Coase and Accommodation: A Reply

Emory Law Journal Emory Law Journal

Coase and Accommodation: A Reply Coase and Accommodation: A Reply

Frederick Mark Gedicks

Part of the Law Commons

Coase and Accommodation: A Reply Coase and Accommodation: A Reply

Cover Page Footnote Cover Page Footnote [1] Michael J. Perry, The Constitution, the Courts, and Human Rights: An Inquiry into the Legitimacy of Constitutional Policymaking by the Judiciary (1982).

COASE AND ACCOMMODATION: A REPLY

Frederick Mark Gedicks

Many years ago, when I was a fresh-faced appointments candidate hoping to teach constitutional law, my dean at USC recommended some reading to ease me into the scholarly flow. One suggestion—which I took—was The Constitution, the Courts, and Human Rights.1 I never imagined its author would become a mentor, colleague, and friend.

Several years into my first appointment at a law school in the rural South, I received a note from Michael (whom I had not yet met) telling me that, in a recent speech, he’d quoted something I’d published—a small but characteristically generous gesture that meant everything to a young scholar toiling in evident obscurity. Michael helped and encouraged me over the years, in ways large and small; I have watched him do the same for many others. He will leave behind scholars as well as scholarship.

There is, to be sure, plenty of scholarship. It is at once inspiring and humbling to see a scholar so fully commit himself to the same problem through a long and successful career. Michael’s work has made a difference—the most a scholar can ask—by providing moral justifications for human rights in a world seemingly bereft of both. Though I have never gotten used to his very long footnotes, I have learned from the depth of what he writes and the care with which he writes it.

While I am honored to participate in this Festschrift for Michael, I am blessed to be his friend. Some years ago, my son Alex passed away while attending a university where Michael was on the faculty. Nicea and I have never

forgotten Michael’s kindness and concern for us and our daughters, then and for years after. He made a difference for us.

INTRODUCTION: RELIGIOUS ACCOMMODATION, THIRD-PARTY HARMS, AND THE COASE THEOREM ................................................................... 1458 I. COASE AND COMMENSURABILITY ............................................. 1464 A. Rights and Incommensurability .......................................... 1466 Incommensurability and Value Choices ............................. 1467 B. II. IDIOSYNCRATIC VALUE AND COASEAN BARGAINING ............... 1469 A. Elasticity of Demand for Religious Observance ................. 1471 B. Anti-discrimination Laws and Idiosyncratic Value ............ 1475 III. HIGH TRANSACTION COSTS AND THE (SUPER) FIRM SOLUTION 1477 IV. THE DUBIOUS ETHICS OF HARMING THIRD PARTIES ................. 1481 CONCLUSION: THE COASE THEOREM SAYS NOTHING ABOUT THIRD PARTY HARMS ...................................................................................... 1485

Economic analysis is no stranger to the Religion Clauses.2 Only recently, however, has it been brought to bear on religious accommodations, which excuse believers from obeying laws that bind everyone else.3 The occasion is the continuing controversy over third-party harms. I and other church-state separationists contend that the Establishment Clause prohibits religious accommodations that impose excessive burdens (“harms”) on those who derive no benefit from the accommodation because they do not engage in the exempted practice (“third parties”):

Like the prototypical established church, cost-shifting accommodations grant a privilege to those who engage in the accommodated practice at

the expense of unbelievers and other nonadherents who do not. Indeed, forcing those who do not belong to a religion to bear the material costs of practicing it is functionally equivalent to taxing nonadherents to support the accommodated faith.4

In economic terms, the third-party harm doctrine prohibits negative externalities created by religious accommodations. “Negative externalities” occur when the costs of an activity are not fully accounted for in its market price because they are “externalized” or shifted to persons who are not market participants—that is, who neither produce nor consume the good produced by the activity.5 A religious accommodation generates negative externalities when it permits its beneficiaries to avoid the material costs of practicing their religion by shifting these costs to a relatively small number of third parties who do not the practice contraception mandate of the Affordable Care Act, for example, shift some costs of observing the employer’s religious anti-contraception beliefs to employees who have no objection to contraception.7 These employees are forced to pay for contraception out of pocket because exemption permits the employer to exclude contraception coverage from the employee health plan.8 The third-party burden

the accommodated religion.6 Employer exemptions from

doctrine would either refuse the employer an accommodation, thereby forcing it to internalize the costs of living out its anti-contraception beliefs by complying with the law, or socialize the accommodation costs—say, through contraceptiononly health insurance funded by tax dollars.9

The third-party harm doctrine was urged as a defense to employer claims for religious exemption from the contraception mandate,10 with limited success.11

The extent to which the doctrine forms part of Establishment Clause jurisprudence remains contested.12

Enter Coase, whose famous article, The Problem of Social Cost, argued that negative externalities associated with economic production are reciprocal.13 A water-polluting factory externalizes the costs of its operations onto nearby residents who rely on the polluted water supply, but those same residents externalize a cost of residential housing—access to clean water—onto the adjacent factory.14 Coase famously observed that when transaction costs are zero,15 persons engaged in competing economic activities will bargain their way to the economically efficient mix of those activities, regardless of where the legal system places liability for externalities.16 The Coase theorem provides that, absent transaction costs, the factory owner and the homeowners will agree to the level of water pollution that maximizes the combined value of factory output and residential housing, irrespective of whether legal responsibility for the pollution is placed on the factory owner or the homeowners.17

religious

restriction of

for government

Invoking the reciprocal character of negative externalities, religious accommodationists have begun to argue that negative externalities from religious accommodations do not justify their denial because externalized harms also flow in the other direction.18 “[T]he mere existence of a religiously caused externality,” claims Stephanie Barclay, “would not, alone, provide a sufficient justification rights” because accommodated believers and third-party bystanders “are always inflicting externalities on each other.”19 Accommodations allow believers to externalize the costs of practicing their religion onto third-parties, the argument goes, but without accommodation, those third parties are permitted to externalize the costs of living out their differing beliefs onto believers seeking accommodation. So, if removal of contraception coverage from an employee health plan is an externalized cost of the employer’s anti-contraception beliefs, borne by its employees, then requiring such coverage is an externalized cost of ensuring employee access to contraceptives, borne by the employer. Harms to third parties from religious accommodations cannot ground the third-party harm doctrine because those very third parties are reciprocally harming those who seek accommodations. Or so accommodationists would argue.

The application of the Coase theorem to religious accommodations is illconceived on both economic and ethical grounds. It fails on its own economic premises, which generate complex puzzles about valuation, bargaining, and government intervention to which accommodationists offer no solution. But worse, it leaves the ethical question unanswered: how can a government that guarantees its citizens religious equality justify actions that force some citizens to pay for the religious practices of others?

To illustrate these criticisms, I will frequently refer to a standard case—that is, the religious accommodations from anti-discrimination laws that protect

LGBTQ customers from denials of service by commercial businesses.20 The standard case asks whether the Coase theorem justifies accommodation of legally prohibited LGBTQ discrimination when motivated by religious belief, despite the harm this externalizes onto LGBTQ persons whom the law would otherwise protect.21 I will argue it does not.

Because Professor Barclay has the most developed Coasean justification of negative externalities from religious accommodations, I will take her position as the accommodationist exemplar. She draws three economic conclusions about the third-party harm doctrine from her Coasean analysis: (1) negative externalities from a proposed or existing accommodation do not justify its denial or invalidation because failure to accommodate likewise entails reciprocal negative externalities; (2) denials of accommodations increase net social harm because of the high cost of forcing believers to comply with burdensome laws; and (3) Coasean analysis supplies a “clinical” method for analyzing conflicts of incommensurable values.22

These conclusions are problematic for three reasons. First, though Barclay acknowledges that conflicting religious and social values are often incommensurable, she does not provide any explanation, let alone an economic one, for how she arrives at her judgments of social costs or how society should weigh incommensurable values in an accommodations setting.23 Second, Barclay’s conclusions rest on the empirically incorrect assumptions that conflicting religious and social preferences are equally idiosyncratic and that the price believers attach to religious values is inelastic.24 Third, Barclay ignores the possibility that, when transactions costs preclude voluntary bargaining, legislative or regulatory intervention (suggested even by Coase) might judicially ordered approximate social efficiency more closely accommodations.25 Finally, a powerful ethical case can be made for the thirdparty harm doctrine, while accommodationists have yet to articulate a plausible normative theory that justifies shifting the costs of practicing a religion from those who believe it to those who don’t.26

than

The application of Coase to reciprocal externalities generated by religious accommodations and anti-discrimination rights raises the problem of incommensurability. Values are “incommensurable” if they cannot be cardinally ordered because they lack a common measure.27 Accommodations from antidiscrimination laws involve two competing rights, freedom of religion and freedom from discrimination,28 but there exists no common measure of valuing these rights to enable meaningful comparisons of their net social cost or benefit. What is the cost to believers of an anti-discrimination law that requires them to act in a manner prohibited by their religious commitments? The cost to LGBTQ persons of permitting believers to treat them unequally in public markets? Richard Posner and Michael McConnell concede that the problem of quantifying liberty values like these is “probably insuperable.”29 Without assigning a value

to each of religious freedom and freedom from discrimination that is expressed in the same terms, one cannot determine the socially efficient mix of each—that is, the amount of accommodation from anti-discrimination laws, if any, that maximizes the sum of religious freedom and freedom from discrimination.30

Coase had little to say about incommensurability in The Problem of Social Cost. He set out to problematize the Pigouvian conventional wisdom that the solution to the problem of externalities is taxing the activity generating them.31 Pigou was squarely within the early tradition of welfare economics, which ignored costs that could not be monetized. By his own account, Pigou studied “that part of welfare that can be brought directly or indirectly into relation with the measuring-rod of money,” excluding entirely “considerations such as justice, freedom, and rights to which money values could not be assigned.”32 Consequently, Coase expressly limited his analysis to the problem of maximizing economic output from competing economic activities when both are efficiently priced by the market.33

Incommensurability was not a problem for Coase because the relative costs and benefits of competing economic activities are reflected in the market prices of their respective outputs. As Joshua Decker observed, “The Coase Theorem works because it distills individual preferences into dollars and cents: parties negotiate in a common medium, not across an incomparable idiosyncratic void.”34 In Coase’s famous analysis of the rancher’s cattle trampling the farmer’s

crops, it is presumed that both the rancher and the farmer are concerned solely with making money and that the social goal is to maximize aggregate output from both ranching and farming.35 So, neither the rancher nor the farmer minds abandoning her activity if she can be compensated with the same sum she would have earned from actually ranching or farming; the rancher is not sentimentally attached to ranching, nor the farmer to farming. Coase’s point works nicely within this limited sphere, but a farmer who loves farming and is too poor to buy a ranch leaves Coase with nothing to say.36

A. Rights and Incommensurability

for Professor Barclay correctly characterizes Coase’s prescription conflicting rights as minimizing social harm and maximizing social gains.37 She also recognizes that this calculation cannot be made when the conflicting rights are not monetized or otherwise compared on a scale of common measurement, referring to Justice Scalia’s observation that comparing incommensurables is “like judging whether a line is longer than a rock is heavy.”38 Yet, having acknowledged incommensurability as a problem, she proceeds to ignore it.

