I. Global justice in liberal political philosophy

Questions of economic justice across borders began to appear on the radars of liberal political philosophy only in the 1970s. This development – like many of the other developments that occurred in liberal political philosophy during and since that decade – was heavily indebted to John Rawls’s work, and specifically to his A Theory of Justice.Footnote ³ In order to indicate what it means to think about global justice within a Rawlsian framework, I must survey briefly some of Rawls’s key ideas that bear on our discussion (though these are likely to be familiar to some readers).

In A Theory of Justice Rawls seeks a public criterion of social justice that can regulate a liberal democracy’s ‘basic structure’ – its ‘major institutions’, including ‘the political constitution and the principal economic and social arrangements’.Footnote ⁴ Rawls derives this criterion from the contractual thought-experiment of the ‘original position’, the parties to which represent democratic citizens. These parties know ‘general facts about human society’ (such as ‘the principles of economic theory … the basis of social organization and the laws of human psychology’).Footnote ⁵ However, a ‘veil of ignorance’ deprives them of any knowledge of the particular individuals they represent and thus ensures that morally irrelevant (in Rawls’s language, ‘morally arbitrary’) features do not influence their decision. Rawls argues that, so situated, the parties will opt for two principles of justice over other alternatives (most importantly, utilitarian alternatives). The first principle demands that all citizens will be accorded equal basic rights and liberties. The second principle, alongside fair equality of opportunity in attaining ‘offices and positions’ in society, contains Rawls’s famous difference principle – the demand that ‘social and economic inequalities are to be arranged so that they are to the greatest benefit of the least advantaged’.Footnote ⁶

Though explicitly designed for an idealised, ‘closed’Footnote ⁷ domestic society, Rawls’s theory seemed to have at least two important global parallels. First, some philosophers have argued that global economic interdependence generates a global basic structure that is a subject of justice in much the same way as the domestic basic structure. Rawls held that the major institutions that comprise the domestic basic structure, such as a society’s property and legal system, should be the main focus of a theory of justice, given their pervasive impact on individuals’ ‘life prospects, what they can expect to be and how well they can hope to do’.Footnote ⁸ Some of Rawls’s main interlocutors have held that the same ought to be true, analogously, for global economic institutions. To introduce a key example that will occupy us at length here, Thomas Pogge, a prominent Rawls student and one of the pioneers of the global justice debate, has considered it simply ‘undeniable that, today and in the foreseeable future, there is a global institutional order that importantly affects the options and incentives societies and their rulers face in their relations with one another and even affects profoundly the domestic institutions and cultures of especially the smaller and weaker societies’.Footnote ⁹ According to the ambitious version of this reasoning, if one endorses Rawls’s domestic theory, one ought to replicate it globally and endorse a global difference principle, holding that global inequalities ought to be arranged so that they are to the greatest benefit of the (globally) least advantaged. According to the more minimal version, even if the demands of global justice fall short of a global difference principle, substantive duties of justice (rather than mere charity) still apply to the global economic order.Footnote ¹⁰

A second important global parallel that Rawls’s interlocutors have drawn from his domestic theory concerns the idea of moral arbitrariness. At the domestic level, Rawls held that the ‘veil of ignorance’ should prevent the distortion of justice by morally arbitrary features – that is, unchosen features that are irrelevant to the assessment of agents’ moral worth, such as their birth into a certain social class and their natural talents. These features reflect nothing but a ‘natural lottery’, and should therefore not be allowed to determine individuals’ income and wealth.Footnote ¹¹ Charles Beitz, another key figure in the global justice debate, has applied this idea to the global level, and argued that the different natural resource endowments of different societies, being a paradigmatic example of ‘the natural lottery’, ought not be allowed to determine societies’ income and wealth. According to this reasoning, just as in the domestic case Rawls writes that we should regard ‘the distribution of natural abilities in some respects as a collective asset’,Footnote ¹² Rawlsians must similarly regard the global distribution of natural resources as a collective asset, and redistribute these resources in accordance with a global difference principle.Footnote ¹³

