We live in a profoundly unequal world. Let's start with wealth. A recent study conducted by Oxfam argues that, by 2016, the combined wealth of the richest 1 per cent of the world's population will be greater than the combined wealth of the remaining 99 per cent. Over the course of the last several years this inequality has been growing. The richest 1 per cent is about to exceed 50 per cent ownership of all global wealth, a share that is up from 2009, when they owned 44 per cent (Oxfam, 2015). Thomas Piketty notes that the wealthiest 0.1 per cent of the global population owns about 20 per cent of all wealth (2014: 438), and he conjectures that the wealth of the wealthiest people in the world has grown 6–7 per cent per year over the past three decades (2014: 435). Global health inequalities are stark. According to the World Health Organization, the average life expectancy in low-income countries is just 57 years, while in high-income countries it is 80. Children in the poorest quintile globally are nearly twice as likely to die before their fifth birthday as children from the richest quintile (WHO, 2011). Educational inequalities are also terrible. In the wealthiest countries almost no one aged 20–24 has less than four years of schooling. In many of the poorest countries of the world the majority of the population that age has had less than four years of schooling. In Niger, the worst case, a full 78 per cent of all people aged 20–24 have had less than four years of schooling (WIDE). Cosmopolitans typically argue that global inequalities such as those in wealth, health and education are unjust. In this chapter, I defend the cosmopolitan view. I discuss some of the arguments made by cosmopolitans in support of their view as well as some of the disagreements that exist among cosmopolitans about when duties of justice exist and about the content of justified principles of distributive justice. I also discuss some cosmopolitan reforms to the existing global order; reforms that would serve to reduce global inequality.
I shall call ‘a relation of justice’ the relationship between two or more persons in which at least one person has a claim of social justice and at least one other has duty in virtue of that claim. Cosmopolitans hold that relations of justice may extend across state borders; relations of justice are not necessarily limited to compatriots; and all those persons who stand in relations of justice are moral equals. Cosmopolitans can allow that compatriots stand in special relations of justice to one another; but cosmopolitans deny that the limit of relations of justice is the compatriot relationship. The form of cosmopolitanism that I shall focus on concerns distributive justice. It is characteristic of these cosmopolitan views that the requirements of the duties of distributive justice are generally no less substantial than the requirements of distributive justice among compatriots, even if they might be different or have a different justification.
I shall assume that, as a first approximation, what matters for distributive justice is the distribution of wealth, income and opportunities. This corresponds to the two parts of John Rawls's second principle of justice (1999: 266). I shall not deny that there might be something more fundamental, say, human capabilities that ultimately matter most of all. But I shall assume that the distribution of wealth and income can serve as a rough approximation for whatever matters most to distributive justice, if it is not ultimately wealth and income. Amartya Sen (1979) criticizes John Rawls's focus on wealth and income by pointing out that two people might be equal in terms of wealth but have very unequal capability sets because the one is able-bodied and the other disabled. One response that Rawls offers to Sen – a response I endorse – is not to deny the force of Sen's point that equality of resources does not ensure that relations are equal in the morally relevant way, but instead to claim that the focus on resources is appropriate at least at a certain level of abstract when the discussion is focused on the design and structure of the basic distributive institutions, and not their day-to-day functioning, which is fine-tuned by legislation and policy (Rawls, 1999: 47–52; 2001: 168–76).
The cosmopolitanism that is under discussion in this chapter, then, holds that persons across state borders stand in relationships of justice with respect to the distribution of wealth, income and opportunities. One remaining characteristic of this view is important. Among the duties that are governed by distributive principles, should there be a case of conflict between what is owed to compatriots and what is owed to non-compatriots, it is not necessarily the case that priority is given to compatriots, on grounds that they are compatriots. A view that held that there were duties of distributive justice that extend across state borders, but which privileged compatriots, would not be in the class of views that I am a discussing as cosmopolitan (Miller, 1998). Cosmopolitanism may allow that there are special duties of justice to compatriots, but it does not allow that among the duties that apply to compatriots as well as non-compatriots, relations between the former necessarily have some feature that requires privileging those who stand in that relation over those who do not stand in that relation to us.
