Is Coercion a Ground of Distributive Justice?

In his rich and stimulating book, Blake argues (among other things) that comprehensive coercion triggers egalitarian obligations of distributive justice. I argue that (1) coercion is not a necessary condition for egalitarian justice to apply; (2) Blake’s use of a moralised conception of coercion is a mistake; (3) coercion is a redundant member of any set of sufficient conditions that might explain why distributive justice applies; (4) Blake’s emphasis on providing conditions for the exercise of autonomy might support a much more cosmopolitan theory of distributive justice.

A basic commitment to the autonomy of all human beings is at the heart of Justice and Foreign Policy. Blake here draws on Raz rather than Kant. In this picture, autonomy is a value realized when three conditions have been satisfied. First, the autonomous have a range of mental capacities sufficient to exercise judgment and choice in the light of reasons. Second, they must have an adequate range of options from which to choose. And third, the autonomous must live independently, free from the domination of others; they must live in accordance with their own view about how that life should be lived. When these three conditions are satisfied, a person can rightly consider herself to be a (partial) author of her life. 1 But Blake's book is not about autonomy. He wants to take the value of autonomy as a given, as something that any liberal has reason to endorse without taking a definitive view about its deeper philosophical grounds. Instead, Blake's attention is on the implications of our commitment to autonomy for our understanding of the content, grounds, and scope of both international toleration and distributive justice. In this response, I will focus on the latter. In short, Blake argues that obligations of distributive justice towards others can emerge in one of two ways. First, they can emerge from a general duty to secure a minimal level of autonomous functioning for all human beings. This, Blake argues, gives us strong reasons to pursue foreign policies that leave no human being with less than is required for minimal autonomous functioning. Given the circumstances of our world, including its division into a world of states, this requires us to support the emergence of democracy in non-democratic states (while being wary of how much we can actually change, especially via misguided military action), to support states that do not have the capacity to sustain minimal autonomous functioning for their citizens, and not to undermine democracy abroad where it has been already established. Second, more demanding obligations of distributive justice -and here Blake means egalitarian justice along the lines of Rawls's second principle -emerge, Blake claims, only in the context of the comprehensive coercion typical of the modern state. The idea, roughly, is that such comprehensive coercion threatens to undermine our autonomy in a way that coercion at the international level does not. International coercion is aimed, for example, primarily at states, and there is no authority internationally that can actively coerce individuals without the mediation of states. Because of this threat to our autonomy, the state therefore owes us a special justification for the way it organizes public and private law that the international order does not. This special justification, in turn, requires the state to secure a broadly egalitarian distribution of prospects, since it must show us that those worse off in the scheme could be no better off in any alternative scheme. This is a rich work that bears much more reflection than I can give it here. There is much with which I agree. For example, I think the arguments for why liberal democratic states have no good reason not to encourage liberal democracy abroad are important, and bear much closer scrutiny. It is a mistake to think that a theory that does not shy away from this conclusion must be committed to crusading policies to intervene militarily in other countries, or to override or disrespect the voices of those who disagree. Quite on the contrary, a liberalism that takes seriously its own commitment to democratic freedom and equality must also take seriously the oppression and resistance of all those around the world whose freedom and equality is not recognized by their own governments. This has important implications, which I cannot explore here, particularly for our attitudes towards the treatment of women in many non-liberal states (as well as in liberal ones!). Blake's critique of Rawls's Law of Peoples is also particularly rewarding, as is the idea that an adequate conception of justice must be sensitive to the contexts in which justice becomes, as it were, a problem. A conception of justice should be, Blake convincingly argues, a response to the injustice we find ourselves faced with us here and now. While more idealizing conceptions of justice -which, for example, try to imagine away our current political world, and try to imagine a world made, as it were, from scratch -are certainly legitimate, this is not the project that Blake wants or claims to engage in. This is not because his conception of justice is a form of non-ideal theorizing, but because this is what ideal theory, correctly understood, itself requires. Indeed, along with Laura Valentini's Justice in a Globalized World, I believe that Justice and Foreign Policy is the most powerful recent statement of a