The “practicability” of his idea of justice is a major concern in his work, most obviously in Political Liberalism. Since then, his principles of justice – the fundament of a just social structure in a pluralistic society – have had to pass the “test” of a shared “overlapping consensus” that consists of a reasoned and public agreement among all citizens. The increasing focus on practicability becomes even more visible if one turns to his Collected Papers, edited by Samuel Freeman. There one finds an almost complete collection of all of his published articles, which Rawls himself has modestly called “experimental work” (ix). The pieces are ordered chronologically, starting with a chapter from his 1951 dissertation and proceeding with an article on utilitarianism, which later became an important source for the conception of “reflexive equilibrium.” After studies on his Theory of Justice, the middle pieces cover the ideas laid out in Political Liberalism on reasonable pluralism, overlapping consensus, and public reason. A much shorter precursor to the now available The Law of Peoples, originally printed in 1993, together with an essay on Hiroshima and an interview, bring the collection to a close, thereby throwing light on the efficacy of his theory in other fields, such as international politics and law.

In The Law of Peoples, various theoretical threads developed over decades are bundled together, while others are quite surprisingly weeded out. In his long essay Rawls transfers his idea of justice to the international level. The main aim is no longer to find the most appropriate conception of justice for the evaluation of social institutions within a democratic society. Rather, the focus now is to put forward a conception of justice that will be most conducive to peaceful institutional cooperation among peoples. In view of the enormous challenges international law has faced during the last ten years, the tightrope walk between reaching for the normative stars and not losing touch with political circumstances on the ground has become rather shaky.

Only two main aspects should be mentioned here. First, the growing number of people who live in extreme poverty or are dying from hunger or disease demands international justice and clearly defined implementation measures. Second, the war in ex-Yugoslavia and, most recently, the terrorist attacks on the political and economic centers of the US have raised the question of whether under certain conditions it is legally justified not to respect a state’s sovereignty. These two aspects, however, differ in one important respect. The general question connected with the war in ex-Yugoslavia is whether the law of peoples does and should allow for intervention in a country that disrespects fundamental human rights. The more general question that arose after the terror attacks in the US is whether a state has a “right to self-defense,” even if the aggressor is not a state and military weapons were not used.

Partly in response to criticisms but also partly reflecting Rawls’s long-sightedness, the new developments in the last edition of The Law of Peoples relate to these two issues. Adding social justice to the law’s agenda surely represents the greatest change. And the extensive supplements Rawls has included concern the problem of how the law of peoples should deal with “outlaw regimes.” But before I go into a more detailed discussion of these two points, I will turn to problems that are inherent in the theory when the liberal idea of a social contract is expanded into a law of peoples.

It is also with the aim of increasing its “practicability” that Rawls develops his law of peoples on two levels. The “ideal theory” justifies the law of peoples, or, more precisely, it justifies moral principles that function as a pattern for the evaluation and interpretation of existing law and for the creation of a new one. In the “non-ideal” part, he focuses on how the ideal principles can be applied and fostered most appropriately among peoples who, for different reasons, are reluctant to respect the law of peoples. The “ideal theory” proceeds in two stages. In a first step, citizens of liberal democratic societies justify principles for a law of peoples. Here the “original position” functions as a citizens’ “device of representation” in the process of reaching agreement on principles on the national level. The original position – briefly said – provides that citizens’ representatives preside over principles of fairness that determine the “basic structure” of society, and that they do this with an impartiality deriving from the fact that they possess only limited knowledge of their prospective social positions. In a further step, the original position is applied to the international level. Here Rawls works the switches in two important ways. In contrast to the arrangement within a society, the subjects to the contract are representatives of peoples, not individuals. Moreover, the justification process does not take place among all peoples right from the beginning, but starts with democratic liberal peoples who interpret and agree on a law of peoples. Only in a second session are non-liberal, or “hierarchical,” peoples integrated.

What exactly characterizes liberal peoples? And what defines peoples as “decent hierarchical” ones? Liberal peoples have a constitutional government. They are internally united through culturally shared elements like language, historical consciousness, and political culture. Outside their borders they act “reasonably,” which means that they “offer fair terms of cooperation” (25) to other peoples – presuming that they are committed to those principles themselves. It is only after this first “meeting” under conditions of an original position that the results are presented to the so-called “decent hierarchical” societies, which do not add anything to the proposed law but simply comply with it. “Decent hierarchical” societies differ from liberal societies in several respects, some of which are leveled out in the commentary Rawls provides to such an extent that they are blurred. A major difference is that governmental institutions in decent hierarchical societies may be influenced by “comprehensive doctrines,” such as religious faiths (64). But Rawls also declares that religious freedom has to be secured to a certain degree, and that judges should be independent and reliable. A further difference is that political equality only exists in a formal sense: it does not apply to individuals but to groups or castes represented in the political and legal systems. Accordingly, political participation is not based on the one-citizen-one-vote principle. It is surprising to read that citizens should nevertheless play a “substantial political role in making decisions.” It remains, moreover, unclear what that means in a context where free elections are not allowed.

The outcome of the first session – the one among liberal societies – is surprising. The peoples agree on principles that should first and foremost ensure peoples’ sovereignty, which actually comes close to state sovereignty (37). They have to respect other peoples’ freedoms – to observe treaties, defend themselves, and be bound by the rules of war. The more interesting part is that they also have to honor human rights. Rawls does not at first specify which human rights the liberal peoples agree on, but he later specifies which human rights decent hierarchical peoples would uphold, and one can assume that they are identical with the human rights in the law of peoples. Among them are the most “basic” human rights, such as the rights to life and liberty, to a “sufficient” measure of freedom of conscience, to property, and to formal equality (65). A new principle on the list is a “duty to assist other peoples under unfavorable conditions.” Rawls’s list is considerably shorter, however, than the 1948 Universal Declaration of Human Rights.

