Contemporary Political Philosophy UCSD; Spring 2013 David O

Draft of 5-17-13

PHIL 167: Contemporary Political Philosophy UCSD; Spring 2013 David O. Brink Handout #5: Nozick’s Libertarian Liberalism

Libertarianism is typically a rights-based conception of moral and political philosophy that takes rights to liberties to be fundamental. Traditional libertarians (e.g. Hayek, Friedman, and Nozick) usually derive competitive capitalism and limited government from this starting point, though a group of recent libertarians (e.g. Otsuka, Steiner, and Vallentyne) has been developing a different form of left-libertarianism. We will focus primarily on right-libertarianism, especially the views of Robert Nozick (1938-2002) in Anarchy, State, and Utopia (1974), because he is the philosophically most sophisticated libertarian. Here is a sample of questions we need to pose for libertarianism. What does it mean to take liberty as the basic value? Is liberty the only basic value? Are all liberties equally valuable? Does recognition of the importance of liberty support a laissez-faire economy or limited government? How revisionary is the libertarian's minimal state, and how should this affect our attitude toward libertarianism? Liberals recognize individual rights against others and the state. Libertarianism is one form of liberalism. Is it the most plausible? In discussing worries for Nozick, we need to ask if they are worries for libertarianism per se, or only for right-libertarianism.

NOZICKIAN LIBERTARIANISM According to Nozick, a minimal state can arise without violating anyone's rights; any more extensive state would violate people's rights to liberty. In particular, the state is entitled neither to engage in paternalism nor to redistribute holdings so as to achieve some pattern of distributive justice.

FOR THE MINIMAL STATE Nozick describes a process by which the minimal state can and should arise without violating anyone’s rights.

1. State of Nature (SN) → Cooperative Protective Associations (CPAs) 2. CPAs → Private Protective Associations (PPAs) 3. PPAs → Dominant PPA (DPPA) (= ultraminimal state (US)) 4. DPPA/US → Minimal State (MS)

Following John Locke (1632-1704), Nozick understands SN to include certain rights.

1. Each individual has a right to act as she sees fit provided that she harms no one else nor deprives others of their liberty. 2. Each individual has a right to be free from harm and interference from others. 3. Each individual has a right to punish those who violate the rights of others, and the person whose rights are violated has a right to exact compensation (ASU: 10).

Because individual private enforcement is not always efficient or reliable, CPAs will be formed. But CPAs suffer from some problems (ASU: 12-13).

(a) Each member is always on call; this is likely to be disruptive for all and painful to smaller, weaker members. (b) Members less in need of services will find that they are carrying more than their share of the protective burdens. (c) It is not clear how disputes among members of the association are to be handled.

The obvious advantages of a division of labor and economies of scale lead to PPAs that offer protection policies, much as insurance companies offer protection against loss. But there will be inter-PPA disputes. There are three possibilities.

1. The PPAs will do battle and one will win; 2. The PPAs will adopt different territories; or 3. The different PPAs will submit to the judgments of a common tribunal.

Under each of these alternatives, there emerges, within a given area, a dominant PPA (DPPA) (ASU: 14-17). The DPPA is the ultraminimal state. It differs from MS insofar as it

(1U) protects only those who have purchased policies (not independents) and (2U) allows independents to punish its members (appropriately) (ASU: 22-6).

By contrast, MS would

(1M) protect all those within its domain and (2M) claim a monopoly on the use of punishment.

But the DPPA will give way to MS. For a DPPA will see such independents as imposing undue risks on its clients and so will prohibit them from self-enforcement of their rights. According to Nozick, clients of the DPPA must compensate independents for the way in which the DPPA's prohibition diminishes the deterrence-value of its own procedures. The way to do this, Nozick claims, is for the DPPA to extend protection to independents and for existing clients to subsidize these new policies (ASU: 108-115).

