After a period of somewhat chaotic construction efforts, the dust is starting to settle on global justice theory. The alternative theoretical options are gaining clear shape. Mathias Risse's excellent On Global Justice is a work of judicious consolidation. He develops a nuanced and complex position that he calls “pluralist internationalism.” Its starting point is the claim that there are several different grounds of justice, that is, reasons for identifying a certain population of people and holding that they have claims of justice against each other, the proper adjudication of which is settled by a certain type of principle. Different principles may apply to different groups of people identified in different ways.
Some grounds of justice are relational; their basis is the relations in which we stand to one another and the social practices in which we are engaged. Some are nonrelational. A profoundly important relational ground of justice is shared membership in a political society. A significant nonrelational ground is shared humanity, simply being one human person in a world along with others. There are other relational and nonrelational grounds. For example, world trade would fall on the relational side, and common ownership of the earth would fall on the nonrelational.
Regarding shared membership and shared humanity, Risse tells an already familiar story, but adds key details that increase its plausibility. Merely being one human person among others engenders thin moral duties. What we owe to one another is the provision of institutional arrangements such that, so far as is possible, every human person is able to meet her basic human needs and flourish at least to some decent extent. In particular, attempts by global egalitarians to derive thick requirements of substantive equality of outcome or of opportunity (supposed to apply across all countries) from thin premises of merely shared humanity are unconvincing. Egalitarianism has a role in the theory of justice, but the scope of egalitarian requirements is confined to each separate country taken one by one. Those who share membership in a state are ensnared in a complementary mixture of coercion and community. The state coerces its members, and state coercion is made effective by reciprocal cooperation among those same members. Where there is shared membership in a political society, how one person fares compared with how others fare raises questions of justice, and the answers ultimately should lead us to embrace broadly Rawlsian egalitarian distributive justice principles.
Embracing these commitments, Risse's pluralist internationalism rejects globalism (the claim that one exhaustive set of distributive justice principles holds at all times and places and does not vary in its fundamental requirements from country to country or from community to community), monism (the claim that there is only one ground of justice requirements), and anti-relationalism (the denial that social relations and the practices in which people are now engaged can be a ground of justice).
Risse's pluralist internationalism accommodates the common-sense conviction that national partiality is morally acceptable and even admirable and mandatory. His theory has no truck with the claim that being part of a like-minded national community united by language and culture and the political aspirations standardly linked to nationalism is participating in a per se valuable social relationship that justifies partiality. According to Risse, community identity is not the basis of acceptable national partiality. At the same time, by affirming plural bases for justice obligations, Risse avoids the blunt implausibility of claiming that, for example, simply being subject to coercion in concert with others somehow triggers the application of the difference principle. Along the same line, Risse's theory is not embarrassed by the evident fact that social relations and density of social connections vary by degree, and that between the comprehensive community of the national state and the sheer fact of being one person among others there are likely to be intermediate sources of justice bonds. Although I criticize Risse's synthesis in what follows, we should pause to acknowledge its multiple, novel, and insightful merits.
Cosmopolitanism
Risse, like other recent global justice theorists, finds nothing of interest in disputes between cosmopolitans and their opponents or among those advocating different versions of cosmopolitanism. The problem with these disputes could be put this way, following a diagnosis offered by Samuel Scheffler: On a weak reading, cosmopolitanism is the claim that we are bound by some moral duties owed to people in virtue of our common humanity. On a strong reading, cosmopolitanism is the claim that we are bound by no duties except those owed to people in virtue of our common humanity. But the weak cosmopolitan claim is obviously, uncontroversially true, and the strong cosmopolitan claim is so wildly implausible as to be not worth discussing.Footnote 1
A nonplatitudinous and yet defensible cosmopolitanism can be identified whether or not strong cosmopolitanism turns out to be defensible. Consider an intermediate position that we can call moderately strong cosmopolitanism. This holds that (1) we are bound by moral duties owed to people just in virtue of their common humanity; (2) we are also bound by special-tie, agent-relative moral duties of partiality that arise in some per se valuable social relationships, including friendship, love, and close family ties; and (3) we are not bound by any special-tie agent-relative moral duties that arise in particular large-scale impersonal social relationships, such as shared membership in a state or shared membership in a national community.