Pogge’s argument concerning a global basic structure and Beitz’s argument concerning ‘the natural resource lottery’ remain the two most important arguments extending Rawlsian principles to the global level, and have shaped much of the ongoing liberal conversation on global justice. In Beitz’s case, this lasting influence is evident in the fact that even authors who depart from his specific Rawlsian formulations often continue to discuss rights to revenue from natural resources, based on the thought that the moral claim that nations can make to this revenue is far weaker than their claim to their ‘non-arbitrary’ social and political capital.Footnote ¹⁴ Pogge’s influence, in turn, has resulted from the radical implications of his conception of a global basic structure, and, more specifically, from his widely debated claim (discussed below) that affluent democracies, through global institutions, not merely fail to help the global poor but actively harm them, sharing active responsibility for their plight.Footnote ¹⁵

The enduring significance of Beitz’s and Pogge’s accounts is interesting from the perspective of social science, since both of these accounts seem – initially, at least – amenable to social-scientific research. Accordingly, I now turn to examine how each of these arguments interacts with social science.

II. Social science and ‘the natural resource lottery’

III. Social science and ‘the global basic structure’

Since the beginning of the millennium, Pogge has been using the idea of a global basic structure to make the claim that global economic institutions ought to prevent the global poor from dropping below the threshold set by the Universal Declaration of Human Rights (UDHR). What makes this claim both dramatic and interesting from a social science perspective, however, is the nature of Pogge’s concomitant moral accusation: since the 1970s, scholars have conceived democracies’ global duties primarily as positive duties to help the global poor, yet Pogge argues that much before helping, citizens of affluent countries – and particularly we, citizens of affluent democracies through our elected governments – are harming the world’s deprived. Through global economic institutions, we not merely fail to help but actively contribute to 50,000 daily deaths from poverty related-causes.Footnote ³⁸

This provocative claim has left few global justice theorists indifferent. Some supporters of Pogge’s views, and even certain critics, have characterised his arguments as ‘groundbreaking’.Footnote ³⁹ Others, however, have dismissed Pogge’s claims as ‘entirely speculative’, insisting that it is implausible to attribute to citizens of affluent countries such ‘barbaric’ crimes.Footnote ⁴⁰ Empirical issues are at the heart of the heated ‘Pogge debate’, and these issues we will occupy us in a moment. Before we turn to them, however, we need to note two things. The first concerns the moral significance of the distinction between negative duties not to harm and positive duties to help. One reason why Pogge wishes to recast the problem of poverty as a matter of negative duties not to harm is that doing so prevents conflict between the demand to ameliorate poverty and ‘the grain of the Anglo-American moral and legal tradition, in which it is often denied that persons have duties to protect and aid other persons in distress’.Footnote ⁴¹ Another reason, specifically with regard to global poverty, is that the negative duties strategy bypasses the dilution through distance that afflicts positive duties – that is, the idea that our positive duties to help become weaker the more we move ‘outward’, away from our near and dear and to ‘distant’ others.Footnote ⁴²

Second, in order to see why these advantages of negative over positive duties depend on empirical considerations, we also need to note Pogge’s actual argument connecting negative duties to poverty and global institutions. Pogge begins this argument by claiming that persons have negative duties to enact and comply with just institutions, since the full content of individual rights and duties cannot be known solely through rules of personal conduct, apart from institutions. Even if I wish merely to satisfy my negative duties not to violate your rights, I cannot do so outside of an institutional framework specifying what your full rights are and what is therefore actually required by my negative duties to respect them.Footnote ⁴³ This, in turn, means that my duty to enact and support just institutions can itself be understood as a negative one. Since we clearly ‘need shared institutions to avoid invading one another’s freedom’, we ‘have a negative duty to comply with such existing institutions, whenever non-compliance can succeed only if like non-compliance by others is constrained’.Footnote ⁴⁴ Once individuals have a negative rather than positive duty towards institutions, they also become right-bearers – and therefore also bearers of negative duties to respect rights – by virtue of cooperating with others under a shared coercive institutional order.Footnote ⁴⁵