QUESTIONS ABOUT THE JUSTIFICATION OF THE MINIMAL STATE Nozick argues that the DPPA can prohibit independents from using their own enforcement procedures, when it deems these unreliable or unfair, provided that it compensates the independents. But this is impermissible if the DPPA deprives independents of their rights (in the state of nature). The DPPA cannot prohibit independents from self-enforcement without violating their rights -- even if they provide compensation. Compare: I can’t saw off your left leg without your permission, even if I pay you compensation. Even if I pay you compensation, I’ve still violated your right. Indeed, this is why I owe you compensation. But then it looks like the DPPA cannot prohibit independents from self-enforcement without violating their rights -- even if they provide compensation (Nozick 3 recognizes some tension between the principle of compensation and natural-rights theory himself (75), but not this one.) Perhaps it's Nozick's view that the DPPA's prohibition does not violate any right to self-enforcement that the individual has. He might claim that DPPA prohibition does not prevent the independent from self-enforcement, it just diminishes the value of self- enforcement to the independent, because (a) the DPPA will retaliate for the independent's exercise of self-enforcement and (b) the deterrent-value of the independent's threat of self- enforcement will, as a result of (a), be greatly reduced (cf. 110). This defense places great weight on the distinction between having a right and being free to exercise it, on the one hand, and the value of having the right and exercising it, on the other hand. We may wonder how much it means to say that rights are inviolable if their inviolability is compatible with their value being completely eroded. Alternatively, Nozick might grant that the DPPA's prohibition infringes the liberty of independents but insist on a moralized conception of rights, in general, and a right to liberty, in particular, according to which a right to x acts as a side-constraint, not on all interferences with x, but rather with unjustifiable interferences with x. Nozick seems committed to a moralized conception of rights to liberty, in any case, by virtue of his conception of them as near absolute side-constraints. Liberty gets restricted in order to prevent serious harms to others and particular liberties (e.g. to private property) may conflict with other liberties (e.g. to make use of any resource as one chooses). If rights are side-constraints, then it seems they must never or very rarely conflict. But then a right to liberty must be either (a) a cluster of rights to particular liberties or (b) a right to liberty from unjustified interference (see below). The independents don't have a right to use unreliable or unfair enforcement procedures, because these would be unjustifiable interferences with the liberty of others. But notice that the DPPA prohibits not just the procedures of independents that are in fact unfair or unreliable, but also rather the procedures that it deems unreliable or unfair or which are simply unknown to it. Perhaps Nozick can claim that this is a justifiable form of interference with the liberty of independents and so violates no (moralized) right that independents possess. But this raises obvious questions. What is to count as a justifiable interference with someone's liberty? And if avoidance of risks that one considers undue is sufficient justification for infringing liberty, perhaps there will be many forms of interference with individual liberty that the state is justified in undertaking. In examining Nozick’s arguments against the extraminimal state later, we will see that libertarians seem required to forego public goods. For present purposes, we can note that because public goods are nonexcludable (the benefits of the good are available to both those who have contributed to the provision of the good and those who have not), their provision is subject to a free-rider problem. Individuals will see that they can enjoy the benefits of the good’s provision without the costs of contributing to its provision. But if enough people reason this way, the good won’t be provided. The immediate relevance of this problem is that creation of a state, even with only Nozick's nightwatchman functions of enforcing contracts and preventing fraud and harm, seems to be a public good itself. Most, if not all, of the benefits of the DPPA accrue to noncooperators (Nozick's independents) as well as to cooperators (clients). If so, creation of the minimal state or even a DPPA would seem to be threatened by free-riding. Nozick discusses this problem briefly (ASU: 113), but it’s not clear that his response is satisfactory.

AGAINST THE EXTRAMINIMAL STATE An entitlement theory of justice in holdings supports the minimal state; alternative theories that would require the state to redistribute holdings involve impermissible restrictions on individual liberty.

THE ENTITLEMENT THEORY The entitlement theory defines just holding in terms of three principles of justice: justice in acquisition, transfer, and rectification (ASU: 150-51).

1. A person who acquires a holding in accordance with the principle of justice in acquisition is entitled to that holding. 2. A person who acquires a holding in accordance with the principle of justice in transfer, from someone else entitled to the holding, is entitled to that holding. 3. A person who acquires a holding in accordance with the principle of justice in rectification is entitled to that holding. 4. No one is entitled to holdings except by (repeated) applications of (1)-(3).

The entitlement theory seems to state the form that a theory of justice should take. Its political implications seem to depend upon its content – specifically, the content of the principles of justice in acquisition, transfer, and rectification. Indeed, one wonders whether any theory of justice couldn’t be represented as a kind of entitlement theory. Given Nozick’s political conclusions, it is surprising how little he says about the content of his entitlement theory.

ACQUISITION Nozick’s theory of justice in acquisition develops Locke's labor theory of property (ASU: 174-82).

