A Common-Law Model for Religious Exemptions

Eugene Volokh

(46 UCLA Law Review 1465 (1999))


Who decides when a religious believer should be granted an exemption from a generally applicable rule? Under the constitutional exemption model set up by Sherbert v. Verner the answer is "judges." Under the statutory exemption model of Employment Division v. Smith the answer is "legislatures."  The Religious Freedom Restoration Act (RFRA) and its recent state analogues have been billed as ways to restore the constitutional exemption model.

In this Article, Professor Eugene Volokh argues that the RFRAs actually implement a third model--a "common-law exemption model" in which exemption decisions are initially made by courts but are ultimately revisable by legislatures--and that this approach is better than both of the other models.  He goes on to (1) explain how state RFRAs (and the federal RFRA, as applied to federal laws) track the traditional common-law development of other subconstitutional claims of right; (2) argue that this protects the RFRAs against some of the criticisms of the Sherbert regime levied by Smith; (3) suggest that drafters of the RFRAs should abandon their current reliance on the "strict scrutiny" test; (4) argue that viewing religious exemption claims through a common-law lens helps show the fallacy of the Sherbert constitutional exemption framework; and (5) suggest that the common-law rights model might be profitably applied to certain other proposed exemption regimes, such as a regime of limited exemptions for news gathering.


I. The State RFRA Model as a Common-Law Exemption Regime
A. The Religious Exemption Problem

B. State RFRAs as a Common-Law Model

C. Common-Law Statutory Exemptions

D. The Common Law and Harm to Others

E. The Practical Effects of a Common-Law Exemption Regime

1. Shifting the Burden of Overcoming Legislative Inertia

2. Shifting the Decision Making from Executive and Local Decision Makers

3. Encouraging a Broader Scope of Exemptions

4. The Difficulty of Predicting Specific Results

F. The Legitimacy of the Common-Law Model After Smith

G. The Problem of Preference for Religion

II. Why State RFRAs Should Not Be Phrased in Strict Scrutiny Terms
A. The Sherbert-Era Regime Was Not a Pure Strict Scrutiny Regime

B. False Analogies to Other Strict Scrutiny Regimes

C. The Impossibility of a Single Formula and the Superiority of a Pure Common-Law Approach

D. Proposal for a Pure Common-Law State RFRA

III. The Difficulties with the Constitutional Exemption Regime

A. Three Kinds of Free Exercise Clause Claims

1. Several Specific Prohibitions on Government Action

2. A General Regime of Exemptions for Conduct Even When It Harms Others

3. A General Regime of Exemptions for Conduct That Does Not Harm Others

B. The Private Rights and Interests of Others

1. The Debates About What Constitutes the Private Rights and Interests of Others

2. The Connection to the Early 1900s Substantive Due Process Debates

C. Implications for the Three Kinds of Free Exercise Claims

D. Objections to the Substantive Due Process Analogy

1. Text and Original Meaning

2. The Relative Rarity of Religious Exemption Claims

3. The Possibility of Legislative Error

4. The Risk of Suppression of Minority Interests

5. The Possible Special Importance of Religiously Motivated Liberty Claims

6. Questioning the Rejection of the Early 1900s Substantive Due Process Cases

7. Does This Analysis Jeopardize Constitutional Rights Protection Generally?

E. Can a Constitutional Exemption Model Nonetheless Be Made to Work?

1. Looking to Exceptions

2. Looking to the Treatment of Majority Religious Groups

3. Looking to the Constitutional Pedigree of the Claimed Countervailing Right

4. Looking to the Facts

5. Developing New Narrow, Specially Justified Rights

F. Implications

1. Rejecting Federal and State Constitutional Exemption Schemes

2. Rejecting Federal Common-Law Exemption Schemes

3. Embracing Legislative Responses to Common-Law Exemption Making

a. The Need to Draw Lines Between Practices

b. The Early 1900s Experience

c. An Example

4. Embracing the Facial Exclusion of Certain Laws from State RFRAs

G. Common-Law Exemptions and Diversity

IV. Applications to Other Claims of Liberty



      When should religious objectors get exemptions from generally applicable laws?  Should a landlady who has religious objections to renting to unmarried couples be exempted from antidiscrimination law? Should a person whose religion views peyote or marijuana use as a sacrament be exempted from drug laws? Should a mother who feels a religious duty not to testify against her daughter be exempted from the legal duty to testify?(1)

      Equally importantly, who should decide these matters? From 1963 to 1990, the Supreme Court took the view that such decisions should ultimately be made by judges. Under what I'll call the constitutional exemption model, the Court concluded that the Free Exercise Clause generally mandated religious exemptions: A law that substantially burdened religious objectors could be applied to them only if it passed "strict scrutiny" (i.e., served a "compelling government interest" in the least restrictive way possible) and it was up to courts to decide whether this test was satisfied.(2)

      In Employment Division v. Smith (1990),(3) the Court reversed itself, shifting to what I'll call the statutory exemption model.  The Free Exercise Clause, the Court held, does not require exemptions; whether an exemption is available should be up to the legislature.

      Following Smith, Congress adopted the Religious Freedom Restoration Act (RFRA), which purported to restore strict judicial scrutiny of all federal and state laws that substantially burden religious practice, mandating religious exemptions wholesale, rather than one by one as under the statutory exemption model.(4) When the Court held RFRA unconstitutional as applied to state laws in City of Boerne v. Flores (1997),(5) supporters of exemptions shifted to proposing RFRA-like statutes in each state; these laws, too, purported to restore strict scrutiny.(6) The rhetoric in support of RFRA and of the state RFRAs focused on restoration of the constitutional exemption model--a return to having courts decide when exemptions are available.(7)

      But, I argue in Parts I.A-D, this state RFRA approach actually implements a different regime, which I call the common-law exemption model. As under the constitutional exemption model, state RFRAs (and the federal RFRA applied to federal laws)(8) let courts decide in the first instance whether an exemption is to be granted. But because RFRAs may be revised by the legislature, the courts' decisions aren't final. Ultimately, the tough calls will be governed by the political process,(9) just as they have been in the common-law system under which American law has generally evolved.(10)

      From this observation flow some other conclusions. First, as I explain in Part I.E, the common-law framework gives exemption supporters some of what they want, despite the possibility of legislative override.  By letting courts take the lead in carving out exemptions for conduct that, in the courts' view, doesn't really harm others, the framework shifts the burden of overcoming legislative inertia to favor exemption supporters. Exemptions would thus be rejected only based on a considered judgment, rather than by mechanical application of laws that may have been enacted without any consideration of their possible effect on religious practices. State RFRAs would also make it especially hard for local governments or executive officials to burden religious exercise, though legislatures can, if they want to, avoid this effect using provisions that I describe in Part I.E.2. And, as I explain in Part I.E.3, state RFRAs may embolden judges to carve out exemptions that they might not have carved out under the constitutional exemption model.

      Second, recognizing that RFRAs follow the common-law model, rather than the constitutional model, shows that Employment Division v. Smith's criticism of the constitutional exemption model does not condemn state RFRAs, or the federal RFRA as applied to federal law. As I explain in Part I.F, while a constitutional exemption mandate would give courts too much final discretionary decision-making power, the common-law tradition shows the legitimacy of giving courts initial discretionary decision-making power.

      Third, the similarity between the question courts should ask when applying the exemption regime and the question they ask when developing the common law--when should people's liberty be constrained by the rights and interests of others?--helps show the error of imposing an across-the-board strict scrutiny test. Obviously the complex pattern of common-law rules couldn't be reduced to any such test; likewise for religious exemptions. As I argue in Part II, instead of mandating such a necessarily misleading formula, legislatures should explicitly instruct courts to develop the religious exemption regime using a common-law process. Though this will give courts fairly little initial guidance, the strict scrutiny test doesn't give courts much sound guidance either, and it gives them some incorrect guidance to boot.

      Fourth (and independent of the first three points, which should be acceptable even to supporters of a constitutional exemption model), the common-law regime is better than the constitutional exemption regime. As Part III argues, the constitutional exemption model was flawed for the same reason that the early 1900s substantive due process regime was flawed. "I may do what I like, so long as I don't harm others" is an undeniably appealing principle, but who is to decide--in religious freedom cases or in substantive due process cases--what constitutes impermissible harm to others, and whether such harm is in fact likely to flow from a certain practice? Is it a harm when I refuse to rent you my property? Do I harm you if I refuse to testify on your behalf when your life, liberty, or property is at stake? Would allowing some marijuana use cause harm by increasing street crime or addiction of children? What about peyote?

      There is no judicially discoverable answer to these questions dictated by logic or by the Constitution. Answering the questions is an exercise of moral and practical judgment about what "harm" means and what actually causes harm, and the rejection of the early 1900s substantive due process jurisprudence tells us that this judgment must ultimately be left to the political process: Judges have no basis for conclusively saying that something that the legislature considers harmful is actually harmless.

      The Free Exercise Clause, like other constitutional provisions, does secure a right to do certain categories of things--for instance, discriminate based on race or sex in hiring clergy--even when they harm others. I have no quarrel with such focused religious freedom protections, or with similar protections provided by other constitutional provisions (such as the Free Speech Clause) to certain kinds of conduct that harms others.

      But as I argue in Part III.A.2, such a constitutionally secured right to harm others can't be based only on the religious motivation for one's conduct. A general Free Exercise Clause(11) right to engage in a wide variety of conduct simply because the conduct is religiously motivated would be normatively acceptable only if it were limited to conduct that does not harm others, and such a limited right, while appealing, is ultimately not judicially administrable.

      U.S. law has thus generally resisted giving judges the last word on what constitutes harm to others, but it has often given them the first word. Most of the basic rules defining the boundary between people's rights and the protection of others' rights and interests--the rules of tort law, contract law, property law, even in part criminal law--were originally crafted by judges exercising their moral and practical judgment about what constitutes harm. Many such rules are still being developed by judges today, sometimes under statutes that instruct judges to carve out common-law exceptions from broader statutory schemes. The state RFRA approach, with its initial delegation of discretionary authority to judges coupled with the possibility of legislative modification, fits well into this common-law framework.(12)

      Fifth, as I argue in Parts III.F.3-4, legislative exclusions of certain laws or government actions--for instance, antidiscrimination laws or prison rules--from state RFRAs often make sense. These exclusions aren't attempts to undermine the exemption regime; rather, they are an inherent part of the common-law-making process, where legislatures review and occasionally supersede judicial decisions about the strength of countervailing private rights and interests.

      Finally, I briefly suggest in Part IV that the common-law model may make sense for some other kinds of individual rights, rights that ought not be constitutionalized but that nonetheless deserve protection through something more than just specific statutory exemptions. Consider, for instance, exemptions for news gathering. Many general laws were created with little thought about the special problems raised when people violate these laws for news-gathering purposes, just as many laws were written with little thought about the special concerns of religious objectors. A legislature may avoid the mechanical application of these laws in situations where they may be inapt by delegating to courts the power to carve out exemptions for news-gathering activities. Courts would then determine, subject to legislative revision, which news-gathering activities should be allowed because they don't sufficiently harm others (perhaps, for instance, lies or certain technical trespasses in the course of an undercover investigation), and which should be forbidden.

      The religious exemption question exposes a gap in our traditional thinking about law. Religious freedom jurisprudence has generally veered between two polar models of the relationship of courts to legislatures. In the Sherbert regime--which the RFRA rhetoric has generally aimed at "restoring"--courts did all the heavy lifting, and what they lifted stayed in place (absent a constitutional amendment). In the Smith regime, the assumption was that courts played almost no role, and all the decisions were up to the legislature.

      But state RFRAs suggest (perhaps unexpectedly) a third model, which assigns both courts and legislatures a role in developing the law of religious exemptions. This model, I argue, will in the long run lead to a better reconciliation of religious freedom claims and countervailing private rights, just as our common-law system of tort, contract, and property law has produced better results than could have arisen under either an entirely constitutionalized or an entirely codified system of private law.

      I.The State RFRA Model as a Common-Law Exemption Regime

      A. The Religious Exemption Problem

      Religious observers cannot be exempted from all generally applicable laws that burden religious practices. Human sacrifice, religiously compelled seizure of another's property, and the like must be punishable. On the other hand, there has often been broad support for having some religious exemptions, for instance from alcohol prohibition, compulsory military service, and oath requirements.(13)

      The questions have thus been: When should such exemptions be granted, and who should decide when they should be granted? Until 1963, the general answer seemed to be that the matter was up to the legislature, but in Sherbert v. Verner,(14) the Court launched the constitutional exemption regime--whenever a law substantially burdened religious practice and couldn't pass muster under strict scrutiny, a religious exemption was constitutionally compelled.

      In 1990, Employment Division v. Smith largely overruled Sherbert and reinstituted the old statutory exemption model.(15) The Free Exercise Clause, the Court held, generally does not compel religious exemptions. Religious objectors should ask the legislature for specific statutory exemptions, as objectors had done (often successfully) in the past.

      But rather than lobbying for specific exemptions from specific laws burdening religion, supporters of the Sherbert model persuaded Congress to enact the 1993 Religious Freedom Restoration Act. RFRA purported to require religious exemptions from all state and federal laws unless the laws passed strict scrutiny; as its name suggests, RFRA was billed as a way to reinstate by statute the old constitutional exemption regime.(16)

      In City of Boerne v. Flores (1997),(17) the Court concluded that RFRA could not constitutionally apply to state governments, because Congress lacked the power to so broadly restrict enforcement of state laws. But Boerne left RFRA in place as a restraint on federal action(18) and didn't preclude the enactment of state RFRAs as restraints on each state government. Following Boerne, exemption advocates have in fact turned to lobbying state legislatures to enact state RFRAs.

      B. State RFRAs as a Common-Law Model

      The debate over RFRA and state RFRAs has operated on the assumption that a state RFRA regime reinstates the constitutional exemption framework--that, as under Sherbert, a religious exemption is judicially guaranteed whenever a law substantially burdens religious practice and can't pass muster under strict scrutiny.(19) And indeed the state RFRAs, though statutes, don't fit the statutory exemption model. Under the statutory model, the legislature decides when a particular exemption should be granted; under state RFRAs, with their relatively vague standards of "compelling government interest" and "least restrictive alternative," the decision is made by a court.

      But state RFRAs, being state statutes, can be modified by the legislatures that enacted them. Thus, under these state RFRAs (or under the federal RFRA as applied to federal law), the answer to the question "When should religious exemptions be granted, and who should decide when they should be granted?" is "When the courts think such an exemption doesn't unduly jeopardize compelling state interests and the legislature hasn't specifically overridden this judgment."(20) This represents a significant and underappreciated change from the constitutional exemption model, under which courts had the final say.

      I call the state RFRA approach a "common-law exemption model" because it embodies two aspects of common-law decision making. Unlike the statutory exemption model, and like the constitutional exemption model, state RFRAs delegate the initial decision to judicial discretion: Both the state RFRAs' text and the existing free exercise strict scrutiny case law are so vague that they rarely bind courts to any particular results.(21)

      But unlike the constitutional exemption model and like the statutory exemption model, state RFRAs leave the final decision to legislative discretion. Say a court concludes that a ban on marital-status discrimination in housing fails strict scrutiny, and a legislature disagrees, perhaps because it believes that tenants have a right to be free from such discrimination. The legislature can then simply enact a new statute: "Whereas we conclude that marital status housing discrimination violates tenants' rights, the statute banning marital status discrimination in housing shall be applied without regard to [the state RFRA]."

      Such a revision would face some political difficulty. It always takes effort to get anything on the legislative agenda, and many legislators might be hesitant to cast what can be characterized as "A vote against religious freedom."(22) But religious freedom isn't the only thing that arouses popular sentiments. Legislators voting on whether to retain the exemption might be reluctant to cast what some will call "A vote in favor of discrimination"; the fact that many state RFRA proposals have themselves been defeated(23) suggests that RFRAs need not be sacred cows. A proposed bill carving out an exemption from a state RFRA would be politically similar to other bills, such as ones that reverse common-law decisions imposing various forms of tort liability: It will pass or fail depending on the breadth, depth, and savvy of its political support and opposition.(24)

      After a while, then, a state's religious exemption law will end up being mostly a creation of the courts, but with legislative modifications. When the legislature concludes that a court was too stingy with exemptions from some statute, it will enact an explicit religious exemption. When the legislature concludes that a court was too generous, it will specifically provide that the statute has no exemption. The result will be much like the tort law, contract law, and property law of many states--a basic body of judge-made law, with alterations imposed by legislators when they disagree with the judges.

      C. Common-Law Statutory Exemptions

      Some may find it odd to hear a statutorily authorized system of exemptions from statutes described as a "common-law" framework; the common law is usually seen as an entirely judicially created construct. But there is ample precedent for this approach. The congressionally enacted Federal Rules of Evidence, for instance, dictate that testimonial privileges are to "be governed by the principles of the common law as they may be interpreted by the courts of the United States in the light of reason and experience."(25) Some environmental laws use the same language to authorize federal courts to develop affirmative defenses.(26) The Copyright Act of 1976 authorizes courts to develop fair-use principles using their own discretion,(27) though it provides more guidance than do the Rules of Evidence.

      The Trademark Anti-Dilution Act also allows a fair-use defense.(28) Though it doesn't specifically tell courts to develop such a defense using their own discretion, such an instruction has to be implied: Because the dilution cause of action is new, courts must create fair-use law on their own. Likewise, Title VII authorizes courts to carve out exemptions from the sex discrimination ban--necessarily using their own discretion--under the rubric of "bona fide occupational qualification."(29) In all these contexts, the legislature has delegated certain decisions to be made initially by judges. State RFRAs are somewhat more sweeping than these statutes, because they apply to the whole body of state law, rather than just to one area, such as copyright, discrimination, or antitrust. Still, the core principle--delegation to courts of the power to create exemptions, coupled with the possibility of legislative power to revise the exemptions--is the same, and seems equally apt regardless of the statute's breadth.(30)

      In these examples, the legislature apparently expected that there would be (1) a minority of situations (2) that it was unlikely to anticipate adequately (3) in which some concern is morally or practically significant enough to justify a result different from what the statute presumptively requires. If these situations were very common, the legislature would probably have felt obliged to consider them more carefully. If they could be adequately anticipated, then the legislature would likely have set forth a statutory answer for them. And if they weren't significantly different from the core case to which the statute would presumptively apply, there would have been no reason to authorize courts to create special rules for them. But when the three conditions above are present, it makes sense for the legislature to let courts make law for such cases as they come up, defining the private rights involved through a common-law approach rather than a purely statutory one.(31)

      As a class, general laws that may incidentally burden religious practice satisfy at least the first two of the three criteria I outlined above. First, they tend to give rise to religious objections only in a small minority of applications. Second, in a country that has many religious sects, considerable creation of new sects and immigration of people from other religious traditions, and a tendency for individuals or small groups to come up with their own interpretations of religious doctrine, the various objections may often be hard to anticipate.(32) Legislators often pass general statutes without considering that what to most people is a modest, easily bearable restraint might be a much greater imposition on a minority religious group,(33) especially when the group is little known or badly organized politically, or when it arrives or arises after the law is enacted. Requirements of bright signs on slow-moving vehicles might be an example(34)--many lawmakers may have voted for such requirements thinking that they provide some safety advantage without imposing any serious burden on motorists, and not realizing that to the Amish the requirements may indeed be a substantial burden.

      The harder question is whether the presence of a religious objection is morally or practically significant: Is there reason for the legislature to think that a result that it sees as right in the absence of a religious objection--for instance, people may not use peyote or discriminate in housing against unmarried couples--may become wrong when the objection is present?

      Some might think that the answer is no--that personal beliefs, regardless of their basis or their importance to the person, usually don't matter.(35) This might be especially true when the private right vindicated by the law is particularly clear: Thus, I think most would agree that a personal religious belief does not give one a right to intrude on others' property, to defame them, or to breach one's contracts with them.

      But the existence and the seeming propriety of many statutory exemptions suggest that at least sometimes an objector's religious beliefs (or, more broadly, conscientious beliefs(36)) should influence the result. The law treats people with conscientious objections to war differently from those who have no such objections, even when the different treatment increases the risk of death for the nonobjectors. Federal law and the laws of many states exempt religiously motivated peyote users, though the general peyote ban is premised on the notion that peyote use tends to cause various harms.(37)

      The legislatures that enacted these exemptions have apparently decided that the presence of conscientious objections can make a difference, and it seems to me--and, in my experience, to most other observers--that at least in some such cases this conclusion is correct. A law that mandates something that is contrary to someone's deep conscientious beliefs imposes a serious burden on the objector, often a much greater burden than it imposes on those who lack such a deep belief on the issue. It's quite right for legislatures to want to ease this burden, especially when easing the burden doesn't harm others.(38) And if this is a sufficiently likely scenario--if conscientious objections can be relevant to whether the law should be applied to the objector, and such objections are both rare and hard to anticipate--then it may make sense for legislatures to let courts create such exemptions wholesale (subject to legislative repeal), rather than requiring objectors to lobby the legislature for specific statutory exemptions.

      D. The Common Law and Harm to Others

      That religious exemption regimes fit well with the common-law tradition should not be surprising. The principle of religious exemptions is often defended on the theory that "people should be free to do what their religions tell them to do, so long as they don't harm others";(39) but of course American law more generally begins (in at least rough outline) with the principle that "people should be free to do what they want, so long as they don't harm others."

      The real debate, as I discuss in more detail in Part III, is about what "harming others" means, and this debate has traditionally been resolved through the common-law model. Judges creating tort law, contract law, and property law have initially defined what constitutes harm to others. Then, when legislatures have disagreed--for instance, when they concluded that alienating the affections of another's spouse shouldn't be treated as an actionable harm,(40) or when they concluded that race discrimination in hiring should be treated as a harm--they have revised the law accordingly. Likewise, under the common-law religious exemption model, determining which religious practices don't harm others and should thus be allowed is initially up to the courts, subject to modification by legislatures.

      Consider an example: claims by churches that their religious uses should be exempted from zoning laws.(41) Traditionally, the legal system has assumed that people are entitled to use their real estate however they please, so long as such use doesn't harm others. Neighbors were protected against harm by the common law through nuisance doctrine, but as legislative bodies concluded that certain uses created harms that the common-law rules might not recognize--for instance, traffic congestion, excessive crowding, or loss of privacy--they modified the nuisance law regime with zoning laws.(42)

      Thus, when religious exemptions are similarly urged on the grounds that people are entitled to follow their religious teachings so long as their actions don't harm others, it likewise makes sense to have courts initially define what constitutes harm, subject to legislative modification. If a court concludes that having a soup kitchen at a neighborhood church, which would normally be an impermissible use under a local zoning ordinance, doesn't really harm neighbors--doesn't infringe their rights or inflict improper externalities on them--it can carve out an exemption. But if the legislature later concludes that the court's decision did not properly consider all the ways in which such a use can inflict harm, the legislature can supersede this judgment, just as zoning laws superseded certain nuisance law principles.

      I am not suggesting that courts must follow common-law rules in deciding on exemptions, for instance by following common-law nuisance doctrine in deciding whether a certain activity should be exempted from zoning laws. Both the judgments about harm embodied in the zoning laws themselves and the special concerns raised when an exemption claim rests on religious grounds may lead the court to create a rule for religious uses that differs both from the common-law nuisance rules and from the statutory zoning rules. But it makes sense for the legal system to make these decisions by relying on the common-law method of initial judicial decision making coupled with the possibility of legislative reconsideration.

      E. The Practical Effects of a Common-Law Exemption Regime

      A common-law exemption regime may generally help religious objectors less than a constitutional exemption regime would, but it gives them significantly more protection than the statutory exemption model does.

      1.Shifting the Burden of Overcoming Legislative Inertia

      Most importantly, the common-law regime shifts the burden of overcoming legislative inertia. To get a statutory exemption, even a fairly uncontroversial one, religious objectors must get space on the legislative agenda, something that may be hard if the religious group is small, unpopular, or not politically well-organized.(43) (The separatist religious beliefs of some groups, such as the Amish, may make it especially hard for them to mobilize politically.) It will also take time; while the objectors are waiting for the legislature to act, they may be forced by the law into what they believe to be sinful behavior.

      And winning in the legislature tends to require not just majority agreement but a mild supermajority:(44) The bill must navigate the legislative committee system and then be accepted by both houses of the legislature and signed by the executive.(45) If the statutory exemption is controversial, it might be rejected even if it has majority support but not the required supermajority.

      Under the common-law exemption framework, if a court rules in favor of an exemption, the objectors--for instance, members of American Indian religions asking for exemption from a peyote ban--will immediately get the relief they need. The legislature could still revisit this question and specifically reject the possibility of exemptions by enacting a flat ban. But to do that the lawmakers would have to confront the fact that the flat ban would hurt a particular religious group, and perhaps confront a record of some months or years during which the exemption was available and the sky didn't fall.

      They would in any event have to make room for the action on the legislative calendar, itself a test of how important the government interest in across-the-board enforcement really is. And they would have to get the mild supermajority needed to get a law through the lawmaking process. Because the common-law exemption regime shifts the legal and political inertia to the religious objector's side, the objecting group would find it somewhat easier to defend itself. At the very least the objectors would have some modest assurance of a relatively thorough hearing, something they might not get if they bore the burden of lobbying for a statutory exemption.(46)

      This shifting of inertia to those who would undo what courts have done, a hallmark of common-law decision making, has its detractors. Many common-law decisions--for instance, broad expansions of tort law--have been criticized by some precisely because they create new legal rules without any requirement of supermajority popular support, and because they are hard to reverse, even if they are in theory reversible.

      Nonetheless, common-law creation of religious exemptions differs significantly from common-law creation of new torts or crimes. New tort or criminal liability generally creates greater government restrictions or mandates on individual action; new religious exemptions, like new common-law defenses more generally, usually create greater individual liberty from government restriction or compulsion.  This greater liberty may not always be good: The liberty could be a liberty to infringe private rights that should be protected. Still, so long as any errors that sufficiently concern the legislature are correctable, the likely errors in the direction of too much liberty from government restriction are generally--though not always--less harmful than the likely errors in the direction of too much government restraint.(47)

      Common-law decision making also breeds uncertainty, especially because common-law decisions are generally retroactive.  But the uncertainties caused by possible creation of religious exemptions should be less troubling than those caused by the possible creation of new crimes or torts. Putting prosecutors at risk of having a case dismissed under a newly minted religious exemption is not as bad as putting people at risk of retroactive criminal or civil liability. Subjecting religious objectors to the risk that their exemption claim would be denied is unfortunate; but such objectors are no worse off than if they had no chance of exemption at all, or than if they were subject to the equally uncertain constitutional exemption regime.

      2. Shifting the Decision Making from Executive and Local Decision Makers

      Under a common-law exemption regime, either the legislature or the courts must explicitly validate executive and local government conduct that interferes with religious practices. While the legislature can exclude statutes from the coverage of a state RFRA, executive agencies and local governments can't do the same for their actions.

      Consider for instance the Sikh requirement that Sikh men wear ceremonial daggers (kirpans). These kirpans may indeed be purely ceremonial--they are often blunted or sown into their sheaths, and might therefore not significantly threaten public safety.(48) Nonetheless, some prosecutors might feel prompted to literally enforce laws banning public carrying of knives, and school officials might, out of concern for discipline untempered by any desire for accommodation, expel any students who wear kirpans to school. A common-law exemption regime would let courts carve out an exemption for certain modes of carrying kirpans that judges believe to be harmless; and while the legislature could eliminate this exemption, individual executive officials could not.

      Likewise, consider zoning laws, which some religious exemption supporters say are often enforced without enough willingness to accommodate religious uses.(49) An ordinance generally limiting an area to purely residential uses might be applied by a zoning board to bar even fairly small weekly religious meetings in people's homes.(50) If one thinks zoning boards too often focus exclusively on the concerns of majority homeowners, the common-law exemption regime may provide a remedy: The regime would let courts carve out exemptions from zoning laws when the judges believe that such exemptions really wouldn't interfere with neighbors' rights or legitimate interests. And while the state legislature could repeal or modify these exemptions, it might take a broader view of the matter than would a local zoning board whose members may be too focused on their neighbors' concerns.

      The common-law exemption model should thus provide considerable, though not complete, comfort even to those who prefer a constitutional exemption model. The Sherbert-era courts' unwillingness to enforce strict scrutiny really strictly(51) suggests that, even under a constitutional exemption regime, courts would only rarely reject considered legislative judgments that a certain law had to be applied uniformly. Nonetheless, exemption supporters hoped RFRAs would at least let courts create exemptions for religious practices that legislatures never thought about, and would give religious objectors a tool to fight rigid-thinking petty bureaucrats, who themselves don't have much democratic legitimacy.(52) The common-law exemption model should in some measure satisfy this desire.

      On the other hand, this preemption of local and executive action could also be a basis for faulting the state RFRA model.  Local governments and executive decision makers are generally given considerable latitude because many people value local self-government, agency expertise, or the elected executive's personal accountability to the voters. State RFRAs would undercut these values.

      In these contexts judicial decision making under state RFRAs departs from traditional common-law decision making. Traditionally, local governments could, within the scope of their powers, supersede some common-law immunities: For instance, even if a court concludes that a certain kind of employment discrimination doesn't constitute common-law wrongful discharge, or a certain kind of land use isn't a common-law nuisance, a city ordinance could nonetheless ban such private action.(53) Similarly, even if a court finds that a certain kind of emission fits within a common-law defense to common-law nuisance liability, a state environmental agency may nonetheless enact regulations restricting the emission.

      Is this restraint on home rule and executive powers wise? It is not enough to respond that letting people do things mandated or motivated by their religion is important--as I discuss below in Part III, the question is who should decide when this important interest prevails over what government officials believe to be the important rights and interests of others.  I argue that when courts disagree with the considered judgment of the legislature on whether a religious exemption should be granted, the legislature should prevail; but who should prevail when courts disagree with the considered judgment of a city council, local voters, or executive officials? The answer must depend on how much one supports home rule or executive discretion, and the analysis is generally the same as it would be in any other debate about the proper scope of local or executive power.

      The standard arguments on these subjects are familiar. The argument for greater home rule is the value of self-government by the majority at the lowest possible level, which takes into account the possibility of differing conditions and differing local cultures throughout the state.(54) The argument against (most famously expressed as to state-federal relations in The Federalist No. 10) is that local governments are more apt to be captured by stable and mostly homogeneous local majorities and are thus particularly likely to undervalue the concerns of minorities.

      The arguments for executive discretion are the need for effective execution of the laws, the greater accountability of a single executive head compared to a plural legislature, and in some contexts greater agency expertise (for instance, as to environmental regulation). The arguments against are that most executive decisions are actually made by unelected bureaucrats, and are thus less democratically legitimate than decisions made by the legislature, and that bureaucrats hired specifically to enforce a particular body of law are more likely than generalist legislators to have a tunnel vision that precludes serious attention to other considerations, such as the interests of religious objectors.(55)

      I am genuinely unsure about how these arguments should ultimately be resolved, but the common-law exemption model provides more flexibility for resolving them than does the constitutional exemption model. To begin with, the legislature could write the state RFRA to say something like "This Act shall not apply to any local legislation, executive order, or agency action that explicitly excludes such application by reference to this Act."(56) This would leave final local and executive authority intact, but would still allow exemptions so long as there were no considered local or executive judgment to the contrary.

      Alternatively, the legislators can adopt a broad state RFRA and then wait and see. If they eventually find that the state RFRA excessively con-

      strains, say, local authority over zoning or executive authority over government employment, they would have several options. They could exempt these decisions from the state RFRA.  They could implement a special statutory exemption scheme for these decisions. Finally, they could provide, as suggested above, that these decisions be governed by the state RFRA only if there were no specific language in the ordinance or regulation stating that the state RFRA ought not apply. The common-law model thus lets the legislature provide whatever level of home rule or executive discretion it chooses.(57)

      3. Encouraging a Broader Scope of Exemptions

      The possibility of reconsideration by the legislature may actually make courts more willing to grant religious exemptions than they would have been under the constitutional exemption regime. A court may be reluctant to accept a close constitutional claim precisely because accepting it would permanently bind the legislature. Even a judge who thinks that granting a religious exemption from peyote law might not cause that much harm, and who thinks the legislature might not have considered this particular question when it banned peyote, may be hesitant to tie the legislators' hands by declaring that religious use of peyote is categorically protected.(58) But under a common-law exemption regime, a judge may be more willing to decide close cases in a claimant's favor, precisely because the decision isn't final.(59)

      Judges applying a common-law exemption regime may also be more willing to grant exemptions as experiments. Concluding that the Constitution requires something leaves judges little room to change their minds. Constitutional text isn't supposed to mean one thing this year and the opposite ten years later--judges are popularly expected to interpret the constitution, not decide based on their own changing attitudes. Changes of mind about doctrine are of course possible (Smith itself is an example), as are changes of mind about application of doctrine.  A court might at first conclude that a certain law fails strict scrutiny because some other alternative could be equally effective, but in a later case, with a better factual record, conclude that the alternative is not effective and that the law now passes strict scrutiny. But these changes are hard to justify to the public; realizing this, judges may be reluctant to grant an exemption in the face of uncertain results, because they know that they themselves will be strongly bound by this in the future.(60)

      Changes in common-law doctrines, on the other hand, are more defensible, because there the judges don't have to claim to be interpreting some unchanging text. A court might grant an exemption based on its tentative conclusion that the exemption--for instance, an exemption for religiously motivated peyote use or religiously motivated light labor by children(61)--won't do much harm to the government interest; but a few years later, after evaluating the exemption in action, the court might change its view.(62) Alternatively, the court may explicitly invite the legislature to review the results of the court-created experiment, and to repeal or modify the exemption if the legislature concludes that the experiment was unsuccessful.