Pogge believes this institutional understanding of rights and correlative duties must be replicated at the global level, among other things since he considers it evident, as noted above, that a coercive institutional global order exists. Specifically, Pogge sees global organisations, from the WTO through the World Bank and IMF to the G8, as the main foundations of this institutional order, skewing it in favour of the global rich against the global poor, in ways ranging from the rules allocating control over natural resources, through regulations of intellectual property rights, to ‘quotas, tariffs, anti-dumping duties, export credits, and huge subsidies to domestic producers … that poor countries are not permitted, or cannot afford, to match’.Footnote ⁴⁶ In these and other ways, Pogge accuses, global institutions limit the poor’s access to UDHR provisions, thus sharing much of the blame for the poverty-related death toll that ‘matches, every three years, the entire death toll of World War II, concentration camps and gulags included’.Footnote ⁴⁷ Pogge insists that ‘minor modifications in the global order that would entail at most slight reductions in the incomes of the affluent’Footnote ⁴⁸ will suffice to generate very different outcomes for the global poor, and secure their access to UDHR provisions. Our failure to pursue such modifications violates our negative institutional duties of justice.

This background should make clear how Pogge wishes to connect global institutions, global poverty, and negative duties. But it also makes clear how Pogge’s argument hinges on empirical elements that are far from consensual. As Wenar emphasises, ‘[I]n order to carry through on his harm argument, Pogge must show not only that current global rules generate very bad consequences, but that we can be fairly certain that different rules would do better.’Footnote ⁴⁹

This thesis, however, faces two key social-scientific problems, paralleling the issues that were apparent in our discussion of the natural resource lottery. The first problem is that, just as in the case of natural resources, there is a gap in social-scientific knowledge. Social scientists simply do not know enough on the impact that global rules have or might have on extreme poverty. Pogge himself, it is important to note, is keenly aware of this gap, emphasising that evidence on the causal role of global institutions with regard to global inequality is ‘sorely lacking today’.Footnote ⁵⁰ Indeed, Pogge explicitly sees his arguments as potential motivators for social scientists who might address the gap: ‘One practical aim of my work’, as he put it recently, ‘is to stimulate more serious social-science research into the impact of global institutional design decisions upon the massive persistence of severe poverty.’Footnote ⁵¹ However, just as in the case of the natural resource lottery, it is not clear that social scientists can fill the empirical gaps of the normative argument.

Pogge’s argument hinges on the question of how extreme poverty would have evolved, had different global institutions been in place. Yet it is far from clear that social science can assess this huge counterfactual.Footnote ⁵² Arguably the most important difficulty here is what we might call the one world problem. We only have one set of global institutions in place at any given time. This means that we cannot compare existing global institutions to another set of global institutions while ‘holding everything else fixed’. Pogge is again aware of the issue: ‘there being only this one world to observe, it is hard to obtain solid evidence about how the overall incidence of poverty would have evolved differently if this or that global factor had been different’.Footnote ⁵³ But how to solve the issue remains unclear, and in the lack of a solution, we remain unable to make confident judgments concerning the causal relationship between global institutions and poverty trends. Whether one believes that global poverty is increasing, decreasing, or remaining the same (itself a much-disputed question),Footnote ⁵⁴ some might argue that the trend is because of global institutions, others might argue that the trend is despite of global institutions, while others still might argue that global institutions are a relatively minor factor either way. Unable to hold everything else constant aside from these institutions, however, it is far from clear that we can confidently prove any of these options.

