A. The challenge of mutual association

No matter what else they do, institutions of property have to help make mutual association beneficial for everybody. Those benefits cannot be taken for granted, because the very possibility of mutual association requires explanation. Everybody needs both freedom to go about life and freedom from interference, but going about life almost always interferes with somebody else. In this respect, a scheme of property rights is like a system of traffic rules,^{Footnote 4} which determines (say) when pedestrians are to be free to keep walking even if it interferes with drivers, and when drivers are to be free to keep driving even if it interferes with pedestrians. Clearly, it is impossible to eliminate interference altogether; more than that, the freedom to interfere is an invaluable resource in social life, so that the only question is how that resource may best be allocated.^{Footnote 5} No matter what else it does, the institution of property has to help make mutual association possible by balancing freedom to interfere with freedom from interference.^{Footnote 6}

George Bramwell made that balance explicit in Bamford v. Turnley (1862),^{Footnote 7} in which one neighbor sued to stop another from manufacturing bricks on his land. Each neighbor wanted the very same things: freedom to interfere, and freedom from interference. The plaintiff wanted freedom to forbid a smoking kiln, even if forbidding it annoyed his neighbor; the defendant wanted freedom to bake bricks, even if his neighbor found the smoke annoying. We share society only by agreeing on tradeoffs between two such freedoms, and these just are the freedoms that constitute property: a Hohfeldian liberty to act even if it interferes with someone else (equivalently, no obligation not to interfere), and a claim against the interference of others (equivalently, an obligation on the part of others not to interfere).^{Footnote 8} The purpose of institutions of property is to balance these claims and liberties—these rights of way—so that mutual association enriches the freedom to go about one’s life, at the least cost of lost freedom from interference.

Bramwell’s opinion in the case identified exactly that purpose:

Bramwell perceived that institutions of property just are for striking a balance between claims against interference and liberties to interfere, by the simple device of giving neighbors equal standing before the law. Any claim of A’s against B will be a like claim of B’s against A; so the less A allows B to enjoy some liberty, the less can A enjoy a like liberty. The upshot of equal standing is that each neighbor wins only by that balance that benefits every neighbor, “a rule of give and take.” So when A is to be afforded no freedom to make bricks, the justification is not that the benefits redound to B but that the benefits redound even to A, by protecting mutual association, where neighbors need be no more than “trifling nuisances” to each other.

Whatever else a society’s institutions of property may do, protecting mutual association is what any such institutions must do. It’s not uncommon for theories of property to start with what a community, once there, might aspire to do. Our approach starts with the recognition that first there has to be a community.

B. Meeting the challenge

Bramwell’s central insight is what I shall call the principle of reciprocal benefits:^{Footnote 10}

The phrase “in the greater scheme of things” adds several things. One concerns time: benefits and harms in the greater scheme of things must be contrasted with any benefits or harms in vacuo. Bramwell recognized that while A would always prefer to do without B’s nuisance today—whether B is the one baking bricks or the one trying to prohibit A’s baking—there are nonetheless some nuisances that eventually even A cannot afford to be protected from. That is why some nuisances must be “ordinary” rather than actionable: in vacuo A may wish he could enjoy complete silence from B, for instance, but given equal treatment, A would have to forgo in the greater scheme of things all the valuable things that he and others might do only by making noise. The reason to limit A’s freedom from B’s interference is that such a limit creates a balance of freedoms for all that is a vastly greater benefit even for A, who faces a future in the sort of community that such a balance of freedoms makes possible.

A further point concerns generality. Not everyone needs to make the same amount of noise; but the question of optimal balance is not how much a particularly quiet neighbor would lose by giving up the freedom to create trifling nuisances.^{Footnote 11} The question is how much one would lose by living in the sort of community where freedom to be a nuisance was limited by the need for that freedom on the part of its least tolerant member. (Likewise, mutatis mutandis, for a community that limited freedom to prohibit nuisances by the need of its most tolerant member.)

Lastly, “in the greater scheme of things” also refers to our interdependence. Ultimately what A wants is the most valuable liberty to go about his life that he can hope for, and so the reason to put up with B’s interference is not just that this increases A’s own freedom as well, but also that it increases freedom for everybody else, all of whom are then at liberty to do and to create the sorts of things that can give A hope of a vastly more attractive future. We put up with the tradeoff of liberties and claims both because of the prosperity that humans create only in multitudes, and because sharing society is enriching for its own sake. Mutual association is what can make it valuable to be at liberty to go about one’s life, and so it is our interconnectedness that gives each of us reasons to allow the rest of us as much liberty as we can afford.