coercion-based view. 2 In this paper, I will focus on those aspects of Blake's argument with which I find myself less in agreement. In particular, I want to focus on Blake's emphasis on coercion, and query whether his commitment to autonomy really can justify such a narrow focus. My response is divided in four sections, covering four independent, but related, objections. My main conclusion is that coercion cannot be a ground for distributive obligations. If we have distributive obligations towards others, they must be grounded independently. I For Blake, comprehensive coercion is a necessary condition for the application of egalitarian norms. Comprehensive coercion involves threats, directed to individuals, issued by public authorities, and 2 See also M. Risse, On Global Justice (Princeton: Princeton University Press, 2012), though it is not clear what role coercion actually still plays in the view, given the emphasis on reciprocity as also a ground of egalitarian justice. Cf. M. Risse, 'What to Say About the State', Social Theory and Practice 32 (2006): 671-698, which holds that comprehensive, direct, and immediate coercion is the sole ground of egalitarian justice. aimed at ensuring compliance with those legal norms covering the domains typically associated with a fully fledged legal system (most significantly, criminal law and all laws touching on the distribution of individual property holdings). 3 In a previous article, I argued that such coercion could not be a necessary condition. I gave an example to which Blake responds in the book. Here is the example: Imagine an internally just state. Let us now suppose that all local means of law enforcement -police, army, and any potential replacements -are temporarily disarmed and disabled by a terrorist attack. Suppose further that this condition continues for several years. Crime rates increase, compliance with the laws decreases, but society does not dissolve at a stroke into a war of all against all. Citizens generally feel a sense of solidarity in the wake of the attack, and a desire to maintain public order and decency despite the private advantages they could gain through disobedience and noncompliance; this sense of solidarity is common knowledge and sufficient to provide assurance that people will (generally) continue to comply with the law. The laws still earn most people's respect: the state continues to provide the services it always has; the legislature meets regularly; laws are debated and passed; contracts and wills drawn up; property transferred in accordance with law; disputes settled through legal arbitration, and so on. 4 The example is meant to trigger our judgment that norms of distributive justice would still apply, even though people are not coerced into compliance. 5 Blake responds in two ways. First, he claims that the thought experiment can't work because the situation described is unlikely ever to be realized in practice. This response seems weak. The point of the example isn't to describe an empirical possibility, but to describe a case in which we hold constant everything but the presence of coercion. All of Blake's own cases involving Borduria and Sylvania (of which there are 16) involve hypothetical cases in which we are asked to imagine mere possibilities. I don't see how the empirical likelihood that any of these cases might be actualized in the world as we know it is relevant. And, in any case, I don't find the case, as described, particularly hard to imagine. After all, most compliance with current legal systems -say in the U. 5 I go on, in the original paper, to describe how implausible it would be to bite the bullet and claim that in this case norms of distributive justice would no longer apply. Could the rich in the society stop paying their taxes simply because police enforcement has stopped, especially once we assume that everyone else in the society continues to pay? responding to the threat of being punished, but because, for a variety of reasons, they find the norms either sufficiently authoritative, or find that non-compliance would be wrong for other reasons (e.g., because, say, murder or stealing is wrong, or out of a sense of fair play -when contemplating, say, cheating on their taxes -or simply out of habit). 6 Indeed, no legal system could survive if individuals sought to disobey whenever they could get away with it for some personal gain; no amount of police or surveillance would be sufficient to prevent breakdown. The case of the post-attack society just asks us to generalize such patterns of compliance, and to assume that sufficient assurance could be achieved through some other means (such as common knowledge regarding the prevalence of such law-abiding dispositions).
The second strategy of response is more promising. Blake grants that egalitarian norms would still apply in the post-attack society, but argues that the example only works to trigger the judgment because there is a more subtle form of coercion at work -a form of coercion which, in turn, is sufficient to generate the very same demands of justice as more typical forms of state coercion (viz. those involving the threat of punishment and police enforcement). The form of coercion Blake has in mind is the threat of exclusion from the protection of the legal system, and hence the threat of being returned to the state of nature. 7 This (implicit) threat, Blake argues, still infringes the autonomy of those in the society, and so still must be justified in the same way.
