Corrective Justice Among States

The debate concerning solidarity and justice among states has missed the key contribution made to international affairs by corrective justice. Unlike distributive justice, which applies within states, corrective justice applies among states. It applies in particular to cooperative arrangements creating interdependence among them. Corrective justice does not require fairness in outcomes. It requires redress in cases of loss caused by unfairness. An illustration of corrective justice among states is the Eurozone’s response to the financial crisis. The assistance offered to the most burdened states was not as an attempt to arrive at fair shares but an attempt to remedy the losses unfairly caused by the mistakes made by the Eurozone as a whole, when designing its basic architecture.

financial assistance programmes helped the burdened states weather the storm so that they all remained in the Eurozone without defaulting on their loans. All of them, however, suffered great economic damage, loss of output and high rates of unemployment.

From Fairness to Solidarity
The Court of Justice has found that these assistance operations were lawful. 4 But what was their moral character? The question remains controversial. Were they manifestations of justice, discharging an obligation of solidarity towards the weaker states? Or were they simply instances of prudence, not strictly required by law or morality, only expedient responses to an emergency? Or were these new loan far too generous, unjustly transferring resources from the responsible to the irresponsible? These are still deeply controversial questions. Some believe that the Eurozone did too little to assist its weakest members. 5 Others think it did too much. 6 All of these positions may rely on some conception of what is fair or unfair. These matters depend on our interpretation of the purpose and nature of the Eurozone agreements, but they also turn on deeper assumptions about the moral obligations arising out of long-term cooperation.
The vocabulary of justification is not optional in this context. Fairness is always a matter closely connected to any interpretation of agreements between two or more parties. This is so both as a matter of the moral character of law in general, as many legal philosophers have shown, 7 but also as a matter of the specific moral requirements of agreements, including international agreements. In many jurisdictions, for example, the law requires 'good faith' in the interpretation and enforcement of contracts, especially in cases where cooperation is for an indefinite period. The distinguished English judge Tom Bingham described the role that fairness in the interpretation of long-term agreements in English and foreign law as follows: In many civil law systems, and perhaps in most legal systems outside the common law world, the law of obligations recognises and enforces an overriding principle that in making and carrying out contracts parties should act in good faith. This does not simply mean that they should not deceive each other, a principle which any legal system must recognise; its effect is perhaps most aptly conveyed by such metaphorical colloquialisms as 'playing fair', 'coming clean' or 'putting one's cards face upwards on the table.' It is in essence a principle of fair open dealing … English law has, characteristically, committed itself to no such overriding 4 See Case C-370/12, Pringle, ECLI:EU:C:2012:756 and Case C-62/14, Gauweiler et al. v Deutscher Bundestag, ECLI:EU:C:2015:400, which was a preliminary reference from the German Constitutional Court in BVerfG, Beschluss des Zweiten Senats vom 14. January 2014-2 BvR 2728/13 -Rn. (1-24). 5 See, for example, A. Mody 2018 and J. E. Stiglitz 2016, 85-144. 6 For example, U. di Fabio 2014, 107-110 andH. W. Sinn 2014, 343. Sinn thought that rather than new loans, Greece should have been given significant debt relief: "The right mixture of debt relief, privatization, and wealth levies could be jointly negotiated in a Paris Club debt conference convened to reset the Eurozone. The European debt crisis has many causes, and creditors and debtors alike share the responsibility. A way to distribute the burden fairly should thus be sought-and it is important that it be found soon", 343. 7 For this general view of law, see R. Dworkin 1986, N. E. Simmonds 2007, R. Alexy 2009. principle but has developed piecemeal solutions in response to demonstrated problems of unfairness. 8 In the USA, the Second Restatement of Contracts requires that "[e]very contract imposes upon each party a duty of good faith and fair dealing in its performance and enforcement". 9 English law takes a narrower view, but it too does not always enforce contracts that parties willingly entered into, if their terms are 'unconscionable' or are the result of undue influence exercised due to unequal bargaining power. 10 So even in England, unfairness can occasionally be a reason for holding a deal unenforceable.
Whatever the position in contract law, the principle of 'good faith' is also an established principle of the interpretation of treaties of international law. Article 26 of the Vienna Convention of the law of Treaties, which bears the title 'pacta sunt servanda' simply states: 'Every treaty in force is binding upon the parties to it and must be performed by them in good faith' 11 . The reference to 'good faith' means that any interpretation is subject to its context and may evolve with the expectations of the parties. 12 The general rules are that a treaty is to be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty 'in their context' and in 'the light of its object and purpose' because 'each of these elements guide the interpreter in establishing what the Parties actually intended, or their common will'. 13 It follows, therefore, that the justification of the assistance programmes as a matter of the law of treaties of the Eurozone can turn on issues of 'good faith' or fairness. It cannot simply be a matter of the narrow reading of explicit technical rules. Whether the law permits such assistance will depend on the context, intention and nature of the parties that entered into the Eurozone agreements as these evolved over time.