Her principal example is religious exemptions from laws that mandate vaccination for childhood diseases as a condition to attending school.39 Such mandates, she argues, entail reciprocal negative externalities: the coercion of religious conscience when parents are forced to choose between in-person public schooling for their children and living out their religious anti-vaccination beliefs, against the degradation of herd immunity from excessive exemption and the consequent increased risk of debilitating illness and death of other people, especially those unable to vaccinate for medical reasons.40

How to decide? Barclay tentatively endorses uniform vaccination mandates without religious exemptions,41 though at the cost of coherence. “This may be one instance,” she speculates, where the “rock is very heavy” and the “stick is

very short”42—an arresting metaphor that nonetheless makes no sense. No quantitative prioritization of a rock’s weight to a stick’s length is possible, however heavy the rock or short the stick. Neither Coase nor economics allows one to coherently conclude that the charge on conscience from uniform vaccination mandates is more or less than the charge on public health in their absence, because the value of freedom of conscience and the value of herd immunity are incommensurable. Barclay’s suggestion that uniform vaccination mandates reduce net social harm is just a guess,43 what economists often (and derisively) refer to as “casual empiricism”—asserting the truth of factual propositions without “a shred of evidence.”44

B. Incommensurability and Value Choices

Of course, people decide among incommensurables all the time, in ordinary life and, more controversially, in the judicial process. Should we vacation in the Indiana Dunes or in Italy, see a movie about dinosaurs or Wittgenstein, make an employer reduce toxic chemicals in the workplace or leave it be?45 These decisions are driven by personal values, not economics. One can rationally believe that a large amount of one incommensurable good is worth more than a small amount of another only if one values the former more than the latter.46 One likes (or can afford) the beach at Lake Michigan better than a foreign country crowded with tourists (despite its spectacular food), one is more interested in prehistoric life than philosophy, one prefers larger company profits to increased workplace safety.

Courts deal with incommensurability by converting all interests into some measure of economic value, however imperfect. In adjudicating tort claims, for example, judges and juries economically value loss of life and limb, pain and suffering, emotional distress, and loss of consortium despite the pervading instinct that mere monetary compensation does not make the plaintiff whole

(though compensation is, of course, better than nothing).47 But even this imperfect option is not available for evaluating the efficiency of religious accommodations. The Religion Clauses flatly prohibit U.S. courts from making theological decisions;48 one can hardly imagine a decision more theologically freighted than economically valuing a believer’s loss of salvation and consequent eternal damnation.49

When courts must evaluate the net cost of religious accommodations, no common measure of valuation is constitutionally available. Without a common measure (or a plausible proxy), one cannot know that the aggregate social benefit of a uniform vaccination mandate exceeds the benefit of one with religious exemptions. Some parents religiously object to vaccination and conventional medicine with such intensity that they choose to let their children die from treatable conditions rather than abandon their faith.50 How does violation of such ferociously held faith compare to the diffuse interest of most parents in preserving herd immunity? Courts typically overcome this impasse by smuggling a reasonableness standard into the analysis, not by comparing the aggregate social value of protecting minority conscience against that of protecting the interests of the collective.51 But, again, this is constitutionally suspect whenever it occurs. It was precisely the unavoidably subjective character of balancing incommensurables like religious fidelity and social harm that Justice Scalia sought to avoid in Employment Division v. Smith.52

Professor Barclay’s endorsement of uniform vaccination mandates reflects her personal preference of the public health interest in herd immunity to the spiritual fidelity of religious vaccine objectors. This is a value judgment that,

even if widely shared, cannot be justified by economics or the Coase theorem in the absence of a dollar valuation of religious fidelity.53

II.

Professor Barclay’s difficulties with incommensurability are compounded by her assumption that the idiosyncrasy of religious values makes Coasean bargaining impossible and the denial of accommodations inefficient.54 “Idiosyncratic” value, by definition, is value unique to one person that is not transferable to others, so market pricing cannot account for it.55 A homeowner might value her residence far above its market value because it has been in her family for generations, or she raised all her children there, or she designed and built it herself, but a rational stranger in the marketplace cannot capture this value and hence will not pay for it.56 As Barclay recognizes,57 idiosyncratic value is a transaction cost and often a decisive one: if the idiosyncratic premium demanded is large enough, the homeowner will never sell because no one will pay a price that incorporates value only the homeowner can realize.

Barclay analogizes the believer to the idiosyncratic homeowner.58 Just as homeowners may decide that there is no price at which they will sell their home,

believers decide they will not abandon religious commitments at any price because no payment or penalty can compensate or justify spiritual infidelity.59 In economic terms, the believer’s demand for religious observance is perfectly price inelastic. “Price elasticity” signifies the relative responsiveness of demand for a good to its price.60 The demand for luxury goods is relatively elastic: when their price rises, demand disproportionately drops. By contrast, the demand for necessities, like food, is relatively inelastic: people need to eat, so when the price of food goes up, even substantially, overall demand for food does not proportionately decline (although consumers might substitute hamburger for steak). Price elasticity tells us that people will forego a vacation to Europe before they starve themselves to death.61

The idiosyncratic value believers place on religious observance, Barclay argues, makes voluntary bargaining unlikely, if not impossible, when laws burden religious exercise:

[R]eligious beliefs can operate as very strong obstacles to a voluntary bargain, where one party has claims of conscience that prevent her from behaving in a way that would allow completion of the transaction. Specifically, some religious individuals believe that certain actions put that individual’s immortal soul at risk. Other actions risk less severe punishment but might still lead to disciplinary actions from the religious community or possibly even removal from the community.62

Many believers, she suggests, are impervious to the price they must pay for faithful religious observance, however high it might go, because the price of abandoning observance is even higher—a ruptured spiritual identity here and divine punishment hereafter.63 In economic terms, she assumes that the price elasticity of believers’ demand for religious observance is almost perfectly inelastic: regardless of price, they maintain the same or nearly the same level of observance until the government coerces them to stop.64 Believers, in short, are always martyrs.

From this premise, Barclay draws two conclusions. First, when religious exercise conflicts with general laws, voluntary bargaining to an efficient solution will usually not take place. Because of idiosyncratic religious valuation, “believers often will not respond in a typical way to penalties or incentives, and voluntary bargains based on market values are unlikely to result.”65 Second, Barclay speculates that the evidently high social cost of forcing believers to adhere to general laws that violate their religious conscious will often outweigh the social benefit of uniformly applying such laws: “Where the government seeks to coerce transactions when such idiosyncratic valuations exist, this can backfire and lead to increased net harm, as Coase warned.”66 In both cases, the Coase theorem is invoked to suggest that the socially efficient solution is government accommodation of believers burdened by general laws.

Both the premise that believers are martyrs and the conclusions that Barclay draws from it are erroneous. Recent empirical work by Netta Barak-Corren contradicts the assumption that believers are martyrs—that is, that their demand for religious observance is so uniformly price inelastic that they will invariably disobey anti-discrimination and other laws that burden their religious beliefs or eschew Coasean bargaining altogether.67 Beyond that, Barclay misapplies the concept of idiosyncratic value in discrimination cases and mistakes its place in U.S. law.68

A. Elasticity of Demand for Religious Observance

Professor Barclay assumes that because religious preferences are inelastic and idiosyncratic, Coasean bargaining will not occur: if exemptions are not forthcoming, no civil or criminal penalty will persuade believers to conform to the anti-discrimination norm, leading to high enforcement costs. Yet, if believers are granted exemptions from anti-discrimination laws, nothing LGBTQ persons can offer is likely to persuade believers to forgo exemptions, resulting in denials of service. This squares well with a corollary of the Coase theorem: when transaction costs are high, the initial allocation of liability for negative externalities will be the final allocation because voluntary bargaining is not

available to shift all or part of it elsewhere.69 The initial grant or refusal of religious accommodation, therefore, will determine who bears the costs of religious coercion and religious LGBTQ discrimination, irrespective of whether the initial grant is socially efficient.

This conclusion again raises the problem of incommensurables—the difficulty of pricing infringements on the liberties of believers and third parties to enable comparison of the costs of different mixes of infringements. On the believer side, Posner and McConnell speculate that shadow pricing might provide a way of quantifying the value of religious freedom.70 “Shadow” or “implicit” prices measure the cost of an activity in terms of its opportunity cost, or the cost of participating in related competing activities.71 Since bread and cupcakes share many ingredients, one can measure the implicit cost of bread in terms of cupcakes by calculating how many cupcakes one must forego if the ingredients are diverted to baking a loaf of bread. Shadow prices are often used to estimate the price of nonmarket activities, like parental childcare or family meal preparation, by measuring the value of compensated labor one foregoes to perform uncompensated tasks.72

In principle, one could employ shadow pricing to estimate the price of a nonmarket activity like religion. One might estimate the price believers place on religious observance by what they are willing to pay in punishment and hardship to remain faithful in the absence of accommodation. Many believers are willing to sacrifice employment to honor religious commitments.73 Before incorporation nationalized Bill of Rights protections, persecuted minorities moved from place to place, often under duress, seeking sympathetic laws and enforcement.74 As discussed above, believers with objections to conventional medical treatment are known to risk their children’s health and even their lives to remain faithful.75

Costs of lobbying or litigating for accommodation also suggest a shadow price. Barclay herself suggests religious commitment can be measured by a believer’s willingness to endure negative publicity stemming from unpopular beliefs.76

Contrary to Barclay’s unsupported assumption of martyrdom, Netta BarakCorren has demonstrated that believers are generally unwilling to incur the penalties associated with violating anti-discrimination laws that burden their religious exercise.77 Based on decision experiments involving employment termination with a randomized group of several thousand believers and interviews with religious managers, Barak-Corren concludes that a substantial majority of believers “were inclined to adhere to the law,”78 though they reacted much differently to damage remedies than to reinstatement and other injunctive relief.79 These conclusions suggest that believer demand for religious observance is not normally inelastic,80 as Barclay maintains.