Though the earth, and all inferior creatures, be common to all men, yet every man has a property in his own person: this nobody has any right to but himself. The labour of his body, and the work of his hands, we may say, are properly his. Whatsoever then he removes out of the state that nature hath provided, and left it in, he hath mixed his labour with, and joined to it something that is his own, and thereby makes it his property. It being removed from the common state that nature hath placed it in, it hath by this labour something annexed to it, that excludes the common right of other men: for this labour being the unquestionable property of the labourer, no man but he can have a right to what that is once joined to, at least where there is enough, and as good, left in common for others [Second Treatise of Government §27; cf. §33].

Nozick raises some good general questions about the labor theory of property, without answering them adequately.

• Why does mixing one's labor with x generate property rights in x? • Why isn't it a way of dissipating one's labor? 5

• Why isn't it a way of acquiring rights to the value added to x?

We might also ask whether in cases of cooperative labor, are acquired rights proportional to percentage of contribution? Nozick does a little better in addressing more specific, internal questions for the labor theory.

• Is labor necessary? o Abundance only requires appropriation; scarcity requires labor. • How to understand the proviso that one's appropriation of x not make others worse-off? o Worse-off in what respect? . Worse-off with respect to x? . Worse-off with respect to x-type goods? . Worse-off in terms of overall utility (drug and beach examples)? o Worse-off compared to what? . No property rights. • We can characterize this interpretation of the proviso in terms of Pareto Superiority. Pareto Superiority is usually understood as a relation among preferences. o X is Pareto Superior to Y iff at least one person prefers X to Y and no one prefers Y to X. o X is Pareto Superior to Y iff at least one person is better off under X and no one is better off under Y. Nozick’s proposal is to understand the proviso so that an initial acquisition is just provided that it produces a state of affairs that is Pareto Superior to the state of nature. • Is the state of nature the appropriate baseline of comparison for the proviso? Why not consider alternative possible distributions of property rights?

To evaluate the appropriate baseline, consider this Dallas inspired shipwreck story. Suppose JR and Miss Ellie are the sole survivors of the shipwreck of a luxury liner and they wash up on an uninhabited island.

• If neither had appropriated the land, and each exercised his or her own powers separately on the common land, their respective utilities (in this state of nature = world 1) would be JR w and ME x, where w > x. • Suppose (world 2), however, that while ME is otherwise occupied (e.g. still unconscious, tending JR's wounds, or searching for food), JR appropriates the land and offers ME a job (e.g. raising vegetables, clearing rocks off the beach, catching fish for dinner). JR divides the proceeds of ME's work between them: JR receives goods and services worth y and ME receives goods (rations) worth z, where y is considerably greater than z, but z is equal to or marginally greater than x.

Is JR entitled to this greater share (y) of the y + z product just because he got there first and had entrepreneurial designs on the island while ME was either unconscious or worrying about JR? Is it just that "nice gals finish last"? 6

• Presumably, if ME had beaten JR to the punch, she could (at the very least) have adopted the same terms of cooperation (whatever they were) and kept y for herself and given JR z (provided z is equal to or greater than w) (world 3). Assume, as is likely, that the division of labor JR institutes in W2 is inefficient (he's not too good an organizer and, of course, he gave all the physically demanding jobs to the weaker ME). • Had ME appropriated first, she would have devised a more efficient division of labor (she might have devised new technology and she certainly would have made JR do a non-negligible share of the hard work) (world 4). Suppose, as seems likely, that the total output in W4 would have been considerably larger than y + z and that ME would have distributed it in a more egalitarian manner -- p to JR and q to herself, where q is equal to or greater than p and p is greater than y.

We can represent these possibilities in the following table.

J.R. Miss Ellie Constraints W1 (state of nature) w (e.g. 7) x (e.g. 4) w > x W2 (JR's appropriation; his division of labor) y (e.g. 15) z (e.g. 8) z ≥ x; y > z; z < w W3 (ME's appropriation; same division of labor as z (e.g. 8) y (e.g. 15) W2) W4 (ME's appropriation; her more efficient division p (e.g. 16) q (e.g. 16) q ≥ p; p > y of labor)

Why should we judge the justice of W2 solely by comparison with W1, ignoring possibilities W3 and W4?

• Is it enough for the proviso to be satisfied at the time of initial acquisition, or must it be satisfied continuously (including after bequest)?

TRANSFER Nozick appears to believe that voluntary exchange is both necessary and sufficient for justice in transfer. But externalities raise questions about such an account of justice in transfer. Externalities are effects (positive or negative) of market exchanges that are not taken into account in the exchange (e.g. not reflected in the price at which goods are bought and sold). Externalities are inefficiencies in market transactions; negative externalities are inefficiencies of overproduction, whereas positive externalities are inefficiencies of underproduction.