But the example can be clarified in a way that makes it evident that no such threat is required to maintain the force of the example. In the original example, I asked the reader to imagine that a small group of rich people (impressed by Blake's argument) had initiated a movement to eliminate steeply progressive taxes. They claimed that there was no longer any coercive enforcement of the law, so they were no longer under an obligation to pay the high taxes required to maintain an egalitarian distribution of shares. It was important here that they were not arguing that they were no longer obligated to pay because, without the coercion, others could not be trusted to pay their fair share. Ex hypothesi, they agreed and accepted that others would continue to pay their taxes (foolishly in their estimation). Rather, their argument was Blake's: comprehensive coercion triggers a special demand for distributive egalitarianism because of the way it infringes autonomy, independently of its further downstream effects. The rich group's argument, I argued, was wrongheaded: why should the presence or absence of legally authorized coercion in a just society determine what we owe to others at the bar of justice? Isn't it sufficient that everyone complies with the law (and that such compliance is common knowledge)? Recall, nothing in this society changes except the state's ability to coerce those on its territory.
Blake writes that this argument only works because we assume (implicitly) that if citizens do not comply with the laws, they will be 'returned to the state of nature'. I interpret Blake here to mean that were citizens to violate the laws, they would be ostracized or face being otherwise stripped of the rights and privileges of citizens. But I don't see why this further assumption is necessary. Suppose, for example, that it is well known that should anyone stop paying the taxes mandated by extant legislation, nothing would happen to them. 8 They would still have all the rights and privileges of citizens; the society has given up on the attempt to punish or penalize noncompliers (though it still expects and calls for their compliance as a matter of law). And imagine that the rich, along with everyone else, still comply with the law out of a sense of solidarity and fair play. I don't see why the mere possibility of violating the law without suffering any sanction ought to change our judgment about the justice of the egalitarian scheme. Again, everything is as before: laws passed, taxes collected and resources redistributed, jobs completed, and so on. Why should the mere fact that people are no longer coerced make any difference to our claims on a fair share of the social product?
The objector might make one last try: 'Yes, but there is still a form of autonomy-infringing coercion in this case, namely 'social pressure': should anyone fail to obey the law, others would complain; such complaints are unpleasant, and so there is still a 'penalty' attached to non-compliance'. I hope it is clear that this response would give away the game. Social pressure is only plausibly seen to constitute a form of autonomy-infringing coercion when it either serves to encourage actions that are morally wrong or when the amount of social pressure applied is disproportionate. Social pressure on wives to be subservient to their husbands, for example, is plausibly seen as autonomy-infringing coercion because it encourages and reinforces the subordination of women. But when social pressure is proportionate and serves to reinforce genuine moral obligations, it is implausible to say that such pressure constitutes autonomy-infringing coercion. Social pressure that functions to support norms against murder, lying, or cheating, for example, is not plausibly seen as autonomy-infringing coercion (or even pro tanto morally wrong), precisely because it (proportionately) discourages morally wrongful conduct. 9 This puts our objector on the horns of a dilemma: if he accepts that social pressure in favor of laws supporting egalitarian norms in our post-attack society are proportionate and do not encourage wrongful conduct, then he must abandon the claim that such social pressure would constitute a form of autonomy-infringing coercion. If, on the other horn, he argues that social pressure in favor of laws supporting egalitarian norms in our post-attack society is either disproportionate or does encourage morally wrongful conduct (because egalitarian norms no longer apply), then he has squarely begged the question. For the argument to go through, we must be able to identify instances of autonomy-infringing coercion first, and then, in virtue of that fact, conclude that egalitarianism therefore applies. But if identifying the presence of autonomy-infringing coercion itself depends on whether egalitarianism applies or not, then we are nowhere. I conclude that Blake's (and the further objector's) response to the counter-example fails. II In this section, I want to raise a related, but more general, problem regarding Blake's conception of coercion. At a crucial point in the book, Blake discusses the possibility of coercion among states. He wonders under what conditions hard bargaining between a richer and a poorer state (that is not, in Rawls's terms, 'burdened') constitutes coercion. 