But what does cosmopolitan distributive justice concern? As a matter of distributive justice, the evaluative terms ‘just’ and ‘unjust’ are sometimes thought to apply both to the actions of individuals and to institutions. Such accounts might be called ‘interactional’ even though they include interactions and institutional effects. An example of an interactional account is G. A. Cohen's argument (2008) that it would be unjust of people to demand higher remuneration as a condition of doing work that would benefit the least advantaged as the application of the difference principle only to institutions allows. I employ a more limited use of the terms ‘just’ and ‘unjust’ such that they apply primarily to the effects of institutions and derivatively to the policies that operate within the scope of institutions. Accounts that take justice to apply primarily to institutions are often referred to as ‘institutional’. These views hold that what matters to justice is how people fare under institutions, not what their pre-institutional circumstance are, and not how individuals behave within the permissible scope of institutions. There are several responses to the interactional account. One relies on the importance of fair public rules for justice in conditions in which persons endorse a plurality of reasonable moral outlooks (Williams, 1998). And Kok-Chor Tan (2012) lays weight on the value of moral pluralism. The idea is that if institutions are arranged justly, as long as people constrain their actions to the rules of the institutions, they are morally free to pursue a variety of morally licit goals, which might range from benefiting their family and friends to developing their talents and pursuing their interests.
There are two different broad approaches for justifying duties of cosmopolitan distributive justice. I distinguish these as ‘non-membership-dependent’ and ‘membership-dependent’. Simon Caney makes an extended non-membership-dependent argument in defence of duties of cosmopolitan distributive justice. Often, arguments of this kind claim that the conditions that justify egalitarianism domestically are applicable globally as well. Here is an outline of Caney's argument:
We can call this ‘the similarly situated argument’. The effect of the similarly situated argument is to recruit the domestic egalitarian to the cause of cosmopolitan egalitarianism. The argument obviously does not appeal to those who believe that the case for egalitarianism also fails at the domestic level. But if it succeeds there, then, according to the similarly situated argument, it should apply globally as well.
The similarly situated argument involves an ambiguity about the nature of the justifying conditions. Premise two invokes moral personality. We affirm domestic egalitarianism (if we do) at least in part because we affirm the equal moral personality of all citizens. But that claim and premises one and three entail the conclusion only if equal moral personality is a sufficient condition for distributive egalitarianism. Premise two speaks of egalitarianism ‘invoking’ universal moral personality, not entailing it. Moreover, it is precisely the entailment claim that will be resisted by those domestic egalitarians who also reject global egalitarianism. They will argue that egalitarianism is justified by such equality only under special relations, such as coercion (Blake, 2001, 2011; Nagel, 2005) or reciprocity (Sangiovanni, 2007), which conditions do not exist between non compatriots.
Membership-dependent accounts of cosmopolitanism invoke features of our partially globalized world to draw the connection between fundamental human equality and distributive egalitarianism. This is sometimes done in explicitly Rawlsian terms by claiming that global interdependence produces conditions relevantly similar to the basic structure of domestic justice. For example, Charles Beitz claims: ‘Assuming that Rawls's arguments for the two principles are successful, there is no reason to think the content of the principles would change as a result of enlarging the scope of the original position so that the principles would apply to the world as a whole’ (1999: 151). As we shall discuss below, we can distinguish the question of whether the existence of interdependence gives rise to duties of distributive justice from the question of whether the appropriate principle of justice is the difference principle.
I also defend a membership-dependent account of cosmopolitan egalitarianism (Moellendorf, 2009). The outline of the argument is as follows:
Each of the four premises requires significant clarification and defence. I turn now to discussing them.
Premise one invokes the idea of justificatory respect. Let's assume that persons possess equal inherent dignity as expressed in the preambles of the major human rights documents of the twentieth and twenty-first centuries. The relationship between the rights enumerated in these documents and the foundational commitment to inherent human dignity is that respecting the latter consists in recognizing the former, which protect and enable dignity. An interpretation of human dignity is required to make this claim plausible. I suggest the following one. Human dignity consists in, among other things, our capacity to act in ways that are responsive to the reasons we have for acting; human rights, by protecting and enabling certain forms of acting (including deliberating and thinking), serve the aforementioned capacity. The freedoms of conscience and association and entitlements to education and medical care all serve the human capacity for practical reason.
The human capacity to act in ways that are responsive to reasons is also respected when the institutions, which constitute the association, are structured by principles that could be reasonably endorsed by persons who participate in the associations constituted by the association. To impose on persons institutions (which are sufficiently robust) that could not be justified to those persons whose conduct is regulated by them would be to fail to respect the dignity of these persons (see also Forst, 2007). It would be to fail to respect their capacity to act in ways responsive to reasons because it would allow their agency to be manipulated by policy in ways that they might reasonably reject.