Additionally, shadow pricing is meaningful only for the first accommodation claimant. Once succeeds—legislatively, administratively, or judicially—its price drops dramatically because other believers can claim it at little risk of punishment or other cost to themselves.81

accommodation

claim

an

As Barak-Corren has shown, when the shadow price of accommodation is high because the availability of accommodation is not clear, believers generally conform to the law rather than risk legal sanctions if their accommodation claim is unsuccessful. But once the shadow price of religious accommodation drops nearly to zero because the availability of accommodation is clearly established, one would expect believers to demand accommodation. Barak-Corren also confirmed this hypothesis, finding that religious denials of goods and services to LGBTQ persons are positively correlated with the availability of religious accommodations.82 This likewise contradicts Barclay’s claim of perfectly inelastic demand for religious observance:

[T]he data unsettle the theory that religious objection is a result of permanent idiosyncratic features of the objectors’ religious identity, features that are unyielding to external influence. Rather, it seems that religious objection is contingent on the seeming availability of an exception. The demand for objection is not fixed, but elastic.83

Contrary to Professor Barclay’s speculation that uniformly enforcing antidiscrimination laws against religious objectors generates high net social costs, Barak-Corren provides empirical evidence that the net cost of uniform enforcement might be quite low—that is, that the aggregate cost of believer compliance with burdensome anti-discrimination laws, as measured by believer behavior, might well be less—perhaps substantially less—than the cost to LGBTQ persons from religious exemptions to such laws.84 If most religious objectors to anti-discrimination norms are not martyrs, one can expect, as BarakCorren found, that they will generally comply with anti-discrimination laws to avoid legal and social sanctions—especially when the remedy is injunctive relief—unless and until someone else shoulders the risks and other costs of pursuing an accommodation.85 Similarly, if religious objectors abandon conformity to anti-discrimination norms only when the cost of seeking accommodation is low or when the remedy is only payment of damages, the aggregate social benefit of uniformly enforced anti-discrimination laws might well exceed the aggregate cost to religious conscience when religious accommodations from such laws are denied. Professor Barclay provides no evidence to suggest otherwise.86

But even if there were such evidence, this argument is a double-edged sword for accommodationists like Barclay. If accommodations are justified when they are shown to generate net social benefits, then their denial is justified when they are shown to generate net social costs, as they often seem to. Government could then deny even accommodations whose costs are widely distributed so that they do not externalize material costs onto third parties, on the ground that they generate net costs to society as a whole. This is surely not normatively attractive to accommodationists or, indeed, to many separationists.87

B. Anti-discrimination Laws and Idiosyncratic Value

Apart from whether idiosyncratically high religious values are price elastic or not, Professor Barclay misapplies the concept of idiosyncratic value and misstates its place in U.S. law. Discussing the Arlene’s Flowers decision,88 Barclay characterizes religiously motivated denials to provide goods and services to same-sex weddings as a collision of idiosyncratically high valuation—the florist’s “idiosyncratic” willingness to endure “financial ruin” to remain faithful to her religious beliefs rather than provide flowers for the customer’s same-sex wedding, against the “idiosyncratic” value the customer placed on having this florist “specifically” provide “this particular service” for his wedding.89

By placing idiosyncratic value on both sides of this conflict, Barclay seeks to sidestep the third-party harm doctrine by invoking the reciprocal externalities of the Coase theorem: since the customer is demanding protection for his idiosyncratic valuation of service by this florist, he is harmed no more than the florist who claims protection for her idiosyncratic valuation of the right to refuse

service to this customer.90 Indeed, the florist suffers a greater harm, the argument goes, because the law strongly prefers voluntary market transactions over government-coerced ones: the customer “must justify the ‘narrow exception to the general rule that economic activity should be organized through voluntary market exchanges and not by asking courts to circumvent the market process.’”91 This libertarian assumption presupposes that the florist’s idiosyncratic valuation of the power not to sell her flowers for use in same-sex weddings is protected by a property or liability right, which begs the question whether a right of believers to accommodation from anti-discrimination laws exists at all;92 after all, one purpose of anti-discrimination laws is precisely to regulate otherwise voluntary market transactions to protect minorities from discrimination.

In fact, the law does not invariably protect idiosyncratic valuations, even in real property. The Supreme Court has long held that “just compensation” for a taking of property under the Fifth Amendment is the property’s “fair market value,”93 or “what a willing buyer would pay in cash to a willing seller,”94 expressly excluding idiosyncratic value from the calculus altogether.95

In any event, only one party in Arlene’s Flowers demanded protection of idiosyncratic value, and it was not the customer. There is nothing unusual about a retail customer’s wanting to purchase wedding flowers from a florist he had previously and regularly patronized. Retail businesses routinely seek to maximize revenue and profits by differentiating their products and services from those of competitors—on price, service, quality, and other bases.96 The value a customer finds in these differentiations is not idiosyncratic because any customer can capture it by paying the business’s price; the value of product and service differentiation is available to any customer rather than being peculiar to

a particular one.97 Indeed, there is something perverse in a business that seeks to attract customers by differentiating itself from competitors, but then seeks legal protection for its refusal to serve some of these very customers by suggesting that they value the differentiation “idiosyncratically.”

When it comes to idiosyncratic value, externalities are not inevitably reciprocal. Here, it is only the florist who is demanding legal protection for idiosyncratic value—the religious value she places on denying public retail goods and services to a class of the public otherwise protected by LGBTQ antidiscrimination laws. Most states have determined this is not a value worth protecting.98 The Coase theorem provides no response; it tells us that there is, in principle, some combination of religious freedom and freedom from discrimination that maximizes the level of both freedoms, but it does not tell us how to value either freedom in respect to the other.

Though Coase had little to say about incommensurability, he had plenty to say about economic efficiency in the face of high transaction costs. Coase was not naïve; he acknowledged that transaction costs are often so high that they generate inefficient results.99 In that event, one cannot rely on the parties’ voluntary bargaining to yield the economically efficient mix of competing activities; one must estimate the economically optimal result in some other way.100

Coase suggested three possibilities. The theory of the firm was one.101 If transaction costs leave competing economic activities at an inefficient equilibrium mix, a firm can capture the unrealized surplus by pursuing both activities. “[I]f I found that Coase’s famous grain farmer and cattle rancher were making foolish decisions with respect to the damage to grain from wandering cattle, I would buy the two enterprises and reap a capital gain from an efficient

reorganization.”102 Thus, any inefficiency in the allocation of water-polluting costs between, say, a polluting factory and adjacent homeowners could be corrected by the factory’s acquisition of the homes or the homeowners’ acquisition of the factory (though holdout, collective action, and other transaction costs might stand in the way). A business that owned both the factory and the housing would displace transaction costs with an internal determination of relative economic value. “Within the firm individual bargains between the various cooperating factors of production are eliminated and for a market transaction is substituted an administrative decision.”103 There are costs attached to the firm’s decision making process, but whenever internal decision making is less costly than external bargaining, one may rely on the relative mix of activities determined by the firm as a proxy for the economic optimum that would have been reached between third-party bargainers in the absence of transaction costs.104 Instead of relying on market transactions to arrive at the economically optimal mix of competing uses, one relies on the firm.

The firm is obviously not a solution to negative externalities produced by religious accommodation or anti-discrimination laws. Religions or individual believers cannot acquire all or most public commercial entities, and no market exists for the acquisition of religions. The theory of the firm, however, points to another proxy for the economic optimum—government regulation. Coase suggests that government, in this instance, acts as a kind of “super-firm,” aggregating and discounting in the law-making process the information and preferences attached to competing uses to arrive at an estimated economic optimization.105 Government regulation, of course, entails its own costs; government risks greater inefficiency than a private firm because it is insulated from market forces and thus more likely to mistakenly value the economic costs and benefits of any particular mix of competing activities.106 Since there is always the possibility that government will compound inefficiencies rather than reduce them, a third response to high transaction costs is nonintervention— leaving external costs where they are found, where they have been externalized.107

Barclay does not develop Coase’s suggestion that, when transaction costs are high, the government-as-super-firm might better approximate the socially

efficient mix of competing noneconomic activities in the law-making process than in litigation.108 Rather, she seizes upon Coase’s skepticism of government efficiency to argue that government intervention entails greater costs than its “doing nothing.” There are at least two problems with this libertarian position.109

First, yet again, it runs afoul of incommensurability. How is one to know that government balancing of, say, burdens on religion from anti-discrimination laws and burdens on LGBTQ persons from allowing such discrimination via accommodations is more or less socially efficient than not enacting antidiscrimination laws in the first place? As before, Barclay offers no answers.

Second, it proves too much. The “something” that government does, that Barclay would have the government not do because it purportedly generates net social costs, has already been done—the government has almost always already acted, as in enactment of anti-discrimination protections and social welfare programs. Barclay, in other words, assumes a libertarian or common-law baseline,110 according to which harms to third-parties are legally cognizable only when they infringe upon rights existing in the state-of-nature or at classical common law.111 This libertarian baseline would erase breathtakingly broad swaths of existing statutory and regulatory protections in the name of religious exemptions. How much would Barclay displace to get to the libertarian state that eliminates the negative externalities she attributes to government action? Her position would justify wholesale religious exemptions from the Civil Rights Act of 1964 and state anti-discrimination laws because of the negative externalities they impose on, inter alia, commercial businesses owned or operated by believers who oppose hiring or serving LGBTQ persons (not to mention African Americans and other racial and ethnic minorities).112 It would similarly justify

religious exemptions from taxes mandated by the Social Security Act, Medicare, and other social welfare programs because of the negative externalities they impose on believers who theologically oppose supporting such programs.113

Presumably, Barclay would not go so far. But if one would displace only some (but not all) legislative or regulatory rights or entitlements with religious exemptions, a theory is required to explain why and how some departures from the state of nature or common law are justified while others are not. Barclay does not offer one.

Legislative or regulatory action might well represent the best proxy for the socially optimal mix of competing rights, especially in case of difficult-to-value liberties like freedom of religious exercise and freedom from LGBTQ discrimination. This remains true even if religious accommodation does not emerge from the process and government enacts a uniform rule; it is conceivable that social harm from discrimination exceeds social harm from accommodations permitting such discrimination, particularly in public contexts like retail businesses, government contracting, and government administration.114 Barclay supplies no reason to believe that judicial exemptions achieve social efficiency better than the legislative and regulatory processes.

As a positive science, economics describes the limitations of scarcity and the consequent impossibility of pursuing all worthy goals at once—hence its nickname, the “dismal science.”115 Though it illuminates the costs and benefits of maximizing a value, economics does not tell us which value to maximize.116

Thomas Nechyba notes the “considerable leap” from economic modeling of human behavior to identifying the good: “If the model does not incorporate all the complexities of who we are, is it useful as we move from considering ‘what is’ to ‘what is good’?”117 Coase himself recognized this: while urging that his analysis be extended beyond aggregate economic output measured by the market, he acknowledged that doing so would involve value choices.118 Indeed, the maximization of economic output is itself only one value that other values might displace in the calculus of social cost.