These general accounts of fairness do not decide the issue of the fairness of the Eurozone. They do, however, put the question of solidarity in context as a specific manifestation of a well-known question. Long-term agreements, be they private or agreements among states, must be interpreted in light of a theory of fairness, but they do not always create obligations of solidarity. Whether fairness requires solidarity will depend on the specific terms and practices of each specific agreement. If the Eurozone agreements do create such obligations, it must be for a reason. Most theorists say that no such reason exists, given how the Eurozone agreements are currently structured.  [1975] QB 326, where Lord Denning summarized the doctrine of 'inequality of bargaining power' as follows: 'By virtue of it, the English law gives relief to one who, without independent advice, enters into a contract or transfers property for a consideration which is grossly inadequate, when his bargaining power is grievously impaired by reason of his own needs or desires, or by his own ignorance or infirmity, coupled with undue influences or pressures brought to bear on him by or for the benefit of the other'. The House of Lords has refused to endorse such a wide principle, however. 11 For an interpretation see Dorr and Schmalenbach 2011), 427-451. 12 For this point, see E. Bjorge 2015, 189, especially at 203-204. 13 The quotes are from the 'Decision regarding delimitation of the border between Eritrea and Ethiopia', XXV Reports of International Arbitral Awards (2002), 83-195, at 109-110. "The meaning of these Treaties is thus a central feature of this dispute. In interpreting them, the Commission will apply the general rule that a treaty is to be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose. Each of these elements guides the interpreter in establishing what the Parties actually intended, or their 'common will', as Lord McNair put it in the Palena award". The German philosopher Jürgen Habermas writes that the present economic rationales of the Eurozone are incompatible with social justice and democratic legitimacy: 'A technocracy without democratic roots would not have the motivation to accord sufficient weight to the demands of the electorate for a just distribution of income and property, for status security, public services and collective goods when these conflict with the systemic demands for competitiveness and economic growth' 14 Habermas contrasts 'technocratic' blueprints for dealing with the crisis and a project for a 'supranational democracy in the core of Europe'. 15 By the term 'democratic roots', Habermas seems to imply that solidarity needs to be supported by the identification of each person with a single political community as its citizen. He says that the EU's 'technocratic' response runs the risk of lacking 'the motivation to accord sufficient weight to the demands of the electorate for a just distribution of income and property, for status security, public services and collective goods when these conflict with the systemic demands for competitiveness and economic growth'. 16 Habermas' argument has a compelling logic, which is shared by standard defences of the welfare state. The argument assumes that citizens of the same state will have the psychological motivation to recognize a duty of solidarity to one another, normally expressed by ideas and principles of social justice. Indeed, the EU treaties recognize the division of the Union into states and an associated division of them into welfare states. 17 As a matter of fact, European citizens do not feel the same identification with those they consider outsiders. Europe's task, therefore, is to expand outward the bonds of community that exist within states. This is why Habermas argues that the Union needs to create its own 'democratic roots' in common identification.
If this argument is right, then the introduction of meaningful solidarity in the European Union presupposes the profound transformation of its institutions. Habermas is clearly aware of the seriousness of the political challenge. He proposes a 'real political union', which will change the circumstances of European citizens. 18 Social justice can then become a component of a sharing in power under the institutions of a political community. In order to achieve solidarity, we will need to abandon borders within Europe to produce a single, integrated political community.
Is this a valid argument? I do not think it is. I do not doubt the positive aspect of the argument. Of course, a political community creates bonds of solidarity. I do wish to challenge, however, the negative part, namely, the claim that solidarity does not apply when persons are not bound by the deep bonds of citizenship. In my view, obligations of solidarity can arise from relations among persons or, indeed, from relations among states and other collective bodies, when they engage with one another in the right way. Solidarity does not always require a relationship of citizenship. Other relations can have the same result. The key to the argument is that obligations of solidarity arise as acts of assistance under a principle of mutual aid or as a principle of redress. These are not matters of justice in distribution but matters of corrective justice. This point has been obscured by most discussions of solidarity in the European Union, where the dominant language is that of distributive claims.

Distributive Claims
The European Union treaties are not silent on solidarity. Article 3(3) TEU provides that the Union shall 'promote social justice' and promote 'economic, social and territorial cohesion, and solidarity among Member States'. Solidarity is also mentioned in Article 122 TFEU in the context of assistance in emergencies, whereby the Council may decide 'in a spirit of solidarity between member states' on measures assisting a state if 'severe difficulties arise in the supply of certain products, notably in the area of energy'. This article, however, was not invoked in the case of the financial assistance programmes.
It is not entirely clear, however, what these general statements mean. Perhaps, like other general statements in this part of the treaties, they do not intend to create clear obligations, moral or legal, on the institutions of the European Union or the member states. 19 They are supplemented, however, by the mention of solidarity in particular policies. Solidarity is mentioned in the context of asylum policies 20 the general 'solidarity clause' in case of natural or man-made disasters, 21 and in the detailed mechanism for distributions of funds to the member states. These funds are very important for the distribution of resources from the wealthier states to the poorer ones. They were created at various points in the history of the Union and are now provided for in the treaties. The European Social Fund is covered by articles 162 to 164 TFEU, Structural Funds are provided for in article 175 TFEU, and a Regional Development Fund is provided by article 176 TFEU. The treaty empowers the EU to distribute funds with the aim of improving employment opportunities; strengthening social inclusion; fighting poverty; promoting education, skills and life-long learning; and developing active, comprehensive and sustainable inclusion policies. 22 Nevertheless, the total amount of spending on such projects is very small, at least in relation to the overall EU economy. It is also very small in proportion to similar funds available within each state for social purposes. In the member states, the equivalent spending ranges between 35% and 58% of gross domestic product. In the total EU budget, the equivalent amount is only about 1% of the gross national income of the 28 member states. 23 According to an estimate made by the Institute of Fiscal Studies, the EU's total spending on structural, cohesion and agriculture and rural development funds accounted for about 38% of the total EU spending in 2014. The scale of redistribution involved is therefore very modest indeed. The direct European Union social spending is about 20 or 30 times less than what occurs within a member state. The small size of these transfers strengthens the argument that solidarity in the European Union is not really a functional principle. Solidarity may be an exclusive preserve of the member states operating only within their separate political communities. Many sophisticated commentators observe that the social dimension of the EU is minimal. Commenting on the 'patchwork' of social justice provisions in EU law, Gráinne de Búrca writes that solidarity had a 'constructive potential' which could gradually 'promote a degree of solidarity and mutual responsibility-however tentative and limited at first-between states, citizens, and other residents within the enlarging European space'. 24 Such cautious endorsements of solidarity, however, cannot hide the fact that social justice is a very small part of the EU. Any authors thus take the view that, since the EU is not a state, the normal rules of distributive justice do not apply to it. Since solidarity or social justice is a principle that only applies within states (or equivalent political communities), no solidarity principle applies in the European Union, at least not yet. 25 In this vein, Professor Christian Joerges, a leading scholar on European Union law, argues that the European Union is incapable of having a proper policy towards social justice. Joerges refers to an argument made by Friedrich Carl von Savigny, the leading German legal scholar of the nineteenth century, that justice applies to private relations but not to the relations between states. In Joerges' reading, Savigny has shown that 'interstate relations … remain in an unruly state of nature governed by power and politics rather than law'. 26 From Savigny's premise, Joerges draws the conclusion that the European Monetary Union is incapable of imposing a uniform architecture of social policy. It does not have the means. Europe has instead created a competitive 'single market' where each state seeks to have a social policy on its own, allowing for great 'socioeconomic diversity', which the institutions of the EU cannot address.