VOLUNTARY EXCHANGE AND NEGATIVE EXTERNALITIES Negative externalities raise the question whether voluntary exchange is sufficient for justice in transfer. Some negative externalities are direct (e.g. pollution). State intervention is the usual remedy.

1. The government may prohibit by law the behavior that produces the negative externality altogether; 2. It may tax the production or sale of the good so as to make production and consumption reflect the full costs of the good; 3. It may regulate production by adopting production standards designed to limit externalities (e.g. industry quality or pollution standards); or 4. The government may allow affected third parties to sue producers for compensation for the costs that the externalities impose on them.

Are such direct negative externalities a problem for the entitlement theory or the minimal state? Well, one might see the imposition of negative externalities as a rights violation, and then one might think that it is a job of the minimal state to protect rights and punish rights violations, perhaps employing any of these remedies, maybe especially (1) or (4). But it’s not clear that every negative externality qualifies as a rights violation. Moreover, if negative externalities are rampant, then there may need to be considerable state interference with otherwise voluntary exchanges. If so, the state that is minimal in the technical sense may not be that minimal in the everyday sense of keeping its nose out of market exchanges. In any case, these direct negative externalities suggest that voluntary exchange is not sufficient for justice in transfer, only voluntary exchanges that do not impose negative externalities. Other negative externalities are indirect. Market exchanges within a system of private property rights redistribute holdings and (can) produce inequalities. These inequalities have negative relational effects on third parties, which, in turn, produce negative intrinsic effects. Suppose A, B, and C initially have equal holdings. But A and B engage in an exchange that advantages A and disadvantages B. Not only is A now better-off than B, but he is also better-off than C. So far, C is only disadvantaged relationally. But if A can use his comparative advantage for political or economic influence, C may be made worse-off than he otherwise would have been. The intergenerational version of this worry in which C is a member of a later generation is a prime example of this sort of indirect negative externality. Here, the worry is that market exchanges can lead over time to a compromise of fair equality of opportunity. One sort of remedy would be limitation on inheritance and bequest and perhaps other forms of redistributive taxation (e.g. progressive income taxes).

VOLUNTARY EXCHANGE AND POSITIVE EXTERNALITIES Positive externalities -- in particular, public goods (PGs) -- raise the question whether voluntary exchange is necessary for justice in transfer. Roughly, a public good (PG) is any object or state of affairs such that if it is available to some member(s) of a group it is available to all others -- including those who have not shared in the costs of producing it (= the 8 good is nonexcludable).1 PGs include clean air, energy conservation, population control, political participation, education, and national defense. Of special interest to us are PGs that meet five conditions.

1. Widespread, but not universal, contribution is both necessary and sufficient for the good to be produced (in the case of goods whose provision is all or nothing) or for large quantities of the good to be produced (in the case of goods whose provision is continuous). 2. If the good is produced, it is available to all, even to non-contributors; it is impossible or impractical to prevent non-contributors from enjoying the good (= the good is nonexcludable). 3. Contribution is a non-negligible cost. 4. The effect of individual contributions on securing the good or on the amount of the good available to individual contributors is negligible; or, if some possible contributions have non-negligible effects (say, because of threshold phenomena), then it is the antecedent likelihood of an individual's contribution being one of these that is negligible. (This condition includes a non-interdependence of behavior assumption.) 5. The amount of the good that each individual would gain from widespread contribution outweighs the costs of his share of contribution.

(3) and (4) imply that the cost of individual contribution exceeds the (expected) marginal return on individual contribution. But then -- despite (5) -- the provision of the good is threatened by at least two problems.

THE FREE-RIDER PROBLEM The first barrier to successful collective action is the free-rider problem. It will be rational not to contribute.

1. Whatever I do -- whether I contribute or not -- either (a) enough others will contribute to produce the good or they will not (if the good's provision is all or nothing) or (b) my contribution will have a negligible effect on the amount of the good available to me (if the good's provision is continuous). 2. My contribution is a non-negligible cost to me. 3. If enough others will contribute, either (a) the good will be provided and my contribution would be a waste or (b) the good will be produced and my contribution will have a negligible impact on the amount of the good available to me.

