolic externalities, 94–95, 102 tagging, 80 takings clause, 85 Taormina Theosophical Community, Inc. v. Silver (1983), 103–7 Targeted Jobs Tax Credit, 178 tax incentives, 144–45, 157, 178, 180–81 teaching evaluations, 165 Theosophy, 103–7 360-degree feedback, 145, 161–62, 163 Tiger Woods Foundation, 125 total exclusion, 29 Toxic Release Inventory, 158 trade secrets, 22–23 tragedy of the commons, 66 transaction costs, 79 Trenton State College (College of New Jersey), 47–48 trespass, 4, 12, 13, 23, 27; necessity in, 31–32 trespass-based exclusion, 14, 45, 76–82, 87, 195

index

TripAdvisor.com, 139, 158, 164, 165, 170–71

voir dire, 146 voting, 157, 191

unbundling, 123–24 University of Chicago, 48 University of the South (Sewanee), 46–47 University of Wisconsin, 48 U.S. Court of Appeals: D.C. Circuit, 96; Second Circuit, 101–2; Fourth Circuit, 96; Seventh Circuit, 108, 109 U.S. Department of Housing and Urban Development, 60 U.S. News and World Report, 172

waiver, 33 Wang, Zhongmin, 128 Washington, D.C., subway, 119–20 Waterford Development, 59–60 wearable communities, 131 welfare recipients, 180–81 Western Maryland College (McDaniel College), 47 wilderness areas, 13 Winfrey, Oprah, 168 Wood, Diane, 108 Woods, Tiger, 122, 124–25 workhouses, 80, 134–35 Work Opportunity Credit, 178

vagueness, of exclusionary criteria, 89–90 van Ryn, Michelle, 154 vestries, 81 Villa Grove, Ill., 49–50 video games, 52–53 Virginia State Pharmacy Board v. Virginia Citizens Consumer Council (1976), 85 virtual reality, 12, 52–53, 71 Voicu, Ioan, 18

Yahoo, 128 Yale University, 102 Yelp, 128, 158 YouTube, 183 Zacahariasse, Jan A., 59–60 Zagat guides, 158, 170 zoning, 27, 29, 88, 120

255

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 20 1 2 3 4 5 6 7 8 9 30 1 2 3 4 5 36x