      Of course, courts will always be somewhat reluctant to reverse recently established doctrine, but so long as it is easier for either courts or legislatures to reverse common-law doctrines than constitutional doctrines, courts under a common-law regime should be somewhat more willing to experiment in the face of uncertainty. And experimentation is particularly important in religious exemption cases, where so much turns on uncertain predictions about an exemption's effect, for instance about whether allowing a peyote exemption will indeed substantially interfere with enforcement of the general ban.(63)

      Judges may also be more willing to grant exemptions under the common-law model because the legislative authorization behind the common-law model makes exemptions easier to defend. Pro-exemption decisions under a constitutional exemption regime say, implicitly or explicitly, "We refuse to apply a democratically enacted statute"; but under a common-law exemption regime, they can be written as "Pursuant to the legislature's command, we apply the democratically enacted state RFRA to carve out a religious exemption." Judges who worry that their decisions might lead to a backlash against the judiciary, or who are philosophically uneasy about the countermajoritarian nature of judicial review, may be hesitant to grant constitutional exemptions but more inclined to grant common-law exemptions.(64)

      These arguments could also cut the other way. Some might argue that a common-law exemption framework would make judges too aggressive at granting new exemptions, and thus too quick to sacrifice the countervailing rights and interests of others, because the ultimate responsibility for the result would be in the legislature's hands.

      But it seems unlikely that judges would systematically overvalue the interests of religious claimants and undervalue the interests of others. True, religious claimants often present viscerally appealing cases, but so does the government, speaking on behalf of people whom the exemption would purportedly hurt. The constitutional exemption regime may have skewed such contests against objectors, because victories for objectors were, if erroneous, more harmful: Such victories could not be undone legislatively whereas victories for the state could be. A common-law exemption regime would eliminate this asymmetry, and thus the distorting incentive to decide against the claimant; and I doubt that it would introduce any new asymmetries in the other direction.

      4. The Difficulty of Predicting Specific Results

      In saying all this, I am not claiming to predict how any particular exemption claims, especially the most contested ones, would come out. I don't know whether and when courts will carve out religious exemptions from housing discrimination laws, zoning laws, or any of the other laws I mention here; nor am I certain about what the right results are in such cases. Likewise, I can't reliably estimate how much a common-law exemption regime will in the aggregate help religious objectors.

      This uncertainty exists partly because the strict scrutiny framework is, as I mentioned earlier, quite vague. More importantly, though, it exists because applying the underlying principle that "people should be free to do what their religions tell them to do, so long as they don't harm others"(65) requires answering an essentially contested question--what "harm to others" means. And because most hotly disputed controversies do involve a conflict between the desires of religious exemption claimants and the countervailing interests of others, the quality of any proposal cannot be measured simply by how much it helps religious objectors.

      As I explain throughout this Article, I think the common-law exemption model is a particularly good way of answering this question, but settling on this model (or, for that matter, the constitutional model or the statutory model) does not dictate exactly what answers will be reached.

      F. The Legitimacy of the Common-Law Model After Smith

      This analysis also suggests the weakness of one common criticism (which I myself have made in the past(66)) of jurisdiction-by-jurisdiction RFRAs: that they implement exemptions "blindly and en masse" and "shrug[] off the careful weighing which the accommodation of religious practices requires."(67) True, the legislatures are acting blindly, in the sense that they are enacting a statute that covers situations that they know they can't anticipate, and they are delegating the initial weighing decisions to courts rather than making the decisions themselves.  But the tradition of common-law decision making, including common-law decision making pursuant to statutory authorization, suggests that such delegation to courts is quite sound. If we assign judges considerable discretion in continuing to design tort law, define evidentiary privileges, evolve fair-use defenses to copyright claims, and the like, why not give them similar discretion for religious exemptions? Nothing in either the relationship of the three branches of government (federal or state(68)) or in the Establishment Clause stands in the way of this sort of delegation.(69)

      Likewise, this analysis undercuts the claim that applying the federal RFRA to federal law improperly or even unconstitutionally contradicts Smith's conclusions about the judicial role.(70)Smith's criticism of the constitutional exemption model says little about the propriety of a common-law exemption regime. As I explain in Part III, I substantially agree with Justice Scalia that "it is horrible to contemplate that federal judges will regularly balance against the importance of general laws the significance of religious practice"(71)--but only because these balancing decisions would bind the political branches.(72) In the common-law context, judges routinely balance claims of liberty against countervailing claims;(73) the very fact that their decisions may be superseded by contrary legislative balancing prevents such judicial balancing from being "horrible to contemplate."

      This does not dispose of all arguments against RFRA; for instance, one could argue that religious objections don't deserve special consideration,(74) or one could claim that courts are for some reason unusually bad at making decisions about religious exemptions. Still, the general concern about delegation to judges should be considerably weakened when judges are given only initial decision-making power, not final power.(75)

      G. The Problem of Preference for Religion

      Finally, one possible objection affects all three religious exemption models: Any regime of religious exemptions by definition prefers those whose actions are motivated by religion over those whose identical actions are motivated by equally deeply held secular beliefs. A purely religious exemption from the draft, for instance, would prefer religious conscientious objectors over secular ones.

      If one thinks exemptions are compelled by the Free Exercise Clause--which talks about free exercise of religion--then it's plausible that any such exemptions should be available only to those whose actions are religious.(76) But this rationale doesn't apply if one rejects the constitutional exemption model: The religious preference embodied in the statutory or statutorily authorized common-law exemptions would be the creation of legislators or judges, not a constitutional mandate. One would have to explain why, as a constitutional matter, such a preference for religion comports with the Establishment Clause, and why, as a policy matter, the preference is morally sound.

      I think the preference is both morally and (less certainly) constitutionally troublesome; the reasons for this have been well addressed elsewhere.(77) This objection may itself justify opposing state RFRAs as they are now written.

      Nonetheless, this objection applies not only to RFRA-type authorizations of common-law regimes, but also to statutory exemption regimes; and one way of defusing the objection--namely, making the exemption available to anyone who has a deeply held conscientious belief--is available in both situations.(78) Draft law has been interpreted to include such a conscientious objector exemption.(79)  Courts have generally interpreted Title VII's religious accommodation requirement the same way (though many such statements might be dictum).(80) A common-law religious exemption regime can similarly be broadened to protect all conscientious objectors, religious or not. Thus, though the concern about preference for religion may lead one to oppose a particular exemption proposal, it isn't an argument against common-law exemption regimes generally.

      II.Why State RFRAs Should Not Be Phrased in Strict Scrutiny Terms

      So far, I have praised state RFRAs in part because they give judges largely complete initial discretion to decide when an exemption should be granted. Both the strict scrutiny test's literal terms and the case law that has emerged under it in religious freedom cases are so vague that they don't meaningfully constrain a judge's range of options,(81) leaving almost unlimited room for judges' own moral and practical judgments about the propriety of granting an exemption. Strict scrutiny in religious freedom cases isn't much of a test--and that's good in a way, because the common-law decision making I praise is a system in which judges are free to come up with their own tests, subject to legislative revision.

      But strict scrutiny is not entirely content-free; some judges might feel to some extent constrained by it. And unfortunately, to the extent that strict scrutiny is a discretion-constraining test, it's the wrong test.

      A. The Sherbert-Era Regime Was Not a Pure Strict Scrutiny Regime

      To begin with, an across-the-board strict scrutiny requirement does not restore the Sherbert-era exemption regime. Rather, it ignores some sound decisions courts reached under that regime.

      As with the case law of many other constitutional provisions (such as the Free Speech Clause), the Sherbert-era constitutional exemption framework was a complex body of law, with not one but several tests. When the government was acting as prison administrator or as commander of military personnel, the religious exemption test--like the Free Speech Clause test--was close to the rational basis framework.(82) Lower courts adopted a similarly deferential test for probation conditions that incidentally interfered with religious practices.(83) When the government was acting as employer, some lower courts likewise adopted fairly (but not entirely) deferential tests borrowed from the Pickering(84) test applied in government employee free speech cases.(85) There was no agreed-on test for the government acting as educator in kindergarten through high school, but courts at least had the option of concluding that the free exercise test--like the free speech test--should be relatively deferential in these cases, too.(86)

      When the government was acting as sovereign, the test was usually strict scrutiny, but not always. For claimants requesting exemptions from generally applicable speech restrictions, the free exercise test was the same as the free speech test, which might differ from strict scrutiny. Content-neutral restrictions on the time, place, or manner of speech, for instance, are only subject to a form of intermediate scrutiny under the Free Speech Clause, and Heffron v. ISKCON(87) held that this same quasi-intermediate scrutiny was applicable to requests for religious exemptions from such restrictions.(88) Similarly, some cases suggested that zoning restrictions were subject to a lower standard of scrutiny.(89)

      A few other cases implemented a more religion-protective standard than strict scrutiny. Though courts have generally held that there is a compelling interest in banning each instance of race discrimination in employment, and that antidiscrimination laws are narrowly tailored to this interest, they have generally held that the Free Exercise Clause bars applying antidiscrimination laws to clergy.(90) As I've argued elsewhere, this rule is in fact a per se ban on interference with a church's reasons for choosing its clergy, not an application of strict scrutiny.(91) Likewise, courts have held that there is a compelling interest in banning each instance of race discrimination in places of public accommodation (setting aside the very small, selective ones);(92) but despite this, it's widely assumed that even big, largely open churches may choose their members free of the constraints of antidiscrimination law. This too must be because there is per se protection for such freedom of choice.

      If judges take seriously the statutory command of strict scrutiny, they may reach results that are much less sound than those mandated by the Sherbert-era cases. Say that a police officer, for religious reasons, refuses to guard abortion clinics,(93) or that a government mailroom worker refuses to deliver materials that he considers sacrilegious. Administrative efficiency is generally not considered a compelling interest under strict scrutiny,(94) which may be one reason that free speech cases have explicitly adopted a more deferential standard for government-as-employer regulations, instead of purporting to apply strict scrutiny.(95) So a court that takes the RFRA text seriously might feel obligated to let police officers choose their beats and let mailroom workers choose what they deliver, thus requiring the government to hire--at considerable expense and cost to morale--more officers and mailroom workers to take up the slack.

      This would be the wrong result. The government should be able to demand that employees do their jobs without having the job requirements pass strict scrutiny, just as the government need not face strict scrutiny when it restricts government employees' rude or disruptive speech to clients or coworkers. The government should be able to control potentially disruptive religious actions by K-12 public school students--such as wearing knives (even blunt ones) to school,(96) or using peyote at school--just as it may control potentially disruptive student speech.(97) The government should be able to restrict religious vigils in government office building lobbies, just as it can restrict political demonstrations in this sort of nonpublic forum. And the government, even when acting as sovereign, should be able to treat religiously motivated speech the same as other speech.

      Of course, courts applying a RFRA might nonetheless allow these restrictions. They may either ignore the text--for instance, applying Pickering scrutiny to government employee claims even though the statute calls for strict scrutiny--or stretch the strict scrutiny commanded by the text to reach what they see as the right result, for instance concluding that the government has a compelling interest in workplace efficiency.(98)  But this is far from certain, and, in any event, why set up a test that courts have to ignore in order to reach the right results?

      B. False Analogies to Other Strict Scrutiny Regimes

      So we see that in many contexts--both when the government acts in a special capacity, and when the government acts as sovereign--Sherbert-era courts used tests other than strict scrutiny. But even when the courts claimed to apply strict scrutiny, they didn't and couldn't apply strict scrutiny as it has become familiar in free speech law, race discrimination law, and other areas.

      Free speech law protects speech even when it causes serious harms.(99) Speeches praising draft resisters may in fact be likely to interfere with the war effort, but they aren't punishable unless they fall within the narrow category of incitement.(100) Racist and sexist speech is constitutionally protected even though it may create or reinforce racist and sexist attitudes and thus lead to illegal discrimination.(101) Distribution of sexually explicit material to adults increases the chances that such material will end up in minors' hands, but despite what the Court has called a "compelling government interest" in shielding minors from exposure to this material, distribution of the speech to adults is protected.(102)

      This is not so with religiously motivated acts. Strict scrutiny, the Court has held, doesn't prevent the government from drafting religious objectors, nor does it prevent the government from applying antidiscrimination law to religious objectors.(103) Speech might harm the war effort or the fight against discrimination more than these claimed religious exemptions would. Nonetheless, the speech is protected but the action--even when religiously motivated--may be banned.

      This difference cannot be explained by any difference in the compellingness of the government interests, which are pretty much the same in the free speech cases and the religious exemption cases. Nor can it be justified on the grounds that there are less restrictive but equally effective alternatives to banning the speech, but no such alternatives to banning the action. One often-suggested alternative, counterspeech, may alleviate some of the harms caused by the speech, but it won't alleviate all, or even close to all.(104) Likewise, the other alternative to restricting the speech--banning the acts of draft evasion or discrimination themselves--will prevent some such conduct, but the bans (especially the discrimination ban, violations of which are notoriously hard to prove) will be substantially underenforced.

      Suppressing the speech, in addition to counterspeaking and banning the action, would prevent some people from even wanting to resist the draft or to discriminate, and would thus probably be considerably more effective than any alternatives would be. Nonetheless, the speech suppression is forbidden, even though the suppression of religiously motivated acts is permitted.

      So long as "strict scrutiny" is used to describe the "strict in theory, pretty much fatal in fact" scrutiny applicable to content-based speech restrictions and to racial and religious classifications,(105) it shouldn't be used to describe the necessarily weaker test applicable to religious exemption claims.(106) When people are asking for freedom not just to speak, or to be treated equally without regard to race, but to act, the law must often intrude on that freedom.

      Using the same term to describe different tests poses three practical problems. First, courts might import the strongly rights-protective traditional strict scrutiny doctrine into religious exemption cases, or at least be influenced by the symbolism of "strict scrutiny" as it has evolved in other areas. This generally hasn't proven to be much of a problem; exemption cases usually seem to apply a considerably looser standard than strict scrutiny as it has been traditionally understood. Nonetheless, the risk of confusion exists.

      Second, courts might export the watered-down religious exemption strict scrutiny into other cases, or, less directly, weaken strict scrutiny in these other cases by diluting its formerly forceful symbolism through its feeble-in-fact application in religious freedom cases.(107) Thus, some courts have held that bans on bigoted speech should pass strict scrutiny under the Free Speech Clause because bans on discriminatory acts pass strict scrutiny under religious freedom principles.(108) Some others have suggested that "the broad public interest in maintaining a sound [and administratively workable] tax system"(109)--recognized as compelling for purposes of denying religious exemptions(110)--might justify restrictions on tax protester speech, though they fortunately did not have to reach the question whether such speech restrictions indeed passed strict scrutiny.(111)

      In my view, these courts erred: Speech restrictions should be less permissible than conduct restrictions, even when the latter incidentally affect religiously motivated conduct. But such an error is to be expected when the tests for speech and conduct are described using the same term.

      Third, promising strict scrutiny, with its historical connotation of extreme skepticism concerning the government action, but delivering something considerably weaker diminishes courts' credibility and risks further alienating religious objectors. It's upsetting enough for people to see their deeply felt claims rejected; it's worse when the legal system claims to greatly respect such claims but in reality readily sweeps them aside. Better to acknowledge that people's religious conduct must necessarily be less protected than their speech or their right to be free from race discrimination than to claim the protection is identical but then in practice protect one right much less strongly than the others.(112)

      C. The Impossibility of a Single Formula and the Superiority of a Pure Common-Law Approach

      As my discussion in Parts III.F.3-4 will show, I don't insist that legislatures delegate all decisions to the courts. I think the legislators should set forth discretion-constraining rules when they think the courts have reached (or fear the courts will reach) incorrect results. But, for the reasons given above, strict scrutiny is just the wrong rule--the Court was wrong to adopt it, and legislatures would be wrong to restore it.(113)

      And no single formula, or even small set of formulas, can be the right rule for all exemption cases. As I argue throughout, any religious exemption regime must reconcile religious objectors' claims with the countervailing private rights and interests of others.(114)  Tort law, property law, and contract law have evolved from attempts at this very sort of reconciliation between liberty claims and others' rights, and these complex bodies of law can't possibly be united under one verbal formula, unless the formula is so broad as to be meaningless.

      Likewise with religious exemptions. Courts can make the initial moral and practical judgments about when a countervailing right or interest should yield to a religious objector's claim, but these judgments can't be reduced to a single test. Better to acknowledge this and explicitly delegate to courts common-law-making authority so that they can generate different tests for different situations.

      Some state constitutions have been interpreted to fairly significantly constrain delegations of legislative power,(115) but my proposal should be permissible even in those states. The principle underlying the nondelegation doctrine is that each branch of government--the legislative, the executive, and the judiciary--may only exercise powers that are within its proper sphere.(116) But as I've argued above in Parts I.B-C, what is delegated here to courts is only the power that Anglo-American courts have exercised for centuries--the power to make common-law rules that are subject to revision by the legislature.

      The absence of a test would, of course, create unpredictability, and thus a chilling effect. For instance, a religiously-motivated peyote user who doesn't know whether a court will carve out a religious exemption to a peyote ban under a vague common-law exemption regime may be deterred from practicing his religious ritual. But the presence of a test such as strict scrutiny wouldn't help. In both situations, there is no clear rule governing the use of peyote until a court renders a decision addressing the specific practice. Strict scrutiny is the worst of both worlds--as I argue above in Parts II.A-B, it's potentially constraining enough to lead to the wrong results, but not constraining enough to lead to predictable results.

      D. Proposal for a Pure Common-Law State RFRA

      I want to suggest, therefore, that the common-law exemption model should be implemented through a statute that explicitly delegates to courts the discretion--unconstrained by strict scrutiny or other such ultimately misleading formulas--to initially decide which exemption claims win and which lose. Such a statute might look something like this (with alternative choices and optional material in brackets):

      WHEREAS the legislature has a duty both to protect people's freedom to practice their religions and to protect people from harms created by others' behavior; and

      WHEREAS some laws inadvertently incidentally restrict religious practice that is not harmful, but other laws must incidentally restrict religious practice in order to prevent harm;

      Be it enacted that:

      (1) The government shall not substantially burden(117) a person's exercise of [religion/conscience] unless

      (a) it demonstrates that such a burden is justified, or

      (b) the legislature has specifically exempted the government action from the coverage of this statute[, or]

      [(c) the local legislative body that authorized the action has specifically stated, by ordinance or comparable official enactment, that the action is to be exempt from the coverage of this statute, or]

      [(d) the executive official or executive branch agency that authorized the action has specifically stated, by executive order, regulation, or comparable official act, that the action is to be exempt from the coverage of this statute].

      (2) Whether a burden is justified shall be decided under the principles of the traditional law of [religious/conscientious] exemptions as they may be developed, expanded, contracted, or modified by the courts of [the jurisdiction] in the light of reason and experience.

      (3) "The traditional law of [religious/conscientious] exemptions" shall refer to the body of law related to religious [or conscientious] exemptions developed by courts and legislatures of the United States and of the several states, including:

      (a) Legal principles developed by courts pursuant to statutes or constitutional provisions requiring religious [or conscientious] exemptions from generally applicable laws.

      (b) Decisions by a legislature or a court to exempt religious [or conscientious] objectors from the scope of a particular statute, common-law rule, or other government action.

      [(4) "Conscience" shall include religion and religious practice and belief, as well as practice and belief that flows from moral or ethical beliefs about what is right and wrong that are held with the strength of traditional religious convictions.]

      A brief explanation:

      1. The statute begins by setting forth the general rule: Substantial burdens on religious exercise are forbidden, unless either a court or the legislature explicitly concludes that they are justified; depending on the legislature's views on the questions discussed in Part I.E.2, a local government body or executive agency can also be given the same power. The statement that legislative action can justify a burden makes clear that the legislature is expected to play a role here--that legislative revisions of the judicially created rules are part of the normal process of crafting religious exemption law, not repeals of any broad religious freedom principle. This recognition of the legislature's authority fits with the theory of the common-law regime, and as I explain in Part III, such legislative authority is actually a good thing.

      2. The statute goes on to direct courts to develop the law of religious exemptions, building on existing decisions but with the explicit authorization to "expand[], contract[], or modif[y]" the principles. The phrase "principles . . . as they may be [developed] by the courts in the light of reason and experience" is borrowed from the Federal Rules of Evidence, which say that testimonial privileges "shall be governed by the principles of the common law as they may be interpreted by the courts of the United States in the light of reason and experience."(118)

      3. The statute authorizes courts to look to religious exemption law developed both by courts and by legislatures in all jurisdictions. Though the common law is generally seen as being built on prior court decisions, looking to principles developed by legislatures makes sense. A considered judgment by a legislature that a particular law (for instance, alcohol prohibition, a draft law, or a peyote ban) can adequately serve its goals despite a religious exemption is a useful data point.(119)

      4. The statute may protect nonreligious conscientious objectors as well as religious ones. The language defining conscience is drawn from Welsh v. United States,(120) which read the conscientious objector exemption to the draft as including nonreligious objectors.(121)

      III.The Difficulties with the Constitutional Exemption Regime

      Even if all I say above is true, why isn't a constitutional exemption regime even better? Much has been written about whether Smith is correct, but I believe the comparison with the common-law exemption regime, coupled with a focus on the common law as the traditional means for defining when liberty must be constrained to prevent harm to others, provides a relatively novel approach to this question. This approach leads me to conclude that Smith rightly rejected the constitutional exemption model and that the common-law model is the best solution, not just a fourth-best alternative to federal constitutional exemptions, state constitutional exemptions (discussed in Part III.F.1), and a broad federal "son-of-RFRA" (discussed in Part III.F.2).

      A. Three Kinds of Free Exercise Clause Claims

      1. Several Specific Prohibitions on Government Action

      We can divide claims for Free Exercise Clause protection--both those that may win under current doctrine and those that will likely lose--into three categories. The first category contains several relatively focused rights. People have a right not to be discriminated against by the government because of their religion, or because of the religiosity of their conduct.(122) Everyone has a right to sincerely express his religious views and even to solicit money based on them without being punished (for instance, by general laws punishing solicitation of money based on misrepresentations) for the supposed falsity of these claims.(123) Ministers generally may not be sued for clergy malpractice or for breach of fiduciary duty to those whom they counsel, even though similar secular counselors may be subject to this sort of liability.(124) Churches have great autonomy in hiring and firing clergy, without constraint from antidiscrimination laws, labor laws, or (possibly) the torts of negligent hiring and retention.(125)

      Churches likewise have autonomy in choosing their members, without restriction by public accommodations laws or creative tort actions for "wrongful excommunication."(126) People are free to spread their religious doctrine even if the doctrine leads those who accept it to act badly (though the Free Speech Clause also bans such government restrictions). And courts may in the future recognize some other similarly narrow Free Exercise Clause principles.

      These protections share four related attributes. First, they focus on particular kinds of behavior (clergy hiring, membership decisions, or spread of religious doctrine) or on particular kinds of government action (determination of the truth or falsity of religious claims or discrimination against religious practice), rather than on all religiously compelled or motivated behavior.

      Second, each of the protections is justified by something more than the religious motivation for the claimant's actions. Church autonomy, for instance, is justified by the importance of unhampered clergy selection and membership selection to the churches' ability to define and propagate their teachings.(127) The bar on discrimination against religious practice is justified by equality concerns; the bars on judicial determination of the truth or falsity of religious claims and on judicial definition of a minister's fiduciary duty to parishioners is justified by the perceived impropriety of getting secular courts involved in religious disputes.(128)

      Third, the protections apply even when the conduct they protect causes certain kinds of harm to others. Though the legal system recognizes employment discrimination as a harm to the applicant or employee, religious institutions nonetheless have a First Amendment right to inflict such harm when hiring or firing clergy.(129) Though the legal system recognizes procuring money based on misrepresentations (even sincere ones) as a harm to the person who pays the money, the First Amendment nonetheless bans the government from preventing this harm when the harm is inflicted through sincere--even if possibly false--religious claims.(130) And fourth, this willingness to protect these rights despite the harms they cause explains why they remain protected quite strongly, rather than under the generally feeble scrutiny provided by the Sherbert-era constitutional exemption regime.(131)

      This set of specific protections is thus similar to the protections offered by other constitutional provisions. For instance, the strongest protection of freedom of speech (strict scrutiny or something similar) applies to a specific set of laws--those that punish speech because of its communicative impact. Within that zone, free speech includes the freedom to inflict certain harms through the communicative impact of speech that would be punishable if inflicted in other ways.(132) I may with impunity urge a political boycott of your business, even if it violates your right, secured by tort law, to be free from intentional interference with business relations.(133) I may express the nastiest opinions about a public figure, even if my speech violates his right to be free from intentionally inflicted emotional distress.(134) I may collaborate with my competitors to lobby for anticompetitive legislation, even when the legal system generally recognizes similar anticompetitive activity as inflicting a harm on others.(135) I may record a song that praises cop-killers, even if it actually leads others to kill.(136)

      Likewise for other provisions. The Compulsory Process Clause secures a narrow right, but one that entitles the bearer to inflict a certain sort of harm on others by interfering in a particular way with another's liberty and privacy. Certainly any private attempt to drag someone against his will into a room, make him talk on pain of imprisonment, and require him to turn over his property for inspection would normally be seen as a gross deprivation of private liberty and property rights. Nonetheless, the Compulsory Process Clause authorizes such infliction of harm.(137)

      Similarly, the privilege against self-incrimination lets a person deny to litigants (and not just to the government) the testimony to which they would normally be entitled.(138) State and, more controversially, federal rights to keep and bear arms let people possess and even carry guns, which many think creates serious improper externalities for others.(139) As with the specific religious freedom protections I mentioned above, these rights provide--within a limited zone--rather strong protection against certain laws, even when this protection shields conduct that causes certain kinds of harm to others.

      The courts haven't arrived at any general theory uniting the focused areas of protection that I outline above and explaining when the Free Exercise Clause protects behavior even though it harms others, just as they haven't endorsed any grand theory about exactly when speech that harms others should be protected by the Free Speech Clause. Such a theory (if sound) would be tremendously valuable, in part because it would explain when it would be proper to create new areas of Free Exercise Clause protection; but to my knowledge none has been developed.

      Still, I am not alone in concluding that these focused areas of protection should be treated specially, and differently from any broader regime of religious exemptions (such as those described in the next two subsections). Some of these focused areas existed before Sherbert,(140) they continue to exist after Smith,(141) and even during the Sherbert era they evolved separately from the constitutional exemption regime.(142) For now, I assert only that certain special concerns--for instance, the need to avoid decision of religious questions by civil courts or the need to preserve the autonomy of religious institutions--do indeed justify protecting certain kinds of religious behavior that harms others.

      2. A General Regime of Exemptions for Conduct Even When It Harms Others

      A second possible kind of free exercise claim is a broad right to do whatever your religion motivates (or, in some formulations of the theory, compels(143)) you to do, simply because of your religious motivation. Taken literally, such a claim is clearly too broad. Surely no court would immunize, for instance, murder or rape simply because the perpetrator acted out of religious conviction.

      Nonetheless, one might argue, while the right to life or to bodily integrity is more important than religious freedom, the right to religious freedom is more important than some other rights.  Thus, the argument might go, religious conduct might be protected even when it harms others in certain ways, perhaps even when it constitutes, for instance, assault, breach of contract, copyright infringement, libel, trespass on private property, or negligent infliction of physical injury.(144) One formulation I have sometimes heard is that religious conduct should be constitutionally protected even when it harms others so long as the harm to others is outweighed by the importance of the conduct to the religious observer.

      But such a claim is normatively unappealing. My relationship with my God may be important to me, but how can it by itself--setting aside any special, narrower justifications, such as those discussed above in Part III.A.1--be a constitutionally sufficient justification for my harming you, even slightly?  From your perspective and the legal system's perspective (even if not from my own), my God is my God, not yours, and the Constitution doesn't give those acting in His name sovereignty over your legally recognized rights and interests.

      The trouble with such a broad religious freedom claim is not that private contract, property, and tort law rights somehow outweigh religious freedom rights--how could we justify such a conclusion?(145) Even if one somehow determines that contract, property, and reputation rights are as important as religious freedom rights, I see no way to explain why they are more important, which would be required to show such an outweighing. Rather, the reason why a general exemption regime can't override these rights must be that any religious freedom right that's solely grounded in the religious motivation for one's actions simply can't extend to actions that impair others' rights or impose improper externalities on others. Whether your countervailing right is a right to life, to bodily integrity, or to something perhaps somewhat less important, such as property or reputation, if it is indeed a right then the religiosity of my motivation can't justify violating it.

      Constitutional protection for inflicting harm on others simply because of the nature of one's motivation (as opposed to other, more focused reasons such as those discussed in Part III.A.1) is virtually unprecedented. The only constitutional principle that even approaches such a regime is the expressive conduct doctrine, which in some formulations supposedly does protect otherwise regulable conduct, even conduct that harms others in certain ways unrelated to the conduct's communicative impact, when it's engaged in for expressive reasons.(146)  But even this doctrine is probably considerably narrower than its broadest statements: The Court has at times suggested that it might apply not to all conduct, but only to conduct that is commonly used as expression.(147) And in any event, the protection this doctrine offers is a fairly weak form of intermediate scrutiny, which has proven quite deferential to government interests, and not the strict scrutiny promised by the Sherbert-era religious exemption cases.(148) If expressive conduct is constitutionally allowed to inflict harm on others, these harms can at most be modest indeed; the prevention of even aesthetic harms, such as the "visual blight" created by billboards, has often been found sufficient to justify restrictions.(149) If this very weak scrutiny were transplanted to religious exemption claims, it would deny exemptions in virtually all the hotly disputed cases.

      The Court is correctly unprepared to protect a wide range of conduct that harms others in noncommunicative ways simply because of the actor's expressive motivation. The communicative reasons for your actions--a sit-in on my property, discrimination against me in hiring or housing, or whatever else--generally can't justify the noncommunicative harms that the actions may inflict on me. Likewise, the religious reasons for your actions can't, by themselves, justify harms to others.

      3. A General Regime of Exemptions for Conduct That Does Not Harm Others

      The most common claim for a constitutional exemption regime is the third kind--a broad right to do what your religion motivates you to do, simply because of your religious motivation, but only so long as it doesn't harm others. Jefferson's defense of religious freedom, for instance, was justified by the argument that someone's "say[ing] there are twenty gods, or no God . . . neither picks my pocket nor breaks my leg."(150) Madison wrote that religion should be "immun[e] . . . from civil jurisdiction, in every case where it does not trespass on private rights or the public peace."(151) Similarly, Michael McConnell, one of the leading authors on free exercise law, argues that we should be "free to practice our religions so long as we do not injure others."(152)

      I intentionally cast the discussion here in terms of harm to other people, not harm to government interests. The traditional focus on the supposed clash between individual rights and countervailing government interests--both under the Sherbert religious freedom regime and under substantive due process--tends to obscure the fact that behind virtually all government interests (at least those asserted when the government is acting as sovereign(153)) lies a claimed right or interest of some individual.(154) The government, after all, neither bleeds nor eats nor desires; its power is enlisted to defend the felt rights and needs of people.(155)

      Any claim of countervailing private right can, of course, be reformulated as a compelling government interest in protecting such a right, and the case law in fact often does this;(156) my focus on the concerns of others is thus entirely compatible with the compelling interest test. Nonetheless, I think that looking at the matter as being primarily about the countervailing rights and interests of individuals better illuminates the inquiry.

      And when exemption proposals are limited to conduct that doesn't harm others--that doesn't invade their rights or impose improper externalities on them--they seem quite appealing. After all, if my religiously motivated actions genuinely harm no one, why shouldn't I be allowed to engage in them?

