Public Accommodations Under the Civil Rights Act of 1964: Why Freedom of Association Counts as a Human Right

University of Chicago Law School University of Chicago Law School

Chicago Unbound Chicago Unbound

Journal Articles Faculty Scholarship

2014

Public Accommodations Under the Civil Rights Act of 1964: Why Public Accommodations Under the Civil Rights Act of 1964: Why

Freedom of Association Counts as a Human Right Freedom of Association Counts as a Human Right

Richard A. Epstein

Follow this and additional works at: https://chicagounbound.uchicago.edu/journal_articles

Part of the Law Commons

Recommended Citation Recommended Citation

Richard A. Epstein, "Public Accommodations Under the Civil Rights Act of 1964: Why Freedom of

Association Counts as a Human Right," 66 Stanford Law Review 1241 (2014).

This Article is brought to you for free and open access by the Faculty Scholarship at Chicago Unbound. It has been

accepted for inclusion in Journal Articles by an authorized administrator of Chicago Unbound. For more

information, please contact unbound@law.uchicago.edu.

1241

PUBLIC ACCOMMODATIONS UNDER

THE CIVIL RIGHTS ACT OF 1964:

HY FREEDOM OF ASSOCIATION

OUNTS AS A HUMAN RIGHT

Richard A. Epstein*

On its fiftieth anniversary, Title II of the Civil Rights Act of 1964 enjoys

widespread social support on all sides of the political spectrum. That support is

fully deserved to the extent that the nondiscrimination in public accommodations

provisions offset the monopoly power of common carriers and public utilities, or

neutralize the abusive application of public power and private violence to sup-

press the free entry of firms that would otherwise target minority customers in

competitive markets.

The subsequent expansion of Title II’s nondiscrimination principle becomes

much more difficult to justify, however, when applied to normal businesses when

segregationist forces no longer hold sway. In particular, these principles are sus-

pect when applied to membership organizations that care about their joint gov-

ernance and common objectives. In these cases, the principles of freedom of as-

sociation should constitutionally protect all groups, even those that do not fall

under the uncertain rubric of expressive associations.

The application of the modern antidiscrimination rules for public accommo-

dations to Christian groups who are opposed to gay marriage on moral principle

represents a regrettable inversion of the original purpose of Title II, using state

power to force these groups to the unpalatable choice of exiting the market or

complying with these modern human rights laws that prohibit any discrimination

on grounds of sexual orientation. These rules should be struck down even if the

other antidiscrimination prohibitions represent a group of settled expectations

that no one today wishes to overturn.

* Laurence A. Tisch Professor of Law, New York University School of Law; Peter

and Kirsten Bedford Senior Fellow, Hoover Institution; James Parker Hall Distinguished

Service Professor Emeritus of Law and Senior Lecturer, University of Chicago Law School.

My thanks to Lauren Barnett for her valuable research assistance, and to the Symposium par-

ticipants for their instructive comments on, and disagreements with, an earlier version of this

Essay.

1242 STANFORD LAW REVIEW [Vol. 66:1241

INTRODUCTION..................................................................................................... 1242

I. THE STATUTORY SCHEME .............................................................................. 1244

II. SUBSTANTIVE SOUNDNESS. ............................................................................ 1246

A. Common Carriers ..................................................................................... 1249

B. The Historical Reality .............................................................................. 1254

III.THE EXPANSION OF TITLE II AND KINDRED STATUTES .................................. 1261

IV.FROM PUBLIC ACCOMMODATIONS TO HUMAN RIGHTS .................................. 1266

A. Jaycees and Boy Scouts ............................................................................ 1266

B. The Small Business Problem .................................................................... 1277

V. GOVERNMENT DISCRIMINATION ..................................................................... 1287

AN UNHAPPY CONCLUSION ................................................................................. 1290

INTRODUCTION

The fiftieth anniversary of the passage of the Civil Rights Act of 1964 has

been, and will continue to be, a source of remembrance and reflection. Like

many people of my generation, I believed then, as I believe today, that its pas-

sage was a defining moment in American culture, which had for far too long

tolerated state-sanctioned segregation backed by massive social intolerance. At

the time, the civil rights movement had priorities that are accurately reflected in

the order of its particular titles. Title I, which dealt with voting, was first and

foremost on everyone’s mind because the overt and systematic exclusion of Af-

rican Americans from the polls was as complete an affront to full citizenship as

anyone could imagine. Next in line was Title II, which dealt with the question

of public accommodations. Title VII, which dealt with employment discrimina-

tion, was way down on the list, as it should have been.

Title II was passed when memories were still fresh of the many indignities

that had been inflicted on African American citizens on a routine basis. It took

little imagination to understand that something was deeply wrong with a nation

in which it was difficult, if not impossible, for African American citizens to se-

cure food, transportation, and lodging when traveling from place to place in

large sections of the country. In some instances, no such facilities were availa-

ble, and in other cases they were only available on limited and unequal terms.

As someone who came of age (quite literally) when the Civil Rights Act was

passed, it is easy to recall how widespread moral outrage propelled the statute

to its passage. The sit-ins in Southern and border states were still fresh in the

memories of the general public.

Those memories come flooding back with

1. See RICHARD A. EPSTEIN, FORBIDDEN GROUNDS: THE CASE AGAINST EMPLOYMENT

DISCRIMINATION LAWS (1992) (critiquing Title VII and calling for its repeal in competitive

labor markets).

2. See Bell v. Maryland, 378 U.S. 226, 239-42 (1964) (remanding a case involving

several criminal convictions resulting from a sit-in protest to the Maryland Court of Appeals

June 2014] PUBLIC ACCOMMODATIONS 1243

more recent historical accounts of the earlier times. Just recently, I read Isabel

Wilkerson’s The Warmth of Other Suns: The Epic Story of America’s Great

Migration, which contains personalized accounts of how difficult it was for

black people, fearing violent retribution, to sneak out of the South on segregat-

ed trains in their efforts to make it to the North.

And Wilkerson’s account of

how Robert Joseph Pershing Foster was unable to find sleeping accommoda-

tions in Arizona on his migration to California in 1953 shows all too vividly

that the practices of segregation extended far beyond the boundaries of the Old

South.

Indeed, it can be taken as a vindication of Title II that its commands

have rarely been the subject of litigation after the initial skirmishes on its con-

stitutionality were decided shortly after passage of the Civil Rights Act.

The

sign of successful legislative reform is its widespread social acceptance, which

has certainly been the case for Title II.

In one sense, therefore, it is proper to treat the fiftieth anniversary of Title

II as occasion for unrestrained celebration of legislation that has both met and

exceeded the expectations at the time of its passage. But at the same time, the

passage of a successful piece of legislation should give rise to at least some

level of reflection about the principles on which that legislation rests and their

soundness for general applicability. On this score, the inquiry goes off on two

branches. For the first, it turns out that the original design of Title II contains its

fair share of conceptual and practical difficulties, relating both to the terms of

its passage on the one hand and its precedential value on the other. The para-

digmatic case of Title II’s application in 1964 was against monopolists who

used their powers of exclusion to limit the options of politically vulnerable per-

sons.

Historically, the dominant white segregationists who controlled the polls,

the police, and all key government positions exercised in combination a level of

state monopoly power that no simple public utility could hope to match. It was

against the backdrop of this unified phalanx that the passage of the Civil Rights

Act of 1964 has to be understood.

For the second, the resurgence of Title II-type obligations under modern

“human rights laws” indicates a serious and regrettable reversal of fortune with

respect to the basic function of this legislation. Under pressure from modern

civil rights advocates, the worm has unfortunately turned, as people have lost

sight of the evils that a public accommodations law should combat. The new

application of the next generation of human rights law has the exact opposite

for reconsideration in light of the supervening passage of city and state public accommoda-

tions laws).

3. See ISABEL WILKERSON, THE WARMTH OF OTHER SUNS: THE EPIC STORY OF

AMERICA’S GREAT MIGRATION (2011).

4. See id. at 206-10.

5. See Heart of Atlanta Motel, Inc. v. United States, 379 U.S. 241 (1964).

6. See id. at 261.

1244 STANFORD LAW REVIEW [Vol. 66:1241

orientation: may the state force small and isolated businesses, often with Chris-

tian beliefs, to violate their bona fide religious beliefs in order to provide ser-

vices in highly competitive market segments? The fact that the two problems

are both seen to justify strong government intervention offers powerful evi-

dence of an unfortunate change in the dominant social attitudes toward the pub-

lic use of force.

In order to work out the development of this theme, I shall proceed as fol-

lows. In Part I, I shall set out in brief fashion the essential structure of Title II,

which, at least in form, has become the template against which all modern hu-

man rights laws are measured. In Part II, I shall examine the substantive sound-

ness of the modern civil rights laws when tested against the standards for anti-

discrimination laws developed in various common law contexts. In so doing, I

shall pay special attention to the controversial critique of public accommoda-

tion laws offered by the late Robert Bork in The New Republic in August

1963.

In Parts III and IV, I shall trace the evolution of public accommodations

laws into human rights laws that first addressed the position of large organiza-

tions like the Jaycees and the Boy Scouts, but later extended their influence to

small, often fundamentalist groups that are frequently powerless to protect

themselves against the rigors of majoritarian political processes. In Part V, I

shall address the new set of dangers that arise when governments, at both the

federal and state levels, use their monopoly powers over highways and other

public facilities to exclude those groups whose internal practices are incon-

sistent with the preferred set of public norms. A civil rights program that at one

time protected individual liberty and choice has by degrees become an instru-

ment of repression in the hands of public and private groups. Even on a cele-

bratory occasion, therefore, it is important to keep our intellectual distance and

subject it to some serious scrutiny, which is what this Essay attempts to do.

I. T

HE STATUTORY SCHEME

In order to put this inquiry into perspective, it is critical to set out the spe-

cific guarantees contained in Title II, paying special attention to their scope and

underlying rationale. The initial section sets out the basic guarantee: “All per-

sons shall be entitled to the full and equal enjoyment of the goods, services, fa-

cilities, privileges, advantages, and accommodations of any place of public ac-

commodation, as defined in this section, without discrimination or segregation

on the ground of race, color, religion, or national origin.”

Once that is established, subsection (b) of the legislation then lists the types

of accommodations that fall within the general ambit of the Act.

7. Robert Bork, Civil Rights—A Challenge, NEW REPUBLIC, Aug. 31, 1963, at 21.

8. Civil Rights Act of 1964, tit. II, § 201(a), 42 U.S.C. § 2000a(a) (2012).

June 2014] PUBLIC ACCOMMODATIONS 1245

Each of the following establishments which serves the public is a place of

public accommodation within the meaning of this title if its operations affect

commerce, or if discrimination or segregation by it is supported by State ac-

tion:

(1) any inn, hotel, motel, or other establishment which provides lodging to

transient guests, other than an establishment located within a building which

contains not more than five rooms for rent or hire and which is actually occu-

pied by the proprietor of such establishment as his residence;

(2) any restaurant, cafeteria, lunchroom, lunch counter, soda fountain, or

other facility principally engaged in selling food for consumption on the prem-

ises, including, but not limited to, any such facility located on the premises of

any retail establishment; or any gasoline station;

(3) any motion picture house, theater, concert hall, sports arena, stadium or

other place of exhibition or entertainment; and

(4) any establishment (A)(i) which is physically located within the premis-

es of any establishment otherwise covered by this subsection, or (ii) within the

premises of which is physically located any such covered establishment, and

(B) which holds itself out as serving patrons of such covered establishment.

These provisions have to be read in light of the narrow exemption that is

found in subsection (e), which reads: “The provisions of this title shall not ap-

ply to a private club or other establishment not in fact open to the public, except

to the extent that the facilities of such establishment are made available to the

customers or patrons of an establishment within the scope of subsection (b).”

In addition to these basic coverage provisions, Title II also contains subsec-

tions (c) and (d), which are intended to secure the proper constitutional basis

for the substantive provisions just set out above. Thus subsection (c) offers a

very broad definition of what it means for the “operations of an establishment”

to “affect commerce,” which includes the service at these establishment of in-

terstate travelers or the use of food, gasoline, or other products that “move[] in

[interstate] commerce.”

Subsection (d) then gives a definition of the meaning

of state action under the provision, which includes all conduct carried out under

the color of state law or custom or required by any state or political subdivision

thereof.

There is no question that these provisions received extra attention in

order to avoid the fate of the Civil Rights Act of 1875,

which had been de-

9. Id. § 201(b), 42 U.S.C. § 2000a(b).

10. Id. § 201(e), 42 U.S.C. § 2000a(e).

11. Id. § 201(c), 42 U.S.C. § 2000a(c).

12. Id. § 201(d), 42 U.S.C. § 2000a(d).

13. Civil Rights Act of 1875, ch. 114, 18 Stat. 335, invalidated by The Civil Rights

Cases, 109 U.S. 3 (1883). Its basic provision read:

[A]ll persons within the jurisdiction of the United States shall be entitled to the full and equal

enjoyment of the accommodations, advantages, facilities, and privileges of inns, public con-

veyances on land or water, theaters, and other places of public amusement; subject only to

1246 STANFORD LAW REVIEW [Vol. 66:1241

clared unconstitutional in the Civil Rights Cases on the ground that

“[i]ndividual invasion of individual rights is not the subject-matter of the [Four-

teenth] Amendment.”

Nor could the Civil Rights Act of 1875 reach those

forms of private action that were outside the scope of interstate commerce as

that phrase was understood at the time, which roughly speaking was limited to

any journey that involved two or more states.

It is commonly overlooked—I

plead guilty to the offense—that Title II does not explicitly apply to retail es-

tablishments,

which thus fall into a limbo between the explicit command of

Title II and its narrower exception, which has been plugged in large measure by

the human rights laws of the next generation.

It is worthwhile examining both

the substantive soundness of the basic provisions of Title II and its constitu-

tional basis in some detail. It is equally important to note that the Commerce

Clause restraint drops out with respect to state human rights laws, even though

the various property right restrictions on government remain very much in play.

II. S

UBSTANTIVE SOUNDNESS

In the run-up to the passage of the Civil Rights Act of 1964, most serious

commentators had little doubt about the moral imperative behind passage of

Title II of the Act. In light of the earlier court precedents, they were far more

worried about its constitutional foundations for congressional action. The nota-

ble exception to that consensus was the ever-contrarian Robert Bork, who in

August 1963 wrote a scathing critique of Title II in the New Republic on the

ground that it offended the basic principle of freedom of association, which he

regarded as a fundamental norm of any sound political order. One key passage

reads as follows:

The legislature would inform a substantial body of the citizenry that in order

to continue to carry on the trades in which they are established they must deal

with and serve persons with whom they do not wish to associate. In part the

willingness to overlook that loss of freedom arises from the feeling that it is ir-

rational to choose associates on the basis of racial characteristics. Behind that

judgment, however, lies an unexpressed natural-law view that some personal

preferences are rational, that others are irrational, and that a majority may im-

pose on a minority its scale of preferences. The fact that the coerced scale of

the conditions and limitations established by law, and applicable alike to citizens of every

race and color, regardless of any previous condition of servitude.

Id. § 1, 18 Stat. at 336.

14. 109 U.S. at 11.

15. See The Daniel Ball, 77 U.S. (10 Wall.) 557, 566 (1870).

16. The point is driven home at great length in Joseph William Singer, No Right to Ex-

clude: Public Accommodations and Private Property, 90 N

W. U. L. REV. 1283, 1288-93

(1996).

17. See infra Part IV.A.

June 2014] PUBLIC ACCOMMODATIONS 1247

preferences is said to be rooted in a moral order does not alter the impact upon

freedom. . . . Of the ugliness of racial discrimination there need be no argu-

ment (though there may be some presumption in identifying one’s own hotly

controverted aims with the objective of the nation). But it is one thing when

stubborn people express their racial antipathies in laws which prevent individ-

uals, whether white or Negro, from dealing with those who are willing to deal

with them, and quite another to tell them that even as individuals they may not

act on their racial preferences in particular areas of life. The principle of such

legislation is that if I find your behavior ugly by my standards, moral or aes-

thetic, and if you prove stubborn about adopting my view of the situation, I am

justified in having the state coerce you into more righteous paths. That is itself

a principle of unsurpassed ugliness.

That passage, especially its last four words, has provoked outrage that has

spanned generations. Although Bork tried to distance himself from this piece,

mere recollection of the phrase helped doom his 1987 nomination to the United

States Supreme Court. Indeed, as recently as two years ago, that 1963 article

formed part of the grounds for the People for the American Way, which played

such a powerful role in derailing Bork’s 1987 nomination, to denounce Mitt

Romney for the bad judgment of naming Bork as one of the co-chairs of the le-

gal advisory committee for his 2012 presidential campaign.

