Author: Maloberti, Nicolás
Date published: April 1, 2011
Journal code: NDRV
My object in this article is to answer the following question: Can we accept the fundamental tenets of classical liberalism and at the same time support the state's raising of immigration barriers? I argue that if we accept these tenets as essentially correct, we should regard immigration barriers as essentially illegitimate.
I do not believe, however, that a direct appeal to individuals' property rights or other fundamental rights, such as the right to associate, is enough to establish such a conclusion because under certain conditions it is permissible to infringe on individuals' rights. Therefore, we must determine whether such conditions validate the raising of immigration barriers. Moreover, we run the risk of ignoring important issues if we focus exclusively on how things should be in an ideal world, regardless of how things are in the world in which we actually live.
Classical liberals believe that the state's role should be limited to some basic functions, such as protection of property rights, enforcement of contracts, and national defense. Under current conditions, however, virtually all existing states have gone beyond such limits, with a corresponding extension of their taxing power. In particular, states impose severe limitations on the use of individuals' property in the form of regulations, licenses, antidiscrimination laws, and so forth. They also bring about massive wealth transfers in the form of corporate subsidies and welfare benefits. From a classical-liberal point of view, these actions create nonideal conditions. The thought that under certain conditions of injustice we should be allowed to act in ways that would otherwise be morally impermissible is not implausible. We consider, then, whether we can make a classical-liberal case for immigration barriers given the injustices inherent in current institutions.
Classical liberals tend to support government's limitation to certain basic, minimal functions. They also tend to stress the importance of institutional restrictions that prevent government from extending its power. Classical liberals have offered alternative rationales in support of their conception of the proper scope and limits of government power, ranging from hedonistic forms of utilitarianism to uncompromising forms of deontology. The most convincing of such rationales, however, share a common ground.
Individual Rights: Guarantors of Sovereignty
The highlight of the ethical component of the most attractive versions of classical liberalism is the fundamental importance assigned to individuals' capacity to lead their own lives, where this capacity is taken to require an allocation of rather stringent and extensive areas of moral freedom in which others should not interfere. Classical liberals tend to reject certain patterns of moral reasoning that would subordinate the individuals' capacity to lead their own lives to the will of other individuals, to the promotion of social benefits, or even to the promotion of the same individuals' self-interest and moral good. Classical liberalism is in this sense an individualistic philosophy. It empowers the individual in a way that communitarian and nationalistic philosophies do not.1
An assignment of ownership rights, both over one's body and over external objects, is generally taken to play a fundamental role in the concrete representation of that commitment to individual sovereignty. The basic thought is that an individual's capacity to lead his own life would be seriously truncated should others have a right either to move or to block the movements of his body or to interfere with the use of the external resources the individual needs in order to achieve virtually any of his purposes. Under a system of collective ownership over external resources, for example, no individual has the liberty to use or possess such resources without everybody else's approval. Whether individuals are capable of leading their own lives thus depends on factors completely beyond their control. If we are genuinely concerned with individuals leading their own lives, we must leave room both for their making all sorts of decisions about how to use their bodies and for their appropriating and using external resources without having to secure anyone's approval. Of course, this reasoning does not entail that individuals should not be allowed to develop jointownership arrangements in a cooperative manner. The point is that individuals must be recognized as having an original, unacquired right not to be precluded from their private acquisition and discretionary disposition of external resources.2
This sort of consideration in favor of private rather than collective ownership, however, is indeterminate among different regimes of private property. It is with regard to this indetermination that the distinction between classical and modern liberals can be made. Modern or egalitarian liberals do favor private rather than collective property. Modern liberalism, in its most prevalent form, tends to maintain that the distribution of external resources should not depend on morally arbitrary factors, such as an individual's superior genetic endowment or his family's socioeconomic status. In general, egalitarian theorists claim that "undeserved" income may be coercively redistributed to compensate for the misfortune of others when this misfortune is similarly undeserved.3 More specifically, they tend to propose a redistribution of the unequal income that results from individuals' unequal circumstances. They do not propose to redistribute the unequal income that results from individuals' different choices.4
Classical liberals, in contrast, tend to disagree with the claim that just because there is no moral reason that individuals' circumstances ought to be as they are, these circumstances are of no moral significance in deciding, for example, whether a particular individual or somebody else should have exclusive control over the income stemming from the use of his talents.5 We may agree that a significant proportion of Lionel Messi's income can be traced to his inborn talents. But classical liberals would remain unconvinced that this fact shows that somebody other than Lionel Messi has a legitimate claim over that income. Moreover, the connection between self-ownership and ownership of external resources might be tighter than modern liberals tend to assume. So it is unclear whether the alleged redistributable nature of the income generated by the use of individuals' superior genetic endowment, for example, is compatible with granting individuals an exclusive right to decide how to use that endowment, a right that is required by any genuine concern for individual sovereignty. If there are reasons why it makes sense to expropriate coercively the talented person's income that stems from the use of his natural talents, for example, why do those same reasons not support coercing the talented person to prevent his attempt to stop producing that income?
Classical liberals have argued on different grounds in support of the system of strong ownership rights they favor, so it would be inaccurate to claim that they regard individuals' rights as "natural rights" if we take this notion to refer to a particular sort of normative rationale. However, the notion of natural rights is also understood in reference to the independent normative status that such rights are taken to have in relation to certain contingent social facts. In this sense, it is uncontroversial that classical liberals regard individuals' rights as "natural." Individuals' ownership rights are not taken to be contingent on any sort of social or cultural recognition. Classical liberalism, then, is clearly incompatible with the idea that the mere fact that a legal system is in force in a certain society is a reason for considering it morally justified. Classical liberalism is also incompatible with the view that the fundamental moral values protected by a system of ownership rights are relative to culture and that individuals may be deprived of such values if the current cultural practices allow it.
A "Constrained" Conception of Human Nature
If the notion of individual sovereignty captures classical liberalism's strongest ethical commitment, Thomas SowelPs (2007) notion of a "constrained" vision captures its most fundamental understanding of the nature of man. According to SowelPs conception of human nature, the moral limitations of man in general and of his egocentricity in particular are neither lamented nor regarded as things to be changed. They are treated rather as inherent facts of life.
According to a "constrained" vision, as Sowell explains, the fundamental moral and social challenge is to make the best out of the possibilities that exist within the bounds set by those inherent facts of Ufe. To expend energies in an attempt to change human nature is seen as basically foolish and usually dangerous. Characterizations of individuals as purely self-interested, in the sense of being concerned only with their own well-being, certainly do not have empirical support. For example, individuals follow moral norms, and under certain conditions those who follow such norms end up worse off than they could have been had they simply ignored them. Furthermore, individuals sometimes go beyond what those norms demand, making significant sacrifices of their own well-being for the sake of others. A "constrained" vision of human nature, however, need not claim that individuals are purely self-interested. The most plausible understanding of classical liberalism's conception of human nature is one that depicts individuals as predominantly self-interested.