      B. The Private Rights and Interests of Others

      1. The Debates About What Constitutes the Private Rights and Interests of Others

      The difficulty is that what constitutes "invading the rights of others" and "imposing improper externalities on others" are hotly contested questions. Do people have a right not to be economically pressured into working at exploitative wages?(157) Do people have a right to be free from various kinds of discrimination in private commercial transactions?(158) Do children have a right--unwaivable by them because of their immaturity--to be educated until age sixteen?(159) Do litigants have a right to require a reluctant witness to testify?(160)

      Do householders have a right to be free from the intrusion and annoyance created by certain kinds of land use by their neighbors?(161) Do businesspeople have the right to be free from various forms of unfair competition--unfair because it's monopolistic or potentially involves misleading consumers, or because it takes advantage of certain cost savings?(162) Does a person's failure to report for the draft impose unjustified externalities on other draft-eligible people?(163) Do creditors have a right to repayment by a bankrupt debtor that precludes the debtor from giving away his property as he sees fit?(164) The answers to these questions can't be deduced through pure logic or traditional constitutional analysis.(165)

      Similarly, there are lively debates about when the likelihood of violation of others' rights by some justifies a restraint on the liberty of all. Most agree that some such prophylactic rules are proper: Consider bans on drunk driving, which are justified by the possibility that a drunk driver might cause injury to another, even though most instances of drunk driving do not actually cause such injury. May the law likewise ban other driving practices--such as driving a horse and buggy without an orange reflector--that foreseeably pose a risk to others, but that need not necessarily cause injury?(166) What about possession and use of alcohol or drugs, which lead some people to act criminally or tortiously, but which others use quite safely?(167) What about possession or carrying of certain weapons, which some people use to violate others' rights but which others use for innocent, even laudable, purposes?(168)

      Neither is there a consensus about whether certain externalities actually flow from certain kinds of behavior, or which of those externalities are improper. Is an employer imposing an improper externality on taxpayers when, by paying employees a lower wage, it "casts a direct burden for [the employees'] support upon the community," on the theory that "[w]hat these workers lose in wages the taxpayers are called upon to pay"?(169) Is an employment agency imposing an improper externality when, through its actions, it increases unemployment, and do such agencies' actions in fact tend to increase unemployment?(170) Is a motorcyclist acting improperly when he rides his motorcycle without a helmet, thus increasing his chances of suffering an injury for which the taxpayers will have to pay?(171) Does drug use really lead to more street crime than would the less restrictive alternative of decriminalization?

      Thus, the debate is not really about whether people should be able to do what they please so long as they don't harm others. The reason for most restrictions on conduct is precisely that people think the conduct does harm others.(172) In my experience, for instance, when people have defended the religious landlady's claimed right to discriminate against unmarried couples in housing rentals, they have rarely claimed that, just because the landlady's beliefs are religious, she is entitled to deny to tenants their private right to equal treatment (the second kind of claim I outlined above). Nor have they generally admitted that the landlady's conduct imposes some real harm on the tenants while arguing that the harm is outweighed by the benefit to the landlady's religious practice. After all, when conduct genuinely causes harm to another--as for instance when the conduct is a trespass on the other's property, a breach of contract, or an infringement of the other's copyright--such a harm, even if relatively modest, can't be justified on the grounds that the person inflicting the harm might get a tremendous spiritual benefit from acting this way.

      Rather, supporters of the landlady's right usually make the third kind of claim--that the landlady's religiously motivated decision should be immune because it doesn't really harm the tenants, since the tenants don't really have a true private right to equal treatment. The debate is thus about who ultimately defines what constitutes "infringement of the private rights of others" or "imposition of improper externalities on others"--the legislature or the courts.

      2. The Connection to the Early 1900s Substantive Due Process Debates

      The power to define, as a final, constitutional matter, the limits of another's lawful private rights or to decide what constitutes an unjustified externality is the very power that proved so troubling in the early 1900s substantive due process cases. Those cases, after all, didn't purport to reject the government's right to protect people against injuries inflicted by others, but they did claim for courts the power to finally define what was an injury and what was not. A minimum-wage law was unconstitutional because the employer wasn't hurting the employee by paying him too little.(173)  A ban on employers' requiring employees to agree not to join a union was unconstitutional because it didn't violate anyone's rights for an employer to discriminate based on a person's union membership.(174) A law allowing certain kinds of picketing was unconstitutional because union members had no right, and could be given no right, to engage in "palpable wrongs" such as "libelous attacks," "abusive epithets," and "unlawful annoyance and . . . hurtful nuisance."(175) A maximum-hours law was unconstitutional because there was no legally cognizable harm, the Court assumed, in a bakery owner's bargaining with employees for a work week longer than sixty hours.(176)

      On the other hand, a requirement that an employer issues a discharged employee a letter "setting forth the nature and character of his service [rendered] and its duration, and truly stating what cause, if any, led him to quit such service" was upheld because it helped preserve the employee's reputation, which "is an essential part of his personal rights--of his right of personal security."(177) A zoning law prohibiting apartment buildings in a certain residential area was upheld because the buildings, with their tendency to create various externalities for neighbors--"street accidents," "noise," "interfer[ence] . . . with the free circulation of air and monopoliz[ation of] the rays of the sun"--"come very near to being nuisances."(178) An alcohol ban was upheld because "the general use of intoxicating drinks" leads (albeit indirectly) to "idleness, disorder, pauperism, and crime."(179) Some claims of externalities and countervailing private rights won and some lost, with little to explain the distinction beyond the moral and practical senses of the judges, and perhaps the degree to which the claims fit with traditional common-law rules.

      The Court ultimately repudiated this substantive due process jurisprudence, because it concluded that the legislature may redefine private rights by, for instance, deciding that people had a right to "protection from unscrupulous and overreaching employers," or to be free from being "exploit[ed] . . . at wages so low as to be insufficient to meet the bare cost of living."(180) Likewise, the legislature may try to prevent more indirect externalities, such as the "burden [imposed] upon the community" for the support of workers who are being paid less than "[t]he bare cost of living."(181)  The legislature wasn't limited to protecting private rights akin to those traditionally secured by the common law, such as reputation or freedom from nuisance. It could also recognize new private rights, even controversial ones, and protect them even though this meant limiting the liberty of people to infringe those new rights.

      A constitutional religious exemption regime would either return courts to identifying their own favored view of what really constitutes others' private rights and interests, and enforcing it contrary to the political branches' views, or would require them to rule that certain violations of others' private rights and interests must be permitted just because the violator acted for

      religious reasons. Imagine, for instance, a would-be employer who comes to court and sincerely says:

      I have a religious obligation to hire this employee for $3 per hour to work in my factory. I interpret the Bible as commanding all Christian employers to hire the needy and to pay them what the employer honestly thinks the employee's labor is worth, no more and no less. "[T]he labourer is worthy of his hire," but no more.(182) These unemployed people whom I wish to hire are needy, and I believe their labor is worth $3 per hour.

      This is an unorthodox view but in our religiously pluralist and nonhierarchical culture, some people probably sincerely hold it, and such individual sincere belief is enough to trigger a religious freedom claim, even if the belief isn't widely accepted by a broader religious group.(183) Many devoutly religious people believe that virtually all their moral views flow in some measure from religious command: Plenty of people believe both that they are morally obligated to help the poor and that it's morally proper for workers to be paid what their labor is worth, and I suspect that at least some believe it morally proper to pay no more and no less than this amount.

      If the employer demands an exemption from minimum-wage laws on the grounds that these laws infringe his general liberty under substantive due process, he will lose.(184) During the early 1900s, he would have won--a court would have concluded that he had the right to hire people at whatever wage they agreed to, because this didn't violate anyone's private rights.(185) But this jurisprudence was rejected because courts recognized that what constitutes "another's rights" is not fixed by the Constitution. Libertarians like me might believe there's no right to be paid "a decent wage," and no right to be free of competition from those who are paid a lower wage. But others disagree and believe that paying someone $3 per hour does violate the employee's rights and the rights of competing employees.(186) Under modern substantive due process law,(187) "the rights of others" are for the legislature, not the courts, to ultimately define.

      A constitutional religious exemption regime, though, would give our employer a facially viable claim. The minimum-wage law bars the employer from doing what he sincerely (though idiosyncratically) believes his religion commands--certainly a substantial burden. If one takes Sherbert seriously, a court would thus have to decide whether the minimum-wage law is the least restrictive means of serving a compelling government interest, a standard even stricter than that applied in the early 1900s substantive due process cases. And deciding whether there is a compelling government interest here depends in large part on one's vision of workers' rights.

      As in Adkins v. Children's Hospital,(188) the court would have to make this moral decision itself, perhaps superseding the judgment reached by the legislature. A court using the "exemptions are mandated unless they harm others" model might tell the legislature:

      Yes, we recognize that you think workers have a private right to be paid a decent wage, akin to the private right to be free from trespasses, theft, defamation, nuisance, and the like--but you're wrong. We have compared these claimed rights, and find that the supposed right to be paid a decent wage is just not in the same league as the others. The Constitution doesn't let people trespass, steal, or defame for religious reasons, but it does let people acting for religious reasons violate this right that you think is so important. And this ranking of private rights is embedded in the Constitution, so you're powerless to change it.(189)

      Alternatively, a court taking the view that exemptions may be mandated even for conduct that harms others, just because of the religiosity of the objectors' motivation, might tell the legislature:

      Even if you're correct that people have a private right to be paid a decent wage, it's permissible for people to violate this private right--though not the right to be free from trespass or defamation or breaches of contract--just because their motivation for violating the right is religious.

      But why should the religious objector's beliefs authorize such a violation of the rights of people who don't share those beliefs?

      Of course, the legislature might not have thought of the minimum wage as securing a private right. It might have just seen it as an attempt at economic engineering, and a narrow court-created religious exemption might not substantially interfere with such an attempt. If this is so, then an exemption might well be reasonable--at the very least it wouldn't, by hypothesis, let employers violate employees' legislatively recognized private rights. A common-law exemption regime, though, could secure such an exemption as well as a constitutional exemption regime would. If the court thinks the law is just an economic engineering measure that doesn't protect private rights or prevent improper externalities, it could carve out a religious exemption. Then, if the legislature agrees with the court, or doesn't think the matter is important enough to reconsider, the exemption would stand.

      But if the legislature does think there is a countervailing private right involved, then under a common-law exemption regime--unlike under the constitutional exemption regime--its judgment could ultimately prevail. The common-law exemption model is thus like the traditional common-law system, where courts recognized a general regime of unregulated wages, but the legislature could step in to revise this regime as it thought necessary. The constitutional exemption model is like the regime of Adkins v. Children's Hospital, where the original common-law judgment was constitutionalized, and thus made unchangeable even in the face of changing attitudes about a private right to a minimum wage.(190)

      Likewise, say a landlady believes it would be sinful for her to rent an apartment to an unmarried couple.(191) Again, if she makes a substantive due process challenge to a law banning marital-status discrimination in housing, she will lose (assuming she falls outside the narrow zone of constitutional protection provided by the right of intimate association).(192) Sure, she might claim she isn't violating anyone's rights by her decision, and again libertarians may agree. No one, they will argue, has a right to rent another's property over the owner's objections.(193) But the legislature has taken a different view--unlike the libertarians, it believes everyone has a right to be treated equally, even in the opportunities to use another's property.(194) And it is not for courts to decide whether the legislature's or the libertarians' moral vision is correct.

      Here too, a constitutional religious exemption regime would require courts to decide the questions that substantive due process jurisprudence suggests they ought not decide.  The law commands the landlady to do something she thinks is religiously prohibited, so the landlady would have to be exempted unless the law passes strict scrutiny.(195) Whether there is a compelling government interest in prohibiting discrimination rests--again, just as with prohibiting trespass, copyright infringement, or assault--on whether a person has the right to be free of such discrimination.(196)

      A court applying a constitutional exemption regime of the "so long as it doesn't interfere with the rights of others" variety will have to resolve this difficult moral question itself, perhaps trumping the legislative decision. "While there is a private right to be free from race or sex discrimination," the court may say, as some courts have indeed said,(197) "there's really no private right to be free from marital status discrimination, no matter what the legislature might conclude." A court applying a constitutional exemption regime of the "even if it does interfere with the rights of others" stripe may have to go further: "Even if there is a private right to be free from marital status discrimination, a landlady who wants to violate this right for religious reasons must be free to violate it, simply because she's religiously motivated." Again, the first of these seems unjustified, and the second seems normatively unattractive.(198)

      As in the minimum-wage example, a court applying a common-law exemption regime may actually give the landlady what she wants.  The principle of the common-law regime is that there is nothing wrong with the court's saying "While there is a private right to be free from race or sex discrimination, there really is no private right to be free from marital status discrimination"--so long as the court doesn't add "no matter what the legislature might say." In fact, this distinction between different kinds of discrimination may be quite reasonable, and is akin to the distinctions courts have traditionally drawn in exercising the moral and practical judgment involved in creating the common law.

      If the legislature agrees with the court that there is no private right to be free from marital-status discrimination, then the court decision will stand. This might happen if, for instance, that particular discrimination ban was aimed more at relieving a broader economic problem of occasional shortages of housing for particular groups (a problem that a narrow religious exemption probably wouldn't exacerbate) than at securing perceived private rights to be free from individual instances of discrimination. But if the legislature takes a different view of private rights, its view could prevail, just as the legislature's view of countervailing private rights prevails over the courts' when the question is the scope of one's right to act free of tort liability.

      Finally, consider the right-to-assisted-suicide cases. In Washington v. Glucksberg,(199) the Court concluded that this claimed right does not fall within the narrow "right of privacy," in which judicial, rather than legislative, judgment is paramount.(200) Given this conclusion, it was easy to defend bans on assisted suicide against substantive due process claims. "Whom do I hurt by assisting a suicide?", Dr. Kevorkian asks. Well, the government might say, allowing assisted suicide by some will create an improper externality for others: It will lead to their being pressured--by greedy or distraught family members, or by greedy or uncaring doctors--into choosing suicide.(201) And because the court can't tell whether such a prediction would be right or wrong, the matter is left to the legislature.(202)

      But say a doctor claims a religious obligation to assist the suicide of someone whose life has lost what the doctor considers to be the proper dignity; or say a patient claims a religious obligation to end his life with a doctor's help. (Some have in fact asserted such obligations.(203)) Under a constitutional religious exemption regime, just as under the substantive due process regime rejected in Glucksberg, the court would have to do what Glucksberg correctly concluded courts shouldn't do.

      One constitutional exemption model would require that the court decide whether, as a moral matter, the supposed externality--pressure on others to commit suicide even when they wouldn't otherwise choose it--is improper, and whether, as a practical matter, the supposed externality is indeed likely to materialize (a hotly contested question). Under the other constitutional exemption model, the court would have to decide whether, even if letting people commit assisted suicide does impose an improper externality on others, they should nonetheless be allowed to inflict this externality simply because their motivations are religious. Neither of these approaches seems sound, and both seem inconsistent with the reasoning of Glucksberg and more generally with the rejection of the early 1900s substantive due process claims.(204)

      Again, a common-law exemption regime would actually let a court carve out an exemption here, if the court concludes that such an exemption wouldn't harm other patients--for instance, if the court disagrees with the prediction that recognizing a right to assisted suicide would actually cause improper pressure on patients. But if the legislature disagrees, either immediately or after looking at the results of several years of this experiment, it can reverse this judgment and prevent what it concludes is harm to others.

      C. Implications for the Three Kinds of Free Exercise Claims

      The above argument thus counsels against a general constitutional exemption regime for religiously motivated action. Such action, when it hurts others, can't be protected just because of its religiosity. And while the notion that "you may do as you please--or as your God pleases--so long as you don't hurt others" is appealing, the ultimate definition of "hurt others" should be made by legislatures (as in the common-law model), not by courts (as in the constitutional model).

      This is what the Court has generally and correctly held in the substantive due process context, outside the narrow right of privacy, which does protect certain private choices even when they harm others--spouses, future children, taxpayers--in certain ways. Courts can't simply constitutionalize the common-law regime of private rights and protectible interests. Under our democratic constitutional structure, the legislature must be able to redefine the proper scope of these countervailing rights and interests, rather than just being stuck with the judges' views on the matter.(205) And concluding that someone has a certain right or interest necessarily limits others' liberty to "do what they please so long as they don't harm others."

      Free Exercise Clause protection should thus cover the first category of free exercise claims: a set of fairly specific protections for particular kinds of conduct and prohibitions on particular sorts of government action (just as substantive due process now essentially covers a set of fairly specific protections for particular kinds of conduct). These protections and prohibitions recognize that in certain cases, religious freedom must include the right to act in certain ways even though this inflicts certain harms on others. Put together, they provide a significant level of constitutional protection, protection that is as strong as that provided by other focused but important constitutional provisions such as the Equal Protection Clause, the Sixth Amendment, and others.  Rejecting a constitutional exemption regime would certainly not render the Free Exercise Clause a nullity. And this set of discrete areas of protection might yet be expanded, if courts find reasons why other kinds of conduct involve special considerations that justify protecting the conduct even when it inflicts certain kinds of harm on others.(206)

      But I do claim that each of these areas of protection, whether currently established or recognized in the future, must have a special justification beyond simply the religiosity of the actor's motivations. The second kind of free exercise claim, which would let people inflict harm simply because their motivations are religious, and the third kind, which would put courts in the untenable position of deciding that something is harmless even when the legislature has concluded that it is harmful, should be rejected.

      D. Objections to the Substantive Due Process Analogy

      Can one nonetheless somehow distinguish a broad "you may do as your religion moves (or compels) you so long as you don't hurt others" regime from the properly rejected "you may do what you want so long as you don't hurt others" framework? My examples in Part II.B.2 were aimed to suggest the contrary, but let me confront some specific counterarguments.

      1. Text and Original Meaning

      If the constitutional text (or the text's original meaning) did indeed secure a right to engage in religiously motivated conduct so long as this doesn't hurt others, I would agree that courts would have to engage in the analysis that I condemn above.(207) Likewise, if the clause's text or original meaning secured a general right to engage in religiously motivated conduct even when it hurts others, I would have to accept this despite my moral unease with such a right.

      But there is considerable debate about whether the text and the original meaning do indeed point to such a mandate; in my view, the debate is either in equipoise or leans slightly against a constitutional exemption regime.(208) And in the absence of a definitive textual or originalist answer, courts must consider other matters, including how the religious exemption regime would fit with the constitutional structure.

      For the reasons given above, I think such a constitutional exemption regime is incompatible with the primarily democratic decision-making structure established by our Constitution. And the arguments I give here can be a tiebreaker in resolving an interpretive debate that otherwise seems to me to be close to a tie.

      2. The Relative Rarity of Religious Exemption Claims

      Sincerely religious objections will be rarer than objections under a broader substantive due process regime, which could be made by all people regardless of the source of their beliefs. But religious objections won't be terribly uncommon, especially because some people who have a pervasively religious worldview base most of their important actions on religious conviction.(209)

      More importantly, if there really is a right not to be paid "exploitative" wages, or not to be discriminated against on certain grounds, or not to be pressured into assisted suicide, then the law must be able to protect this right whenever it's jeopardized, even if it's jeopardized by only a relatively few religiously motivated employers, landlords, doctors, or patients. There is broad agreement on this, I think, in cases that involve traditional common-law harms: Even if, for instance, only a few people will trespass on others' property for religious reasons, or defame them for religious reasons, or breach contracts with them for religious reasons, the victims of such behavior remain entitled to protection from it.  The same must apply to newly legislatively recognized rights, such as the right to be free from discrimination or from unfair wages. Even if demands for religious exemption from these laws will be rare, each such demand will still be an attempt to inflict what the legislature has concluded is a harm.

      3. The Possibility of Legislative Error

      Legislatures will often be imperfect guardians of "true" religious liberty, and imperfect definers of what "really" constitutes the rights and interests of others; legislatures will often even be imperfect representatives of the popular will. I have no desire to romanticize the often ugly, unprincipled, and improperly influenced legislative process. Courts, if given the power, would sometimes do better, both at defining harms and at gauging public desires.

      But the same is true for other aspects of individual liberty, such as the liberty to earn a living in your chosen occupation, to ingest whatever substances you please, or to use your property as you see fit.(210) Most agree that such liberties exist, and many might condemn certain legislative decisions as infringing these liberties. Despite this, because the exact scope of these claims of liberty is profoundly contested, the best way to read our democratic constitutional framework is as leaving it to the political process to ultimately decide what is harmless and thus a proper exercise of liberty, and what is harmful and thus impermissible--or so the rejection of the early 1900s substantive due process cases suggests.  The democratic process is imperfect, but it is a more legitimate place for resting such choices than are the courts, which of course are imperfect too.(211)

      4. The Risk of Suppression of Minority Interests

      Is a constitutional exemption regime nonetheless required in order to prevent unfair majority impositions on the interests of powerless religious minorities? There are three possible bases for such a claim.

      First, one could argue that a constitutional exemption regime is needed to prevent seemingly neutral actions that are actually caused by simple hostility towards the exempted religious group because of who they are--for instance, a legislature's disliking the Rastafarians or the Amish and wanting to do these religious groups harm just because of the groups' beliefs. This sort of intentional discrimination, though, is already actionable under the Equal Protection Clause, the Establishment Clause, and the post-Smith Free Exercise Clause.(212)

      True, such intentional discrimination may often be hard to prove, and thus may often go unremedied, which is regrettable. Nonetheless, if our concern is preventing intentional discrimination, I see no reason why the Court should adopt a stronger prophylactic rule to prevent religious discrimination than it has for race or sex discrimination, where it requires evidence of actual discriminatory intent despite the difficulty of getting such evidence.(213)

      Second, one could argue that under a statutory exemption regime, legislatures might not grant exemptions to religious minorities because they are too small to lobby effectively for space on the legislative agenda. But the common-law exemption regime avoids this problem, because it lets courts grant exemptions without any need for the objectors to get the legislature's attention.

      Third, one might argue that even under a common-law regime the legislature would be particularly likely to repeal certain religious exemptions because of hostility--perhaps mistaken hostility--to the challenged practice. Legislators who are used to alcohol consumption may be more bothered by marijuana or peyote consumption.  Legislators who are used to churches' discriminating based on religion in hiring may balk at landladies' discriminating based on marital status among renters.

      But hostility to a particular practice, which is usually justified by the belief (right or wrong) that the practice causes harm, is a presumptively proper basis for legislative action. It's the legislature's job to identify the conduct it believes harms others, and to take steps to restrict or prohibit such conduct. This is why legislatures ban housing discrimination, the paying of low wages, assisted suicide, and other practices.

      Legislatures may indeed err in their determinations, as I think they have in many respects, especially if the practice is unfamiliar. Still, so long as there is no objective standard by which courts can determine whether or not the legislature erred, judges ought not (for reasons I've discussed above) replace legislative definitions of harms with their own.  And minority religious groups may not claim a right to do what the majority, acting through the legislative process, has concluded is harmful, any more than minority ideological groups may claim such a right.

      Thus, while laws that facially discriminate against certain religious groups because of their religious identity are properly viewed with deep constitutional suspicion,(214) laws that prohibit certain behavior and therefore rest on a discrimination between what the majority thinks is harmful and what the majority thinks is harmless are the essence of democratic decision making. The laws may end up being burdensome to the minority--whether a religious minority, an ideological minority, or an occupational minority--

      that takes a different view of the laws' propriety and that wants to act in ways the majority sees as harmful. But this sort of burden on minorities is inevitable and generally proper.

      Finally, one can argue that religious objectors especially need judicial solicitude because, in the words of Carolene Products,(215) they are more likely than other objectors to be subject to "prejudice" because they are "discrete and insular minorities."(216) I'm not sure, though, that the factual predicate for this argument is accurate. Most religious groups in the United States today do not lead particularly discrete and insular lifestyles, and while engaging in practices that society thinks harmful is likely to set one apart and lead to condemnation (whether or not "prejudiced"), that's true whether or not the reason for the practice is religious.

      But beyond this, even Carolene Products speaks only of "statutes directed at particular religious, or national, or racial minorities,"(217) not at statutes which apply to all but incidentally burden some groups more than others; such a focus makes sense when one is concerned about the political-process failure on which the theory of "discrete and insular minorities" rests. Most general laws--as opposed to laws that are in fact directed at certain religious groups--do not burden only a discrete or insular minority. Laws banning marital-status discrimination in housing burden landlords generally; zoning laws burden millions of homeowners; marijuana laws burden many more secular users than Rastafarians; Sunday closing laws burden non-Sabbatarian consumers and many profit-conscious non-Sabbatarian businesspeople as well as Sabbatarians. Some general laws may indeed have an effect that's more limited to religious objectors, but as a class, most general laws against which religious exemption claims are raised burden many people who don't belong to insular religious minorities.

      This relative generality of burden does not mean the laws are sound, and it does not change the fact that the religious objectors may subjectively feel the burden much more acutely than do the other objectors. But it does suggest that we shouldn't be particularly worried in this context about the "discrete[ness] and insular[ity of] minorities [being] a special condition, which tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching judicial inquiry."(218)

      5. The Possible Special Importance of Religiously Motivated Liberty Claims

      Another possible distinction between courts' second-guessing legislative judgments of harm as to general liberty claims and courts' doing the same as to religious exemption claims rests on a perceived difference in the strength of the underlying rights.  General liberty claims, the argument would go, simply aren't that important, and the impropriety of courts' deciding as a constitutional matter that some conduct "isn't really harmful" therefore justifies judicial abstention in substantive due process cases. But religious freedom claims are so important--because they involve conduct that is more fundamental to people's identities or involve more painful imposition of government authority--that this importance outweighs courts' reluctance to determine what's really harmful and what's not.

      I don't think, however, that this distinction can work. To begin with, it underestimates the burdens that general conduct restrictions can place on people.  Minimum-wage laws or professional-licensing laws may make it much harder for some people to earn a living. Antidiscrimination laws may interfere with people's choices about whom to associate with. Bans on debt adjusting(219) or employment agencies(220) may prevent people from following their chosen occupations, something that to many people is a very important part of their identities.  Assisted suicide bans may deny people the ability to make the most important choices about their existence. Of course, many people won't feel some of these burdens particularly severely, but the same may be true of many not very devout religious people, who might be bothered but not deeply affected by the bans on certain practices in which their religion moves them to engage.

      But beyond this, I think judicial overriding of legislative decisions about what constitutes harm is more than just hard; I think it's basically improper, and remains improper even where legislative judgments about harm cause genuinely serious burdens on people.

      Thus, in the examples discussed in Part III.B.2, it seems to me that a court may not permissibly decide--even if people's deeply held religious beliefs are involved--that the Constitution views paying a subminimum wage or discriminating in housing as "really" harmless. If a legislature concludes that there is a private right to have equal access to rental housing or not to be paid "exploitative" wages, I don't see a basis for the court to conclude the contrary. The common-law model would let a court carve out an exemption for a religious observer if it's not clear that the legislature really thinks there is a countervailing private right involved. But I believe a court may not go beyond that and decide that the Constitution dictates, at least for religious exemption purposes (even if not for other purposes), that the legislatively recognized private right isn't a "real" right.

      What about the possibility that the presence of deeply held religious beliefs, while not allowing the court to deny the legitimacy of the countervailing private right, would allow the court to conclude that the claimant's beliefs outweigh the private right? This would be another way of saying that the Free Exercise Clause does give the claimant a right to harm others--here, by violating this particular countervailing private right--simply because he is religiously motivated to do so. And, for the reasons given in Part III.A.2 and throughout Part III.C, I believe that this is not a sound conclusion for courts to draw.

      6. Questioning the Rejection of the Early 1900s Substantive Due Process Cases

      One could accept the similarity between the early 1900s Court's attempt to fashion general liberty principles and a constitutional exemption regime's commitment to religious liberty principles, but argue that it cuts the other way--that it proves Lochner v. New York,(221)Coppage v. Kansas,(222) and the other cases were right, at least in principle (though perhaps not in application), and that courts should be in the business of rejecting "wrong" claims of harm made by legislatures.

      I confess that I have some sympathy for this claim, as I do for the whole Lochner era. The notion that wise, impartial, nonpartisan courts can and should protect our liberty against spurious though popular claims of harm is quite appealing. But after reflecting on the cases I discuss above, and seeing the difficulty--maybe the impossibility--of coming up with any general, principled theory of what constitutes harm (breach of contract? defamation? copyright infringement? monopolization? interference with another's patent monopoly? alienation of affections? annoying uses of residential property? insider trading?), I've reluctantly abandoned that particular hope.

      In any event, this particular debate has already been amply covered elsewhere; and, rightly or wrongly, the anti-Lochner forces seem ahead enough that I feel comfortable resting, for purposes of this Article, on their conclusions.

      7. Does This Analysis Jeopardize Constitutional Rights Protection Generally?

      Would my analysis, if taken seriously, jeopardize the protection of all constitutional rights, such as free speech, the right of privacy, and others? As I've argued in Parts III.A.1 and III.A.2, I believe free speech and other constitutional rights do deserve strong protection, just as I believe the focused Free Exercise Clause rights I discuss in Part III.A.1 deserve strong protection. Such protection, though, must be founded on a recognition that the constitutionally protected behavior often does harm others, and is protected despite the harm it causes.(223)

      The main question for each substantive constitutional right is thus: When does the right immunize even harmful behavior? What makes the broad formulation of religious freedom different from the other rights is that it potentially applies to a vast range of conduct that can harm others in many ways, and it applies solely on the grounds that the actor is religiously motivated. I think this is not a claim that the Constitution ought to recognize, and I don't believe the text or the original meaning endorses such recognition.(224) The Constitution does give people the right to compel others' testimony, even when that causes harm, or to communicate facts and ideas, even when that causes harm via the communicative impact of the speech, or to discriminate in hiring clergy, even if employment discrimination is seen as a harm. But it doesn't give people the general right to inflict harm simply because of their religious motivations.

      I recognize that the answer to the question "when does the right immunize even harmful behavior?" will be controversial. The debates about the right to privacy, and especially the rights to abortion and assisted suicide, are classic examples of such controversy.(225) My main claim, though, is that one can't avoid asking this question by simply saying "Oh, the Constitution protects only harmless behavior, and courts can figure out, as a final matter, what really is harmful and what isn't." That, it seems to me, would be repeating the core error made in the early substantive due process cases.

      E. Can a Constitutional Exemption Model Nonetheless Be Made to Work?

      Might the above overstate the difficulties inherent in determining what constitutes others' countervailing private rights or interests?  Could there be some principled way for courts to make such a determination?  I address several candidates in turn.

      1. Looking to Exceptions

      Some commentators suggest that courts can determine when a law is genuinely needed to prevent a certain harm by inquiring whether the law covers all of that sort of harm. If it doesn't--if it, for instance, exempts certain kinds of secular conduct--then, under this theory, the legislature can't really say the conduct is per se harmful.

      Thus, the argument goes, the legislature can't claim a compelling interest in preventing marital-status discrimination in housing when it "explicitly sanctions such discrimination" by itself discriminating against unmarried couples in intestate succession, workers' compensation death benefits, evidentiary privileges, and family accident insurance.(226) Likewise, the legislature can't claim a compelling interest in requiring humane slaughter when it "exempts any person slaughtering and selling 'not more than 20 head of cattle nor more than 35 head of hogs per week.'"(227)

      This approach is appealing because it seems to suggest a way for courts to determine which activities are really harmful by looking at the legislature's own pronouncements, rather than substituting their own moral or practical judgment for the legislature's. So long as the legislature has a "pattern of exemptions [for secular interests]" or "explicitly sanctions [the harm]" in supposedly similar contexts, this "shows that the legislature's goals do not require universal application"(228)--that the purported harm to others that the legislature is trying to prevent isn't really a harm.

      But virtually all laws, including those widely seen as aiming at quite serious harms, contain many secular exceptions. As Doug Laycock, a proponent of the "look to the exceptions" approach, himself points out, "[t]he federal employment discrimination laws do not apply to Congress or to employers with fewer than fifteen employees,"(229) and actually there are many more exceptions than that.(230)  Trespass law is full of exceptions--consider adverse possession, necessity, law enforcement, and so on.(231) The duty to testify when subpoenaed is subject to many exceptions in the form of testimonial privileges.(232) Statutory-rape laws often except acts committed by someone who is close enough in age to the minor, or acts committed by the minor's spouse.(233) Breach of contract law has exceptions galore. The Copyright Act contains one operative section followed by fifteen sections of exceptions.(234) Even the bans on intentional homicide have exceptions--execution of a lawful sentence, killing in war, police killing of a dangerous fleeing felon, killing in self-defense or in defense of another, and disconnecting life-sustaining equipment at a patient's request.(235)

      I take it, though, that the presence of these exceptions doesn't justify a general license for religious objectors to discriminate based on race or sex in employment, to trespass, to refuse to testify, to commit statutory rape, to breach contracts, to infringe copyrights, or to kill. The presence of these exceptions, coupled with the absence of corresponding religious exemptions, does show "that the legislature values the exempted secular activities more highly"(236) than the religious activities (among many other activities). But this may be perfectly proper.