The principle of reciprocal benefits identifies a justificatory burden that institutions of property must meet, whatever else they may also be justified in doing. More precisely, the principle of reciprocal benefits reveals when an institution is a bad one. The question is not whether an institution makes a community so bad that it is not worth belonging to at all, though. A community can withstand many a bad institution and still be worth belonging to, taking all things together. Rather, the principle of reciprocal benefits helps us identify those institutions about which the best that we can say is that we might withstand them, even as they reintroduce some measure of the very problems of shared existence that property is meant to help solve.

C. Incidents of property

Property is not things but freedoms: freedoms to act without interference, even if one’s actions interfere with another. But what freedoms are these? Or, more precisely, what freedoms are the incidents of property? The answer, as should now be clear, is that the incidents of property are whatever they have to be in order for institutions of property to create reciprocal benefits within mutual association. And so the incident of property that is easiest to see is the freedom to restrict access to what is owned (ius possendi). It is, after all, in virtue of this freedom that there can be property at all.^{Footnote 12} However, in order for persons to do and to create things of value with their property, for themselves and their neighbors, property must also include the freedom to use what one owns (ius utendi), as well as the freedom to dispose of it (ius abutendi), as for instance by transferring it. A property right is always some configuration of freedoms of these three types.^{Footnote 13}

Persons have property rights in their labor in exactly this sense. By labor I don’t mean “sweating toil,” except by coincidence; I mean the human capacity (“the strength and dexterity of one’s hands”) to reconfigure factors that might advance one’s aims into resources that actually do.^{Footnote 14} A right against expropriation of that capacity is freedom to restrict access; a right to engage in work is freedom to use that capacity; and a right to exchange services with others is freedom to dispose of that capacity.

Distinguishing possession from use and disposition will be important for understanding the nature and variety of takings. For instance, the taking of an easement across an owner’s land, as for the construction of a common road, is more precisely a taking of the owner’s freedom to restrict access to that strip of land, in exchange for the market value of that freedom. But where the joint venture is to preserve a scenic view, owners may be forced to contribute the freedom to build on their land (a right of use); where the joint venture is to preserve the peace of a residential neighborhood, owners may be forced to contribute the freedom to transfer their land to commercial users (a right of disposition).

And so the distinction between possession, use, and disposition will also be important for framing restrictions on labor, such as occupational licensing. In particular, observe that rights of use and disposition can be taken even without any seizure or dispossession. Of course, the sort of taking that is easiest to grasp is the condemnation of the right of possession; land can be seized, labor can be forced. But even when property is in a thing, that property can be taken whether that thing is seized or not: rights of use and of disposition can be taken without any taking of possession, and such rights are property. Crucially, the rationale for limiting takings of possessory rights is the same for limiting takings of rights of use and disposition. And crucially, it is use and disposition—specifically, the liberty to do work and to make one’s work available to others—rather than possession that are at issue in cases of occupational licensing.

D. Scope of property

Institutions of property balance claims against interference with liberties to interfere. The principle of reciprocal benefits states what justifies the rules that constitute institutions of property, including the public power to take property. So the justification for taking is that even those from whom property is taken are better off in the greater scheme of things for living in the sort of community that such a power to take property makes possible.

This justification can be applied to the question of where the claims and liberties that constitute property must not extend, and therefore what arrangements of rights are not to count as takings, in the first place. The dangers of being underprotected from interference with one’s actions are easy to see, but Bramwell’s insight was that since protection is reciprocal, in the greater scheme of things overprotection is dangerous too. In fact, it is this observation that made Bramwell’s opinion memorable for introducing the legal concept of “ordinary nuisance.” Neighbor A may wish today that he could forbid B’s annoying outdoor cookouts, but A could not afford such protection in the greater scheme of things. Neighbor B’s cookout, though no doubt a nuisance, must be regarded as only an ordinary nuisance against which A is to have no claim in that time and place. More broadly, every institution of property must admit such a thing as interference that is not actionable wrong, or what is known in the common law as damnum absque injuria.

The other side of this coin, of course, is that property cannot extend so far as liberty to create “extraordinary nuisance”—those sorts of interference from which neighbors must be able to expect protection. If A would be overprotected by freedom from B’s ordinary nuisance, so too would A be overprotected by freedom to create extraordinary nuisance.