Joerges adds that the various economies have diverged rather than converged since the creation of the Union. Europe's idea of 'a highly competitive social market economy' results in the undermining of social justice and solidarity everywhere in the EU. He says that 'the socioeconomic diversity of the Union was treated with benign neglect and an institutional framework with the potential to manage the implication of this move was not established'. 27 He finds this 'fateful' for the 'prospects of Social Europe'. 28 The only policy that, for Joerges, is working at EU level is that emanating from the European Central Bank, which in his view is insensitive to the concerns of social justice. Hence, the 'Maastricht arrangement was an illdefined political compromise, rather than a sustainable accomplishment of constitutional validity and strength', which has led to 'authoritarian managerialism'. 29 The only way of overcoming this stalemate, for Joerges, is through the creation of a 'transnational democracy' through federal institutions. He insists that such a transformation is urgent, because the present arrangements do not provide a solid basis for solidarity and social justice. 30 A similar view has been taken by Floris de Witte, who has offered an extensive and original discussion of the question of social justice in the EU. De Witte argues that 'justice is an associative commitment: it is born and sustained by the interaction between particular groups of citizens'. 31 Nevertheless, de Witte's discussion locates justice firmly within the framework and architecture of institutions of the state. De Witte argues that the existence of the European 24 G. de Búrca 2005, 9. 25 This is a view reflected in the writings of influential legal scholars. It is the view, for example, defended by A. Williams 2010), A. Somek 2015, 295-310, andN. Walker 2015, 247-258. 26  Union 'ensures that the associative connections between citizens across borders are incorporated within the pursuit of justice on the national level'. 32 In effect, the EU brings citizens closer to one another, by putting them 'in a social relationship' in the sense that 'all EU citizens are part of the incipient and ill-defined European polity' but also on 'the national level, where citizens of different nationalities engage with each other through a whole range of social interactions'. 33 But de Witte concludes that since the EU is in effect an international project, it does not have the potential for justice: 'The EU does not dispose of the institutional sophistication required in order legitimately to articulate an autonomous conception of supranational justice…. In consequence, the EU cannot engage in the definition of supranational ideas of equality, fairness, or distributive justice in a legitimate way'. 34 Surprisingly perhaps, Joerges and De Witte are in agreement with the former Finance Minister of Germany, Wolfgang Schäuble. In a widely leaked 'non-paper' to the ministers of the Eurogroup on the occasion of his departure from his post in November 2017, Dr. Schäuble gave a detailed account of his overall assessment of the Eurozone, where he also expresses scepticism about the potential of the European Union to achieve social justice. It lacks both of the mechanism for redistribution and the required institutions of democratic accountability. The paper states that the Ministry's view fiscal responsibilities and fiscal control belong together. The paper observes that there are only two ways in which the EU could maintain the symmetry of responsibilities and control: Either we transfer parts of national sovereignty and control of fiscal rules to the EU level ("Euro Finance Minister"), together with a greater democratic legitimacy. This would certainly require EU treaty changes to be credible. Or we agree on an intergovernmental solution. 35 The real choice here appears to be between full union and loose integovernmentalism. The 'non-paper' by the Finance Ministry, just like Joerges, De Witte and Habermas, relies on the assumption that an intergovernmental institutional setting is an inappropriate institutional arrangement for the operation of solidarity. The paper therefore proposes further political union through treaty change as a priority for the future. But, as long as 'there is little willingness for treaty changes, we should follow a pragmatic two-step approach: Intergovernmental solution now, to be transposed into EU law later on'. 36 In summary, the views I have examined by Habermas, Joerges and De Witte make the following negative claim: they suggest that solidarity is strictly limited to the national domain and cannot develop in an international context. I will now argue that this assumption is mistaken. I reject the premise that solidarity applies only within states and that solidarity does not apply to international or intergovernmental structures. In order to understand this point, we need to expand our scope beyond familiar assumptions about the domestic welfare state. I will argue that solidarity is not identical to distribution. Solidarity has in fact two distinct grounds, 32 Ibid. 33 de Witte 2008, 123. 34 de Witte 2008 Available at: https://www.scribd.com/document/361120275/German-finance-ministry-non-paper-on-Eurozone-reforms. 36 These views seem to correspond to standard views among German economists. It is common place in German economics departments that within the Eurozone 'ever member country has to do its homework and that relying on bailouts from others distorts incentives'. See Brunnermeier et al. 2016, 97. one in distributive justice and one in corrective justice. Only distributive justice is restricted to the domestic case.