1 Sometimes people add the further stipulation that the goods be non-rivalrous – such that one person’s consumption of the good does not reduce the amount of the good available for others. But few goods are genuinely non-rivalrous in this sense. Even if my consumption of clean air leaves you with plenty or all you desire, it does reduce the total amount of clean air available for consumption by you and others. In any case, I don’t see that the goods that concern us need be non-rivalrous if they satisfy the other five conditions. 9

4. If enough others won't contribute, either (a) the good will not be produced and my contribution would be a waste or (b) a very small amount of the good will be produced and my contribution will have a negligible impact on the amount of the good available to me. 5. Hence, either way, the cost of my contribution exceeds the (expected) marginal return on my contribution. 6. Hence, I should not contribute.

As many have noted, the free-rider problem is an n-person prisoner's dilemma.2

OTHERS Contribute Don't Contribute Benefits No Benefits Contribute 2 4 Costs Costs ME Don't Benefits No Benefits Contribute 1 3 No Costs No Costs

For every rational individual, non-cooperation strictly dominates cooperation; noncooperation is the rational strategy for me whatever others do. If so, rational individuals cannot, in principle, secure PGs.

THE ASSURANCE PROBLEM The second barrier to successful collective action is the assurance problem. An individual who does not intend to free ride may nonetheless be willing to contribute to the production of the good iff she is assured that others will contribute as well (NW box preferred to SW box). But as long as there is reason to believe that enough others will not contribute (e.g. because they will free ride), the individual will not contribute. In competitive markets, one will gain a competitive advantage on other firms by free riding; so free-rider problem will be the dominant barrier to PGs in the competitive markets. The usual remedy for these collective action problems is for the state to provide the good itself and fund the costs of production through compulsory taxation (e.g. as in the case of education, international and domestic defense, public parks).

PUBLIC GOODS AND THE MINIMAL STATE But the free-rider and assurance problems suggest that we may not produce public goods relying only on market exchanges. The libertarian might claim that in that case we must forego public goods. But that seems a very steep price to pay. In fact, a traditional

2 For the classical PD-situation, substitute "stay silent" for "contribute" and "rat" for "don't contribute" and put the following values in boxes (1)-(4): (1) I go free, you get 12 yrs; (2) we each get two yrs; (3) we each get 10 yrs; (4) you go free, I get 12 yrs. 10 liberal rationale for the state is precisely to solve various sorts of collective action problems. The state intervenes to supply the good that rational individuals cannot and funds the provision of these goods through mandatory taxation. Since the goods in question are genuine goods that outweigh the cost of production, most citizens should welcome this role for the state. But it is an extraminimal function for the state to perform. True, it is not one of extraminimal functions that most concern Nozick -- paternalism or redistributive taxation -- but it is an extraminimal function nonetheless.

RECTIFICATION Until we know "how far back we must go in wiping clean the historical slate of injustices," how do we know that the entitlement theory supports a minimal state? For instance, should we rectify for violations of property rights that occurred in American slavery and in the appropriation of Native American lands? Demands for rectification are often met with the reply that historical injustices were practiced by our ancestors against the ancestors of current African Americans or Native Americans and that our generation has not practiced these injustices against current African Americans or Native Americans. These facts are supposed to show that it would be unjust to enrich current African Americans or Native Americans at the expense of descendants of those who wronged their ancestors.

Assume that A1 perpetrated significant injustice against B1, that A2 and B2 are descendants of A1 and B1, and that A2 has committed no injustices against B2. Does justice require A2 to compensate B2 for A1’s injustices against B1 or would this itself be unjust?

Some would say that justice does not require, but rather forbids, reparations for these reasons. But it would only be unjust to ask the current generation to make reparations if they deserved their holdings. But if the holdings of descendants of those who committed injustices depend on those prior injustices, then those descendants are not entitled to their current holdings and it would not be unjust to them to require them to pay reparations. A2 does not have to commit injustices to owe reparations to B2 if A2’s advantages in relation to B2 depend on A1’s injustices against B1. Likewise, if current African Americans or Native Americans are worse off as the result of injustices committed against their ancestors than they would other wise be, especially in ways that put them at a competitive disadvantage against the descendants of those who committed injustice against their ancestors, then they can coherently demand reparations from those who have unjustly benefited from past injustices even if they have not committed injustices themselves. This is not an answer to Nozick’s question about how far back we must go in correcting past injustices. What should we say about more distant descendants, such as A96? One might claim that the effect of earlier injustices peters out over time inasmuch as later generations depend on more intervening variables. But why shouldn’t our metaphor be that inequalities often snowball over generations, increasing, rather than diminishing, their influence? Perhaps larger expanses of time just make for epistemological difficulties about calculating the economic effects of injustices. Even if that were true, it wouldn’t show that we shouldn’t concern ourselves with rectifying past injustices. Moreover, how 11 big an epistemological problem the passage of time raises is surely a partly empirical question about the resources of economic modeling.