Gregory Kavka provides an illuminating explanation of this notion, analyzing the thesis of predominant egoism in terms of four propositions (1986, 64-80). The first proposition is that for most people in most situations the altruistic gain/personal loss ratio needed to motivate self-sacrificing action reliably is large. The second proposition claims that the number of people for whom altruism and other non-self-interested motives override self-interested motives is small. The third proposition posits that for the average person the number of situations in which non-self-interested motives override personal interest is small. Finally, the fourth proposition holds that the scope of altruistic motives that are strong enough to override self-interest is for most people equally small - this is the sense most naturally associated with David Hume's notion of limited generosity. According to Kavka, an individual's generosity "is confined to concern for family, close friends, close associates, or particular groups or public projects to which the individual is devoted" (1986, 65).
A "constrained" conception of human nature does not merely include a particular understanding of human motivation as predominantiy self-interested. It also includes a particular understanding of individuals' capacity to deliberate and act on those motives. In this respect, individuals are not regarded as fully rational and perfectly infallible. Yet they are not regarded as essentially irrational. If we referred to the previous account of human motivation as one of predominant egoism, we might refer to this account of human action as one of predominant rationality. Individuals are regarded as having the capacity to realize what their ultimate motives actually are, to deliberate about the alternative actions available to realize such motives, and to act on the basis ofthat deliberation. Individuals are regarded, for the most part, as having the capacity to foresee how effective their actions will be in bringing about their goals. Individuals often make mistakes, but they also learn from them, especially when the costs of such mistakes, measured in terms of the sacrifice of their own goals, are high.
As Adam Smith observed, each individual has "a principle of motion of its own" ( 1982, 234). According to a constrained conception of human nature, such a principle is both easily discernible and rather ubiquitous. People care about their own self-interested well-being and for the most part act in ways that tend to promote it. As Smith famously declared, it is not from the benevolence of the butcher or the baker that we expect our dinner, but rather from their regard for their own self-interest. A constrained conception of human nature is fully aware of the important implications that follow from Smith's insights. Should the pursuit of the butcher and baker's self-interest be constrained in regard to their producing and selling their products on their own terms, other individuals might end up with no dinner in prospect.
Coercion and Government
A simple but fairly definite institutional conclusion follows from classical liberalism's commitment to the value of individual sovereignty and its constrained conception of human nature: individuals should be allowed to use force as a means of protecting their private spheres of moral freedom against others' opportunistic behavior. Classical anarchists believed that coercion was virtually always morally impermissible. Peter Kropotkin ( 2002) claimed that any law inflicting penalties is an abomination that should cease to exit. Michail Bakunin (1953) argued that even when the state commands the good, it undoes that good by commanding it. Classical liberals do not share these beliefs. Self-interest has a "dark side" that needs to be constrained. In fact, it is as a result of the constraints on such dark-side actions that we expect the butcher and baker to produce and trade their goods. Within a constrained conception of human nature, it is only by changing the corresponding relative payoffs that entrepreneurial activities can be reallocated from predation to production and trade. Coercion is of major importance in producing that change.
An influential trend in classical liberalism, most prominently represented in the writings of John Locke, holds that in the absence of the state, individuals have exclusive control over the coercive means necessary to secure their ownership rights. Thus, in addition to those ownership rights, individuals also have enforcement rights - that is, rights of control over a range of actions whose purpose is to deter, rectify, and punish others' infringing actions. This particular way of understanding the permissibility of the use of coercion is certainly consistent with classical liberalism's core commitments. If individuals' capacities to lead their own fives should not be made contingent on anyone else's will, the individuals' capacity to protect themselves from unjustified aggressions should not be contingent on public officials' decisions to perform their alleged functions. In particular, in the absence of government individuals should not be expected to abstain from punishing their offenders because such abstention will make them more likely targets of future acts of aggression.
In general, however, regardless of whether any special emphasis is placed on the individualistic, natural character of people's enforcement rights, classical liberals tend to believe that the government should administer such coercive constraints and that doing so should be its primary and virtually sole function. If, following Locke, we believe that individuals have a right to enforce their own rights, the question of the state's legitimacy involves the question of how the state can acquire a right to monopolize the administration of individuals' enforcement rights. It also involves the question of why exactly it is permissible for the state to infringe on individuals' ownership rights.
The state threatens its subjects with harm should they decide to exercise the full range of their enforcement rights. Private individuals are not allowed, for example, to inflict just punishment on those who violate their rights. But the state also taxes its subjects to fund the costs involved in the collective provision of the enforcement services it does not allow individuals to provide privately. If private individuals have rights to their persons and property, rights that are taken to constrain government's legitimate activities, should they not be free to decide whether their rights are to be protected, if at all, by somebody else rather than by the public officials? Why should individuals not be free to allocate whatever amount of their own resources, if any, for purposes of their protection?
The most plausible classical-liberal answer to the problem of political legitimacy appeals to the nonabsolute character of individuals' rights.6 What distinguishes classical liberalism from the alternatives is not an alleged belief in the absolute nature of individual rights, but rather the stringency of the conditions imposed on the legitimate infringements on those rights. Classical liberals allow for rights infringements only in extraordinary circumstances: in general, only if the extraordinary nature of certain circumstances renders such infringements necessary to secure the benefits that those same rights provide in normal circumstances. If a well-functioning system of justice administration is taken to constitute a public good in the sense that markets cannot supply it, then classical liberals will regard the necessary infringements justified insofar as the rights people have are simply not worth having without a system that prevents and rectifies violations to such rights.
If we are committed to the value of individual sovereignty, and if we believe that individuals are predominantly self-interested and rational, we surely will think that an assignment of enforceable and distinct spheres of jurisdiction to each individual is of fundamental importance. Coercive barriers raise the cost of prédation and thus decrease its profitability in relation to production and trade. Individuals are therefore better able to lead their own lives. If this same concern leads us to justify a compulsory, monopolistic administrator of such coercive barriers, such as a state, a new problem will certainly arise: that of constraining the relatively selfish and rational individuals who compose the state.