Generalizing, in order to create reciprocal benefits, institutions of property must assign each person no greater a claim against interference than will benefit him or her in the greater scheme of things. Put another way, each of us is owed protection against overprotection. Arranging rights so as to provide such protection is therefore no taking of property. It is no taking when A is barred from using his property in ways that put filth onto B’s property; it is no taking when A is barred from contracting with B to sell his services as a hit-man.

E. Protection of property

The principle of reciprocal benefits also applies to how claims against interference are to be protected where they do extend.^{Footnote 15} Different protections alter the balance of claims against interference with liberties to interfere. A property-rule entitles A to refuse to transfer the protected freedom to B and, equivalently, forbids its transfer to B without A’s permission. By contrast, a liability-rule transfers the protected freedom from A to B for compensation as determined by public authority. The takings power just is the power—the freedom to alter other freedoms—that the public authority exercises when it transfers property by way of liability.

It’s tempting to say that property is only as real as it is absolute—and in particular, as it is immune to takings—but a moment’s thought shows why that cannot be so. Along with everybody else, A is put at some risk if B is allowed to drive a car; by licensing B anyway, the public authority transfers A’s freedom from that risk to B without having to get A’s permission, provided that B compensate A, as the public authority deems appropriate, in the event that B harms A. A liability-rule offers less protection than a property-rule, but more protection is not always better: here, A would be overprotected by a property-rule, because the costly negotiations between every A and every B would preclude vehicular travel, the benefits of which swamp the associated risk in the greater scheme of things, even for A. To create reciprocal benefits, institutions of property must assign each person no greater protection of property than will benefit him or her in the greater scheme of things.

Of course, less protection is not always better either. The reason to protect A with a liability-rule is to reduce transaction costs that would otherwise preclude transfers that A would lose from forgoing. But if it could be determined ex ante that it is A in particular that B risks injuring, then B should have to negotiate with A for the transfer; neither A nor B can afford to live where a neighbor can be forced to sell his or her claim against injury to whichever neighbor desires to purchase it. A liability-rule would therefore underprotect A, and for the same reason it would extend B’s property further than even B could afford. So in order to create reciprocal benefits, institutions of property must also assign each person no less protection of property than will benefit him or her in the greater scheme of things.

The standard rationale for occupational licensing is the same as that for licensing drivers. Occupational licensing is an arrangement of property rights, such that one may not use or transfer one’s labor in certain ways without first obtaining formal public permission in the form of a license, in the name of improving quality of service.^{Footnote 16} Permission may depend on academic credentials, hours of training, licensing exams, bonding, and so on, often with a licensing fee.^{Footnote 17} In theory, licensing provides the public good of protecting the public from poor services from which they cannot feasibly protect themselves.^{Footnote 18} For instance, patients cannot check the credentials of EMTs in the midst of an emergency; patients can check the credentials of a family physician, but third parties they might infect cannot.^{Footnote 19} Of course, beyond some ceiling on risk A becomes worse off, and so licensing provides the public good of a reasonable ceiling on that risk. Even A is better off if B may be licensed; even B is better off if he has to be licensed.

Licensing therefore bears the same justificatory burden as all other property institutions do, a burden described by the principle of reciprocal benefits. In particular, there are two things that together justify such takings. First, it matters that the stakes are high. In vacuo, A would always rather avoid B’s nuisance; but in the greater scheme of things there are some nuisances of B’s that A should have no protection from at all, for A’s own sake. Stakes are “high” when, as Bramwell observed, even B would be worse off for an institution that permitted neighbors to create such nuisances—that is, when the nuisance is in this sense “extraordinary.” Incompetence that poses some severe threat is no ordinary or “trifling” nuisance, though, and it is for this reason that A is to be protected from B in the first place. And second, it matters that transaction costs are high, as when patients don’t know in advance which EMTs will attend them, and third parties don’t know which patients may be carrying poorly treated contagion. This explains why A’s protection from B is by liability-rule rather than property-rule, which would leave A and B to negotiate the transfer for themselves, to the detriment of both.

F. From property to takings

As promised, I have not offered a theory of property. The approach to takings that I’ve sketched here starts from two ideas: one, that property is not things but freedoms, so that to alter those freedoms just is to take property; and two, that the justification for taking property is to benefit those from whom property is taken, in the greater scheme of things (the principle of reciprocal benefits). This is not a theory of property, but a theory about what a theory of property has to be. Even so, appreciating what institutions of property have to do is enough to help us understand what the takings power has to be. Restrictions on labor are takings of property in labor, and the burden of justifying such takings is the same as it always is.