JUSTICE IN HOLDINGS Nozick's official taxonomy of theories of justice in holdings is this:

Theories of Justice in Holdings

Historical End-State

Entitlement Patterned

End-state theories are current time-slice principles that make the justice of a set of holdings depend only on the sum and/or distribution of value. Historical theories make the justice of holdings depend upon past circumstances of the distribution in some way. A principle is patterned if it makes the justice of a distribution depend upon some historical property of distributees, rather than their past actions (e.g. need, effort, merit, IQ). Entitlement theories say that the justice of holdings depends upon the past actions of the holders and how they came by their holdings. But this taxonomy is puzzling. Nozick seems to think that patterned theories have more in common with end-state theories than with entitlement theories. His criticism of patterned theories is supposed to apply to Rawls’ Difference Principle, an end-state view. Moreover, we might note that some patterns are historical and others are not. Distributing according to effort would be historical. But it’s not clear if this would be true if we distribute according to moral merit. It would certainly not be true if we distributed according to need or IQ or attractiveness. So perhaps we need to distinguish between historical and non-historical patterned principles, which should lead us to revise Nozick’s taxonomy.

Theories of Justice in Holdings

Historical Ahistorical

Historically Non-historically Entitlement End-state Patterned Patterned

WILT CHAMBERLAIN Nozick believes that voluntary exchanges are justice preserving: if D2 is unjust, then D1 must have been unjust. Because D1 satisfies the pattern, and liberty upsets any pattern (ahistorical principle), all patterned (ahistorical) principles must be mistaken.

1. Voluntary exchanges are justice preserving. 2. Hence, if D2 is unjust, then D1 must have been unjust. 3. D1 can satisfy any pattern (non-historical principle). 4. Voluntary exchange can upset any pattern (non-historical principle). 5. Hence, all patterned (non-historical) principles must be mistaken.

Nozick likens principles of justice in transfer, which are supposed to be justice preserving, to rules of valid inference, which are supposed to be truth preserving. In effect, his argument about liberty upsetting patterns is really a modus tollens argument. This is what is supposed to be demonstrated in the Wilt Chamberlain example. The argument raises two issues. First, is (4) true? D2 upsets the pattern of income distribution; so, for example, it upsets the pattern requiring strict equality of income. But does it upset all patterns? What about the pattern of resource or “starting gate” equality? Second, if a pattern determines the justice of particular transactions directly, does anyone hold a patterned theory? Rawls's Difference Principle applies to the basic structure of society, not to individual market exchanges (e.g. TJ 87-88). It will favor principles that permit market exchanges but require certain kinds of redistributive taxation. Presumably a utilitarian would favor economic and legal regimes that allow market exchanges. Equality of resources requires starting gate equality but does not require that subsequent transactions preserve this equality.

ARE ALL OPERATIONS OF THE MARKET MORALLY INNOCENT? Market exchanges within a system of private property rights redistribute holdings and (can) produce inequalities. Such inequalities can have negative relational effects on those who become comparatively worse-off. They affect people's relative levels of well- being, and this affects their bargaining position insofar as it allows some to hold out for better terms of cooperation. Hence, these inequalities produce further inequalities and make the worse-off worse-off than they would otherwise have been. If so, negative relational effects of market exchanges produce negative intrinsic effects. Should parties to these exchanges (e.g. basketball fans) participate on these terms if they calculate indirect, as well as direct, costs? These indirect costs include negative externalities, affecting people not party to the exchange, whether intragenerational or intergenerational (161). These negative externalities appear to run afoul of Nozick's own prohibition on uncompensated sacrifices (32-3). Moreover, the negative externalities seem unfair, especially in the intergenerational case. It may seem that one should not be disadvantaged as the result of factors outside of one's control. The worry here is that market exchanges in a laissez-faire system lead to concentrations of wealth that can compromise fair equality of opportunity. Think about Potterville in Frank Capra’s Christmas classic It’s a Wonderful Life. 13

Concerns about fair equality of opportunity might lead us to try to control the effects of the social lottery, perhaps through inheritance taxes and perhaps progressive income taxes, if we can’t otherwise protect against the effects of inequality on education, social networking, and economic clout. Must we also try to control the effects of the natural lottery? Nozick thinks not. He denies that entitlements must be "earned all the way down" (225). He thinks that correcting for the effects of the natural lottery is a reductio ad absurdum of correcting for the effects of the social lottery. But is it? We can maintain that both the social and natural lotteries are morally arbitrary. We need to distinguish four claims.