      To begin with, some of these exceptions may reflect a legislative judgment that the private right protected by the law can't be defined in a single phrase. A legislature might not, for instance, think that everyone has a right to life as such; rather, it might think that the right is inherently bounded by concerns about self-defense and the like. Likewise, a legislature might not think that people have a right to be free from marital-status discrimination generally, but only from marital-status discrimination in housing and some other fields.  A legislature might not think that people have a right to be free from all sex discrimination in employment, but only from sex discrimination where sex is not a bona fide occupational qualification.(237) As Fred Schauer has pointed out, what looks like the exception and what looks like the rule often turns on largely irrelevant factors, such as whether existing legal language happens to contain a single phrase to describe some morally significant concept.(238)

      Other exceptions reflect legislative judgments that in some situations the private right is particularly weak, a countervailing interest is particularly strong, a ban is unlikely to be effectively enforced, or there is an adequate nonlegal check on the harmful conduct.  We see this, for instance, in employment discrimination law, which excepts small companies (presumably based on freedom of association concerns); which until recently exempted overseas discrimination (presumably based on a perceived need to follow local laws or customs, or on notions that certain rights weaken outside U.S. borders); and which has been interpreted to allow certain discrimination against men and whites (based on a concern that barring such discrimination would have harmful side effects).(239)

      The legal system thus often defines our private rights using somewhat complex, seemingly exception-filled rules, but this doesn't make them any less worthy as private rights, and doesn't mean that others should be able to violate these rights simply because they feel religiously motivated or compelled to do so. As I argued above, I may feel that God obligates me to do something that harms you, and that God's commands are superior to the state's; but if you don't feel the same way, because you follow a different vision of God (or no vision at all), there is no justification for requiring, as a constitutional matter, that the state in effect let me impose the burden of my religious obligations on you, solely because I feel them to be my religious obligations. And the magnitude of other limitations that are imposed on you and your rights doesn't diminish this basic point.

      The legislature might well "value [the various] exempted secular activities more highly"(240) than it values other activities, including religious ones--such as race discrimination by a businessperson who feels a religious duty not to mix with people of a certain race(241)--but the examples suggest that this isn't inherently wrong.  One could, of course, argue that courts should allow this only when there are good enough reasons for the secular exemptions; but such an inquiry would get us back to courts' substituting their moral and practical judgments about what constitutes "true" harm to others for those of the legislature, as they determine which secular exemptions are indeed based on good enough reasons. And, as I have argued above, this is an enterprise in which courts should not engage.

      2. Looking to the Treatment of Majority Religious Groups

      What about using as a benchmark the way the law treats, or would treat, the religious interests of majority religious groups? The theory here (expounded in a characteristically fine article by Michael McConnell) is that "claims of minority religions should receive the same consideration under the Free Exercise Clause that the claims of mainstream religions receive in the political process."(242)

      Thus, the argument goes, in the early 1800s, when Catholics were a far smaller group in the United States than they are today, a New York trial court was right to create a constitutionally rooted priest-penitent privilege--"the exemption was required in order to maintain neutrality between the Protestant majority and the Catholic minority,"(243) not because Protestants already had a clergy privilege (they didn't), but because the legal system protected practices that were equally central to the religious beliefs of the Protestant majority.  The court was correct to "pose[] and answer[] the hypothetical question: Is the government's interest in compelling testimony so strong that it would interfere with a Protestant sacrament in order to achieve it?"(244) Again we see the promise of an approach that can substitute objective inquiry for a court's subjective moral judgment,(245) and with egalitarian appeal to boot.

      This egalitarianism, though, is radical to the point of impossibility. McConnell asks, in criticizing United States v. Lee's(246) denial of exemptions from social security taxes, "Who can doubt that there would be exceptions to social security (or, more likely, no social security) if mainstream Christians were forbidden by their religion to participate?"(247) Well, yes, if most Americans were forbidden by their religion from participating in a social program, the program probably wouldn't exist. But that is true of all laws. Who can doubt that there would be exceptions to antidiscrimination law (or, more likely, no antidiscrimination law) if mainstream Christians were required by their religion not to associate with people of other races?(248) Who can doubt that there would be a religious exemption from trespass law if mainstream Christians were required by their religion to trespass onto private property to, for instance, view apparitions of the Virgin Mary? There would even be exceptions to murder law if the dominant religious group viewed certain kinds of murder--whether human sacrifice, the burning of widows, infanticide, or killing of the old--as a sacrament.(249)

      McConnell argues that there is a limiting principle to the proposal: "A country . . . [probably would not] allow trespass or interference with the private rights of others. A government interest is sufficient if it is so important that it is not conceivable that the government would waive it even if the religious needs of the majority so required."(250) But what constitutes "the private rights of others" is itself defined by the majority's laws.  True, a country wouldn't allow (at least on a more than marginal scale) interference with what most of its voters think are the private rights of others. But if the religious needs of the majority required the toleration of certain trespasses or even of certain killings, then the majority would probably see the private rights of others as inherently limited by its religious needs, just as we see the right to life or the right to freedom from trespass as limited by other considerations.(251) And then of course the government would happily waive any contrary interests it might have.

      In a democracy, the majority's judgments about what constitutes private rights get enacted into law, whether or not the judgments are religious. Had a different group been in the majority, the law would of course have been different, but we can't displace the majority's judgment simply because of this. For better or worse, we live in a country where paying taxes (social security or otherwise) is seen as a moral duty, and race discrimination and trespass are seen as moral wrongs. This does flow from the fact that the majority holds no moral beliefs--or religious beliefs--to the contrary, but this fact can't justify exemptions for others who hold different moral or religious beliefs.

      3. Looking to the Constitutional Pedigree of the Claimed Countervailing Right

      Might we profitably distinguish private rights that are analogous to constitutionally secured rights from those that aren't? Thus, the argument might go, the government may deny religious exemptions from race and sex discrimination laws because the ban on such private discrimination is grounded in the Fourteenth Amendment. No religious exemptions need be given from laws that protect life, liberty, or property, because these rights are themselves constitutionally rooted. The federal government may enforce tax laws without regard to religious exemptions because the Constitution explicitly grants Congress the power to levy taxes,(252) and may uniformly enforce copyright law, which is specifically authorized by the Copyright and Patent Clause. On the other hand, courts must grant exemptions from bans on private sexual-orientation discrimination or marital-status discrimination, because the Constitution generally tolerates such discrimination when done by the government.(253) Again, this approach offers the hope that courts can decide what constitutes "the rights of others" by reference to some source other than their own moral views.(254)

      But which constitutional rights are secured against government action by the Constitution has little to do with which statutory or common-law rights deserve security against private action, including religiously motivated action. Protections against government action were included in the Bill of Rights because of specific fears that the government would violate certain rights, and the perceived need for a constraint beyond what the tort system or the political process would normally provide. Some well-acknowledged private rights, for instance the freedom from negligently inflicted bodily injury (including death), the freedom from trespass on open land, the right not to have one's property taken even if the taker offers compensation, the right not to be defamed, the freedom from malicious prosecution, and the general freedom to do what one pleases without fear of violent retribution aren't secured against government action.(255) On the other hand, many of the rights that are secured against the government--for instance, the right not to be punished because of one's speech--have no private-law analogs.(256) What the government may not do to me and what you may not do me are two different things.

      Looking to the scope of Congressional powers as a way to identify private rights or compelling government interests is even less profitable. Anything Congress does must be an exercise of an enumerated power, whether the tax power, the copyright power, the commerce power, or some other power; in every religious freedom claim raised against the federal government, the government interest has a constitutional pedigree.  Conversely, some of the most important private rights and corresponding government interests--for instance, the interests in preventing death, injury, and theft--aren't generally sources of federal power at all.(257)  The federal powers were created not because the interests they defend were seen as more important than other interests, but because the interests were seen as particularly suited to federal legislation. Courts thus get no useful clues about the existence of a private right or the strength of a government interest from its mention in a grant of congressional power.

      4. Looking to the Facts

      Would it be proper for courts to get out of the business of evaluating the supposed existence of a countervailing private right--or the supposed compellingness of a government interest--and instead focus only on whether the right or the interest is in fact implicated? Thus, for instance, McConnell points out that "[e]vidence in the Smith case showed that ingestion of peyote by members of the Native American Church is not dangerous and does not lead to drug problems or substance abuse."(258)  Stephen Pepper likewise claims that "[t]he children not going to school in Yoder were not being harmed in any concrete, measurable way."(259) This approach would be a retreat from strict scrutiny, but it would theoretically offer some constitutional protection to religious objectors without allowing courts to trump legislative value judgments.

      But this puts unwarranted faith in courts' ability to answer questions on which there is no consensus even among top experts. Would some drug decriminalization increase or decrease drug-related harms?(260) Which educational techniques help children and which hurt them?  Is the carrying of concealed weapons a menace to society or a deterrent to crime?(261) If leading social scientists and policymakers disagree on these matters, it seems unlikely that a trial before a lay judge or jury can produce a reliable answer. And if we again look at the early 1900s substantive due process cases, we see the problems created by courts' asserting the power to pass final judgment on such hotly contested factual questions.(262)

      Furthermore, at least some of the supposedly objective practical judgments about whether the harm actually takes place ultimately prove to be subjective moral judgments about what constitutes harm. For instance, does exempting Amish children from compulsory high school education harm them in a concrete, measurable way? If one thinks children have a right to an education that prepares them to earn a living in the social group in which they'll probably end up, then the Amish children aren't harmed by the exemption.(263) But if one thinks children have a right to an education that maximizes their options in life and helps them decide for themselves whether they should follow the plans their social group has for them, then at least some Amish children probably are harmed.(264)

      Like many harms, including for instance the harms of trespass or defamation, the harm may not be particularly concrete, but it is nonetheless real (assuming one makes the moral judgment that children do indeed have this right). And though we might not be able to measure this harm--which is largely invisible so long as the Amish parents get the exemption, since it is a harm to Amish children who might have chosen a different life had they been educationally prepared(265)--we may still have strong reason to believe that it exists. So the matter primarily turns not on the practical judgment about facts, but on the moral judgment about the nature of the countervailing right.

      In some situations, the legislature may have actually made no factual judgment about whether a particular kind of religious conduct would cause harm. It may have, for instance, concluded that peyote is generally harmful without considering whether it might be considerably less harmful when used in a religious ceremony; or it may have concluded that children generally have a right to an education, without considering whether the situation might be different for children in a separatist religious community. As I discussed above, this may justify instituting a common-law exemption regime and letting courts carve out exemptions subject to the possibility of legislative correction. But when the legislature responds by specifically rejecting exemptions, this response represents a moral and factual judgment about the scope of countervailing rights and interests that courts ought not supersede.

      Sometimes, of course, the evidence truly is one-sidedly in the claimant's favor. In most such cases, the common-law exemption regime would reach the right result, because the legislature will interpret the evidence the same (by hypothesis, correct) way that the courts do, and won't repeal any judicially created exemptions. But in some cases, the legislature might consciously reach a result that you and I might agree is just factually wrong.

      Nonetheless, while it's appealing to let courts right this wrong, this appeal comes from the assumptions that you and I are smart enough to see the facts, that the courts are equally smart, and that the legislature is too dumb or too prejudiced or too concerned about dumb or prejudiced voters. Sometimes these assumptions will be true, but sometimes they will be false: Sometimes it will be the judges and those who read and write law review articles who are wrong, and the legislators and voters who are right. Given this, it seems to me more democratically legitimate to leave the ultimate call, even on the practical questions, to the political process's determination of harm and not the judicial process's.

      5. Developing New Narrow, Specially Justified Rights

      Finally, some commentators, instead of suggesting a single broad principle that purports to cover all possible claims of countervailing private rights and government interests, suggest a set of narrower principles that would categorically protect certain conduct or rule out certain claims of harm. Thus, for instance, McConnell suggests that the government should not be able to "protect[] the members of the religious community" from harms that flow "from the consequences of their religious choices."(266) Pepper argues that we should "rule out administrative inconvenience and administrative costs as interests sufficient to justify impingement on religious conduct unless they are substantial in proportion to the overall administrative costs of the governmental program or conduct at issue."(267)

      I am not entirely persuaded by these particular proposals. For instance, consider Tony & Susan Alamo Foundation v. Secretary of Labor,(268) which rejected a constitutional exemption from minimum-wage law. McConnell argues that the Court was mistaken because "if members of the Alamo religious movement are inspired to work for the glory of God for long hours at no pay, their neighbors are not injured and the government has no legitimate power to intervene."(269) But as the Alamo decision pointed out, minimum-wage laws are partly premised on the notion that subminimum wages "affect many more people than those workers directly at issue in this case and would be likely to exert a general downward pressure on wages in competing businesses."(270) If one takes the view that this downward pressure is an improper externality (I do not, but American political institutions generally do) then it is not accurate to say that "the putative injury [of the Foundation's actions] is internal to the religious community."(271)

      I am also not sure exactly where administrative costs should stand in the analysis. When religious conduct imposes a financial burden on the government, it imposes a burden on you and me: "The government" isn't the ultimate subsidizer here--taxpayers are. I'm not sure that my religious beliefs can justify my imposing such costs, however indirect, on other taxpayers.

      Nonetheless, I have considerably more sympathy for these sorts of focused proposals, especially McConnell's, than I do for the other, broader ones. As I mention in Part III.A.1, the Free Exercise Clause authorizes some specific protections for religious practices despite the harm that such practices may inflict on others, so long as there is some extra justification for such protection beyond just the objector's religious motivation. Perhaps such a justification is provided by the limited effect of a particular activity on those who haven't voluntarily joined the religious community coupled with the importance of intracommunity activities to the preservation of religious communities as an alternate source of social power and moral authority, or by the fact that a particular exemption is only a matter of slight expense to any particular taxpayer.

      Whether or not, though, these particular proposals are correct, reading the Free Exercise Clause as creating such focused rights to act in potentially harmful ways strikes me as much sounder than reading it as creating a broad right to inflict harms just because one is acting for religious reasons, or as giving courts the power to determine what "really" constitutes a harm and what does not.

      F. Implications

      1. Rejecting Federal and State Constitutional Exemption Schemes

      If the above analysis is right, then religious exemption regimes should continue to be implemented as state statutes. The Court ought not overrule Smith, state courts ought not interpret their own state religious freedom provisions as creating a constitutionally compelled exemption regime (unless the provision's text or original meaning clearly mandates exemptions),(272) and state legislatures and voters should reject proposed state constitutional amendments that would require exemptions.(273)

      State constitutional mandates aren't as bad as federal ones, because they may generally be changed with a smaller supermajority; but even such state mandates are, by design, significant restraints on the political process. As I've argued above, if legislatures deliberately conclude that a certain form of conduct harms the private rights or interests of others, this conclusion should be enforced. Just as the federal courts recognized this in retreating from their early 1900s substantive due process jurisprudence--a retreat that, as I've argued above, also suggests the impropriety of a federal constitutional religious exemption model--so state courts have recognized it in generally retreating from the analog of this early 1900s case law under state constitutions.(274)

      The proper mechanism for implementing any state-level exemption regime is thus normal legislation, such as a state RFRA, with its significant but mild supermajority requirement for legislative override of court-granted exemptions. States are certainly free as a matter of federalism to impose greater restraints on their legislatures (subject to the Establishment Clause concerns mentioned in Part I.G), but I think it would be a mistake for them to do so.

      2. Rejecting Federal Common-Law Exemption Schemes

      Federal "son-of-RFRA" bills that mandate religious exemptions from large categories of state laws (for instance, from state laws governing behavior that touches interstate commerce(275)) would create a federalized common-law exemption framework, as RFRA itself did. Such a rule would be better than the constitutional exemption model, but it would still be considerably less flexible and leave less room for experimentation than the state common-law approach. Judicial value judgments embodied in grants of religious exemptions could only be modified by congressional action, and it's generally much harder to get Congress's attention than a state legislature's. Moreover, courts would have to make their decisions without the information that state-by-state experimentation can provide about whether an exemption from a particular law does in fact harm those whom the law was meant to protect.

      As I acknowledged in Part I.E.2, decentralized decision making isn't always best; there are, for instance, sound arguments against letting city- and county-level legislative bodies override judicially created exemptions. Nonetheless, I do think that localism and experimentation are important values, and that leaving religious exemption decisions entirely to the distant and unwieldy federal legislative process would interfere too much with these values. (276)

      Because of this, I believe an exemption framework in which the final calls are generally left to state legislatures, subject (as I'll mention below) to federal preemption in particular areas where Congress specifically concludes preemption is necessary, is a sound mix of localism and protection from purely local passions.(277) Here I think my views are buttressed by the evolution of the modern American legal system, in which most important decisions about contract law, tort law, property law, corporate law, criminal law, family law, and similar areas remain governed by state law--again, subject to federal preemption in certain areas when Congress specifically concludes that state law has reached improper results.

      Implementing state RFRAs in all fifty states, of course, would require tremendous effort,(278) but a federal son-of-RFRA could implement a basically state common-law model without requiring exemption supporters to fight all that legislative inertia: The federal statute could specifically exclude those states that have their own state RFRAs or state constitutional provisions establishing an exemption regime,(279) or it could specifically exclude those state statutes that explicitly say that they should not be covered by the federal son-of-RFRA.(280) As with the pure state RFRA approach, both options would initially leave the question of exemptions to courts, but would let state legislatures supersede any judicial decisions with which the legislatures disagree.(281)

      It may seem odd to have a federal statute be so easily avoidable by state legislatures, but, for the reasons given above, the point of these federal statutes would not be to transfer final decision-making authority from state legislatures to federal judges or to Congress. Rather, it would be to shift the burden of overcoming state legislative inertia and to make sure that religious exemption requests are seriously considered. If a state legislature deliberately decides that a certain exemption ought not be granted, then its judgment, not the courts' judgment or Congress's judgment, should prevail.

      Federally mandated exemptions from narrow sets of state laws might sometimes be sensible, just as narrow federal preemption of state law is sometimes sensible.  If Congress concludes that a certain kind of state commercial regulation--for instance, a requirement that certain businesses that operate in interstate commerce be open seven days a week, with no exception for Sabbatarians--excessively burdens religious objectors, it can require religious exemptions from such a regulation. Here Congress would have itself made a value judgment, and implementing such a judgment might well be a prudent exercise of the power to regulate commerce among the states.  Such action falls within the statutory exemption model, in which a legislature creates explicit, specific exemptions from specific laws. But if Congress isn't prepared to make its own value judgments about state regulations, it should leave the judgment to state-by-state decision making, rather than transferring it under a federal son-of-RFRA to federal judges.

      3. Embracing Legislative Responses to Common-Law Exemption Making

      a. The Need to Draw Lines Between Practices

      My analysis also suggests that there is nothing inherently improperly discriminatory about legislative repeals of some exemptions created by courts, or about legislative enactments of some exemptions that were rejected by courts.  Such exemption-by-exemption decisions would neither violate constitutional equality guarantees(282) nor be inherently unfair.(283)

      Lines between practices (as opposed to lines between different sects engaging in the same practice) must be drawn. There can be no religious exemptions from murder laws, battery laws, or theft laws. If there is to be a religious exemption from peyote laws or housing discrimination laws, some decision maker must distinguish one set of practices from another.

      If the process of deciding which exemptions should be granted "d[id] not necessarily, contrary to the assertion in Smith, involve the exercise of policy discretion or balancing of interests,"(284) or if it could be done in a "more principled" way by judges than by the legislature,(285) there might be reason to leave the test entirely in the courts' hands. We do rely on courts, for instance, to answer relatively objective questions such as whether a claimant is actually sincere in his beliefs or whether a law compels a person to perform an act that he believes religiously forbidden.

      But as I have argued above, there is no principle, in the sense of a more or less objectively applicable rule, that can resolve the tough questions inherent in decisions about exemptions.  Should peyote, marijuana, and alcohol be treated the same or differently? What about race discrimination, marital-status discrimination, and sexual-orientation discrimination? Theorists of liberty look for a general principle that can define the scope of the countervailing rights of others or can tell us when externalities become improper,(286) but they haven't yet found one that is so clearly proven that courts can impose it on the legislature as a constitutional command.(287) If the questions in exemption cases are questions of principle, they are so only in the sense of calling for the application of moral principles; and in our democratic system, the moral principles of legislators as to what is a countervailing private right or an improper externality should generally prevail over the moral principles of judges.

      b. The Early 1900s Experience

      Moreover, as with the notion that the Constitution requires accommodation of religious exemption claims, the notion that the Constitution requires equal treatment of different kinds of religiously motivated conduct runs up against the experience of the early 1900s. During those years, the Court often concluded that the Equal Protection Clause required different practices to be treated the same way. One case, for instance, struck down statutes that awarded attorney fees to shippers in successful lawsuits against railroads, but didn't likewise give fees to railroads when they prevailed.(288) Another struck down a tax that treated short mortgages differently from long mortgages.(289) Another struck down a statute that barred injunctions in certain labor disputes but allowed injunctions in other cases.(290)

      The problem, as the Court ultimately recognized in rejecting these holdings, is that whether two kinds of conduct should be treated alike calls for the same sort of moral and practical judgment about countervailing private rights and interests that is called for by the decision about whether certain conduct should be restricted.(291)  Different kinds of conduct have different effects on the participants and on third parties.  To the 1915 Court, a prevailing shipper and a prevailing railroad may have seemed identically situated for purposes of an attorney fee award, but the legislature had come to a different conclusion. And, regrettably, the courts have no principled way of determining when the differences are great enough to justify different treatment.

      Exemptions--whether constitutional, statutory, or common law--for a particular practice should be evenhanded as to the religious groups who engage in the practice: The government ought not be able to grant a sacramental wine exemption from Prohibition to Catholics but not Jews.(292) It is true that even here there may be relevant differences between some groups; for instance, if Catholics tend to have fewer drinking problems per capita than Jews, one might argue that applying Prohibition to Jews passes strict scrutiny and applying it to Catholics fails. Making such judgments, though, about the relative merits of various religious groups is a particularly dangerous task for courts. These differences, even if real, ought not be considered by our judicial system.(293)

      But there is no escaping the need to distinguish among different practices. The question is: Who should ultimately distinguish them, courts or legislatures? Given the absence of any principle by which they can be objectively distinguished, the ultimate decision is better left to the political process. And as I've argued above in Part III.D.4, though this process is biased in favor of the moral and practical judgments of the majority and against the judgments of minorities (whether religious or ideological), this preference for the majority's judgments about what constitutes harm to others is both inevitable and proper in our democratic system.

      c. An Example

      Let me illustrate the above with a concrete example. Say a court applying a state RFRA carves out a religious exemption from a ban on housing discrimination based on marital status or sexual orientation. To do this, the court must conclude that there is no compelling interest in preventing each incident of such discrimination, a conclusion that will probably be based on a judgment that in its view the tenants have no private right to equal treatment.(294) Furthermore, the court must conclude that there is no compelling interest in making sure that people have access to some housing in each locality, or that if there is such an interest, so few landlords will claim the religious exemption that it won't seriously interfere with this interest. This would reflect a judgment that declining to rent to people based on their marital status or sexual orientation, even if this leaves them with difficulty finding housing, doesn't qualify as inflicting an improper externality on them, or that religiously motivated refusals to rent will not in practice make it hard for people to find housing.

      If legislators conclude otherwise, on either normative or empirical grounds, then it makes sense for them to repeal or modify(295) the exemption created by the court decision. This wouldn't be an attack on religious freedom or even an attack on religious exemptions more broadly.(296)  No one claims that all or even almost all important laws should be subject to religious exemptions, and someone must draw the line between those exemptions that should be granted and those that shouldn't. There is nothing inherently wrong with the legislature's drawing the line in a place different from where a court would.

      Nor would this be improper repression of a minority religion that insists on such a practice.  True, the legislature has judged this practice more harshly than it has judged other practices, such as religious discrimination in hiring by religious organizations,(297) or race discrimination in rental of very small owner-occupied apartment buildings.(298) But it is the legislature's job to determine which practices are more harmful than others, and there is no objective standard that courts can apply to reject such a determination. If legislators conclude that certain conduct unacceptably trammels the rights or interests of others, it is proper for them to forbid such harmful conduct.

      In fact, it might be helpful for a state RFRA to say clearly that the legislature expects to modify or repeal some judge-made exemptions in the future, for instance by saying (as Part II.D suggests), "The government shall not substantially burden a person's exercise of religion unless (a) it demonstrates that such a burden [passes strict scrutiny], or (b) the legislature has specifically exempted the government action from the coverage of this statute." Such an explicit statement would help fend off future political claims that legislative alterations of the exemption regime are "contrary to the spirit of the state Religious Freedom Restoration Act." If my analysis is right, the possibility of such alterations is a fundamental part of any sound exemption regime.

      4. Embracing the Facial Exclusion of Certain Laws from State RFRAs

      Likewise, it may sometimes make sense for a legislature to exclude certain classes of laws from a state RFRA when the state RFRA is enacted. If, for instance, a legislature concludes that antidiscrimination laws do generally protect a private right to equal treatment, the legislature may reasonably implement this value judgment by specifically providing that the state RFRA will not apply to this law.(299) Similarly, if a legislature believes that prisoner claims for exemption from prison rules should be treated differently from claims by law-abiding citizens for exemption from general laws, it should say so in the state RFRA.(300)

      The common-law framework is aimed at letting courts make the initial value judgments in situations about which the legislature hasn't thought, or at least hasn't decided. If the legislature reaches a decision on such a situation--and especially if it anticipates, perhaps because of cases in other states, that courts might arrive at a different result--it should enact this judgment up front, rather than risking court decisions that (in its view) would allow invasions of private rights or impositions of improper externalities.(301)

      At the same time, the legislature should avoid carving out exceptions that are so broad that they cover situations about which it might not have seriously thought. For instance, exempting all antidiscrimination laws from a state RFRA may be a mistake: Even if the legislature really believes that private discrimination based on race, sex, and religion generally violates private rights or imposes improper externalities, this need not necessarily be true of discrimination based on (say) marital status or age. Waiting for court decisions about a specific kind of discrimination claim lets the legislature make a judgment that is responsive to the moral and practical considerations of each particular problem, and that is informed by the courts' considered opinions rather than by an overbroad "discrimination bad, equality good."

      Legislative avoidance of anticipatory exclusions, and deferral of the decisions until the courts have considered the matter, is actually easier under state RFRAs than under harder-to-change state constitutional amendments or a federal son-of-RFRA. The harder a regime is to alter, the more rides on the legislature's getting it right the first time: If the legislators fear that courts will grant excessive exemptions in a certain area, any extra difficulty of revising such decisions creates an extra incentive to carve out that area up front, even if the carving out is overbroad. But when any corrections can be made by normal state legislation, the legislature can confidently let courts confront the matter in the first instance, at least unless the legislature is indeed certain about what it wants the result to be.

      G. Common-Law Exemptions and Diversity

      These arguments in favor of the common-law exemption regime--and against the statutory and constitutional exemption regimes--track some important points about religious and conscientious diversity and American law.

      Setting aside all the other baggage that the term "diversity" often carries,(302) there is a core to claims of diversity on which there is broad agreement. First, there is the fact of diversity of views on what God or conscience demands. Many reasonable, well-meaning people have very different attitudes about what foods one is morally or religiously barred from eating, how one should dress, how one should behave in business life and personal life, what are the proper roles of men and women, and many other questions. Maybe there is one right position on each of these issues, or at least on the most important ones, but there is certainly little agreement on what that position might be.

      Second, there is the related fact that many of us generally haven't thought hard about how people with unusual religious opinions view the world. Part of the reason may be that there is so much diversity--not just two or three sects, but dozens or hundreds. Another part may be that the diversity is dynamic, with new groups arising, tiny groups growing, big groups splitting, groups moving within the country, other groups moving into the country, and existing groups changing their views. Another may be that many religious or cultural groups intentionally or unintentionally isolate themselves to some extent, or at least keep some of their practices comparatively quiet. And another part may be that some views are so alien to our own, so irrational by our standards, that we tend not to think seriously about them.(303)

      Third, there is the fact of diversity of views on what constitutes harm to others, and thus when the law should step in to protect the others from the behavior that causes such harm. Just as what constitutes the good life is an essentially contested question, so is what constitutes good government and good protections of private rights against others' trespasses--hence the debates about minimum wages, antidiscrimination law, assisted suicide, drugs, guns, and many other topics.

      The common-law exemption regime is more responsive than the statutory exemption regime to the diversity of religions and consciences because statutory rules are often written with only a particular group of belief systems in mind.  Should driver's licenses have to include photographs? A typical legislature would answer this question by considering how much the requirement would burden typical motorists, how much it would protect other motorists, and how administratively convenient it would be. The legislature may be reluctant to require that, say, every driver's license include a fingerprint, because many motorists would see that as an invasion of privacy or as something of an accusation of criminality; but it may be willing to require that every driver's license include a photo because, after all, why would anyone be bothered by that?

      But this broad rule, designed with the mainstream in mind, may be unexpectedly burdensome to people whose religious or conscientious views are outside the mainstream, for instance those who believe that carrying a driver's license with one's photograph violates the Biblical prohibition on graven images.(304) A statutory exemption regime would require them to suffer these effects until they can get the legislature to exempt them, something the objectors may find very hard to accomplish, especially when they are only a small group. A common-law regime, on the other hand, gives courts more flexibility to accommodate the unexpected interests of diverse groups, at least where courts think the accommodation wouldn't harm others.

      The common-law exemption regime is also more responsive than the constitutional exemption regime to the other sort of diversity--the diversity of views about what constitutes harm. As Justice Holmes said in Lochner, the Constitution does not "embody a particular economic theory," but is rather "made for people of fundamentally differing views."(305) People throughout the country, and their elected representatives, may have views on where harm to others starts (and where the freedom to act so long as one doesn't harm others therefore stops) that differ from the Court's views. A constitutional exemption regime would give the Justices the power to impose their beliefs on local, state, or national majorities whose beliefs are fundamentally different. A common-law exemption regime, on the other hand, would let legislators enact, on due consideration, their views of what constitutes harm to others, and would thus enable diverse moral communities to defend what they think are important private individual rights and interests.

      Of course, one could argue that, in respecting the diversity of communities' views about what constitutes harm, the common-law exemption regime insufficiently protects the diversity of religious and conscientious desires--just as one could argue that, in better respecting the diversity of communities' views about the proper relationship between employer and employee, the rejection of Lochner insufficiently protects the diversity of private desires about how to use one's labor or property. Earlier in Part III, I tried to respond to this argument by explaining why I think such judicial deference to legislative views about harm is the better approach.  The common-law regime, though, does significantly protect diversity of religious and conscientious views in the many cases where they can be protected without interfering with any considered legislative judgment about protecting others against harms.

      This also connects to the questions, discussed in Part I.E.2 and Part III.E.2, about whether RFRA should be imposed (to the extent constitutionally permissible) by the federal government,(306) by state governments in a way that binds all local governments, or by state governments with a provision that local governments can override any judicially crafted exemptions. I generally think, for the reasons mentioned already, that the case for deferring to state-by-state diversity is more compelling than the case for deferring to locality-by-locality diversity. The more important point here, though, is that any prescription on this matter, whether mine or some other, will be related to one's views on the propriety of diverse local cultures enacting into law their own judgments of harm.

      IV. Applications to Other Claims of Liberty

      A common-law model like the one I describe may also make sense for some other claims of liberty that courts have refused to constitutionalize. I tentatively outline a few such possibilities here as potential areas for future inquiry, though I won't go into them in depth.

      Consider for instance the claimed right to gather news. Some news-gathering techniques may violate generally applicable laws. Investigative journalists may try to get information by lying about who they are, which may be civilly actionable fraud.(307) They may smuggle cameras into places that aren't open to the public, such as the kitchens of restaurants that are being investigated for possible unsafe food handling, which may violate invasion of privacy laws, unconsented taping laws, or trespass laws.(308) They may set up a sting operation to find evidence that someone is taking bribes, selling drugs, or distributing child pornography, which may itself involve violations of bans on bribery, drugs, or child pornography.(309) They may threaten the subject of a certain damaging story with publishing the story as is unless he agrees to answer questions about his conduct, which may violate broadly worded blackmail laws.(310)

      Courts have rightly rejected claims of a constitutionally mandated news-gathering exemption from such generally applicable laws.(311) If the legislature deliberately concludes that all fraud, unconsented taping, participation in child pornography transactions, and threats of embarrassing revelations violate private rights or impose improper externalities, the courts ought not set these conclusions aside, even in the interest of more effective news gathering.(312)

      On the other hand, when the legislatures enacted these laws, they may not have thought about the rare person who violates the laws for news-gathering purposes. As with religious exemptions, (1) there's a fairly rare claim of exemption (2) that the legislature may not have fully anticipated and (3) that involves what some may see as a morally or practically significant extra factor--the news-gathering purpose.(313) It may make sense for the legislature to delegate to courts the power to carve out limited exemptions for news gathering, subject to legislative revision.