Supplying the Defect of Better Motives
Classical liberals do not assume that public officials will work for the public interest. On the contrary, following Hume, they believe that in politics "every man ought to be supposed a knave" ( 1987, 42). This basic assumption guides the construction of the "distrusting" institutional arrangement that classical liberals favor. The goal of such an institutional arrangement is, in James Madison's words, to supply "the defect of better motives" (Hamilton, Jay, and Madison  2001, 269). The assumption that motivates this sort of institutional arrangement does not differ from the assumption that leads us to support the coercive enforcement of individual rights in the private sphere. We expect the butcher and baker to produce and trade their goods rather than to prey on others' production only when the net benefits of predation are lower than those of production. Production, in contrast to predation, tends to benefit others. Public officials' best opportunities to advance their own personal well-being are ideally also made coincidental with what is best for the advancement of the "common good."
The main such formal institutional arrangements include constitutions and bills of rights, which provide a focal point that facilitates the collective action of the governed and allows for commitment by public officials; periodic elections of representatives, which generate a degree of responsiveness from public officials to the citizens' will; and separation of powers, under which the state is divided into branches with different areas of responsibility, each branch having a capacity to check abuse by the other branches. Federalism, however, is the sort of institutional arrangement that deserves special mention in the context of a discussion of immigration barriers.
A federalist organization rests on a territorial division of authority that neither the subunits nor the center can alter unilaterally. The political subunits' powers ideally extend to less than the full extent of the economy in which resource movement is free of trade barriers; the center is supposed to guarantee the absence of such barriers. In defense of this splitting of sovereignty, appeals to the same sort of idea involved in the doctrine of separation of powers are not uncommon. Constitutional allocations of powers to subunits as well as the increased opportunities for citizen participation in public decision making are taken to protect individuals from the center. The center's authority to intervene in subunits is taken to protect individuals against subunit authorities. Despite the undeniable importance of these sorts of considerations, the promotion of jurisdictional competition must be regarded as federalism's most important feature in constraining government.
Because political subunits have no power to tax or regulate the whole area in which trade is free, a federal arrangement enhances resource owners' ability to move away from detrimental taxes and regulations. As James Buchanan points out, "The categorical difference between market and political interaction lies in the continuing presence of an effective exit option in market relationships and its absence in politics" ( 2001, 69). Federalism, as Buchanan himself argues, can be understood as the institutional arrangement that is closer to the market ideal by allowing for the workings of competition.
As in the market, individuals retain an exit option to the extent that they can shift among the separate political jurisdictions. Inefficiencies in law enforcement, regulatory favors, and excessive fiscal burdens - all expected outcomes of a monopolistic state run by "constrained" human beings - are kept in check by competitive pressures. In markets, individuals shift firms when dissatisfied with a product. This shift is a purely self-interested action by which individuals defend their own welfare. Yet the unintended by-product of consumers' decisions to switch may induce recovery of declining firms. What markets do for firms, federalism does for political units such as states and provinces. Even citizens who might never consider migration are protected by those who are sensitive to the differential political treatments in each political unit. The possibility of observing such differential treatments by itself exerts discipline on potentially deviant political units.
Immigration Barriers under Ideal Conditions
From a classical-liberal perspective, the question of the moral status of immigration barriers is basically the question of whether such barriers belong in that "distrusting" sort of institutional arrangement taken to realize the value of individual sovereignty given the constraints imposed by human nature.
As already mentioned, however, the distinction between "ideal" and "nonideal" theory seems to be particularly relevant in this context. The distinction attempts to capture two different sorts of arguments made in support of immigration barriers (or of any other policy, for that matter). At one level, a "nonideal" justification of immigration barriers appeals to the existence of other policies or restrictions that are considered unjustified, such as, in the case of classical liberalism, an extensive system of wealth redistribution. Immigration barriers are taken to be justified insofar as they prevent a further extension of such a system and by doing so prevent a greater number of illegitimate transfers. References to past injustices and the consequences of such injustices might also be understood as nonideal considerations. An "ideal" justification, in contrast, regards immigration barriers as justified regardless of these sorts of considerations. In general, a nonideal form of justification can also appeal to the limitations of what individuals can be expected to know or do. To a great extent, however, classical liberalism treats human limitations as aspects of the human condition that need to be accommodated within a theory of the ideal.7
In the next section, I consider whether immigration barriers are consistent with the previously identified normative and empirical considerations that constitute the common ground of classical liberalism (or at least of its more attractive versions) when assessed under ideal conditions. I later explore the case of nonideal conditions.
Immigration Barriers as Rights Infringements
We might think that immigration barriers are merely a manifestation of the right of association and its corresponding right to exclude others. Christopher Wellman writes, "[J]ust as an individual has a right to determine whom (if anyone) he or she would like to marry, a group of fellow-citizens has a right to determine whom (if anyone) it would like to invite into its political community. And just as an individual's freedom of association entitles one to remain single, a state's freedom of association entitles it to exclude all foreigners from its political community" (2008, 110-11). From a classical-liberal perspective, groups of individuals, such as clubs, do have such rights of association that entitle them to exclude others. Contrary to what this line of argument presumes, however, the state does not qualify as a club in a morally relevant sense. In the case of immigration barriers, the state prohibits individuals from deciding whether to exercise the right to exclude others without those individuals' having agreed to abide by any sort of collective decision.8
Immigration barriers rest on the threat of sanctions that public officials issue against citizens who engage in many kinds of transactions with foreigners that are entirely permissible when engaged in with their fellow citizens, such as hiring for employment. Immigration barriers also involve, of course, deportation of foreigners to their countries of origin. According to classical liberals, individuals' ownership rights entail the permissibility not merely of using resources, but also of exchanging them by mutual consent. Immigration barriers prevent individuals from doing so, and thus they preclude individuals from the opportunity to improve their fives by using what classical liberals regard as morally legitimate means.
In a classical-liberal perspective, immigration barriers are not seen as infringing on any alleged positive right of foreigners to have an adequate standard of living. Neither are they seen as clashing with an alleged obligation of the state's citizens to aid those in need. Both the egalitarian concern about the moral arbitrariness of the unequal life prospects of individuals residing within different state boundaries and the left-libertarian appeal to the common ownership of the earth are also absent in the classical-liberal case against immigration barriers. The case for seeing immigration barriers as clear instances of infringements on individuals' rights is exactly the same case that classical liberals make regarding any government restriction on trade: government interference precludes individuals from exercising their so-called negative rights. As Michael Huemer (2010) has argued, immigration barriers do not just fail to confer a benefit to foreigners, thereby allowing them to live in poverty or exploitation without causing any of such harm. Immigration barriers are not analogous to a failing of charity. On the contrary, immigration barriers actively restrain people from reaching a market where people are willing to trade something for what they have to exchange.