Premise two delimits the kind of associations that are constrained by justificatory respect. Not all associations are concerns of justice, which require that their constituting institutions be justifiable to the members. Voluntary associations, such as many churches in liberal societies, or clubs of various sorts, would usually be ruled out. When one freely joins such a club, one usually cannot complain of an injustice in its governing structure if one is not treated as an equal member. Four conditions are individually necessary and jointly sufficient for the associations to generate duties of social justice. First, the association should be sufficiently strong, where strength is measured according to the following three indices: how enduring the association is, how comprehensively its institutions are governed by norms and how regularly it affects the highest-order moral interests of its members. This would not include many clubs organized around common interests. Second, the association should be non-voluntary in the sense that there is no reasonable alternative to participation in the association. Many religious organizations and clubs would not satisfy this condition. Third, the association should be constitutive of the background rules that regulate members’ public lives. Rules that regulate what one can do as member of a club would not be included. And fourth, the rules must be able to be brought under the control of the members. If there were nothing that could be done to change the rules, then they would not be appropriate concerns of justice.
The second premise of this argument states an important part of what I call ‘the principle of associational justice’. That principle holds that duties of justice exist between persons who have a moral duty of equal respect to one another if those persons are co-participants in an association of the requisite kind, one that is relatively strong, largely non-voluntary, constitutive of a significant part of the background rules for the various relationships of their public lives, and governed by institutional norms that may be subject to human control.
The principle of associational justice makes a controversial claim, namely that people can come to bear special duties through non-voluntary relations. Samuel Scheffler summarizes two kinds of objections to that claim. One is ‘the voluntarist objection’ (2001: 54). According to this objection, if there are associative duties in virtue of membership it is only insofar as membership can be adequately characterized as an instance of a contractual relation broadly understood. A worry that supports this objection is that the existence of non-voluntary associative duties would seem to constrain what some people, but not everyone, are at liberty to do. Some people are picked out as bearing duties that others do not bear, and they are picked out by considerations other than those that they voluntarily chose to assume. When a person makes a promise, it is widely agreed that she has a special duty to the promisee, a duty others do not have, to fulfil the promise. That duty arises out of the voluntary action of the promiser. Now, the principle of associative justice claims that special duties owed by members to other members of certain kinds of associations can arise non-voluntarily. In response to the voluntarist objection: first, it seems to be a moral feature of the world that non-voluntary special duties can morally constrain people. That idea conforms with common-sense morality in a number of ways, for example that family members have special duties to one another. Second, to focus only on the moral constraints imposed by these non-voluntary relations is to miss an important part of the picture. Each duty bearer is also normally an entitlement bearer as well. She has good moral reason to expect that, as a member of the relation, she will be able to claim that other members fulfil their duties to her. And as the example of the family suggests, the moral constraints that exist as a result of non-voluntary special duties are not necessarily regrettable, since they may constitute morally valuable relations.
The second objection noted by Scheffler is ‘the distributive objection’ (2001: 56). The worry here doesn't have to do with the burdens on the duty bearer, but with the benefits to the person owed the duty. Associative duties pick out some people as beneficiaries, but not others. Against a background commitment to egalitarianism, this feature of associative duties might seem worrying. The person who is a beneficiary does not seem to deserve special benefits, nor does she necessarily have greater needs. How then can the benefits be justified? The question can also be asked in the case of the person who benefits as a result of the fulfilment of special duties in a voluntary relationship. In that case, it might suffice to answer that the beneficiary voluntarily contracted with another person, who, as a result of the contract, owes the beneficiary a good or a service not owed to others. Regardless of whether that answer is sufficient, it's not available in the cases covered by the principle of associational justice, which applies in particular to non-voluntary relationships. Still, the force of the distributive objection depends upon the content of the background commitment to egalitarianism. The principle of associational justice assumes a background duty of equal respect. Now, as long as that commitment alone does not entail strict distributive egalitarianism, and according to the argument under discussion the commitment does not, then the principle of associative justice is compatible with the background commitment to equal respect. Moreover, as noted in the previous paragraph, part of what a person loses in virtue of being bound by duties of justice, she regains as being a beneficiary as well.
The third premise states that justificatory respect establishes an egalitarian presumption regarding the distribution of the goods and powers produced by members of an association if the association generates duties of justice and if the association is what I call ‘a common good association’. A common good association is one in which there are no pre-associational valid claims on the goods and powers produced by the joint efforts of the members. The idea is that the only kind of distribution that could be justified to members of a common good association is one that is presumptively egalitarian. If a justification must be made to everyone, then there would be a strong, but defeasible, presumption in favour of equality. The presumption could perhaps be defeated, if there were morally relevant differences in needs or capacities, if persons acted in ways – say criminally – to forgo equality, or if, following Rawls's difference principle, it were to the advantage of the least well-off. Recall the worry of the distributive objection: if equal respect for persons alone entailed duties of egalitarian distributive justice, then the principle of associative justice would seem to contradict the requirements of equal respect. I am now denying the antecedent of that conditional. The connection between equal respect and egalitarian distributive justice is characterized here as requiring both associations that satisfy the principle of associative justice and that are common good ones. This makes explicit the way that the distributive objection is deflected.