Distributive and Corrective Justice
Although the distinction between distributive and corrective justice is old and well understood, it has not, for the most part, been deployed in the debates concerning international justice. 37 It appears that when scholars refer to justice in the international context, they only have in mind distributive claims. If distributive claims cannot be justified, then justice is irrelevant. 38 These views seem to work on the presupposition, or at least the mental picture, of justice relying on a central distribution of a good to beneficiaries according to criteria of need or merit. This is the typical model of the distribution of fair shares outlined by Aristotle and repeated many times in modern philosophy. It is the standard model of distributive justice as justice in distribution. John Rawls too identifies social justice with distributive justice when he says on the 'subject of justice' that: 'the primary subject of [social] justice is the basic structure of society, or more exactly, the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation' (emphasis added). 39 In the case of the state, the distribution takes place through central taxation and social welfare schemes. The distribution is not from state resources, but effectively from the mostly better off taxpayers to the least well off, as they receive public assistance through public funds or through freely available services, such as education and health care.
It is clear that such a model of justice cannot apply to the European Union. First, there is no central taxing authority collecting public funds. Second, there is no central spending power of such funds. Third, there are not in place any institutions with the appropriate democratic powers for deciding on the appropriate test of distribution, since the European Parliament cannot have such powers. Fourth, we do not have clear institutions for the accountability of such decisions. Fifth, we do not have the underlying 'community' or 'demos' which would see the public support the transfer of funds from one part of the community to another and give meaning to mechanisms of participation as wemore or lesshave in the conventional welfare systems of the states. For these reasons, the advocates of political union as a precondition of institutions of solidarity, such as Habermas, di Fabio and Joerges, must be right to rule out a general distributive principle for the European Union under the present institutional arrangements. If social justice makes sense only within a territorial state, then it can have no application to the European Union as it is today.
This argument, however, is not the end of the matter as far as solidarity is concerned. It fails to take into account the way in which solidarity works in other ways. The relations between states just like the relations of persons are subject to a second dimension of justice, which since Aristotle we call 'corrective justice'. Corrective justice does not provide for the distribution of something from a central source or state but accounts for the just relations among two or more parties in the event of cooperation, exchange or reparation. 40 Corrective justice addresses injustice by restoring the original positions between a person that suffered a loss and the person who gained a profit from the other's expense. The just redress is the arithmetic mean between the part of the earner and the part of the loser. 41 In Ernest Weinrib's apt description, the organizing idea is that of correlativity. The 'elements of liability under corrective justice can be explicated only in terms of concepts whose normative force applies simultaneously to both parties'. 42 Unlike distributive justice, corrective justice takes the parties to be equal. For corrective justice 'liability involves a conception of fairness that recognizes the equal normative status of the two parties and treats their normative positions as mirror images of each other'. 43 These ideas have generated some very sophisticated arguments about the substance of the law of tort, contract and unjust enrichment. There is no need to rely on these theories in any detail here. What we need for the purposes of our argument is only the idea that states are independent agents, similar to corporations or other collective agents that are subject to private law when they are cooperating towards a common project, sharing in the process rights, obligations and risks. Such relations create mutual moral obligations from each state to all others that are analogous to legal obligations arising in contract law. I say that are merely analogous and not identical, because there is no contract law in the international sphere. Strictly speaking, since there is no central power of enforcement in international law, and since all legal obligations are to some extent provisional under public international law, there cannot be contract law in the strict sense. Under international law, states can have obligations of law to one another, but they cannot rely on the jurisdiction of a common court to enforce them. So the analogy with contract is incomplete. Still however, states do owe each other duties on the basis of their agreements.
That states have mutual obligations to one another which is not an original thought. In the Law of Peoples, Rawls argues that the international community must be based on principles of fairness that apply to states. He argues that such principles would have been adopted by wellordered peoples in a hypothetical original position for the 'law of peoples'. In Rawls' account, the relevant agents are the peoples and not their citizens. He argues that inequalities between states will be unjust 'because of their unjust effects on the basic structure of the Society of Peoples, and on relations among peoples and among their members'. 44 He further argues that in such a scenario, the parties will 'formulate guidelines for setting up cooperative organizations, and will agree to standards of fairness for trade as well as to certain provisions for mutual assistance'. 45 Nevertheless, the relevant principle of justice does not presuppose a central distributing power. It applies in a decentralized way among the various states, as a selfimposed constraint on their foreign policies by way of 'cooperative organizations' and other agreements.
In this account, the principles of international justice bind states in their mutual relations: they create obligations to exercise a particular kind of policy or to discharge their natural duties to each other. A principle of mutual aid is part of these obligations. One of the eight principles of the law of peoples is that 'peoples have a duty to assist other peoples living under 40 For a general historical account of Aristotle's distinction between distributive and corrective justice see, I.  Ibid. 44 J. Rawls 1999a, 113. 45 J. Rawls 1999a unfavourable conditions that prevent their having a just or decent political and social regime'. 46 The obligation to assist binds peoples but exists only up the point at which societies overcome 'unfavourable conditions'. This is not strictly speaking a matter of the distribution of resources between the parties, but the one off lifting of our fellow human beings from a state of destitution. Rawls' principle of aid is solidarity in assistance, not solidarity in achieving fair shares. 47 Although Rawls does not use the term 'corrective justice' for this type of bilateral obligation, it is clear that what he has in mind for the Law of Peoples is very different from the kind of justice that applies to the basic structure of a single political community. It is not a principle of distribution of fair shares but a bilateral relationship crated by the face-to-face encounter of one person or one state with another. The claim arises because of desperate need of the burdened person or state, not because of a pattern of distribution. This is a different moral relationship to that of citizenship. When we encounter each other as citizens, we look at the whole picture in relation to a vertically organized system of authority to which we both contribute. We then raise the question of the existence of inequality in the distribution of 'primary goods' among us buy that vertical system of authority. In the international domain, by contrast, our encounters are structured in a purely horizontal way by way of 'cooperative' agreements, without a vertically organized system of authority guiding our actions or resolving our disputes. The relevant domain of justice here is created by horizontal relations only. Rawls presents the obligations of solidarity or justice in the international case as arising out of the reasonable terms under which states would join a system of international law as equals. Rawls' argument introduces in this way the idea that fairness should be a precondition of general international law, whatever else agreements may establish as binding obligations.