A. The challenge of the takings power

The takings power just is the power that the public authority exercises when it transfers property by way of liability. By a power is meant the freedom to alter other freedoms; for A to lack such a power as against B is for B to have an immunity.^{Footnote 20} So we must say, more precisely, that property is a liberty to interfere plus a claim against interference, together with some configuration of powers and immunities with respect to altering such liberties and claims. By definition, all transfers of property, including takings—transfers by legislative fiat—just are transfers of such claims and liberties.^{Footnote 21} So, since the public takings power just is a power to alter property, the configuration of relations that constitute institutions of property in a given society depends at the fundamental level on the scope of that power.^{Footnote 22}

The reason to vest such power in the public authority at all is that a property right may be more valuable to its owner if it is protected with a liability-rule, owing to the sometimes prohibitive costs of making mutually beneficial transfers of freedoms protected with a property-rule (Section 1.E). But by the same reasoning the takings power cannot be a license to take; the power that lowers the risk of overprotection also increases the risk of underprotection. The power to alter property invites abuse, whether as a means of doing favors for cronies or of lowering the price of implementing a well-meaning ideal. Put another way, the purpose of the takings power is to provide public goods, yet such use of that power is also a public good: we would all benefit from it, but incentives to provide it are rare.^{Footnote 23} If our problem in the absence of a takings power is overprotection, underprotection is our problem in the face of it.

For this reason there is an important asymmetry between property- and liability-rule protections of property generally: the prohibitive cost of creating a mutually beneficial rearrangement of rights is the special reason to protect with a liability-rule, whereas property-rule protections need no special reason at all. There is a presumption of property-rule protection;^{Footnote 24} a liability-rule is appropriate only where a claim against interference would be overprotected by a property-rule, given the costs of transacting for a transfer of that claim.

As there is a presumption of property-rule protection, so there is a presumption against takings, which is to say that a taking is appropriate only for special reasons—namely, when transaction costs are such that a property-rule would overprotect the holder of property in the greater scheme of things. And so the Fifth Amendment to the Constitution of the United States says, “nor shall private property be taken for public use, without just compensation.” The Fifth Amendment acknowledges the necessity of a special clause for liability rules rather than property rules—the necessity, that is, of a takings clause and not a “property clause.”

The takings power solves a problem, but it is also itself a problem for strong limitations to solve.^{Footnote 25} What institutions of property a community turns out to have will depend on a public takings power as constrained by such limits,^{Footnote 26} and so takings face the justificatory burden of creating reciprocal benefits by ensuring that those from whom property is taken are better off in the greater scheme of things, owing to the institution that licenses the taking. This justification can be applied to the five chief questions about the scope of any takings power:

Let’s consider the first two questions first, which concern appropriate occasions for exercising the takings power (Sections II.B and II.C) and in licensing in particular (II.D–II.E). We can then consider the remaining three questions, concerning compensation, in Section III.

B. What liberties are to count as property in the first place?

The answer to this first question depends on what liberties cannot be property, because of the threat they would pose to shared interests in general health and safety. Mutual association benefits everybody insofar as no one is to be at liberty to act with “injury to his neighbour.” Such prohibitions are not takings, since ex hypothesi there is no such liberty to be taken.

Some public acts that might look like takings—such as the seizure of a power plant when this is necessary to prevent contaminating a water supply^{Footnote 27}—are therefore not takings at all. In U.S. constitutional law such acts are known as exercises of the “police power,” that is, the authority to prevent serious threats to public health and safety. The appearance of a taking by the police power is illusory, however, because that power prohibits only what one was not at liberty to do in the first place.^{Footnote 28} Of course, if there is great risk of underprotection from nuisance in the absence of the police power, there is also great risk of overprotection when the police power is interpreted too broadly, forbidding nuisances that would have created even greater benefits in the greater scheme of things.