Nonetheless, it is

worthwhile in placing this passage in context to see the strengths and weak-

nesses of Bork’s broadside against Title II. The story yields a divided verdict.

Let us start with the positive. There is little doubt that Bork is correct on

his major premise that it is difficult to conceive of how a society can flourish if

it does not respect the basic principle of freedom of association—the funda-

mental right that all individuals, regardless of race, sex, or age, have to choose

the persons with whom they choose to do business. The logic behind this free-

dom of association principle runs as follows: Any legal system has to contain

rules that first keep people apart (so that they will not kill each other). Yet at

the same time, people have to be able to engage in various relationships with

other individuals in order to reap the gains from voluntary interaction.

Those gains can come about in two distinct ways. The first is through sim-

ple acts of exchange whereby one person sells goods or services to another. The

basic economic logic of that exchange is that that the seller values the cash (or

other nonmonetary consideration) received more than the goods or services sur-

rendered, while the buyer values the goods or services received more than she

does the cash used to pay for them. Both sides thus benefit from the transaction,

and each is in a position to make better use of the consideration received than

of that surrendered. The seller can choose to use his added wealth to acquire or

18. Bork, supra note 7, at 22.

19. See, e.g., Jamie Raskin, Borking America: What Robert Bork Will Mean for the

Supreme Court and American Justice, P

EOPLE FOR AM. WAY, http://www.pfaw.org/media-

center/publications/borking-america-20 (last visited June 8, 2014).

1248 STANFORD LAW REVIEW [Vol. 66:1241

make new goods for sale, and the buyer can use her goods either for consump-

tion or resale. A successful transaction sets the stage for further beneficial in-

teractions with each other and with third parties. Legal systems may huff and

puff about the exact conditions for valid exchanges, but they all recognize the

category of sales and, along with it, similar two-party arrangements like leases,

mortgages, bailments, and the like.

The second source of potential gains does not involve transfer, but coop-

eration. Two or more individuals can pool their capital or labor in some form of

partnership where they divide the gains as they see fit, typically in accordance

with their respective contributions. These arrangements are governed by a full

set of legal and social norms. The constant need for cooperation requires a

higher level of affinity between parties. Many of their obligations are cast in the

slippery but necessary terms of good faith; in essence, treat the welfare of your

partner as equal to your own. One reason why parties are allowed to select their

associates is that they pick those whom they tend to trust, which in turn creates

a social glue that reduces the burden on the purely legal sanctions that are put

in place to protect each trading partner against the potential default of others.

It is, of course, an open question just how much trust is involved in differ-

ent kinds of relationships, or indeed among different relationships in the same

class. As a rough generalization, people care a lot more about the choice of

their long-term business associates than they do about the personal identity of

people with whom they trade standard goods, at least in situations involving

mass-produced goods. Hiring a worker is not like walking through the checkout

line at Walgreens. But these general rules, even if sensible first approximations,

are often subject to important exceptions. Some partnerships are large associa-

tions where the element of personal trust is relatively small. Some sales con-

tracts dealing with customized goods may require constant cooperation be-

tween buyers and sellers in order to successfully execute the transactions. In

other cases, the element of trust is needed in large purchases when payment

comes before performance or vice versa. Some service contracts involve close

relations; others do not. The point here is not to catalog exhaustively all the

permutations that can and do arise in well-developed voluntary markets. It is

enough in this setting to know that the parties themselves are sensitive to these

issues, so that the selection of terms and conditions can be left safely in private

hands. The more personal the relationship is, the greater the selectivity in the

contractual parties. It makes perfectly good sense that the sound selection of

contracting parties can substantially reduce the risk of contractual breakdown

by calling into play a set of informal sanctions that will not work with perfect

strangers.

The most important goal of public policy in this context is to make sure

that government officials remember how little they know about the internal dy-

namics of the full range of voluntary transactions. These transactions are all en-

tered into by different parties, each with its own internal structure, history, and

personnel. The organization makes its judgments with a full range of infor-

June 2014] PUBLIC ACCOMMODATIONS 1249

mation that is not obtainable by outsiders no matter how hard they work. Any

internal conflicts of interest—so called agency costs

—of which there are

many, are tiny in comparison with the conflicts of interest between the business

as a whole and its regulators who have no stake in the financial viability of the

firm. Decentralization of public institutions thus is the order of the day.

It is therefore correct for any sound legal system to grant individuals free-

dom of choice in the full range of exchange and cooperative contexts. At one

level, Bork’s position was so correct as a theoretical matter that it is hard to see

what the shouting was about in 1963. But shouting there was, and even when

that subsides, there are valid sources of criticism. I will not talk at length about

Bork’s misuse of natural law theory, except to say that orthodox natural law

theory lends no support to the proposition that “some personal preferences are

rational, [and] that others are irrational.”

Quite the opposite, it assumes that

these preferences are personal and subjective and devotes its energy to articu-

lating the set of rules in which these preferences can be maximized in the famil-

iar fashion—controlling aggression and allowing for voluntary cooperation

along lines that rest on a deeply utilitarian foundation.

But if we put his un-

necessary philosophical digression to one side, there remain two important

questions on which Bork falls short. The first is getting a proper theoretical un-

derstanding of the duties of common carriers. The second is adjusting social

theory to take into account that the sometime manifest imperfection of power

structures throughout the United States, but concentrated in the Old South, is

about the institutional distribution of public powers. Both of these require some

real explanation.

A. Common Carriers

One striking feature of Bork’s 1963 article was that it was written at the

time that he was immersed in his articulation of a general theory of antitrust

law. The central tenet of that theory, which he developed in connection with

Ward Bowman, Jr., was that the single worthwhile objective of antitrust law is

the control of monopoly practices. Thus in one famous article, Bork wrote that

the only value worth defending in the antitrust law was “the maximization of

20. See Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial

Behavior, Agency Costs, and Ownership Structure, 3 J.

FIN. ECON. 305, 308-09 (1976).

21. See F.A. Hayek, The Use of Knowledge in Society, 35 AM. ECON. REV. 519, 526

(1945).

22. Bork, supra note 7, at 22.

23. See Richard A. Epstein, The Natural Law Influences on the First Generation of

American Constitutional Law: Reflections on Philip Hamburger’s Law and Judicial Duty, 6

J.L. PHIL. & CULTURE 103, 112 (2011); Richard A. Epstein, The Utilitarian Foundations of

Natural Law, 12 H

ARV. J.L. & PUB. POL’Y 713, 713 (1989) (pointing out the correspondence

in key areas).

1250 STANFORD LAW REVIEW [Vol. 66:1241

wealth or consumer want satisfaction.”

That statement was echoed in many

other articles that he wrote during this period, some alone and some with

Bowman, whom he had first met when the two men were still at the University

of Chicago.

Normatively, I think that there is a great deal to be said for this approach.

The office of antitrust law is to distinguish between competition and monopoly

and to make sure that legal intervention promotes the former and hampers the

emergence of the latter. When Bork entered the antitrust field in the early

1960s, most mainstream antitrust scholars thought that the law should serve all

sorts of collateral objectives, including protecting small businesses against the

rigors of competition.

Bork did much to counteract those diversions in the

law, arguing against the common belief that the simultaneous pursuit of multi-

ple objectives counts as some kind of social benefit. Bork was right to insist

that the embrace of multiple inconsistent objectives often leads to intellectual

confusion and the perpetuation of unwanted social losses.

It is therefore a matter of great irony that the most glaring defect in Bork’s

critique of Title II stems from his failure to recognize that antitrust law is not

the only body of legal rules that is directed toward the control of monopoly. Of

equal, or perhaps greater, importance is the legal response to the problem of

natural monopoly that emerged somewhat earlier in the English law than did

antitrust law, here in connection with those businesses that were, for good rea-

son, deemed “affected with a publick interest.”

The key historical text for

these purposes is Lord Chief Justice Matthew Hale’s De Portibus Maris, which

introduced that phrase into Anglo-American law. The basic point was that in

some situations the party who sets up a wharf or a crane may have, either by

nature or operation of law, a monopoly in providing a particular good or ser-

vice. In those cases, customers have no clear alternative place to go, so that the

refusal to deal takes on far greater weight than it does in a purely competitive

industry where there are many easily available options to purchase the same (or

a very similar) good or service from a rival merchant. In those latter cases, it is

24. Robert H. Bork, Legislative Intent and the Policy of the Sherman Act, 9 J.L. &

ECON. 7, 7 (1966); see also Ward S. Bowman, Jr., Toward Less Monopoly, 101 U. PA. L.

REV. 577, 612 (1953); Ward S. Bowman, Jr., Tying Arrangements and the Leverage Prob-

lem, 67 Y

ALE L.J. 19, 21 (1957).

25. For some of that history, see Richard A. Epstein, Bork’s Bowman, “Not Gone, but

Forgotten,” 79 A

NTITRUST L.J. (forthcoming 2014) (on file with author).

26. See, e.g., Robinson-Patman (Anti-Price Discrimination) Act of 1936, Pub. L. No.

74-692, 49 Stat. 1526 (codified as amended at 15 U.S.C. § 13 (2012)) (seeking to deny large

purchasers, then called chain stores, the advantages of volume purchases to the extent that

they hurt smaller competitors).

27. MATTHEW HALE, De Portibus Maris, in A TREATISE, IN THREE PARTS (c. 1670), re-

printed in 1 A

COLLECTION OF TRACTS RELATIVE TO THE LAW OF ENGLAND FROM

MANUSCRIPTS 1, 78 (Francis Hargrave ed., London, T. Wright 1787).

June 2014] PUBLIC ACCOMMODATIONS 1251

not even possible to practice price discrimination because the high demanders

will by definition have an opportunity to acquire the same good or service

somewhere else at the competitive price. At this point, it can be said with some

confidence that the market supplies, via its rules on free entry and exit, all the

protection that any person needs.

Those conditions do not hold with wharfs and cranes, and so the rule is

otherwise. As Hale explains:

If the king or subject have a publick wharf, unto which all persons that

come to that port must come and unlade or lade their goods as for the purpose,

because they are wharfs the only licensed by the queen, according to the stat-

ute of I. El. cap II. or because there is no other wharf in that port, as it may fall

out where a port is newly erected; in that case there cannot be taken arbitrary

and excessive duties for cranage, wharfage, [etc.,] neither can they be

inhanced to an immoderate rate, but the duties must be reasonable and moder-

ate, though settled by the king’s license or charter. For now the wharf and

crane and other conveniences are affected with a publick interest, and they

cease to be juris privati only . . . .

. . . But in that case the king may limit by his charter, and license him to

take reasonable tolls, though it be a new port or wharf, and made publick; be-

cause he is to be at the charge to maintain and repair it, and find those conven-

iences that are fit for it, as cranes and weights.

From this short passage, so is the business of general rate regulation born.

The owner of these wharfs and cranes, or what became to be known as “essen-

tial facilities,”

is first subject to a limit on what he can charge. But by the

same token, he is entitled to receive sufficient revenues to allow him to main-

tain his business. That basic condition in turn requires that he be allowed to re-

cover the costs of his operation, both sunk and continuing, and make a competi-

tive rate of return on that investment, lest he take his capital elsewhere. In

modern terms, the obligation is to respect the commitments to charge reasona-

ble and nondiscriminatory rates, typically called RAND or FRAND obliga-

tions, where the F stands for “fair.”

The former requirement is to squeeze out

the monopoly profits without exposing the regulated industry to the risk of con-

fiscation. The latter requirement is intended to make sure that the monopolist

does not use its power over price to discriminate between more and less fa-

vored customers, including such divisions by race.

28. Id. at 77-78. I talk about these developments at great length in Richard A. Epstein,

The History of Public Utility Rate Regulation in the United States Supreme Court: Of Rea-

sonable and Nondiscriminatory Rates, 38 J. SUP. CT. HIST. 345, 346-50 (2013).

29. See Phillip Areeda, Essential Facilities: An Epithet in Need of Limiting Principles,

58 A

NTITRUST L.J. 841 (1989) (discussing the essential facilities doctrine).

30. See, e.g., Microsoft Corp. v. Motorola, Inc., 696 F.3d 872, 876, 877 & n.2 (9th Cir.

2012) (using and explaining this terminology).

1252 STANFORD LAW REVIEW [Vol. 66:1241

I have written at great length of the many pitfalls that befall any effort to

set the rates needed to avoid the twin risks of monopolization by the firm and

confiscation by the state.

But in these instances, much of the difficulty is

eliminated by the nature of the problem. In dealing with routine passenger traf-

fic, there is no reason to discriminate in the treatment that is given to two cus-

tomers on the same train or bus. Any system of rate regulation can set the ap-

propriate tariffs. But unlike cases where the cost of service depends on the na-

nature of the goods that are shipped, there is a strong presumption that there is

no reason to vary the rates charged to different customers in order to prevent

any form of cross-subsidization, which, although widely practiced, is not ap-

propriate here any more than it is in competitive markets. The straightforward

application of the nondiscrimination rule offers a powerful response to the dan-

gers of racial segregation, without having to resolve the separate question of

what rates should be regarded as reasonable. It is for that reason that the prob-

lem of setting rates for ordinary passengers is relatively easy, at least compared

to the business of setting rates for public utilities, where different classes of us-

ers require, arguably, very different forms of treatment because of major differ-

ences in the cost of providing service to them. To be sure, any modern transpor-

tation (or hotel or leasing) system takes steps to allow rates to shift in response

to changes in demand and cost, so that no one expects that all passengers sitting

on the same airplane have paid the same price for their tickets. But if time of

departure or time of ticket purchase or method of purchase matter in this mar-

ket, race surely does not. It is hard to imagine any principled exceptions to this

general rule, but, if any occur, they should be dealt with only when they arise,

and not in setting the basic business framework. For example, unruly persons

can be expelled from public transportation facilities for misbehavior, and if

they engage in repeated offenses, they can be barred from use of the facilities

altogether, even if that expulsion policy has some differential effect by race. On

matters of sex, the historical record is a bit more complicated. In some docu-

mented instances, the boorish behavior of male passengers justified the use of

special ladies’ cars to forestall various forms of sexual harassment.

The topic of common carriers deals extensively with various exercises of

monopoly power, and thus overlaps with the topics of antitrust that preoccupied

Bork at the time he wrote his New Republic article. To be sure, the antitrust so-

lutions do not work well with common carriers. One major office of antitrust

law is intended to prevent cooperation between firms that can operate as inde-

pendent entities in competition with each other, which is why controlling car-

telization is the chief objective of that body of law, as Bork himself repeatedly

31. Epstein, supra note 28, at 346-50.

32. See BARBARA YOUNG WELKE, RECASTING AMERICAN LIBERTY: GENDER, RACE,

LAW, AND THE RAILROAD REVOLUTION, 1865-1920, at 280-82 (2001).

June 2014] PUBLIC ACCOMMODATIONS 1253

stressed.

But what is astonishing is that Bork never connected the dots be-

tween these two areas of law. The simple point here is that the nondiscrimina-

tion provision should only apply to those cases where firms exert monopoly

power over certain markets, which includes the usual forms of public transpor-

tation. But elsewhere, doctors, lawyers, plumbers, hair stylists, and a thousand

other occupations, none of which are mentioned by Bork, should in principle be

out from underneath the antidiscrimination laws. Indeed, it is standard econom-

ic theory that sellers cannot price discriminate in competitive markets with

standard goods because the disadvantaged purchasers will shift to another sell-

er. To be sure, there are some tiny frictions in any real market, but these are so

small that they can be safely disregarded in this context. It is therefore clear

that we do not have to worry about the specter of forced association in the

common trades with which Bork was so clearly concerned. He should have ac-

cepted the antidiscrimination principle of Title II, so long as it was confined to

the monopoly-like situations to which it originally applied at common law.

Stated otherwise, the general antitrust law is itself a principled limitation on the

basic tenet of freedom of association insofar as it tells producers in the same

industry that they cannot collude to reduce output and to raise prices. There is

no reason why Bork should not have accepted that same principle for common

carriers. It was his then-extreme version of libertarianism, not in evidence in

the antitrust context, which got him into much of his trouble on questions of

principle.

It is, of course, the case that Bork repudiated his 1963 article in his book

The Tempting of America, where he wrote: “My position was incorrect be-

cause . . . there are no general principles to decide competing claims of associa-

tion and nonassociation.”

As was often typical of Bork, he leaped from one

extreme to the other. In this context, the monopoly control position offers a

principled way to deal with the rival claims. By moving to a kind of theoretical

skepticism about this issue, Bork adopted a position that links all too well with

his later affection for pure majoritarianism. If it turns out that no theory can re-

solve this profound difference, then the political majority might as well have

the last say on that issue, at least until the winds of fortune shift in the opposite

direction. In contrast, the monopoly control theory does not shift with the polit-

ical winds, an important consideration given the massive shifts in political

power since the passage of the Civil Rights Act of 1964, as is evident from the

most cursory account of the historical realities leading up to the passage of that

law.