Few theories of the good are uncontroversial in the contemporary United States. Still, if any value is firmly established in our constitutional jurisprudence, it is equality of citizenship. Neither the states nor the federal government may deny to any person the “equal protection of the laws.”119 Equality of citizenship is especially clear in Religion Clause jurisprudence. Government may not make a person’s religion the basis for prosecution under law,120 or eligibility for political office or other positions of public trust,121 or distribution of public benefits, licenses, or privileges.122 Government may not impose burdens on particular religions or all religions that it does not impose on the citizenry in general,123 nor may it prefer belief to unbelief.124 Deep disagreements persist

about how to achieve equality of citizenship and resolve the claims of citizens to equal freedom of religion and belief, but the values themselves are, to employ Michael Perry’s evocative phrase, constitutional “bedrock.”125

The moral roots of religious equality are many and varied. Some are themselves religious, deriving from Christianity and Judaism.126 Within the Western Enlightenment tradition, Kant condemned the use of others solely as instruments to one’s own ends.127 More recently, Rawls argued that rational agents in an original position, lacking knowledge of their place in a future society, would not agree to a system or society which systematically subordinates some people to the interests and choices of others.128 Perry maintains that the metaphorical “spirit of brotherhood” expressly invoked in contemporary human rights documents invests each human being with “moral equality,”129 including freedom from discrimination on the basis of one’s religious beliefs or lack thereof.130

Their starting points and reasoning vary, but all of these justifications for equality share a common commitment to reciprocity. Kant disallows laws that cannot coherently be applied to everyone, whether they bestow burdens or benefits.131 Reciprocity is at the heart of Rawls’s work; his first principle of justice is the equal right of persons “to the most extensive scheme of equal basic

liberty compatible with a similar scheme of liberties for others,” while the second allows departures from socio-economic equality only when this is “reasonably expected to be to everyone’s advantage.”132 Perry would not allow government to treat any citizen as if she were “morally inferior”—”less worthy than someone else” and thus permissibly saddled with burdens for the benefit of others.133

The metaphor of family relationships on which Professor Perry relies—the “spirit of brotherhood”—illuminates the importance of reciprocity. All parents have heard the classic protest that a sibling has gotten special treatment: “It’s not fair!” Most parents deny it; it would be unusual—at least in a functional family—for parents to respond that the favored child is, in fact, morally better than his siblings, more worthy of gifts and deserving of privileged treatment, even at the others’ expense.134 As Rawls observed, “Members of a family commonly do not wish to gain unless they can do so in ways that further the interests of the rest.”135 To treat a family member in the “spirit of brotherhood” is to treat that person as having the same moral worth as other family members and possessed of the expectation that care and concern for her will not be systematically subordinated to everyone or anyone else in the family.136

The third-party harm doctrine reinforces the reciprocity of rights implied by equal citizenship but rejected by many believers. Resistance to LGBTQ antidiscrimination laws is an obvious example. These believers claim an asymmetric entitlement: a right of exemption from laws protecting freedom from discrimination while insisting for themselves on the very anti-discrimination protections they deny to others. Fulton v. City of Philadelphia, handed down just last year and celebrated by accommodationists, permits religious social service agencies to discriminate against LGBTQ persons even though they exercise delegated state authority, operate with government money, and themselves claim the protection of anti-discrimination laws.137

Asymmetric believer claims for accommodation not only violate the reciprocity at the heart of equal citizenship, theyalso, to return to economics, constitute classic rent-seeking behavior. “Rent-seeking” is the effort to create or preserve market advantages through government fiat instead of competition.138 As Paul Horwitz has suggested, the effort of religious institutions to obtain both constitutionally guaranteed access to government funds and constitutionally guaranteed religious accommodations from anti-discrimination laws might well be a form of rent-seeking, allowing “churches to enhance both their status and their revenue base by providing social services while being paid by the state to do so” and, simultaneously, granting them “immunity from any regulatory conditions connected to those public funds.”139

“[I]t seems quite incredible,” Rawls observed, “that some citizens should be expected, on the basis of political principles, to accept still lower prospects of life for the sake of others.”140 For all they have written against the third-party doctrine, accommodationists have yet to offer an ethical theory that justifies its rejection. The Coase theorem is no help. It cannot tell us the socially optimal solution to conflicts between religious exercise and LGBTQ anti-discrimination protections, nor does it show that denials of accommodation are socially inefficient. Far from providing a “clinical” approach to the problem of religious accommodations, application of the theorem to third-party harms betrays a bias against anti-discrimination rights and social welfare entitlements that have

immeasurably improved the lives of many Americans, and a bias in favor of judicial intervention over legislative and regulatory bargaining. The theorem cannot identify (nor was it designed to) those social values that define when an external cost may even count as a “cost.”

Most critically, Coasean analysis has nothing to say about the political-moral question to which the third-party harm doctrine is an answer: should religious accommodations empower believers to impose the costs of living their religion on others who believe and live differently? As Professor Perry argued, the “spirit of brotherhood” that animates the international human rights guarantees “entails the right to moral equality: the right of every human being to be treated as the moral equal of every other human being.”141 This right forbids the government to treat any of its citizens in a way that “discounts if not disregards, their welfare or abilities—thereby failing to treat them as moral equals.”142 To force someone to pay for the religious commitments of another is to make her morally subordinate to that other, a mere means to the fulfillment of someone else’s religious ends.

Even when applied correctly, then, Coasean analysis fails to explain, let alone to justify, why a Constitution that guarantees equal citizenship and religious equality should permit the government to force some to bear the burdens of religious exercise by others.

--- FOOTNOTES ---

This essay is available in Emory Law Journal: https://scholarlycommons.law.emory.edu/elj/vol71/iss7/6

PREFACE

 Guy Anderson Chair & Professor of Law, Brigham Young University Law School. J.D., University of Southern California; B.A. (economics), M.A. (comparative studies–philosophy), Brigham Young University. I am indebted to my BYU law and economics colleagues Paul Stancil and Cree Jones for helping me relearn economic concepts I had not thought about for many years, and to Michael Ransom, recently retired from the BYU Economics Department, for suggesting a reliable text on microeconomics. I also received helpful comments and criticisms from my other BYU Law colleagues, from participants in workshop sessions sponsored by the Midwestern Law & Economics Association and the Nootbaar Institute of the Pepperdine Caruso School of Law, and from Netta Barak-Corren, Stephanie Barclay, Tom Berg, Alan Brownstein, Andy Koppelman, Chip Lupu, Nate Oman, Brandon Paradise, Micah Schwartzman, and Bob Tuttle. Iantha Haight and Melanie Coleman of the BYU Hunter Law Library provided their usual outstanding work with sources. Errors that remain are mine.

1 MICHAEL J. PERRY, THE CONSTITUTION, THE COURTS, AND HUMAN RIGHTS: AN INQUIRY INTO THE

LEGITIMACY OF CONSTITUTIONAL POLICYMAKING BY THE JUDICIARY (1982).

TABLE OF CONTENTS

INTRODUCTION: RELIGIOUS ACCOMMODATION, THIRD-PARTY HARMS, AND THE COASE THEOREM

2 See, e.g., RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW § 29.8, at 972–74 (9th ed. 2014); Karl Bade, Comment, Scylla, Charybdis & Adam Smith: An Economic Analysis of the Religion Clauses, 39 DEPAUL L. REV. 1235 (1990); Paul Horwitz, Freedom of the Church Without Romance, 21 J. CONTEMP. LEGAL ISSUES 59, 89–95 (2013); Keith N. Hylton, Yulia Rodinova & Fei Deng, Church and State: An Economic Analysis, 13 AM. L. & ECON. REV. 402 (2011); Eric A. Posner, The Legal Regulation of Religious Groups, 2 LEGAL THEORY 33, 60–62 (1996); Michael W. McConnell & Richard A. Posner, An Economic Approach to Issues of Religious Freedom, 56 U. CHI. L. REV. 1 (1989).

3 See, e.g., County of Allegheny v. ACLU, 492 U.S. 573, 613 n.59 (1989); see also Mark Tushnet, Accommodation of Religion Thirty Years On, 38 HARV J.L. & GENDER 1, 2 (2015) (defining religious accommodation as government adjustment of its general programs to relieve adverse effects of such programs on religious belief).

4 Frederick Mark Gedicks & Rebecca G. Van Tassell, RFRA Exemptions from the Contraception Mandate: An Unconstitutional Accommodation of Religion, 49 HARV. C.R.-C.L. L. REV. 343, 363 (2014) [hereinafter Gedicks & Van Tassell, RFRA Exemptions].

I first suggested the third-party harm doctrine in response to legal challenges to the contraception mandate of the Affordable Care Act. See Frederick Mark Gedicks, Establishment Clause Limits on Permissive Accommodations of Religion: RFRA and the Contraception Mandate, Panel Presentation at the University of Virginia College of Law, International Consortium for Law and Religious Studies Conference (Aug. 23, 2013). Rebecca Van Tassell (now Powell) and I elaborated the doctrine in two subsequent papers. See Gedicks & Van Tassell, RFRA Exemptions, supra; Frederick Mark Gedicks & Rebecca Van Tassell, Of Burdens and Baselines: Hobby Lobby’s Puzzling Footnote 37, in THE RISE OF CORPORATE RELIGIOUS LIBERTY 323 (Micah Schwartzman et al. eds., 2016) [hereinafter Gedicks & Van Tassell, Of Burdens and Baselines]. Andrew Koppelman and I discussed it further in another series of papers. See Frederick Mark Gedicks & Andrew Koppelman, Invisible Women: Why an Exemption for Hobby Lobby Would Violate the Establishment Clause, 67 VAND. L. REV. EN BANC 51 (2014); Frederick Mark Gedicks & Andrew Koppelman, Response, The Costs of the Public Good of Religion Should Be Borne by the Public, 67 VAND. L. REV. EN BANC 185, 186–87 (2014); Andrew Koppelman & Frederick M. Gedicks, Is Hobby Lobby Worse for Religious Liberty than Smith?, 9 U. ST. THOMAS J.L. & PUB. POL’Y 223 (2015).

5 THOMAS J. NECHYBA, MICROECONOMICS: AN INTUITIVE APPROACH 468–69 (2d ed. 2017); WALTER NICHOLSON & CHRISTOPHER SNYDER, MICROECONOMIC THEORY: BASIC PRINCIPLES AND EXTENSIONS 683 (12th ed. 2017); see also Julian Reiss, Public Goods, STANFORD ENCYCL. PHIL. (July 21, 2021), https://plato. stanford.edu/entries/public-goods/ (“Externalities are effects of economic transactions on individuals that are not party to the transaction.”).

6 Gedicks & Van Tassell, RFRA Exemptions, supra note 4, at 358–62. 7 Cf. Zubik v. Burwell, 578 U.S. 403, 408 (2016) (per curiam) (vacating lower court decisions that had

denied categorical exemptions from the contraception mandate to religious nonprofit entities).