We can see this problem at play when we examine specific central issues, such as the relationship between global and domestic institutions. Arguably, in order for Pogge’s argument to work, it must be the case that reforms in global institutions will significantly increase the likelihood of reforms in poor countries’ domestic institutions. But it is far from obvious that this is indeed the case. In an essay titled ‘Philosophy, Social Science, Global Poverty’, Joshua Cohen, for example, insists that there is ‘no reason to accept the claim’ that

My point here, however, is not to endorse ’s criticism as such. Rather, I merely wish to point out that it is extremely difficult for social science to construct reliable models that will support either Cohen’s pessimism or Pogge’s optimism regarding the causal force of global institutions. The enormous number of potentially confounding factors involved in the global economy poses structural difficulties towards creating such models.

More specifically, the combination between the one world problem and the ever-growing complexity of the global economy yields at least three obstacles to social-scientific analysis, all of which are evident when we ask how global institutions impact domestic ones. First, even if we could be confident that certain global institutions have a negative impact on the domestic institutions of poor countries, it is very hard to assess their relative weight in comparison to many other likely causes. Thus for example, many global dynamics that are in principle independent of global institutions, and are not always completely amenable to institutional interventions (such as dramatic fluctuations in the supply and demand for certain valuable commodities), can themselves be a significant cause of institutional crises in poor countries. Yet it is hard to assess their relative weight in comparison to global institutional factors.

Second, and very much related, global institutional reforms necessarily interact with an extraordinary number of causal levers, which in turn interact amongst themselves in unknown and potentially undesirable ways. This makes the effects of global institutional changes extremely difficult to predict.

Finally, all of these problems are compounded by the fact that not only existing global institutions might be novel: certain proposals for global reforms might be even more novel – indeed, unprecedented. One of Pogge’s reform proposals, for instance, involves ending the ability of odious regimes to borrow funds and sell natural resources in their peoples’ name.Footnote ⁵⁶ This proposal (which will also occupy us below) in effect requires an extraordinary number of countries – specifically in Pogge’s scheme, all of the world’s democracies – to sanction a very large number of regimes. Predicting what would be the result of such reform is extremely difficult not only because of the number of factors that might be involved, but also because such a global policy has never been carried out before: we have even less comparative evidence to work with in such a case than we do in other global contexts. The combination of one world, multiple factors and unprecedented change is especially difficult for social scientists to assess.

IV. What ought to be done? Reform under uncertainty

All of this brings us to a practical question, which applies both to Beitz’s proposal of global resource redistribution, and to Pogge’s charge of global harm: what ought global justice theorists do when they not only lack social-scientific evidence that seems necessary to motivate certain reforms, but also face serious obstacles in attaining such evidence?

One answer that may initially seem appealing is that global justice theorists should retain the reform proposals that suffer from social-scientific uncertainty. The thought here would be that it is worthwhile to pursue some reforms even given deep uncertainty, whether because of the mere chance that they would achieve important gains for the world’s poor and oppressed, or because of the hope that these reforms will be worthwhile even if their benefits would be limited. Thus for instance, even if a global redistribution of natural resource wealth, or reform of the WTO, will not achieve much by way of combating global poverty, it would still achieve something.

I do not wish to deny that this strategy will make sense in some cases. But we should be cognisant of its important limitations. One limitation concerns again the distinctive structural circumstances of global politics, and specifically the structural problem of global collective action. In the enduring lack of a global sovereign, no state can have assurance that other states will join it in undertaking reforms meant to improve the condition of the world’s poor and oppressed. This structural fact makes the deep uncertainty regarding the impact of many global reforms particularly burdensome, since it allows each state to excuse itself from undertaking the sacrifices involved in solitary action. Global collective action problems can run deep even where there is reasonable certainty that joint efforts by multiple states will actually achieve important moral aims: even under such happy circumstances, each state may often morally excuse itself from the sacrifices of solitary action, by claiming that, lacking credible commitment from other states, such sacrifice would be ineffective. But when there is deep social scientific uncertainty as to whether important moral aims will be realised even in the best case scenario, of full cooperation by a critical mass of states, it is especially easy for each state to provide moral excuses for its refusal to take action.Footnote ⁵⁷