      These exemptions shouldn't protect news gatherers in every case, just as a common-law religious exemption regime shouldn't protect religious objectors in every case. News-gathering purposes shouldn't, for instance, authorize reporters to break into buildings, or to do things that help criminals get away with their crimes. But courts might conclude that certain lies

      told by investigative journalists in trying to get a story--such as falsifying a resume to get a job at a company that one is investigating--shouldn't be actionable as fraud,(314) or that certain audiotaping which would be invasion of privacy if done for some motives shouldn't be actionable if done to investigate misconduct,(315) or that journalists should be entitled to certain exemptions from the general duty to testify or to turn over records when subpoenaed.(316)

      As with common-law religious exemptions, the legislature could then revisit these questions and, if it wanted to, eliminate or cut back the judicially crafted exemptions. But to do this it would have to seriously consider the special problems raised by applying the law to news gathering, something it might not have done when the law was originally enacted.

      Likewise, a similar approach--partly adopted in Florida and Texas(317)--may protect landowners from excessively burdensome land-use regulations. As with religious practice, many accept the notion that everyone should be able to use his land as he sees fit (unless he is compensated for loss of this right) so long as his use doesn't harm others; but as with religious practice, there are lively debates about what qualifies as harm to others. It may thus make sense for the legislature to let courts order compensation for any sufficiently large loss of value to land (say, over 25%) caused by land-use regulations, if the court concludes--subject to legislative override--that the prohibited land use didn't infringe others' rights or inflict improper externalities on them. Such a statute would fit the three conditions outlined above for when common-law judicial exemptions are sensible: (1) Many land-use rules generally impose relatively small burdens in most applications, but large burdens in a few cases; (2) the magnitude of these burdens may often be hard to foresee, since often so much turns on the particular characteristics of each parcel; and (3) a legislature may reasonably conclude that the presence of an unusually large economic burden on a particular landowner is a special factor that may counsel in favor of a special result in these relatively unusual, hard-to-anticipate situations.

      This approach, however, probably isn't appropriate for liberty claims more generally. For instance, a RFRA-like statute that lets courts carve out exemptions for private consensual sexual practices would make little sense, because the prohibitions on such practices are typically deliberate. Such a statute wouldn't call on courts to carve out exemptions for the (1) unusual, (2) uncontemplated cases that (3) pose special moral or practical problems which differ from those raised by the core case to which the laws apply. Rather, it would ask courts to overturn considered (albeit perhaps wrong) legislative judgments that were intended to apply to the very situations in which the defense is raised. The best way to legislatively protect sexual rights is for the legislature to determine which restrictions are unsound and then repeal them, not to leave the matter to the courts.



      "The world has never had a good definition of the word liberty," Lincoln said in 1864:

      We all declare for liberty; but in using the same word we do not mean the same thing. With some the word liberty may mean for each man to do as he pleases with himself, and the product of his labor; while with others the same word may mean for some men to do as they please with other men, and the product of other men's labor. Here are two, not only different, but incompat[i]ble things, called by the same name--liberty. And it follows that each of the things is, by the respective parties, called by two different and incompat[i]ble names--liberty and tyranny.(318)

      This, in a paragraph, is the tragedy facing libertarians, whether those committed to liberty broadly or to religious liberty specifically. Most people believe in liberty, and it's tempting for each person to see the Constitution as a broad charter of liberty that prohibits tyrannical legislation. But each of us believes in a liberty--and its flip side, the body of private rights and interests that none of us is properly at liberty to infringe--that may differ from the one in which our neighbors believe.

      We don't like to see our cherished claims of liberty against government restraint put to mere popular vote, but we don't like to see others at liberty to violate what we believe to be our cherished private rights. Few of today's debates have the moral or political salience of the slavery controversy, but most do involve some incompatibility, similar in kind if not quite in degree, between competing conceptions of liberty: Consider, for instance, the liberty to use one's property as one sees fit versus the freedom from discrimination: "two not only different but incompatible things, called by the same name--liberty."

      The lesson of the early 1900s substantive due process experience is that the final call on the definition of "liberty" and "the rights of others," outside certain narrow areas, must therefore be made through the political process. Liberty writ large may not be enforced by judges as a matter of constitutional command. At the same time, the lesson of the common-law tradition is that judges may play a valuable role in initially defining "liberty" and "the rights of others," and that shifting this initial authority to judges can help protect liberty.

      The common-law religious exemption model can't--and shouldn't--protect claims of liberty of action, grounded solely in the religious motivation for one's actions, against the considered judgment of the democratic process. But it can protect them against legislative inertia, against the mechanical application of rules that were designed without any thought about religious objectors, and against local and bureaucratic overreaching. This isn't all that one side in the debate may hope for, but neither is it all that the other side may fear. And especially because we often find ourselves on different sides of debates about different exemption claims, this compromise may be the best we can reasonably expect.

      1 For examples of cases in which such claims were raised, see infra Part III.B.

      2 See, e.g., Wisconsin v. Yoder, 406 U.S. 205, 214-15 (1972); Sherbert v. Verner, 374 U.S. 398, 406-07 (1963).

      3 494 U.S. 872 (1990). Smith preserved the constitutional exemption mandate in two sets of cases: where there is an "individualized governmental assessment of the reasons for the relevant conduct," as in Sherbert, and where there is a hybrid claim brought under both the Free Exercise Clause and another constitutional protection, as in Yoder. See id. at 881, 884. While these exceptions will be important in some cases, I largely set them aside here.

      4 42 U.S.C. § 2000bb (1994). "Government shall not substantially burden a person's exercise of religion" unless "application of the burden to the person . . . is in furtherance of a compelling governmental interest; and . . . is the least restrictive means of furthering that compelling governmental interest," id. §§ 2000bb-1(a), (b); this "least restrictive means"/"compelling governmental interest" inquiry is a strict scrutiny test.

      5 521 U.S. 507 (1997).

      6 See S.B. 604, 1998 Reg. Sess. (Ala. 1998) (state constitutional amendment approved by the voters in the Nov. 1998 election); Conn. Gen. Stat. Ann. § 52-571b (West 1998); 1998 Fla. Sess. Law Serv. ch. 98-412 (West); 775 Ill. Comp. Stat. Ann. 35/10 (West 1999); R.I. Gen. Laws § 42-80.1-3 (1998); see also H.B. 2421, 43d Leg., 1998 Reg. Sess. (Ariz. 1998); A.B. 1617, 1997-98 Reg. Sess. (Cal. 1997), vetoed Sept. 27, 1998; H.R. 865, 144th Gen. Assembly, 1997-98 Reg. Sess. (Ga. 1997); H.B. 2916, 77th Leg., 1998 Reg. Sess. (Kan. 1998); H.B. 167A, 1998 Spec. Sess. A (La. 1998); S.B. 515, H.B. 1041, 1998 Reg. Sess. (Md. 1998); S.B. 678, H.B. 4376, 89th Leg., 1997 Reg. Sess. (Mich. 1997); H.B. 1470, 2d year of the 155th Sess. of the General Court (N.H. 1997), defeated Feb. 18, 1998; S.B. 321, A.B. 903, 208th Leg. (N.J. 1998); S.B. 5673, A.B. 8499, 220th Annual Leg. Sess. (N.Y. 1997); H.B. 3469, 70th Leg. Assembly, 1999 Reg. Sess. (Ore. 1999); H.B. 2068, 181st Gen. Assembly., 1997-98 Reg. Sess. (Pa. 1997); H.B. 5045, 112th Sess. of Gen. Assembly (S.C. 1997); S.B. 3054, H.B. 3051, 100th Gen. Assembly, 1998 Reg. Sess. (Tenn. 1998); H.B. 550, 1997-98 Leg. Sess. (Vt. 1997-98); H.B. 1, 1998 Leg. Sess. (Va. 1998). The language of the statutes generally closely tracks the language of the federal RFRA. See, e.g., 1998 Fla. Sess. Law Serv. ch. 98-412 (West) ("The government shall not substantially burden a person's exercise of religion [unless] application of the burden to the person . . . [i]s in furtherance of a compelling governmental interest; and . . . [i]s the least restrictive means of furthering that compelling governmental interest.").

      7 Besides the name of the statute, consider also Douglas Laycock, The Religious Freedom Restoration Act, 1993 BYU L. Rev. 221, 221 ("[RFRA] would enact a statutory version of the Free Exercise Clause."). Laycock helped draft RFRA and has been one of its leading academic advocates. Later in the article he acknowledges that "[u]nder the statute, the judicial striking of the balance is not final. If the Court strikes the balance in an unacceptable way, Congress can respond with new legislation [either providing a broader religious exemption or a narrower one]." Id. at 253. But this, to him, is not a virtue but a vice:

      The possibility of amendments to RFRA should be reassuring to a Court that doubts its own capacity to authoritatively balance competing interests, but it is also the weak spot of the legislation. . . . Congress should resist such amendments; it should recognize that RFRA is a quasi-constitutional statute that should not be lightly amended.

      Id. at 254.

      One exception to this rhetoric of restoration is Michael Stokes Paulsen's excellent A RFRA Runs Through It: Religious Freedom and the U.S. Code, 56 Mont. L. Rev. 249, especially 255-56 (1995); his analysis, though, generally deals with rather different questions from the ones I face here.

      8 I will generally use "state RFRAs" as somewhat imprecise shorthand for "state RFRAs as applied to state laws, and the federal RFRA as applied to federal laws." It would be technically more accurate to talk about "jurisdiction-by-jurisdiction RFRAs," but "state RFRAs" is shorter and, on the whole, clearer.

      9 Of course, a constitutional decision technically isn't final either--it can be overridden by a constitutional amendment--but the federal amendment process is so onerous that constitutional decisions are in practice almost always binding on the political branches. See infra note 46.

      10 As the discussion in the text suggests, I use "common law" in the general sense of "law made initially by judges but subject to statutory override," rather than in the specific sense of the particular body of contract, tort, and property law that Anglo-American courts have evolved over the past several centuries.

      11 My arguments about the federal Free Exercise Clause also apply to analogous state provisions. See infra Part III.F.1.

      12 Parts of my argument here may be reminiscent of the "legal process" school. See, e.g., Gary Peller, Neutral Principles in the 1950's, 21 U. Mich. J.L. Reform 561 (1988) (describing and criticizing the "legal process" school); id. at 596-97 (describing how the school distinguishes judicial value judgments in common-law cases from similar judgments in constitutional cases, a distinction that is closely mirrored by my discussion in Part I and Part III). Much ink has been spilled both criticizing and defending this school, and I don't intend to enter this general debate here. I do want to mention two basic points: First, my criticism of judicial value judgments is limited to constitutionalized judicial determinations of whether a particular harm is really a harm, determinations that arise only in certain constitutional contexts, see infra Part III.A.1; I do not endorse some of the broader proceduralist criticisms of judicial protection of individual rights generally. Second, I try to address the applicability of one especially forceful critique of the legal process school--the argument that deference to democratic legislative judgment is inappropriate when the legislative judgment is biased against the interests of particular groups--in Part III.D.4.

      13 See, e.g., National Prohibition Act, 21 U.S.C. § 16 (1919) (repealed 1933); Military Selective Service Act of 1967, 50 U.S.C. App. § 456(j) (1994); cf. U.S. Const. art. I, § 3, cl. 6; art. II, § 1, cl. 8; art. VI, § 3; amend. IV (providing for affirmations as alternatives to oaths).

      14 374 U.S. 398 (1963). There were hints of the Sherbert regime in Supreme Court opinions as early as 1940, see Cantwell v. Connecticut, 310 U.S. 296, 303-04 (1940), but also contrary statements, see, e.g., Chaplinsky v. New Hampshire, 315 U.S. 568 (1942):

      [E]ven if the activities of the appellant which preceded the incident could be viewed as religious in character, and therefore entitled to the protection of the Fourteenth Amendment, they would not cloak him with immunity from the legal consequences for concomitant acts committed in violation of a valid criminal statute.

      Id. at 571.

      15 494 U.S. 872, 890 (1990). For some qualifications, see supra note 3.

      16 See supra note 7.

      17 521 U.S. 507 (1997).

      18 See, e.g., In re Young, 141 F.3d 854 (8th Cir. 1998). Some courts have assumed, mostly without discussion, that Boerne struck down RFRA even as to federal action, see, e.g., Robinson v. District of Columbia Gov't, No. CIV. A. 97-787, 1997 WL 607450, at *1 n.1 (D.D.C. July 17, 1997), but that has to be wrong. Boerne relied on Congress's lack of power to control state law, but Congress must have the power to carve out exceptions from its own laws. See also infra Part I.F.

      19 See supra note 7; infra note 70.

      20 Of course, in states whose constitutions have been read as creating a constitutional exemption regime, such legislative overrides of state RFRA decisions may do little good, since courts will often reach the same results under the non-legislatively-overridable state constitutional provisions as they did under the state RFRA. See infra Part III.F.2 (discussing why state constitutional Free Exercise Clauses should not be interpreted as creating a constitutional exemption regime).

      21 See William P. Marshall, The Case Against the Constitutionally Compelled Free Exercise Exemption, 40 Case W. Res. L. Rev. 357, 411 (1989-90) ("The cases [under strict scrutiny in religious freedom cases] have been inconsistent . . . ."); Michael W. McConnell, Religious Freedom at a Crossroads, 59 U. Chi. L. Rev. 115, 170 (1992) ("The [strict scrutiny] test . . . provided little guidance to legislatures or lower courts . . . ."); Ellis West, The Case Against a Right to Religion-Based Exemptions, 4 Notre Dame J.L. Ethics & Pub. Pol'y 591, 597, 608 (1990) (describing the strict scrutiny test in religious freedom cases as leading "to unpredictable results" and being "vague enough that courts can manipulate it to justify almost any decision they make"); articles cited infra note 106 (pointing out that the test claimed to be strict scrutiny but really wasn't). But see infra Part II (discussing whether strict scrutiny may, despite its vagueness, constrain courts' discretion to some degree).

      22 As I explain in Part II.D, legislatures can to some extent blunt this argument by writing the state RFRA in a way that specifically contemplates future legislative revision. If the state RFRA on its own terms applies "unless the legislature has specifically excluded the government action from the coverage of this statute," then it becomes slightly harder to argue that such legislative exclusions are "against the spirit of the state Religious Freedom Restoration Act."

      23 The California RFRA was vetoed, the Illinois RFRA was partially vetoed, and the New Hampshire bill was voted down in the House. See, e.g., A.B. 1617, 1997-98 Reg. Sess. (Cal. 1997); H.B. 2370, 90th Gen. Assembly (Ill. 1997); H.B. 1470, 2d Legis. Sess. (N.H. 1998). Many other bills, cited supra footnote 6, died in their respective state legislatures. The Illinois veto was legislatively overriden, and the California law may get reenacted, but the governors' willingness to veto the bills shows that such bills are seen as part of the normal legislative process, not as motherhood-and-apple-pie laws that no political official could safely oppose.

      24 Cf. Laycock, supra note 7, at 229 ("In each request for a legislative exemption, churches are likely to find an aroused interest group on the other side, and they will be trying to amend that interest group's statute."). Laycock is using this as an argument in favor of a constitutional exemption regime, but his specific claim here--applicable equally to his argument and to mine--is that "You're voting against religious freedom!" isn't always a devastating political charge.

      As my colleague David Sklansky has pointed out, it would be interesting to know how often legislators in fact override judicial rulings under state RFRAs, and how effective the argument that "you shouldn't tamper with this important statute" is in fighting such moves. State RFRAs are, however, too new for any empirical inquiry into these questions.

      25 Fed. R. Evid. 501.

      26 See 42 U.S.C. §§ 6928(f)(4), 7413(c)(5)(D) (1994); N.C. Gen. Stat. §§ 130A-26.1(e), 143-215.6B(g), 143-215.88B(d), 143-215.94X(g), 143-215.114B(e) (1998).

      27 See 17 U.S.C. § 107.

      28 See 15 U.S.C. § 1125(c)(4)(A).

      29 42 U.S.C. § 2000e-2(e)(1); see Eugene Volokh, The California Civil Rights Initiative: An Interpretive Guide, 44 UCLA L. Rev. 1335, 1369-74 (1997) (describing the bona fide occupational qualification principles that courts have developed using their own moral and practical judgment, and the continuing moral and practical judgments that application of these principles requires).

      30 Cf. Guido Calabresi, A Common Law for the Age of Statutes (1982). Judge Calabresi's work is the deepest recent treatment of common-law decision making against a statutory backdrop, but his proposal is focused on the problem of statutory obsolescence. RFRAs and the other delegations to the judiciary mentioned in these paragraphs are aimed at the somewhat different problem of dealing with exceptional cases in which the legislature foresees that its explicit statutory balance may be inapplicable.

      31 This answers the question that Judge Calabresi asks in a similar context: "What justifies court power to make temporary rules and thereby to assign the burden of overcoming inertia and of getting those rules revised?" Id. at 93. Sometimes the legislature, faced with the foreseeable likelihood of unforeseeable events, may sensibly leave these events for courts to deal with in the first instance.

      32 Of course, they're not always hard to anticipate: When they're pretty obvious, as with the draft law or the Prohibition Act, it makes sense for the legislature to enact an explicit statutory exemption.

      33 Cf. Thomas C. Berg, What Hath Congress Wrought? An Interpretive Guide to the Religious Freedom Restoration Act, 39 Vill. L. Rev. 1, 21 (1994) ("The political branches . . . may not be aware of the effect of a law on a religious minority . . . ."); Laycock, supra note 7, at 225 (asserting that some burdens on religious practice arise "because the legislature was unaware of [the group's] existence and failed to provide an exemption"); Laycock, supra note 7, at 229 (discussing the case of someone who had a religious objection to having her photograph on her driver's license, and claiming--probably correctly--that "[i]t would be nearly impossible for any legislature to know in advance about a believer like [this claimant] and to enact an exemption for her"); Mark Tushnet, "Of Church and State and the Supreme Court": Kurland Revisited, 1989 Sup. Ct. Rev. 373, 400 ("In core applications, where up to the point of litigation legislators have simply overlooked the impact of a neutral rule on some religious believers, the doctrine of mandatory accommodation [i.e., a constitutional exemption regime] plainly produces 'better' outcomes.").

      34 See, e.g., State v. Miller, 538 N.W.2d 573, 579 (Wis. Ct. App. 1995) (creating a religious exemption from such a requirement).

      35 See, e.g., Christopher L. Eisgruber & Lawrence G. Sager, Why the Religious Freedom Restoration Act Is Unconstitutional, 69 N.Y.U. L. Rev. 437, 448-49 (1994) [hereinafter Eisgruber & Sager, Why RFRA Is Unconstitutional]; Christopher L. Eisgruber & Lawrence G. Sager, The Vulnerability of Conscience: The Constitutional Basis for Protecting Religious Conduct, 61 U. Chi. L. Rev. 1245, 1260-70 (1994) [hereinafter Eisgruber & Sager, Vulnerability of Conscience].

      36 See infra Part I.G.

      37 See 21 C.F.R. § 1307.31 (1998); state statutes cited in Smith v. Employment Div., 763 P.2d 146, 148 n.2 (Or. 1988), rev'd, 494 U.S. 872 (1990).

      38 For more detail on this and other arguments, see Douglas Laycock, Religious Liberty as Liberty, 7 J. Contemp. Legal Issues 313, 316-19 (1996).

      39 See infra Part III.A.3 for more discussion of this.

      40 See, e.g., Cal. Civ. Code § 43.5 (West 1982) (abolishing tort claims for alienation of affections, criminal conversation, seduction of a person over the age of consent, and breach of promise of marriage); cf. Veeder v. Kennedy, No. 20360, 1999 WL 92620 (S.D. Feb. 24, 1999) (refusing to abrogate the alienation of affections tort).

      41 See infra note 161.

      42 See, e.g., Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 394-95 (1926).

      43 See Teresa Stanton Collett, Heads, Secularists Win; Tails, Believers Lose--Returning Only Free Exercise to the Political Process, 20 U. Ark. Little Rock L.J. 689, 694 (1998) (pointing out that small sects may not even be able to get legislators' attention); Douglas Laycock, The Remnants of Free Exercise, 1990 Sup. Ct. Rev. 1, 57 ("Legislative exemptions are hard to get because the political dynamic of the modern regulatory state recognizes no natural limits to the pursuit of secular values."); id. at 58 (stressing the magnitude of political battles for some statutory exemptions, and their possible corrosive effect); McConnell, supra note 21, at 139 ("Those groups whose beliefs are least foreign and least offensive to the mainstream, and those with the largest numbers and greatest visibility, will be better able to protect themselves than will the smaller, more unpopular groups.").

      44 See, e.g., Maxwell L. Stearns, The Public Choice Case Against the Item Veto, 49 Wash. & Lee L. Rev. 385, 416 n.175 (1992) (describing the legislative process as creating a supermajority requirement). This need not be supermajority enthusiasm, only supermajority acquiescence.

      45 Or, of course, by a supermajority in each house if the executive vetoes the bill. In one state, Nebraska, the bill need only be approved by the unicameral legislature and the governor.

      46 In a sense, of course, even a federal constitutional exemption regime would only shift the burden of overcoming legislative inertia; a sufficiently unpopular court decision could still be overcome by a constitutional amendment. But there is a huge difference between the dramatic supermajority needed to amend the Constitution and the much smaller supermajority needed to overturn a common-law decision. The experience of the early 1900s, which I discuss in more detail in Part III.C, attests to this: Progressives and populists did have to overcome a considerable amount of legislative inertia to overturn the common-law regime of relatively unregulated business life, but this difficulty is generally considered a proper obstacle that reformers must surmount in a government of checks and balances. The constitutionalization of these common-law principles, on the other hand, was improper, because it created a vastly greater obstacle to the popular will. Cf. infra Part III.F.1 (discussing state constitutional exemption regimes).

      47 Of course, some errors in the direction of too much liberty may be bad even if they are quickly corrected by the legislature; for instance, even a few months during which religiously motivated genital mutilation of girls is allowed may permanently injure many girls, in violation of their rights to bodily integrity. Nonetheless, precisely because most Americans see the practice as so obviously harmful, it seems extremely unlikely that American courts would erroneously grant such an exemption in the first place.

      48 Cf. Cheema v. Thompson, 67 F.3d 883, 886 (9th Cir. 1995) (holding that Sikh children had the right under RFRA to wear ceremonial daggers to school); State v. Singh, 690 N.E.2d 917, 921 (Ohio Ct. App. 1997) (Painter, J., concurring) (suggesting that RFRA allowed Sikhs to wear ceremonial daggers in public, and condemning the prosecutor for bringing a prosecution based on such conduct). But cf. People v. Singh, 516 N.Y.S.2d 412, 416 (Crim. Ct. 1987) (holding that the Free Exercise Clause did not allow Sikhs to wear ceremonial daggers in public).

      49 See, e.g., The Religious Liberty Protection Act of 1998: Hearing on H.R. 4019 Before the Subcomm. on the Constitution of the House Comm. of the Judiciary, 105th Cong. (1998) (statement of Douglas Laycock) (visited May 31, 1999) <http://www.house.gov/judiciary/222498.htm>; Laycock, supra note 43, at 48-49, 49 n.203.

      50 See, e.g., Grosz v. City of Miami Beach, 721 F.2d 729, 733 (11th Cir. 1983), vacated and remanded, 82 F.3d 1005 (11th Cir. 1996) (holding that the claim may be brought again under RFRA's tougher standards).

      51 See infra note 106.

      52 See, e.g., Frederick Mark Gedicks, RFRA and the Possibility of Justice, 56 Mont. L. Rev. 95, 95 (1995) (asking why it made sense for RFRA supporters to "work so hard to restore a test that had never made all that much difference," and concluding that "the compelling interest test made a difference at the administrative and trial level, where many religious exemption cases are decided, even if the test was largely ineffective in appellate litigation"); Laycock, supra note 7, at 225; Douglas Laycock & Oliver S. Thomas, Interpreting the Religious Freedom Restoration Act, 73 Tex. L. Rev. 209, 244 (1994) ("Bureaucrats may be more likely to accommodate religious exercise when they know that a federal statute requires them to do so in most cases, and by giving religious claimants the bargaining leverage of a viable claim in court, RFRA encourages out-of-court settlements.").

      53 Cf. Gary T. Schwartz, The Logic of Home Rule and the Private Law Exception, 20 UCLA L. Rev. 671 (1973) (discussing the limits of home rule powers).

      54 For instance, Colorado's Amendment 2, see Romer v. Evans, 517 U.S. 620, 623-24 (1996), was prompted by some liberal Colorado cities' banning of sexual preference discrimination when the majority of Colorado voters in other areas opposed such bans.

      One can also support home rule by stressing the greater ease of moving from one local jurisdiction to another than from one state to another. This option of exit, often to a place only a few miles away, makes local laws somewhat less burdensome than state laws: If a religious group can't build the church it likes in one municipality, it might often be able to build it in another nearby. If a neutral local speech restriction incidentally bans a certain form of religious proselytizing, religious objectors might be able to proselytize a few miles away. Nonetheless, I doubt that this argument ultimately proves much, because what one locality may do, several nearby localities may likewise do for the same reason. This might mean that to avoid the local laws, a religious objector would have to move to an entirely different area, leaving behind friends, extended family, and work.

      Some state constitutions give local governments limited constitutional protection from interference by the state; the general rule, though, is that state laws that touch on matters of "statewide concern" still supersede contrary local ordinances, and I suspect that laws protecting what many see as an important individual right would qualify as relating to a statewide concern. See, e.g., Welter v. City of Milwaukee, 571 N.W.2d 459, 463 n.4 (Wis. Ct. App. 1997) (holding that protecting city employees' vested pension rights is a matter of statewide concern); Jacobberger v. Terry, 320 N.W.2d 903, 906 (Neb. 1982) (holding that protecting city residents' right to have their vote fairly considered in city elections is a matter of statewide concern); Midwest City v. Cravens, 532 P.2d 829, 834 (Okla. 1975) (holding that protecting city employees' rights to bargain collectively is a matter of statewide concern).

      55 See, e.g., Laycock, supra note 7, at 225 (asserting that some burdens on religion arise "because the secular bureaucracy is indifferent to [objectors'] needs"); Laycock, supra note 43, at 57 ("[A]genc[ies] tend to be single-mindedly focused on the benefits of their legislation; it is not for them to balance competing interests. Thus, those who . . . enforce much modern legislation tend to be wholly unsympathetic to claims that religious liberty requires an exemption.").

      56 This text is modeled on a provision of the federal RFRA: "Federal statutory law adopted after [the date of the enactment of this Act] is subject to this chapter unless such law explicitly excludes such application by reference to this chapter." 42 U.S.C. § 2000bb-(3)(b) (1994). The federal RFRA provision was aimed at a somewhat different problem from the one I discuss in the text, but its language seems suitable for my purposes.

      57 One can imagine even more possibilities. Fans of administrative agency decision making, for instance, could propose an administrative exemption model, in which the legislature exempts certain agencies--environmental agencies, for instance--from the normal operation of a state RFRA, but then requires the agencies to consider requests for exemptions in the first instance, subject to relatively deferential review by the judiciary, and of course subject to the possibility of revision by the legislature itself. This would retain the perceived advantages of administrative decision making in certain highly technical areas, while at the same time aiming to fight agencies' tunnel vision by making clear that the agencies must consider religious exemption requests alongside their more traditional concerns. Cf., e.g., American Indian Religious Freedom Act, 42 U.S.C. § 1996 (requiring administrative agencies to consider American Indian religious exemption claims, though not imposing any particular obligation to defer to those claims); Exec. Order No. 13,007, 3 C.F.R. § 196 (1996) (setting up regulations under the act). The legislature might even require administrative agencies to systematically go through their regulations and the statutes they enforce, identify possible burdens on religious belief, and, where appropriate, create new regulations that would accommodate those burdens. (I thank Jennifer Friesen for this suggestion.)

      I have no opinion on whether this particular approach would be sound, though I suspect one's view on this would depend on one's broader faith in agency judgment and expertise; I only want to point out that the common-law exemption model, implemented as it is by an enabling state RFRA statute, has room for many variations.

      58 Cf. Calabresi, supra note 30, at 11-12 (criticizing doctrines under which statutes are too easily struck down as unconstitutional, because such doctrines unduly shackle the legislature, and ultimately proposing a common-law model instead).

      59 Cf. Paulsen, supra note 7, at 255-56 (arguing that even if it was proper for courts to defer to congressional judgments under a constitutional exemption regime, courts should be less deferential under a common-law regime in which Congress had the option to exempt a challenged law from RFRA but declined to do so).

      60 Cf. Washington v. Glucksberg, 521 U.S. 702, 789 (1997) (Souter, J., concurring) ("[E]xperimentation . . . should be out of the question in constitutional adjudication displacing legislative judgments . . . .").

      61 Cf., e.g., the rather harmless-seeming child labor involved in Prince v. Massachusetts, 321 U.S. 158, 160-61 (1944).

      62 See, e.g., Gary T. Schwartz, The Beginning and the Possible End of the Rise of Modern American Tort Law, 26 Ga. L. Rev. 601, 647-99, especially 657, 697-98 (1992) (describing changes in tort law, including some outright--or close to it--reversals of fairly recently adopted doctrines).

      It is true that here the common-law regime would be set up by statute, and that courts at times seem to take stare decisis especially seriously in statutory interpretation cases. See generally William N. Eskridge, Jr., Overruling Statutory Precedents, 76 Geo. L.J. 1361 (1988). But as Justice Stevens pointed out, this principle generally has not been applied and ought not be applied when the legislature frames "statutes in sweeping, general terms, expecting the federal courts to interpret them by developing legal rules on a case-by-case basis in the common law tradition." Guardians Ass'n v. Civil Serv. Comm'n, 463 U.S. 582, 641 n.12 (1983) (Stevens, J., dissenting); cf. Eskridge, supra, at 1376-77 (discussing this point).

      63 Compare, e.g., Employment Div. v. Smith, 494 U.S. 872, 911-12, 916 (1990) (Blackmun, J., dissenting) (arguing that there is "no evidence that the religious use of peyote has ever harmed anyone," and that "the availability of peyote for religious use" wouldn't interfere with the fight against drug trafficking), with id. at 905 (O'Connor, J., concurring in the judgment) (accepting the argument that "the possession and use of controlled substances, even by only one person, is inherently harmful and dangerous . . . regardless of the motivation of the user," and that "uniform application of the criminal prohibition at issue" is essential to "preventing trafficking in controlled substances").

      64 I stress the "may be"; what I outline here is a force that I think will likely push judges mildly towards providing more exemptions under the common-law model than under the constitutional model, but there may be more powerful forces that cut against judicial recognition of exemptions under either model.

      65 See supra Part I.D; infra Part III.

      66 Compare California RFRA Legislation: Hearing Before the Cal. Assembly Judiciary Comm., Oct. 4, 1997 (Cal. 1997) (statements of Eugene Volokh), and Letter from Eugene Volokh to California Senate Judiciary Committee (June 24, 1998) (on file with author), with Letter from Eugene Volokh to Assemblyman Baca's Office (July 22, 1998) (on file with author) (repudiating the claim that there is something inherently wrong with leaving the initial exemption decisions to courts, but reaffirming prior criticism of the exclusion of secular conscientious objections, cf. infra Part I.G, and of the strict scrutiny test, cf. infra Part II).

      67 Marci A. Hamilton, The Religious Freedom Restoration Act is Unconstitutional, Period, 1 U. Pa. J. Const. L. 1, 11-12 (1998). Hamilton was speaking about the federal RFRA, including the federal RFRA as applied to federal law, but her criticism would equally apply to state RFRAs.

      68 See infra notes 115, 116 and accompanying text (explaining why this model poses no difficulty under state constitutional separation of powers jurisprudence).

      69 As I mention in Part I.G, the Establishment Clause may condemn exemption schemes, whether statutory or common law, that prefer religious objectors over conscientious objectors, but that is a separate question from whether there is something especially unsound about common-law exemption regimes as opposed to statutory ones.

      70 Cf. Joanne C. Brant, Taking the Supreme Court at Its Word: The Implications for RFRA and Separation of Powers, 56 Mont. L. Rev. 5, 6 (1995) ("Because Smith represents the Court's reasonable . . . refusal to undertake the task of balancing religious liberties against neutral government regulations, Congress cannot override that refusal."); id. at 17 ("Smith warns us that application of the compelling interest test in the free exercise context places courts in the position of making arbitrary and unprincipled choices."); Daniel O. Conkle, The Religious Freedom Restoration Act: The Constitutional Significance of an Unconstitutional Statute, 56 Mont. L. Rev. 39, 65 (1995) ("In mandating the very process of balancing to which the Supreme Court so strongly objected, RFRA . . . rejects the Court's institutional reasoning no less than it rejects the Court's substantive understanding of religious freedom."); Eisgruber & Sager, Why RFRA Is Unconstitutional, supra note 35, at 445 (saying that Congress insisted with RFRA "that the Court return to a test that it has abandoned for good reason"); id. at 464 (arguing that RFRA is "antagonistic . . . to the Court's judgment about the workability--the judicial manageability--of a particular doctrinal approach to the policing of state and local governmental policies"). All these quotes, except perhaps the last, suggest that RFRA is unconstitutional or at least unwise even as to federal statutes. See also Laycock, supra note 7, at 253 (arguing that the "institutional concerns [underlying Smith] do not apply to RFRA," in part because "Congress, rather than the Court, will retain the ultimate responsibility for the continuation and interpretation of [the] decision [to provide religious exemptions]").