10 This is an important point to clarify for Blake because the extension of his coercion-based account to the international realm depends on it. If there is objectionable coercion in cases of this kind, then states would have a special obligation to compensate or otherwise rectify the consequent infringement of the subjugated state's citizens. While this may not be as thoroughgoing a set of obligations as those incurred at the domestic level (depending on how we ultimately characterize the coercion), they would still be significant. So, given the absence of systemic police enforcement or individual punishment at the international level, 11 we need to know when pressure or influence counts as coercive (and hence as objectionably autonomy-infringing). Blake notes that the solution turns on whether a particular proposal 12 -such as, for example, 'we, A, will continue to trade with you, B, only if you allow our military forces access to your bases when operating in foreign theaters' -is an offer or a threat. Answering this further question, in turn, requires determining what baseline should be used in judging whether B has been relevantly harmed or benefitted by the proposal. Should we, for example, use how well off B was before the proposal as a baseline? How well off B would be in the normal, expected course of events without the proposal (given what we know about situations of this kind)? 13 Or how well off B morally ought to be? For reasons first 10 p. 119. 11 Blake notes the exception, namely international criminal law. 12 Or, we might add, whether the influence or pressure exerted by a particular set of proposals to which the existence of an institution or regime gives rise counts as coercive. 13 For example, does it matter whether A was planning to stop trading with B for independent reasons before they made the proposition? adduced by Nozick, Blake sides with the last option. 14 Blake's account of coercion is therefore moralized: one counts as having been coerced only when the proposal would leave one worse off than one morally ought to be. In the example just given, A would thus count as coercing B only if A had a prior moral obligation to trade with B. This serves Blake well because it allows him to deploy his account of a state's foreign obligations -which are, as I mentioned above, based on a general duty to maintain basic conditions for autonomous flourishing -to determine whether particular international transactions count as coercive or not. On Blake's moralized view, a particular international transaction or institution counts as coercive only when it can be shown to undermine the basic capacity for autonomous flourishing of those affected by transaction or institution. If Blake is right, this not only simplifies the task of assessing the justice of foreign policy but also blocks off an important and common line of objection, namely that international coercion is so profound and pervasive that it fundamentally infringes the autonomy of citizens the world round (which would entail that egalitarian obligations would be global in reach).
Adopting a moralized view of this kind comes, however, at a cost. While it may help to clarify when international transactions and institutions are plausibly seen as coercive, it undermines his domestic account. Recall that Blake's paradigmatic example of comprehensive state coercion is criminal law. 15 But suppose a state's criminal law penalizes only morally wrong acts, and suppose further that sentencing judgments are fully proportionate, and punishment is administered fairly. If that is the case, then, on a moralized view, the state does not coerce you when it threatens to throw you in jail if you commit murder. 16 Because you lack a right not to be imprisoned in such circumstances, you are made no worse off than you morally ought to be. And if that is true of the criminal law, then it is also true of any just law, including just tax and property law. Take laws proscribing tax evasion, for example. Let us imagine that the tax system is just overall, and that the laws declare that anyone caught in violation of such laws will either have to pay a (proportionate) fine or be subject to (proportionate) criminal penalties. Because (ex hypothesi) the tax system is just, and the fines and penalties proportionate, no one who either complies with the law or who is caught and punished or fined for evading taxes is morally worse off than they ought to be. And if no one is morally worse off than they ought to be, then, according to a moralized account, they have not been coerced. This is a serious problem for Blake's view for two reasons. First, as with the social pressure objection, it looks as if, on this view, we need to know whether we have egalitarian obligations before we can identify whether a particular tax system is coercive in the first place, which would straightforwardly undermine Blake's claim that coercion is an independent ground of egalitarian justice. Second, it looks like the view would have the absurd consequence that egalitarianism was required in all but the most just states (whose laws could not, on a moralized view, count as coercive).
Blake is alive to this worry, and, in a footnote, he writes: My response, I think, is to point out that much of the law involves restrictions on actions that are not malum in se -much of what we do through law is to define who shall hold what entitlements, rather than simply to prevent force or fraud.