Alongside the issue of global collective action, there is also the matter of opportunity costs. Because there are always opportunity costs to prioritising any given reform proposal over others, the mere hope that some reform might have limited benefits for the world’s poor and oppressed is often not a sufficiently compelling empirical foundation on which to base a normative call for political action. In such a case, the normative political theorist (at least one who is practically inclined) is better off shifting focus elsewhere, to discuss moral principles related to different potential reforms that have a greater chance of making a tangible real-world impact. Furthermore, a strategy that settles on the hope for limited global gains ignores not only practical but, for lack of a better term, ideational opportunity costs. There are ideational opportunity costs involved in pushing for reforms that rely on certain moral principles if other potential reforms that rely on conflicting principles can be more effective.

To make these claims (especially the latter) concrete, consider the conflicting moral principles underlying, on the one hand, Beitz’s proposal for a global redistribution of natural resource wealth, and on the other hand, the various proposals made in recent years to end the essentially neocolonial practice under which dictators enjoy customary rights to sell their state’s natural resources to any willing buyer.Footnote ⁵⁸ There are clearly conflicting normative principles at play here: the reform of dictators’ ‘resource privilege’Footnote ⁵⁹emphasises the idea of national ownership over natural resources, an idea that the Beitzian argument wishes to undermine. The key thought underlying recent philosophical discussions of dictators’ trading privileges is that a state’s resources belong to its people rather than to rulers, and that dictators (and corporations) who are transacting in state resources without obtaining proper consent from their people are therefore violating national property rights. Rather than contrasting the claim of national ownership with the claims of humanity as a whole, reform of dictators’ trading privileges focuses on the contrast between the claims of sovereign peoples and the claims of those who wield effective political power. Insofar as the grounding rationales for these two proposals are in tension with one another, pursuing a global redistribution of natural resource wealth has a clear philosophical opportunity cost, in undermining the normative case for national resource reform.

In light of these difficulties, I would like to propose an alternative normative strategy for dealing with the structural uncertainty of global social science. This strategy sees the problems of global social science as an important reason to examine normative arguments for global reform that are far more independent of empirical and predictive uncertainty. These arguments, in turn, focus on self-referring moral reasons that liberal polities have against entanglement in certain foreign practices.

Two examples should suffice to give a sense of what such self-referring reasons can look like. First, consider the divestment of almost all liberal democracies from apartheid, culminating in the late 1980s. Divestment could have been justified through consequentialist reasoning, as a way to precipitate the collapse of the regime. But it is difficult to prove that apartheid would not have collapsed at about the same time as it actually did, even if the regime was less isolated (as was desired, for example, by the Reagan administration preaching ‘constructive engagement’).Footnote ⁶⁰ This difficulty provides a further reason to examine an alternative, self-referring moral perspective on disengagement from apartheid. One might argue that liberal democracies had moral reasons pertaining to their own integrity to divest from a blatantly racist regime, independently of the social-scientific question as to whether divestment would be effective in precipitating apartheid’s demise.

A second example has to do, once more, with dictators’ ‘resource privilege’. One way to see an end to dictators’ ability to sell their peoples’ natural resources is simply as a means to destabilise dictators’ grip on power. But this perspective once again brings us to extremely complex social scientific questions, such as whether even a wide-scale boycott of dictators’ resource exports can indeed be effective in weakening them, instead of simply producing ‘rallying behind the flag’, and imposing costs primarily on the population, as the sanctions literature often warns. A self-referring perspective, however, would emphasise the thought that democracies have principled reasons to end their own complicity in dictators’ theft of their peoples’ resources, independently of the social-scientific uncertainty over whether doing so will also destabilise the regime. One could argue that even if political and economic circumstances in most dictatorships will remain the same as a result of liberal democracies prohibiting natural resource transactions, such a prohibition would still be morally required, just as ending colonialism would have been morally required even if one could not prove that doing so will make the victims of colonialism better off.Footnote ⁶¹