33. Bork, supra note 24, at 7.

34. ROBERT H. BORK, THE TEMPTING OF AMERICA: THE POLITICAL SEDUCTION OF THE

LAW 80 (1990). My thanks to Brian Landsberg for this useful reference.

1254 STANFORD LAW REVIEW [Vol. 66:1241

B. The Historical Reality

As is evident from the text of Title II, its focus on luncheon counters in res-

taurants—small businesses in highly competitive industries—indicates the

weak fit between the traditional theory behind the antidiscrimination principle

and the institutions to which it was targeted. If all that mattered was pure theo-

ry, individual owners of various diners could discriminate to their hearts’ con-

tent, as their customers would simply move over to some rival establishment

down the street that was prepared to extend them service. Yet it is very clear

that pure theory does a very poor job describing the historical practices of the

Old South, where, to say the least, the prospect of free entry as a check against

monopoly power was a snare and a delusion. It is important to ask the question

of why the standard economic theory of competition failed so abjectly in prac-

tice. That failure does not come from technical mistakes within the theory. Ra-

ther, the explanation depends critically upon the totalitarian nature of the Old

South, and much of the rest of the United States, which undercut the assump-

tion of free entry and exit on which the economic theory rests.

The point here is quite simple. So long as people are free to leave, they

cannot be exploited by those who wish to continue to employ their services.

Conversely, free entry into the South by Northern employers could have bid up

the wages for labor above the rates that Southern employers acting in concert

(and backed by the force of the state), the Ku Klux Klan, and a host of informal

groups and gangs could have set. The basic insight is that a competitive market

will work well in both directions because the fortunes of an oppressed group

under open market conditions depend only on the attitudes of those who are

most favorably disposed to it. The opinions of others groups do not matter since

they are no longer in the hiring market. The conditions are very different when

either private violence or political power sets wages and other employment

terms. With the former, it is the least favorably disposed group that can deter-

mine the terms of trade. With the latter it is, roughly speaking, the median

voter.

These avenues of political control were effectively exploited in the South

prior to 1964. I wrote a short account of these laws in my 1994 article Standing

Firm, on Forbidden Grounds, which answered my many critics as follows:

The Emigrant-Agent Laws imposed heavy taxes on those agents who sought

to encourage black labor to leave the South for more gainful employment

elsewhere. If there were no market pressures operating for the exit, then pre-

sumably these statutes would have been unnecessary. As it was, they were un-

able to stem the flow of migrant labor from the South, but not for want of try-

ing. Vagrancy laws were used as a means for forcing blacks to enter into labor

contracts in the first place, by attacking so-called idleness, and contract-

enforcement laws in Mississippi “required Negroes to enter into labor con-

June 2014] PUBLIC ACCOMMODATIONS 1255

tracts by a specific day each January,” and laws in other Southern states re-

quired employers to grant discharge to their black laborers before they could

be hired by other employers.

The standard blackboard account of markets does not delve into the opera-

tion of these rules but assumes, if only implicitly, that public institutions are

well ordered. In this context, “well ordered” means that these institutions en-

force the background norms of tort and criminal law so as to prevent various

forms of private aggression against all citizens within the state, all of whom

have, of course, full rights of access to the political system to help choose the

people who will enact and enforce the laws. It also assumes that the major

agencies of the state distribute the essential services such as gas, electricity, and

telephone service on equal terms to all citizens. It further assumes that state

public institutions will not turn a blind eye to private acts of violence that are

intended to prevent other individuals from exercising their normal rights of as-

sociation in the marketplace. The dysfunctional nature of Southern labor mar-

kets needs no further elaboration. No one can claim that this system of South-

ern governance was “well ordered.”

The stress on aberrant social institutions is not in evidence in the contribu-

tion of Samuel Bagenstos to this Symposium. Instead, his defense of Title II

depends on his “postrealist” views of private property when he writes:

In a postrealist world, we understand that regulation is not incompatible with

private ownership. Laws delimiting the rights and obligations of property

owners and those with whom they deal do not, in Paul’s phrasing, make the

government the owner of the property. Indeed, the institutions of property and

contract depend on background legal rules delimiting those rights and obliga-

tions and enforcing them in cases of breach.

The inspiration for this passage was Rand Paul’s objection to Title II of the

Civil Rights Act on pure freedom of association grounds. At the time of Paul’s

statements,

and since then,

I have taken Senator Paul to task on exactly the

35. Richard A. Epstein, Standing Firm, on Forbidden Grounds, 31 SAN DIEGO L. REV.

1, 42-43 (1994) (footnotes omitted) (quoting William Cohen, Negro Involuntary Servitude in

the South, 1865-1940: A Preliminary Analysis, 42 J.S. HIST. 31 (1976), reprinted in

MERICAN LAW AND THE CONSTITUTIONAL ORDER: HISTORICAL PERSPECTIVES 317, 322

(Lawrence M. Friedman & Harry N. Scheiber eds., enlarged ed. 1988)).

36. Samuel R. Bagenstos, The Unrelenting Libertarian Challenge to Public Accom-

modations Law, 66 S

TAN. L. REV. 1205, 1222 (2014).

37. Richard A. Epstein, Rand Paul’s Wrong Answer, FORBES (May 24, 2010,

12:48 PM), http://www.forbes.com/2010/05/24/rand-paul-rachel-maddow-opinions-column

ists-richard-a-epstein.html.

38. Richard A. Epstein, My Rand Paul Problem: Why Classical Liberalism Is Superior

to Hard-Core Libertarianism, D

EFINING IDEAS (Feb. 3, 2014), http://www.hoover.org/

publications/defining-ideas/article/167496. It turns out that Senator Paul does not categori-

cally oppose the income tax, which on that issue at least means that he is not a hard-core lib-

ertarian, although his proposal is far less progressive than the current law. See Richard Ep-

1256 STANFORD LAW REVIEW [Vol. 66:1241

two points that are in issue here—the economic structure of common carriers

and the deplorable institutional arrangements in the United States at the time of

the passage of the Civil Rights Act. Bagenstos points out that property and con-

tract depend on legal rules for their articulation and enforcement, but the con-

ventional rules of property and contract promote the protection of individual

zones of autonomy and the security of exchange, both of which were utterly ab-

sent in the South. Unfortunately, civil rights laws displace these beneficial

common law institutions with a set of rules on forced exchange in cases where

monopoly issues don’t apply. The objection to these rules is that they lead to

negative-sum games, not that they regulate property. The thought that forced-

association laws “may serve broader interests in democracy, freedom, and the

operation of a system in which individuals have an opportunity to acquire and

exchange property”

is whistling in the dark. Market competition poses no

threat to democratic institutions. But partisan struggles over forced associations

do.

Civil rights laws grounded in the problem of monopoly power supply a far

firmer basis for economic and political stability alike than the misinformed

postrealist critique.

Far from any universal structural defect in competitive

markets, these background considerations of institutionalized segregation make

it painfully clear that competitive markets in the standard sense are not easily

maintained or supported. Start with a voting system that is wholly skewed to

members of one race who bear extensive levels of animus to individuals of an-

other race. Any allusion to the system of equal justice before the law is an illu-

sion in the face of systematic political domination. The reason why Justice John

Marshall Harlan mounted so powerful an appeal to the norm of “color-blind”

justice in Plessy v. Ferguson is that this elementary principle had been violated

in connection with the imposition upon carriers and customers alike of the re-

quirement that black and white passengers be carried in separate railroad cars,

or at least in separate sections of the same car.

The imposition of

antimiscegenation laws is a violent affront to the ordinary principles of freedom

stein, Re: My Rand Paul Problem, RICOCHET.COM (Feb. 11, 2014, 8:57 AM),

http://ricochet.com/main-feed/Re-My-Rand-Paul-Problem.

39. Bagenstos, supra note 36, at 1222.

40. For one defense of this proposition, see RICHARD A. EPSTEIN, HOW PROGRESSIVES

REWROTE THE CONSTITUTION 84-111 (2006) (explaining the damage that cartelization has

imposed on labor and agricultural markets). To see how rate regulation done right can ad-

dress the monopoly problem, see Epstein, supra note 28, at 345-46.

41. 163 U.S. 537, 559 (1896) (Harlan, J., dissenting) (“Our Constitution is color-blind,

and neither knows nor tolerates classes among citizens. In respect of civil rights, all citizens

are equal before the law. The humblest is the peer of the most powerful. The law regards

man as man, and takes no account of his surroundings or of his color when his civil rights as

guaranteed by the supreme law of the land are involved.”), overruled by Brown v. Bd. of

Educ., 347 U.S. 483 (1954).

June 2014] PUBLIC ACCOMMODATIONS 1257

of association, which apply as much (if not more) to marriage as to any other

relationship. The simple point here is that the exercise of state power was ex-

erted in a race-conscious fashion that blocked rights of free association of all

citizens white and black.

Unfortunately, the dangerous consequences of Plessy are often obscured by

extensive discussions of the then-fashionable tripartite distinction between civ-

il, social, and political rights to which Bagenstos devotes much of his Essay. He

quotes Mark Tushnet for the basic distinction as follows: “The core civil rights

included the rights to sue and testify; social rights included the right to select

one’s associates; voting was the central political right.”

As an analytical mat-

ter, the civil-social distinction is not quite right because it has nothing to do

with the line between legal and social obligations. The more accurate line

seems to be that in this context, civil rights refers to rights that deal with the en-

forcement of claims within the legal system, and social rights with the ability to

exercise rights of free association that people can enjoy without having to go to

court to enforce them. But the distinction is said to gain traction because of the

use to which Justice Brown put it in Plessy:

The object of the amendment was undoubtedly to enforce the absolute

equality of the two races before the law, but in the nature of things it could not

have been intended to abolish distinctions based upon color, or to enforce so-

cial, as distinguished from political equality, or a commingling of the two rac-

es upon terms unsatisfactory to either. Laws permitting, and even requiring,

their separation in places where they are liable to be brought into contact do

not necessarily imply the inferiority of either race to the other, and have been

generally, if not universally, recognized as within the competency of the state

legislatures in the exercise of their police power.

Yet this passage makes a mockery of the notion of freedom of association

implicit in the notion of social rights when it equates law “permitting, and even

requiring” these interactions. Laws permitting the freedom of association, at

least outside the common carrier situation, are consistent with the libertarian

framework, and in the South their most important application would have been

to protect blacks and whites who sought to associate together from state sanc-

tion or private abuse. But, as Bagenstos notes, the true crime of Plessy is that it

“required race segregation in public conveyances and in schools,”

a pro-

foundly anti-libertarian conception. The distinction between civil and social

rights is a sideshow. The key notion is that, properly qualified by common car-

42. Bagenstos, supra note 36, at 1210 (quoting Mark Tushnet, The Politics of Equality

in Constitutional Law: The Equal Protection Clause, Dr. Du Bois, and Charles Hamilton

Houston, 74 J. AM. HIST. 884, 886 (1987)) (internal quotation marks omitted).

43. Plessy, 163 U.S. at 544.

44. Bagenstos, supra note 36, at 1212.

1258 STANFORD LAW REVIEW [Vol. 66:1241

rier rules, the principle of freedom of association is fully protected by the

common law rules of property and contract.

It remains therefore critical to note the threats to the social rights of all in-

dividuals posed by segregation. Southern states systemically augmented their

powers by how local officials ran their local governments, especially in their

control over various common carriers and public utilities. Governments exer-

cise power over the ability of businesses and residences to access roads, elec-

tricity, gas, and water. In each of these cases, governments have control over

the kinds of institutions with monopoly power that are properly subject to gen-

eral regulation. But the regulation that is required offers open and equal access

to all persons. It is not a system that allows for political actors to deny service

covertly to those private firms that do not toe the segregationist line. One rea-

son why the Southern system remained as tight as it did was that most local

businesses supported the status quo ante and thus had nothing to fear if they

implemented segregationist policies. But at the same time, new businesses, es-

pecially from the North, that might have been tempted to enter the local market

to take advantage of depressed black wages were well advised to stay out lest

local government intrigue cut their key services at the most inopportune time.

Of course, the local utility could apologize profusely when service was disrupt-

ed or curtailed, but what remedy lies when the legal system is itself dominated

by champions of segregation?

The disruption of public services was in turn supplemented by the use and

threat of private violence against those who tried to stand up to the dominant

political forces. A friendly gesture by a white person to a black friend could

lead to dangerous consequences. As youngsters in the North in the 1950s, with

no direct connection to the South, we knew well that so simple an act as pur-

chasing gas from a black-owned filling station could lead to a beating or worse.

At this point, it is only necessary to put together the pieces. The practice of

freedom of association cannot survive in a society that has corrupt electoral in-

stitutions, corrupt provision of public services, corrupt use of public force, and

unrestrained use of private violence. The hard question in these settings is to

ask exactly what legal changes should be made. In one sense, the thought that

some nondiscrimination principle could gain hold through legislation seems

laughable. Indeed, it was only because federal legislation could work, with

much huffing and puffing, to override state legislation that the local monopoly

was broken. In this regard, it is worth noting that the 1962 decision in Baker v.

Carr provides powerful evidence of the tenacity of those who held undue pow-

er to insulate themselves from institutional reform.

The party that controls the

local franchise can continue to redirect power and resources to its favored con-

45. 369 U.S. 186, 190-96 (1962) (detailing how the 1901 apportionment formula re-

mained in place for sixty years notwithstanding huge population shifts).

June 2014] PUBLIC ACCOMMODATIONS 1259

stituency. It was only by going outside the system that the legislative solution

had a chance.

There were three possible lines of attack against these entrenched segrega-

tionist institutions. The first of these was the use of private rights of action and

public prosecution against those who used force. In speaking of the remedial

issues, Bagenstos asks: “Even as of 1964, if the problem was a cartel that en-

forced discrimination by businesses via threats, violence, and harassment, why

is the proper libertarian response not to directly target the threats, violence, har-

assment, and monopoly, so that business owners will be truly free to choose

whom to serve?”

No one should claim that these direct actions should be pre-

cluded. The harder question is whether they should be regarded as sufficient, as

his question seems to imply. Bagenstos’s way of setting up the problem fails to

grasp the difficult remedial choices of under- and overenforcement. Private

rights of action against a variety of individual persons are hard to enforce, es-

pecially in a system where the levers of power are in the hands of the champi-

ons of segregation who approve of the current arrangement. The matter there-

fore required federal intervention, which worked much better than any form of

case-by-case remedy through the tort system. To be sure, there may be some

overbreadth in relying on federal intervention, but the rapid extent to which the

major retailers and restaurants fell into line showed that the broader response

was the best option. It is only a hard-line libertarian who would rely solely on

ex post damage actions or criminal sanctions in the individual case. The classi-

cal liberal approach is far more flexible on the choice of remedies.

At this juncture only two lines of attack remained, and there was no reason

not to use both of them. The first of these was to go after embedded sources of

political power. The second was to prevent its most vicious manifestations. On

the first point, it was manifestly the right decision to start the attack on segrega-

tion through a reform of the electoral system. At the time that Title I was

passed, nothing fancy was intended or needed. The explicit exclusion by race

was a staple of Southern artifice, and it did not matter whether it was done

overtly or stealthily. The practices had to stop. In dealing with these remedial

choices, the challenge is always to thread the needle between remedies that are

too tough and those that are not tough enough. This issue of remedial design is

important for answering the question, for example, of whether to have special

preclearance rights in Title I of the Civil Rights Act.

Interestingly enough, the

preclearance procedures of the Voting Rights Act of 1965,

have no analog in

46. Bagenstos, supra note 36, at 1226-27.

47. See Pub. L. No. 88-352, tit. I, § 101, 78 Stat. 241, 241-42 (codified as amended at

42 U.S.C. § 1971 (2012)).

48. Pub. L. 89-110, §§ 4-5, 79 Stat. 437, 439 (codified as amended at 42 U.S.C.

§§ 1973b-1973c), invalidated by Shelby Cnty. v. Holder, 133 S. Ct. 2612 (2013). The con-

troversies here continue with the recent Supreme Court decision in Shelby County.

1260 STANFORD LAW REVIEW [Vol. 66:1241

the context of public accommodations. Instead, the major issues of

overenforcement in the public accommodations context relate only to the

choice between public and private enforcement or some combination thereof.