In this view, the horizontal principle of fairness is a background condition for the continuing legitimacy of the international community as a whole, including the multilateral treaties among its members. The best way in which we can understand such a non-distributive principle of fairness is as a principle of corrective justice, similar to that giving rise to a duty of mutual aid. It arisesjust as it does with personsfrom the very fact of the vulnerability of states towards natural disasters or other misfortunes. 48 If a prosperous state is in a position to provide aid to a burdened state without risk to itself, it has a duty to do so. A similar principle of corrective justice covers the cooperative practices of international law in the following way. If states engage with one anotheras they shouldin setting up complex cooperative arrangements in the course of international relations, then they are bound by the terms of fairness towards one another, just like motorists sharing a road are bound by the rules of negligence for accidents. We often call this aspect of justice 'cooperative justice'.
The distinction between distributive and corrective justice has one important consequence for the European Union. It suggests that the simple disjunction drawn by the Habermas and 46 J. Rawls 1999a, 37. 47 There is an analogy with the idea of a duty of mutual aid among strangers. Such a duty is ethically fundamental outside any legal or institutional framework. I follow here B. Herman 1984, 577-602, who applies the idea of the 'dependence' of human beings to the process of the categorical imperative and concludes: 'The duty of mutual aid has its ground in the facts that we are dependent beings with ends it is not rational for us to forgo: ends set by "true needs" whose satisfaction is a necessary condition for the exercise of rationality. As we are rational agents, we set ends… As a person's true needs are those which must be met if he is to function (or continue to function) as a rational, end-setting agent, respecting the humanity of others involves acknowledging the duty of mutual aid: one must be prepared to support the conditions of the rationality of others (their capacity to set and act for ends) when they are unable to do so without help' (p. 597). See also B. Herman 2012, 391-411. 48 By analogy with dependent human beings, as in B. Herman 1984. Joerges between independence and solidarity was not correct. Remember that they assumed that there were only two ways forward as far as the future of the Eurozone was concerned: either a decentralized and international fiscal order in which each country was exclusively liable for its own debt or a fully integrated fiscal union in which spending powers are transferred to a European authority. The distinction necessarily assumed that there could not be a sharing of burdens or a practice of solidarity in an international union, without some form of integrated political union. This is correct, only as far as distributive justice is concerned. If there is no single political power to tax the parties and distribute the shares justly, then it makes no sense of speaking of 'just shares'. The question is very different if we are asking about something else. Solidarity in the form of mutual aid does make sense without a central authority doing the sharing. Corrective justice that arises from international cooperation does not require political union.
The question that corrective justice asks is not one of fair shares, but of fair redress for a wrong. If parties provide redress to one another for the wrongs they committed, they exhibit solidarity outside a distributive framework. If I drive carelessly and I cause a cyclist to lose their balance, I owe them a duty of reparation, even if it is only an apology. If I have injured them I must provide appropriate redress. If I ignore my moral obligation to another person in need of aid, I showamong other thingsa lack of solidarity. It is the same with states. A state may owe to another state some form of redress, for example, the return of a cultural artefact stolen by some of the first state's citizens (e.g. the Parthenon marbles looted by Lord Elgin) or compensation for an oil spill caused by a military vessel. This sense of solidarity is entirely familiar from private law. There is no reason for not following the same principle in international relations.
The best example of this kind of obligation in international relations is the granting of debt forgiveness as an obligation under the doctrine of 'odious debt' in international law. 49 Under certainvery demandingconditions, the debts assumed by a corrupt government may not be held to bind a subsequent democratic government. The doctrine of 'odious debt' holds that the debt must be unlawful and therefore invalid. A lender should not seek to exploit a dictator's greed for one's personal profit. Contract law has a similar doctrine about contracts whose object is unlawful. Under that doctrine, Cuba and the USA refused to recognize Cuba's debts that had been taken by the Spanish colonial government just before Cuba's independence. 50 The same applied to loans taken by the Costa Rican government in the dying days of the dictator Federico Tinoco's corrupt regime. 51 In all these cases, fairness in substance was deemed more important than the terms of an existing agreement, which was considered invalid.
Of course, the theory of 'odious debt' cannot conceivably apply to any member of the European Union. The issues are entirely different. All the members of the Eurozone are democracies and have democratically approved the original accession to the European Monetary Union as well as any amendments to the treaties. They have borrowed in the international markets exercising their constitutional powers in fiscal policy. Nevertheless, the rule of odious debt may help us understand the basic principle, namely, that fairness 49 R. Howse 2007, J. King 2016. 50 See R. Howse 2007 Aguilar-Amory and Royal Bank of Canada claims (Great Britain v Costa Rica) (1923) 1RIAA 371. See R. Howse 2007, 11-12. applies more generally as a background principle for all international institutional arrangements. 52 The general point is this. Ιf cooperative fairness creates international obligations under a principle of corrective justice in some cases of international interaction, then it can also create obligations in a process of market integration such as that provided for by the European Union treaties. In the most direct case, if a state suffers a loss because of the wrongful action of another state, it is then entitled to redress, and failure to provide it is a failure to exhibit appropriate solidarity. Indeed, the European Union has created exactly such mechanisms. It provides redress for violations of the law through the Commission and the Court of Justice. But we are entitled to ask if corrective justice creates further moral obligations of fairness or solidarity among states on the basis of their cooperation. We need to refine the argument by asking in particular who can be responsible for the joint decisions of a cooperative agreement such as the European Union.