To prevent misunderstanding, I should emphasise that taking seriously such principled reasons for global reform, does not mean adopting an indifference towards the consequences of reform.Footnote ⁶² We can see this through the examples just given. If there had been, at the time, credible evidence that divestment from apartheid will threaten the basic subsistence of apartheid’s victims, then liberal democracies’ self-referring moral reasons to divest from apartheid would have been outweighed by opposing consequentialist considerations. Similarly, at present, if ending natural resource transactions with a certain dictatorship will have the consequence of producing absolute deprivation for its population, there will clearly be strong reasons in favour of continued transactions. In other words, highlighting self-referring reasons for reform is not the same as claiming that these reasons are always going to be decisive, regardless of consequences. Rather, the claim is that such reasons warrant independent attention: that the moral reasons for global reform are not exhausted by consequentialist considerations.

There are likely to be normative theorists who will find even this moderate approach to be overly self-absorbed and too removed from the practical needs of the world’s poor and oppressed. But this likelihood only makes it all the more necessary to debate the normative questions that self-referring moral arguments trigger. It is important, in particular, to discuss the way in which a society’s global conduct may or may not derive from ‘moral duties to itself’, from its own integrity or ‘clean hands’.

Global justice theorists have yet to undertake such a discussion. Yet a recognition of the deep limitations of global reforms dependent on (structurally limited) social science should provide an important reason to pursue it. It is worthwhile, in turn, to note both some of the practical pay-offs that will arise if the discussion will vindicate self-referring reforms, and the new kinds of normative questions that will have to take centre stage in this discussion.

One important practical pay-off of self-referring global reforms is their dissolving of the aforementioned collective action problems endemic to international politics. Hinging reform of dictators’ trading privileges, for instance, on the ability to destabilise them, also means making such reform dependent on successful collective action by multiple boycotting countries. If the goal is to destabilise the dictatorships basing their rule on natural resource revenue, each state can exempt itself from any duty to take solitary action by claiming ‘ineffective sacrifice’ – pointing out that the moral gains from such action will be miniscule unless others join, but the cost (in terms of lost access to valuable resources) will be significant. Furthermore – the excuse will continue – were that state to act alone, it will only be the ‘sucker’ incurring absolute and relative costs in relation to its competitors.Footnote ⁶³ Yet when reform of dictators’ trading privileges is justified through a prohibition on complicity in theft, such collective action excuses disappear. To take the analogy, one’s duty not to steal from the person next door, for example, holds independently of the possibility that other neighbours will be stealing from that person (even if these neighbors are also competitors), as well as of the possibility that one’s clean hands will not alter the neighborhood’s overall condition.

Another, related payoff of self-referring reforms has to do with the significance of blocking self-seeking rationalisations. As long as powerful actors can attach social-scientific uncertainty to many global reform proposals, we are bound to see a proliferation of self-serving moral justifications for why it is morally permissible (or even morally necessary) to avoid reform. Just as the Regan administration sought to explain how continuing economic ties with South Africa represented ‘constructive engagement’ rather than merely profit-maximising, so can powerful oil corporations seek to explain why oil trade with dictators actually has a chance of improving the condition of their victims, accordingly suggesting that these victims may be disposed to consent to the continuation of such trade (in much the same way, for instance, that many ordinary consumers of cheap clothing rationalise their purchases of goods produced in sweatshops). It should be obvious, however, that a situation in which agents can permissibly benefit from actions that would normally constitute grave wrongs, simply as long as their doing so could be somehow conjectured to fit the victims’ interests, would pose the danger of a slippery slope, since it would create a deeply problematic structure of incentives for agents’ conduct: if agents would be allowed to benefit from the violations of victims’ rights as long as this could somehow be shown to be in line with the victims’ interests, then we are bound to see many agents seeking ‘loopholes in morality’.Footnote ⁶⁴ In contrast to all this, insofar as they are largely independent of social scientific uncertainty, self-referring reforms leave very little room for such loopholes.