As mentioned earlier, however, the most plausible versions of classical liberalism do not regard individuals' rights as absolute. The problem with immigration barriers is that individuals are prohibited from exercising their right to associate with or exclude others based on their own judgments. But individuals are also prohibited by the state from exercising their enforcement rights, and they are not allowed to allocate whatever amount of resources they want, if any, for their own protection. In other words, individuals do seem to be prevented from acting within the scope of their rights, and yet we might believe that the state is justified in doing so. The question, then, is: Can immigration barriers be justified by appealing to the same sorts of considerations that may justify government's basic enforcement functions? Can these barriers be seen as justified infringements on individuals' ownership rights?
The conditions of infringement that classical liberals allow are rather stringent. Again, in very general terms, infringements on rights are allowed only when they are necessary to secure an environment in which individuals can enjoy the benefits that those same rights are taken to provide. The continuous fears that immigrants depress wages, displace workers, and threaten the very existence of the nation's culture are certainly at odds with a classical-liberal perspective in general and with the specific rationale that might be advanced to justify any particular infringement on individuals' rights. Classical liberals have been among the most fervent advocates of the elimination of all privileges that groups of individuals secure by imposing legal barriers to entry into the market, thereby artificially increasing the value of their goods and services. And regardless of how much value classical liberals might recognize in the preservation of, for example, any given human language, they will certainly object to language laws such as those in Quebec, which establish, among other things, a consumer right to be informed and served in French. Private -property owners, such as restaurant owners, must surely be free to decide the language in which they will print their menus. Parents must be free to decide the language in which their own children are educated. Justifying immigration barriers on "cultural-preservation" grounds presumably will then have to be seen as equally objectionable. If the preservation of our culture does not justify forcing people to use a specific language, it cannot justify forcing people to limit their economic exchanges with those who share the same cultural traits.
Cultural Homogeneity and the Mobility Constraint
We might believe, however, that cultural homogeneity has great importance in the establishment of a stable and effective system of government and thus in securing the enjoyment of individuals' ownership rights. According to this line of thought, the inherent value of certain cultural practices, such as speaking in French, does not matter as such. What matters, rather, is the instrumental value of some basic shared agreement in enabling individuals to live peacefully with one another. Michael Walzer claims that without immigration barriers, "there could not be . . . historically stable, ongoing associations of men and women with some special commitment to one another and some special sense of their common life." Thus, "[t]he restraint of entry serves to defend the liberty and welfare, the politics and culture of a group of people committed to one another and to their common life" (1983, 39).
How important is this commitment to one another in an assessment of the moral status of immigration barriers under ideal conditions? Must cultural homogeneity be regarded as a necessary condition for the minimal state that classical liberals advocate? We may well doubt that such is the case. After all, liberalism differs from other political philosophies, as Chandran Kukathas notes, "in that it rejects the idea of an organic and spiritually unified social order in which the interests of the individual are brought into perfect harmony with the interest of the community" (2003, 2). Partisans of immigration barriers certainly need not endorse the idea of an "organic and spiritually unified social order." Yet Kukathas's observation is still relevant. One of the central purposes of the strong private-ownership rights that classical liberals defend is that these rights provide a way of dealing with the multiple and sometimes conflicting ends that individuals seek. Classical liberals argue that by allocating jurisdiction over an extensive array of actions to each individual, the need for agreement on the value of those actions is avoided to a significant extent.
Thus, the alleged inherent destabilizing force of cultural diversity is questionable once we understand the nature of the demands that a classical-liberal regimen imposes on citizens. Although it is reasonable to expect that the demands of an extensive redistributive system might clash with individuals' natural identification with those who share certain cultural traits, it is doubtful that any comparable conflict will arise in the absence of such a system. Moreover, a classical-liberal regimen does not support any policy of forced integration, such as racial quotas in employment and education. Under a classical- liberal regime, people are not forced to enter into any exchange against their will. We might expect that the typical social tensions we tend to associate with a multicultural society would be greatly diminished by strict observance of individuals' freedom of association.
Despite the importance of these previous considerations in denying the significance of cultural homogeneity in a classical-liberal regimen, they do not illuminate the most fundamental shortcoming of a case for raising immigration barriers, which is the paramount importance of the mobility constraint that will be forgone by the raising of immigration barriers.
In this sense, the fundamental shortcoming of a classical-liberal case for immigration barriers under ideal conditions is analogous to the public-choice critique of the conventional theory of fiscal federalism. According to this conventional theory, the functions of government should be assigned to different levels in accordance with the spatial properties of the public-good externalities associated with the carrying out of these functions. Such an assignment would be advisable, however, only under a set of "implicit assumptions about politics" (Brennan and Buchanan  2000, 204). Once we recognize people's constrained nature, federalization may be preferable even in the presence of significant externalities, given the competitive pressures gained by ensuring that individuals have available options among the separate taxing-spending jurisdictions. Similarly, even if we grant the positive effects of cultural homogeneity on the quality of governance, we still need to consider, in evaluating the permissibility of infringing individuals' ownership rights by raising immigration barriers, how pubHe officials' opportunities to advance their own personal well-being are affected. Cultural homogeneity can be secured only by reducing competitive pressures among political regimes because immigration barriers raise the cost of individuals' mobility. Yet competitive pressures must surely be considered vital within a "distrusting" institutional system of the sort that classical liberals advocate.
Democracy, although necessary to guarantee peaceful successions, is hardly adequate to constrain what in a classical-liberal perspective are unjustifiable extensions of government involvement. Democracy brings redistributive pressures, and because individual votes lack decisiveness, few individuals find it profitable to invest in the costly process of understanding the unintended effects of popular policies. Extensive systems of subsidies, trade restrictions, burdensome regulations, and high taxes, for example, have important effects on prosperity. Yet individuals wrongly think that such government interventions are able or necessary to increase their collective welfare. Individuals prefer misguided policies simply because they tend to be more intuitive, and interest groups prefer such policies because they benefit from them. Politicians must cater to these preferences in order to win elections.9 In general, deviating from electoral policies is profitable to the politician only as an attempt to strike a balance between doing what the median voter thinks will benefit him and what actually does.10
The indecisiveness of voting rather than the voters' irrationality fies at the root of "democratic failure" - that is, democracy's tendency to adopt and maintain harmful policies. Individuals do not tend to adopt erroneous beliefs when the quality of those beliefs makes a difference in their own lives - for example, when hiring an architect rather than an actor for designing their homes. They tend to do so only when the personal costs of being wrong are low - for example, when voting for policies that economists oppose, but actors support. Because their own votes do not matter, and they will always get the policies that others decide to have, it rarely makes sense for them to engage in the costly process of revising their biases and flawed mental models of the social world.11 If migration were costless, however, individuals would vote with their feet and choose the policies that they themselves want to have. If such were the case, the supply of misguided policies would decrease because people tend to reject erroneous beliefs when those beliefs prove to be clearly harmful in their own lives.