C. For what reasons may property be taken?

The takings clause of the Fifth Amendment to the U.S. Constitution limits the takings power to takings “for public use.” The rationale for the public-use test is that there must be special reasons to take,^{Footnote 29} which rule out transfers of property from A to B for the private benefit of B. For that reason, takings are justified only as necessary^{Footnote 30} to provide public goods, like roads and common-carrier utilities.^{Footnote 31} Takings for public goods are consistent with the rationale of taking from A, not just to benefit B, but to benefit even A. Takings for other purposes, though, cannot be so justified: insofar as goods that come from them are not public, such takings are transfers from some for the benefit of others.^{Footnote 32}

Of course, a lot depends on when a reason is “special” or a use is “public.” What should be clear is that a taking—protection by a liability-rule—is justifiable when property would be overprotected by a property-rule. A half-century of lugubrious Supreme Court decisions notwithstanding, public use cannot be so “broad and inclusive” as to be the same as “a conceivable public purpose.”^{Footnote 33} The justification for taking is never that the public authority thinks the world would be a better place if B had what is A’s.^{Footnote 34}

D. Licensing as taking

Restrictions on labor, injury to neighbors aside, are takings of property in labor. But several points require clarification before we proceed. First of all, it is no taking not to guarantee employment, even though employment can increase the value of property in labor. Recall that the principle of reciprocal benefits is defensive: it defends A from transfers to B except insofar as such transfers benefit A. The principle never entitles B to a transfer from A so as to benefit B, which would make property the very problem it is meant to help solve. So, however much B’s not finding work with A may be harmful (damnum) to B, it can be no wronging (injuria). It is a taking, not when the public authority refuses to make zero-sum transfers, but when it does not allow positive-sum transfers.

Second, in treating licensure as a taking there is no assumption that workers have any claim to work within the particular field to which licensing restricts entry. The freedom that is taken in cases of occupational licensure is again not a claim to work, but a liberty to employ one’s labor, “harm to one’s neighbour” aside, without let or hindrance.

Finally, and crucially, we must also appreciate that when there is an argument against restricting A’s labor, the argument has nothing to do with what options A would have had in some hypothetical baseline to which he is entitled. Rather, the argument is that everyone would enjoy the reciprocal benefits of living in a society made possible by an arrangement of property rights within which A would have greater liberty to employ his labor.^{Footnote 35}

E. Licensing and the public-use test

Now, the standard rationale for occupational licensing is that even though restrictions on the use and disposition of labor can raise prices by restricting entry into an occupation, nonetheless certain restrictions make everyone better off in the greater scheme of things as those higher prices are offset by superior services.^{Footnote 36} And that is to say, in the language of the Fifth Amendment, that licensing is justified insofar as it is a taking of property for public use.

Unfortunately, takings-by-licensing in the United States increasingly fail to pass the public-use test. In the United States, licensed workers have gone from 5 percent of the workforce in the 1950s to about 25 percent today,^{Footnote 37} and only about one-third of the expansion is attributable to changes in the nature of work.^{Footnote 38} And while prices have risen alongside licensing in the United States,^{Footnote 39} in general improvements in the services provided haven’t followed.^{Footnote 40} Worse, high prices and low provider density lead some consumers to lower-quality substitutions (including do-it-yourself) and some to doing without, resulting in worse services actually received.^{Footnote 41}

The reason licensing has done so little to benefit the public is that takings in this area of public law have seldom been for public use. The common practice of “grandfathering” incumbents against new standards doesn’t help if stakes are high;^{Footnote 42} the common practice of state-specific licensing actually raises transaction costs.^{Footnote 43} And neither stakes nor transaction costs can explain where licensing has actually expanded: licensing to thread eyebrows (Louisiana), to braid hair (Arkansas, Iowa, Missouri, Washington), to sell caskets (Alabama), to coach dieters (Florida), or to broker room-shares (Pennsylvania).^{Footnote 44}

These takings of property in labor all share a combination of low stakes and low transaction costs. Where the stakes are low, protecting consumers by licensing risks raising prices by more than the protection is worth to consumers. Where transaction costs are low, it is better left to consumers and providers to negotiate tradeoffs of credential and price. Not only is there no strict public-use basis for licensing in these circumstances—which creates no public goods but is, if anything, a public nuisance. There is not even a loose public-use basis: licensing serves no “conceivable public purpose” that providers couldn’t achieve by obtaining voluntary certification, say, and that consumers couldn’t achieve by scrutinizing provider reputation.^{Footnote 45} Left to their own devices, consumer A and provider B can each factor the value of the credential to themselves into their negotiations for the price of the service. Outside the public-use situation, both providers and consumers would be better off without any liability-rule transfer of liberty from provider to consumer, which underprotects providers and overprotects consumers.