These advantages may still not convince those philosophers who have very strong intuitions against self-referring moral concerns, seeing such concerns as befitting only ‘moral dandies’.Footnote ⁶⁵ But, with normative political philosophy as with empirical social science, mere intuition is not enough. An actual in-depth discussion is necessary, and such discussion will have to cover multiple sets of questions.

One set of questions has to do with whether and why there is a difference in the moral force of self-referring moral reasons in different contexts. Thus for instance, if, as I suspect is the case, many philosophers would be willing to concede that liberal democracies had a duty to divest from apartheid for the sake of their own moral integrity, but far fewer philosophers will say the same about a duty to disengage from kleptocrats, what argument can justify this difference?

Another set of questions has to do with the relationship between personal and political morality. To stick with our running examples, philosophers who intuitively resist ‘clean hands’ global reforms will be hard-pressed to deny that when the agent in question is an individual person, the duty not to steal a victim’s property is independent of what other agents do. What then (if anything) is different when the property belongs to a people and is effectively stolen by a corporation colluding with a dictator, in a process legalised by a state?

A related set of questions concerns the interaction of integrity and responsibility. If democracies stop certain global practices out of concern for their own integrity, are they morally responsible for the negative consequences that their decisions may produce for ‘distant’ victims of poverty and oppression? Specifically, if democracies as third parties end ties with a morally abhorrent regime out of concern for their own moral integrity, but, losing their leverages vis-à-vis the regime, end up seeing it mistreating its subjects even further, can democracies be held morally responsible for this worsening of the subjects’ condition?

Finally, and perhaps most importantly, global justice theorists need to direct philosophical energies to the very idea of self-referring moral reasons. Such reasons, whether understood in terms of clean hands, duties to self, or integrity, have enjoyed considerable attention in philosophical discussions of individual ethics, including well-known discussions of the relationship between one’s moral integrity and one’s complicity in the harms that third parties inflict on innocent victims.Footnote ⁶⁶ Yet these moral reasons and relationships have received virtually no attention from philosophers discussing global justice. There currently exists no sustained body of philosophical work discussing in what way a political community as a collective agent can have integrity in a way that parallels the integrity of an individual person, nor (for that matter), is there a discussion of how personal integrity relates to a society’s sense of collective integrity (or lack thereof).Footnote ⁶⁷ Answering all these questions represents one of the most important tasks of global justice theorists. The more global justice theorists are cognisant of the social scientific obstacles facing some of the key reforms they have been discussing, the more reason they will have to turn to this task.

V. Instead of a conclusion: The broader view

In order to be complete, the normative research agenda to which I have just gestured will obviously have to be elaborated in much greater detail. Rather than attempt such elaboration here, however, I want to conclude with some remarks on the broader implications of my claims, going beyond the specific proposals for global reform that I have discussed.

First, some readers might think that the points I have made regarding the limitations of global social science have domestic parallels. Should we therefore adopt similar epistemic humility about the expected consequences of domestic policies? And should such humility, in turn, similarly lead us to give a central place to non-consequentialist factors in domestic policymaking?