I am conceptualizing a “market for religious fidelity,” which encompasses costs and benefits of observing one’s religion, including costs attached to violation of the contraception mandate. An exemption from

the mandate reduces the price of religious fidelity to the employer (by relieving it of otherwise applicable penalties for violating the coverage mandate) at the expense of the employee (who must now pay for contraception out of pocket). A cost of religious fidelity, in other words, has been “externalized”—it is borne by employees who do not share the employer’s anti-contraception beliefs and thus receive no benefit from the lower price of fidelity, instead of being borne by the employer, who receives all the benefits of fidelity at a lower price. In short, the costs of fidelity are no longer fully reflected in its market price (or, what amounts to the same thing, the employee is forced to bear some of the costs of a transaction—exemption from the mandate—to which the employee was not a party and from which the employee derives no benefit). Under the standard economic definitions, this is an externality. See supra note 5 and accompanying text; see also Joshua A. Decker, Markets in Everything and Another View of the Cathedral: Religious Freedom and Coasean Bargaining, 26 STAN. L. & POL’Y REV. 485, 491 n.34 (2015) (characterizing loss of contraception coverage from religious exemption as a “negative externality” borne by employees); Developments in the Law—Reframing the Harm: Religious Exemptions and Third-Party Harm after Little Sisters, 134 HARV. L. REV. 2186, 2199 & n.84 (2021) (arguing the same). One can argue that an exemption from the mandate, while perhaps unfair to employees, does not externalize the costs of fidelity because this cost is fully internalized by the wage market—i.e., an exempt employer must raise wages to account for the value of the lost coverage or risk losing employees to other employers who pay a higher effective wage by including coverage in their health plans. On this view, shifting the cost of exemption from the mandate is the result of the imperfectly competitive character of many wage markets; if employment is “sticky” or subject to transactions costs—because, say, employees would not easily find comparable employment elsewhere—anti-contraception employers may consistently pay less than the market price for religious fidelity at the expense of their employees, who would consistently make less than the effective market wage that includes coverage. Though we disagree on this point, I am indebted to Nate Oman for pressing it and prompting me to clarify my position.

9 This is essentially how the Court decided the Hobby Lobby case. See Burwell v. Hobby Lobby Stores, Inc., 537 U.S. 682, 692, 730–35 (2014) (extending already-existing nonprofit religious accommodation to closely held business corporation because it relieved shareholders of obligation to supply objectionable contraceptives while preserving beneficiaries’ access without cost sharing and requiring the government to pay any resulting expense); see also id. at 729–30 (suggesting that direct provision by government of objectionable contraceptives would have been a less restrictive alternative to the mandate).

10 See Brief for Church-State Scholars Frederick Mark Gedicks et al. as Amici Curiae Supporting Petitioners at 3–6, Hobby Lobby, 537 U.S. 682 (Nos. 13-354 & 13-356); Frederick Mark Gedicks, Paying for the Boss’s Beliefs, WASH. POST, Jan. 20, 2014, at A15.

Justice Kennedy endorsed the doctrine in Hobby Lobby, 537 U.S. at 738–39 (Kennedy, J., concurring), as did Justice Ginsburg in a dissent joined by three other Justices, id. at 764, 768–69 (Ginsburg, J., dissenting). See also Holt v. Hobbs, 574 U.S. 352, 370 (2015) (Ginsburg, J., concurring in the grant of exemption because it would not harm third parties). Kennedy, however, also joined the Hobby Lobby majority opinion, which acknowledged the doctrine but limited it to extreme and unlikely situations. See Hobby Lobby, 537 U.S. at 729 n.37; Frederick Mark Gedicks, One Cheer for Hobby Lobby: Improbable Alternatives, Truly Strict Scrutiny, and Third-Party Employee Burdens, 38 HARV. J.L. & GENDER 153, 168–72 (2015). And, of course, neither Kennedy nor Ginsburg is any longer on the Court.

12 Critics include KATHLEEN A. BRADY, THE DISTINCTIVENESS OF RELIGION IN AMERICAN LAW: RETHINKING RELIGION CLAUSE JURISPRUDENCE 259–65 (2015); Stephanie H. Barclay, First Amendment “Harms,” 95 IND. L.J. 331, 353–55, 362–64 (2020); Thomas C. Berg, Religious Accommodation and the Welfare State, 38 HARV. J.L. & GENDER 103, 131–32 (2015); Marc O. DeGirolami, Free Exercise by Moonlight, 53 SAN DIEGO L. REV. 105, 131–35 (2016); Marc DeGirolami, On the Claim that Exemptions from the Mandate Violate the Establishment Clause, MIRROR JUST. (Dec. 5, 2013), https://mirrorofjustice.blogs.com/mirrorofjustice/2013/ 12/exemptions-from-the-mandate-do-not-violate-the-establishment-clause.html; Carl H. Esbeck, Do Discretionary Religious Exemptions Violate the Establishment Clause?, 106 KY. L.J. 603 (2018); Christopher C. Lund, Religious Exemptions, Third-Party Harms, and the Establishment Clause, 91 NOTRE DAME L. REV. 1375 (2016); Gene Schaerr & Michael Worley, The “Third Party Harm Rule”: Law or Wishful Thinking?, 17 GEO. J.L. & PUB. POL’Y 629 (2019). For a cogent defense of the doctrine and synthesis of previous arguments, see Micah Schartzman, Nelson Tebbe & Richard Schragger, The Costs of Conscience, 106 KY. L.J. 781 (2018). For other defenses, see, e.g., NELSON TEBBE, RELIGIOUS FREEDOM IN AN EGALITARIAN AGE 49–70 (2017); Kara Loewentheil, When Free Exercise Is a Burden: Protecting “Third Parties” in Religious Accommodation Law, 62 DRAKE L. REV. 433 (2014); Douglas Nejaime & Reva B. Siegel, Conscience Wars: Complicity-Based Conscience Claims in Religion and Politics, 124 YALE L.J. 2516, 2529–33 (2014). 13 R. H. Coase, The Problem of Social Cost, 3 J.L. & ECON. 1 (1960). 14 See NECHYBA, supra note 5, at 483–84. 15 Coase defined “transaction costs” as expenses incurred in overcoming obstacles to agreement, such as discovering with whom one has to bargain, informing that person one wishes to bargain and on what terms, engaging in actual negotiations, drawing up an agreement if negotiations are successful, monitoring and enforcing the agreement, and so on. Coase, supra note 13, at 15; see also NECHYBA, supra note 5, at 761 (defining transaction costs as “barriers that keep people from getting together to bargain their way out of an externality problem”); NICHOLSON & SNYDER, supra note 5, at 438 (same as “costs associated with making market transactions”).

16 Coase, supra note 13, at 3–7. The “economically efficient mix” of two competing activities is the

maximum value of the combined production from both activities. Id. at 15–16.

Ironically, it was not Coase but George Stigler who coined the “Coase theorem.” R.H. COASE, THE

FIRM, THE MARKET, AND THE LAW 158 (1988).

18 See, e.g., Stephanie H. Barclay, An Economic Approach to Religious Exemptions, 72 FLA. L. REV. 1211, 1215–18 (2020); Jason J. Meulhoff, Case Comment, A Ministerial Exemption for All Seasons: Our Lady of Guadalupe School v. Morissey-Beru, 140 S.Ct. 2049 (2020), 45 HARV. J.L. PUB. POL’Y, 465, 479 & n.59 (2022). Additionally, Thomas Berg has relied on Professor Barclay’s argument in unpublished work and Laurence Solum “highly recommended” it. See Laurence Solum, Barclay on an Economic Approach to Religious Exemption, LEGAL THEORY BLOG (Oct. 21, 2020), https://lsolum.typepad.com/legaltheory/2020/10/barclay-onan-economic-approach-to-religious-exemptions.html.

19 Barclay, supra note 18, at 1228; e.g., Meulhoff, supra note 18, at 475 n.59 (arguing that negative externalities from the ministerial exception are reciprocal, because removing “a minister harms that employee, but being forced to keep an unwanted minister also harms the religious group’s ability to function in the manner it deems best”); see also id. at 1217 (“[T]he existence of a religious externality would not provide a sufficient normative ground for government restriction of rights” because of the “ubiquitous nature of competing and complex reciprocal externalities involved in the protection of any legal rights.”).

20 Cf. Fulton v. City of Philadelphia, 141 S. Ct. 1868 (2021) (holding requirement that city contractors not discriminate on the basis of sexual orientation inapplicable to Catholic social service agency that refused to certify same-sex couples as adoptive or foster parents); Masterpiece Cakeshop, Ltd. v. Colo. C.R. Comm’n, 138 S. Ct. 1719 (2018) (holding state law prohibiting sexual-orientation discrimination inapplicable to evangelical baker who refused to bake a cake celebrating same-sex union).

21 Again, I am conceptualizing a market for religious fidelity, with an analysis similar to that used for exemptions from the contraception mandate. See supra note 8 and accompanying text. In this context, one cost of fidelity is the legal penalty a believer would incur for violating anti-discrimination laws when religious commitments preclude her from selling goods or services to a person belonging to a protected class like LGBTQ persons. When the government exempts a commercial business owner from the obligations and penalties provided by such laws, a cost of religious fidelity is thereby externalized onto LGBTQ persons whom the laws would otherwise protect from discrimination. One can argue that exemptions from anti-discrimination laws do not generate externalities in a perfectly competitive market because the goods and services that an exempt business refuses to sell to members of a protected class are available elsewhere at little additional cost. This presupposes that anti-discrimination laws protect only bare access to goods and services, excluding from protection, for example, rights to personal dignity or equal citizenship. See, e.g., Frederick Mark Gedicks, Dignity and Discrimination, 46 BYU L. REV. 961, 966–69 (2021) (equal citizenship); Marvin Lim & Louise Melling, Inconvenience or Indignity? Religious Exemptions to Public Accommodations Laws, 22 J.L. & POL’Y 705 (2013) (personal dignity). If mere access to goods and services were all that anti-discrimination laws protected, however, regimes of separate-but-equal LGBTQ access would be constitutional, in contradiction to well-settled and universally acclaimed equal protection jurisprudence. Cf. Brown v. Board of Educ., 347 U.S. 483, 495 (1954) (finding racially segregated “separate-but-equal” public education is “inherently unequal”); see also Newman v. Piggie Park Enterprise, 390 U.S. 400, 402 n.5 (1968) (summarily rejecting believer’s free exercise defense to unlawful denial of restaurant service to African Americans). Accordingly, I will assume that when government grants religious exemptions to commercial businesses from anti-discrimination laws that protect LGBTQ customers, the availability of identical goods and services elsewhere does not eliminate all the externalities generated by exemption.

22 Barclay, supra note 18, at 1217–18. I take “clinical” to suggest “scientific” in the sense of procedurally

and evidentially constrained.

I. COASE AND COMMENSURABILITY

23 See infra Part I. 24 See infra Part II. 25 See infra Part III 26 See infra Part IV. 27 Henrik Andersson & Nien-hê Hsieh, Incommensurable Values, STANFORD ENCYCL. PHIL., https:// plato.stanford.edu/archives/fall2021/entries/value-incommensurable/ (July 14, 2021). For simplicity, I have passed over distinctions among the many versions of incommensurability and the related concept of incomparability. See id. §§ 1–2.