      71 Employment Div. v. Smith, 494 U.S. 872, 889 n.5 (1990).

      72 To be precise, I don't condemn such balancing (whatever exactly that may mean) in the abstract. Rather, I condemn balancing decisions in which courts conclude that something which the legislature sees as harmful is really harmless.

      73 Common-law claims are traditionally seen as the province of state judges, but federal judges often engage in the creation of common law too, either in areas of federal authority (such as admiralty law), or in areas where a statute leaves them substantial lawmaking discretion (such as antitrust law, or, especially before the Copyright Act of 1976, copyright law). See, e.g., Frank H. Easterbrook, Workable Antitrust Policy, 84 Mich. L. Rev. 1696, 1702 (1986) ("Back in 1890 Senator Sherman and his colleagues protested the Sugar Trust and other malefactors and told the judiciary to do something about it. They weren't sure just what. Their statute does not contain a program; it is instead a blank check.").

      74 See, e.g., Eisgruber & Sager, Why RFRA Is Unconstitutional, supra note 35, at 453-54.

      75 I argue here that the possibility of legislative correction makes state RFRAs more legitimate and more attractive as a policy matter, but it may also make them politically less attractive to some lawmakers, who may dread the prospect of regularly being called on to decide whether to repeal particular judicially created religious exemptions--something that may often cut across the legislators' usual political bases and put them in no-win political situations that they'd rather avoid. These legislators may thus prefer a constitutional exemption regime, which may let them plead that their hands are tied because they can't repeal the judicially created exemption, even though the regime means less policymaking power for them. Cf. Mark A. Graber, The Nonmajoritarian Difficulty: Legislative Deference to the Judiciary, 7 Stud. Am. Pol. Dev. 35 (1993) (arguing that legislators are often delighted to punt certain questions to the judiciary, to be resolved as a matter of constitutional law). I am indebted to my colleague Daniel Lowenstein for pointing this out to me.

      76 Cf. Laycock, supra note 38, at 314-15 ("[T]he Constitution does give special protection to liberty in the domain of religion . . . . 'Because the Constitution says so . . . ' should be sufficient reason to vigorously protect religious liberty."). But cf. id. at 331 (arguing that deeply held conscientious belief should be treated as religion for constitutional purposes, and thus presumably for RFRA purposes too). For commentators who do support an exemption regime limited to religious objectors, see, e.g., Thomas C. Berg, The Constitutional Future of Religious Freedom Legislation, 20 U. Ark. Little Rock L.J. 715, 720 (1998); Timothy L. Hall,Omnibus Protection of Religious Liberty and the Establishment Clause, 25 Cardozo L. Rev. (forthcoming July 1999); Michael Stokes Paulsen, God is Great, Garvey is Good: Making Sense of Religious Freedom, 72 Notre Dame L. Rev. 1597 (1997) (reviewing John H. Garvey, What Are Freedoms For? (1996)).

      77 See, e.g., City of Boerne v. Flores, 521 U.S. 507, 536 (1997) (Stevens, J., concurring); Eisgruber & Sager, Why RFRA Is Unconstitutional, supra note 35, at 453-54; Marshall, supra note 21, at 388-94; West, supra note 21, at 600-04.

      78 See Ira C. Lupu, Why the Congress Was Wrong and the Court Was Right--Reflections on City of Boerne v. Flores, 39 Wm. & Mary L. Rev. 793, 809 (1998).

      79 See, e.g., Welsh v. United States, 398 U.S. 333, 343-44 (1970).

      80 See, e.g., Protos v. Volkswagen of Am., Inc., 797 F.2d 129, 137 n.4 (3d Cir. 1986) ("The breadth of the 'exemption' afforded by Title VII is underscored by the fact that in defining religion, the EEOC has used the same broad definition as the Selective Service employs for conscientious objector purposes."); Nottelson v. Smith Steel Workers, 643 F.2d 445, 454 n.12 (7th Cir. 1981) (noting the same broad definition); Redmond v. GAF Corp., 574 F.2d 897, 901 n.12 (7th Cir. 1978) ("We believe the proper test to be applied to the determination of what is 'religious' under § 2000e(j) can be derived from the Supreme Court decisions in [Welsh] and [United States v. Seeger, 380 U.S. 163 (1969)], i.e., (1) is the 'belief' for which protection is sought 'religious' in person's own scheme of things, and (2) is it 'sincerely held.'"); Ali v. Southeast Neighborhood House, 519 F. Supp. 489, 490 (D.D.C. 1981) ("Sincere beliefs, meaningful to the believer, need not be confined in either source or content to traditional or parochial concepts of religion. [Welsh.] See also [Seeger] for the definition of 'religious training and belief' as applied to a conscientious objector claim, which definition is no less appropriate here."); Wondzell v. Alaska Wood Prods., Inc., 583 P.2d 860, 866 n.12 (Alaska 1978) ("In order to avoid the danger of unconstitutionality we would interpret [the state statute] to accord the same privileges to all sincere conscientious beliefs, whether or not they are accompanied by a belief in a supreme being."); Kolodziej v. Smith, 682 N.E.2d 604, 607 (Mass. 1997). But see Seshadri v. Kasraian, 130 F.3d 798, 800 (7th Cir. 1997) (Posner, J.) (concluding that Title VII doesn't apply when "the plaintiff's belief, however deep-seated, is not religious").

      81 See supra note 21.

      82 See O'Lone v. Estate of Shabazz, 482 U.S. 342, 349 (1987) (prisons); Goldman v. Weinberger, 475 U.S. 503, 507 (1986) (military); see also Ira C. Lupu, Of Time and the RFRA: A Lawyer's Guide to the Religious Freedom Restoration Act, 56 Mont. L. Rev. 171, 191-98 (1995).

      83 See, e.g., United States v. Juvenile No. 1 (LWQ), 38 F.3d 470, 473 (9th Cir. 1994) (following United States v. Consuelo-Gonzalez, 521 F.2d 259, 264 (9th Cir. 1975) (en banc), which held in the free speech context that probation conditions are constitutional if they "can reasonably be said to contribute significantly both to the rehabilitation of the convicted person and to the protection of the public"); United States v. Tolla, 781 F.2d 29, 36 n.3 (2d Cir. 1986) (same); State v. Emery, 593 A.2d 77, 79-80 (Vt. 1991) (same).

      84 Pickering v. Board of Educ., 391 U.S. 563, 568 (1968).

      85 See, e.g., Baz v. Walters, 782 F.2d 701, 708 (7th Cir. 1986); Philadelphia Lodge No. 5 v. City of Philadelphia, 599 F. Supp. 254, 258 (E.D. Penn. 1984) (seeming to apply an even more deferential standard than Pickering); Doherty v. Wilson, 356 F. Supp. 35 (M.D. Ga. 1973) (adopting a Pickering-like approach); Barlow v. Blackburn, 798 P.2d 1360, 1366 (Ariz. Ct. App. 1990) (seeming at times to apply Pickering and at other times to apply strict scrutiny).

      86 Compare Mozert v. Hawkins County Bd. of Educ., 827 F.2d 1058, 1073 (6th Cir. 1987) (Boggs, J., concurring in the judgment) (arguing that when parents object that a school's curriculum burdens their children's religious beliefs by exposing them to material to which their religion forbids exposure, the standard of review should be even more deferential than under Tinker v. Des Moines Indep. Community Sch. Dist., 393 U.S. 503 (1969)), with, e.g., Spence v. Bailey, 465 F.2d 797 (6th Cir. 1972) (applying strict scrutiny in a similar situation); Moody v. Cronin, 484 F. Supp. 270 (C.D. Ill. 1979) (same).

      87 452 U.S. 640 (1981).

      88 See id. at 654; see also Geoffrey R. Stone, Constitutionally Compelled Exemptions and the Free Exercise Clause, 27 Wm. & Mary L. Rev. 985, 994-96 (1986) (discussing "the special embarrassment that exists when free speech and free exercise claims coalesce").

      89 See Grosz v. City of Miami Beach, 721 F.2d 729, 733 (11th Cir. 1983), vacated and remanded, 82 F.3d 1005 (11th Cir. 1996) (holding that the claim may be brought again under RFRA's tougher standards); Congregation Beth Yitzchok of Rockland, Inc. v. Town of Ramapo, 593 F. Supp. 655, 659-60 (S.D.N.Y. 1984) (following Grosz).

      90 Some courts have concluded that the Free Exercise Clause protects clergy-hiring decisions even after Employment Div. v. Smith, 494 U.S. 872 (1990), which rejected strict scrutiny of generally applicable laws under the Free Exercise Clause. See, e.g., EEOC v. Catholic Univ. of Am., 83 F.3d 455, 462 (D.C. Cir. 1996); Young v. Northern Ill. Conference of United Methodist Church, 21 F.3d 184, 185 (7th Cir. 1994); Scharon v. St. Luke's Episcopal Presbyterian Hosps., 929 F.2d 360, 363 (8th Cir. 1991); Van Osdol v. Vogt, 908 P.2d 1122, 1127 (Colo. 1996); Porth v. Roman Catholic Diocese, 532 N.W.2d 195, 200 (Mich. Ct. App. 1995); Geraci v. Eckankar, 526 N.W.2d 391, 401 (Minn. Ct. App. 1995) (shifting somewhat confusingly among the Free Exercise Clause, RFRA, and the state religious freedom guarantee); Jocz v. Labor & Indus. Review Comm'n, 538 N.W.2d 588, 596 n.13 (Wis. Ct. App. 1995); see alsoSmith, 494 U.S. at 882 (suggesting that free exercise claims might survive if linked with freedom of association claims). Other courts have held that clergy-hiring decisions are protected by RFRA.  See, e.g., Powell v. Stafford, 859 F. Supp. 1343, 1347 (D. Colo. 1994); see also Rayburn v. General Conference of Seventh-Day Adventists, 772 F.2d 1164, 1169 (4th Cir. 1985) (protecting clergy-hiring decisions under the pre-Smith constitutional exemption regime); cf. NLRB v. Catholic Bishop of Chicago, 440 U.S. 490, 506-07 (1979) (suggesting that clergy-hiring decisions are protected by the Free Exercise Clause under the pre-Smith constitutional exemption regime). The Establishment Clause might also protect churches in this context. See Rayburn, 772 F.2d at 1169 ("To subject church employment decisions . . . to Title VII scrutiny would also give rise to 'excessive government entanglement' with religious institutions prohibited by the establishment clause . . . .").

      91 The Court has held that the interests in stopping race discrimination in education and public accommodations are compelling. See Roberts v. United States Jaycees, 468 U.S. 609, 625 (1984); Bob Jones Univ. v. United States, 461 U.S. 574, 595 (1983). Lower courts agree that the same is true in employment, even employment by church institutions. See, e.g., EEOC v. Fremont Christian Sch., 781 F.2d 1362, 1368 (9th Cir. 1986); McLeod v. Providence Christian Sch., 408 N.W.2d 146, 151 (Mich. Ct. App. 1987); Minnesota ex rel. McClure v. Sports & Health Club, Inc., 370 N.W.2d 844, 852 (Minn. 1985). What could be more narrowly tailored to these interests than a prohibition on such discrimination, even applied to the selection of clergy? Cf.Roberts, 468 U.S. at 628-29 (concluding that in prohibiting discriminatory practices "in the distribution of publicly available goods, services, and other advantages," the state antidiscrimination law "'responds precisely to the substantive problem which legitimately concerns' the State and abridges no more speech or associational freedom than is necessary to accomplish that purpose") (citation omitted).

      92 See, e.g., Roberts, 468 U.S. at 620-21.

      93 See Rodriguez v. City of Chicago, 975 F. Supp. 1055 (N.D. Ill. 1997) (rejecting such a claim, without reaching a strict scrutiny analysis), aff'd, 156 F.3d 771 (7th Cir. 1998).

      94 See, e.g., City of Richmond v. J.A. Croson Co., 488 U.S. 469, 508 (1989) (applying the "compelling interest" test in the race classification context); Carey v. Population Servs. Int'l, 431 U.S. 678, 691 (1977) (applying the same test in the right to privacy context).

      95 See, e.g., Waters v. Churchill, 511 U.S. 661, 672 (1994) (plurality opinion) (stressing that the substantive First Amendment rules must be different when the government acts as employer).

      96 Cf. Cheema v. Thompson, 67 F.3d 883, 886 (9th Cir. 1995) (holding that Sikh children had the right under RFRA to wear ceremonial daggers to school).

      97 Public primary and secondary education does have an important component of the government acting as sovereign as well as of the government acting as regulator of its own property: The government doesn't just offer students access to government-run schools, but also compels parents to send children to some school, which for poor children whose parents can't afford private education translates into a legal compulsion to attend public school. (I thank Doug Laycock for forcefully reminding me of this in an e-mail responding to an earlier draft of this Article.)

      Nonetheless, for the reasons I allude to in the text, the government must have considerably more latitude in controlling student behavior in such schools than it does in controlling private behavior outside government property. Strict scrutiny might be a sound legislative solution if it is clear that under strict scrutiny the government need only exempt the children of objecting parents from compulsory education requirements (either entirely or as to individual classes); but it is not a sound solution if, as in Cheema, it is interpreted to require the government to accommodate objections within the public school itself.

      98 Cf. Lumpkin v. Brown, 109 F.3d 1498, 1501 (9th Cir. 1997) (holding that the mayor of San Francisco could, despite RFRA, remove a minister from the city Human Rights Commission based on the minister's public religious condemnation of homosexuality, because the removal "served a compelling governmental interest--the preservation of the integrity of [the city's] antidiscrimination policies").

      99 See infra Part III.A.1.

      100 See Eugene Volokh, Freedom of Speech, Permissible Tailoring and Transcending Strict Scrutiny, 144 U. Pa. L. Rev. 2417, 2432-36 (1996).

      101 See id. at 2436.

      102 See id. at 2437-38; Eugene Volokh, Freedom of Speech, Shielding Children, and Transcending Balancing, 1997 Sup. Ct. Rev. 141, 141-42.

      103 See Bob Jones Univ. v. United States, 461 U.S. 574, 603-04 (1983); Gillette v. United States, 401 U.S. 437, 460 (1971); EEOC v. Fremont Christian Sch., 781 F.2d 1362, 1368-69 (9th Cir. 1986); McLeod v. Providence Christian Sch., 408 N.W.2d 146, 151-52 (Mich. Ct. App. 1987); Minnesota ex rel. McClure v. Sports & Health Club, Inc., 370 N.W.2d 844, 852-53 (Minn. 1985).

      104 See Volokh, supra note 100, at 2434.

      105 In fact, as I have argued elsewhere, id. at 2444-52, the strict scrutiny regime in Free Speech Clause cases and racial equal protection cases is actually considerably more rights-protective than strict scrutiny as it is normally defined; it protects speech even when suppressing the speech would be the least restrictive means of serving a compelling state interest. But even if "strict scrutiny" inaccurately describes what goes on in such cases, the tests used in these cases are commonly given this label, and using strict scrutiny language in religious exemption cases will necessarily suggest an equivalence with the level of protection available in those areas.

      106 See, e.g., Eisgruber & Sager, Vulnerability of Conscience, supra note 35, at 1247 (calling strict scrutiny "strict in theory but feeble in fact"); Ira C. Lupu, The Trouble with Accommodation, 60 Geo. Wash. L. Rev. 743, 756 (1992) (calling it "strict in theory, but ever-so-gentle in fact"); Michael W. McConnell, Free Exercise Revisionism and the Smith Decision, 57 U. Chi. L. Rev. 1109, 1127 (1990) (saying that "[t]he 'compelling interest' standard is a misnomer" because the actual test the Court has applied is more lenient); James E. Ryan, Note, Smith and the Religious Freedom Restoration Act: An Iconoclastic Assessment, 78 Va. L. Rev. 1407, 1413-37 (1992) (examining the "rise and fall of the compelling interest test" in Free Exercise Clause cases); see also Berg, supra note 33, at 10, 32 (acknowledging that "[s]ome [Supreme Court Free Exercise Clause] cases had applied the 'compelling interest' test but only half-heartedly," and concluding that the test "seems to promise more than it can realistically deliver," but ultimately still defending it); Robert D. Kamenshine, Scrapping Strict Review in Free Exercise Cases, 4 Const. Commentary 147, 154 (1987); Marshall, supra note 77, at 369; Stone, supra note 88, at 994.

      107 See Employment Div. v. Smith, 494 U.S. 872, 888 (1990) (making this point); Eisgruber & Sager, Why RFRA Is Unconstitutional, supra note 35 (same); cf. Ira C. Lupu, The Failure of RFRA, 20 U. Ark. Little Rock L.J. 575, 597-98 (1998) (expressing concern that having the same strict scrutiny test in prison cases and nonprison cases may water down strict scrutiny in the latter).

      108 See Robinson v. Jacksonville Shipyards, Inc., 760 F. Supp. 1486, 1536 (M.D. Fla. 1991) (concluding that sexually offensive speech in the workplace may be punished, in part because even "if the speech at issue is treated as fully protected, . . . [o]ther first amendment rights, such as the freedom of association and the free exercise of religion, have bowed to narrowly tailored remedies designed to advance the compelling governmental interest in eradicating employment discrimination"); McClure, 370 N.W.2d at 849-50 (holding that punishing health club owners for offensive religious speech doesn't violate the Free Speech Clause, and citing Bob Jones University in support).

      109 Kahn v. United States, 753 F.2d 1208, 1217 (3d Cir. 1985).

      110 See United States v. Lee, 455 U.S. 252, 260 (1982).

      111 See, e.g., Kahn, 753 F.2d at 1217; Todd v. United States, 613 F. Supp. 552, 555-56 (D. Mont. 1985).

      112 Cf. Tushnet, supra note 33, at 379 ("A 'doctrine' that is not enforced is not a happy thing for the Court to have. It generates litigation, as people present claims for exemption based on the purported 'doctrine' of Sherbert, only to find that those claims are routinely denied.").

      113 Cf. Ira C. Lupu, Statutes Revolving in Constitutional Law Orbits, 79 Va. L. Rev. 1 (1993).

      One negative consequence of the availability of judge-made constitutional terminology is the repression of debate and careful consideration of alternatives. The dynamics of this phenomenon are not difficult to understand. Whether a legislature is focused on "restoring" some doctrine, extending it to a new setting, or building upon constitutional concepts in a related field, precrafted judicial language may arrive with an aura of respectability, precise reflection, continuity of thought, and professional recognition. Moreover, questioning the language closely may expose new controversy or bring to light conflict buried in the existing formulation. Thus, a rush to adopt prefabricated constitutional concepts may undermine the processes of debate and deliberation and result in drafting choices being made too quickly and without reflection.

      Id. at 23; see also id. at 53-66.

      114 See, e.g., Parts I.D, III.

      115 See, e.g., Fields Jeep-Eagle, Inc. v. Chrysler Corp., 645 N.E.2d 946, 953-54 (Ill. 1994); City of Okla. City v. State, 918 P.2d 26, 29 (Okla. 1995).

      116 See cases cited supra note 115.

      117 On reading this section, Chip Lupu suggested that it might be a mistake to enact a "substantial burden" threshold, and that it may be better to leave courts with more room to innovate in this area. See Ira C. Lupu, Where Rights Begin: The Problem of Burdens on the Free Exercise of Religion, 102 Harv. L. Rev. 933, 960-89 (1989) (outlining a proposed alternative to the existing substantial burden jurisprudence). I tentatively think that the substantial burden threshold is fairly sensible, though it has at times been applied incorrectly, see, e.g., infra note 195. Nonetheless, nothing in my broader proposal turns on my endorsement of the substantial burden threshold; if Lupu is right, then his alternative framework could certainly be substituted into my proposed statute.

      Note also that this proposal doesn't take a stand on whether exemptions should be granted only for behavior that is religiously or conscientiously compelled, or for behavior that is religiously or conscientiously motivated. See sources cited infra note 143.

      118 Fed. R. Evid. 501. The phrase also appears in the statutes cited supra note 26.

      119 See, e.g., Jaffee v. Redmond, 518 U.S. 1, 12-15 (1995) (referring to state legislative adoption of a psychotherapist-patient privilege in determining whether federal courts should adopt it under Rule 501). But see id. at 26, 30-31 (Scalia, J., dissenting) (arguing that common-law decision making should ignore statutes enacted in other jurisdictions, because such statutes are more likely to be the product of interest-group power than of common-law "reason and experience").

      120 398 U.S. 333 (1970).

      121 See id. at 339-40.

      122 See, e.g., Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 532-33 (1993); McDaniel v. Paty, 435 U.S. 618, 626 (1978); Fowler v. Rhode Island, 345 U.S. 67, 69-70 (1953); Columbia Union College v. Clarke, 159 F.3d 151 (4th Cir. 1998); Peter v. Wedl, 155 F.3d 992, 996-97 (8th Cir. 1998); Hartmann v. Stone, 68 F.3d 973, 979 (6th Cir. 1995).

      123 See United States v. Ballard, 322 U.S. 78, 86-87 (1944).

      124 See, e.g., Dausch v. Rykse, 52 F.3d 1425, 1429 (7th Cir. 1994) (clergy malpractice); In re Pleasant Glade Assembly of God, No. 2-98-222-CV, 1998 WL 745954, at *2-*4 (Tex. App. Oct. 22, 1998) (same); Doe v. Evans, 718 So. 2d 286, 291-94 (Fla. Dist. Ct. App. 1998) (breach of fiduciary duty); Amato v. Greenquist, 679 N.E.2d 446, 453-54 (Ill. App. Ct. 1997) (same); H.R.B. v. J.L.G., 913 S.W.2d 92, 98-99 (Mo. Ct. App. 1995) (same); Langford v. Roman Catholic Diocese, 677 N.Y.S.2d 436, 438 (N.Y. Sup. Ct. 1998) (same); L.C. v. R.P., 563 N.W.2d 799, 800-03 (N.D. 1997) (same). But see Moses v. Diocese of Colorado, 863 P.2d 310, 314, 321-23 (Colo. 1993) (accepting breach of fiduciary duty claim); F.G. v. MacDonel, 696 A.2d 697, 703-05 (N.J. 1997) (same); Martinelli v. Bridgeport Roman Catholic Diocesan Corp., 10 F. Supp. 2d 138, 149-57 (D. Conn. 1998) (same); Sanders v. Casa View Baptist Church, 898 F. Supp. 1169, 1176-77 (N.D. Tex. 1995) (same). These claims generally arise out of sexual relationships between a clergyman and a parishioner whom he is counseling; in secular counseling contexts, similar breach of fiduciary duty claims are fairly widely accepted because such relationships are seen as harmful and not fully consensual.

      125 See cases cited supra note 90 (holding that the Free Exercise Clause bans the application of employment discrimination law to clergy); NLRB v. Catholic Bishop of Chicago, 440 U.S. 490, 504-07 (1979) (suggesting likewise as to labor law). Compare Swanson v. Roman Catholic Bishop, 692 A.2d 441, 445 (Me. 1997) (holding likewise as to negligent supervision law), and L.L.N. v. Clauder, 563 N.W.2d 434, 447 (Wis. 1997) (same), with Moses, 863 P.2d at 323-29 (rejecting a Free Exercise Clause defense to a negligent hiring and supervision claim), and Kenneth R. v. Roman Catholic Diocese, 654 N.Y.S.2d 791, 795-69 (App. Div. 1997) (rejecting the same defense to a negligent supervision and retention claim). Cf. also, e.g., Higgins v. Maher, 258 Cal. Rptr. 757, 761-62 (Ct. App. 1989) (holding that the Free Exercise Clause bars tort claims based on the disciplining of a church official); Tidman v. Salvation Army, No. 01-A-01-9708-CV00380, 1998 WL 391765, at *5-*7 (Tenn. Ct. App. July 15, 1998) (likewise). See generally Douglas Laycock, Towards a General Theory of the Religion Clauses: The Case of Church Labor Relations and the Right to Church Autonomy, 81 Colum. L. Rev. 1373 (1981); Ira C. Lupu, Free Exercise Exemption and Religious Institutions: The Case of Employment Discrimination, 67 B.U. L. Rev. 391 (1987).

      126 Cf., e.g., Marks v. Estate of Hartgerink, 528 N.W.2d 539, 544-45 (Iowa 1995) (wrongful excommunication); Korean Presbyterian Church v. Lee, 880 P.2d 565, 569-70 (Wash. Ct. App. 1994) (same).

      Under Roberts v. United States Jaycees, 468 U.S. 609, 621 (1984), large and unselective organizations seem to lack a general freedom of association right to exclude others based on race, sex, and the like. The right of churches to impose such exclusions--and I've heard no arguments denying the existence of such a right--would thus probably be a special protection provided by the Free Exercise Clause. Cf. Charles Sumner's Civil Rights Bill, Cong. Globe, 42d Cong., 2d Sess. 3274 (1872) (purporting to ban race discrimination "by trustees and officers of church organizations"), reprinted in Kurt T. Lash, The Second Adoption of the Free Exercise Clause: Religious Exemptions Under the Fourteenth Amendment, 88 Nw. U. L. Rev. 1106, 1154 & n.213 (1994).

      127 See, e.g., Rayburn v. General Conference of Seventh-Day Adventists, 772 F.2d 1164, 1167-68 (4th Cir. 1985); cases cited supra note 124.

      128 See United States v. Ballard, 322 U.S. 78, 86-88 (1944).

      129 See supra notes 90-91 and accompanying text.

      130 See, e.g., Ballard, 322 U.S. at 85-88; see also Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 542-43 (1993) (preventing the government from treating religiously motivated conduct worse than identical secularly motivated conduct); McDaniel v. Paty, 435 U.S. 618, 629 (1978) (same).

      131 See supra note 106.

      132 Note the importance of the phrase "through the communicative impact": Free speech jurisprudence offers far less protection to speech that harms through its noncommunicative impact; for instance, interfering with another's business by passing out leaflets urging a political boycott is protected speech, while interfering with the business by making very loud noises outside it is not.

      133 See, e.g., NAACP v. Claiborne Hardware Co., 458 U.S. 886, 912-15 (1982).

      134 See, e.g., Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 50-57 (1988).

      135 Cf. Eastern R.R. Presidents Conference v. Noerr Motor Freight, Inc., 365 U.S. 127, 136-38 (1961) (holding that a publicity campaign aimed at producing government action adverse to another's interests may not be the basis of antitrust liability).

      136 See Davidson v. Time Warner, Inc., No. Civ. A. V-94-006, 1997 WL 405907, at *15 (S.D. Tex. Mar. 31, 1997) (concluding that even if such a claim is viable under standard negligence law, it is prohibited by the First Amendment).

      137 Cf. Braham v. State, 571 P.2d 631, 644-45 (Alaska 1977) (discussing the applicability of the Compulsory Process Clause to subpoenas duces tecum, and citing cases).

      138 See infra note 160.

      139 See infra note 168; Eugene Volokh, "A Right to Keep and Bear Arms in Defense of Himself" (in progress). Forty-four state constitutions secure a right to keep and bear arms; most of them unambiguously specify an individual right, and in my view all of them must be read that way--rights in a state constitution must be protections against state government action, not protections for a state-run body. See, e.g., N.H. Const. pt. 1, art. 2-a (enacted 1982) ("All persons have the right to keep and bear arms in defense of themselves, their families, their property and the state."); Dano v. Collins, 802 P.2d 1021, 1022-23 (Ariz. Ct. App. 1990) (suggesting that the Arizona Constitution protects the right to carry guns openly, though not concealed); Kellogg v. City of Gary, 562 N.E.2d 685, 699 (Ind. 1990) (suggesting that a license to carry a handgun is guaranteed to law-abiding citizens by the state constitution's recognition of a right to keep and bear arms); State v. Rosenthal, 55 A. 610, 611 (Vt. 1903) (holding that the Vermont Constitution prohibits a ban on carrying concealed weapons). Vermont is today the one state where a person may carry a concealed weapon even without a license.

      140 See, e.g., United States v. Ballard, 322 U.S. 78, 86-87 (1944) (upholding the right to make false but sincere religious claims).

      141 See, e.g., recent cases cited supra notes 122 (right to be free from discrimination based on religion or religiosity), 125 (right to discriminate in choice of clergy), 126 (right to select or reject members).

      142 See, e.g., Rayburn v. General Conference of Seventh-Day Adventists, 772 F.2d 1164, 1167-68 (4th Cir. 1985) (protecting a church's right to discriminate in choosing clergy even in the absence of a claim--usually required in Sherbert-era Free Exercise Clause cases--that there is a religious compulsion or motivation for the discrimination); text accompanying notes 90-92 (arguing that a church's right to discriminate in choosing clergy is protected by a per se rule, which provides more protection than does the strict scrutiny analysis usually required in Sherbert-era Free Exercise Clause cases). The same is true of the right secured by United States v. Ballard.

      143 See Mack v. O'Leary, 80 F.3d 1175, 1178 (7th Cir. 1996) (Posner, J.) (discussing the dispute and citing cases); Laycock, supra note 43, at 24-27 (same); Lupu, supra note 78, at 806 (same); Laycock & Thomas, supra note 52, at 231-34 (discussing the dispute); Steven C. Seeger, Note, Restoring Rights to Rites: The Religious Motivation Test and the Religious Freedom Restoration Act, 95 Mich. L. Rev. 1472 (1997).

      144 See, e.g., American Life League, Inc. v. Reno, 47 F.3d 642 (4th Cir. 1995) (involving a claimed religious freedom right to block the entrance to another's property); Urantia Found. v. Maaherra, 895 F. Supp. 1329 (D. Ariz. 1995) (involving a claimed religious freedom right to infringe another's copyright); Urantia Found. v. Maaherra, 895 F. Supp. 1335 (D. Ariz. 1995) (same); Bridge Publications, Inc. v. Vien, 827 F. Supp. 629 (S.D. Cal. 1993) (same); Park v. Schlitgen, No. C 97-1813SI, 1999 WL 138887 (N.D. Cal. Mar. 9, 1999) (discussing a religious freedom challenge to an involuntary manslaughter charge based on a death that happened in a prayer session); People v. Jones, 697 N.E.2d 457, 460 (Ill. App. Ct. 1998) (discussing a claimed religious freedom right to beat one's wife); Hester v. Barnett, 723 S.W.2d 544 (Mo. App. 1987) (involving a claimed religious freedom right to defame). All these claims were rejected by the courts.

      145 Cf. infra Part III.E.2 (explaining why the Constitution itself can't be read as imposing the ranking).

      146 See Clark v. Community for Creative Non-Violence, 468 U.S. 288, 293-94 (1984); United States v. O'Brien, 391 U.S. 367, 376-77 (1968); Brown v. Louisiana, 383 U.S. 131, 141-43 (1966) (plurality opinion); Garner v. Louisiana, 368 U.S. 157, 199-203 (1961) (Harlan, J., concurring).

      147 See, e.g., O'Brien, 391 U.S. at 376 ("We cannot accept the view that an apparently limitless variety of conduct can be labeled 'speech' whenever the person engaging in the conduct intends thereby to express an idea."); Wisconsin v. Mitchell, 508 U.S. 476, 484 (1993) (endorsing this); Texas v. Johnson, 491 U.S. 397, 404 (1989) (same) (citing O'Brien); Clark, 468 U.S. at 293 (same); Spence v. Washington, 418 U.S. 405, 409 (1974) (same). I believe this position is correct, though I'm uncertain about what its exact scope should be.

      To be protected as expressive conduct, conduct must at least be generally perceived by most listeners as expressive, Spence, 418 U.S at 411, another constraint that the proposed religious exemption regime wouldn't have.

      148 See, e.g., Kent Greenawalt, Free Speech in the United States and Canada, 55 Law & Contemp. Probs. 5, 27 (1992) ("[T]he O'Brien test may already be so weak that its formal abandonment is unnecessary."); Laycock, supra note 43, at 20 (describing both the expressive conduct test and the time, place, or manner restriction test as "deferential"); Lupu, supra note 107, at 592-93 ("[J]udges are uncomfortable with any claim of exemption from general law. . . . The symbolic speech cases, like U.S. v. O'Brien, are only rhetorically supportive of pro-exemption arguments; when claims of free speech privilege are pressed against speech-neutral, generally applicable laws, speech loses.") (footnotes omitted); Susan H. Williams, Content Discrimination and the First Amendment, 139 U. Pa. L. Rev. 615, 654 (1991) (describing the expressive conduct test as "really just a weak version of the [time, place, or manner restriction] test").

      149 See, e.g., Members of the City Council v. Taxpayers for Vincent, 466 U.S. 789, 804-05 (1984) (concluding that under the O'Brien test, aesthetic interests were sufficient to justify the challenged restriction); Rappa v. New Castle County, 18 F.3d 1043, 1066-67 (3d Cir. 1994) (same); Outdoor Media Dimensions Inc. v. State, 945 P.2d 614, 620 (Or. Ct. App. 1997) (same).