Structural Responsibility
The principle of mutual aid applies between two parties that encounter each other in a certain way. The case of international cooperation in trade agreements is more complex. The parties organize their relationship based on common rules and monitored by multi-lateral institutions. Here, the argument for a principle of cooperative fairness must take a modified form. By entering the founding treaties, the member states jointly created institutions of cooperation, such as the Commission, the Council, the European Parliament andmore recentlythe European Central Bank. Such institutions make decisions in the name of the organization as a whole. Some of these decisions are highly consequential. They may determine a member states' social and economic future. These common institutions are not entirely independent nor do they enjoy comprehensive powers. The member states retain economic decision-making and exercise some element of control through the Council of ministers (and through their delegates at the European Parliament). The member states remain, therefore, key decisionmakers and are to that extent responsible for what the EU does. 53 At the same time, we cannot hold the states responsible for everything that the EU does. But since the member states created the basic structure, such as the single market and the free movement of persons, goods, services and capital and if we can detect some economic outcomespositive or negativewith these structures, then the member states can be held to be jointly responsible for them. This is not an entirely original thought. Something similar may already exist in the structures of global economy. It is common ground that the current structures of economic globalization are the conscious result of policy decisions made by the leading economic powers, for example, the global financial architecture, the World Trade Organization and numerous other multilateral agreements that make trade possible. 54 The philosopher Thomas Pogge argues, for example, that world poverty is partly the result of the joint actions of 52 An attempt to interpret Greece's debt as 'odious debt' is made in Bantekas and Vivien 2016, 539-565. In my view the argument cannot possibly succeed. Greece's democratically elected governments created the debt mountain in the period between 1980 and 2010. 53 In this sense, we must consider the European Union to be an international and not a federal project. developed states, which make it possible for dictators and corrupt politicians to siphon their illgotten gains to tax heavens around the world. 55 The most extensive argument in this respect is offered by the American philosopher Aaron James. In his book, Fairness in Practice describes the process of creating institutions of international trade as an international 'social practice', a practice of 'mutual reliance on common markets' which, in his view, creates 'a distinctive class of fairness responsibilities'. 56 These responsibilities go beyond the explicit commitments of the parties. They are structural in that they require that the structures of international trade meet certain requirements of 'structural equity'. James argues that the states that create these structures of international trade are jointly responsible for any harm and unfairness that such structures bring about. In James' argument, the states' negotiation has, in effect, the same effect as direct legislation. 57 The states are structurally responsible for their processes of negotiation, just as much as if they were directly legislating the resulting consensus. What are the structural principles of fairness that bind the states, when they act in this way? For James, these principles require 'due care' for those who unfairly lose out from the system overall, principles of 'fair distribution among states' and 'fair distribution within states'. 58 The states that participate in the international social practice of trade are jointly responsible for setting up mechanisms for compensation to the losers of free trade, as well as mechanisms for maintaining equality in the distribution of benefits among states and among populations within states.
An argument along these lines concerning structural responsibility is clearly applicable to the European Union, where the member states have legislated the terms of their own cooperation. They have created legally binding treaties and created enforcement mechanisms. Their actions have gone much beyond a mere 'practice' of international trade, as described by James. In the case of the European Union the structural responsibility of the states is even more direct than that of the creators of diffuse systems of cooperation, such as the World Trade Organization or systems of international arbitration.
Cooperative structural fairness applies thus to the member states in two stages. At the first stage, which we may call 'basic fairness', the principle of fairness asks if an agreement to cooperate, given its formation and substantive content, is fair overall at the point the parties first enter it. This test requires, for example, that when the parties enter into some agreement to cooperate, they do so willingly and having been in a relatively even bargaining position to one another. The principle also assumes that, absent special circumstances, the cooperating parties would strike an agreement by which they would receive a fair return on their investment over time. Otherwise, the agreement might have been the result of exploitation or undue pressure.
At a second stage, which we may call 'fairness in practice', the principle asks if the parties acted fairly as their agreement unfolded in practice. This aspect of fairness asks of the content of the respective obligations of the parties after the agreed rules are put into effect. John Rawls explains the relevance of fairness in practice in the following way. He says that fairness creates obligations on the participants in a mutually advantageous cooperative venture to continue to acquiesce by its terms. 59 As Rawls puts it: 'we are not to gain form the cooperative labors of others without doing our fair share'. 60 This goes beyond the basic fairness that was in place at the original starting point. Whether we are fairly treated by others depends on how they treated us in the process of cooperation over time. Although some of the ongoing cooperation will be based on the original rules and commitments, the relevant obligations will also arise from the conduct, practices and expectations created and relied upon by others after the cooperative project started. So a cooperative project may be fair to begin with but can fail to remain fair because of the subsequently unreasonable conduct of one of the parties.

Basic Fairness
I now turn to a fuller discussion of the idea of basic fairness in the interaction among states. My focus remains the framework of multiparty cooperation and principles of corrective justice. We are not asking about the ideal distribution of shares or about a certain ratio of return to investment. No such ratio of 'fair return' exists in the absence of a central distributing mechanism. The outcomes of any trade agreements are determined by economic practice, or in effect by a mixture of political judgement, prudent decision-making and, of course, chance. Because cooperative justice is a matter of bilateral relationship structured around the parties' actions, what we are looking for is a standard of wrongness in bilateral relationships. What is fair and what is unfair in cooperation among states? Basic fairness rules out terms of cooperation that impose unjustified inequalities, or terms that are the result of unfair imposition, monopolies, cartels and other competitive restrictions. 61 Any such terms must be void and unenforceable. Can we organize these intuitive wrongs in a more coherent whole, so as to cover less obvious but still real unfairness?