Blake is here referring to the distinction between moral wrongs that are malum in se (morally wrong in themselves, i.e., independently of any legal regulation) and those that are malum prohibitum (wrong because they are prohibited, i.e., morally wrong only because a law has proscribed them). Blake's thought appears to be that adopting a moralized view of coercion is not a problem because only laws that legitimately regulate mala in se come out as not coercive on such a view (but then what of most criminal law?). On the other hand, laws that legitimately create mala prohibita -such as, for example, tax, traffic, and most property law -can still be viewed as coercive because they regulate areas in which there are no (pre-existing) moral wrongs. If this is right, then the state still has a job to do in providing a special, more demanding justification to those subject to laws governing areas that are mala prohibita -and therefore, most importantly, a special justification for the distribution of property entitlements. The problem is that Blake never explains why the distinction between mala prohibita and mala in se should matter in determining the moral baseline for identifying instances of coercion. In both cases, there are moral wrongs; in both cases, threats are issued which come with penalties. Why the difference? Perhaps the idea is that, in the case of actions that are mala prohibita, citizens have, by definition, 'pre-existing' moral permissions that the law then proscribes. So when the state changes our normative situation by prohibiting such previously permissible actions, it owes us a special justification for doing so that it wouldn't owe us had we never had the permissions in the first place. The intuition here seems to trade on the fact that the law's prohibition harms us by setting back our interests in doing whatever it is that was morally permitted before the law settled the case; the harm is key in explaining both why the prohibition is coercive, and hence why it is pro tanto wrong.
But what are the morally permissible actions that the state now makes it impossible for us to act on? In what sense, exactly, does the state harm us when it passes (just) laws? The trouble is that, especially with regard to property and tax law, it is not clear that we have morally relevant 'pre-existing' moral entitlements with respect to any particular bit of property or income that are thwarted by the (just) state's laws in that area. The laws create property entitlementsincluding entitlements to income -that would not have existed in the absence of the law. 17 If this is correct, then the only moral permissions that we must give up in a fully fledged legal system are our general moral permission to acquire property in a state of nature (by, say, mixing our labor with it) and our general permission to roam the territory unhindered by property restrictions. The key question is: In what sense does the state harm us when it legislates in such a way as to make what was previously permissible impermissible? The moralized view of coercion says that we need to take moralized view of the baseline in determining whether we have been harmed. 18 So we should ask: Does the state leave us worse off 17 Blake appears to agree at p. 94; but cf. p. 92. 18 Notice that if we take a non-moralized view, then only a radical anarchist would be able to argue that we were better off before the just state started making laws. But Blake argues that we have a moral obligation (rather than merely a permission) to create coercive institutions, precisely because of their good effects on autonomous flourishing, so the anarchist position is unavailable to him anyway. than we morally ought to be when it begins passing (just) laws? The answer must be no. For recall that Blake argues (among his four basic principles), that we have a moral obligation to set up coercive institutions precisely because such institutions are required to make autonomous lives possible. This entails that while we have moral permissions in the state of nature, we don't have any moral immunities to being coerced by just states. So by passing laws, and changing our moral permissions, the just state does not harm us in any relevant moralized sense. If that is right, then, once again, the just state cannot be understood to coerce us when legislating the mala prohibita. The distinction between mala prohibita and mala in se cannot be used, I conclude, to answer the objection.
One might wonder whether Blake has in mind a more libertarian view, according to which, for example, our pre-tax income or property is considered to be ours, and taxation is considered as a taking, in which what is ours is given to someone else. On this view, taxation would be coercive, even on a moralized view. This is because it would constitute a pro tanto violation of our pre-tax entitlements. At one point, Blake writes: The law of taxation is clearly coercive. Federal income taxation plainly involves the taking away of previously earned resources from individuals. As above, this form of law seems properly regarded as a putative violation of the liberal principle of autonomy -it gives us, in essence, a choice between surrendering our good or our lives…. [S]uch taxation is presumptively wrong until justified through the giving of reasons which could not be reasonably rejected by those who face the taxation. (92) But in what sense do we have any entitlement to what we earn before taxes, such that the tax counts as taking something that is already rightfully ours rather than simply defining what is to be rightfully ours in the first place? For the argument to go through, some form of libertarianism is required to get the pro tanto wrongness of any tax law going. 19 But if such libertarianism is required, then it looks unlikely that any kind of egalitarianism could be justified, except as a way of rectifying past wrongs in transfer or acquisition. Given the existence of such strong rights to pre-tax income, why wouldn't the conclusion simply be that redistributive taxation is a form of slavery?