The line of reasoning that holds that equal respect entails egalitarian distributive duties only under conditions that satisfy the principle of associational justice and the requirements of a common good association provides a plausible response to Pablo Gilabert's claim that the universal scope of the moral equality of persons affirmed by cosmopolitans requires that the scope of requirements of justice also be universal (2011: 589). As my discussion of premise four will suggest, I agree with the likes of Gilabert and Caney, who defend cosmopolitan principles of distributive justice. But, for the reasons just canvassed, I think that they are wrong in claiming that cosmopolitan principles of distributive justice follow from a commitment to the moral equality of persons without consideration of the nature of the association in which people are related. We do not violate the moral equality of persons when we maintain that we have special duties only to some persons, and that we may claim special entitlements from some persons and not others. As it happens, given the contingencies of globalization we have special duties of distributive justice that have global scope.
The fourth premise simply states that the four conditions laid out in premises two and three apply to the global economic association. I now adumbrate some reasons for believing that this is plausible for each of the four conditions of premise two. With respect to the first condition, the global economic association seems strong in light of the three indices of strength. The association created by the processes of economic globalization is enduring. It is a structural feature of capitalist economic development that has gathered pace recently with technological changes, but that has been observable since at least the early colonial area. Additionally, it is governed primarily by the norm of competition for market share that requires firms to innovate ceaselessly and to reduce production costs. But norms of governance are also in place, especially through the regulatory framework established by the WTO (World Trade Organizaton) and the municipal exclusionary property regimes that are implicitly recognized in all international trade. The globalization of trade, investment and finance has had profound effects on the highest-order interests of persons, including spreading market norms that dictate whether and how a state may succeed in poverty-eradicating human development and imposing rules of governance through the WTO that affect, for example, a state's ability to succeed in development by means of trade. Finally, the effects of the international economic crisis perpetuated by the home-lending crisis in the USA in 2007–8 illustrates the manner in which the highest-order interest of persons are affected by the international economic association. For example, according to the World Bank, net capital flows to developing countries fell 20 per cent in 2009 to $598 billion and were a little over half the 2007 peak of $1.11 trillion (World Bank, 2011: 1).
Consider the second of the four conditions of the principle of associational justice, namely non-voluntariness. There is no reasonable prospect for extensive economic development and growth other than adopting a capitalist economy (Cohen, 2001: 207–15). And the gains from international trade can be immense. Moreover, although states are formally free either to join the WTO or not, the overwhelming majority of states see no reasonable prospect to development outside the WTO. Regarding the third condition, there is no longer any doubt that the pressures of global competition constrain the regulatory regimes of states, and that the constraints affect a range of issues, from employment opportunities and job security to the intellectual property law of a state. These are associated with highest-order human interests in well-being, security and health. The fourth condition requires that norms be subject to control. Market competition can, of course, be limited, directed or counterbalanced by deliberate public policy. And obviously, WTO rules can be amended.
There seems to be plenty of empirical evidence to support the claim that the global economic association satisfies this condition. (See also Tan, 2012: 155 for an abbreviated form of this kind of argument.) This argument does not deny the existence of special duties of justice between compatriots in virtue of their co-membership in either a national economy or a state. It is not an argument that duties of egalitarian justice exist exclusively at the global level (see also Moellendorf, 2011a).
The argument above differs from a prominent kind of anti-cosmopolitan argument that seeks to limit relations of egalitarian justice (Blake, 2001, 2011), or perhaps of social justice generally (Nagel, 2005), to the state on grounds that a necessary condition of such relations is that the persons in them share membership in a coercive state structure. These might be called ‘coercion accounts’. Michael Blake and I have had an extended debate on these points, but I cannot fully summarize the arguments here (see Blake, 2011; see also Moellendorf, 2009, 2013). Suffice it to say that Blake's considered opinion is that the best explanation for why distributive egalitarianism is required within states is that states are vertically coercive, in the sense that the coercive agency is established by citizens; that kind of coercion cannot be removed and, instead, it requires justification; and the existence of distributive egalitarianism renders the vertical coercion justified. Internationally, states coerce one another horizontally, that is with respect to the norms of the international system. In principle, that kind of coercion can be eliminated, so rather than justifying it we should seek means to eliminate it (Blake, 2011). In contrast, I argue that the power of egalitarianism to justify coercion, if one supposes that it does do that (which I doubt), does not entail that coercion is necessary to make the case for egalitarianism (Moellendorf, 2009, 2013). I argue that international investment, production, trade and finance bring people into a common good association of the kind characterized by the principle of associational justice; and that suffices for duties of egalitarian distributive justice, assuming justificatory respect (Moellendorf, 2009). A different kind of response to Blake would be to argue that the kind of coercion relevant to relations of justice exists in international institutions as well as in the state (Abizadeh, 2007; Hassoun, 2012). I do not dispute that claim, but nor do I rely on it, because I am not convinced that the existence of vertical coercion is necessary to justify the duties of egalitarian distributive justice.