Yet it would be a mistake to think that it was sufficient to prevent long-

term abuse without tackling the short-term problem of the systematic denial of

service in many places in the Deep South and elsewhere. The key point here is

to note that many of the strongest supporters of Title II were the large firms that

would be regulated by it. It is not that these firms thought it was good in itself

to surrender power to the federal government. Rather, it was their clear percep-

tion that only federal intervention could hold at bay those local officials and lo-

cal citizens that could be expected to shut down their operations by hook or

crook if the firms tried to integrate their facilities. Viewed in this light, the

dominant motive for the passage of Title II came from parties whose basic

commercial interests were undermined by segregation and who wanted gov-

ernment protection at the federal level against the depredations by public and

private forces at the local level.

Thus just as the railroads did not want to be

subject to state segregation laws in the 1890s, so in the 1960s, the next genera-

tion of businesses wanted to be out from under the thumb of those dangerous

and reactionary elements who prevented them from integrating their facilities,

which they desperately wanted to do.

Viewed from this perspective, the question with respect to Title II was

whether it went too far or did not go far enough. Even those who thought that

Ollie’s Barbecue had in principle the perfect and absolute right to turn away

customers on account of race had to recognize that the first order of business

was to make sure that those firms that wanted to integrate could not be blocked

from doing so by state officials and private violence.

In this respect, the de-

fense of Title II at the time of its enactment was on decidedly second-best

grounds: a statute that is not needed in a perfect world is much needed in an

imperfect one. The weight of history makes it impossible to leap from state-

imposed segregation to a perfectly voluntary market. In light of the past, it was

49. See Richard H. McAdams, The Origin, Development, and Regulation of Norms, 96

ICH. L. REV. 338, 375 n.127 (1997) (collecting sources).

50. See Katzenbach v. McClung, 379 U.S. 294, 304 (1964) (rejecting Commerce

Clause objections to Title II on the ground that Congress had a rational basis for finding that

racial discrimination in restaurants affected interstate commerce). The Commerce Clause

argument here, which is said to follow from Gibbons v. Ogden, 22 U.S. (9 Wheat.) 1 (1824),

bears no relationship to Gibbons, which was limited to interstate journeys. But if the stakes

were high in Wickard v. Filburn, 317 U.S. 111 (1942)—whether to uphold key provisions of

the Agricultural Adjustment Act—they were unbearable in this case. For my views on the

correct reading of the Commerce Clause, see, for example, Richard A. Epstein, A Most Im-

probable 1787 Constitution: A (Mostly) Originalist Critique of the Constitutionality of the

ACA, in T

HE HEALTH CARE CASE: THE SUPREME COURT’S DECISION AND ITS IMPLICATIONS

28 (Nathaniel Persily et al. eds., 2013).

June 2014] PUBLIC ACCOMMODATIONS 1261

better to move too fast rather than to go too slow. No one today need reject the

judgment that was made some fifty years ago.

The situation on the ground in 1964 therefore called for extensive measures

to undo a broken and corrupt system. But remember that none of those violent

episodes around the time of the passage of the Civil Rights Act of 1964 did an-

ything to undermine the basic relationships that as a matter of first principles

should govern a society with sound public institutions, including properly

elected public officials, a disinterested police force, a professional public ad-

ministration, and strict control over private violence. The politics that sur-

rounded Title II of the Civil Rights Act did not clearly distinguish between the

first- and second-best rationales for its implementation, for at the time there

was no pressing need to do so. But times change, and background conditions

can change with them. On each of the key measures—voting, public admin-

istration, policing, and private violence—the abuses were largely eliminated, in

large part due to some of the extraordinary measures imposed in the immediate

aftermath of the Civil Rights Act.

The major transformations over the past fifty years should spark a norma-

tive reexamination of the scope and impact of Title II. In this regard, it is useful

to set the two rationales for the application of Title II against the various insti-

tutions that are covered by it. The best practical argument for Title II was that it

functioned as a corrective against private force and public abuse in government.

The best theoretical argument for Title II was that an antidiscrimination rule

was needed to offset institutions that wielded monopoly power. In principle,

that rationale is as good today as the day it was uttered. But the economic land-

scape has changed in ways that have eroded the strength of traditional monopo-

lies, at least those not backed by government force. One feature of advanced

technology is that it reduces the number of firms and industries that hold mo-

nopoly power. In the Middle Ages, the one inn on the road from London to Ox-

ford held a monopoly position. But today the view that any hotel, motel, or inn

possesses monopoly power is odd in an age when customers typically move by

car, plane, or bus, with lots of choices of where to spend the night. The eating

establishments listed with such precision in Title II

also operate in a competi-

tive industry, which is surely the case with the movie theaters, concert halls,

and arenas that are found everywhere. The original justifications for civil rights

enforcement have become weaker. Yet ironically, the scope of the law has be-

come ever more extensive.

III. T

HE EXPANSION OF TITLE II AND KINDRED STATUTES

With the adoption of Title II, it is important to see how its statutory defini-

tion of a public accommodation squares up with the traditional account of a

51. See 42 U.S.C. § 2000a(b).

1262 STANFORD LAW REVIEW [Vol. 66:1241

common carrier. The expansive nature of Title II’s definition becomes clearer

when its scope is contrasted with the smaller class of businesses that were cov-

ered under Hale’s “affected with a publick interest” test before that test’s de-

mise in Nebbia v. New York in 1934.

To keep Nebbia in perspective, it is crit-

ical to contrast it with Chas. Wolff Packing Co. v. Court of Industrial Relations,

decided in 1923.

Under the conventional wisdom of the day, the classification

of a business as one affected with a public interest was thought to give the state

the power to regulate its rates so long as those rates were not confiscatory. The

precise holding in Wolff was that a small packing house in Kansas, whose wag-

es a state statute sought to regulate, was not “affected with a public interest”

because it fell into none of the traditional regulated categories.

The first two

categories involved the traditional common carriers and inns. But all the action

in this case took place with respect to a third category:

Businesses which though not public at their inception may be fairly said to

have risen to be such and have become subject in consequence to some gov-

ernment regulation. They have come to hold such a peculiar relation to the

public that this is superimposed upon them. In the language of the cases, the

owner by devoting his business to the public use, in effect grants the public an

interest in that use and subjects himself to public regulation to the extent of

that interest although the property continues to belong to its private owner and

to be entitled to protection accordingly.

After offering that definition, Chief Justice Taft, writing for a unanimous

Court, listed cases illustrating the types of concerns that were covered.

These

started with grain elevators, which were thought to have a “‘virtual’ monopoly”

in Munn v. Illinois,

a case that played a key historical role by transforming

Allnutt v. Inglis into a principle of American constitutional law.

The list also

52. 291 U.S. 502, 536-37 (1934) (“[T]here is no closed class or category of businesses

affected with a public interest . . . . So far as the requirement of due process is concerned, . . .

a state is free to adopt whatever economic policy may reasonably be deemed to promote

public welfare . . . .”).

53. 262 U.S. 522 (1923), abrogated by Nebbia, 291 U.S. 502.

54. See id. at 543 (describing the packing house); id. at 533 (describing the state stat-

ute); id. at 535 (listing the traditional regulated categories); id. at 539, 544 (suggesting that

the packing house probably does not fit into any of those categories).

55. Id. at 535.

56. Id. at 535-36.

57. 94 U.S. 113, 131 (1877); see also Brass v. North Dakota ex rel. Stoeser, 153 U.S.

391, 399-405 (1894) (applying Munn and Budd v. New York, 143 U.S. 517 (1892), to uphold

North Dakota’s regulation of grain elevators); Budd, 143 U.S. at 543 (explicitly reaffirming

Munn and upholding New York’s regulation of grain elevators); Spring Valley Water Works

v. Schottler, 110 U.S. 347, 354 (1884) (citing Munn for the proposition that local govern-

ments may regulate “one who enjoys a virtual monopoly”).

58. See Munn, 94 U.S. at 151-52 (citing Allnutt v. Inglis, (1810) 104 Eng. Rep. 206

(K.B.) 210-11; 12 East 527, 538-40) (“From this case it appears that it is only where some

June 2014] PUBLIC ACCOMMODATIONS 1263

included cases affirming state regulation of matters such as water companies

(which should in any event be lumped with public utilities), banks subject to

customer runs,

more controversially, large insurance companies doing busi-

ness in a competitive market,

and rent-controlled housing units, but only in

times of emergency.

Thus, Chief Justice Taft’s definition was much narrower

than the one that was eventually adopted in Title II.

In light of the argument developed in Part II of this Essay, the key question

then arises whether the peculiar constellation of political and legal forces pre-

vents free entry into particular geographical markets. Racially motivated zoning

laws may have those effects. But ordinarily, population mobility counteracts

any nascent form of monopoly power in ordinary business establishments. The

number of business choices is very high so that the case for Title II becomes

quite thin over many areas of its operation. Nor can it be said that a strong set

of formal and informal sanctions makes it impossible for any particular firm to

serve members of minority groups. We are long past the days when Robert Jo-

seph Pershing Foster could not find accommodations in Arizona for the night

out of fear of what others might do.

The competitive market works well when supported by well-ordered public

institutions, whose formation undercuts the case for retaining Title II. But in the

area of its original application there is no need whatsoever to take on that hero-

ic battle. Within the class of institutions originally targeted by Title II in 1964,

its effect has caused no visible inconvenience. Some of these public accommo-

dations remain common carriers, at least for some fraction of their markets, as

with certain airline and railroad routes. But even those institutions that have no

or weak monopoly power still resemble common carriers in certain key ways.

Look at the way in which admission to a movie theater or amusement park is

privilege in the bestowal of the government is enjoyed in connection with the property, that

it is affected with a public interest in any proper sense of the terms.”).

59. See Van Dyke v. Geary, 244 U.S. 39, 47-48 (1917).

60. See Noble State Bank v. Haskell, 219 U.S. 104, 113, amended by 219 U.S. 575

(1911).

61. See Ger. Alliance Ins. Co. v. Lewis, 233 U.S. 389, 411-15 (1914). I regard this

case as wrong on rates because of competition, but proper on security of premiums given the

risk of default.

62. See Block v. Hirsh, 256 U.S. 135, 155 (1921). Chief Justice Taft joined the dissent

in this five-to-four decision, see id. at 158 (McKenna, J., dissenting), which was thought to

rest on short-term emergencies but has become massively abused under New York stabiliza-

tion laws, in which the emergency comes like clockwork every three years. The current law

calls for the end of control only when the vacancy rate reaches five percent in any category.

See Emergency Tenant Protection Act of 1974, N.Y.

UNCONSOL. LAW § 8623 (McKinney

2014). The provision was last reenacted in 2011, extending the law for another four years.

Rent Act of 2011, ch. 97, § 1-a, 2011 N.Y. Laws 767, 767 (codified as amended at

NCONSOL. § 8623).

63. See WILKERSON, supra note 3, at 206-10.

1264 STANFORD LAW REVIEW [Vol. 66:1241

determined: it is strictly and solely by the ability to pay, subject to certain limi-

tations on age (for admission) and sex (for washrooms). Even though these in-

stitutions are not common carriers, their proprietors have no interest in review-

ing customer resumes before selling tickets to the latest show. Quite simply, it

is in their interest to take all comers on the same terms just as common carriers

have traditionally done. Loutish or menacing behavior can still result in ejec-

tion with cause, so that these operations proceed more or less without a hitch.

Why repeal a benign statute in order to raise symbolic hackles? The contrast

between the common applications of Title II and the intrusive and mischievous

preclearance procedures under sections 4 and 5 of the Voting Rights Act could

not be more vivid.

The potential application of Title II is fraught with far greater risk, howev-

er, when enforcement challenges organizations that do not operate lunch coun-

ters and movie theaters, for in these cases rival interests come into play. That

principle was recognized in the run-up to Title II in connection with one hard-

fought exemption from Title II: an exception for “an establishment located

within a building which contains not more than five rooms for rent or hire and

which is actually occupied by the proprietor of such establishment as his resi-

dence.”

Each of these specific conditions was the result of a hard-fought

compromise over the status of Mrs. Murphy’s much-debated boarding house.

To get the pulse of the times, read this uneasy account of the struggle from the

vantage point of the passionate defender of Title II:

Already there was grave concern over the wide-spread newspaper talk that

the public accommodations section would be gutted, possibly by an exemption

for “small” public accommodations. “Mrs. Murphy’s boarding house” with a

few rooms was one thing, for her right of privacy cut across the [black per-

son’s] right to a room. But a general exception for all small public accommo-

dations was something else again. A [black] laborer entering a small diner

could be quite as hungry as a [black] banker seeking service at the Waldorf

[Hotel in New York].

The scope of that concession was limited to situations where people were

justly concerned about their privacy and safety when asked to live and work in

close proximity with each other. People do check references before accepting

boarders. The point of this exception under Title II was to carve out the cases

where personal preferences matter most, while leaving subject to the law the

large inn or hotel, where the impersonal nature of the business makes for few of

64. Civil Rights Act of 1964, tit. II, § 201(b)(1), 42 U.S.C. § 2000a(b)(1) (2012).

65. Joseph L. Rauh, Jr., The Role of the Leadership Conference on Civil Rights in the

Civil Rights Struggle of 1963-1964, in T

HE CIVIL RIGHTS ACT OF 1964: THE PASSAGE OF THE

LAW THAT ENDED RACIAL SEGREGATION 49, 55 (Robert D. Loevy ed., 1997) (alterations in

original).

June 2014] PUBLIC ACCOMMODATIONS 1265

these close interactions. Basically, the deal worked, and the exception has had

no significant role to play.

The same type of concern was also raised with respect to “a private club or

other establishment not in fact open to the public, except to the extent that the

facilities of such establishment are made available to the customers or patrons

of an establishment within the scope of subsection (b) [which defines the insti-

tutions covered by Title II].”

Clubs are different beasts because they all have

distinct organizational objectives, membership rules, and complex internal gov-

ernance structures. These organizations never relate to their members in the

same way that hotels, restaurants, and movie theaters relate to their customers.

In clubs and other voluntary organizations, people care a lot about whom they

associate with. The type of open seating that works on airplanes won’t work in

organizations that choose to assign seats at their annual banquet. No defender

of the nondiscrimination provision believes that all these organizations should,

or could, run on the same principles as the traditional business.

The original application of public accommodations laws to common carri-

ers, narrowly defined, covers those cases where claims for associational prefer-

ences are at their weakest. At that point, the social losses, if any, created by im-

posing a nondiscrimination norm hardly seem to be of any concern, given the

smooth implementation of the program once the restrictions were in fact put

into place. Indeed, it is hard to tell whether Title II protected the preferences of

individuals who opposed segregation but were afraid to speak out, whether it

added legitimacy to the nondiscrimination principle, or whether it just matched

the change in public sentiments reflected by the widespread and growing politi-

cal support for the statute itself.

The next question is how the broad reach of the new antidiscrimination

principle plays out in other cases that are covered by a very broad definition of

what counts as a public accommodation. Take the example of private clubs, or-

ganizations which are generally open only to members, which means that

someone has to take the time to decide who is in and who is out. The member-

ship process itself gives firm evidence that people care more about who is a

member of their club or church than they do about the identity of the person

next to them in the checkout line. Put otherwise, the question is whether these

private institutions should be treated like the common carriers of old when the

arguments of freedom of association that cut against the characterization grow

far stronger. Not only is there a complete absence of monopoly power, but

there is also a concern with internal operations that just does not arise in the

earlier civil rights cases. Nonetheless, the next generation of antidiscrimination

cases ignored these critical differences.

66. § 201(e), 42 U.S.C. § 2000a(e).

1266 STANFORD LAW REVIEW [Vol. 66:1241

IV. FROM PUBLIC ACCOMMODATIONS TO HUMAN RIGHTS

A. Jaycees and Boy Scouts

The first of the new wave of cases was the 1984 decision in Roberts v.

United States Jaycees, which involved a decision by the Commissioner of the

Minnesota Department of Human Rights to strike down the Jaycees’ decision

to limit itself to male members only.

The moniker “human rights” shows how

far the law has moved from the original concern of the nondiscrimination rule

for common carriers developed at common law. Under the new definition, hu-

man rights do not include rights of property and free association, but impose an

obligation not to discriminate against outsiders in certain businesses or accom-

modations

—including retail stores,

which are not explicitly covered by Ti-

tle II.

In his Essay, Bagenstos claims, “As Joseph Singer has shown exten-

sively, the common law doctrine before the Civil War in many jurisdictions at

least plausibly prohibited any discrimination by any business holding itself out

as serving the public.”

Bagenstos, however, provides no citation to a particu-

lar passage that supports that conclusion, which Singer tentatively defends on a

mix of normative and historical grounds that in the end are not sufficient to dis-

place the common view to the contrary.

It is also unclear what it means under

67. 468 U.S. 609, 614-15 (1984).