Moreover, in a hypothetical scenario of costless migration, there would be strong incentives for zero or minimal wealth redistribution and for efficient protection of ownership rights. An exit option imposes credible constraints on reduced protection of private-property rights, but also on government's predatory and redistributive nature. With costless migration, any attempt by government to raise taxes on the most productive part of the population in order to redistribute funds to the less productive part would result in outmigration of the most productive individuals to a nonredistributive jurisdiction. Tax and regulatory burdens that are not compensated by benefits in services would create outmigration, and such a threat would force government to pursue policies that attract rather than repel productive individuals. This pressure would be felt especially in nations lacking access to valuable natural resources and therefore dependent for their fiscal needs on their citizens' productive activities. Only under a collusive agreement would political units be able to engage in redistribution, but such agreements are difficult to enforce. The defection of one single political unit would attract productive resources from all of the other units.
Contrary to this hypothetical situation, migration is not costless. There are significant moving costs, and people have personal preferences about location. In the context of this discussion, however, the important point to note is the significant extent to which immigration barriers raise these costs. Under a system of relatively closed borders, only those who have virtually nothing to lose are willing to do what it takes to avoid detection in order to enter other countries and live under conditions of illegality after their arrival. Those who have virtually nothing to lose are not, however, those whose "shrugging" will affect governments' decisions the most.
Within a system of relatively closed borders, each government has greater taxing power than within a system of relatively open borders, and this power is precisely what enables government to extend its actions beyond the functions assigned by classical liberalism. Furthermore, a system of relatively closed borders gives government wide latitude in choosing the quality of the basic functions classical liberals want government to serve. Therefore, it is difficult to believe that the benefits that might be gained by cultural homogeneity will not be offset by diminished competitive pressures on government when assessed along the same dimension: the effectiveness of securing individuals' ownership rights as conceived by classical liberalism.
Immigration Barriers under Nonideal Conditions
I have argued that under ideal conditions immigration barriers constitute infringements of the strong ownership rights that classical liberalism takes to be fundamental in guaranteeing the exercise of individual sovereignty. The possibility of justifying such an infringement on grounds of necessity, such as those taken to justify the very existence of government, lacks persuasiveness. In certain extraordinary circumstances, individuals' rights might be permissibly infringed if such infringement were necessary to procure the sort of environment in which the benefits that those rights provide can be enjoyed securely. Yet immigration barriers are not merely unnecessary to that effect; they are detrimental because they raise individuals' mobility costs and thereby lower the competitive pressures on governments.
The concerns appealed to in nonideal justifications, however, are usually absent in the ideal sort. We might agree that states never should have raised immigration barriers. The reality is, however, that they have done so. Furthermore, governments have also deviated in significant ways from the classical-liberal ideal. Although the demand for an additional examination of the issue based on these sorts of grounds is always legitimate, the appeal to nonideal conditions is also ultimately unable to make a compelling case for immigration barriers. Again, we may acknowledge that under certain conditions of injustice we are allowed to act in ways that would otherwise be morally impermissible. However, once we understand both the nature of such special conditions and the importance of immigration barriers in relaxing the constraints that government face, the case for seeing immigration barriers as justified by nonideal conditions is unconvincing.
Immigration Barriers and Welfare Transfers
In order to assess the plausibility of a case for immigration barriers that takes into account the illegitimacy of certain present arrangements, such as an extensive welfare state, we must first determine how exactly such injustices affect what we take as ordinary moral demands. Which moral demands, if any, may we ignore if necessary to cope with deviations from an ideal of justice? What conditions must be satisfied for us to ignore such demands permissibly? Although a comprehensive account of such a difficult issue lies outside the scope of this article, it is possible to identify certain basic requirements that any plausible comprehensive account must include and to characterize certain uncontroversial necessary conditions under which it is permissible to deviate from ordinary moral demands without claiming that meeting such conditions is sufficient for deviating in such a manner.
In the case under consideration, a recognized infringement of certain individuals' rights (the raising of immigration barriers) is performed as a means of reducing the likelihood of an extension in existing rights violations (the growth of the welfare state). The following three conditions seem to suggest themselves.
First, there must be no "morally innocent" alternative to the alleged preventive infringement that is equally effective in reducing the likelihood of the extension of rights violations. If we can prevent an injustice while acting within the constraints imposed by individuals' rights, nothing can justify overstepping these constraints.
Second, the preventive infringement must be of lesser moral significance than the expected extension of the existing rights violation. We might think that under certain extraordinary circumstances it is permissible to inflict some form of limited harm on innocent individuals if by doing so a catastrophe involving many more individuals will somehow be averted. However, we would not approve of inflicting a significant harm to a comparable number of innocent individuals in order to protect those facing the catastrophe because doing so will merely shift the costs from one group to another.
Third, acting in this preventive manner must not have a considerable dynamic effect in contributing to a perpetuation of the conditions of injustice that are currently faced. The importance of this consideration is evident if we think about governments' common refusal to negotiate with terrorists. Although the terrorists' demands, such as the release of "political" prisoners, is usually considered less significant than the credible threats that terrorists make, such as the killing of innocent individuals, giving in to their demands will only increase the incentive for making such threats. Having a policy prohibiting negotiation will tend to protect innocent individuals from being used in such manner.
In assessing the argument that finks the welfare state to the justification of immigration barriers, we might think, with the foregoing three conditions in mind, that the first condition (regarding the unavailability of an alternative, non-rightinfringing option) is not met, given the possibility of restricting forced exchanges - in particular, welfare payments - to citizens. Yet concerns about the political feasibility of creating a category of "second-class" residents might be warranted. At the moment, I am conceding that in the absence of that possibility the lowering of immigration barriers will indeed increase the likelihood of an expansion of the welfare state. But this concession is debatable. Cultural diversity might lower levels of trust among groups and decrease support for redistribution. In addition, an increasing proportion of noncitizens among the poor might lessen political support for redistribution by making the median voter better off relative to the population as a whole.12 However, assuming that indeed there is no other choice but ( 1 ) to remove immigration barriers and cause an increase in the number of forced exchanges that sustain the welfare state and (2) to create immigration barriers and avoid such an increase, it is doubtful that a plausible case can be made in favor of the satisfaction of the other two conditions stated previously for reasons that have already been advanced to a great extent in the analysis of immigration barriers under ideal conditions.