As discussed above, Caney favours a non-membership-dependent account of cosmopolitan distributive justice. He rejects accounts of cosmopolitan egalitarian distributive justice, like the account I have defended above, on grounds that they face an unsatisfactory dilemma (2011: 523–5). Either they draw a sharp line between the associations in which duties of distributive justice exist and associations in which they do not, or they allow duties to vary with features of the association. The first horn is implausible, he claims, since failing to satisfy the appropriate requirements in only one way would make the association one in which there are no distributive duties. But, according to Caney, the second horn permits gradual variance of duties of justice as associations become more robust, rendering the account necessarily imprecise and incomplete. Hence, any such view is implausible.
How do I reply? Premise two of my argument above takes the existence of duties of justice to depend on the following four conditions: (i) the strength of the association, (ii) its lack of voluntariness, (iii) the association constituting a significant part of the background rules that govern members' public lives, and (iv) that it can be brought under the collective control of members. These requirements are all threshold empirical conditions. The view might be thought of, then, as falling on the first horn of Caney's dilemma. But it could avoid doing so, if beyond some low threshold, satisfaction of the empirical requirements admits of degrees and the moral principles are scaled to the degree of satisfaction. I claim that strength depends on how enduring the association is, how comprehensively its institutions are governed by norms, and how regularly it affects the highest-order moral interests of its members. These considerations admit of degrees, even after the point at which the association is sufficiently strong. There could be more and less strong associations beyond the threshold. The same claims apply also to the constitution of a significant part of the background rules of public life and being subject to collective control. Both may be fulfilled in degrees beyond the satisfaction of a low threshold. My argument seems rather to fall upon the second horn of Caney's dilemma. But how much damage does that do? A. J. Julius offers a reply that seems to deflect the force of the charge: any account of distributive justice that does not register the importance of both national economies and partial global integration seems insensitive to morally important facts of the current global condition. ‘A continuous transition between the standards succeeds in acknowledging both circumstances even as it ensures that the international difference principle eventually displaces the national one’ (Julius, 2006: 191). Caney may be right that all such accounts currently lack precision and completeness, but what is lost in that regard seems to be made up for in registering the morally relevant features of our partially globalized world.
Cosmopolitans often defend the global application of the difference principle (Pogge, 1989; Beitz, 1999; Moellendorf, 2002). Brock contends that this is mistaken. Cosmopolitans should rather endorse a basic needs principle, which would ensure that everyone is ‘adequately positioned to enjoy the prospects for a decent life’ (Brock, 2009: 52). Brock claims that experimental evidence refutes Rawls's original position argument in support of the difference principle. People in a number of different countries were put in simulations modelling the original position, and they consistently ranked the difference principle lower than a basic needs principle. The trouble with Brock's rejection of Rawls's justification, however, is that it does not address Rawls's arguments for why the difference principle would be chosen (Moellendorf, 2011b: 263). Nor is there any assurance that the experimental subjects understood and sought to evaluate Rawls's arguments. Hence, their choice in experimental conditions should not be considered a decisive refutation of the Rawlsian arguments for the difference principle.
Nonetheless, there may be other reasons to doubt the appropriateness of the difference principle for global distributive justice in current circumstances. The requirement of the principle of associational justice is that the rules of the association must be subject to collective control. There is a certain lack of capacity for doing that in conditions of partial globalization. Global and international political capacity to regulate the terms of international trade, investment and finance is significantly weaker than domestic power to do so within home markets. More than that, the political organizations that might legislate that capacity do not even exist. Hence there is reason to doubt the power to arrange municipal property law, international trade treaties, and the institutions of international investment and finance so as to ensure that they approximate a distribution that is, over time, to the maximum benefit of the least advantaged.
The trouble with the difference principle is that it does not offer a plausible norm against which to judge current distributions, because, in a partially globalized world, we lack the capacity to arrange institutions in a way that would seem to be even close to satisfying it (Moellendorf, 2009). Still, as Julius seems to suggest in the quotation cited above, assuming the process of continual global integration, the eventual satisfaction of the difference principle could be a moral goal worthy of our allegiance insofar as it will one day be possible to arrange institutions to approximate it, and justificatory respect would demand that. The greatest collective control currently exists in matters where state capacity can be used and where international institutions are strongest. In these areas, the case for duties that seek to reduce inequalities seems the strongest, even if satisfying the difference principle may not be required.