68. The Minnesota Human Rights Act at issue in Roberts, for example, imposed an

obligation not “[t]o deny any person the full and equal enjoyment of the goods, services, fa-

cilities, privileges, advantages, and accommodations of a place of public accommodation

because of race, color, creed, religion, disability, national origin or sex.” Id. at 615 (quoting

INN. STAT. § 363.03(3)) (internal quotation marks omitted).

69. Minnesota’s statute applied (and continues to apply) to every “business, accom-

modation, refreshment, entertainment, recreation, or transportation facility of any kind.” Id.

(quoting MINN. STAT. § 363.01(18)); accord MINN. STAT. § 363A.03(34) (2013) (retaining

that same language).

70. See Singer, supra note 16, at 1288-93.

71. Bagenstos, supra note 36, at 1225.

72. Singer writes:

Businesses other than inns and carriers dealt in necessities such as salt, food, materials to

make clothes, and services such as medical care. Denial of such goods and services would

have caused great hardship precisely because there was often no more than one general store

or doctor in the area, thereby constituting as much of an effective monopoly as the inn or

stagecoach. In addition, many businesses other than inns and common carriers were required

to obtain licenses or franchises from the state in order to operate.

Singer, supra note 16, at 1292-93. But even if there were no close substitutes, the services in

question were not of the standardized variety.

It is exceedingly doubtful that the physician was treated as a common carrier. See, e.g.,

Hurley v. Eddingfield, 59 N.E. 1058, 1058 (Ind. 1901) (“In obtaining the state’s license

(permission) to practice medicine, the state does not require, and the licensee does not en-

gage, that he will practice at all or on other terms than he may choose to accept. Counsel’s

June 2014] PUBLIC ACCOMMODATIONS 1267

this common formulation for any business to “hold[] itself out as serving the

public.” The standard rule on common carriers prohibited them from making

any contrary public reference. But as to ordinary retail shops, if they posted a

sign regarding whom they refused to serve, that action itself might well have

removed them from the class of public accommodations.

But the legal relevance of this diversion is uncertain. Even though it is

crystal clear that the Minnesota Human Rights Act covers retail establishments,

the fit between the Jaycees and the statute still looks less than ideal, given the

objectives of the Jaycees to “foster the growth and development of young

men’s civic organization in the United States.”

The Jaycees also entertained a

variety of similar objectives, which surely qualify it for treatment as a tax-

exempt organization under § 501(c)(4) of the Internal Revenue Code, which

applies to “[c]ivic leagues or organizations not organized for profit but operated

exclusively for the promotion of social welfare.”

Notwithstanding the Jaycees’ extensive social program, the Commissioner

took the position that the Jaycees was covered by the Minnesota Human Rights

Act.

The Jaycees then mounted its constitutional challenge against the Act in

federal court, which in turn certified the question to the Minnesota Supreme

Court. The state supreme court held that the Jaycees was covered by the statute

because the Jaycees organization

(a) is a “business” in that it sells goods and extends privileges in exchange for

annual membership dues; (b) is a “public” business in that it solicits and re-

cruits dues-paying members based on unselective criteria; and (c) is a public

business “facility” in that it conducts its activities at fixed and mobile sites

within the State of Minnesota.

The manifest differences between organizations like the Jaycees, with their

substantive programs, and the standard movie theater is too evident to require

any serious discussion, for these organizations are not just open to the public

upon payment of a fee. When the case reached the U.S. Supreme Court, Justice

Brennan was alert to the serious issues raised, but he nonetheless overruled the

Eighth Circuit, which had struck down the statute,

by holding that the Jaycees

fell on the public accommodations side of the line. In Justice Brennan’s view,

the only escape from that classification involved intimate associations on the

ground that only family relations give rise to “deep attachments and commit-

analogies, drawn from the obligations to the public on the part of innkeepers, common carri-

ers, and the like, are beside the mark.”). Hurley is not cited in Singer’s article.

73. Roberts, 468 U.S. at 612 (quoting the Jaycees’ bylaws).

74. I.R.C. § 501(c)(4)(A) (2012).

75. Roberts, 468 U.S. at 615.

76. Id. at 616 (quoting U.S. Jaycees v. McClure, 305 N.W.2d 764, 768-74 (Minn.

1981)).

77. See U.S. Jaycees v. McClure, 709 F.2d 1560, 1561 (8th Cir. 1983), rev’d, 468 U.S.

609.

1268 STANFORD LAW REVIEW [Vol. 66:1241

ments” that require a “high degree of selectivity” and “seclusion” and thus mer-

it legal protection.

He concluded:

As a general matter, only relationships with these sorts of qualities are likely

to reflect the considerations that have led to an understanding of freedom of

association as an intrinsic element of personal liberty. Conversely, an associa-

tion lacking these qualities—such as a large business enterprise—seems re-

mote from the concerns giving rise to this constitutional protection.

Yet why? No one quarrels with the view that intimate personal associations

should not be subject to any antidiscrimination norm. Could it really be a crim-

inal offense to deliberately refuse to marry a person because of his or her reli-

gion or national origin? But nothing that Justice Brennan wrote addressed ei-

ther of two key questions about the extension of that principle beyond these

core cases. Thus he did not explain why this law should go beyond the standard

definition of public accommodations to ordinary businesses that operate in

competitive environments. Nor did he explain why the law, if it does go beyond

those businesses, should extend to organizations like the Jaycees. In dealing

with both questions, the relationship of employer to employee, and of employ-

ees to each other, is worlds apart from the standard firm-to-customer relation-

ship. It is for just this reason that the provisions regulating discrimination in

employment are, to say the least, far more difficult to apply than those set out

in Title II.

As a working generalization with many exceptions, civic leagues like the

Jaycees have to govern relationships of large numbers of individuals with wide

variations in tastes and sentiments. Any decision to let the state force an associ-

ation of one person or group on an organization necessarily has serious nega-

tive consequences for the persons within the organization who lose some con-

trol over their organization when subjected to this state imposition. The range

of sentiments and tastes is vastly expanded by the change in membership,

which makes it harder to reach consensus on matters of common concern. The

original members, who may have given years of service to their organization,

are thus left with the choice between running an operation in ways that com-

promise their principles, leaving the organization, or shutting it down al-

together.

Yet the need to permit one group to break down the doors of another insti-

tution cannot be justified on the ground that they have no place else to go.

There are no transactional barriers to letting the new members in, and just that

change in membership policy would happen if in fact the change in composi-

tion were a win-win situation for the outside applicants and the current mem-

bers. The fact that this transformation does not happen across the board is evi-

dence that the Minnesota Human Rights Act imposes win-lose types of deals,

78. Roberts, 468 U.S. at 619-20.

79. Id. at 620.

June 2014] PUBLIC ACCOMMODATIONS 1269

which are always harder to maintain given that one side is always intent on al-

tering the deal in order to improve its terms of trade. The results are not unique

to these so-called public accommodations, but also extend to any organization

where the mandate is imposed on the one side for the benefit of the other. This

happens, for example, when employers have to deal with unions on a good faith

basis, even when the union has (as the exclusive representative of the work-

ers

) monopoly power over the firm, including the power to abrogate preexist-

ing contracts with workers.

A similar result happens under rent-control stat-

utes that require landlords to renew their leases against their will with terms

that promise them far less by way of rent than a market rate. Indeed, we should

expect nothing less than this kind of antisocial behavior. The party subject to

the imposition is right to protest the use of force directed its way, so that it will

use every means within the law (and, regrettably, some beyond it) to rid itself

of a losing contract, either by breaking off the relationship or changing its terms

of trade. It is for this reason that ordinary contracts at will, which are termina-

ble at any time by either side, routinely have great durability. A set of constant

incremental adjustments allows both sides to share in the gains, which reduces

the desire of either side to defect from the agreement.

Of course these con-

tracts often do dissolve as one party finds that it no longer wants to deal with

the other. But these are clean breaks that are not marked by prolonged struggle

involving strikes, lockouts, and other refusals to deal, which are routine in labor

law cases.

Quite simply, external force by the state is always necessary to

keep win-lose arrangements from blowing apart, which introduces new levels

of social uncertainty and new layers of administrative oversight.

The question is whether there is reason to incur these long-term destabili-

zation costs. Letting one organization exclude potential members does not pre-

vent outsiders from joining or forming countless organizations that do not use

sex or race or any similar criterion for admission. Indeed, it was surely the case

that many members of the Jaycees in 1984 were also members of organizations

that admitted women as full and equal members. There were in all likelihood

many women’s organizations that did not admit men on equal terms. It is no

credit to the United States to limit by legislation the diversity and types of or-

ganizations that can be formed for all sorts of nonbusiness reasons to those

80. 29 U.S.C. § 159(a) (2012).

81. See J.I. Case Co. v. NLRB, 321 U.S. 332, 337 (1944) (imposing a collective bar-

gaining agreement on dissenting employees by abrogating preexisting contracts).

82. See generally Richard A. Epstein, In Defense of the Contract at Will, 51 U. CHI. L.

REV. 947 (1984).

83. See, e.g., Boys Mkts., Inc. v. Retail Clerks Union, Local 770, 398 U.S. 235, 237-

38 (1970) (enforcing a no-strike clause by injunction, notwithstanding the anti-injunction

provisions of the Norris-LaGuardia Act); Am. Ship Bldg. Co. v. NLRB, 380 U.S. 300, 301-

02, 318 (1965) (allowing an employer to “lock[] out” his employees during a labor dispute in

order to bring economic pressure in support of his bargaining position).

1270 STANFORD LAW REVIEW [Vol. 66:1241

groups which meet the approval of the Minnesota Commissioner of Human

Rights, who shows scant respect for the rights of individuals to form their own

organizations as judged by their own rules and standards.

Consistent with this view, it would be a deep mistake to require any large

all-women association to admit men. Philosophically, it could well be more

dangerous to decide that such organizations were entitled to a pass from the

general antidiscrimination law because of the legacy of discrimination against

women, which is then viewed as creating a historic debt that can never be paid

off. The world is a better and more confident place if voluntary organizations

can choose their own admission requirements and change them in accordance

with the wishes of their membership. Indeed, in a world of free association, no

organization is exempt from changing social pressures, and each must adjust its

membership provisions to take into account that social risk. At a guess, all-men

organizations pay a higher social price for exclusion than comparable groups

for women. The members and boards of such organizations have to think long

and hard before they decide to reduce the potential membership base of their

organizations. And should they make the decision to preserve their exclusive

membership lists, the only persons who should be in a position to challenge that

decision are their own members, in accordance with their internal rules. In a

heavily competitive environment, the fact that some group is able to become

and stay large only means that it has adopted a successful formula for govern-

ing its internal affairs. Its success does not mean that it has to forfeit control

over the policies that have guided it to date.

The decision to treat the Jaycees as a business was made in large part be-

cause Justice Brennan had his eye on the status of large business operations that

are right now subject to extensive antidiscrimination norms. Justice Brennan

was well aware that the principle of freedom of association finds no place un-

der either Title VII of the Civil Rights Act of 1964 or the collective bargaining

procedures outlined under the National Labor Relations Act.

There is no mo-

nopoly justification for either of these flat restrictions on the principle of free-

dom of association;

nor is there any easy way to administer the appropriate

good faith standards of negotiation. There are of course vast differences in how

the two statutes operate, and each therefore can be subjected to forms of specif-

84. Pub. L. No. 74-198, 49 Stat. 449 (1935) (codified as amended at 29 U.S.C. §§ 151-

169).

85. See, e.g., Coppage v. Kansas, 236 U.S. 1, 11-14 (1915) (striking down a state law

that required employers to engage in mandatory collective bargaining with unions); Adair v.

United States, 208 U.S. 161, 172-73 (1908) (holding that the Fifth Amendment protects the

right to enter or refuse to enter into contracts). For my defense of these decisions, see Rich-

ard A. Epstein, A Common Law for Labor Relations: A Critique of the New Deal Labor Leg-

islation, 92 Y

ALE L.J. 1357, 1370-75 (1983).

June 2014] PUBLIC ACCOMMODATIONS 1271

ic criticisms that could not be lodged against the other.

The Civil Rights Act

often relies on disparate impact theories that have high error rates in determin-

ing improper motive. The National Labor Relations Act often provides insuffi-

cient protection for dissenting members of the bargaining unit.

Yet it is a mistake in a discussion of this sort to concentrate on these differ-

ences. What matters is the common mistakes that make both of these regimes a

threat to individual liberty and a massive drain on the productive capacities of

the residents of United States, who are duty-bound to enforce laws that make

their businesses less responsive to the needs of their shareholders, customers,

suppliers, and, all too often, their present and prospective employees.

In grappling with these thorny issues, it is clear that the dim voice of Rob-

ert Bork did not echo in the ears of Justice Brennan, who at no point sought to

apply any general test of social welfare to see if the changes wrought by the

Minnesota Human Rights Act counted as a social improvement under either the

Pareto or Kaldor-Hicks tests for social welfare.

At this juncture, moreover, it

does not matter which of these two tests is chosen, because rules like the Min-

nesota Human Rights Act generate a large set of social losses through the de-

stabilization of competitive markets with administrative systems that are more

costly, less efficient, and more vulnerable to political intrigue than the ordinary

firms that they displace. Nor is it sufficient to say that these laws produce some

unquantifiable positive externalities for third persons in society at large, since

the laws displace common law rules which produce even greater benefits by

doing something that no system of government coercion can hope to accom-

plish: namely, increasing the opportunities for voluntary trade with all sorts of

third persons. The rationale for freedom of association is not limited to intimate

or small associations. It extends to every organization under the principle that

mutual gain through cooperation is a principle that lies at the heart of a free so-

ciety. It should surely rank high on the list of human rights.

The other cases in the Roberts line illustrate the same weaknesses of its tri-

partite classification of associations. In Board of Directors of Rotary Interna-

tional v. Rotary Club of Duarte,

the Supreme Court upheld California’s Un-

86. I have raised these specific criticisms in numerous places. See generally, e.g.,

PSTEIN, supra note 1 (critiquing Title VII); RICHARD A. EPSTEIN, THE CASE AGAINST THE

EMPLOYEE FREE CHOICE ACT (2009) (critiquing the National Labor Relations Act).

87. See, e.g., Vaca v. Sipes, 386 U.S. 171 (1967) (denying workers control over their

own individual grievances). As a student, I criticized Vaca in Note, Individual Control over

Personal Grievances Under Vaca v. Sipes, 77 Y

ALE L.J. 559 (1968).

88. The last qualification is needed because unions can only survive with the support

of a majority of their workers, which leads them to adopt highly inefficient work rules, hurt-

ing some current employees and keeping out prospective employees.

89. For the standard definitions of Pareto and Kaldor-Hicks efficiency, see, for exam-

ple, R

ICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 17-20 (8th ed. 2011).

90. 481 U.S. 537 (1987).

1272 STANFORD LAW REVIEW [Vol. 66:1241

ruh Civil Rights Act, which required Rotary Clubs to admit women, over

claims based both on due process “freedom of private association” and on First

Amendment “freedom of expressive association.”

Similarly, in New York

State Club Ass’n v. City of New York, the Supreme Court, speaking through

Justice Byron White, rejected a federal constitutional challenge to New York

City’s Human Rights Law, which prohibited discrimination based on race,

creed, sex, and other grounds by any “place of public accommodation, resort or

amusement,” but specifically exempted “any institution, club or place of ac-

commodation which proves that it is in its nature distinctly private.”

That

basic requirement was in turn subject to an exemption for certain clubs with

fewer than four hundred members.

In both of these post-Roberts cases the

legislature based its human rights law on the state’s compelling state interest in

eliminating discrimination by various clubs and other public accommodations

in order to improve chances of equal opportunity for women and minorities.

Initially, it seems clear that the organizations at issue in New York State

Club Ass’n could not credibly be described as “distinctly private” as that term

is used in New York City’s Human Rights Law. But it hardly follows that this

law can be justified by the threadbare efforts to identify the “compelling state

interest” that justifies this limitation on associational freedom. There is of

course no reason to think that all clubs and groups practice discrimination on

grounds of sex or race so that some, indeed many, membership opportunities

are available to all women and minority groups, who also have, and often exer-

cise, the option to form their own exclusive groups. The freedom of association

pertains to these groups as well, so that there can be no claim of formal inequal-

ity across various lines of race and sex. Nor should the government ever take

the position that under the First Amendment the only reason to exclude indi-

viduals is because they do not share the aims and ends of the organization,

whatever those may be. It is exceedingly difficult for outsiders to understand

the culture and ethos of private organizations, which may place great weight on

the very traits that public accommodations laws regard as irrelevant. But the

question still remains why any outsiders have the right to pass judgment on the

membership rules and social behaviors of groups of which they are not a part.