For those who are prevented from migrating by acts of mutual consent into nations that have established relatively stable institutional systems, immigration barriers usually condemn them to live in nations with defective institutional systems under which individuals lack incentives to save and invest. Although we might have a special obligation of charity to the "nearest and dearest," duties of justice, such as those imposed by the recognition of an individual's rights, are not limited in such a way. For classical liberals, nationality is not a basic moral category. In regard to respecting individuals' rights, the fact that some individuals might be of different nationalities than ours is morally irrelevant. So the injustice that immigration barriers create cannot be justified as a means of precluding an extension of the injustice of the welfare state. The extension of the welfare state is regrettable, yet the raising and maintaining of immigration barriers entail an injustice of a similar significance, at least for those who might otherwise escape from dire situations. Some individuals gain protection from some forced exchanges, but only by shifting a heavy burden, in the form of blocked exchanges, onto other, equally innocent individuals.
The considerations already advanced in support of the importance of relatively open borders in constraining government from acquiring greater taxing power are relevant for assessing the satisfaction of the third condition regarding the dynamic effects of immigration barriers. Interest groups and rent seekers can obtain transfers at the expense of the rest of the population only when there are significant constraints to mobility. Mancur Olson's (2000) distinction between "roving" and "stationary" bandits is helpful in this context. Lacking the prospect of future interactions with their victims, roving bandits will take everything from them. Stationary bandits, in contrast, need to take into account how their actions affect their victims' productive decisions. How much will be available to steal depends on how much income individuals produce. In Olson's words, the stationary bandit must become "a benefactor to those he robs" (2000, 9). A tax rate of much less than 100 percent as well as provision of basic public goods and an environment relatively secure from thefts by anyone but the bandit himself become paramount. In this sense, stationary bandits are preferable to roving bandits. Yet if multiple stationary bandits face moving victims, the victims cease to be treated as such in a very important sense, and they become clients whom the bandits must now satisfy rather than merely rob. The raising of immigration barriers protects stationary bandits from having to deal with this scenario.
If there is no other option available, choosing a lesser injustice might be advisable if by doing so we do not contribute to perpetuating the underlying conditions of injustice. However, granting that the choice is only between (1) a welfare state with closed borders and (2) a welfare state with open borders, it is unclear that we must favor the former option. First, in no clear sense are we choosing a lesser injustice, at least when dealing with immigrants whose conditions in their home country are dire. Second, immigration barriers create an acknowledged, significant increase of individuals' mobility costs and therefore should be seen as a major contributor to perpetuating the existing rights violations that sustain the welfare state.
The Sinking Boat
We might believe that immigration barriers are justified as a means not of preventing an extension of the welfare state, but rather of securing the sort of Uberai institutions that have been established, albeit imperfectly, in the countries to which most individuals want to migrate. The fear is that immigrants will bring with them political and economic attitudes hostile to liberal culture.
In the late nineteenth century, Henry Sidgwick, who thought that free immigration is implicitly assumed in the economic arguments for free trade,13 stated this sort of concern. He thought that the cosmopolitan ideal was the ideal of the future. According to this ideal, the business of the state is "to maintain order over the particular territory that historical causes have appropriated to it, but not in any way to determine who is to inhabit this territory, or to restrict the enjoyments of its natural advantages to any particular portion of the human race." He thought, however, that under present conditions "the casual aggregates that might result from perfectly unrestrained immigration would lack internal cohesion." Furthermore, the "efficient working of the political institutions of different States presupposes certain characteristics in the human beings to whom they are applied; and a large intermixture of immigrants brought up under different institutions might inevitably introduce corruption and disorder into a previously well-ordered State" ( 1919, 309, emphasis added).14
James Buchanan has expressed similar concerns. In a brief article entitled "A Two Country Parable," he considers the following conceptual experiment:
There are two physically separated locational units (call them "countries") that are descriptively identical and also inhabited by equal numbers of humans, who are possessed of identical potential capacities. Throughout the historical period to be considered, there is no economic relationship between inhabitants of the two countries that extend across boundaries. Each unit exists in autarky.
Over some period of time, one of the two countries, call it A, either through an evolutionary process not well understood or though deliberate design and construction of constitutional constraints, comes to be organized in such a fashion that facilitates the production (and potential consumption) of a relatively high level of economic value (as determined by the preferences of inhabitants) and also facilitates relatively high increases in this level of value over time. By comparison, the other country, call it B, although initially equally endowed, does not experience a similar evolution and/or does not undergo a process of successful design and construction of constitutional parameters, effective as measures by the potential for facilitating the generation and growth of economic value. As a result of these disparate histories, the economy of country B produces a relatively low level of value, as measured by the preferences of its inhabitants, and this level increases relatively slowly through time, if at all. (1995, 61)
The question Buchanan poses to himself is the same one we are dealing with here: "Do those who find themselves to be inhabitants (citizens) of A have any obligation, grounded in principles of justice, to allow those who seek entry to become members?" (1995, 61). He claims that the conceptual experiment offers the appropriate framework for answering this question. I believe, on the contrary, that the experiment fails to draw attention to what ultimately matters in dealing with the sort of concerns that both Sidgwick and Buchanan have.
Buchanan claims that the conditions stipulated in his conceptual experiment "offer the most favorable set of circumstances under which potential immigrants might advance claims for relocation based on principles of justice." He argues, however, that even in this idealized setting no such claims can be supported because "[t] he entry of an immigrant into an ongoing social-political-legal-economic order, with a defined membership, an experienced history, and a set of informal conventions, necessarily modifies the structure of 'the game' itself, the complex and ill-understood set of interpersonal and intergroup relationships that generates the pattern of results that are observed by participants" (1995, 62). According to Buchanan, the fragility of those ill-understood parameters that have made some countries relatively rich justifies restrictions on the entry of full members in an ongoing political community.
This sort of concern seems to rely on a highly plausible moral principle. Such a principle will hold, for example, that shipwreck victims may be prevented from reaching a rescue boat and those who already reached it from helping others aboard if such preventions are necessary to keep the rescue boat afloat. Because no one will be better off if the boat sinks, it is difficult to deny that such restrictions are indeed permissible despite the suffering and possible death of those kept in the sea and their lack of responsibility for bringing about their dire condition.