1. I am entitled to possess my natural endowments, though they are unearned. 2. I am entitled to exercise my natural endowments (in acceptable ways). 3. I am entitled to benefit from the productive exercise of my natural endowments. 4. I am entitled to all the benefits that I can get others to concede to me from the productive employment of my natural endowments.

(1)-(3) do not imply (4). In effect, Nozick assumes that trying to control for inequalities in natural and social endowments that are outside one’s control would lead to interfering with rights of self-ownership, perhaps by leveling natural endowments (by weakening the strong, lobotomizing the intelligent, or disfiguring the attractive or at least by appropriating their talents). But the person concerned with fair equality of opportunity need not meddle with self-ownership. She wants only to enforce a proviso about the terms on which one can benefit and generate property rights in things and worldly resources from the productive employment of one’s natural talents.

NOZICK'S MORAL THEORY Nozick seems to have a direct argument against the extraminimal state from his assumptions that rights are side-constraints and that our most basic right is a right to liberty. For any state that deprives people of liberty for paternalistic reasons, to redistribute wealth for purposes of fair equality of opportunity or for other distributive justice reasons, or to provide public goods will violate side-constraints. But that moral argument is only as strong as the foundations on which it is built. Nozick assumes that rights are side-constraints and that the only rights we have are to any and all liberties. Each of these assumptions deserves scrutiny.

WHY SIDE-CONSTRAINTS? Nozick, we saw, contrasts constraints and goals and insists on understanding rights as constraints, rather than important goals. This, we said, was like Pettit’s contrast between promoting a value and honoring it. If I’m promoting x, I may sometimes need to forego or deny x in one place in order to better promote it overall. But if I honor x, I must observe it or respond to it on every occasion I can. For instance, we said Mill promotes autonomy by denying agents the right to sell themselves into slavery. If we were to honor personal autonomy, instead, we would let them sell themselves into slavery (while advising them not to). Which approach better fits our intuitions? I think that the cases are mixed. Sometimes we think of rights as side-constraints, but in other cases we do not. 14

1. Organ Harvesting. B is otherwise healthy and in the hospital for hangnail removal. C-G are in the hospital in desperate need of organ transplants, but there are no available donors. Assume that the surgeon could save C-G by harvesting B’s organs, against his will, and transplanting them. Most of us think that non-consensual organ harvesting would impermissibly violate B’s rights, even though it would save five innocent lives. 2. Hunting Trespass. You have been seriously injured in a hunting accident and need immediate medical attention. The only telephone for miles is in an unoccupied cabin. Most of us think that it is permissible for me to trespass and violate the cabin owner’s property rights breaking down the door of her cabin to contact EMT and get medical aid to my friend. Here we violate a property right to save another’s life. 3. Torturing the Terrorist’s Child. Suppose that a terrorist has planted a bomb in NYC that will kill and injure thousands, perhaps millions. We have her in custody. She is a determined and accomplished terrorist who has fulfilled all her terrorist threats in the past. Her only weakness is her daughter, whom we also have in custody. The only way we can prevent a tremendous loss of life and widespread violation of rights is to torture her innocent daughter to induce her mother to reveal the location of the bomb. On these facts, it seems permissible to torture the daughter just enough to get her mother to reveal the location of the bomb and the means of defusing it. Here we violate the right not be tortured of one innocent person to prevent even more serious rights violations of thousands/millions. 4. Killing the Terrorist’s Shield. Same terrorist. She’s advancing to detonate the bomb but is using her daughter as a shield. The only way we can stop her is for our sniper to shoot the daughter in the head, piercing her mother’s heart. On these facts, it seems permissible to kill the innocent shield to prevent the terrorist from killing thousands/millions. Here we violate one innocent person’s right to life to prevent comparably serious rights violations of thousands/millions.