I do think that we should practice epistemic humility in the domestic context as well, and be wary, even in this context, of what Jon Elster for instance calls ‘excessive ambitions’ in the social sciences.Footnote ⁶⁸ Such humility seems warranted given the enduring relevance of classic warnings, as to the limited ability of social science to predict future political and economic developments.Footnote ⁶⁹ Epistemic humility seems warranted, furthermore, in light of social-scientific disputes even regarding past policies whose consequences may initially seem straightforward. Moreover, given the degree to which consequentialist political theories hinge upon social-science to yield practical prescriptions, I believe that the deep and enduring limitations of social science provide, already at the domestic level, an important reason not to adopt an all-out-consequentialist position on public policy.Footnote ⁷⁰

Yet, notwithstanding all of these concerns, when it comes to domestic politics, I also want to qualify the demand for epistemic humility and its non-consequentialist implications. I do not want to suggest that this demand should apply just as forcefully as in the global case. Alongside some complex normative reasons,Footnote ⁷¹ there are also simple empirical and strategic reasons. Empirically, there is a better chance of finding reliable natural experiments in the domestic as opposed to the global level.Footnote ⁷² And strategically, I do not want to give the impression that in order to accept my arguments regarding the limitations of global social science, one must also accept a parallel position on domestic issues. Global concerns are important enough to stand on their own, and I therefore explicitly want to keep the argument ecumenical. It is less crucial for me to show that the points I have laid out might also apply to domestic politics. It is more crucial to show that these points are cogent at the global level, regardless of what one thinks about the domestic level.

These remarks, however, still leave aside an important issue, with which I will close. Even if one accepts the independent significance of global issues, it may appear that the self-referring reasons for global reform sketched above render this focus too narrow. The reason is that, even if acting on such reasons will be important as a way of retaining democracies’ own ‘integrity’ or clean hands, this will not by itself provide any clear future path to improve the state of the world as a whole. So rather than solving problems of global justice, some may accordingly fear that my (tentative) proposals only seem to back away from these problems.

I think that this fear is unwarranted, for at least two reasons. First, taking seriously democracies’ self-regarding moral reasons is not equivalent to giving up on efforts to realise a more just world. Rather, such a strategy can be understood as highlighting moral constraints that ought to accompany such efforts. Even someone who thinks, with Rawls, that we must always take as our practical compass a vision of perfect justice to which we try to transition, will presumably also accept that we ought to follow ‘only courses of action that are morally permissible’Footnote ⁷³ on the way there. The claims of integrity to which I have gestured can be seen as one way of following this kind of warning – of identifying which courses of action are not permissible as part of our efforts to achieve a more just world.

Second, more fundamentally, I am sceptical about the prevalent thought that we must always have a vision of perfect justice – in this case, perfect global justice – as a practical guide. The claim that ‘ideal’ visions of perfect justice are a necessary practical compass remains extremely influential in political philosophy in general, and in Rawlsian political philosophy in particular.Footnote ⁷⁴ It leads many philosophers to hold that in circumstances where social science is unable to tell us how to best transition to a perfectly just world, we are condemned to mere ‘muddling through’, and even risk ‘an abandonment of the goal of any systematic theoretical guidance of political practice’.Footnote ⁷⁵ In my view, however, this prevalent position is problematic.

For one thing, this position underestimates just how frequently the limitations of social science – in the global context at least – force us to ‘muddle through’.Footnote ⁷⁶ Furthermore, this position underestimates our ability to develop, without recourse to visions of perfect justice, systematic normative accounts of complex political situations. Finally, at the most foundational level, the prevalent insistence on treating visions of perfect justice as a compass underestimates the degree to which these visions themselves are not fixed, but may shift over time. This, at least, seems to follow if we believe (with Rawls) that even ideal visions of perfect justice have to account for ‘limits of practical possibility’ to have practical value – to serve as ‘realistic utopias’Footnote ⁷⁷ rather than mere utopias. If we hold this belief, then we need to recognise that concrete political choices – including our self-referring moral choices under the haze of social-scientific uncertainty – might shift the limits of practical possibility which condition, from the outset, the content of our visions of a perfect world.Footnote ⁷⁸ If this argument is cogent, then it requires a much more flexible approach to global justice and injustice than the one that many philosophers take. But the task of establishing this argument, I shall leave for another day.