There are many ways in which we can interpret fairness in trade agreements. One set of was provided, as we saw above, by Aaron James with reference to the world trade order. James argues that the relation among states must be subject to a principle of 'international relative gains' according to which: 'gains to each trading society, adjusted according to their respective national endowments (e.g. population size, resource base, level of development), are to be distributed equally, unless unequal gains flow (e.g., via special trade privileges) to poor countries'. 62 Why should the gains be in principle equal? James argues that 'the gains of trade are socially created, by the joint practice of market reliance'. 63 This of course is the starting point for the responsibility of states, although not a sufficient reason for equal shares. He goes on to say: 'Because each trading country has a morally relevant interest in greater rather than lesser national income gains, equal treatment requires equal distribution of gains, unless we can specify a relevant difference among participating countries'. 64 James identifies two such 'relevant' differences as possible grounds for inequality of gain. First, relevant endowments of each country, such as population size, natural resources and degree of economic development. Second, 'inequality of gain is fair if greater benefits flow to people who are worse off in absolute terms'. 65 I find James' arguments for equal shares and for requiring priority for the worse off unconvincing. What James recognizes as the 'endowments' exception cannot be limited only to the features he sets out. Other considerations are relevant too. The performance of a state in trading with other states does not depend only on the state structures but also on the success of private parties: producers, entrepreneurs and workers that produce relevant tradeable products that appeal to consumers abroad. Economic success or failure cannot therefore be imputed only to each state and its institutions. As long as the background terms of fair competition are respected, the outcome of economic competition will be the result of the 'endowments' that James describes as well as of the capacityflexibility, agility and resourcefulnessof firms in producing competitive products and services. This is why James' principle of equality in outcomes seems to me at odds with his general framework. He seems to leave out the element of uncertainty inherent in economic competition. If we allow for the effects of economic competition, equality of outcomes cannot be a default position.
A second argument for fairness is provided by Andrea Sangiovanni, who has offered a sophisticated argument about fairness in trade specifically in the context of the European Union. 66 Sangiovanni takes an 'internationalist' view on the justice that applies to the EU. He believes that the best argument for justice in the EU proceeds from the reciprocal commitments the states have made to each other. Sangiovanni argues that: 'According to reciprocity-based internationalism, demands for social solidarity at all levels of governance can be understood as demands for a fair return in the mutual production of important collective goods'. 67 Fair return is required because in effect the states have jointly produced the collective goods brought about by the European Union.
Sangiovanni's proposal is to ask what would be a rational insurance policy in case the member states considered the terms of their cooperation in advance of entering into this cooperative project. He asks '…what do member states and their citizens owe one other as a fair return for the mutual provision of these goods and the mutual exposure to these risksgoods and risks made possible by opening up their markets, societies and polities to the joint control and supervision of both supranational actors and intergovernmental decision-making?' 68 The answer for Sangiovanni is that the fair return 'which members states owe one another, under member state solidarity, is given by the level at which each state would insure against the potential losses identified above had they known the distribution of risks but not their place in the distribution'. 69 Sangiovanni argues that by asking how states would have insured against the 'risks intrinsic to the project of European integration, had they not known what state they would have turned out to be', we eliminate the advantage at the bargaining table of the European social contract obtained by the fact that member states know their relative positionincluding their level of development, population size, welfare regime, type, etc. 70 This argument is ambiguous, however. It is not clear if it proposes 'fair shares' as distributive shares, or a system of minimum insurance against failure. A question of what fair shares are appropriate is, as we have seen, a distributive question. It argues for an appropriate allocation among the various parties. Insurance is by contrast a corrective question. It is not an allocation of shares but a provision of redress.
It seems to me that the distinction between distribution and insurance is not clear in Sangiovanni's argument. The aim of insurance is to provide reparation for some loss insured against. It is not the allocation of fair or equal, or less unequal, distributive shares, all things considered. On the one hand, he speaks of an argument from insurance, but on the other, he is also seeking fair shares. I believe that his argument needs amendment to make clear that what is relevant is a basic form of insurance and not a form of distribution. If we are seeking to model our fair background conditions of basic fairness on the basis of a hypothetical insurance scheme, we must envisage returns triggered by a type of wrongful action, not merely inequality in outcomes. The redress must correspond to the wrongful loss actually incurred, or loss in comparison to the position one would have been without the wrong.
Here is how, I believe, the argument for basic fairness among states must be restated. Remember that the relevant principle is corrective justice as it applies to cooperative projects. It is one a principle of distributive shares. We proceed by identifying the threshold of wrongful conduct which generates a one-off claim for redress from one party to another, not a general pattern of fair distribution that should be imposed when it fails to materialize in real life. I propose the following formulation as the appropriate standard of background cooperative fairness in trading among states, which I call the 'symmetry principle': The Symmetry Principle: An agreement to cooperate for the purposes of international trade in goods and services is unfair and potentially unenforceable, if it is shown to create asymmetrical opportunities for gain and risks of loss for the parties involved, taking into account the parties' original position, endowments and prospects as the agreement was reached.
Following the principle of symmetry, a structure of cooperation will be unfair, if it creates asymmetrical opportunities for gain and risks of loss. The symmetry here refers both to the level of risk but also the gravity of the injury it may cause. Irreparable harms must be given much higher value than temporary harms. Similarly, permanent gains (e.g. those associated with education or long term health) are to be assessed differently from transient gains, economic or social.