28 See, e.g., Daniel O. Conkle, Equality, Animus, and Expressive and Religious Freedom Under the American Constitution: Masterpiece Cakeshop and Beyond, in LES MUTATIONS DE LA DE LA LIBERTÉ D’EXPRESSION EN DROIT FRANÇAIS ET ÉTRANGER (Panthéon-Assas ed. 2021).

29 McConnell & Posner, supra note 2, at 51; see also Decker, supra note 8, at 502 (“[R]eligious practices

are hard to value and compare: . . . it is hard to monetize them.”).

30 See infra Part I.A. 31 Coase, supra note 13, at 1–2; see Steven G. Medema, Non-Welfarism in the Early Debates over the Coase Theorem: The Case of Environmental Economics, in WELFARE THEORY, PUBLIC ACTION, AND ETHICAL VALUES: REVISITING THE HISTORY OF WELFARE ECONOMICS 208, 208–09 (Roger E. Backhouse et al. eds., 2021).

Barclay incorrectly maintains that Coase “discredited” the Pigouvian approach (and, by implication, the third-party harm doctrine). See Barclay, supra note 18, at 1263–64, 1268. To the contrary, economists continue to teach Pigouvian taxes as one of several plausible solutions to the problem of negative externalities. See, e.g., NECHYBA, supra note 5, at 747–48; NICHOLSON & SNYDER, supra note 5, at 691–93.

32 Roger E. Backhouse, Antoinette Baujard & Tamotsu Nishizawa, Introduction, in WELFARE THEORY, PUBLIC ACTION, AND ETHICAL VALUES, supra note 31, at 1,1 (quoting A.C. PIGOU, THE ECONOMICS OF WELFARE 11 (1920 ; see also Backhouse, Baujard & Nishizawa, supra, at 7 (“The old welfare economics of Marshall and Pigou has been said to restrict welfare-relevant information to utility or economic welfare, believed to be measurable in money.”). Pigou later extended his analysis to include equity and other noneconomic values, see id. at 9–10, but this did not figure into Coase’s analysis.

33 See Coase, supra note 13, at 43 (“In this article, the analysis has been confined . . . to comparisons of the value of production, as measured by the market.”); see also id. at 2 (When straying cattle trample crops on an adjoining farm, “[t]he nature of the choice is clear: meat or crops. What answer should be given is, of course, not clear unless we know the value of what is obtained as well as the value of what is sacrificed to obtain it.”); id. at 15 (“The economic problem in all cases of harmful effects is how to maximise the value of production.”).

34 Decker, supra note 8, at 502.

35 Coase, supra note 13, at 2–3. 36 See infra Part I.B. I am grateful to Andy Koppelman for this example. 37 Barclay, supra note 18, at 1217 (quoting Coase, supra note 13, at 26, 44). 38

Id. at 1247, 1247 n.230 (citing Bendix Autolite Corp. v. Midwesco Enters., 486 U.S. 888, 897 (1988)

(Scalia, J., concurring .

Id. at 1247–48. As Barclay recognizes, most states do not mandate vaccination as a matter of law but as a condition to accessing public benefits, such as public schools, and attending sports, entertainment, and other large and crowded venues. Id. at 1247–49 & n. 245.

40 Id. at 1250–51. I assume with Barclay that parental interests are coextensive with those of their children. 41 She conditions this conclusion on the availability of educational alternatives and a finding that herd

immunity is actually threatened by exemptions. See id. at 1251–52.

Id. at 1251.

42 43 See id. at 1252 (Requiring objecting parents to home school their unvaccinated children “seems

normatively defensible if the goal is to decrease net harm.”).

44 Milton Friedman, Lange on Price Flexibility and Employment: A Methodological Criticism, 36 AM. ECON. REV. 613, 624 (1946); see also Robert Premus & James A. Swaney, Modern Empiricism and Quantum Leap Theorizing in Economics, J. ECON. ISSUES 713, 720 (1982) (criticizing the tendency of many economists to “invert the scientific process by developing fully formulated theoretical structures based on casual empiricism, hunches, and unrealistic assumptions”).

45 Cass R. Sunstein, Incommensurability and Valuation in Law, 92 MICH. L. REV. 779, 798 (1994). 46 Cf. id. at 803–04 (arguing that it is rational to choose one incommensurable good over another if one

values it intrinsically rather than merely instrumentally).

47 See, e.g., John L. Diamond, Rethinking Compensation for Mental Distress: A Critique of the Restatement (Third) §§ 45-47, 16 VA. J. SOC. POL’Y & L. 141 (2008); Stanley Ingber, Rethinking Intangible Injuries: A Focus on Remedy, 73 CAL. L. REV. 772 (1985); Joseph H. King, Pain and Suffering, Noneconomic Damages and the Goals of Tort Law, 57 SMU L. Rev. 163 (2004).

48 See, e.g., Emp. Div. v. Smith, 494 U.S. 872, 886–87 (1990); Thomas v. Rev. Bd., 450 U.S. 707, 716 (1981); United States v. Ballard, 322 U.S. 78, 86 (1944); see also Watson v. Jones, 80 U.S. (13 Wall.) 679, 728 (1871) (decided at federal common law).

49 As I elaborate in Part II.A, infra, believers may themselves betray by their actions that they value religious fidelity rather less than one might imagine. Nevertheless, courts are prohibited from calculating this value externally.

50 Cf. 1 Cor. 6:19–20 (Revised Standard Version) (“Do you not know that your body is a temple of the Holy Spirit within you, which you have from God? You are not your own; you were bought with a price. So glorify God in your body.”). I am grateful to Chip Lupu for suggesting this text.

51 See generally Anne D. Lederman, Note, Understanding Faith: When Religious Parents Decline Conventional Medical Treatment for Their Children, 45 CASE W. RSRV. L. REV. 891 (1995) (parsing applications of various government-interest tests in the context of religious free exercise).

52 See Smith, 494 U.S. at 888–89 & n.5 (1990).

IDIOSYNCRATIC VALUE AND COASEAN BARGAINING

53 A potentially fruitful economic approach to this problem is public good analysis. A “public good” is one that, once produced, may be enjoyed irrespective of whether one pays for it, such as national defense, a freeway system, or a fireworks display. NECHYBA, supra note 5, at 1040–41; NICHOLSON & SNYDER, supra note 5, at 695–96. The efficient solution to payment for public goods is complex and usually entails remedies for market failure. See NECHYBA, supra note 5, at 661; NICHOLSON & SNYDER, supra note 5, at 697–99, 702–03. One common solution is government production of the good, funded by tax dollars or other public financing. See NECHYBA, supra note 5, at 674–75; NICHOLSON & SNYDER, supra note 5, at 701–03.

Herd immunity produced by vaccination might be usefully conceptualized as a public good produced by government vaccination mandates for two reasons. First, the high rate of vaccination necessary to achieve herd immunity prevents circulation of the disease among the entire population, including the unvaccinated who do not pay for or “free ride” on this benefit. Second, a relatively few unvaccinated free-riders—often as little as five percent of the population—can degrade and ultimately destroy herd immunity for everyone. See, e.g., Coronavirus Disease (Covid-19): Herd Immunity, Lockdowns and Covid-19, WORLD HEALTH ORG., https://www. who.int/news-room/questions-and-answers/item/herd-immunity-lockdowns-and-covid-19 (Dec. 31, 2020) (describing different thresholds for herd immunity). If there is reason to believe that affording religious exemptions would reduce the vaccinated percentage below the level necessary to maintain herd immunity, the economics of public goods would justify refusing religious exemptions. Noneconomic arguments, of course, could point in other directions.

54 Barclay, supra note 18, at 1231-33. 55 See Idiosyncratic, OXFORD ENGLISH DICTIONARY ONLINE, https://www-oed-com.erl.lib.byu.edu/view/ Entry/91048?redirectedFrom=idiosyncratic#eid (last visited May 1, 2022) (defined as “attributable to individual disposition, susceptibility, or character”).

56 See generally John Fee, Eminent Domain and the Sanctity of Home, 81 NOTRE DAME L. REV. 783

(2006) (discussing the problems idiosyncratic value poses for eminent domain law).

57 See Barclay, supra note 18, at 1232. 58

Id. at 1233–34.

Id. at 1227–28, 1234–35.

59 60 See NECHYBA, supra note 5, at 637; NICHOLSON & SNYDER, supra note 5, at 159. 61 Concepts related to price elasticity of demand include income and cross elasticity of demand. See NICHOLSON & SNYDER, supra note 5, at 159–61. These concepts suggest that demand for accommodation to protect religious fidelity might well be affected by changes in believer income or the availability of religious substitutes; for simplicity, however, I have confined the analysis to price elasticity (which, in any event, is thought to be the most important elasticity concept in microeconomics). See id. at 160.

62 Barclay, supra note 18, at 1234. 63

Id. at 1236 (discussing Christopher C. Lund, Martyrdom and Religious Freedom, 50 CONN. L. REV.

959, 965 (2018 .

Id. at 1234–36.

Id. at 1234; see also id. at 1237 (“One need not agree with or even understand religious beliefs to recognize them as obstacles to a voluntary transaction . . . . [T]he important takeaway is that religious beliefs sometimes make voluntary transactions highly unlikely, in part because the transaction is perceived as much costlier by the religious individual than others might perceive it.”).

Id. at 1235; see also id. at 1237 (“If . . . religious individuals are unwilling to perform the voluntary transaction, government penalties or coercion might simply increase net harm rather than facilitate efficient transactions.”).

67 See infra Part II.A. 68 See infra Part II.B.

69 See Coase, supra note 13, at 15–16. 70 McConnell & Posner, supra note 2, at 52. 71 NICHOLSON & SNYDER, supra note 5, at 193. 72 See id. at 193–94. 73 See, e.g., Sherbert v. Verner, 374 U.S. 398 (1963) (Seventh Day Adventist terminated for refusal to

work Saturday shifts).

74 Seeking a place to practice their religion in peace, nineteenth-century Mormons left Ohio voluntarily for Missouri; within Missouri, they were driven from Jackson County to Clay County, then to Illinois, and, finally, to the interior West, from which there was nowhere left to go. Joseph Smith Abandons Ohio, HISTORY, https://www.history.com/this-day-in-history/joseph-smith-abandons-ohio (Jan. 11, 2022). Amish in the United States likewise repeatedly moved during the nineteenth and early-twentieth centuries in search of state laws and governments sympathetic to their counter-cultural way of life. See Brief for Respondents at 14–15, Wisconsin v. Yoder, 92 S. Ct. 1526 (1972) (No. 70-110); Donald B. Kraybill, Amish, EVERY CULTURE, https://www. everyculture.com/multi/A-Br/Amish.html (last visited May 1, 2022).