An alternative principle of distributive justice is fair equality of opportunity. Sometimes, cosmopolitans endorse a globalized version of that principle as well (Caney, 2001; Moellendorf, 2009). Equality of opportunity requires equalizing opportunities to possess goods of some specified kind among persons with approximately equal endowments of some specified sort. Typically, the ideal requires the mitigation of inequalities of wealth and educational access inherited due to social circumstance at birth in order to ensure approximate equality of opportunities for success among the equally talented. Usually, egalitarians argue that the relevant social circumstance to mitigate is the social class of the child's family. But in the global context, the circumstances might include country of origin as well.
Gillian Brock argues that attempts to extend the principle of fair equality of opportunity globally face a dilemma (2009: 61–2). Either they must employ a version of equality of opportunity that relies on favoured social positions from a particular culture without appropriate global scope, or they must allow cultural variation about what counts as a favoured position, which produces an account of equality of opportunity that is too weak to rule out significant disadvantages and discrimination. I believe that there is a viable third horn, namely to identify positions, powers and goods, which are drawn from an interpretation of the association in which the question of justice arises, and which therefore have a broader relevance than merely to participants of a particular culture. This requires an interpretive argument about the position, powers and goods that the association distributes (Moellendorf, 2009: 73–7). If there are such association-specific positions, powers and goods, they constitute the material for egalitarian concern that is established by justificatory respect applied to the common good global economic association.
As an alternative to equality of opportunity Brock supports a principle requiring that there be a decent set of opportunities for all. She seeks to support the principle by contrasting it with equality: ‘If faced with the option of equal but poor life options, or a situation in which, though there is some inequality of access, everyone has access to developing a range of skills sufficient for earning a living (or meeting needs) with dignity and delight, we should choose the second situation’ (2009: 62). This, however, does not seem relevant. The issue between egalitarians and Brock does not concern the choice between whether boys and girls should have equally poor access to education, or better but unequal access. Rather, it concerns the choice between the latter – better educational opportunities for boys but still decent ones for girls – and the option of equal opportunities for boys and girls to a good education. In rejecting equality of opportunity, Brock would seem to be indifferent between those two options. In contrast, the proponent of equality would defend the option of equalizing opportunities to a good education for boys and girls.
Several feasible reforms to both states and international organizations would promote greater global equality and therefore would appear to be required by cosmopolitan distributive justice, unless there prove to be compelling reasons of justice that militate against them. I discuss three such reforms below. One attractive feature of cosmopolitan distributive justice, in contrast to state-centric or nationalist accounts of distributive justice, is that the cosmopolitan accounts have the theoretical resources to pick out these three areas as sites of existing injustice; that, I believe, conforms to many people's considered judgments about these matters. Ultimately, however, remedying global inequality would require a new global institution that has the effect of redistributing wealth from the richest to the poorest. So, I finish by discussing proposals for a global tax.
Country of birth confers significant privileges and is therefore an important factor in global inequality. Branko Milanovic's studies of global income distribution show that at least half of the variability in a person's global income percentile rank can be explained by her country of birth's level of human development and the inequality of income within the birth country. If that is correct, there are three ways a person's global income percentile rank might be improved: her efforts might receive greater remuneration; her country might improve its development level; or she might immigrate to a more highly developed country (Milanovic, 2015). Liberalization of immigration restrictions in wealthy countries is fully within the scope of state power. Reducing barriers to immigration could allow people from less developed countries to become better-off. According to a World Bank report, an increase in immigration from developing to high-income countries would increase the labour force of the latter by 3 per cent from 2002 to 2025, and it would generate large aggregate increases in global welfare. When adjusted for prices, the increase in aggregate global income would be $356 billion or 0.6 per cent. Moreover, the disaggregated gains of this policy would go disproportionately to developing countries, where incomes would increase by 1.8 per cent, compared to 0.4 per cent in high-income countries. But for the new immigrants to high-income countries, the cost-of-living-adjusted income gains would increase on average by nearly 200 per cent, due to differences in wages between the countries of origin and destination. And there would be income gains in developing countries due to decreased labour market competition and increased remittances (World Bank, 2006). Immigration liberalization is within the capacity of states to affect; it would promote the equalization of opportunities globally; it is, therefore, required by a cosmopolitan approach to distributive justice.
