Libertarianism presents itself as a simple, clear, and principled view. It appears to provide a moral basis, in the value of individual liberty, for a specific political program of limited government and low taxes. The moral significance of liberty seems obvious even to those who believe it is not the only thing that matters. But the claim of the libertarian political program to be founded on this value is illusory. Three lines of thought lead to conclusions that might be seen as libertarian. But none of these shows that respect for the value of individual liberty should lead one to support the political program of low taxes and limited government that libertarians are supposed to favor.


One route to libertarian conclusions appeals to an idea of productive efficiency. As Hayek argued, the market is, in an important range of cases, a more efficient mechanism for deciding what to produce than decisions by any central planner. This is so for two reasons. The first is the flow of information: no planner could acquire information about what consumers want to buy as efficiently as the market does. The second is capture by interests: decisions by state-owned industries are likely to be guided by the interests of those who run or work in those industries rather than by the goal of efficient overall production. Where they apply, these arguments are powerful. As recent financial crises show, however, these considerations do not lead to the conclusion that government regulation is always a bad thing. And even Hayek would not deny that government intervention is needed in the case of externalities such as pollution and climate change. The considerations just mentioned provide some guidance about how to deal with these problems, but they provide no reason for thinking that they should be dealt with by simply leaving it to the market.

Whatever policies they support, however, these considerations are not based on the value of liberty for an individual. This argument assigns individual liberty only an instrumental value: it is important only as a means to economic efficiency. “Efficiency” sounds important. But efficiency is only as important as the goal that is efficiently promoted. The value that Hayek’s argument takes as fundamental is the satisfaction of individual preferences. More specifically, it is the satisfaction of preferences that can be expressed through the market, the weights given to these preferences being determined by individuals’ willingness and ability to pay for their satisfaction. Since individuals with more money are willing to pay more for the satisfaction of a given preference, this means in practice that what is maximized is the satisfaction of preferences weighted by the wealth and income of those whose preferences they are.

Many of the factors affecting the degree of control individuals have over their lives—such as the legal system and the organization of the economy—are not the subject of preferences expressed through the market. To the degree that they are not, market outcomes will not be sensitive to the value individuals place on their own liberty. For example: The productive efficiency of a market economy depends importantly on its ability to shift resources from industries that are no longer needed or efficient—such as typewriters manufacturers in an era of the computer—to those making products for which there is greater demand—such as computers and software to use on them. This efficiency is attained at a cost to workers, who must find new employment when such changes occur. Workers who are constantly subject to such disruption have less control over their lives than they would in a more stable society. To determine what system is to be preferred, some decision must be made about how to balance the conflicting values of productive efficiency and individuals’ control over their lives. The market itself does not answer this question, since the choice between different systems is not something that individuals express a preference about through their market behavior.


A second, quite different view is what might be called “motorcycle-helmet libertarianism,” which gives fundamental place to the value of having control over how one’s life goes in important respects. The idea of control that this line of thinking appeals to is not a right but a value—something that individuals have reason to want. The importance of the difference between rights and values is demonstrated by an argument of Robert Nozick’s. In Chapter 8 of Anarchy, State, and Utopia, Nozick considers, as a possible objection to his view, that in a society of the kind he is recommending some people would lack control over their lives in important respects. In a skillful rhetorical move, he responds to this objection by asking whether there is “a right to have a say about what affects you,” and he quickly and convincingly shows that there is no such right. As he puts it, few things affect your life more deeply than whom you marry. So a right to have a say over what affects you would include a right to have a say about whether your beloved will marry someone else, thereby becoming unavailable to marry you. But clearly you have no such right.

An unregulated market leaves many workers with little control over their lives. Their liberty also matters morally.

What this argument shows is that the objection Nozick is considering should not put it in terms of a supposed right. It does not show that having control over one’s life in certain respects is not an important value that needs to be taken into consideration in deciding what rights people have. Indeed, Nozick himself seems to appeal to such a value when he says that his system of “libertarian” rights is appropriate for us because we are “distinct individuals each with his own life to lead.”

The distinction involved here is one of several that can be referred to, somewhat misleadingly, as between positive and negative rights. As I have said, however, it is not a distinction between two kinds of rights but between rights and considerations that must be taken into account in justifying them. The lesson to draw from it is not that there are no “positive rights”—rights to particular benefits—but rather that not every desirable thing that is relevant to justifying rights can be directly transformed into a “right to” realize that thing.

Recognizing control as an important moral value leads to the question of what system of rights—what set of laws and policies—would best secure this important form of control for everyone, since everyone counts morally. It may seem to industrialists that an unregulated market provides the greatest freedom, because regulation and taxation reduce their ability to do what they want. But as I have mentioned, an unregulated market leaves many workers with little control over some important aspects of their lives, and their liberty also matters morally. So an argument appealing to the moral importance of control over one’s life must take both of these facts into account, along with others.