Is moderately strong cosmopolitanism viable? To answer this question we need to turn to the concept of shared membership in a state, Risse's basis for the justification of profound and far-reaching relational social justice principles that he associates with Rawlsian egalitarianism. I am going to claim that when we examine Risse's views on shared membership, we turn up reasons to embrace moderately strong cosmopolitanism.
Shared Membership in a State
The pluralist internationalist accepts a roughly two-tier account of global social justice. As to the first tier, there are minimal justice duties of common humanity that each of us owes to every other human person. These global-scope duties are sufficientarian in character; each person is entitled to access to circumstances adequate for a good enough quality of life. As to the second tier, within each country the social relation of fellow citizens conjures up stringent social duties owed by each member to every other to provide equal opportunity or maximal advantages for the worst off (or something along this line).
Note that Risse's two-tier account may gain unearned plausibility by its claim that within each nation-state justice demands distributive equality of some sort. A simple argument can be pressed that starts with this claim: within each separate country, egalitarian distributive justice principles hold, but it is incorrect to hold that egalitarian justice principles hold across the globe, and it is incorrect to hold that no distributive justice principles hold across the globe, so there must be distinct sets of distributive justice principles, one with national scope and one with global scope.
It is indeed possible to maintain that no egalitarian justice principles hold across the board. But there is another possibility that merits highlighting: maybe the correct distributive justice principles hold across the globe but are not egalitarian. That is, none of them demands that everyone get the same or has the same. If this is so, then the implausibility of equality as a global justice requirement is no objection to globalism.
To focus our thinking on this point, notice two alternative nonegalitarian and nonrelational conceptions of justice or fundamental morality that might occupy the space that in Risse's worldview is occupied by a two-tier or split-level conception. One is a modified Lockean natural moral rights theory. On this view, the rock-bottom foundation of morality is the claim that each person has certain natural moral rights, along with a natural moral duty to respect the rights of all others and to promote their fulfillment.
Another candidate conception of justice that eschews any egalitarian (or for that matter any essentially comparative justice) requirements is prioritarian welfarist act consequentialism. On this view, one morally ought always to do some act among the alternatives available that would bring about an outcome no worse than the outcome of anything else one might instead have done. The value of outcomes is a function of the welfare or well-being accruing to individual persons that the outcomes contain, and the value of obtaining a benefit for a person is greater the larger the well-being gain it elicits, and greater the lower the person's lifetime well-being would otherwise be, absent this benefit. These two moralities do not exhaust the nonrelational alternatives. Here I am simply pointing out that the denial of substantive equality is a live option in moral theory.
Let us now turn to Risse's considerations for two-tier justice—equality for fellow citizens, sufficiency for foreigners. Risse singles out two features characterizing the relationship among members of a state that work in tandem to render state membership normatively unique: the state coerces those who inhabit its territory, and at the same time the state is a cooperative endeavor. That is, citizens work together to maintain the state.
State coercion according to Risse is marked by two types of immediacy: “legal” and “political.” By “legal immediacy” Risse means that “state enforcement agencies have direct, unmediated access to bodies and assets” and that the regulation of our lives by state enforcement is pervasive. By “political immediacy” he refers to the fact that “states provide the environment in which basic rights are, or fail to be, realized” (p. 27).
Why suppose, however, that the facts of legal immediacy—assuming the claims are roughly correct—provide reasons for the claim that state membership is a ground of justice? Suppose all law enforcement was executed by complex, indirect, causal chains, like Rube Goldberg machines. Would this somehow invalidate the claim that state membership is a ground for justice? I want to claim the answer is “no.” No doubt the pervasiveness of law enforcement is reason to think it is important that law enforcement should not violate moral standards and should advance the morally important goals that are proper functions of the state. But the fact of pervasiveness does not help orient us toward what these correct moral standards and proper functions are.