A. What form is compensation to take?

It is important to see, first of all, that compensation for takings can be either in cash or in kind.^{Footnote 46} The road that connects A’s farm to the marketplace may be worth at least as much to A as the easement taken, so that the resulting road is in-kind compensation for the taking.^{Footnote 47} Further cash compensation may overprotect A, by making the road harder for the public to provide.

Compensation in cash and in kind are not mutually exclusive, since in-kind benefits may be less than the full value of the taking.^{Footnote 48} That said, too strict an accounting of A’s gains and losses can overprotect A too. For a like taking, B may benefit more than A does; C may benefit for no taking at all. Even so, when the marginal benefits of takings remain very large, A will do better not to get in the way. Rectifying differential gains and losses is costly, and is worth doing only as marginal benefits diminish.^{Footnote 49}

B. When is compensation for takings to be made?

Since taking from A is supposed to benefit A, all takings should be compensated. But it is important to see that A’s neighbor B also benefits from a restriction that makes it less perilous for A to do, to create, and to possess things that other people value, which is how humans create prosperity. Even when a taking benefits the whole public, some member of the public must pay disproportionately more. Turning it into a grave misfortune for anyone to have what enough others want is a practice that makes losers of all of them in the greater scheme of things.

Universal compensation also protects the community from takings that move resources to lower-valued uses; even good intentions can destroy wealth when they don’t have to pay their own way. Consider the 1992 Supreme Court case Lucas v. South Carolina Coastal Council.^{Footnote 50} When the Council rezoned beachfront property to prevent new construction, the combined value of Mr. Lucas’s two undeveloped lots fell from $1,200,000 to almost nothing; the Supreme Court found that Lucas was owed compensation under the takings clause. Question: if the rezoning—which just is an alteration of protected liberties—was for public use, should the public have to compensate Lucas for what the public had (let’s suppose) good reason to do?

Yes, as the sequel of the case illustrates. After purchasing Lucas’s two lots, the Council sold them—to a housing developer—for $392,500 each, even though a neighbor had offered $315,000 for one of the lots in order to keep it undeveloped. The Council’s purpose may have been public use, but when public money was on the line, the Council decided $77,500 per lot was too much to pay for that use, even after arguing before the Supreme Court that $600,000 per lot was not too much when Mr. Lucas’s money was on the line.^{Footnote 51} Even for public-use takings, a universal compensation rule therefore serves as a safeguard against moving property rights from higher- to lower-valued uses.^{Footnote 52}

The reason for a takings power is, again, to make everybody better off for the licensed transfers, and the reason to compensate for takings is that that way of licensing transfers is what it takes to make everybody better off in the greater scheme of things. Put another way, even A can benefit from an institution that transfers freedom from A to B, but even B loses from an institution in which A is not compensated for the transfer.

C. What is the appropriate level of compensation?

The reasons to compensate are also reasons to compensate completely. Providing public goods may benefit everybody (an in-kind compensation), but any residual cost of provision is still a taking. For the same reason, the cost of a taking must be reckoned at the market value of what is taken, as it was prior to the taking; otherwise, the power to compensate below market value becomes a power to take at a discount at someone else’s expense.^{Footnote 53}

D. Licensing and compensation for takings

Recall that the standard rationale for licensing is that it can pay its own way, compensating everybody—even workers excluded by the license—with enormous in-kind benefits of improved general health and safety. To be sure, licensing has high administrative costs, it raises prices, and it makes providers scarcer; but the transfer of certain freedoms to use and dispose of labor from workers to consumers moves them to where they are ultimately more valuable for everyone. The case for this rationale is strong for workers like Emergency Medical Technicians (EMTs), where nothing short of licensing would create those benefits. But when states can require cosmetologists to complete an average of 372 hours of training, as compared with 33 hours for the average EMT,^{Footnote 54} it is no surprise that licensing yields few benefits in quality. It is difficult to deny that in the end, such licensing for no public use is a net transfer from both consumers and excluded workers to protected incumbent workers and salons. To the extent that any in-kind benefits of the taking are insufficient compensation for the taking, the licensing requirement transfers freedoms to use and dispose of labor from higher- to lower-valued uses.