If we ask what conditions are most important for having meaningful liberty—meaningful control over one’s life—in a modern society, one of the first things that comes to mind is education, which enables one to understand one’s choices and to acquire the skills needed to pursue them, including the skills needed participate in the market economy. A second important factor is a strong social safety net, including unemployment benefits, which enable people to plan responsibly for having a family despite the uncertainties of employment in an efficient market economy. Neither of these is part of the “low taxes and limited government” program normally favored by libertarians. Perhaps a revised libertarianism might incorporate these policies, along with other measures needed to give meaningful liberty to all. As commonly understood, however, the libertarian political program should seem responsive to the value of individual liberty to only one group of people: those who believe that they have no need of such policies in order to exercise control over their lives, and imagine that no one else needs these things either (or else that it does not matter whether they have them).


It may be said that what is objectionable about laws requiring motorcyclists to wear helmets is not that when such laws are in force people lack control over important aspects of their lives. The objection is rather that such laws deprive people of control over their lives in a particular way, by coercively telling them what to do. So the basis of libertarianism might be taken to lie in the idea that no one should be coercively told what to do. As stated, this way of putting the objection is overly broad. Enforcing any law involves coercively telling people what to do. Certainly this is true of property laws, which libertarians favor. So the libertarian idea is a narrower one, that no one should be coercively told what to do as long as he or she is not violating the rights of others.

This intuitively appealing idea is the third route to libertarian conclusions, which starts not from the value of control over one’s life but from an idea of non-interference, given content by an enumerated list of rights. Since this idea is to serve as a test for the legitimacy of laws and social institutions, it is important that the rights in question be “natural” in the sense of not depending, for their own validity, on the legitimacy of institutions that establish them. Foremost among these rights claimed by libertarians are property rights and rights not to be subjected to force or violence. If every government action beyond the protection of such rights is objectionable coercion, then respecting these rights may seem to lead to policies of limited government and low taxes favored by libertarians.

The question, then, is why we should think that, independent of any law or social institution, people have these rights, and only these rights, and that these rights are the only basis of justified coercion. The strongest reason for thinking this seems to me to be that the existence of these rights seems to be the best explanation of the wrongfulness of certain actions that do seem to be clearly wrong, independent of any law or institution. Being attacked by a murderer, or being captured and enslaved are good examples of such wrongs. So does having the crops that one has raised in order to live through the winter taken away by a band of armed marauders. Examples of this kind seem to support the idea that there is a natural right to property.

Property rights require an institution that creates, defines, and enforces them, and is justified by the benefits it brings to all.

This brings us to a second interpretation of the distinction between negative and positive rights: Negative rights are rights not to be interfered with; positive rights are rights to be provided with certain benefits. A mechanism of enforcement is a positive benefit, whatever rights are being enforced. But the idea is that coercive enforcement is legitimate only if what is being enforced is simply that people not interfere with others who are not interfering with them.

But property rights go beyond mere rights to non-interference. We can see the difference by looking more closely at the example of the crop-stealing marauders. To explain why their action is wrong, we do not need to appeal to a right to property. The wrong is adequately explained as a violation of a narrower right not to be interfered with. To put the matter in Lockean terms: we have the right to act on the things of the world in order to preserve and improve our lives, as long as, in so doing, we do not encroach on others’ ability to do likewise. Others ought not to interfere with our doing this, and if they try to do so we are justified in using force to prevent them. The assumption in this example is that clearing land and growing crops in order to survive did not encroach on anyone, hence it is wrong for others to interfere with this.

These ideas of rightful use, wrongful interference, and rightful defense account for what Locke called natural property rights. But property rights as we commonly understand them are much stronger. They involve not only the right to use the things one owns, and to exclude others from taking them, whether or not we would not suffer from this loss. And property rights also include the power to give others similar rights over a thing, by transferring it to them.

By permissibly using something, I can make it wrong for you to take it, on Lockean grounds, because you would be interfering with my use. By ceasing to use it, and leaving it with the intention that you will use it, I can make it the case that you will not wrong me by using or destroying it. These ideas, included in the right of non-interference as I have construed it, are extremely plausible. But it is extremely implausible to think that I can, by an exercise of my will, confer upon you the right to exclude anyone else from the use of a thing, and give you the power to transfer this right to yet other people. Having this power would make me an odd kind of moral legislator. As David Hume argued, property rights require an institution that creates, defines, and enforces them, and is justified by the benefits it brings to all affected by it. It follows that if there are property rights that can be coercively enforced, justifiable coercion is not limited to the enforcement of “natural” rights. So a rights-based idea of mere non-interference does not provide a foundation for libertarian politics.


There are no property rights independent of some institution defining them, but it is generally agreed that there should be such an institution. The question is what form this institution should take. So-called “defenders of property rights” are best understood as arguing that it would be in some way better—more conducive to economic productivity, for example—for our institutions to define these rights in one way rather than another.

The threshold question here is how property rights must be defined in order to be justifiable to all who are required to accept them. Above that threshold there is the question of which of the various justifiable systems of property rights we should most prefer under current conditions. Considerations of productive efficiency—more specifically, the considerations of information flow and interest-group influence that I mentioned in Section 1—clearly have a role in determining the answer. So does the value, mentioned in Section 2, of individuals having control over their lives. Libertarians are correct in calling attention to these considerations, although there is no reason to believe that they are the only things that are relevant to deciding what form our institutions should take. Moreover, since these two ideas are distinct, and do not always point in the same direction, it is misleading to lump them together under a single heading of concern for “liberty.”