Of course, no group should be immune from criticism because of its poli-

cies of either admission or exclusion. But the threshold to engage in those ac-

tivities, always open to all, is far lower than the threshold that has to be crossed

to engage in acts of state coercion that are in fact far more intrusive than the

gathering of names that was rightly invalidated in NAACP v. Alabama ex rel.

91. See id. at 544 & n.2, 545-47.

92. 487 U.S. 1, 4 & n.1, 5 (1988) (quoting N.Y.C., N.Y., ADMIN. CODE §§ 8-102(9),

-107(2)) (internal quotation marks omitted).

93. Id. at 6 (quoting ADMIN. § 8-102(9)).

94. See id. at 5-6.

June 2014] PUBLIC ACCOMMODATIONS 1273

Patterson.

In these cases, it is the wrong inquiry to ask whether the admission

of new members will influence the public positions of particular bodies, which

may or may not be the case. The key point is that any change in membership

can lead to changes in the governance structures and internal norms that current

members prize. The ultimate defense of freedom of association is that it does

not require group members to justify their actions to legislatures who may not

share in their beliefs. In most cases of compelling state interest one restricts

speech for fear of imminent violence or systematic fraud. But it is dangerous

business to assume that exclusion blocks the economic advancement of women

and minorities or that open admission advances it. Like the Jaycees, the New

York State Club Association and the Rotary Club advanced positions that prob-

ably lost internal support of group members shortly after they were announced.

But far from justifying state coercion, that process of internal transformation of

membership rules offers a reason why compulsory membership rules should

not be imposed by government: there is no compelling state interest in chang-

ing norms that often change by voluntary means, especially when none of the

clubs in question have anything close to the monopoly position that normally is

needed to justify the application of an antidiscrimination norm. The effect of

these laws is to reduce the richness and heterogeneity of our voluntary institu-

tions, without any assurance that the forced membership rules will have the

slightest effect on the desired levels of economic integration and economic par-

ity—goals that have proved as elusive after the passage of these statutes as be-

fore.

Given the way the law has developed, however, the logic of voluntary as-

sociation does not apply in any across-the-board fashion. Roberts and its prog-

eny have made it clear just how contingent freedom of association is on the

predilections of judges. Nonetheless, that willingness to extend human rights

laws to new situations has not foreclosed further development of the rules limit-

ing freedom of association. In Roberts, the Jaycees did not put much emphasis

on the distinctive nature of its operation. Chances are it would have changed

course quickly anyhow. It is harder to be confident about that judgment with

Rotary Club of Duarte. But there are other larger organizations that have a

stronger sense of mission, which does cut against the antidiscrimination law.

That clash came to a head in Boy Scouts of America v. Dale, which asked

whether the New Jersey Law Against Discrimination prohibited the Boy Scouts

from excluding James Dale, an assistant scoutmaster, on the grounds of his

95. 357 U.S. 449 (1958) (holding that the First Amendment right of freedom of asso-

ciation prevented the state attorney general from forcing disclosure of the NAACP’s mem-

bership lists).

1274 STANFORD LAW REVIEW [Vol. 66:1241

homosexual orientation.

The relevant law is broad with respect to the organi-

zations that it wishes to cover when it provides in part: “All persons shall have

the opportunity . . . to obtain all the accommodations, advantages, facilities, and

privileges of any place of public accommodation . . . without discrimination

because of . . . affectional or sexual orientation . . . .”

The New Jersey Su-

preme Court held that the Boy Scouts were indeed covered by the statute even

though, as an organization, they had no fixed place of operation.

A quick glance at the New Jersey Supreme Court’s opinion shows no

qualms whatsoever about this state-law finding, so powerful in its mind was the

antidiscrimination norm. But take the standard principles of freedom of asso-

ciation seriously, and the organization doesn’t look like a common carrier in

light of its mission statement:

It is the mission of the Boy Scouts of America to serve others by helping to

instill values in young people and, in other ways, to prepare them to make eth-

ical choices over their lifetime in achieving their full potential.

The values we strive to instill are based on those found in the Scout Oath

and Law:

Scout Oath

On my honor I will do my best

To do my duty to God and my country

and to obey the Scout Law;

To help other people at all times;

To keep myself physically strong,

mentally awake, and morally straight.

Scout Law

A Scout is:

Trustworthy Obedient

Loyal Cheerful

Helpful Thrifty

Friendly Brave

Courteous Clean

Kind Reverent.

96. 530 U.S. 640, 644 (2000). For an earlier statement of my views, see generally

Richard A. Epstein, The Constitutional Perils of Moderation: The Case of the Boy Scouts, 74

S. CAL. L. REV. 119 (2000).

97. Id. app. at 661-62 (quoting N.J. STAT. ANN. § 10:5-4). The language cited was in

effect in 2000 when Dale was decided, and the language remains the same today. See N.J.

STAT. ANN. § 10:5-4 (West 2014).

98. See Dale v. Boy Scouts of Am., 734 A.2d 1196, 1210-11, 1230 (N.J. 1999), rev’d,

530 U.S. 640; see also Dale, 530 U.S. at 646, 661 (construing and reversing the New Jersey

Supreme Court’s opinion).

99. Dale, 530 U.S. at 649 (internal quotation marks omitted) (quoting the Boy Scouts’

mission statement).

June 2014] PUBLIC ACCOMMODATIONS 1275

Not many railroads could remain in business if their passengers all had to

sign on to that program. Nor can one claim with a straight face that the Boy

Scouts looks like an ordinary for-profit business. Nor is it possible to subscribe

for a second to the odd property-based view that all that Dale sought from the

Boy Scouts was some form of an “easement” to join their organization.

100

easement only allows entrance onto the property that would otherwise amount

to a trespass. It permits, for example, individuals a right of way to walk or ride

across the land of another without recrimination. But no easement has ever al-

lowed a party to participate in the operation or governance of a servient tene-

ment, which is exactly the claim that Dale made against the Boy Scouts.

It is necessary to take a step back to see why the claim of freedom of asso-

ciation is strong here: just ask what would happen if whites in Alabama had

claimed an “easement” to join the NAACP, which would have been a thousand

times worse than the actions demanded in that case by the state’s attorney gen-

eral—disclosure of membership lists that he could easily use to gin up grounds

for retaliation against the membership.

101

It was the perception of that risk

which made the Court’s decision to reverse the Alabama court’s order one of

the most warmly greeted of its time, even though it is a bit of a stretch to

wrench the notion of “freedom of association” out of the constitutional provi-

sion that protects the right of the people “peaceably” to assemble.

102

The better

way to approach that issue is to posit that the right to freedom of speech only

makes sense if people can cooperate in a common venture to create and ad-

vance that speech. Accordingly, the First Amendment protection is linked to

the standard justifications for freedom of association, which allow individuals

to pool their resources in all lawful endeavors through joint effort.

This point gains real immediacy in this case because we know that the Boy

Scouts is an organization that is selective about its choice of members and the

value structures to which they adhere. To treat an organization that has exten-

sive social and instructional activities as a common carrier is even more

strained here than it was with Roberts, given the huge internal divisions within

the organization regarding whether to keep the ban on gay scouts and troop

leaders. At this point the losses to the insiders are truly large, which is one rea-

son why its internal blocs stand their ground, even against each other. The ul-

timate question therefore was whether New Jersey could apply its aggressive

antidiscrimination law against the Boy Scouts, as the state court was prepared

to do, taking note of the costs of discrimination on those excluded, but without

reckoning on the benefits to the organization itself.

100. Cf. Louis Michael Seidman, The Dale Problem: Property and Speech Under the

Regulatory State, 75 U.

CHI. L. REV. 1541, 1552-53 (2008) (proposing this view of Dale and

discussing its implications).

101. See NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 466 (1958).

102. See U.S. CONST. amend. I.

1276 STANFORD LAW REVIEW [Vol. 66:1241

When the issue came to the Supreme Court, the calibers of analysis under

modern constitutional law were changed. The decision of the New Jersey Su-

preme Court was reversed by a five-Justice majority because the Boy Scouts

was not, in the Justices’ eyes, some nondescript, no-account organization like

the Jaycees. Given the bifurcated structure of American constitutional law, the

Boy Scouts could claim a higher level of protection as an expressive organiza-

tion by virtue of its explicitly moral orientation, which raised the case, perhaps,

from the area of undifferentiated social liberties to that of a preferred freedom.

In dealing with this issue, Justice Stevens in his dissent exhibited the fatal con-

fusion of claiming that only an organization that is passionate about its beliefs

is entitled to claim special protection on freedom of association grounds. Un-

fortunately, that position then creates the unpleasant absurdity that a change of

internal policy after major deliberation could strip an organization of its First

Amendment rights.

103

The decision thus has to be up or down on that question,

and it was up by the small majority of five-to-four where the majority thought

that the expressive nature of this organization exempted it from the New Jersey

Law Against Discrimination.

The question is what to make of this decision. To Bagenstos, Dale repre-

sents the camel’s nose under the tent. He thinks that it is the first step on a long-

term effort to use the First Amendment to dismantle Title II by insisting that all

types of association are expressive and thus entitled to receive First Amend-

ment protection, on the ground, say, that “[s]erving an African American cus-

tomer in a restaurant side by side with white customers sends the message of

equal citizenship of blacks and whites at least as strongly as admitting gay

members to the Boy Scouts sends the message that homosexuality is accepta-

ble.”

104

It is worth noting that no one in the thirteen years after Dale has sought

to make this argument to overturn Title II. Indeed, the common response of

Dale defenders on this point, as embodied in the work of Seana Shiffrin and

Dale Carpenter, is to insist that the line between expressive and business asso-

ciations is clear enough that this slippage will never take place.

105

Their sepa-

rate efforts to draw just this line are undoubtedly in good faith because they are

both strong defenders of Title II as it applies to commercial situations, as

103. See Dale, 530 U.S. at 676 (Stevens, J., dissenting) (“[The Boy Scouts] never took

any clear and unequivocal position on homosexuality. . . . At a minimum, a group seeking to

prevail over an antidiscrimination law must adhere to a clear and unequivocal view.”). For

criticism of Justice Stevens’s dissent, see Epstein, supra note 96, at 138; and Seana Valen-

tine Shiffrin, What Is Really Wrong with Compelled Association?, 99 N

W. U. L. REV. 839,

845-51 (2005).

104. Bagenstos, supra note 36, at 1229.

105. See Dale Carpenter, Expressive Association and Anti-Discrimination Law After

Dale: A Tripartite Approach, 85 M

INN. L. REV. 1515, 1517-18 (2001); Shiffrin, supra note

103, at 876-77.

June 2014] PUBLIC ACCOMMODATIONS 1277

Bagenstos himself notes.

106

Given the attitude of the courts, I think that it is

fair to say that politically this line will stand against future attacks.

Even that prediction, however, does not answer the normative question of

whether that line ought to stand once we move away from the restaurant situa-

tion to other situations that give rise to more serious disputes over policy objec-

tives and admission requirements for organizations like the Boy Scouts. On this

score, Bagenstos has to explain why it is better to force the Boy Scouts to admit

gay members, when in the aftermath of that decision members engaged in an

immense internal struggle and worked hard to find the proper response to this

burning dispute. Today gay members will be admitted to the Boy Scouts,

107

and the organization will not, or at least has not, imploded, its membership hav-

ing gone down only six percent.

108

So the question remains, which resolution is

preferable? A command-and-control operation from the states and the federal

government or the imperfect give-and-take process of internal deliberation that

produced results more successful than anyone on either side of the debate had

dared to hope. It was only the decision in Dale that allowed this second, less

coercive resolution.

B. The Small Business Problem

The remaining issue is whether or not the category of expressive organiza-

tions could, or should, extend to other commercial firms of all descriptions. I

have already indicated that it is highly unlikely that anyone would wish to ap-

ply it to restaurant situations. But the same is not true with respect to a wide

range of employments, some of which involve intimate, or at least deeply per-

sonal, interactions between employers and employees. In dealing with this

106. Bagenstos, supra note 36, at 1231.

107. Adele M. Banks, A Boy Scout Schism?, ONFAITH (May 28, 2013),

http://www.faithstreet.com/onfaith/2013/05/28/will-there-be-a-mass-exodus-of-religious-

groups-from-the-scouts/20866 (“Now that Boy Scout delegates have taken their long-

awaited vote and permitted openly gay Scouts, will there be a mass exodus by religious

groups?”). The answer to the article’s question is probably “no.”

108. A recent article noted:

The Boy Scouts of America experienced a modest dip in membership in 2013, but not the

mass exodus that some social conservatives predicted while it was considering changing its

membership policy on gays.

Boy Scouts membership fell by 6 percent last year, leaving it with nearly 2.5 million

youth members and 960,000 adult members.

Reasons for the attrition, which is slightly greater than the 4 percent losses in 2012 and

similar-sized declines in several previous years, are not fully understood but are likely related

to the divisive vote on admitting openly gay youths to Scouts as well as a 60 percent increase

in annual membership dues.

Cheryl Wetzstein, Boy Scouts Shrink 6 Percent After 1 Year of Allowing Gay Members,

ASH. TIMES (Feb. 12, 2014), http://www.washingtontimes.com/news/2014/feb/12/boy-

scouts-shrink-6-percent-after-1-year-of-allowi.

1278 STANFORD LAW REVIEW [Vol. 66:1241

question, Bagenstos takes issue with the position advanced by Carpenter and

Shiffrin, and insists that the line between ordinary and expressive employments

will be exceedingly difficult to maintain in practice given the wide range of

business types along a continuum.

109

Indeed, I think that Bagenstos is right on

just this point for a wide range of establishments: think of health care personnel

and personal trainers, contexts in which most people have deep preferences as

to the people with whom they interact. There is no question that many custom-

ers use information about race, national origin, religion, sex, and sexual orien-

tation to decide whom to patronize. These choices are bundled in with deep no-

tions of identity politics. I know of no one who thinks that this set of customer

choices should be limited by any human rights law. So why have a different

rule on the opposite side of the relationship, where the preferences may in some

cases be just as intense?

To be sure, many people will say that businesses have a profit motive that

makes them less willing to act on these preferences. But that is precisely the

point. Knowing that in most cases they will be eager for business, the sensible

result is to let them choose their own customers so that these choices will sort

themselves out quite nicely. Where these identity issues do matter—largely in

small firms, one suspects—the businesses may pay a financial price in order to

act in accordance with their own beliefs. There is no need therefore to reserve

any such associational privilege under the First Amendment to “expressive” ac-

tivities.

It should be clear that I don’t think that Carpenter and Shiffrin are making

some kind of strategic retreat because “they recognize that a frontal attack on

Title II of the Civil Rights Act is a political nonstarter.”

110

I think that they are

making their argument in complete good faith because they believe that the dis-

tinction is both correct and sustainable. On that issue, however, I think that

Bagenstos is correct—but only conceptually—that the line will not hold. Simi-

larly, my own position is not a sneaky attempt to challenge Title II. It is a

frontal assault that rests on the explicit and emphatic defense of the older view

that only the presence of monopoly power should trigger a generalized obliga-

tion of universal service on nondiscriminatory terms. In my view, any concep-

tual effort to subdivide associations into three separate categories, some of

which receive higher levels of constitutional protection than others, is doomed

109. See Bagenstos, supra note 36, at 1208 (“[O]ngoing legal developments—both in

the area of public accommodations law itself and in the litigation surrounding the Affordable

Care Act’s ‘contraception mandate’—are poised to undermine this expressive-commercial

distinction. If these challenges succeed, Dale’s freedom-of-association principles will threat-

en the core of public accommodations law—including, perhaps, Title II itself.” (footnote

omitted)).

110. Id.

June 2014] PUBLIC ACCOMMODATIONS 1279

to face not only serious line-drawing issues, but also the greater sin of intellec-

tual incoherence.

Here is why. The standard theory of freedom of association writ large is

that the members of the organization get to determine its purposes, its mode of

operation, and the composition of its membership. So long as transaction costs

are low, as they typically are, the refusal to deal does not represent any form of

market failure, but only the revealed preferences that the gains to the outsiders

are smaller than the perceived losses to the insiders. We make that conclusion

not because we collectively weigh their preferences on some exquisite scales of

justice, but because we observe that there has been no deal.

In contrast, the modern view applies antidiscrimination laws to those or-

ganizations so long as they do not meet some state-determined standard of “ex-

pressiveness,” which flies directly in the face of that traditional understanding,

and thus leads to the obvious question: what should be done in those cases in

which standard business associations—family partnerships, private corpora-

tions, and small firms with single owners—are organized to serve both reli-

gious and secular ends, or whose principals seek to take into account their own

religious beliefs in choosing the goods and services that they offer to the mar-

ket? This pattern of behavior is perfectly consistent with neoclassical econom-

ics, which recognizes that parties may choose to take the gains from coopera-

tive activities in nonpecuniary forms. It is not as if these organizations will go

out of business; after all, they are not looking for the last dollar in market trans-

actions. But it does spell trouble down the road for these groups when faced

with challenges as to how they interact with their customers.