      150 Thomas Jefferson, Notes on the State of Virginia 159 (William Peden ed., 1955).

      151 Letter from James Madison to Edward Livingston (July 10, 1822), in 9 The Writings of James Madison 100 (Gaillard Hunt ed., 1901).

      152 McConnell, supra note 106, at 1128; see also Berg, supra note 106, at 35 (arguing that "government's role in matters of religion is limited to preventing harm to others"); Stephen Pepper, Taking the Free Exercise Clause Seriously, 1986 BYU L. Rev. 299, 333 (1986) ("The free exercise clause is written to limit governmental conduct and to give shelter from governmental impingement; it is not a license to impinge on private third parties. For example, if your religious shrine is in my backyard, your free exercise right does not trump my property right and allow you to trespass."). Pepper and Berg believe that these countervailing private rights and harms to others are quite rare; see infra note 154 for my explanation of why I think they are quite common, and in fact present in most of the Court's religious exemption cases.

      153 When the government is acting as landlord or employer, its interests may often be less directly connected to the interests of particular people; but in such cases, traditional Free Exercise Clause jurisprudence, like Free Speech Clause jurisprudence, has in any event mandated much more deferential tests than strict scrutiny. See supra Part II.A. For this reason, the discussion in this part focuses on the government acting as sovereign.

      154 Consider for example some of the Court's main Free Exercise Clause cases (limited for convenience to the government acting as sovereign, in situations where a religious exemption was sought and there was little doubt that the neutral law imposed a substantial burden on free exercise). Bob Jones University v. United States, 461 U.S. 574, 577-79 (1983), involved antidiscrimination law, which is generally seen as securing a private right to be treated equally in certain transactions. Wisconsin v. Yoder, 406 U.S. 205, 207-13 (1972), and Prince v. Massachusetts, 321 U.S. 158, 160-61 (1944), involved compulsory-education laws and child-welfare laws, which are usually defended as means to protect a child's unwaivable right to be educated and not to be "exploited." See Yoder, 406 U.S. at 244-45 (Douglas, J., dissenting) (arguing that children who may wish to "break from the Amish tradition" have a "right . . . to be masters of their own destiny," for which a normal high school education may be necessary); Prince, 321 U.S. at 170 (saying that parents are not free "to make martyrs of their children before they have reached the age of full and legal discretion when they can make that choice for themselves"). United States v. Ballard, 322 U.S. 78, 79 (1944), though not quite an exemption case, involved a prosecution for getting money under false pretenses, conduct that is generally seen as infringing another's rights. NLRB v. Catholic Bishop of Chicago, 440 U.S. 490, 492-95 (1979), decided on statutory grounds but based on a constitutional judgment, involved what many see as the right of union members to bargain collectively. See, e.g., National Labor Relations Act, 29 U.S.C. §§ 157-158 (1994) (specifically referring to this as a "right," enforceable against private abridgement).

      United States v. Lee, 455 U.S. 252, 254 (1982), and Gillette v. United States, 401 U.S. 437, 439 (1971), involved not the rights of others but improper externalities on others: The Lees of the world, by not paying taxes, and the Gillettes, by not going to war, increase the burden on other taxpayers and on other draftees; in fact, tax avoiders are often condemned precisely for increasing the burden on others who have to take up the slack. See, e.g., Texas Monthly, Inc. v. Bullock, 489 U.S. 1, 14 (1989) (plurality opinion) ("Every tax exemption constitutes a subsidy that affects nonqualifying taxpayers, forcing them to become indirect and vicarious donors."); Bob Jones Univ. v. United States, 461 U.S. 574, 591 (1983) (in the Free Exercise Clause context) ("When the Government grants exemptions or allows deductions all taxpayers are affected; the very fact of the exemption or deduction for the donor means that other taxpayers can be said to be indirect and vicarious 'donors.'"); Harry G. Balter, Fraud Under Federal Tax Law 11 (1951) ("[H]e who does not pay his legally due share of taxes, is automatically shifting his share, and thus adding to the already heavy burden of the honest taxpayer."); articles cited infra note 163 (making a similar point about the draft). Likewise for Employment Division v. Smith, 494 U.S. 872 (1990), and Reynolds v. United States, 98 U.S. 145 (1878): Drug use and polygamy are generally condemned in large part because of the concern, well-founded or not, about the harm these practices supposedly cause others, whether third parties or co-religionists who are socially or economically pressured into the practice. See, e.g., Reynolds, 98 U.S. at 166, 167-68 (arguing that polygamy tends to "lead[] to the patriarchal principle, . . . which, when applied to large communities, fetters the people in stationary despotism," and that polygamy causes "evil consequences" to "innocent victims" such as "pure-minded women and . . . innocent children").

      Finally, the unemployment compensation claimed in Sherbert v. Verner, 374 U.S. 398 (1963), and in the three other unemployment cases came indirectly out of the employer's pocket: Unemployment compensation is generally experience-rated, so an employer's unemployment tax payments are tied to the number of claims the employer has had to pay out. Sherbert's win equaled a loss of property for her employer. Cf. Eisgruber & Sager, Why RFRA Is Unconstitutional, supra note 35, at 454. Pepper contends that "[i]n Sherbert, Yoder, Lee, [and] Bob Jones, . . . granting the claimed exemption does no direct harm to any identifiable third party." Pepper, supra note 152, at 333; see also Berg, supra note 106, at 38-39. But whether or not the harms are direct, they are quite real.

      Of course, these cases were generally argued as compelling interest cases, not countervailing private concern cases; as I mention in the text, all claims of countervailing private concern can be expressed as claims of compelling government interest in defending such concerns. My point is simply that the supporters of the general law in each of these cases would almost certainly view the law as in large part justified by the interests of particular people, and not just "government interests."

      155 Cf. Stephen E. Gottlieb, Compelling Governmental Interests and Constitutional Discourse, 55 Alb. L. Rev. 549, 552 (1992) ("The term compelling governmental interests is a misnomer. The interests themselves can hardly be governmental; they must be public.").

      A few laws involve people's attempt to defend the felt rights and interests of entities other than people; this may be true of laws that ban animal cruelty, or that protect the environment even in situations where no human has a medical, economic, or even aesthetic interest in such protection. It may also be in some measure true of laws that aim to protect fetuses (within the scope permitted by the Court's right to privacy jurisprudence), and of laws that aim to protect certain interests of the recently deceased, for instance the right to dispose of property pursuant to a will; these laws, though, may generally also be seen as attempts to prevent harm to living people.

      I don't believe these laws pose any special difficulties for my analysis, in part because I think the legislature's power to define who or what is entitled to protection from harm is comparable to the legislature's power to define what harm is. Cf. Michael C. Dorf, Truth, Justice, and the American Constitution, 97 Colum. L. Rev. 133, 156-58 (1997). Still, even if I'm wrong, and these cases must be treated differently, they involve only a small subset of the controversial religious freedom claims.

      156 See, e.g., Bob Jones Univ., 461 U.S. at 602-04.

      157 Compare West Coast Hotel Co. v. Parrish, 300 U.S. 379, 398-99 (1937) (accepting a legislative judgment that employees may be "the ready victims of those who would take advantage of their necessitous circumstances" and "exploit[] . . . workers at wages so low as to be insufficient to meet the bare cost of living, thus making their very helplessness the occasion of a most injurious competition"), withid. at 408 (Sutherland, J., dissenting) (rejecting "the assumption that the employee is entitled to receive a sum of money sufficient to provide a living for her"); cf. Adkins v. Children's Hosp., 261 U.S. 525 (1923) (involving the same debate, but with the opposite result).

      158 Compare Coppage v. Kansas, 236 U.S. 1, 19 (1915) (striking down a law that banned employers from demanding that their employees not join a union, on the grounds that a union member has "no inherent right to [join a union] and still remain in the employ of one who is unwilling to employ a union man"), with id. at 32-33 (Day, J., dissenting) (contending that the legislature was empowered to protect "the legal right" to join a union "by preventing an employer from depriving the employee of [the right] as a condition of obtaining employment" because there is "no reason why a State may not, if it chooses, protect this right, as well as other legal rights"); compare also, e.g., Swanner v. Anchorage Equal Rights Comm'n, 874 P.2d 274, 284 (Alaska 1994) ("Because Swanner's religiously impelled actions trespass on the private right of unmarried couples to not be unfairly discriminated against in housing, he cannot be granted an exemption from the housing anti-discrimination laws."), with id. at 287 (Moore, C.J., dissenting) ("I am not willing to place the right to cohabitate on the same constitutional level as the right to freedom from discrimination based on either innate characteristics--such as race or gender--or constitutionally protected belief, such as freedom of religion.").

      159 Compare Yoder, 406 U.S. at 222 (rejecting the claim that a compulsory education law was required "to protect children from ignorance," on the grounds that the Amish community "has been a highly successful social unit" with "productive and very law-abiding" members, and that it educates its children, albeit in an unconventional way), with id. at 244-45 (Douglas, J., dissenting) (arguing that children who may wish to "break from the Amish tradition" have a "right . . . to be masters of their own destiny," for which a high school education may be essential).

      160 Compare In re The Grand Jury Empaneling of the Special Grand Jury, No. 98-6415, 1999 WL 150880 (3d Cir. Mar. 19, 1999) (holding that the Free Exercise Clause didn't allow a religiously motivated refusal to testify against a family member, at least in this case), and In re Doe, 842 F.2d 244, 245-48 (10th Cir. 1988) (same), with In re Greenberg, 11 Fed. R. Evid. Serv. 579 (D. Conn. 1982) (holding the opposite), and In re The Grand Jury Empaneling, 1999 WL 150880, at *12 (McKee, J., dissenting) (same). Cf. Grossberg's Parents Ask to Keep Talks Confidential, Newark Star-Ledger, Nov. 26, 1997, at 43.

      The parents of Amy Grossberg, the college student accused of killing her newborn in Delaware . . . argued in court papers that talks with their daughter should be kept secret and that it is a violation of their right to the free exercise of religion [for prosecutors] to force them to divulge information. Rabbi Joel Roth, a legal expert at the Jewish Theological Seminary in New York City, confirmed yesterday he wrote an affidavit for the Grossbergs, stating that "under Jewish law, a mother and/or a father are not allowed to give testimony against their child in any legal proceeding."

      Id. These cases involved government attempts to compel testimony in criminal investigations, but similar exception claims could have equally easily arisen in a civil suit. Cf. Sharma v. Zollar, 638 N.E.2d 736, 742 (Ill. App. Ct. 1994) (discussing "[litigants'] rights of compulsory process and confrontation of witnesses" in noncriminal, administrative cases where the litigant's property is on the line).

      161 See, e.g., Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 394-95 (1926) (stressing neighbors' interest in being free from "street accidents," "noise," and other interferences that "come very near to being nuisances"); Grosz v. City of Miami Beach, 721 F.2d 729, 738 (11th Cir. 1983) (stressing neighbors' interest in being free from "crowds, noise and disturbance"), vacated and remanded, 82 F.3d 1005 (11th Cir. 1996); Laurie Reynolds, Zoning the Church: The Police Power Versus the First Amendment, 64 B.U. L. Rev. 767, 774, 794, 819 (1985) (discussing "negative externalities" that zoning law tries to prevent, such as "traffic hazards, congestion, noise, [and] unacceptable density"). Some courts have held that religious freedom principles trump zoning regulations. See, e.g., Stuart Circle Parish v. Board of Zoning Appeals, 946 F. Supp. 1225, 1237-40 (E.D. Va. 1996); Western Presbyterian Church v. Board of Zoning Adjustment, 849 F. Supp. 77, 79 (D.D.C. 1994); Jesus Ctr. v. Farmington Hills Zoning Bd. of Appeals, 544 N.W.2d 698, 703-05 (Mich. Ct. App. 1996). Others have held that zoning regulations trump religious freedom claims. See, e.g., Daytona Rescue Mission, Inc. v. City of Daytona Beach, 885 F. Supp. 1554, 1560 (M.D. Fla. 1995).

      162 See, e.g., Costello Publ'g Co. v. Rotelle, 670 F.2d 1035, 1048-50 (D.C. Cir. 1981) (discussing but not deciding a religious freedom defense to an antitrust claim); cf. Jay Burns Baking Co. v. Bryan, 264 U.S. 504, 517 (1924) (striking down a state statute mandating standard weights for loaves of bread, notwithstanding the dissent's claim that "[t]he purpose of the Nebraska standard-weight bread law is to protect buyers from short weights and honest bakers from unfair competition"); Danneskjold v. Hausrath, 82 F.3d 37, 40 (2d Cir. 1996) (explaining that "one of the primary purposes of the [minimum-wage law is] preventing unfair competition among employers if some were able to pay less than the minimum wage").

      163 See, e.g., Lisa Belkin, Quayle Fields Another Question About Admission to Law School, N.Y. Times, Sept. 10, 1988, § 1, at 8 ("Outside the plant about 150 protesters carried signs that said such things as 'Draft Dodger Quayle--Who Died in Your Place?'"); Michael Kenney, Author Writes of War as the "Ultimate Crisis," Austin Am.-Statesman, July 25, 1995, at E2 ("'I very much regret not having served in Vietnam,' [author and former draft dodger Mark Helprin] said in an interview. And the reason why, 'to cut to the chase, is that I am physically very able, but in not serving, someone may have died in my place.'"); Greg Trevor, Veterans Still Feel Left Out, Ashbury Park Press, Oct. 14, 1996, at A1 ("Clinton 'wouldn't bite the bullet. I wonder who died in his place,' said [Korean War] veteran Larry Schaeffer . . . .").

      164 Compare In re Young, 82 F.3d 1407, 1420 (8th Cir. 1996) (implying the answer is "no," by rejecting the notion that "the interests of creditors" justify rescission of such transfers), rev'd and remanded for reconsideration in light of City of Boerne v. Flores, 521 U.S. 507 (1997), reaffirmed on remand, 141 F.3d 854 (8th Cir. 1998), with In re Hodge, 200 B.R. 884, 907 (Bankr. D. Idaho 1996) (answering the question "yes," and stressing "the rights of Debtors' creditors to payment of their just claims"); Brief for Intervenor United States at 30, In re Young (No. 93-2267) (eventually withdrawn by the United States) (taking a similar view); Daniel Keating, Bankruptcy, Tithing, and the Pocket-Picking Paradigm of Free Exercise, 1996 U. Ill. L. Rev. 1041, 1044 (arguing that the Free Exercise Clause ought not be read "to allow what amounts to a form of picking a private third party's pocket"); Jonathan C. Lipson, First Principles and Fair Consideration: The Developing Clash Between the First Amendment and the Constructive Fraudulent Conveyance Laws, 52 U. Miami L. Rev. 247, 282 (1997) ("Every case in which a religious donation is challenged as a fraudulent conveyance involves a choice between the right of free exercise and a trespass on private rights."); Bruce W. Megard, Jr., Tithing and Fraudulent Transfers in Bankruptcy: Confirming a Trustee's Power to Avoid the Tithe After City of Boerne v. Flores, 71 Am. Bankr. L.J. 413, 427 (1997) ("Free Exercise rights should not be construed to force others to pay for the debtor's religious beliefs, and . . . should carry no greater weight than the legally cognizable rights of innocent third parties.").

      Cf. Hunt v. Hunt, 648 A.2d 843, 854 (Vt. 1994) (concluding that religious freedom principles sometimes bar use of contempt sanctions to force payment of child support).

      165 In particular, I don't think it's profitable to distinguish "positive" rights here from "negative" ones. The rights I discuss here are private rights--rights to government protection from certain kinds of harmful conduct by others--so in that sense they are positive rights (granted by statute or common law) to enlist the government's aid in restraining others' behavior. And while some of these rights are usually talked about as the right to have the government prevent certain private conduct and others are usually talked about as the right to have the government compel certain private conduct, this distinction generally breaks down on closer examination. The right to prevent breaches of contract is a right to compel performance of contracts; the right to compel others to rent their property to you without regard to your marital status is a right to prevent others from discriminating based on marital status in their decision about renting property to you.

      166 Cf. State v. Miller, 538 N.W.2d 573, 577-79 (Wis. Ct. App. 1995) (concluding that such a ban would serve a compelling interest, but striking down the requirement because the state hadn't disproven the effectiveness of alternatives proposed by the objectors).

      167 As to alcohol, compare Commonwealth v. Campbell, 117 S.W. 383, 385, 387 (Ky. 1909) (striking down an alcohol possession ban on the grounds that the legislature may not restrict "what a man will drink, or eat, or own, provided the rights of others are not invaded," and that private alcohol possession "can by no possibility injure or affect the health, morals, or safety of the public"); State v. Williams, 61 S.E. 61, 66-67 (N.C. 1908) (same); Ex parte Brown, 42 S.W. 554, 556-57 (Tex. Crim. App. 1897) (same); State v. Gilman, 10 S.E. 283, 284-85 (W. Va. 1889) (same); and Commonwealth v. Smith, 173 S.W. 340, 342-43 (Ky. 1915) (applying Campbell to strike down an alcohol possession ban even when the ban had an exception for possession in private residences), with, for example, Mugler v. Kansas, 123 U.S. 623, 662 (1887) (concluding that there are substantive limits on the legislative power to restrict production and sale of goods, but that an alcohol ban was justified because "we cannot shut out of view the fact, within the knowledge of all, that the public health, the public morals, and the public safety, may be endangered by the general use of intoxicating drinks; nor the fact, established by statistics accessible to every one, that the idleness, disorder, pauperism, and crime existing in the country are, in some degree at least, traceable to this evil").

      As to marijuana, compare United States v. Bauer, 84 F.3d 1549, 1559 (9th Cir. 1996) (vacating a conviction and remanding for inquiry on whether a marijuana ban passes strict scrutiny as applied to religious marijuana users), and Ravin v. State, 537 P.2d 494, 513 (Alaska 1975) (striking down under the Alaska Constitution a ban on private possession of marijuana), with, for example, State v. Balzer, 954 P.2d 931, 941 (Wash. Ct. App. 1998) (concluding that "the 'social realities' of marijuana's effects and dangers" justify banning even sacramental uses, and citing many other cases that have reached this conclusion). The Ravin court acknowledged that the state may regulate "activities of the individual which affect others or the public at large," including ones that do not "have a present and immediate impact on the public health or welfare," but it held that marijuana use "will [not] result in numbers of people becoming public charges or otherwise burdening the public welfare"; and, the court reasoned, while "the need for control of drivers under the influence of marijuana" justifies a ban on possession or ingestion of marijuana in cars,

      given the relative insignificance of marijuana consumption as a health problem in our society at present, . . . the potential harm generated by drivers under the influence of marijuana, standing alone, [does not] create[] a close and substantial relationship between the public welfare and control of ingestion of marijuana or possession of it in the home for personal use.

      Ravin, 537 P.2d at 509, 511.

      As to peyote, compare Employment Division v. Smith, 494 U.S. 872, 905-06 (1990) (O'Connor, J., concurring) (concluding that a religious exemption for peyote users may lead to serious health effects and to an increase in illegal drug trafficking); Peyote Way Church of God, Inc. v. Smith, 742 F.2d 193, 196 (5th Cir. 1984) (describing the leakage of peyote from the church's supplies, supposedly kept only for church-run rituals, to nonmembers); Brief for Petitioners at 14 n.8, Employment Division v. Smith (No. 88-1213) (citing medical evidence that peyote "can cause temporary psychosis[,] severe hypertension, . . . convulsions [and] death or respiratory failure"); and Reply Brief for Petitioners at 14 & nn.27-28, Employment Division v. Smith (No. 88-1213) (citing sources that report that children sometimes attend the peyote ceremony and participate in the peyote ingestion), with, for example, Employment Division v. Smith, 494 U.S. at 913, 916 (Blackmun, J., dissenting) (concluding that peyote is not harmful because of "[t]he Native American Church's internal restrictions on, and supervision of, its members' use of peyote" and that "[t]here is . . . practically no illegal traffic in peyote"), and Whitehorn v. State, 561 P.2d 539, 544 (Okla. Crim. App. 1977) (concluding that allowing a peyote exemption wouldn't jeopardize compelling government interests).

      Note that, even if peyote consumption is generally carefully monitored in the Native American Church today, and this monitoring prevents the possible harmful consequences of peyote use, "denominational practices, and indeed [the practices of] individual believers, even in long-standing religions, can and do change. [Religions and individuals] change the nature of their religious beliefs, they change the nature of their doctrine, and that is the very essence of freedom of religion and belief." Respondent's Oral Argument at 23, Employment Division v. Smith (No. 88-1213), available in 1989 U.S. TRANS LEXIS 94 (Att'y Gen. David B. Frohnmayer, for the State of Oregon).  Thus, unless the government closely monitors Native American Church rituals--something that would raise obvious problems of its own--the government can have little assurance that peyote is not in fact being used by minors or being abused in ways that can lead to various costs to others. This argument was brought to my attention by Garrett Epps's excellent To an Unknown God: The Hidden History of Employment Division v. Smith, 30 Ariz. St. L.J. 953, 1010-11 (1998).

      168 See supra note 48 (citing cases involving a religious freedom defense to a ban on carrying knives); Nicholas J. Johnson, Beyond the Second Amendment: An Individual Right to Arms Viewed Through the Ninth Amendment, 24 Rutgers L.J. 1, 2-3 (1992) (concluding that any risks of abuse by private gun owners can't justify restricting such ownership, even setting aside the Second Amendment). Of course, the Second Amendment and state constitutional rights to keep and bear arms might explicitly protect such conduct even when it risks creating harm for others. See generally, e.g., Sanford Levinson, The Embarrassing Second Amendment, 99 Yale L.J. 637 (1989); L.A. Powe, Jr., Guns, Words, and Constitutional Interpretation, 38 Wm. & Mary L. Rev. 1311 (1997); Eugene Volokh, The Commonplace Second Amendment, 73 N.Y.U. L. Rev. 793 (1998). Here I ask only what the rules would be under general liberty or religious liberty principles.

      169 West Coast Hotel Co. v. Parrish, 300 U.S. 379, 399 (1937) (using this argument to uphold a minimum-wage law).

      170 Compare Adams v. Tanner, 244 U.S. 590, 593 (1917) (striking down a ban on employment agencies on the grounds that "there is nothing inherently immoral or dangerous to public welfare" about them, and that "[o]n the contrary, such service is useful, commendable, and in great demand"), with id. at 604 (Brandeis, J., dissenting) (arguing that a legislature could conclude that employment agencies "actually serve to congest the labor market and to increase idleness and irregularity of employment").

      171 Compare People v. Fries, 250 N.E.2d 149, 151 (Ill. 1969) (striking down, on substantive due process grounds, a requirement that motorcyclists wear helmets, and stating that the law is a "regulation of what is essentially a matter of personal safety"), with Robotham v. State, 488 N.W.2d 533, 540 (Neb. 1992) (upholding a helmet law because injuries impose externalities on taxpayers, because helmets "help to prevent accidents by lessening the likelihood that the cyclist will lose control of the cycle after a blow to the head," and because "'all users of a highway have . . . a definite interest in how serious are the consequences, not only to themselves but to others, of any accident in which they may become involved,' if for no other reason than that [Nebraska is a] comparative negligence state[]") (citations omitted), and People v. Kohrig, 498 N.E.2d 1158, 1163-65 (Ill. 1986) (making similar arguments in the context of a mandatory seat-belt law, and overruling People v. Fries).

      172 Some such restrictions also incidentally punish conduct that doesn't have the potential to harm others. If forced to seriously consider such incidental effects, the legislature might agree that the restriction may have been written more broadly than necessary and need not be applied in certain situations; it's therefore appealing to let courts carve out religious exemptions in such cases. But the common-law exemption regime would generally suffice for this purpose: It would let courts create exemptions when they think the exemptions wouldn't harm others, and such exemptions would then remain the law unless the legislature concluded that they would indeed cause harm.

      I recognize that some restrictions are generally defended as protecting purely moral interests, and are rarely seen as preventing specific harm to others. Bans on public nudity, even in places that are open only to those who voluntarily choose to see such nudity, cf. Barnes v. Glen Theatre, Inc., 501 U.S. 560 (1991), might be an example, though I think the ultimate concern there is indeed with harm to others, albeit harm supposedly indirectly caused by corruption of public morals. But even if my arguments don't apply to this narrow set of cases, I believe the overwhelming majority of the disputed religious exemption cases do involve eminently credible claims of harm to others. See supra note 154.

      173 See Adkins v. Children's Hosp., 261 U.S. 525 (1923).

      The ethical right of every worker, man or woman, to a living wage may be conceded. . . . [B]ut the fallacy of the [minimum-wage law] is that it assumes that every employer is bound at all events to furnish it. . . . Certainly the employer, by paying a fair equivalent for the service rendered, though not sufficient to support the employee, has neither caused nor contributed to her poverty.

      Id. at 558; cf. id. at 562 (Taft, C.J., dissenting) (concluding that minimum-wage laws are permissible in order to prevent the "overreaching of the harsh and greedy employer," a view that the majority opinion implicitly rejects).

      174 See supra note 158.

      175 Truax v. Corrigan, 257 U.S. 312, 327-28 (1921). But see id. at 354 (Brandeis, J., dissenting) ("What methods and means are permissible in [the] struggle of contending forces [involved in competition between businesses or between capital and labor] is determined in part by decisions of the courts, in part by acts of the legislatures. The rules governing the contest necessarily change from time to time.").

      176 See Lochner v. New York, 198 U.S. 45, 64 (1905).

      177 Prudential Ins. Co. v. Cheek, 259 U.S. 530, 545-46 (1922). Note that the statute went beyond the normal scope of the right to reputation, as traditionally protected by defamation law; the statute "convert[ed] what otherwise might be . . . a 'moral duty,' into a legal duty." Id. at 546. Today, this might be analyzed as a First Amendment "compelled speech" case, but the Cheek Court treated it only as a substantive due process case, because it concluded that the First Amendment was not applicable to the states through the Fourteenth Amendment.

      178 Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 394-95 (1926).

      179 Mugler v. Kansas, 123 U.S. 623, 662 (1887). But see state cases cited supra note 167 (striking down state alcohol bans on the grounds that private consumption of alcohol does not invade the rights of others).

      180 West Coast Hotel Co. v. Parrish, 300 U.S. 379, 398-99 (1937).

      181 Id. at 399.

      182 Luke 10:7. I don't suggest that most Christians would take this view (from what I've seen, this verse is generally quoted as setting a lower bound), only that some might.

      183 See Thomas v. Review Bd., 450 U.S. 707, 715-16 (1981) (holding that the right to exemption "is not limited to beliefs which are shared by all of the members of a religious sect," but applies even to idiosyncratic religious beliefs). A claim of an employee's religious duty not to comply with minimum-wage law was in fact raised in Tony & Susan Alamo Foundation v. Secretary of Labor, 471 U.S. 290 (1985); the Court disposed of the case by concluding that the particular claimants had a way to obey the law without violating their religious beliefs, but such a result was made possible only by the specific nature of these particular claimants' beliefs. See id. at 303-05. Had the claimants' beliefs been slightly different, the Court would have had to engage in an analysis very similar to the one I describe in the text.

      184 See West Coast Hotel Co., 300 U.S. at 397-400.

      185 See Adkins v. Children's Hosp., 261 U.S. 525, 545-48, 560-61 (1923).

      186 Cf. Tony & Susan Alamo Found., 471 U.S. at 302 (stressing that minimum-wage law is needed to protect even those employees who supposedly consent to working for less than the minimum wage, because otherwise "employers might be able to use superior bargaining power to coerce employees . . . to waive their protections under the Act," and that sub-minimum-wage payments "affect many more people than those workers directly at issue in this case and would be likely to exert a general downward pressure on wages in competing businesses").

      187 I set aside here the narrow area of the right to privacy, which affects only a small subset of government regulations, and which does let people act in ways that may harm others (for instance, imposing externalities on a child and on society by educating the child less effectively than one might, or imposing externalities on taxpayers by having more children than one can afford to raise oneself). I am no great fan of the right to privacy, for many of the reasons I give with regard to an asserted right to act in religiously motivated ways; but it's at least quite narrow, and it imposes judicial supremacy only with respect to a distinctly limited set of government regulations.

      188 261 U.S. 525 (1923).

      189 See Cass R. Sunstein, What's Standing After Lujan? Of Citizen Suits, "Injuries," and Article III, 91 Mich. L. Rev. 163, 187 (1992) (describing the New Deal laws that were ultimately upheld by the Court as "generally reflect[ing] a democratic judgment that the new interests now protected by statute--the interests of consumers, listeners, poor people, and so forth--should receive no less protection than the interests traditionally protected by the common law," and arguing that "[t]he New Deal had itself been a wholesale attack on the idea that common law interests deserved special constitutional status").

      190 Cf. Peller, supra note 12, at 576-77 & n.18 (discussing the substantive due process cases' constitutionalization of common-law judgments).

      191 See, e.g., Thomas v. Anchorage Equal Rights Comm'n, 165 F.3d 692 (9th Cir. 1999) (accepting the claim, under a Free Exercise Clause "hybrid rights" rationale); Swanner v. Anchorage Equal Rights Comm'n, 874 P.2d 274, 284 (Alaska 1994) (rejecting such a claim); Swanner v. Anchorage Equal Rights Comm'n, 513 U.S. 979, 981-82 (1994) (Thomas, J., dissenting from denial of certiorari) (suggesting that the claim should have been accepted); Smith v. Fair Employment & Hous. Comm'n, 51 Cal. Rptr. 2d 700 (1996) (three Justices concluding that the law didn't substantially burden religious exercise, three Justices concluding that it did burden religious exercise and that it failed strict scrutiny, and one Justice not reaching the question); Jasniowski v. Rushing, 685 N.E.2d 622 (Ill. 1997) (reversing, with no opinion, a lower court's rejection of such a claim), rev'g 678 N.E.2d 743 (Ill. App. Ct. 1997); Attorney Gen. v. DeSilets, 636 N.E.2d 233, 242-43 (Mass. 1994) (concluding that the result depends on factual findings); McCready v. Hofius, 586 N.W.2d 723 (Mich. 1998) (rejecting such a claim), vacated and remanded, No. 108995, 1999 WL 226862 (Apr. 16, 1999) (appearing to reverse course, with little explanation); State v. French, 460 N.W.2d 2, 9-11 (Minn. 1990) (accepting such a claim).

      192 See Roberts v. United States Jaycees, 468 U.S. 609 (1984).

      193 Compare the rationale from Coppage v. Kansas quoted supra note 158.

      194 "To be treated equally" is shorthand here for "to be treated without regard to the particular attribute that is the subject of the antidiscrimination law."

      195 The plurality in Smith v. Fair Employment & Housing Commission concluded that the law didn't burden religious freedom because the landlady could have sold her property and gone into another line of business, but that can't be right: Surely a regulation that requires religious believers to change their occupation must count as a substantial burden.  A law that required service stations to be open seven days a week would, I assume, clearly impose a substantial burden on a Sabbatarian who felt obligated to close the station on Saturdays, even though he could avoid the burden by selling his service station and becoming an office worker. The landlady example is structurally identical.

      196 Cf. Sunstein, supra note 189, at 191 ("When Congress creates a cause of action enabling people to complain against racial discrimination, consumer fraud, or destruction of environmental assets, it is really giving people a kind of property right in a certain state of affairs.").

      197 See, e.g., Smith v. Fair Employment & Hous. Comm'n, 30 Cal. Rptr. 2d 395, 404-06 (Ct. App. 1994), rev'd, 51 Cal. Rptr. 2d 700 (1996).

      198 One could try to defend the second position on the grounds that the landlady's interests outweigh, simply because of their religiosity, the relatively slight importance of the private right to be free from housing discrimination; but, for the reasons given supra Part III.A.2, such an approach is normatively unappealing. The landlady may think that God requires her to discriminate against the tenants, but the tenants doubtless have a very different view of what God commands. It is wrong for the legal system to authorize one person to injure what we acknowledge to be the rights of others based solely on a subjective religious belief that the others do not share; and this remains true even when those rights are among the less weighty ones we have--for instance, the right to be free from discrimination, trespass, copyright infringement, defamation, or breach of contract, rather than, say, the right to life or bodily integrity.

      199 521 U.S. 702 (1997).

      200 See id. at 731-35; supra note 187.

      201 See Glucksberg, 521 U.S. at 736; see also Nelson Lund, Two Precipices, One Chasm: The Economics of Physician-Assisted Suicide and Euthanasia, 24 Hastings Const. L.Q. 903, 914-16 (1997); Martha Minow, Which Question? Which Lie? Reflections on the Physician-Assisted Suicide Cases, 1997 Sup. Ct. Rev. 1, 21-23.