Very often, a country's access to finance from the International Monetary Fund (IMF) during times of economic crisis comes with conditions requiring the country to liberalize its labour markets. The required labour market liberalization takes many forms, including increasing retirement ages, weakening collective bargaining, lowering minimum wages, reducing unemployment benefits and loosening firing procedures. About one quarter of all IMF loans since 1987 have required changes that deregulate the labour market of the borrowing country (Caraway et al., 2012). The IMF justifies labour liberalization as promoting growth and employment (Furceri et al., 2012). That justification, however, is made primarily on the basis of computer modelling exercises, which lead IMF researchers to make rather hedged prognoses, especially in light of the experienced negative effects of such policies, including reduced growth, recession and increased unemployment: ‘To the extent reforms enhance credibility and confidence, some of the short-run negative effects may be countered. In any case, the full impact on growth and employment are likely to materialize only over a medium- to long-term horizon’ (Barkbu et al., 2012: 13). Moreover, whatever the merits of that justification on its own terms, it simply neglects the more important issue of the distributive effects of the liberalization policies, namely that they tend to increase inequality.
One comprehensive study employing regression analysis over a 32-year period of a sample of 110 countries shows that ‘even controlling for the fact that countries participate in IMF programs under bad economic conditions, the inherent effects of programs are negative on income distribution’ (Vreeland, 2002: 122). Indeed, the programs seem to contain a class bias. Although labour suffers a decrease in income, capital experiences an increase, which offsets its losses due to diminished growth (2002: 133). A reform of IMF policy to prohibit making access to finance during economic crises conditional on labour market liberalization is easily made. And it would stop the inegalitarian effects of such conditionality. It is, therefore, required by cosmopolitan distributive justice.
Although a central aim of the WTO is to liberalize multilateral trade, and although it claims to prize agreements that give adjustment time to less developed countries, the WTO has no principled commitment to sequencing trade liberalization by requiring developed countries to reduce their trade barriers first. There are, however, good reasons to believe that developing economies must engage in practices that support and protect their industries, especially industries producing for export, if these countries are to achieve higher levels of human development. The classic nineteenth-century argument to this effect is made by Friedrich List (1966) on the basis of an historical survey, revealing that the development policies of Europe and North America relied on measures to protect infant industry in almost every successful case of development. Ha-Joon Chang (2002) provides a twenty-first-century supplement to List's argument with recent studies. Chang argues that Japan and the newly industrialized countries of East Asia all followed the same kind of infant industry support and protection strategy as North America and Europe. Now, trade policies that promote human development are good because the constitutive elements of human development – education, health and income – are valuable. But more can be said about why a trade regime that does not stymie human development is a requirement of cosmopolitan justice. Recall the research of Milanovic (2015) about the importance of the level of human development of a person's country of birth to her income. A trade regime that promotes human development by allowing developing states to protect and support their infant industries is also a concern of justice because it promotes more equal opportunities for income among persons regardless of their country of origin.
I began this chapter by citing research that indicates a vast and growing inequality of wealth globally. Although the three reforms discussed above are feasible and would serve to reduce some of the existing global inequalities, primarily by promoting the well-being of some of those who are disadvantaged, they do not seek to reduce inequality by constraining income and wealth. In this final section I consider three proposals that would do that. The financial transaction tax, or Tobin Tax, would tax income made on short-term currency speculation. Milanovic (2005) recommends a global income tax; Piketty (2014) proposes a global, progressive tax on capital.
The Tobin Tax, initially proposed by James Tobin as a disincentive to international currency speculation (1982: 43), involves taxing short-term speculative investments in international money markets. Currency speculation can produce destabilizing fluctuations in the value of currencies in developing countries. Such fluctuations can affect the ability of a country to export its goods, to pay for imports and to pay its debts. Unforeseen fluctuations compromise the capacity to develop rational development plans utilizing export and debt. An appropriately designed tax could serve the development aims of many developing countries to seek to mitigate such fluctuations. And there is no great social good served by short-term currency speculation. So, there is no reason not to deter it.
Tobin and others also suggest that the proceeds that the tax generated could be used for global poverty reduction (Eichengreen et al., 1995). If the tax could be fine-tuned to be applicable only to short-term speculative investments in currency, it would not deter long-term capital investment that serve development ends. There is reason to believe that a very low tax could be effective at generating considerable revenue. Since the benefits of global poverty reduction would be derived from an activity that produces negative economic externalities, there seem to be good moral reasons to support it. The feasibility of the tax is not entirely clear. A Swedish tax on the sale and purchase of equity securities and fixed income securities was dropped a few years after implementation because revenues were low and the volume of trading declined significantly. But the causes of these events are not entirely clear (Wroebel, 1996). A forthcoming European-wide experiment will be interesting: 11 EU states have agreed to implement it in early 2016. The Tobin Tax might serve important aims of cosmopolitan justice, by facilitating development and combatting poverty. If the European experiment demonstrates that it can work successfully, then it should be supported on grounds of cosmopolitan justice.