This too has been the pattern in the United States.^{Footnote 55} Licensed workers often earn a substantial premium over their unlicensed counterparts, even after controlling for other variables.^{Footnote 56} Worse, such premiums are regressive: premiums rise disproportionately in higher-paying occupations, and meeting licensing requirements is hardest for the least advantaged, such as poor workers and workers with a criminal record.^{Footnote 57} Unlicensed workers also face lower rates of employment and employment growth, compared with licensed workers performing similar work.^{Footnote 58} Some licensing requirements protect only those incumbents in a given state, by requiring minimum residency periods, or by rejecting licenses from out of state. These requirements make it harder to move for work,^{Footnote 59} which is especially burdensome for military spouses and immigrants.^{Footnote 60}

Where the stakes are low, where transaction costs are low, or both, transferring liberty from workers to consumers via licensing is a taking of property from excluded workers for at best insufficient in-kind compensatory benefits. Where stakes are low, the taking is a net loss for the excluded worker, even if transaction costs are high. Where transaction costs are low, consumers and providers can negotiate tradeoffs of credential and price, even if stakes are high.

Unfortunately, the legal treatment of licensing in the United States has been no more satisfying than its handling of the takings clause. Licensing routinely transfers wealth from consumers and excluded workers to incumbents, but lawmakers are seldom held accountable for such transfers, which usually go unseen and are always off-budget. The fact that it was in the 1950s that licensing began to expand throughout the United States is probably no accident, considering the Supreme Court’s 1955 decision in Williamson v. Lee Optical of Oklahoma, Inc., which upheld an Oklahoma law restricting the making, repair, and fitting of eyeglasses to licensed ophthalmologists and optometrists.^{Footnote 61} The Court’s decision established a general deference to state legislatures to regulate work, requiring states to show only some rational basis for the regulation.^{Footnote 62} And states can almost always discharge this burden: for instance, in Niang v. Carroll (2018) the Court of Appeals for the 8th Circuit upheld a Missouri law requiring cosmetology licensing for hair-braiders, accepting with a straight face that the law “furthers legitimate government interests in health and safety.”^{Footnote 63} Such lawsuits have had little success, whether tried under the due process and equal protection clauses of the Fourteenth Amendment^{Footnote 64} or under antitrust law (the Sherman Act);^{Footnote 65} and in general, licensing requirements once passed have been very difficult to repeal.^{Footnote 66}

E. Use, disposition, and compensation

If charges of arbitrariness and favoritism won’t stick in cases like these—cases of meritless takings of rights of use and disposition of labor—it is hard to see what would, short of a substantial shift in judicial reasoning.^{Footnote 67} But for that very reason, it is important to work toward, if not a comprehensive, then certainly a unified understanding of the justificatory burden for all legislative transfers of property, whether that is property in land or whether it is “the property which every man has in his own labour.”

Unfortunately, it is with respect to use and disposition of property that U.S. takings doctrine is especially in a shambles. The problem, in a word, is that the Supreme Court has no coherent framework for recognizing takings of use and disposition, as opposed to dispossession, and so treats restrictions on use and disposition only when they are so egregious as to be equivalent to dispossession. The failure began with the Court’s very first effort on such takings, in Pennsylvania Coal v. Mahon (1922).^{Footnote 68} Pennsylvania Coal owned the right to mine underground, and H. J. Mahon owned the right to build on the surface. However, Pennsylvania Coal also owned the so-called support rights—that is, by contract the surface rights included no right against subsidence—whereas a new Pennsylvania law required mineral owners to guarantee support to surface owners, that is, to transfer their support rights, and thereby to bear enormous restrictions of their right to mine.^{Footnote 69} In his opinion for the majority, Justice Holmes wrote that the Pennsylvania law violated the takings clause because that law had gone “too far.” But while Holmes’s opinion sounds like a victory for protecting use, in fact it was a defeat: going “too far” meant restricting use so far as to wipe out so much value as to be tantamount to dispossession. Ever since, the Court has interpreted taking as dispossession only, and interprets restrictions on use and disposition as takings only when they are equivalent to dispossession. So, the Court’s position has been that as long as restrictions of use and disposition leave an owner with some scintilla of value, there is no dispossession, and therefore no taking, and therefore no compensation.^{Footnote 70} Short of a total wipeout, rights of use and disposition are in the public domain.^{Footnote 71}

I have not tried to develop a theory of takings, but only to give some idea of how takings are framed when our focus is on what has to obtain for an institution of property, including takings, to create reciprocal benefits. It is an advantage of the approach to property and takings I have sketched here that, despite its humble bases, it is determinate enough to explain what is muddled about the Court’s approach over the last century.