The political immediacy idea is that if state coercion is done in a certain way, this contributes to the realization of basic rights for those affected by the coercion, and if state coercion is done in other ways, this impedes the realization of basic rights for those affected by the coercion. So if we accept basic rights, we should hope that state coercion is carried out in ways that advance rather than impede their fulfillment. I do not yet see how this takes us closer to seeing what special moral rights, basic or otherwise, people have in virtue of being members of states.
Taken by itself, coercion is an unlikely candidate for the role of triggering the applicability of principles of justice that would otherwise be inert. Let us accept that when one person gives another orders backed by an effective threat that leaves the second no reasonable alternative but to obey, some justification is owed to the coerced individual. But the justification need not invoke benefits to the coerced person. Suppose you threaten me with penalties if I steal from you, and this threat effectively coerces. Here it would be an adequate justification of what you are doing to point out that I am anyway duty-bound to refrain from stealing, and this is a duty of a kind that is apt for enforcement, and no special moral license is required for permission to engage in enforcement by coercing.
Granted, state coercion often bears little resemblance to this scenario. States impose some particular version of contract and tort and administrative law out of many that might instead have been selected, and choosing one particular version rather than another has differential effects on people's interests. So no doubt this type of coercion also requires justification, but again, it does not follow that the justification has to consist of pointing to benefits to the coerced. (Maybe some coercion is justified by the fact that it effectively facilitates the coerced people to help outsiders, such as distant needy strangers, in ways that they are anyway morally obligated to undertake.)
Perhaps coercion is justified sometimes—but only if one compensates the person who is coerced. But those, like Risse, who see the coercion standardly employed by states as triggering the application of special principles of justice not otherwise applicable have not provided arguments purporting to show that standard state coercion is one of those cases in which coercion is unjustified unless accompanied by compensation to those disadvantaged by it, much less arguments purporting to show that the compensation must take the form of the difference principle.
Turning to cooperation and reciprocity, we find that matters are even more complicated. Reasonably well-functioning states depend on the allegiance of citizens and on their disposition to comply with legal rules even when the threat of penalties for noncompliance is absent. Consider the Hart-Rawls principle of fairness, originally formulated to help explain political obligation. The key idea is that if some members of a group cooperate to provide goods that are nonexcludable (if any group members consume some of the good, none can be excluded from consuming some), and especially if they provide goods that are nonoptional (if any in the group consume some of the goods, all must consume some), the mere receipt of benefits can obligate one in reciprocity to pay one's fair share of the costs of the provision of the goods. Moreover, the obligations thus incurred can be legitimately enforceable. The Hart-Rawls principle tells us in effect that you should pay a fair share of the costs of the public goods you consume for essentially the same reason that you should pay for the groceries or other private goods you purchase, except that in virtue of the special characteristics of public goods (in particular, nonexcludability and nonoptionality), provision of them via voluntary exchange in free markets will generally not be efficient, so public provision and state enforcement can be a reasonable delivery system.
None of this goes a step toward supporting the grand view of public goods provision articulated by some global justice theorists, including Andrea Sangiovanni and Risse. According to the grand view, fellow members of the state cooperate to supply you with the basic goods of security and safety and mutual trust that make it possible for you to exercise agency and pursue a decent plan of life. The obligation of reciprocity incurred by receipt of these precious benefits is nothing less than acceding to egalitarian principles of justice.
Let us grant that basic security and safety are important public goods. But the networks of cooperation are many and various and do not coincide neatly with state borders. Living in San Diego, I am far more dependent for my basic safety on the law abidingness of foreigners in northern Mexico than of fellow Americans living in Maine or the hollows of Kentucky, and I am more reliant on the effectiveness of the law enforcement efforts of northern Mexican states than of law enforcement efforts in much of the United States. Complex webs of cooperation and public goods provision—which may be local, regional, national, and international—generate various and sundry obligations of reciprocity to pay one's fair share of costs.