The general view of property on which Blake seems to rely strikes me as implausible for similar reasons as those adduced by Murphy and Nagel: We are all born into an elaborately structured legal system governing the acquisition, exchange, and transmission of property rights, and ownership comes to seem the most natural thing in the world. But the modern economy in which we earn our salaries, own our homes, bank accounts, retirement savings, and personal possessions, and in which we can use our resources to consume or invest, would be impossible without the framework provided by government supported by taxes. This doesn't mean that taxes are beyond evaluation -only that the target of evaluation must be the system of property rights that they make possible. We cannot start by taking as given…. some initial allocation of possessions -what people originally own, what is theirs, prior to government interference. 20 But the 'everyday libertarianism' rejected by Murphy and Nagel cannot be Blake's view. He does not hold the implausible view that somehow redistribution is justified as a kind of compensation for the violation of pre-existing property rights. 21 After all, if we did have prior rights to pre-tax income and property, then the state would owe the rich more than the poor, since it takes more from the rich than from the poor. I conclude that Blake's adoption of a moralized view is a mistake.

III
In the first section, I argued that comprehensive coercion is not necessary. In the second section, I argued that it was a mistake for Blake to adopt a moralized view of coercion. At this point, we might wonder two things. First, we might wonder whether there is any deeper reason to endorse the intuitions governing our response to the post-attack counterexample. Why can't Blake just bite the bullet in that case, and say that the group of rich people envisaged in the counterexample has a point? Second, we might wonder whether Blake could abandon both the claim that coercion is moralized and the claim that coercion is necessary. That is, can Blake argue that non-moralized coercion is only a sufficient condition for distributive justice? 22 In this section, drawing on previous work, I will argue that coercion (whether or not we take a moralized view) cannot be a sufficient condition for egalitarian obligations to apply. In brief, I will contend that we have no reason to believe that coercion can explain either the content or the scope of our distributive obligations. Such obligations must be grounded independently. If I am right, then this helps to explain our intuitions in the post-attack counter-example: we react the way we do to the counter-example because we rightly assume that egalitarianism, if it applies, must apply for reasons that have nothing to do with the presence (or absence) of coercion.
Here is a schematic summary of Blake's argument (here taken as a sufficient condition which is neutral between moralized and nonmoralized views): (1) Coercion is pro tanto wrong (because it violates autonomy).
(2) Those whose will have been coerced are therefore owed a special, more stringent justification for the coercion. (3) Basic social and political institutions massively coerce subjects by enforcing a vast array of legal rules that shape the full extent of their life and liberty, including how they may acquire, transfer, and so on, property. (4) Those forced to live by this pattern of rules are therefore owed a special, more stringent justification for the resulting distribution than those who are not. (5) This special, more stringent justification, to be successful, requires the pattern of rules to realize a more demanding set of socioeconomic standards (e.g., egalitarian standards) among those have been coerced.
The key to the view is to explain how (5) follows from (4). When Blake, for example, writes that comprehensive coercion is justified only if no alternative feasible scheme could make the worst-off 22 I here leave aside the other changes that Blake's theory would need to make in order to accommodate this revision. Notice, for example, that if coercion were merely a sufficient condition, then Blake couldn't exclude other ways in which egalitarianism might be generated at the international level. among the coerced any better off, we should wonder: Why would it be unreasonable to reject a set of distributive principles that is limited in scope only to those coerced? And why must the justification take the specific form of the difference principle (or other similarly egalitarian standard)?