The principle of symmetrical risks tries thus to capture the idea proposed by Sangiovanni through the insurance approach. This is the thought that an agreement is fair if the parties would have been able to agree it as sufficiently beneficial to them, had they entered it without full knowledge of their circumstances. 71 The idea of symmetry of risks seeks to reflect the idea of reciprocity for states that are greatly unequal in size and economic size. Reciprocity entails a form of insurance for all parties: we ask what costs the member states would be willing to pay to offset the risks generated by the project of cooperation. 72 There are also similarities here with a principle that Aaron James discusses by the name of principle of 'reciprocity of risk', which he eventually rejects. 73 James' idea involves only risk of loss, not opportunity for gain. Yet, his account of reciprocity is the same idea as that behind the 'symmetry principle'. One of the reasons he gives for rejecting the 'reciprocity of risk' 71 An analogy may be made between this argument and Rawls' argument for the making of a constitution in a constitutional convention, in the Theory of Justice. This stage of law-making, for Rawls, was the second of a 'four stage sequence' where the parties were aware of their individual circumstances (whereas they were not in the 'original position' concerning the basic structure. 72 The logic of an insurance calculation is the same as set out by A. , 18. 73 A. James 2012 principle is that 'current trade practice is clearly a situation of non-reciprocal risk'. This, however, is not a reason to reject the principle. He also notes that symmetry of risks is compatible with terrible losses for both sides, if they are symmetrical. 74 This argument fails, however, for another reason. If a cooperative agreement is open to massive losses of the kind that James has in mind, it will not be struck at all, at least not if we rule out imposition, deception and the like.
The underlying idea behind the symmetry principle is a requirement of reciprocity. In the philosophical literature, reciprocity commonly refers to the requirement that one returns a benefit they have received from another. 75 In his defence of reciprocity as an ideal of private law, Arthur Ripstein introduces it as follows: 'The root idea, fundamental to both fair terms of interaction and the idea of responsibility, is one of reciprocity, the idea that one person may not unilaterally set the terms of his interactions with others'. 76 This is a distinct matter from that of keeping a promise. A promise or an exchange of promises creates obligations by virtue of itself alone. Reciprocity creates obligations by virtue of rendering a benefit to another, irrespective of a promise. The economist Serge-Christophe Kolm, for example, begins his wide-ranging study of reciprocity with a definition that stresses that reciprocity goes beyond a 'binding exchange agreement'. 77 Reciprocity applies beyond agreements, when, for example, no agreement exists or the one that exists has failed to meet a fair measure of equal return among the parties. In those cases, the reason to offer somethingor the motivation-is independent of any promise or other undertaking. Some key examples Kolm discusses, for example, are the reciprocity of giving and receiving gifts or reciprocity in family relations.
Philosophers today deploy reciprocity as a political ideal when they discuss the design of a social contract or give a similar egalitarian basis for social life. John Rawls, for example, explicitly discussed reciprocity alongside legitimacy. He links the idea of a reasonable person with a recognition of the value of reciprocity as follows: 'Reasonable persons [are moved by a desire for] a social world in which they, as free and equal, can cooperate with others on terms all can accept. They insist that reciprocity should hold within that world so that each benefits along with others'. 78 In light of the disagreements, we expect to have with others about the terms of cooperation, legitimate political power of one person over another requires that: 'our exercise of political power is fully proper only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason'. 79 For this view, political power must fulfil a criterion of reciprocity, in that citizens must reasonably believe that all can reasonably accept a particular set of institutions. Reciprocity is therefore at the heart of corrective justice. As Aristotle put it, voluntary exchanges or instances of cooperation raise issues of corrective justice regarding voluntary transactions. 80 For Aristotle, justice in voluntary transactions involves not simply keeping their parties to their word by way of a formal equalityas in a simple commercial contractbut at the same time by being willing to preserve proportionate equality in the fruits of joint activity or cooperation. 81

Conclusion
When the European Monetary Union was first proposed, it hoped that the common currency would create conditions of economic and political convergence among the member states. 82 Looking back, it is clear that this aim has not been achieved. The Eurozone today is fragmented between states of the 'core' that have low unemployment, high rates of investment and healthy growth and states of the 'periphery', which have high unemploymentespecially among the younglow levels of savings and investment and very low rates of growth. Some of the most distinguished commentators have linked this failure to the Eurozone's original design. 83 The American economist Kenneth Rogoff, who has co-authored one of the leading studies of financial crises, 84 has given a categorical rejection of the Eurozone's design. He has said that 'the problem at the heart of the euro crisis' is that 'the eurozone is a half-built house' and that 'it was a catastrophic mistake to put monetary union ahead of fiscal and political union'. 85 The absence of central fiscal policy, for Rogoff, meant that European policy makers did not have the tools to address the crisis. He concluded: 'Monetary policy is simply one side of fiscal policy. Monetary union without fiscal union is an accident waiting to happen'. 86 Similarly, the distinguished economist C. Fred Bergsten wrote that 'the European crisis is rooted in a failure of institutional design' and that 'the absence of crucial policy tools constrained Europe's ability to reach a solution quickly, triggering severe market reactions that continue to this day'. 87 If these economists are right, then there is a clear case of 'structural responsibility' burdening all the original members of the Eurozone. They must all be held responsible for creating an economic practice with potentially disastrous consequences for some of them. Twenty years after the Euro came to being, the economies of Greece, Italy and Spain are now caught in a vicious circle of low growth, high indebtedness and high unemployment. 88 These states cannot now leave the Euro without running even more serious risks of economic collapse, which would set back their own economic recovery even further. This is a loss that demands redress, if it has been caused by an unfairly burdening overall structure. If we accept these leading economists' arguments, we can see that the financial assistance programmes were at least part of a required redress for unfairness.
Given that the crisis had systemic roots in a collective failure, the rescue operations can be seen as the appropriate moral response. Since the loss was caused by the design of the cooperative project, the principle of structural responsibility must take effect. The programmes are manifestations of morally required solidarity under a symmetry principle: they compensated the burdened member states for the loss caused by the structural flaws of the Eurozone's design (although they, arguably, also took into account those member states' own contributory culpability in not doing enough to avoid the worst aspects of the crisis during the boom years). If this argument is correct, then the financial assistance programmes were not merely exercises in self-preservation and economic prudence. They were also expressions of a new form of solidarity generated by the economic cooperation of European states under common principles of corrective justice.