75 See supra note 61 and accompanying text.

76 See Barclay, supra note 18, at 1234–35. 77 Netta Barak-Corren, Does Antidiscrimination Law Influence Religious Behavior? An Empirical

Examination, 67 HASTINGS L.J. 957, 992–96 (2016).

Id. at 993.

78 79 Barak-Corren’s results suggested that damage awards erode adherence to anti-discrimination norms going forward because religious employers perceive damages as the purchase of a license to discriminate. By contrast, religious employers perceive reinstatement of a wrongfully terminated employee as a conflict of values on which the government is unwilling to compromise, resulting in their greater adherence to anti-discrimination norms going forward. See id. at 1009–12.

80 Netta Barak-Corren, A License to Discriminate? The Market Response to Masterpiece Cakeshop, 56

HARV. C.R.-C.L. L. REV. 315, 365 (2021).

81 See McConnell & Posner, supra note 2, at 52. 82 Barak-Corren, supra note 80, at 362–66. Based on field surveys of retail businesses offering wedding

goods and services, Barak-Corren found that LGBTQ discrimination substantially increased following the Supreme Court’s judgment in favor of the objecting believer in Masterpiece Cakeshop. See id. at 345–47. AntiLGBTQ discrimination increased even in jurisdictions that had no anti-discrimination laws, suggesting that increased denials of service to LGBTQ persons were motivated by the expressive effect of the law—for example, the Constitution exempts believers from religiously burdensome anti-discrimination laws—and not just by fear of legal sanctions. See id. at 355–57, 359–60; see also id at 365 (“[W]edding vendors changed their behavior in the absence of any state penalty and absent any likelihood of enforcement.”).

83 Barak-Corren, supra note 80, at 365. 84 Of course, one can alter this conclusion considerably by manipulating what counts as a cost. Charles Barzun and Michael Gilbert, for example, conclude that LGBTQ persons are the “least cost avoider[s]” in conflicts between religious wedding vendors and LGBTQ customers, because wedding goods and services are almost always available for purchase from another willing vendor. Charles H. Barzun & Michael D. Gilbert, Conflict Avoidance in Constitutional Law, 107 VA. L. REV. 1, 38–41, 45 (2021). They can make their analysis work in this context, however, only with a contestable move that excludes from the definition of “cost” infringements on citizenship, dignity, and reasonable expectations of participants in public markets. See id. at 27–30. But see supra note 21 (suggesting that anti-discrimination laws protect equal citizenship and personal dignity in addition to the bare access to goods and services); cf. Aaron Tang, Harm-Avoider Constitutionalism, 109 CAL. L. REV. 1847, 1890–91 (2021) (arguing in other constitutional contexts that courts should defer to the claimant’s definition of harm to determine the proper cost avoider).

85 See Barak-Corren, supra note 80, at 365–66. 86 Barclay points to the withdrawal of Catholic social service agencies from the fostering and adoption

market in jurisdictions that insist on uniform application of LGBTQ anti-discrimination laws. Barclay, supra note 18, at 1235–37. She provides no evidence, however, that other social service agencies, including other religiously sponsored ones, did not move into the market space vacated by Catholic agencies to take up some or all of the slack in supply of fostering and adoption services. Nor did she supply any evidence suggesting that the aggregate social cost here, if any, was not compensated by the aggregate social benefit of increased freedom of fostering and adoption applicants from LGBTQ discrimination.

I am grateful to Alan Brownstein for this point. I emphasize here that I am challenging only the descriptive implications of Barclay’s argument that (1) believers are always martyrs, (2) the social costs of martyrdom are exceptionally high, therefore (3) denials of accommodation are necessarily inefficient. Though the social cost of accommodation is a factor to consider— indeed, the third-party harm doctrine itself seeks to avoid a social cost—I reject the normative position implied by the argument from reciprocal externalities, that efficiency alone should govern the availability of accommodations.

88 State v. Arlene’s Flowers, Inc., 193 Wash.2d 469 (2019) (upholding application of anti-discrimination law to florist who refused on religious grounds to provide flowers for same-sex wedding of longtime customer), cert. denied sub nom. Arlene Flowers, Inc. v. Washington, 141 S. Ct. 2884 (2021).

89 Barclay, supra note 18, at 1253.

Id. at 1252; see also id. at 1253–54 (citing Todd J. Zywicki, Libertarianism, Law and Economics, and the Common Law, 16 CHAP. L. REV. 309, 321 (2013 (“[A]pplying Coasean economic theory to a conflict caused by idiosyncratic valuations would generally not justify the government forcing a transaction to occur.”).

Id. at 1254 (quoting Zywicki, supra note 90).

91 92 Cf. Posner & McConnell, supra note 2, at 10 (cautioning that it is “far from certain” that “members of

[a] minority faith have, as it were, a property right in their religious beliefs”).

93 Olson v. United States, 292 U.S. 246, 253–55 (1934). 94 United States v. Miller, 317 U.S. 369, 374 (1943). 95 See United States v. Miller, 317 U.S. 369, 375 (1943) (“[A]lthough the market value of the property is to be fixed with due consideration of all its available uses, its special value to the condemnor as distinguished from others who may or may not possess the power to condemn, must be excluded as an element of market value.”). See generally Brian A. Lee, Just Undercompensation: The Idiosyncratic Premium in Eminent Domain, 113 COLUM. L. REV. 593 (2013) (arguing that paying idiosyncratic premiums under the Takings Clause is unnecessary and would be unjust).

96 NECHYBA, supra note 5, at 329–30; NICHOLSON & SNYDER, supra note 5, at 535–41.

III. HIGH TRANSACTION COSTS AND THE (SUPER) FIRM SOLUTION

97 The situation would be different if the florist routinely provided products and services to the

complaining customer that she did not provide to others at any price.

98 Elizabeth Sepper has shown that the vast majority of state public accommodations laws prohibiting discrimination on the basis of race, ethnicity, and sex by retail commercial establishments provide only narrow religious exemptions or no such exemptions at all. See Elizabeth Sepper, The Role of Religion in State Public Accommodations Laws, 60 ST. LOUIS U.L.J. 631, 653-58 (2016). This pattern has generally held with respect to exemptions for religiously motivated LGBTQ discrimination, with exemptions only for religious organizations with respect to venues or services they provide at the celebration or solemnization of marriages to which they religiously object. See id. at 659–62.

99 Coase, supra note 13, at 15–16. 100 101

Id. at 16. Id. at 16–17.

102 George Stigler, Two Notes on the Coase Theorem, 99 YALE L.J. 631, 631 (1989). 103 Coase, supra note 13, at 16. 104 105 106 107

Id. at 16–17. Id. at 17. Id. at 17–18. Id. at 18.

108 Professor Barclay does endorse one legislative solution, the so-called “Utah Compromise,” as having correctly weighed the incommensurable interests of believers and LGBTQ persons to arrive at a socially optimal solution. Barclay, supra note 18, at 1242–43. The Compromise statutorily prohibited LGBTQ discrimination in housing and employment and guaranteed the availability of persons in county offices to perform same-sex marriages, while affording robust religious accommodations. See UTAH CODE ANN. §§ 17-20-4(2), 34A-5-106, 57-21-5 (2015). Regrettably, the conditions that yielded the Compromise rarely exist and when they do, the compromises they produce are easily undone by the courts.

The bargaining conditions that led to the Compromise are akin to those existing in a bilateral monopoly, whose signal characteristic is that buyer and seller must deal only with each other or not at all. See, e.g., ROGER D. BLAIR & JEFFREY L. HARRISON, MONOPSONY IN LAW AND ECONOMICS 123–45 (2010); NICHOLSON & SNYDER, supra note 5, at 491, 589; Fritz Machlup & Martha Taber, Bilateral Monopoly, Successive Monopoly, and Vertical Integration, 27 ECONOMICA 101 (1960); James N. Morgan, Bilateral Monopoly and the Competitive Output, 63 Q.J. ECON. 371 (1949). The conditions under which the Utah Compromise was struck were analogous to cooperative bargaining between bilateral monopolists. After the Supreme Court constitutionalized LGBTQ access to civil marriage in Obergefell v. Hodges, religious conservatives in Utah—notably The Church of Jesus Christ of Latter-day Saints and its leaders and members, who comprise more than sixty percent of the population and nearly ninety percent of the state legislature—faced a future of uncertain access to religious accommodations in light of likely judicial expansion of LGBTQ rights. See Lee Davidson, Latter-Day Saints Are Overrepresented in Utah’s Legislature, Holding 9 of Every 10 Seats, SALT LAKE TRIBUNE, https://www.sltrib.com/news/politics/ 2021/01/14/latter-day-saints-are/ (Jan. 19, 2021). At the same time, LGBTQ persons in Utah lacked any common law or statutory protection against private discrimination, possessing only the bare access to civil marriage ordered by the Court. Both sides were deeply committed to their positions, but both faced future uncertainties about which available information was symmetric. LGBTQ activists in Utah could obtain LGBTQ antidiscrimination protections only by bargaining with the Church, and the Church could obtain legal protections for its observance of traditional marriage norms only by bargaining with Utah LGBTQ activists—an apparent (if short-lived) bi-lateral monopoly. It is doubtful such a confluence of circumstances will often (or ever) recur. No socially conservative religion dominates state politics like Latter-day Saints do Utah’s; religious conservatives elsewhere face considerable collective action costs, as do LGBTQ activists. Uncertainties have also shifted since 2015; most states have enacted LGBTQ anti-discrimination laws and LGBTQ activists are consequently now less willing to compromise these protections. See Nondiscrimination Laws, MOVEMENT ADVANCEMENT PROJECT, https://www.lgbtmap.org/equality-maps/non_discrimination_laws (last visited May 1, 2022).

Finally, even successful bilateral legislative or regulatory bargains are easily undone by the courts in the current constitutional climate. Increasing judicial receptivity to religious accommodations has made it difficult to trust that legislative and regulatory compromises will hold up when they are challenged in later free exercise litigation, often by the very parties who struck the original bargains. Micah J. Schwartzman, Judicial Compromise and Political Uncertainty (or, What If You’re Not Sugar Ray Robinson?), BALKINIZATION (July 22, 2020), https://balkin.blogspot.com/2020/07/judicial-compromise-and-political.html. Judicial reversal of regulatory bargains struck over the contraception mandate is a cautionary example. E.g., Zubik v. Burwell, 578 U.S. 403 (2016); Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014); see Leslie C. Griffin, The Catholic Bishops vs. the Contraception Mandate, 6 RELIGIONS 1411, 1419–25 (2015).