There are, I believe, two intuitively plausible answers that might help in explaining the jump from (4) to (5). On one variant, (5) follows from (4) because more demanding norms of distributive justice are understood as outweighing the initial wrong; on another variant, such standards are understood as compensating for the initial wrong. According to the Compensation variant, by coercing you, I infringe your right to autonomy, and hence I owe you special compensation for the wrong -a compensation paid in the currency of more demanding socioeconomic standards. 23 You have no reason to complain about being made in one sense worse off (by being coerced) because, all things considered, the more demanding distributive standard makes you much better off overall. According to the Outweighing variant, my infringement of your right to autonomy is not compensated but outweighed (without 'moral residue') by the urgency or weightiness of the general interests protected by the more demanding distributive standard.
In Blake's 2001 article, it was unclear which variant Blake adopted. In his book, he also doesn't explicitly come down in favor of one or the other, but, at least at one point, he seems to adopt Outweighing. He writes: '[A]ppropriate conditions of hypothetical consent [model the way in] which the moral harm of coercion might be nullified' (96). In the following, I will therefore assume the Outweighing reading of the argument. In previous work, I have argued at greater length that this interpretation cannot work. 24 In this section, I provide a foreshortened version of that discussion, and explore its implications for justice in foreign policy in the next. 23 The structure of the view is similar to discussions of 'moral residue' in the theory of rights. Many believe that when I break into your log cabin to protect myself from a storm, I infringe your property right, and hence owe you compensation as a result. See, e.g., J. When someone has been coerced, Blake is right that we must offer him a special justification for the coercion. But what form must that special justification take? Suppose your son would like to borrow your car to go to a party with friends, but you are not sure whether to give it to him. I offer you advice, suggesting that you should; after all, he is a trustworthy and reliable young man. My offering advice (whatever its content) is not something that requires any special justification. At most, I may have an obligation to give you sincere advice. But now suppose that instead of offering you advice, I coerce you -for example, by threatening to break your armin such a way as to ensure that you will give the car to your son, whether you want to or not. Now suppose that I coerce you in an effort to take possession of the car to rescue the lives of ten people who are drowning. We believe that threatening to break your arm in the latter case (where people are drowning) is justifiable, whereas in the former it is not. How did we make up our minds?
The general form of an act of coercion is this: A coerces B to do some action u in order to attain some end C. In wondering whether the coercion is justified, we then ask: does B have a morally weighty reason to u sufficient to outweigh or 'nullify' the wrong involved in coercing her to u? To answer this question, Outweighing asks us to tally up the moral weightiness of A's interests in B's u-ing, the moral weightiness of B's interests in not-u-ing, and the moral weightiness of the interests significantly affected by the pursuit of end C. We then see whether the net moral weightiness of the interests involved on both sides of our ledger outweighs the moral weight of B's interest in not being coerced (however that interest is construed). 25 To illustrate: The moral weight of my (A's) interests in your giving the car to your son (u) is (barring unusual circumstances) not very strong. The moral weight of your son (and his friends') interest in driving the car (C), it also seems plausible to assume, is also not very strong: at most, getting around will be more difficult. But your (B's) interest (your fear that he will get in an accident) in not giving the car to your son (not-u-ing) is much more morally weighty, as is the moral weightiness of your interest in not being coerced. Therefore we conclude, it is unjustifiable for me to coerce you. It should be clear that the same analysis can easily be used to demonstrate that coercion is justifiable in the case of the ten drowning people.
But here's the problem: the moral weightiness of A's interests in B's u-ing, the moral weightiness of B's interests in not-u-ing, and the moral weightiness of the interests significantly affected by the pursuit of end C can be identified without any reference to whether or not A coerces B. They remain, that is, constant across both scenarios in which coercion is present and those in which it is not. To be sure, whether the course of action is ultimately justified hinges on whether B has been coerced or not. But the moral weight of B's interest in not being coerced serves as a counterweight only once the net moral weightiness of the other interests has been determined. The moral weights assigned to the various interests involved are therefore not a function of how, or even whether, the ends at stake are attained by coercing B.