Both the shipwreck story and Buchanan's parable, however, ignore the importance of establishing whether a causal link exists between the expected outcomes that we are exhorted to prevent and the preventive measures themselves. In particular, it is implicitly assumed that acting in the suggested way does not create the same conditions whose prevention is taken to justify acting in such a way. Buchanan's parable, in particular, is silent regarding the origin of the autarky between country A and country B. Yet the same parable might be told otherwise by specifying the reasons for the original autarky in terms of high mobility costs, such as those created by immigration barriers. In that case, it must surely be important that in the absence of such artificially imposed mobility costs it is difficult to imagine how the result that Buchanan envisions might come about in any reliable manner.
Contrary to what is assumed by the shipwreck story, in which shipwrecks are not caused by our decision to prevent more victims from coming aboard, the very existence of immigration barriers creates, perhaps not fully but to a significant extent, the difficulties that propel massive numbers of individuals to migrate away from their home places. In the absence of such barriers, poor governance is kept in check by individuals' willingness to move. In other words, a system of relatively open borders creates the conditions under which constrained individuals would be expected to realize and protect the value of individual sovereignty in a more or less reliable way. Lacking such an institutional global environment, only extraordinary historical contingencies might produce the sort of political culture required to restrain the expansion of government beyond the classical-liberal scope. Thus, although a case might be made for a restricted, gradual elimination of immigration barriers, given Sidgwick and Buchanan's concerns, these concerns can hardly justify a move in the opposite direction because immigration barriers reproduce the situation to which we must appeal as a means of justifying those barriers themselves, and this form of justification surely must be regarded as unacceptable.
Does the foregoing case against immigration barriers stand in a world in which most states are expected to disregard it? Should any particular country lower its own immigration restrictions even when it knows that other countries will not do the same? Perhaps an excuse can be made for certain actions if, given others' unjustifiable behavior, some individuals would have to bear a great burden should they decide to comply with the moral demands that are being ignored. Yet few countries, if any, can excuse their restrictive immigration policies on such grounds. The usual "economic" case for immigration should be found normativer/ relevant in this context. Contrary to what is implied by common trends of reasoning, economies tend to absorb immigrants by expanding job opportunities rather than by displacing native workers, in part because immigrants and native workers are rarely perfect substitutes. Their complementarity results in specialization, which in turn yields an increase in productivity. Yet even if we do not account for such complementarity, it would be surprising if an increase in the labor supply were to produce stagnation and misery rather than an expansion of the overall economy. In this particular regard, an increase in population because of immigration is no different from an increase in population because of human reproduction. Price changes and perhaps a gradual process of adjustment should suffice for coping with concerns regarding a potentially higher rate of population increase because of immigration.
If man does not take the trouble to work, nature condemns him to the suffering of privation. "What means can he adopt to avoid both?" According to Frederic Bastiat, "There remains now, and there will remain, only one way, which is, to enjoy the labor of others" ( 1995, 143). The existence of government does not alter this reality. Although government might reduce the likelihood of private prédation, it also provides an organized, formidable mechanism of rent extraction. If the cost of exit is sufficiently high, predominantly selfish and rational individuals will take advantage of such a mechanism. Immigration barriers raise the cost of exit to such an extent that few individuals can afford to incur it. The state certainly becomes, in Bastiat's words, "the great fiction through which everybody endeavors to live at the expense of everybody else" ( 1995, 144).
Because of this expected outcome, a case for immigration barriers on classicalliberal grounds is unpersuasive. Considered under ideal conditions, immigration barriers constitute an unjustified infringement on individuals' ownership rights because it is difficult to identify a purpose that such an infringement might serve that would outweigh the disadvantages created by eliminating important competitive pressures on governments. Considered under nonideal conditions, the problem is, roughly, that immigration barriers cannot be seen as the choice of a lesser evil in the face of either an expected extension of the redistributive state or an expected threat to liberal institutions. On the contrary, because immigration barriers relax the constraints governments face, they should be seen as a major contributor in creating the conditions for the perpetuation of the sort of political arrangements that classical liberals oppose. If individual sovereignty is to be protected, the state's sovereignty over a particular territory should not include a prerogative to determine who is to inhabit it. I do not claim that there should be no borders or that anyone willing to entry a country should be allowed to do so.15 I claim, rather, that from a classical-liberal perspective there are no good reasons for preventing the entry of those who merely intend to advance their own well-being within the constraints imposed by respect for other people's rights.16
1. Classical liberals tend to believe that usually allowing people to exercise their capacity to lead their own lives is compatible with and indeed enhances promotion of social benefits. Yet this correspondence is not the ground on which the exercise of such a capacity is primarily valued.
2. For an illuminating discussion of this point, see Mack 2010.
3. See, for example, Rakowsky 1991, Dworkin 2000, and Arneson 2006.
4. Some liberal-egalitarian theorists argue, however, that this popular articulation of the egalitarian idea is inadequate. See, for example, Anderson 1999.
5. On this issue, see, for example, Nozick 1974, Miller 1982, Lomasky 1987, and Schmidtz 2006.
6. The state, after all, is not a voluntary association. The problematic nature of the state seems to arise precisely from the fact that it does not infringe on the ownership and enforcement rights only of those who explicitly consent to the corresponding actions.
7. Thus, an ideal theory need not be Utopian. In John Rawls's understanding, an ideal theory of justice gives an account of a "realistic utopia," which is the best we can realistically hope for (1999, 11-23). Allen Buchanan also emphasizes the realistic aspect of his conception of an ideal theory in arguing for a feasibility requirement, according to which a theory is feasible "if and only if the effective implementation of its principles is compatible with human psychology, human capacities generally, the laws of nature, and the natural resources available to human beings" (2004, 61). For a recent discussion of this topic and other aspects of the distinction between ideal and nonideal theory, see Simmons 2010.
8. Wellman acknowledges that the nonvoluntary nature of political states "can raise complex problems for those who would defend a state's right to political self-determination." He argues, however, that there are some "unpalatable implications that follow from denying a country's right to freedom of association" (2008, 112). According to Wellman, if one denies that legitimate states, such as Canada, have such a right, it is hard to understand why it would be wrong for the United States to annex Canada forcibly if an overwhelming majority of Americans wanted to merge with Canada and an equally high proportion of Canadians preferred to maintain their independence. For a response to Wellman, see Fine 2010.