      202 See Glucksberg, 521 U.S. at 736.

      203 See Brief of 36 Religious Organizations, Leaders, and Scholars as Amici Curiae in Support of Respondents, Vacco v. Quill, 521 U.S. 793 (1997) (No. 95-1858), and Glucksberg (No. 96-110) (making this argument); Winthrop Drake Thies, Shall the Dying Be Denied Their Religious Freedom?, Newark Star-Ledger, Feb. 6, 1997, at 26.

      [W]here the dying patient, in accordance with his religious beliefs, concludes that a loving and compassionate God has by the terminal process called him; and where the patient's physician believes that God asks him to play the part of the good Samaritan in releasing the dying person from suffering through death, we may not in effect say, "Your religion is different from mine and thus not worthy of respect. You can't die the way you religiously feel called upon to die. And you, Doctor, can't respond to your patient the way your religion calls on you to respond, although your suffering patient cries out for your help." The First Amendment guarantees to each of us--including each dying person and his physician--the free exercise of our religions.

      Id.; see also McIver v. Krischer, No. CL-96-1504-AF, 1997 WL 225878, at *11 (Fla. Cir. Ct. Jan. 31, 1997) (finding a right to assisted suicide under the state constitutional right to privacy, and stating that a doctor "can determine his own ethical, religious, and moral beliefs in declining or agreeing to assist"), rev'd, 697 So. 2d 97 (Fla. 1997).

      204 Even supporters of a right to assisted suicide don't generally argue that the right should be recognized solely because of the claimant's motivation; rather, they argue that the right rests on the importance of the decision to commit suicide. See generally Compassion in Dying v. Washington, 79 F.3d 790 (9th Cir. 1996) (en banc), rev'd sub nom. Washington v. Glucksberg, 521 U.S. 702 (1997).

      205 Certain constitutional provisions--for instance, the Contracts Clause, the Ex Post Facto Clause, and the procedural component of the Due Process Clause--by their nature "constitutionalize" existing legal rules, but in a distinctly limited way: The constitutional limits on a legislature's power to impair contracts that were valid when entered, to punish conduct that was legal when it took place, or to deprive people of property rights granted under existing law only apply to past conduct and to existing contracts and property rights. The legislature remains free to redefine the legal rules of contract law, criminal law, and property law for future contracts, future conduct, and newly granted property rights, which makes these constitutional provisions much lighter constraints on legislative discretion than the substantive component of the Due Process Clause was in the early 1900s.

      Likewise, while the Takings Clause in some measure limits the legislature's power to redefine existing property rights (though when the redefinition falls short of actual transfer of property or interference with the right to exclude, the Takings Clause has been read as providing only a very slight limitation, see, e.g., Andrus v. Allard, 444 U.S. 51 (1979); Penn Cent. Transp. Co. v. City of New York, 438 U.S. 104 (1978)), the legislature can escape these limits by paying compensation. The compensation requirement is a significant burden on the legislature, but one that's considerably smaller than a total prohibition on such legislative action would be; moreover, many interferences with private conduct, even if they are classified as takings, in practice cause so little economic loss that the compensation requirement poses little difficulty. See, e.g., Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419 (1982).

      206 Cf. infra Part III.E.5.

      207 I take the view that a right whose text and original meaning is clear is binding even when we no longer find it appealing. See, e.g., Eugene Volokh, The Amazing Vanishing Second Amendment, 73 N.Y.U. L. Rev. 831, 831 (1998). But because my argument ultimately turns on the relative unclarity of the text and original meaning of the Free Exercise Clause, those who are less committed to following the text and original meaning should be even more willing to accept my conclusions.

      208 Compare City of Boerne v. Flores, 521 U.S. 507, 537 (1997) (Scalia, J., concurring in part), and Philip A. Hamburger, A Constitutional Right of Religious Exemption: An Historical Perspective, 60 Geo. Wash. L. Rev. 915 (1992) (arguing that the Free Exercise Clause was not originally seen as providing compelled exemptions), withBoerne, 521 U.S. at 548 (O'Connor, J., dissenting) (taking the opposite view), Michael W. McConnell, Freedom from Persecution or Protection of the Rights of Conscience?: A Critique of Justice Scalia's Historical Arguments in City of Boerne v. Flores, 39 Wm. & Mary L. Rev. 819 (1998) (same), Michael W. McConnell, The Origins and Historical Understanding of Free Exercise of Religion, 103 Harv. L. Rev. 1410 (1990), and Lash, supra note 126 (taking a view similar to McConnell's, but more influenced by the original meaning of the Fourteenth Amendment than of the Free Exercise Clause).

      209 For instance, it seems extremely likely that any devoutly religious person who chooses assisted suicide would do so at least in part based on his religious beliefs.

      210 Cf. Commonwealth v. Campbell, 117 S.W. 383, 385, 387 (Ky. 1909) (striking down an alcohol possession ban on the grounds that the legislature may not restrict "what a man will drink, or eat, or own, provided the rights of others are not invaded," and that private alcohol possession "can by no possibility injure or affect the health, morals, or safety of the public").

      211 Legislators and executive officials have their own independent duty to decide what the Free Exercise Clause means, and to act on this decision even when courts have made clear that they will not read the clause as compelling exemptions. See generally Paul Brest, The Conscientious Legislator's Guide to Constitutional Interpretation, 27 Stan. L. Rev. 585 (1975); Sanford Levinson, Identifying the Compelling State Interest: On "Due Process of Lawmaking" and the Professional Responsibility of the Public Lawyer, 45 Hastings L.J. 1035 (1994); Lawrence Gene Sager, Fair Measure: The Legal Status of Underenforced Constitutional Norms, 91 Harv. L. Rev. 1212 (1978). What this duty means is a fascinating question, but one I won't go into here.

      212 See Miller v. Johnson, 515 U.S. 900, 911 (1995) (Equal Protection Clause) (dictum); Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 532 (1993) (Free Exercise Clause); Larson v. Valente, 456 U.S. 228 (1982) (Establishment Clause).

      213 See Personnel Adm'r v. Feeney, 442 U.S. 256 (1979); Washington v. Davis, 426 U.S. 229 (1976); cf. City of Boerne v. Flores, 521 U.S. 507, 530 (1997) ("RFRA's legislative record lacks examples of modern instances of generally applicable laws passed because of religious bigotry. . . . [A]s one witness testified, 'deliberate persecution is not the usual problem in this country.' . . . '[L]aws directly targeting religious practices have become increasingly rare.'"). But see Laycock, supra note 49 (arguing that deliberate discrimination against religion is fairly common in zoning cases).

      214 See supra note 122.

      215 United States v. Carolene Prods. Co., 304 U.S. 144 (1938).

      216 Id. at 152 n.4.

      217 Id. (emphasis added) (citations omitted).

      218 Id.

      219 See Ferguson v. Skrupa, 372 U.S. 726 (1963) (upholding a law that banned everyone except lawyers from becoming debt adjusters).

      220 See Adams v. Tanner, 244 U.S. 590 (1917) (striking down, under substantive due process, a law that banned employment agencies).

      221 198 U.S. 45 (1905).

      222 236 U.S. 1 (1915).

      223 See supra text accompanying notes 132-139.

      224 See supra Part III.A.2. .

      225 As I mention in note 183, I don't mean my reference to the right to privacy as an endorsement of the right.

      226 Swanner v. Anchorage Equal Rights Comm'n, 874 P.2d 274, 289 (Alaska 1994) (Moore, C.J., dissenting); see also, e.g., State v. French, 460 N.W.2d 2, 10 (Minn. 1990) (making the same argument).

      227 Laycock, supra note 43, at 50; cf. Garrett Epps, What We Talk About When We Talk About Free Exercise, 30 Ariz. St. L.J. 563, 597 (1998) (arguing that truly generally applicable laws should usually not be subject to constitutionally mandated exemptions, but saying that such laws "will be quite rare," because to qualify they "cannot contain a structure of exemptions permitting some secular or religious exemptions but denying others"); Paulsen, supra note 7, at 254, 267-70 (taking a similar view).

      228 Laycock, supra note 43, at 50-51.

      229 Id. at 50. After Laycock's article was written, Title VII was amended to apply to Congress. See 2 U.S.C. § 1311(a) (Supp. 1995-97) (enacted 1995).

      230 See infra note 239.

      231 Cf., e.g., Model Penal Code § 3.02 cmt. 1 (1985) (describing the necessity defense, and giving as an example a person lost in a storm who takes refuge in a house).

      232 See, e.g., Cal. Evid. Code §§ 950-1063 (West 1995). For claims of a general religious exemption (not limited to a clergy-penitent privilege) from a duty to testify, see sources cited supra note 160.

      233 See, e.g., N.Y. Penal Law § 130.25 (McKinney 1999) (providing such limitations). Statutory-rape laws sometimes phrase these exclusions as facial limitations, rather than as exceptions, but surely that can't be dispositive; the important thing is that the legislature has consciously decided to make statutory rape cover something less than all sex with people (or even just with women) under the age of consent.

      234 See 17 U.S.C. §§ 106-121 (1994); see also id. § 1008 (exempting home copying of recorded music). Other provisions, such as the fact that copyrights eventually lapse into the public domain, see id. §§ 302-305, or the fact that private performances and displays are not considered infringements, see id. § 106, might likewise be seen either as exceptions to the ban on unauthorized copying or more generally as evidence that the Copyright Act explicitly sanctions certain kinds of unauthorized copying.

      235 This last is indeed an exception from the general ban on killing of another, and not just "not killing at all." Intentionally disconnecting life-sustaining equipment without a patient's request would normally be considered a homicide, cf. People v. Phillips, 414 P.2d 353, 358 (Cal. 1966) (implying that unconsented-to shortening of a patient's life may be homicide), and absence of consent is not an element of a prima facie homicide case, cf. Gospodareck v. State, 666 So. 2d 835, 842 (Ala. Crim. App. 1993).

      236 Laycock, supra note 43, at 51.

      237 See 42 U.S.C. § 2000e-2(e).

      238 See Frederick Schauer, Exceptions, 58 U. Chi. L. Rev. 871, 873-75 (1991).

      239 . See 42 U.S.C. §§ 2000e(b), 2000e-1(b) (exempting businesses with fewer than 15 employees and exempting discrimination in foreign countries when required by foreign law); EEOC v. Arabian Am. Oil Co., 499 U.S. 244 (1991) (exempting, under the original version of the Civil Rights Act, discrimination in foreign countries); United Steelworkers v. Weber, 443 U.S. 193 (1979) (exempting some programs that give preference to racial minorities); see also 42 U.S.C. § 2000e-2(i) (exempting certain preferences for American Indians). Compare Quaring v. Peterson, 728 F.2d 1121, 1127 (8th Cir. 1984) (concluding that a religious objector to photographs should be able to get a driver's license without a photograph, and that "[b]ecause the state already allows numerous exemptions to the photograph requirement, the . . . argument that denying Quaring an exemption serves a compelling state interest is without substantial merit"), aff'd by an equally divided court, 472 U.S. 478 (1985), withid. at 1128 (Fagg, J., dissenting) ("[T]he limitations placed upon these provisional operators [who are exempted from the photograph requirement] deny any meaningful comparison with the state's regular license holders.").

      240 Laycock, supra note 43, at 51.

      241 Cf. cases cited infra note 248.

      242 McConnell, supra note 152, at 1147.

      243 Id. at 1148.

      244 Id.

      245 McConnell specifically suggests this approach as an alternative to the "arbitrariness" and the "unacceptabl[e] subjectiv[ity]" of pre-Smith supposed "strict scrutiny," and as a "principled basis" that prevents judges from being "forced into the sort of free-wheeling balancing of incommensurate interests that the majority feared in Smith." Id. at 1144-45; cf. Jesse Choper, Securing Religious Liberty 93, 95 n.154 (1995) (likewise acknowledging that the pre-Smith framework "involves a comparison of incommensurables that creates serious problems of judicial prerogative in constitutional adjudication," and specifically endorsing McConnell's focus on whether "the governmental interest [is] so important that the government would impose a burden of this magnitude on the majority in order to achieve it").

      246 455 U.S. 252 (1982).

      247 McConnell, supra note 106, at 1148; see also Respondent's Oral Argument at 26, Employment Division v. Smith, 494 U.S. 872 (1990) (No. 88-1213), available in 1989 U.S. TRANS LEXIS 94 ("[If] Indian people were in charge of the United States [government] right now, . . . you might find that alcohol was the [banned] substance and peyote not listed at all."). Likewise, if Rastafarians were in charge of the U.S. government, marijuana would probably be legal; if traditionalist Muslims were in charge of the U.S. government, sex discrimination would probably be legal.

      248 Cf., e.g., Fiedler v. Marumsco Christian Sch., 631 F.2d 1144 (4th Cir. 1980) (involving such a claim made by a nonmainstream Christian); Brown v. Dade Christian Schs., Inc., 556 F.2d 310 (5th Cir. 1977) (same).

      249 Cf. Berg, supra note 106, at 38 (making a similar point); William P. Marshall, In Defense of Smith and Free Exercise Revisionism, 58 U. Chi. L. Rev. 308, 316-17 (1991) (same).

      250 McConnell, supra note 152, at 1148.

      251 See supra text accompanying notes 231, 235.

      252 See, e.g., Epps, supra note 227, at 589.

      253 See, e.g., Stephen M. Bainbridge, Student Religious Organizations and University Policies Against Discrimination on the Basis of Sexual Orientation: Implications of the Religious Freedom Restoration Act, 21 J.C. & U.L. 369, 389-90, 393-94 (1994).

      254 Cf. Eugene Volokh, Freedom of Speech and the Constitutional Tension Method, 3 U. Chi. L. Sch. Roundtable 223 (1996) (discussing a similar argument in the Free Speech Clause context).

      255 See County of Sacramento v. Lewis, 523 U.S. 833 (1998) (holding that negligent injury doesn't violate the Constitution); Albright v. Oliver, 510 U.S. 266 (1994) (same as to malicious prosecution); Oliver v. United States, 466 U.S. 170, 183 (1984) (same as to trespass on open fields, noting that "[t]he law of trespass . . . forbids intrusions upon land that the Fourth Amendment would not proscribe"); Paul v. Davis, 424 U.S. 693 (1976) (same as to defamation). As to the general freedom to do what one pleases, the government may use the threat of violent retribution embodied in the criminal law to restrain a wide range of conduct--selling drugs, engaging in price-fixing, and the like--which private parties could not violently restrain.

      256 Some states impose some limited restrictions on employers' punishing employees for certain kinds of speech and political activity, see, e.g., Cal. Lab. Code § 1101 (West 1998), but these are a fairly narrow exception to the general point I make in the text.

      257 The Enforcement Clause of the Fourteenth Amendment gives Congress the power to prevent deprivations of "life, liberty, or property without due process of law," but even this goes only to deprivations by the state, not by private parties. See City of Boerne v. Flores, 521 U.S. 507, 524 (1997) (citing approvingly The Civil Rights Cases, 109 U.S. 3 (1883)).

      258 McConnell, supra note 152, at 1135 (emphasis added).

      259 Pepper, supra note 152, at 334; see also Berg, supra note 106, at 38 (taking a similar view).

      260 Even in the narrow context of peyote, there is considerable factual disagreement on how harmful a religious exemption would be, see sources cited supra note 167, and the uncertainty is complicated by the fact that "what government knows or can know about private religious drug use is highly uncertain." Reply Brief for Petitioners at 11, Employment Division v. Smith, 494 U.S. 872 (1990) (No. 88-1213).

      261 See cases cited supra note 48. Compare John R. Lott, Jr., More Guns, Less Crime: Understanding Crime and Gun Control Laws (1998) (concluding that allowing all law-abiding adults to get permits to carry guns decreases violent crime), and William A. Bartley & Mark A. Cohen, The Effects of Concealed Weapons Laws: An Extreme Bound Analysis, 36 Econ. Inquiry 258 (1998) (confirming Lott's findings), with conventional wisdom (in certain circles) to the contrary.

      262 Compare, e.g., Ribnik v. McBride, 277 U.S 350, 357 (1928) (concluding that a law fixing maximum prices for employment agency services was unconstitutional, in part because there is "no reason for applying a [special] rule" for such services), with id. at 362-74 (Stone, J., dissenting) (arguing that such services are in fact different from other commodities, and more prone to various abuses); compare Weaver v. Palmer Bros. Co., 270 U.S. 402 (1926) (concluding that a law banning use of second-hand rags as bedding stuffing was unconstitutional because such use doesn't cause any health problems), with id. at 415 (Holmes, J., dissenting) (taking the opposite view); compare Lochner v. New York, 198 U.S. 45, 58 (1905) (concluding that a law setting maximum working hours for bakers wouldn't actually improve their health), withid. at 69 (Harlan, J., dissenting) (taking the opposite view).

      263 See Wisconsin v. Yoder, 406 U.S. 205, 224 (1972).

      264 See id. at 244-45 (Douglas, J., dissenting).

      265 Cf. id. at 245 n.2 (Douglas, J., dissenting).

      266 McConnell, supra note 106, at 1145; see also Akhil Reed Amar, The Bill of Rights: Creation and Reconstruction 43, 255-56 (1998) (suggesting that "some religious practices that affect only the religious community itself (with no externalities imposed on religious nonbelievers)" should be constitutionally immunized from regulation); id. at 386 n.100 (tentatively suggesting that even voluntary human sacrifice might be shielded by the Free Exercise Clause); Akhil Reed Amar, Intratextualism, 112 Harv. L. Rev. 747, 820 & n.285 (1999) (suggesting that the Fourteenth Amendment's incorporation of the Free Exercise Clause should be read as protecting practices "[w]here only co-religionists are involved" and not ones that "impose[] a harm on nonbelievers," with "harm" used "in the sense elaborated by John Stuart Mill and his followers").

      267 Pepper, supra note 152, at 335.

      268 471 U.S. 290 (1985).

      269 McConnell, supra note 106, at 1145.

      270 Tony & Susan Alamo Found., 471 U.S. at 302.

      271 McConnell, supra note 106, at 1145; cf. id. at 1145 n.159 (describing this principle as linked to "the economic concept of externalities," and arguing that "[w]here the government is preventing the imposition of negative externalities, its interest generally overrides free exercise claims, but otherwise (except in special circumstances) it does not"); Michael W. McConnell, The Rule of Law and the Role of the Solicitor General, 21 Loy. L.A. L. Rev. 1105, 1109 (1988) (talking further about this case).

      272 Cf. Jennifer Friesen, State Constitutional Law: Litigating Individual Rights, Claims, and Defenses 235 n.181, 249-50, supp. 22 (2d ed. 1996 & Supp. 1997) (citing cases interpreting state constitutions as mandating an exemption regime); Angela C. Carmella, State Constitutional Protection of Religious Exercise: An Emerging Post-Smith Jurisprudence, 1993 BYU L. Rev. 275 (generally discussing this subject); Daniel A. Crane, Beyond RFRA: Free Exercise of Religion Comes of Age in the State Courts, 10 St. Thomas L. Rev. 235, 245-47 & nn.71-79 (1998) (citing cases interpreting state constitutions as mandating an exemption regime).

      273 See, e.g., S.B. 604, 1998 Reg. Sess. (Ala. 1998) (state constitutional amendment approved by the voters in the Nov. 1998 election); S.B. 66, 1998 Reg. Sess. (Fla. 1998); H.R. 865, 144th Gen. Assembly (Ga. 1997).

      274 A mild version of the early 1900s substantive due process jurisprudence does exist in some states, see, e.g., Russell v. Town of Pittsford, 464 N.Y.S.2d 906 (App. Div. 1983); Zoning Hearing Bd. v. Lenox Homes, Inc., 439 A.2d 218 (Pa. Commw. Ct. 1982), but this seems to be only slightly less deferential than a rational basis test.

      275 See, e.g., Religious Liberty Protection Act of 1998, S. 2148, H.R. 4019, 105th Cong. (1998).

      276 The "race to the bottom" argument that is sometimes made against state-level corporate and environmental law doesn't apply to the religious-exemption question. The core of the argument is that businesses will flock to states with looser regulations, and therefore states will compete with one another to have regulations that are as lax as possible, even if this means underprotecting their citizens from fraud or pollution. Thus, one state's decision to have a particular regime will indirectly hurt citizens of other states, which is seen as a justification for having a federal, rather than just state-by-state, legal regime. But see, e.g., Alexander Volokh et al., Race to the Top: The Innovative Face of State Environmental Management (Reason Pub. Policy Inst., Policy Study No. 239, Feb. 1998) (arguing that some such races end up increasing the welfare of state residents rather than decreasing it).

      But I suspect that very few people who are either the potential beneficiaries of religious exemptions (e.g., landlords who want to discriminate against unmarried couples) or potential victims of such exemptions (e.g., unmarried couples who want a place where they can rent free from discrimination) will move from state to state in search of a better exemption regime. This makes it unlikely that any state will change its laws to attract such migrants, which in turn makes it unlikely that any other state will feel the pressure to "race to the bottom" (or to the top) to keep pace.

      Occasionally, a state might indeed relax its laws in order to attract a large religious community that is so dissatisfied with its current residence that many of its members feel the need to move--one could imagine something like this happening with a modern replay of the Mormon migrations of the 1800s. Even then, it's not clear that there would be a race to the bottom; other jurisdictions might not feel any real pressure to keep pace. But in any event, the situations where one state's religious-exemption regime leads to this sort of harmful externality on another state should be exceedingly rare.  They might justify a federal solution when they come up, but they don't justify having a federal regime for all religious exemptions, including the overwhelming majority that won't have any serious effects on other states.

      277 As I mention in Part I.E.2, I'm agnostic about whether state legislatures should also delegate to local legislatures considerable power to override judicially crafted exemptions; here I stress only that, regardless of one's position on this question, Congress shouldn't entirely preempt the area and reserve the override power exclusively to itself.

      278 See, e.g., Berg, supra note 76, at 725.

      279 The statute might, for instance, say: "This Act shall not apply to any states whose courts are generally authorized, by the state constitution or by a state statute, to grant religious exemptions from laws of general applicability, whether or not the state constitution or state statute contains any exceptions."

      280 Such a statute might say: "This Act shall not apply to any state or federal statute that explicitly excludes such application by reference to this Act." Cf. RFRA, 42 U.S.C. § 2000bb-3(b) (1994) ("Federal statutory law adopted after the date November 16, 1993 is subject to this chapter unless such law explicitly excludes such application by reference to this chapter.").

      Any such statute would of course have to be limited to laws that are subject to some enumerated Congressional power, such as the power to regulate interstate commerce.

      281 There is one significant difference between the two statutes: Under the first statute, religious freedom lawsuits will be brought in some states under a state RFRA and in other states under the federal son-of-RFRA; under the second statute, religious freedom lawsuits will generally be brought under the federal son-of-RFRA (except where a state RFRA provides more protection than the federal statute). This means that the first approach would lead to several potentially different (though likely quite similar) bodies of judicially crafted exemptions--much as the various states have subtly different bodies of judicially crafted tort law--while the second approach would lead to a single body of federal judicially crafted exemption law. Both approaches would allow variation among the states as a result of legislative decisions, but the first would allow more variation among state judicial decisions.

      282 But cf. Ira C. Lupu, Reconstructing the Establishment Clause: The Case Against Discretionary Accommodation of Religion, 140 U. Pa. L. Rev. 555, 585-87 (1991) (arguing that legislative discretion in deciding which practices to accommodate violates constitutional principles of religious equality).

      283 But cf. Berg, supra note 76, at 724 (arguing that statutory exemption schemes are inherently unfair because of the risk of religious favoritism); Laycock, supra note 227, at 14-15 (same).

      284 Lupu, supra note 282, at 602.

      285 See Laycock, supra note 227, at 14-15 ("[A] line based on compelling interest is likely to be more principled than a line based on who can get the sympathetic attention of enough legislators.").

      286 Cf., e.g., Randy E. Barnett, The Structure of Liberty: Justice and the Rule of Law (1998); Richard A. Epstein, Takings: Private Property and the Power of Eminent Domain (1985).

      287 See West, supra note 21, at 604 (suggesting that there is no way for the Court to "draw the line in any sort of fair and principled manner").

      288 See Atchison, Topeka, & Santa Fé Ry. Co. v. Vosburg, 238 U.S. 56 (1915); see also Gulf, Colo. & Santa Fé Ry. Co. v. Ellis, 165 U.S. 150 (1897) (striking down a statute that allowed attorney fees against railroad company defendants but not against other defendants).

      289 See Louisville Gas & Elec. Co. v. Coleman, 277 U.S. 32 (1928).

      290 See Truax v. Corrigan, 257 U.S. 312, 334, 337 (1921). But see id. at 352 (Pitney, J., dissenting) ("Doubtless the legislature . . . concluded that in labor controversies there were reasons affecting the public interest for preventing resort to the process of injunction and leaving the parties to the ordinary legal remedies . . . .").

      291 See Railway Express Agency Inc. v. New York, 336 U.S. 106 (1949).

      292 See Board of Educ. of Kiryas Joel Village Sch. Dist. v. Grumet, 512 U.S. 687, 714-15 (1994) (O'Connor, J., concurring in part and concurring in the judgment).

      293 Chip Lupu, on reading a draft of this Article, suggested that one advantage of an exemption regime administered entirely by courts is that it's proper for courts (but not legislatures) to sometimes make judgments based on the characteristics of each particular sect. Under this view, Wisconsin v. Yoder may properly be read as applying only to the Amish, based on their particular track record, and not to all religious parents who want to stop schooling their children at age 14.

      But to the extent that Yoder creates such a sect-specific exemption, I think it's mistaken: The Supreme Court, like any other government institution, has no business discriminating among sects and distinguishing the wholesome from the suspicious. See, e.g., Thomas v. Review Bd., 450 U.S. 707, 715-16 (1981). At most, the Court may set forth neutral criteria governing when an exemption should be granted (for instance, a parent may withdraw a child from high school only if there is convincing evidence that the child is likely to live his life in a community where a high school education isn't particularly valuable), something that legislatures can equally do.

      294 If the proposal I make supra Part II.D is adopted, the court would just decide whether or not tenants have a private right to equal treatment without using the language of "compelling interests."

      295 Legislators might, for instance, limit the exemption to marital-status discrimination cases on the grounds that tenants have no private right to equal treatment without regard to sexual orientation or marital status, that it is nonetheless improper for landlords to act in a way that makes it very hard for people to get housing, and that discrimination in fact makes it very hard for homosexual couples to get housing but doesn't make it that hard for unmarried heterosexual couples. Some might think this sort of preference for sexual-orientation discrimination claimants over marital-status discrimination claimants is politically unlikely, but if the state banned sexual-orientation discrimination in the first place, then it's likely to be fairly sympathetic to sexual-orientation discrimination claimants. A legislature in such a state might plausibly and without much controversy conclude that homosexuals face more discrimination because of their homosexuality than unmarried or married or widowed or divorced people face because of their marital status.

      296 But see Laycock, supra note 7.

      The possibility of amendments to RFRA should be reassuring to a Court that doubts its own capacity to authoritatively balance competing interests, but it is also the weak spot of the legislation. . . . Congress should resist such amendments; it should recognize that RFRA is a quasi-constitutional statute that should not be lightly amended.

      Id. at 254.

      297 See 42 U.S.C. § 2000e-1(a) (1994).

      298 See id. § 3603(b)(2).

      299 See, e.g., Assembly Committee on Judiciary, AB 1617 Bill Analysis (visited May 31, 1999) <http://www.leginfo.ca.gov/pub/97-98/bill/asm/ab_1601-1650/ ab_1617_cfa_19980113_170132_asm_comm.html>, at 12-13 (discussing proposed exceptions--ultimately not included in the final bill--for antidiscrimination laws and zoning and land-use rules).

      300 See Lupu, supra note 107, at 598 (arguing that "prison cases should be made subject to an explicitly different standard than non-prison cases, so that the results in the latter are not dragged down by the interpretations in the former"); Daniel J. Solove, Note, Faith Profaned: The Religious Freedom Restoration Act and Religion in the Prisons, 106 Yale L.J. 459, 471-72 (1996) (describing congressional debate on this issue).

      301 I thus disagree with the view expressed by Doug Laycock, among others, that a RFRA "can work only if it is as broad as the Free Exercise Clause, establishing the fundamental principle of religious liberty and leaving particular disputes to further litigation." Laycock, supra note 7, at 221.

      302 This discussion does not rely, for instance, on judgments about governmental attempts to ensure diversity of viewpoints through racial or religious classifications. See generally Eugene Volokh, Diversity, Race as Proxy, and Religion as Proxy, 43 UCLA L. Rev. 2059 (1996).

      303 These reactions need not be wrong: Ignorance of others' worldviews may be quite rational--we can't learn everything about everything, especially about views that we may correctly think are somewhat kooky or at least inconsistent with our best understanding of what's right. My claim here is only the descriptive one that we often are ignorant of others' views, or at least of the details of those views, the justifications for those views, and the strength with which the views may be held.

      304 See, e.g., Quaring v. Peterson, 728 F.2d 1121 (8th Cir. 1984), aff'd by an equally divided court, 472 U.S. 478 (1985).

      305 Lochner v. New York, 198 U.S. 45, 75-76 (1905) (Holmes, J., dissenting).

      306 I refer here to a basic federal son-of-RFRA, as opposed to the more complex federal alternatives, which I describe in Part III.F.2, that do provide for state-by-state variations.

      307 See, e.g., Desnick v. ABC, 44 F.3d 1345 (7th Cir. 1995) (rejecting such a claim); Special Force Ministries v. WCCO Television, 584 N.W.2d 789, 793-94 (Minn. Ct. App. 1998) (accepting such a claim).

      308 See, e.g., Cal. Penal Code § 632 (West 1998) (forbidding a person from recording his conversation with another if the other doesn't consent and if that person wants the conversation to remain private); Special Force Ministries, 584 N.W.2d at 792-93 (treating certain kinds of surreptitious videotaping as an actionable trespass).

      309 See, e.g., Neil A. Lewis, Radio Writer Pleads Guilty in Child Pornography Case, N.Y. Times, July 7, 1998, at A10 (describing a reporter's claim that "he was trafficking in lewd photos of minors on the Internet as part of his First Amendment right to research an article on the subject" and that "the only way for [him] to penetrate the secretive world of fanciers of child pornography was to pretend to be one himself," and the trial court's rejection of this argument on the grounds that "[a] press pass is not a license to break the law").

      310 See, e.g., D.C. Code Ann. § 22-3852 (1998) ("A person commits the offense of blackmail, if, with intent . . . to cause another to do . . . any act, that person threatens . . . [t]o expose a secret or publicize an asserted fact, whether true or false, tending to subject any person to hatred, contempt, or ridicule . . . .").

      311 See, e.g., Cohen v. Cowles Media Co., 501 U.S. 663 (1991) (rejecting a proposed news-gathering exception from contract law); Zurcher v. Stanford Daily, 436 U.S. 547 (1978) (likewise as to the duty to turn over documents); Branzburg v. Hayes, 408 U.S. 665 (1972) (likewise as to the duty to testify); Le Mistral, Inc. v. CBS, 402 N.Y.S.2d 815, 817 (App. Div. 1978) (same); Prahl v. Brosamle, 295 N.W.2d 768, 780-81 (Wis. Ct. App. 1980) (likewise as to trespass).

      312 Of course, as with the Free Exercise Clause, if I were persuaded that the text or the original meaning of the provision did protect news-gathering practices that violated private rights or imposed improper externalities, then I would have to endorse courts' carving out constitutional news-gathering exemptions from generally applicable laws.

      313 For a more detailed discussion of this three-pronged analysis, see supra Part I.C.

      314 See, e.g., Desnick v. ABC, 44 F.3d 1345 (7th Cir. 1995) (Posner, J.).

      Investigative journalists well known for ruthlessness promise to wear kid gloves. They break their promise, as any person of normal sophistication would expect. If that is "fraud," it is the kind against which potential victims can easily arm themselves by maintaining a minimum of skepticism about journalistic goals and methods.

      Id. at 1354.

      315 See, e.g., Shulman v. Group W Prods., Inc., 955 P.2d 469, 493 (Cal. 1998).

      316 See, e.g., Cal. Const. art. I, § 2(b); Cal. Evid. Code § 1070 (West 1998).

      317 See Private Property Rights Protection Act, Fla. Stat. Ann. § 70.001 (West 1998); Private Real Property Rights Preservation Act, Tex. Gov't Code Ann. § 2007.002-.043 (West 1999); Kimberle E. Dodd, And Justice For All: The State Experience With Property Rights Legislation (Competitive Enter. Inst. Envtl. Studies Program, 1998).

      318 Abraham Lincoln, Address at a Sanitary Fair, Baltimore, Md. (Apr. 18, 1864), in Abraham Lincoln: His Speeches and Writings 748-49 (Roy P. Basler ed., 1946) (emphasis in original). Today, the idea that liberty could include the right to own slaves sounds ridiculous, but of course in Lincoln's time the Supreme Court had held that the Due Process Clause secured such a right, and made it unconstitutional for Congress to ban slavery in a territory. See Scott v. Sandford, 60 U.S. (19 How.) 393, 450 (1857).