Milanovic proposes a progressive global income tax scheme that could be expected to be effective in reducing global inequality because of both the burden it places on high earners and the distribution it directs to low earners (Milanovic, 2005: 159–62). Milanovic's income tax proposal involves the coordination of state institutions for taxation. Participating states with high per capita GDPs add a small levy to their already existing income tax schemes. The funds are collected by participating rich states and transferred to an international organization, which oversees their distribution to participating per capita poor states. Transfers of this sort would lay a responsibility on persons who have been successful in using their natural or social good fortune to their market advantage.
There is, however, a potential drawback to the arrangement outlined above. In states that have lower than average per capita GDPs, but high inequality, there may be some persons who are wealthier than persons in states with higher than average per capita GDPs. So, a scheme that simply sent tax revenue from the higher than average per capita countries to the lower than average could carry out regressive transfers from a poorer person in a rich state to a richer person in a poor state. In order to reduce the possibility of this, Milanovic advocates transfers from per capita rich countries only to per capita poor ones that are also relatively egalitarian. This would render it less likely that there would be rich people who would benefit from the transfers of the poor in a more wealthy country. Milanovic's proposal avoids the injustice of regressive taxation, then, at the price of failing to be entirely global in scope. The poor who have the misfortune of living in very inegalitarian states, such as Brazil and South Africa, have no entitlement under the scheme against the rich of the per capita rich states. This limitation, however, may not be all bad. It could have positive effects on political developments within states. It provides an incentive for political elites to address inequality, and it gives an additional argument to domestic social movements on behalf of egalitarian causes.
A proposal that involves taxing politically powerful high earners probably suffers more severely from problems of political feasibility than the Tobin Tax, especially since the latter is also likely to affect a smaller class of people. The feasibility problems of the global income tax are somewhat mitigated by the fact that the tax could be partially implemented before all of the per capita rich countries participated in it. The successful example of partial implementation might build political support for the proposal. The proposal is certainly worthy of support on cosmopolitan grounds, although it is unclear whether the potential political problems are surmountable.
The central attraction of Piketty's proposal for a global tax on capital is that the tax could get to the main source of inequality far more effectively than an income tax. Because the very wealthy only take a very small proportion of their wealth as income, typically much smaller than the percentage by which their wealth grows, inequality of wealth can grow even with a progressive income tax (2014: 525). What is more, given the high levels of private wealth, even a tax at a modest rate would bring in significant revenue for social welfare policies that support people living in poverty and with low incomes. In Europe, a tax of 2.5 per cent would generate revenues equivalent to 2 per cent of Europe's GDP (2014: 528).
Piketty readily admits that the idea of a global, progressive wealth tax is utopian, since it would require all countries to cooperate by agreeing to, and enforcing, a tax schedule applicable to wealth everywhere. But it is nonetheless useful to consider for a number of reasons. Because presumptively it would be very effective at reducing global inequality in wealth, it can serve as a standard against which to measure the effectiveness of other proposals (2014: 515). Moreover, even a modest tax, Piketty claims, would have important virtues. It would generate reliable information about the global distribution of wealth by requiring the reporting of wealth (2014: 518–19) and by requiring states to broaden their agreements about sharing banking data (2014: 520). The latter would remove the possibility of hiding money away in remote countries. Finally, the aim of moving to a global, progressive tax on capital could possibly be served by first developing regional capital taxes, say in the European Union (2014: 527–30).
Since the political feasibility of a global, or even regional, tax on capital is uncertain, it's not plausible to claim that cosmopolitan justice requires this institution. But it certainly would be desirable. So, there are good reasons of cosmopolitan justice to advocate for such a tax. And should successful campaigning for a global or regional tax on capital sufficiently improve its feasibility, then justice would require that it be instituted.
Our world is remarkably unequal. Cosmopolitans typically reject this inequality. They seek a more egalitarian world. According to the account discussed most fully in this chapter (Moellendorf 2009), the requirement of equality derives from respect for the equal inherent dignity of all persons. Respect for this dignity requires that the social order be justifiable to those who are its members. Justice is about that kind of respect. The institutional order that permits the massive inequalities of our world could not possibly be justified to the great many people who fare so poorly while a very small percentage does so fabulously well. According to cosmopolitans, our world is not only remarkably unequal, it is deeply unjust.1