This point was brought home with great vividness in the recent New Mexi-

co decision, Elane Photography, LLC v. Willock, where the state supreme court

applied its human rights law to a business owned by a fundamentalist Christian

who refused to take pictures at a commitment ceremony—gay marriages are

now recognized in New Mexico but were not at the time—between two lesbi-

ans, Vanessa Willock and Misti Collinsworth.

111

It is worth stressing the up-

side-down quality of the New Mexico law. Antidiscrimination laws should be

condemned because they force unwilling associations. The moral case for

same-sex marriage rests on the very libertarian principles that are offended by

111. 309 P.3d 53, 58-59 (N.M. 2013), cert. denied, 134 S. Ct. 1787 (2014). In another

recent development, a Colorado administrate law judge refused to dismiss an antidiscrimina-

tion charge against a bakery that would not supply a wedding cake for a same-sex ceremony.

The intrusion here seems smaller, but the whole point of a strong freedom of association

principle is to keep the legislature or the courts from drawing these lines. See Craig v. Mas-

terpiece Cakeshop, Inc., No. CR 2013-0008 (Colo. Civ. Rights Comm’n Dec. 6, 2013)

(ALJ), available at https://www.aclu.org/sites/default/files/assets/initial_decision_case_

no._cr_2013-0008.pdf, aff’d, No. CR 2013-0008 (Colo. Civ. Rights Comm’n May 30,

2014) (final agency order), available at https://www.aclu.org/sites/default/files/assets/

masterpiece_--_commissions_final_order.pdf.

1280 STANFORD LAW REVIEW [Vol. 66:1241

the New Mexico human rights laws. Marriage is an intimate form of associa-

tion that the state cannot forbid by the use of its monopoly power over the li-

censing system. Nor should any broad reading of the morals head of the police

power upend that conclusion.

112

But for these purposes, the New Mexico Supreme Court took the sound-

ness of the law as given. Thus, as the facts unfolded, Willock put in a request

that Elane Photography be the photographer for the event, which Elane

Huguenin, co-owner and lead photographer, refused on clear religious grounds.

At that point, Collinsworth followed up with an e-mail asking about Elane Pho-

tography’s availability, without disclosing that her request was for a same-sex

commitment ceremony. Once Huguenin replied that the company was availa-

ble, she had irrefutable evidence that the earlier refusal was based on grounds

of sexual orientation.

113

The couple found services elsewhere, as one would

expect in a competitive market as large as Albuquerque. But when the matter

was turned over to the New Mexico Human Rights Commission, Elane Photog-

raphy was hit with a $6,637.94 award in attorney’s fees,

114

which the plaintiffs

later waived.

115

But what of it? If this violation stands, Elane Photography is

then at risk of future complaints by other potential claimants, and of fines from

the New Mexico Human Rights Commission that can be collected. In effect,

the result of this case could be to force Elane Photography to choose between

violating its religious beliefs or going out of business, all as a result of a lawsuit

that never should have been brought in the first place. Elane Photography is a

classic version of the test-case setup.

And for what? In looking at this case, it is clear that this was a deal that

neither side wanted. Just imagine if Elane Photography had taken the job with-

out telling Willock and Collinsworth of Huguenin’s religious preferences, and

Huguenin then had showed up bedecked with bold Christian paraphernalia and

signaled to all the guests that she did not approve of the relationship. Why in-

vite anyone to a commitment ceremony who is not committed to the cause of

which it is a part? On ceremonial occasions like this, the line between the per-

sonal and impersonal starts to blend. Notwithstanding this context, however,

the New Mexico Supreme Court had no difficulty in deciding that “a commer-

cial photography business that offers its services to the public, thereby increas-

ing its visibility to potential clients, is subject to the antidiscrimination provi-

sions of the [New Mexico Human Rights Act (NMHRA)] and must serve same-

112. See generally Richard A. Epstein, Of Same Sex Relationships and Affirmative Ac-

tion: The Covert Libertarianism of the United States Supreme Court, 12 S

UP. CT. ECON. REV.

75 (2004).

113. Elane Photography, 309 P.3d at 59-60.

114. Elane Photography, LLC v. Willock, 284 P.3d 428, 433 (N.M. Ct. App. 2012),

aff’d, 309 P.3d 53, cert. denied, 134 S. Ct. 1787.

115. Elane Photography, 309 P.3d at 60.

June 2014] PUBLIC ACCOMMODATIONS 1281

sex couples on the same basis that it serves opposite-sex couples.”

116

It stressed

that the standard here is the same as it is with discrimination on grounds of

race. It then rejected the First Amendment defense “because the NMHRA does

not compel Elane Photography to either speak a government-mandated mes-

sage or to publish the speech of another.”

117

With rare exceptions,

118

liberal

commentators commended the decision on the ground that the antidiscrimina-

tion laws would be gutted if all sorts of expressive commercial organizations

could claim exemption from its operation.

119

Libertarian groups have taken the

opposite position, holding that the act of photography itself constitutes a form

of expressive behavior that is indeed protected by the First Amendment.

120

116. Id. at 59.

117. Id.

118. See, e.g., Andrew Koppelman, Gay Rights, Religious Accommodations, and the

Purpose of Antidiscrimination Law, 88 S.

CAL. L. REV. (forthcoming) (on file with author).

Andrew Koppelman writes:

Businesses that serve the public, such as wedding photographers, should be exempted, but

only if they are willing to bear the cost of publicly identifying themselves as discriminatory.

That cost will make discrimination rare almost everywhere. Employers—some of whom also

object to recognizing same-sex marriages—should not however be allowed to discriminate in

providing benefits for their employees, such as denying health insurance to same-sex spous-

es. You can find another wedding photographer, but you only have one insurance plan. . . .

I’ve been a gay rights advocate for more than 25 years. Here, for the first time, I make

common cause with my longtime adversaries. I’ve worked very hard to create a regime

in which it’s safe to be gay. I’d also like that regime to be one that’s safe for religious

dissenters.

Id. I welcome Andrew Koppelman’s vigorous defense of Elane Photography, but have these

observations about his overall position. First, I don’t think that the notice requirement will

deter businesses like Elane Photography. They will instead announce that they welcome eve-

ryone’s business on all matters except gay weddings, which they refuse to do on grounds of

religious conscience. Many people might come to them because they sympathize with the

position. Second, I don’t think that Koppelman points to a viable distinction between wed-

ding photographers and employers. To be sure, there are fewer goods bundled into a short-

term stint with a photographer, but there is also extensive competition in labor markets.

Workers can shift jobs, and they can surely acquire specialized coverages outside the em-

ployment relationships. Labor markets are not monopolistic, so that freedom of association

should remain the rule.

119. See Adam Serwer, The New Plot to Take Down Gay Rights, MSNBC (Sept.

25, 2013, 5:40 AM), http://www.msnbc.com/msnbc/the-new-plot-take-down-gay-rights

(“‘It’s impossible to draw a line here between an expressive commercial service and a non-

expressive commercial service. Almost any commercial interaction we have today involves

the expression of words or pictures or talking,’ said Josh Block, an attorney with the Ameri-

can Civil Liberties Union. ‘People once argued that they had a religious or associational

right to refuse to serve black people in an integrated manner, and those claims were reject-

ed.’”). So if the line cannot be drawn, why use the antidiscrimination laws at all? In most

cases they are not needed given business imperatives, so underenforcement is no insuperable

obstacle, even if overenforcement is.

120. Ilya Shapiro, We Support Gay Marriage but Oppose Forcing People to Support It,

ATO AT LIBERTY (Nov. 2, 2012, 1:45 PM), http://www.cato.org/blog/we-support-gay-

marriage-oppose-forcing-people-support-it.

1282 STANFORD LAW REVIEW [Vol. 66:1241

That position was argued with great conviction by Dale Carpenter, Eugene

Volokh, and the Cato Institute, who claimed that photography is an expressive

activity that should receive First Amendment protection,

121

relying on both Jus-

tice Jackson’s famous opinion in West Virginia State Board of Education v.

Barnette

122

and the Court’s well-known decision in Wooley v. Maynard.

123

De-

spite the Supreme Court refusing to grant certiorari, I commend these efforts

because of their laudable attempt to shoehorn associational freedoms into the

First Amendment, but recognize all too well that Bagenstos has a field day in

explaining all the complications that will result in policing the line: “As anyone

who has ever hired a caterer for a wedding, bar mitzvah, or other occasion

knows, catering has inevitably expressive elements.”

124

The typical wedding

must have dozens of contractors.

125

Is it only the camera crew that is protected?

What about the makers of the wedding cake and the car valet staff?

Bagenstos’s colorful prose should, however, lead him to endorse the exact

opposite conclusion. Why face these issues of classification as a matter of pub-

lic law when the parties in this fiercely competitive business can sort matters

out themselves? There are few error costs, given market pressures, in defending

freedom of association across the board, and massive disruptions caused by de-

fending the NMHRA, which tramples over the all-too-human right of freedom

of association. As a matter of first principles, there should be no need to locate

these cases on the right side of the Dale line that for legal purposes ought not to

exist in the first place. Normatively, the correct rule is that freedom of associa-

tion is a generalizable value that holds in all competitive markets; the effort to

apply the antidiscrimination laws in that domain is a giant form of overreach,

no matter whether the lines of difference are race, religion, or sexual orienta-

tion. This position applies a fortiori to those persons who reject a request for

services on bona fide religious grounds, but it is not limited to them. This posi-

tion also applies to all sorts of services, not some subclass like photography that

may just be given preferred First Amendment status on freedom of speech

grounds. There is virtually zero risk of systematic exclusion when competitive

substitutes are available, so that using the broad freedom of association princi-

ple produces few error costs. If most organizations regard these distinctions as

abhorrent, it is all the more important to allow those who differ to go their sepa-

rate ways. Customers have lots of options to choose from, while the practition-

ers of certain beliefs have few choices of their own if forced to engage in prac-

121. See Brief of Amici Curiae Cato Inst., Eugene Volokh, & Dale Carpenter in Sup-

port of Petitioner at 3, Elane Photography, LLC v. Willock, 134 S. Ct. 1787 (2013) (No. 13-

585), 2013 WL 6665006.

122. 319 U.S. 624 (1943).

123. 430 U.S. 705 (1977).

124. Bagenstos, supra note 36, at 1235.

125. See id.

June 2014] PUBLIC ACCOMMODATIONS 1283

tices that they find offensive to their religious beliefs in order to stay in busi-

ness.

Nor is it possible to sugarcoat the New Mexico court’s ruling by claiming

that Elane Photography brought the law down upon itself when it opened its

business to the public. The argument, which has a long history in civil rights

litigation, claims that the phrase “offers its services to the public” carries with it

the clear implication that Elane Photography accepted the duty not to discrimi-

nate the moment it opened its doors for business. Put otherwise, no one who is

in favor of this law thinks that it could be avoided if below “open for business”

was the sentence “we reserve the right to choose our own customers.” (Ironical-

ly, the New Mexico Supreme Court said that Elane Photography could post a

notice that it will respect the law that it doesn’t agree with.)

126

Note that this

sentence could not be posted by a common carrier that indeed has just that duty

to serve on reasonable and nondiscriminatory terms. That difference in the two

settings is recognized by the common law distinction between offers and invita-

tions to treat,

127

such that it is not credible to say that Elane Photography has

somehow voluntarily offered its services to any or all members of the public.

That outcome represents, moreover, the standard position in competitive

markets that each person—whether merchant or customer—is the master of his

offer; the firm has the unqualified right to turn down any business proposition,

just as the potential customer can choose for whatever reason the firms that it

wants to deal with. Surely Willock has no duty to accept Elane Photography’s

offer to work on any project whatsoever if she takes offense at its policy not to

photograph gay marriages—no questions asked. That right is shared by all oth-

er members of the public. No one claims that someone who puts out a request

for services has committed herself to hire a contractor without regard to his re-

ligious or sexual preferences. It is therefore odd to posit some “humiliation and

dignitary harm”

128

as a trump on the side of a disappointed customer, without

recognizing that the mandated services now impose humiliation and dignitary

harm on business proprietors who are also human beings: why else is Elane

Photography fighting this case? These issues are hot. They can spur widespread

boycotts and other actions against the firm—think of Chick-fil-A.

129

The deci-

sion to override private associational preferences based on the allegation of soft

126. Elane Photography, 309 P.3d at 59.

127. See Lefkowitz v. Great Minneapolis Surplus Store, Inc., 86 N.W.2d 689, 691

(Minn. 1957); Moulton v. Kershaw, 18 N.W. 172, 173 (Wis. 1884).

128. Elane Photography, 309 P.3d at 64.

129. See, e.g., Kim Severson, A Chicken Chain’s Corporate Ethos Is Questioned by

Gay Rights Advocates, N.Y.

TIMES (Jan. 29, 2011), http://www.nytimes.com/2011/01/30/us/

30chick.html?pagewanted=all (describing gay rights advocates’ attacks on the firm for its

anti-same-sex marriage position).

1284 STANFORD LAW REVIEW [Vol. 66:1241

harm thus founders. Those harms are created on the one side in the attempt to

limit them on the other.

The correct analysis therefore requires looking at both sides of the relation-

ship before making some judgment on both social efficacy and dignitary harms.

That point is also correct on more general grounds. The first point is that it is

likely that Willock and Collinsworth could not pay enough money to Elane

Photography to overcome its objection. Nor could Elane Photography pay

enough to Willock and Collinsworth to make them change their minds. This

dispute is not over money, but fundamental values. But what lesson can be

learned from the observation that it is difficult for two parties to bargain them-

selves out of impasses? In this situation, the correct default rule is that the two

sides go their separate ways.

That choice of default rule becomes more salient as the number of potential

parties increases. Thus if Elane Photography is under a duty to serve all com-

ers, it cannot not afford to buy off the long queue of gay and lesbian couples

that come to its door asking for financial payment. But if Elane Photography

has the right to exclude for any reason, then selective admission of chosen per-

sons becomes the norm on which free association can rest. So giving owners

the right to exclude others reduces the bargaining complications that would

otherwise ensue, which in turn eases the path to competitive markets as others

jump in and offer to serve the customers that Elane Photography will not. The

point here applies across the board. It is not an answer to say that by this logic

Elane Photography could refuse to serve black customers, which is within its

rights, just as it is for any firm to refuse to serve white customers, or to refuse

to serve any firm that does not engage in a systematic policy of nondiscrimina-

tion. Once again, the principle of freedom of association applies across the

board once the issues of monopolization and the abuse of state power are put to

one side. In all cases, competitive firms will fill the supposed gap.

The overall system works far better with strong property rights and of

course the strong protection of individual autonomy in all personal dealings. It

is for that reason that in ordinary property arrangements elsewhere, it is the du-

ty of the outsider to win consent, not the duty of the owner to buy off all out-

siders that she does not wish to admit. So the standard rule is that the cattle

owner has to keep the cattle out of the farmer’s land, and not the other way

around. Armed with the right to exclude, the landowner can decide which, if

any, outsider to let in to graze on his land.

130

Those who do not like the situa-

tion can go elsewhere. Elane Photography does not have any dominant market

position in Albuquerque, New Mexico. Huguenin should be allowed to run her

business in peace.

130. Kenneth R. Vogel, The Coase Theorem and California Animal Trespass Law,

16 J. L

EGAL STUD. 149, 172-74 (1987).

June 2014] PUBLIC ACCOMMODATIONS 1285

The question then arises as to what should be done going forward. In one

sense the damage was already done once the New Mexico Human Rights

Commission put the issue in play. In retrospect, it should have just ducked the

issue by invoking the constitutional doctrine of avoidance given the inevitable

clash of this human rights law with religious liberties. At this point, the matter

might have been defused without inviting frontal assaults on Title II. But once

on the books, Elane Photography provoked many conservative and religious

organizations to back an extension of Arizona’s Free Exercise of Religion

Act

131

to cover not just religious assemblies or entities, but also ordinary indi-

viduals, associations, partnerships, and corporations. One such bill, Arizona

Senate Bill 1062,

132

passed, only to be vetoed by Governor Jan Brewer who

stated, correctly in my view, that the bill “has the potential to create more prob-

lems than it purports to solve.”

133

She did so on the urging of prominent politi-

cal leaders—including Arizona Republican Senators John McCain and Jeff

Flake, and former Republican presidential nominee Mitt Romney—as well as

major corporate and business interests, “including Delta Air Lines, the Super

Bowl host committee and Major League Baseball,” which feared a backlash

against the state.