109 A third is that it might not be a correct reading of Coase in the absence of market and firm solutions when transaction costs are high. To be sure, Coase opined that economists tend to exaggerate the social benefits of government intervention. Coase, supra note 13, at 18. But he also suggested that, while “direct government regulation will not necessarily give better results than leaving the problem to be solved by the market or the firm,” “equally there is no reason why, on occasion, such governmental administrative regulation should not

lead to an improvement in economic efficiency. This would seem particularly likely when . . . a large number of people are involved and . . . therefore the costs of handling the problem through the market or the firm may be high.” Id. Coase might well have found, therefore, that the solution to conflicts between religious exercise and LGBTQ anti-discrimination rights calls precisely for legislative or regulatory rather than judicial intervention. 110 State-of-nature and classical common-law baselines for measuring harm have been rejected by accommodationists and separationists alike. See, e.g., McConnell & Posner, supra note 2, at 6–8; Elizabeth Sepper, Religious Exemptions, Harm to Others, and the Indeterminacy of a Common Law Baseline, 106 KY. L.J. 661 (2017); see also Gedicks & Van Tassell, Of Burdens and Baselines, supra note 4 (arguing for a “socialwelfare baseline” as the most plausible contemporary measure of third-party harm).

111 When it comes to government benefits, like funding, Barclay abandons this baseline, maintaining that religion is entitled to access government benefits without satisfying the conditions for access, even when benefits are not part of a natural-law or classical common-law baseline. See Barclay, supra note 18, at 1212–14 (criticizing on economic grounds exclusion of Catholic Social Services from government adoption and fostering funding because it rejected LGBTQ anti-discrimination conditions) (discussing Fulton v. City of Philadelphia, 922 F.3d 140 (3d Cir. 2019), rev’d, 141 S. Ct. 1868 (2021 ; see also Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246 (2020) (holding the exclusion of religious K-12 schools from state scholarship funds was unconstitutional religious discrimination); Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012 (2017) (same with regard to exclusion of such schools from state playground safety program).

112 The Supreme Court has held that the provision of Title VII of the Act, prohibiting discrimination “because of” sex, encompasses discrimination on the basis of sexual orientation and gender identity. See Bostock v. Clayton County, 140 S. Ct. 1731 (2020). Bostock will likely prompt federal courts to give the same reading to prohibitions of sex discrimination in other federal statutes, such as Title IX of the Education Amendments Act, and may influence state courts to read state laws prohibiting “sex” discrimination in a similarly expansive

IV. THE DUBIOUS ETHICS OF HARMING THIRD PARTIES

manner. See Tarrant Cty. Coll. Dist. v. Sims, 621 S.W.3d 323, 328–29 (Tex. Ct. App. 2021) (interpreting state statutory prohibition on “sex” discrimination to encompass sexual-orientation and gender-identity discrimination, on the authority of Bostock). Religious exemptions from racial anti-discrimination protections were rejected out of hand by the Supreme Court. See Newman v. Piggie Park Enters., 390 U.S. 400, 402, 402 n.5 (1968); see also Bob Jones Univ. v. United States, 461 U.S. 574 (1983) (denying tax exemption to Christian university and K-12 school because of racially discriminatory policies).

113 Cf. United States v. Lee, 455 U.S. 252 (1982) (refusing exemption from payment of Social Security taxes to an employer with religious objections to social welfare programs because exemption would reduce employee benefits).

114 See, e.g., Thomas C. Berg & Alan Brownstein, Giving Our Better Angels a Chance: A Dialogue of Religious Liberty and Equality, 21 J. APP. PRACT. & PROCESS 325, 356 (2021) (Religious organizations may properly discriminate with their own money, but special harm is generated when they are permitted to discriminate with government money.) (argument of Brownstein).

115 AGNAR SANDMO, ECONOMICS EVOLVING: A HISTORY OF ECONOMIC THOUGHT 67, 67 n.6 (English

trans. 2011).

116 See, e.g., POSNER, supra note 2, at 16 (“Because economics cannot tell us whether the existing distribution of income or wealth is good or bad, just or unjust, the economist cannot issue mandatory

prescriptions for social change.”).

117 NECHYBA, supra note 5, at 717. 118 Coase, supra note 13, at 43 (“In this article, the analysis has been confined . . . to comparisons of the value of production, as measured by the market. But it is, of course, desirable that the choice between different social arrangements for the solution of economic problems should be carried out in broader terms than this and that the total effect of these arrangements in all spheres of life should be taken into account. As Frank H. Knight has so often emphasized, problems of welfare economics must ultimately dissolve into a study of aesthetics and morals.”). Professor Knight was Coase’s colleague at the University of Chicago during the mid-twentieth century and a distinguished economist in his own right who wrote on, inter alia, the relationship of economics to politics and morality. See, e.g., FRANK HYNEMAN KNIGHT, THE ETHICS OF COMPETITION AND OTHER ESSAYS (1935); Frank H. Knight, Abstract Economics as Absolute Ethics, 76 ETHICS 163 (1966). As this fuller context makes clear, Coase is referring here to the inevitability of choosing among theories of value and accounts of the good once one ventures beyond comparisons of economic output, and not, as Barclay incorrectly claims, to “downstream” or “secondary” externalities. See Barclay, supra note 18, at 1223.

119 U.S. CONST. amend. XIV; Bolling v. Sharpe, 347 U.S. 497 (1954) (applying the Equal Protection

Clause to the federal government through the Fifth Amendment Due Process Clause).

120 E.g., AMERICAN BAR ASSOCIATION, CRIMINAL JUSTICE STANDARDS FOR THE PROSECUTION FUNCTION 3-1.6 (4th ed. 2017), https://www.americanbar.org/groups/criminal_justice/standards/ProsecutionFunctionFourth Edition/.

121 E.g., McDaniel v. Paty, 435 U.S. 618 (1978); Torcaso v. Watkins, 367 U.S. 488 (1961). 122 E.g., Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012 (2017); Zelman v.

Simmons-Harris, 536 U.S. 639 (2002).

123 E.g., Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 521 (1993); Larson v.

Valente, 456 U.S. 228, 229 (1982).

124 E.g., McCreary County v. ACLU, 545 U.S. 844, 860 (2005); Epperson v. Arkansas, 393 U.S. 97, 104

(1968); Everson v. Bd. of Educ., 330 U.S. 1, 18 (1947).

125 MICHAEL J. PERRY, A GLOBAL POLITICAL MORALITY: HUMAN RIGHTS, DEMOCRACY, AND

CONSTITUTIONALISM 60 (2017).

126 See, e.g., Leviticus 19:34 (Jewish Publication Society) (“The stranger that sojourneth with you shall be unto you as the home-born among you, and thou shalt love him as thyself; for ye were strangers in the land of Egypt.”); id. at 24:22 (“Ye shall have one manner of law, as well for the stranger, as for the home-born.”); Luke 10:27 (Revised Standard Version) (“Love . . . your neighbor as yourself.”).

Immanuel Kant, Groundwork of the Metaphysics of Morals (1785), in PRACTICAL PHILOSOPHY 37, 80 (Mary J. Gregor ed. & trans., 1996) (“So act that you use humanity . . . always at the same time as an end, never merely as a means.” (emphasis omitted .

128 See, e.g., JOHN RAWLS, A THEORY OF JUSTICE 13 (rev. ed. 1999) [hereinafter RAWLS, THEORY OF JUSTICE] (“Offhand it hardly seems likely that persons who view themselves as equals, entitled to press their claims upon one another, would agree to a principle which may require lesser life prospects for some simply for the sake of a greater sum of advantages enjoyed by others.”); see also JOHN RAWLS, POLITICAL LIBERALISM 41 (1999) (A “political conception of justice . . . means that citizens do not think there are antecedent social ends that justify them in viewing some people as having more or less worth to society than others.”).

129 PERRY, supra note 125, at 47. 130

Id. at 73 (“[T]he right on which the countries of the world have converged protects the freedom to practice one’s morality without regard to whether one’s morality is religiously based . . . .” (emphasis omitted ; id. at 80 (“That the free exercise right is an antidiscrimination right is not disputed” in American constitutional law.).

131 Kant, The Metaphysics of Morals (1797), in PRACTICAL PHILOSOPHY, supra note 127, at 363, 387 (“Any action is right if it can coexist with everyone’s freedom in accordance with a universal law, or if on its maxim the freedom of choice of each can coexist with everyone’s freedom in accordance with a universal law.”); see also id. at 379 (“The categorical imperative . . . is: act upon a maxim that can also hold as a universal law.”).

132 RAWLS, THEORY OF JUSTICE, supra note 128, at 53. Rawls expressly labels the second part of the second principle a “conception of reciprocity,” and a “principle of mutual benefit.” Id. at 88; see also id. at 29 (“[W]e think of a well-ordered society as a scheme of cooperation for reciprocal advantage regulated by principles which persons would choose in an initial situation that is fair . . . .”).

133 See PERRY, supra note 125, at 57 (“[W]hat government may not do is deny a benefit to anyone or impose a cost on anyone . . . based on the view (or on a sensibility to the effect) that she is morally inferior: less worthy than someone else, if worthy at all, of being treated ‘in a spirit of brotherhood.’” (emphasis omitted ; see also id. at 60–61 (“If it is not “rational”—reasonable, plausible—to believe that a particular instance of the government’s disadvantaging some persons relative to some other persons serves a “legitimate” government interest—if it is not “rational” to believe that a particular instance of such disadvantaging serves . . . the common good—then presumably government . . . is simply “playing favorites” (by disfavoring some persons relative to some others) . . . .”).

In the Bible, at least, this never turns out well. See Genesis 4 (relating the story of Cain, who killed his brother Abel after God rejected Cain’s offering but accepted Abel’s); Genesis 37 (relating the story of favored son Joseph, sold into slavery by his jealous older brothers).

135 RAWLS, THEORY OF JUSTICE, supra note 128, at 90; see also id. at 91 (the difference principle

corresponds to “fraternity.”).

136 This is more my conclusion than Perry’s, though he did join the Hobby Lobby amicus brief urging the Court to adopt the third-party harm doctrine. See Brief for Amici Curiae Church-State Scholars Frederick Mark Gedicks et al., Burwell v. Hobby Lobby Stores, Inc., 537 U.S. 682 (2014), 2014 WL 333891, at *1–*2; see also Michael J. Perry, Two Constitutional Rights, Two Constitutional Controversies, 52 CONN. L. REV. 1597, 1611 (2021) (The right to moral equality forbids government to treat any of its citizens in a way that “discounts, if not disregards, their welfare or abilities, thereby failing to treat them as moral equals.”); id. at 1643 (“[C]onstitutionally protected moral freedom [includes] freedom from government action withholding benefits bestowed on others who make a different choice in exercising their moral freedom.”).

CONCLUSION: THE COASE THEOREM SAYS NOTHING ABOUT THIRD-PARTY HARMS

137 141 S. Ct. 1868 (2021). 138 NECHYBA, supra note 5, at 542; NICHOLSON & SNYDER, supra note 5, at 492, 495. 139 Horwitz, supra note 2, at 131. 140 RAWLS, THEORY OF JUSTICE, supra note 128, at 154.

141 PERRY, supra note 125, at 56. 142 Perry, supra note 136, at 1611.