Someone might comment that I have conceded too much ground to the reciprocity argument for national partiality. After all, I have granted that receipt of benefits from a cooperative scheme supplying public goods triggers obligations of reciprocity, and surely some such cooperative schemes might have a scope that coincides with national borders. Hence, the moderately strong cosmopolitan loses the argument on my own showing. So it might be claimed.
The moderately strong cosmopolitan has room for multiple responses. First, a consequentialist will flatly deny that there is a nonderivative reciprocity obligation. Second, even accepting the reciprocity obligation as pro tanto binding, it generates reasons to reciprocate to all (and only) voluntary participants, a group that will not be coextensive with the full set of fellow co-nationals. Third, and in this context most important, there is a gap between the idea that one should pay a fair share of the costs of (nonoptional) public goods one receives and the idea that this fair share amounts to the introduction of egalitarian principles of distributive justice. Consider the classic example of peaceful farmers living in an isolated valley and menaced by bandits. Some farmers initiate a sentry and armed guard system that brings about increased safety, benefitting all valley residents. So recipients of safety should pay a fair share of the costs of provision. This does not somehow bring about a general egalitarian norm of distributive justice binding all valley residents.
Others claim that it is neither state coercion nor cooperation regarding public goods provision by itself that triggers a special justice duty of partiality toward co-nationals, but rather the combination of the two. I for one do not see the “moral magic” in the suggested combinations. It would be helpful, in this impasse, for those who see the proposed combination as specially reason-generating to help the rest of us by supplying some hints as to why and how this might be so.
Egalitarian Justice, Equality as per se Morally Valuable
Risse suggests that the two-tier account of distributive justice that he proposes can be filled in, on its nation-state side, by a broadly Rawlsian theory of distributive justice. He does not argue for this; it is an assumption.
Following G. A. Cohen, Derek Parfit, and others, I find the Rawlsian treatment of equality as a moral requirement within a nation, regarded as a cooperative scheme for mutual advantage, to be profoundly puzzling and ultimately unacceptable. With Parfit, I believe we should reject distributive equality; and with Cohen, I believe we should reject justice as fundamentally regulating a cooperative scheme for mutual advantage.
Chipping away at the claim that distributive justice at the level of the nation-state requires substantive equality of some sort does not undermine relationalism. Perhaps the (possibly nonegalitarian) principles of distributive justice that apply to us vary in the stringency of their requirements depending on the density of social connections among people or some other feature of their social relations.
The Alleged Moral Magic in Relations: Nozick's Thought Experiment
Robert Nozick once proposed a thought experiment that bears on the plausibility of Risse-style relationalism. Nozick imagined individual human persons living isolated on islands, one person per island. The individuals do not interact, but could help one another. Seeing that the resources of my island are sparse and that my talents at extracting goods from my island's resources are meager, you could place resources in an unmanned boat that would drift with the tides to my island. Nozick suggested that if no one has any enforceable justice obligations to aid anyone else in this scenario, then simply initiating interaction by making mutually profitable voluntary trades does not trigger the application of the difference principle or any other distributive justice requirements. One who holds that there is a strong beneficence component to morality can agree with Nozick's hypothetical claim but deny the antecedent. That is, in the scenario Nozick describes, the individuals who are not interacting but could aid others can be under enforceable obligations to do so. If we had no obligations to others absent interaction, the initiation of voluntary trading would not cause obligations to others to arise, beyond such requirements as refraining from theft and fraud and breach of contract. But we do have positive obligations absent interaction.
A relationalist could respond that Nozick is misled here by the paucity of interaction he has introduced. If justice requirements vary, perhaps discontinuously rather than by some continuous function, then we could allow that merely trading some bananas for figs once would not plausibly introduce strong distributive justice requirements, but initiating long-term trade relationships and mutual dependency would do so. Maybe just a bit of interaction introduces just a bit of relational justice, perhaps too small to notice. But change the example to involve large-scale interaction and perhaps social relations of other sorts and Nozick's thought experiment ceases to support the hypothetical claim he aims to establish.