If this is right, then it means that coercion cannot bound the scope or fill the content of our distributive obligations. This is because the moral weightiness of our various interests in receiving a share of the social product are determined independently of whether coercion is the offing or not. To illustrate, consider that, just as in the case of the car and the ten drowning people, we do not need to take into account only the interests of the person or persons coerced. If the coercive act serves morally weighty third-party interests, then those can be used to justify the coercion all-things-considered. The interests of ten people in survival are, for example, very morally weighty, and, ex hypothesi, certainly weighty enough to threaten you in order to get use of your car. But if that is true, then, analogously, the fact that only citizens have been coerced by domestic law does nothing to narrow the range of interests that should be considered in justifying any particular act of coercion. It is one thing to say that coercion requires special justification, and another to say that it justifies treating the coercion-independent interests of those coerced differently than those not coerced. Blake's view only succeeds by eliding the distinction. Consider the following passage (which reflects a common pattern in the literature): 'We have to give all individuals within the web of coercion [call it group X], including those who do most poorly, reasons to consent to the principles grounding their situation by giving [group X] reasons they could not reasonably reject -a process that will result in the material egalitarianism of the form expressed in the difference principle, since justifying our coercive scheme to those least favored by it will require that we demonstrate that no alternative principle could have made them [emphasis mine] any better off' (95-96). But here we may wonder: why should only the interests of group X be considered in justifying the coercion of those in group X? Why wouldn't it be sufficient to say that those coerced, namely group X, have no reasonable objection to the coercion because the interests of others (in our case, nonmembers) are sufficiently weighty? Why couldn't we replace the last, italicized 'them' with some other group, Y (rather than the group X intended)? I do not see how appeal to some kind of hypothetical consent eliminates that possibility.
Blake's view must thus presuppose a prior set of entitlements (what I have called 'morally weighty interests') rather than provide grounds for a new one. They leave us, therefore, with no explanation of how and why the morally relevant features of coercion fix the content and scope of the obligations in question. At most, an independently justified set of distributive obligations affects the kinds of coercion that might be permissible in the circumstances, but no moral features of the coercion itself explain which obligations apply and why they only hold among those whose will has been bent. Subjection to coercion must there be a redundant part of any set of sufficient conditions which together grounds egalitarianism.

IV
In this section, I want to draw some implications of the arguments I have made in Sections I-III for Blake's overall account of justice in foreign policy. Recall that Blake conceives of his view as working out the consequences of his liberal commitment to four principles: (1) All individuals have equal moral status in virtue of their capacity to act as autonomous agents.
(2) All individuals have a right to those political institutions sufficient to protect their equal rights as autonomous agents. (3) All individuals have a right to have the coercive actions of such political institutions justified to them as autonomous agents.
Note that nothing I have said thus far puts into question any of these principles. I have been happy to accept them for the sake of argument. What I have tried to challenge is the inference from (3) to a domestically constrained egalitarianism. Indeed, once we put that connection in question, we might reasonably wonder: Given the foundational emphasis on autonomy, why focus so narrowly on only two ways in which our capacity for autonomous functioning can be set back, namely via comprehensive coercion (as in Blake's case for domestic egalitarianism), and via policies that leave us with less than some minimal threshold (as in Blake's case for a just foreign policy)?
Our capacity for autonomy can be set back, after all, in all sorts of ways, and to greater and lesser degrees. If our obligation to secure conditions for autonomous flourishing is truly global (as Blake avers), then, once we have broken the connection between coercion and distributive justice, why not take a much more globalist view? Why not say that we have an obligation, through whatever institutional means are at our disposal, to secure conditions for the autonomous functioning of every human being, wherever they live? On this view, domestic as well as supra-, inter-, and cross-national institutions and policies would simply be instruments for realizing the autonomous functioning of every human being. Of course, much more would need to be done to work out such a position. For example, we would still need to define some distributive principle to determine a principle for selecting among different feasible institutional schemes or policies: should we select the feasible scheme or policy that is 'closest' to a globally egalitarian pattern of autonomous functioning (or alternatively, an egalitarian set of social relations)? Or the one closest to a prioritarian solution? Or sufficientarian? And so on. The important point is that on such a globalist position coercion would be only one way in which autonomous functioning could be set back. I will not pursue this proposal any further since I do not think Blake would endorse it, and I also do not think it should be endorsed for independent reasons. I mention it only to point out a natural way in which one might develop the four principles Blake begins with once we abandon the narrow focus on coercion.