9. For a convincing account of this process, see Pincione and Tesón 2006.
10. On this point, see Caplan 2007.
11. This does not mean, however, that voters cannot revise those biases in nontrivial ways. The decline of racist preferences might be one such example. Guido Pincione and Fernando Tesón's theory of "discourse failure" might help us understand when change is more likely. The authors contrast what they call "vivid" theories with "opaque" theories. Clear examples of vivid dieories are those that rest on zero-sum analysis. Positive -sum analysis and invisible-hand explanations are instances of opaque theories. The former group of theories is easier to understand. The latter group demands a greater effort because they tend to be counterintuitive. Given the indecisiveness of voting, voters tend to endorse the former group of theories radier tíian the latter. The idea of racial equality might have been rejected at some point, but it is not terribly difficult to understand. The case against economic protectionism, however, is not usually grasped easily, and attitudes regarding free trade still rest on pre-Ricardian mercantilist notions that were refuted two hundred years ago. For a detailed account of the phenomenon of discourse failure - that is, "die public display of political positions that are traceable to truth-insensitive processes" (Pincione and Tesón 2006, 17) - see chapter 2 of Pincione and Tesón 2006.
12. This claim is made in McCarty, Poole, and Rosenthal 2006.
13. Sidgwick says, "In order that the advantages of complete freedom of exchange among nations may be fully realized, it is necessary that labour should move with perfect ease from country to country to meet the changes that are continually likely to occur in the industrial demand for it" ( 1919, 308).
14. The concern is not with allowing immigrants who are expected to take part in social practices that entail the violation of individual rights. A straightforward case may be made against the admission of such individuals.
15. For an account of the service that borders provide according to classical liberalism, see Lomasky 2001.
16. Security concerns are valid, but they are often misplaced when brought to bear on immigration. Many more people move from one country to another as tourists than as immigrants, and virtually no one advocates the elimination of tourism. If a person is considered safe to be awarded a tourist visa, it is difficult to see how he can possibly be denied residence on security grounds. On this point, see Kukathas 2004.
Anderson, Elizabeth. 1999. What Is the Point of Equality? Ethics 199: 287-337.
Arneson, Richard. 2006. Luck Egalitarianism: An Interpretation and Defense. Philosophical Topics 32: 1-20.
Bakunin, Michail. 1953. The Political Philosophy of Bakunin. Edited by G. P. Maximoff. New York: Free Press.
Bastiat, Frederic.  1995. The State. In Selected Essays on Political Economy, edited by George B. de Huszar, 140-51. Irvington-on-Hudson, N.Y.: Foundation for Economic Education.
Brennan, Geoffrey, and James M. Buchanan.  2000. The Power to Tax: Analytical Foundation of a Fiscal Constitution. Vol. 9 of The Collected Works of James M. Buchanan. Indianapolis, Ind.: Liberty Fund.
Buchanan, Allen. 2004. Justice, Legitimacy, and Self-Determination. Oxford, U.K.: Oxford University Press.
Buchanan, James. 1995. A Two Country Parable. In Justice in Immigration, edited by Warren Schwartz, 63-66. Cambridge, U.K.: Cambridge University Press.
_____ .  2001. Federalism as an Ideal Political Order. In The Collected Works of James Buchanan, vol. 18: Federalism, Liberty, and the Law, 67-78. Indianapolis, Ind.: Liberty Fund.
Caplan, Brian. 2007. The Myth of the Rational Voter. Princeton, N.J.: Princeton University Press.
Dworkin, Ronald. 2000. Sovereign Virtue. Cambridge, Mass.: Harvard University Press.
Fine, Sarah. 2010. Freedom of Association Is Not the Answer. Ethics 120: 338-56.
Hamilton, Alexander, John Jay, and James Madison.  2001. The Federalist: The Gideon Edition. Edited by George W. Cary and James McClellan. Indianapolis, Ind.: Liberty Fund.
Huemer, Michael. 2010. Is There a Right to Immigrate? Social Theory and Practice 36: 429-61.
Hume, David.  1987. Essays: Moral, Political, and Literary. Indianapolis, Ind.: Liberty Fund.
Kavka, Gregory. 1986. Hobbesian Moral and Political Theory. Princeton, N.J.: Princeton University Press.
Kropotkin, Peter.  2002. Law and Authority. In Anarchism: A Collection of Revolutionary Writings, edited by Roger Baldwin, 195-218. Mineola, N.Y.: Dover.
Kukathas, Chandran. 2003. The Liberal Archipelago. Oxford, U.K.: Oxford University Press.
_____ . 2004. The Case for Open Immigration. In Contemporary Debates in Applied Ethics, edited by Andrew I. Cohen and Christopher Heath Wellman, 193-206. Maiden, Mass.: Blackwell.
Lomasky, Loren. 1987. Persons, Rights, and the Moral Community. Oxford, U.K.: Oxford University Press.
_____ . 2001. Toward a Liberal Theory of National Boundaries. In Boundaries and Justice, edited by David Miller and Sohail H. Hashmi, 55-78. Princeton, N.J.: Princeton University Press.
Mack, Eric. 2010. The Natural Right of Property. Social Philosophy and Policy27: 53-78.
McCarty, Nolan, Keith T. Poole, and Howard Rosenthal. 2006. Polarized America: The Dance of Ideology and Unequal Riches. Cambridge, Mass.: MIT Press.
Miller, Fred, Jr. 1982. The Natural Right to Private Property. In The Libertarian Reader, edited by Tibor Machan, 275-87. Totowa, N.J.: Rowan and Littlefield.
Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books.
Olson, Mancur. 2000. Power and Prosperity New York: Basic Books.
Pincione, Guido, and Fernando Tesón. 2006. Rational Choice and Democratic Deliberation. Cambridge, U.K.: Cambridge University Press.
Rakowsky, Eric. 1991. Equal Justice. Oxford, U.K.: Oxford University Press.
Rawls, John. 1999. The Laws of People. Cambridge, Mass.: Harvard University Press.
Schmidtz, David. 2006. Elements of Justice. Cambridge, U.K.: Cambridge University Press.
Sidgwick, Henry.  1919. The Elements of Politics. London: Macmillan.
Simmons, A. John. 2010. Ideal and Nonideal Theory. Philosophy and Public Affairs 38: 5-36.
Smith, Adam.  1982. The Theory of Moral Sentiments: The Glasgow Edition of the Works and Correspondence of Adam Smith. Indianapolis, Ind.: Liberty Fund.
Sowell, Thomas. 2007. A Conflict of Visions. New York: Basic Books.
Walzer, Michael. 1983. Spheres of Justice. New York: Basic Books.
Wellman, Christopher Heath. 2008. Immigration and Freedom of Association. Ethics 119: 109-41.
Acknowledgments: I thank my colleagues at Liberty Fund for a very helpful discussion of this manuscript. I also thank Ezequiel Spector, Katherine Erbeznik, and an anonymous referee for their comments and criticisms.
Nicolás Maloberti is a fellow at Liberty Fund.