This would be a possible move, but not a plausible one. To see the continuing force of Nozick's thought experiment, imagine thick trade relations developing among relatively well-off isolated islanders. Sally does not behave dishonestly toward Samantha, her trading partner, but continues to bargain hard and treat Samantha as owed no more and no less than honest trading. A Nozickian would say Sally is morally at liberty to do whatever she wishes with the wealth she gains from trading with Samantha, providing she does not use the wealth to engage in wrongful harming of anyone. A social justice theorist who includes a strong beneficence requirement in a natural moral rights doctrine will hold that Sally may have strict enforceable duties to others who are singled out as morally apt recipients of beneficence. Now consider the relations between Sally, Samantha, and Sarah. Sarah is badly off, and hence more eligible for aid than Samantha.
Here is a test case for relationalism of the Risse variety. I say Sally owes nothing special to Samantha and is within her rights, and perhaps strictly morally required, to bargain tough in dealings with Samantha in order to accumulate more resources for aiding Sarah. Let Sarah be just above whatever level of quality of life Risse and followers regard as the sufficient level that we owe one another under the human rights requirements that are called into play by common humanity or common personhood. So Risse-style common humanity requirements do not exert any moral force on Sally's decision-making, but the density of social connections between Sally and Samantha is supposed to bring about greater moral demands on each one.
I want to claim that the social relations between trading partners do not trigger strong moral requirements, but the neediness and disadvantage level of other people do provide direct and strong moral reasons to improve their condition. Let there be many Sallys, Samanthas, and Sarahs, and let the Sallys and Samanthas be bound together under a single nation-state. None of this changes the moral relations among the islanders, or the moral verdict prompted by Nozick's simple thought experiment.
Suppose now that rich and poor live in one society and, according to relationalists, are bound together by egalitarian justice bonds. Finding the ties onerous, the rich propose seceding, and no longer desire being caught in a web of cooperation and coercion with the poor. Forming Richland, they now owe egalitarian justice only to fellow advantaged types, and they trade with former compatriots (they export their clothes to be laundered in Poorland) at whatever level is just below whatever the relationalist theory under review says is the trigger for egalitarian justice ties.
Maybe the relationalist theory under review denies it is acceptable, once social relations are in place, to undo them by secession. This would need some backing argument; it is not obvious this move succeeds. Suppose it does. Then imagine that at an earlier time in history, genetically favored individuals who expect that they and their descendants will be on the average more talented and better off than the genetically less favored scrupulously avoid interaction with the disfavored of a sort that would trigger obligations of egalitarian justice according to relationalist theory. So Richland and Poorland arise without the need for secession. For what it is worth, I entirely lack the relationalist intuition that the history of relations and the character of social relations matter for what we fundamentally owe one another.
Of course, my disagreement with Risse on this key issue does not amount to a refutation of his position. Everyone will agree that social relations matter to what we owe to one another: we have duties to our parents and siblings that we do not have toward strangers. The nonrelationist claims that these relational duties are in every case explainable as being useful tools for better fulfilling nonrelational moral claims. For example, a utilitarian will hold that we should always do what maximizes the general welfare, and the general welfare will be boosted if we all accept special duties to care for immediate family members. The relationist will hold that being in certain social relations can trigger fundamental moral duties that are not just instruments for fulfilling other norms. To see better whether relationalism or its rejection is correct, we need to become clear about what are its most plausible specifications and how these might be supported. In this enterprise we should all be grateful for Risse's careful elaborations.
To conclude, much recent global justice theory consists of a search for plausible arguments supporting what some regard as an overwhelmingly plausible common-sense conviction: we owe more to fellow countrymen than to mere foreigners. In my estimation, this search has come up with empty pockets. Maybe we should entertain the possibility that our common-sense convictions regarding the moral significance of nation-state membership are wrong, and the failure of the search for solid underpinning arguments is explained by the incorrectness of the common-sense convictions we are trying to buttress. Risse's book is so far the most sophisticated elaboration and defense that any political theorist has developed of these common-sense convictions concerning national partiality. His book is an excellent achievement, wherever the truth of this complicated matter may lie.