134

There is at least one deep irony in this collective rejection of Arizona Sen-

ate Bill 1062. The firms and individuals who opposed the legislation had no de-

sire whatsoever to take advantage of its protections. That fact alone explains

why the bill itself posed no threat to the established patterns of business. In-

deed, if the governor had signed off on the bill, all companies that were uneasy

with the new state of affairs could have taken binding pledges to continue to do

business just as if the legislation had never passed. Their behavior makes out

131. Ch. 332, 1999 Ariz. Sess. Laws 1769 (codified as amended at ARIZ. REV. STAT.

ANN. §§ 41-1493 to -1493.02 (2014)). Arizona’s Free Exercise of Religion Act is modeled

off the Religious Freedom Restoration Act (RFRA) of 1993, Pub. L. No. 103-141, 107 Stat.

1488 (codified as amended at 42 U.S.C. §§ 2000bb-2000bb-4), invalidated by City of

Boerne v. Flores, 521 U.S. 507 (1997). RFRA was held unconstitutional as it applies to the

states in City of Boerne on the ground that it sought to give primacy to Congress’s interpreta-

tion of the substantive guarantees of the Fourteenth Amendment, when the last word on that

question belonged to the Supreme Court. See 521 U.S. at 536. But the law continues to apply

against the federal government, where no constitutional issue is raised. See Gonzales v. O

Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 424, 439 (2006). In addition, at

last count, nineteen states have enacted state analogs to RFRA. See Reid Wilson, Mississippi

Passes Arizona-Style Religious Freedom Bill, W

ASH. POST (Apr. 1, 2014, 9:18 PM),

http://www.washingtonpost.com/blogs/govbeat/wp/2014/04/01/mississippi-passes-arizona-

style-religious-freedom-bill.

132. S.B. 1062, 55th Leg., 2d Sess. (Ariz. 2014).

133. Catherine E. Shoichet & Halimah Abdullah, Arizona Gov. Jan Brewer Vetoes Con-

troversial Anti-Gay Bill, SB 1062, CNN (Feb. 26, 2014, 11:13 PM EST),

http://www.cnn.com/2014/02/26/politics/arizona-brewer-bill.

134. Id.

1286 STANFORD LAW REVIEW [Vol. 66:1241

the case for why the legislation is in principle unnecessary. But these firms

were only interested in returning to the settled set of social expectations that

antedated the legislation.

There is much pragmatic sense in their position. As I have argued else-

where, constant practice may give rise to a prescriptive constitutional right that

is formed in much the same way as the long use of a right-of-way can create a

prescriptive easement.

135

But as with prescription, the right should only go as

far as the established social practice, which has never covered situations like

Elane Photography, where the equities between the parties lie so much in favor

of the firm. No doctrine of settled expectations should sanction this new legis-

lative development. Indeed, the case for protection starts with the observation

that now Elane Photography is part of a discrete and insular minority under

footnote four of United States v. Carolene Products Co.

136

Its minority reli-

gious views need constitutional space between it and the relentless ambitions of

an ascendant gay rights movement that seems to have quickly forgotten that its

members were once on the receiving end of the unthinking and abusive exer-

cise of state criminal law. Nothing in the court’s decision in Elane Photography

allays this fear of a new wave of authoritarian abuse.

This pragmatic plea for the status quo ante leaves untouched the question

of how this issue should be resolved as a matter of first principles. On that

score there is no reason to be coy or pragmatic. As a matter of first principles,

Title II of the Civil Rights Act should be regarded as unconstitutional as ap-

plied to all voluntary organizations in well-functioning markets. Civil rights

laws are turned upside down when used to harass small businesses with minori-

ty viewpoints. At this point, the Robert Bork of 1963, who was so wrong about

public accommodations in the Old South, is unhappily vindicated by the care-

less way in which statutory duties to serve are extended far beyond their origi-

nal purpose of coping with monopoly power in common carrier situations.

135. RICHARD A. EPSTEIN, THE CLASSICAL LIBERAL CONSTITUTION: THE UNCERTAIN

QUEST FOR LIMITED GOVERNMENT 68-71 (2014).

136. 304 U.S. 144, 152 n.4 (1938). Recall the basic passage, which was written in the

context of a futile challenge to a filled milk regulation, just after the 1937 revolution:

It is unnecessary to consider now whether legislation which restricts those political pro-

cesses which can ordinarily be expected to bring about repeal of undesirable legislation, is to

be subjected to more exacting judicial scrutiny under the general prohibitions of the Four-

teenth Amendment than are most other types of legislation . . . .

Nor need we enquire whether similar considerations enter into the review of statutes di-

rected at particular religious, or national, or racial minorities: whether prejudice against dis-

crete and insular minorities may be a special condition, which tends seriously to curtail the

operation of those political processes ordinarily to be relied upon to protect minorities, and

which may call for a correspondingly more searching judicial inquiry.

Id. (emphasis added) (citations omitted).

June 2014] PUBLIC ACCOMMODATIONS 1287

V. G

OVERNMENT DISCRIMINATION

The decision to pick on small and isolated groups is not just a function of

antidiscrimination law as it applies to private parties. It has also arisen in con-

nection with government actions that overtly and consciously discriminate

against small groups that wish to retain their organizational preferences. In the-

se cases, the dominant question involves the actual or potential application of

the doctrine of unconstitutional conditions.

137

That principle states in its canon-

ical form that even if the state has the power to grant or deny a certain privi-

lege, it may not have the power to grant that privilege contingent on certain

conditions that are constitutionally suspect. The scope of the doctrine is broad

because it arises whenever the government exerts monopoly power or control

over some essential facility normally open to the public at large. In constitu-

tional language, these cases of government ownership often involve the opera-

tion of public forums, including streets, parks, and halls, where people are al-

lowed to congregate. In all of these cases it has long been held that

[w]herever the title of streets and parks may rest, they have immemorially

been held in trust for the use of the public and, time out of mind, have been

used for purposes of assembly, communicating thoughts between citizens, and

discussing public questions. Such use of the streets and public places has,

from ancient times, been a part of the privileges, immunities, rights, and liber-

ties of citizens.

138

The reference to the public trust in this case is the signal that fiduciary du-

ties are involved so that the state cannot pick and choose between the groups

that are allowed to engage in political activity on public premises. In essence,

there is no particular obligation to open the premises to all comers, but there is

a duty to act in a fair and nondiscriminatory way with respect to said groups,

whose conduct is only subject to time, place, and manner restrictions.

It thus becomes clear that in allowing groups to enter onto the public high-

way, the relevant government authorities cannot discriminate among potential

applicants on the grounds of their own political or associational preferences.

For example, in Hurley v. Irish-American Gay, Lesbian & Bisexual Group of

Boston the question was whether the Massachusetts public accommodations

law could require the South Boston Allied War Veterans Council to include the

gay, lesbian, and bisexual group in its parade, to which the answer was “no.”

139

The state as the holder of the monopoly resource—here the streets—is under

the nondiscrimination duty to all comers, including those private groups that

discriminate on grounds that offend the Massachusetts statute. In effect, Hurley

137. For my views, see RICHARD A. EPSTEIN, BARGAINING WITH THE STATE 3-12 (1993).

138. Hague v. Comm. for Indus. Org., 307 U.S. 496, 515 (1939) (opinion of

Roberts, J.).

139. 515 U.S. 557, 569 (1995).

1288 STANFORD LAW REVIEW [Vol. 66:1241

is a modern instantiation of the earlier and proper application of the nondis-

crimination principle against the state.

A narrow majority of the Supreme Court lost sight of this lesson in Chris-

tian Legal Society Chapter of the University of California, Hastings College of

the Law v. Martinez, where the scene of action switched to Hastings College of

Law, which acted to exclude a tiny Christian Legal Society (CLS) chapter from

various privileges that were extended to other Hastings groups on the ground

that CLS required its members to affirm that they adhered to the teachings of

Jesus Christ and were opposed to “unrepentant homosexual conduct.”

140

There

is no question that a private institution in Hastings’s position should be allowed

freedom over whom to admit and the terms on which that admission should be

based. There is no doubt that if Hastings sought to exclude fundamentalist

Christians from its student body, that total ban would amount to a form of

viewpoint discrimination that would be struck down. But as a public institution

that doles out public subsidies, it is much more closely analogous to a state

monopoly because of its ability to tap into public funds to support its opera-

tions. The question then arises why, in this “limited public forum” to which on-

ly Hastings students were invited,

141

it could exclude this tiny group from the

usual benefits and privileges that Hastings bestowed on more powerful groups,

including Outlaw, a society of lesbian, gay, bisexual, and transgender stu-

dents.

142

In defending that result, Justice Ginsburg insisted that Hastings could ex-

clude from its list of “Registered Student Organizations” all groups that refused

to adopt an “accept-all-comers policy” on the ground that said policy “encour-

ages tolerance, cooperation, and learning among students.”

143

But she never

explained why an organization had to admit into membership individuals who

did not share its fundamental mission. CLS was a tiny group vulnerable to po-

litical crosswinds in a sea of hostile students. It was quite willing to allow oth-

ers to attend its meetings, but not to join its governance structure, which is ex-

actly right. The organization could not survive once a fifth column was

guaranteed membership rights, which it could use to stack the organization with

its own supporters.

144

Justice Ginsburg thought that it was one thing merely to

exclude people from benefits and another to impose direct losses upon them.

145

140. 130 S. Ct. 2971, 2980 (2010) (internal quotation marks omitted). For my views,

see generally Richard A. Epstein, Church and State at the Crossroads: Christian Legal Soci-

ety v. Martinez, 2009-2010 CATO SUP. CT. REV. 105 (2010).

141. Christian Legal Soc’y, 130 S. Ct. at 2991-92.

142. 1 Joint Appendix at 236-45, Christian Legal Soc’y, 130 S. Ct. 2971 (No. 08-1371),

2010 WL 372139 (listing “Registered Student Organizations”).

143. Christian Legal Soc’y, 130 S. Ct. at 2978-79 (internal quotation marks omitted).

144. See id. at 3019-20 (Alito, J., dissenting).

145. See id. at 2986.

June 2014] PUBLIC ACCOMMODATIONS 1289

She would never take that position for law schools that wanted to exclude so-

cialists from the faculty or student body. But the choices that are allowed to

private institutions in competitive markets are not available to state-run institu-

tions that receive subsidies and other forms of public support for which there is

no private alternative available to CLS or anyone else.

The key point about the doctrine of unconstitutional conditions is that it

limits the power of state choice when the state acts as a common carrier in the

use of its monopoly power. The only justifications that allow for discrimination

are those that the state could apply through direct regulation of private activi-

ties. It could refuse to allow groups to join if they posed the threat of force or

fraud. Otherwise the only difference between the open public forum in Hague

v. Committee for Industrial Organization and Hurley and the limited public fo-

rum of Hastings is that admission (under nondiscriminatory rules) can be re-

stricted to potential law students, who as a group have privileges above and be-

yond those granted to outsiders. Within this framework it is clear that the great

tragedy of Christian Legal Society is that it allows the government to act as a

discriminating monopolist against weak and vulnerable groups, precisely be-

cause, as in Elane Photography, the government forces those small groups to

toe its own collectivist line on discrimination. How quickly we forget.

Nor is this an isolated instance. One current battle with the health care law

is whether the statute can force religious institutions to supply contraception

and abortion services to female employees even if it is inconsistent with the re-

ligious beliefs of the organization, usually as represented by its dominant fami-

ly shareholders.

146

That debate is extensive and wide ranging, and it is often

cast in terms of whether the “boss” should be able to determine the sexual prac-

tices of female employees.

147

But the reality is that the “boss” in this case is the

government that seeks to force organizations to make expenditures that they

don’t want in order to serve state but not private ends. In this case too, the cor-

rect response is for the government to understand that it cannot condition its

grants and regulations on any idea that suits its fancy. Instead it must follow the

widely rejected principle that in exercise of its monopoly power, it cannot deny

government benefits to practices that it could not ban directly,

148

which covers

this case. Antidiscrimination laws should never be used as a government club

against recalcitrant individuals in their everyday lives.

146. See, e.g., Hobby Lobby Stores, Inc. v. Sebelius, 723 F.3d 1114 (10th Cir.) (en

banc) (presenting this issue), cert. granted, 134 S. Ct. 678 (2013).

147. For a brief account, see EPSTEIN, supra note 135, at 470-75.

148. See Bob Jones Univ. v. United States, 461 U.S. 574, 603-04 (1983) (conceding

that the school’s decision to ban interracial dating was constitutionally protected, but that its

right to a religious tax exemption was not).

1290 STANFORD LAW REVIEW [Vol. 66:1241

AN UNHAPPY CONCLUSION

In the English language the word “discrimination” once had two meanings,

the first of which is in danger of being lost. In its positive sense, discrimination

refers to persons of taste and discrimination who can draw the right lines for

setting preferences of places to go, things to like, and people to choose as com-

pany. In its negative and now standard sense, discrimination is always against

some group, where it is thought that some bad or invidious motive accounts for

the behavior.

There is some truth to both of these definitions, and the task of legal an-

alysis is to decide when this or that form of discrimination should be allowed. It

is very risky to permit any single group, public or private, to be the arbiter of

good or bad taste. So the ultimate lesson to learn here is to try to decentralize

the use of power in good Hayekian fashion,

149

which in turn requires an extra

dollop of suspicion in organizing social relationships. The common law re-

sponse to this development, which was to impose duties of nondiscrimination

on parties with monopoly power over relatively commoditized goods and ser-

vices, was a good first cut into this problem. It meant that people could not be

kept out of railroads and off the electrical grid, but it also allowed all private

groups to select their own members and govern their own organizations when

they provided uniquely differentiated services in competitive markets. That rep-

resents in my view the correct division between government regulation and pri-

vate freedom of association.

Those principles in part drove the earlier applications of Title II of the Civ-

il Rights Act, where a combination of public abuse of essential facilities and

private violence posed a mortal threat to the individual liberties of vulnerable

citizens, often on grounds of race. Title II of the Civil Rights Act offered a sen-

sible first response in this respect. But with the increase in power of the civil

rights movement, the more recent applications of antidiscrimination law have

had a very different purpose. There have been many aggressive attempts to

achieve state-mandated nondiscrimination over a wide range of business and

social institutions that are bereft of monopoly power and that have distinctive

purposes and objectives that rightly resist any form of state standardization. In

this sphere, the situation is topsy-turvy. Antidiscrimination law now is a threat

to the diversity of our private institutions. It allows the state to impose nondis-

crimination obligations on weak and powerless individuals, institutions, and

firms that only wish to be left alone. It then compounds the mischief by insist-

ing that its key control over basic public facilities allows it to impose its will on

private institutions that are powerless to resist its combination of direct controls

and fines. It is indeed a sorry state of affairs that a great norm intended to blunt

149. See Hayek, supra note 21, at 526.

June 2014] PUBLIC ACCOMMODATIONS 1291

private power has now become a tool to allow all-too-powerful institutions to

stamp out those groups that oppose their vision of the good society.

I have no doubt that any group that wishes to buck the dominant social sen-

timent must face the risk that it will lose popularity and business to people who

are offended by its views. The application of public opinion in just that regard

is an essential safeguard of our social institutions, but only so long as those

same sentiments are allowed to the minority group that wishes to distance itself

from the majority. There should be free and open competition in the world of

ideas and behavior. But it also must be remembered that this process need not

generate a unity of views on key questions of gay rights any more than it must

generate a uniform view of whether Macs are better than PCs, or the reverse.

The interplay of social forces is well equipped to figure out what distribution of

power goes to which group.

The modern application of public accommodations and human rights laws

pushes too far and too fast in the wrong direction. There is little doubt that sex

discrimination has become a dirty word in many quarters. It is equally clear that

the opponents of gay marriage are losing the public debate as well. It is surely

correct to support gay marriage for the reasons noted above. The state has a

monopoly over marriage licenses and should not discriminate between various

candidates for marriage, which is why libertarian organizations are and should

be firm supporters of requiring the state not to discriminate between couples in

its exercise of monopoly power. But the state becomes the source of discrimi-

nation when it fines and punishes those who wish to exercise their own rights

of association. Those rights should be broad and firm. They should not depend

on whether an organization is or is not expressive, is or is not religious in orien-

tation, or is or is not engaged in political speech. Our authoritarian human

rights acts have to be recalibrated so that they protect one of the most funda-

mental of human rights—the right to associate, or not to associate, with people

of one’s own choosing. It is that lesson that we have to relearn on the fiftieth

anniversary of the Civil Rights Act of 1964.

1292 STANFORD LAW REVIEW [Vol. 66:1241