Cases like Organ Harvesting lend some initial support to side-constraints. But there are many other cases in which it seems permissible and perhaps obligatory to promote, rather than honor, value and to minimize rights violations. Hunting Trespass and Torturing the Terrorist’s Child involve cases where it is permissible to sacrifice one right for a more important right or a larger number of more important rights. Nozick could accommodate such cases if he allowed aggregation across kinds of rights but not within a category of rights. Property rights can be violated to save lives, and a person can be tortured to prevent many killings. But Killing the Terrorist’s Shield introduces minimization within a category of rights -- the right to life. These cases should make us pause before accepting side-constraints. As Nozick himself notes, side-constraints are paradoxical. If x is so important, why not maximize x or minimize violations of x (28-33)? Nozick's own answer is to appeal to the separateness of persons and the Kantian demand that we treat all agents as ends and never merely as means. Notice that Kant does not forbid treating others as means, only treating them as mere means. Does the person who tortures the terrorist’s child to prevent the deaths of thousands/millions treat the child as a mere means? That would be to treat the child as a tool that deserves no intrinsic concern. But then we would see no reason not 15 to torture, no reason to stop and switch tactics if another safer one (e.g. truth serum) becomes available. But that is not how we view her. We recognize that she has a right not to be tortured, which would normally be decisive. We regret torturing her and proceed reluctantly because this is the only available means to prevent widespread loss of life. If any less intrusive means becomes available to accomplish the same objective, we will gladly abandon the girl’s torture. In such circumstances, we treat the girl as a means, but not, I think, as a mere means. If it is impermissible to torture her (which I doubt), it is not because, as Nozick claims, we would be treating her as a mere means. In their discussions of the separateness of persons Rawls and Nozick both appeal to inviolability. Nozick claims that we can only recognize the inviolability of persons by recognizing individual rights and embracing a side-constraint conception of rights. We can grant that inviolability (in this sense) requires side-constraints, but then we should note that inviolability has a price. It denies moral considerability to those who could be saved or have their rights protected by violating the rights of another. We want to take seriously the fundamental interests of each. But do we want to endorse inviolability at the price of considerability?

WHICH SIDE-CONSTRAINTS? Nozick’s argument doesn’t just assume that rights are side-constraints. It also assumes that the principal right that people have is a right to liberty. This seems to be a right to any and all liberties. Nozick does not distinguish between important and unimportant liberties. Is this assumption about the content of individual rights plausible? Is there a right to liberty per se? Even the defender of the minimal state recognizes that there is no right to do harm. But he must recognize other limitations on the right to liberty as well. Private property rights restrict the liberty of non-owners. Moreover, all liberal states recognize nuisance laws that restrict the liberty of citizens to be disruptive and that place time, manner, and place restrictions on conduct. These examples show that there is no right to liberty per se. Indeed, if rights are side-constraints, they must be near absolute guarantees against interference. But then there could be no simple right to liberty.

1. Rights are side-constraints. 2. Side-constraints on conduct are near absolute prohibitions of that conduct. 3. Hence, rights can rarely if ever conflict. 4. Rights to liberty per se (i.e. to be free to do as one wishes) conflict with other rights we have (e.g. against harm by others, to own property). 5. Hence, we don't have rights to liberty per se. 6. If rights are side-constraints, we must either (a) recognize only rights to basic liberties or (b) moralize liberty, so that we have a right to be free from unjustified interferences with our liberty.

Either way, we lose the direct argument from a rights-based conception to libertarianism and the minimal state. The minimal state may not be necessary to secure basic liberties or to prevent unjustified interference with liberties. Is there only a right to liberties? Nozick recognizes only negative rights to be free from interference and claims that any extraminimal state would violate these negative 16 rights. That depends on which negative rights we have. But what if we also had certain positive rights to fair equality of opportunity, to public goods, to basic medical care, to a good and affordable education, to a decent standard of living, or work for those who want it? These would be positive rights. By the previous argument, we know that recognition of positive rights would have to curtail our negative rights. But if we have positive rights and the state must honor these, then this is likely to require an extraminimal state. How doe we decide which rights we have? Nozick attempts to justify libertarian side-constraints by appeal to their role in a meaningful life in which rational agency is exercised (49-50). As Mill argues, certain basic liberties are essential to a meaningful life for progressive beings with higher, rational capacities. But it’s not clear that all liberties are equally essential to a meaningful life. Freedom of speech, conscience, association, and occupational choice will be essential liberties. But the liberty to drive both ways down a street, to choose not to wear a seatbelt, or to dispose of one’s gross income as one sees fit may be less essential. Moreover, this rationale may support recognizing positive rights, to education, healthcare, public goods, or fair equality of opportunity.