The first question that immediately comes to mind is why Rawls constructed the domestic and transnational original positions differently. One wonders why, instead of reasonable citizens, one finds in the latter a similar ideal-typical construction of peoples. The justification of the principles of a law of peoples has been delegated to peoples’ representatives who only know what citizens share, such as a common language or history and a democratic political order. They are not expected to know that citizens have different conceptions of the good (40). So, first of all, Rawls ignores the important social condition of “pluralism” that he had stressed in Political Liberalism. This is surprising because, by neglecting the “fact of pluralism,” he drops the rather “realistic” assumption of pluralism as a basis for justifying transnational principles.

There is a further problem with his arrangement: the legitimation of the principles of law would be much stronger if it were the individual citizens – or their representatives – who agreed on the law of peoples. For then the justification of the law of peoples would rest on one of the basic principles of democracies, namely the principle of political self-determination. According to this principle, citizens need only submit to rules over the establishment of which they had some procedural influence. It is not understandable why citizens of liberal democracies would choose to do otherwise.

In any event, there must be a reason why Rawls thinks that they would prefer the minimal version. In the earlier edition of The Law of Peoples, he offered a fairly clear answer: He fears an ethnocentric bias should the individual be the main element in the construction of the original position. And surely, even though the first round only takes place among liberal democracies, the results could only occur with a view to a second round in which the hierarchical societies join in. But doubts remain. First of all, one may wonder whether by overcoming the “western” bias he nevertheless presupposes too much. Just think about the list of human rights that, even though reduced, still demands respect for individual freedoms and political participation rights.

But let us accept the presumption that the justification of the law of peoples begins with peoples and not individuals. Another, more pressing objection arises when one focuses on the two-fold application of the original position. If the representatives of liberal democracies have to adopt the perspective of the decent hierarchical societies anyway, then why are there two rounds at all, and not just one in which all societies are represented at the same time? On the other hand, if one really wants to learn what kind of laws liberal and hierarchical societies would agree on separately, then one could reasonably expect a different outcome. One can imagine that representatives of the liberal democratic peoples would promote a law of peoples that complies with principles they have already accepted on the national level. There is no reason why they should not prefer a law that includes a transnational difference principle, a human rights charter that embraces political rights and guarantees a more far-reaching right to free speech, as well as a notion of equality that demands that everybody be treated equally – to mention only some of the rights one already finds in the UN Charter.

The more serious problem is that Rawls misses the opportunity to set up a law of peoples that can function as a normative ideal, at least among liberal peoples. This goes together with a further incoherence between the two levels of original positions that also affects the notion of “reasonableness.” Whereas the citizens in the first original position are reasonable in the sense that they accept limits on reaching a consensus through public reason, the representatives of the peoples are not equipped with an awareness of this limitation. This affects their method of justification as well as the content of the law of peoples.

If persons act reasonably, this leads them to recognize that some of their reasonable presumptions are not acceptable to others. And this is the case not because they are not committed to being reasonable, but, on the contrary, because they judge reasonably. In liberal democratic societies, normative conflicts can occur that cannot be solved through public reasoning. There are good reasons for two or more alternatives. Abortion is one example: does equal respect for “human life” have priority over a woman’s right to self-determination? Another example is the use of embryos for research purposes. Here one of the questions is whether the potential to cure people of serious illnesses takes precedence over the right to life. In both cases, good reasons cancel one another out and one cannot come to a reasonable but only to a temporarily justified agreement. That is an effect of what Rawls calls the “burdens of judgment.” The upshot is that despite their disagreements, the parties involved accept that their reasonable judgments are not necessarily acceptable to others. It is part of being reasonable that persons can understand that the incommensurability of ideas of the good life is not reason enough to force others to adopt certain ethical convictions.

On the international level, however, “reasonableness” acquires another meaning. The realm of reasonable disagreement expands dramatically, encompassing parties who do not share the same notion of equality or a similar idea of political participation or free speech. Rawls can characterize his proposal as reasonably just because he dismisses specific aspects of the liberal idea of justice, such as a substantive notion of equality, which does not appear on his agenda at all. So “reasonable pluralism” has turned into a kind of plain pluralism, which is no longer based on a consensus between reasonable citizens or peoples. The agreement now takes on the shape of a modus vivendi, an agreement that is the result of bargaining processes and a balance of interests. The advantage of a modus vivendi is that one can reach an agreement relatively easily, even among very diverse parties; the disadvantage is that such agreements depend on current power constellations and thus can also fall apart easily. Rawls, however, has to pay a high price for circumventing the criticism of ethnocentrism. He gives up a position that allows for a critical evaluation of international treaties and a law of peoples. He is left without any arguments against those who would deny, for example, an enlargement of the proposed law of peoples.

Let us return briefly to those issues that challenge international law today: first, growing world-wide poverty and the role that international law could play in combating it. In 1998, some 1.214 billion people lived on less than one dollar a day, while a further 2.8 billion had a daily income of no more than two dollars. The “duty to assist other peoples” – an important step towards global justice – recently integrated into his law of peoples shows that Rawls is aware of this. The nature and subjects of the assistance are precisely described (106ff): those societies should be assisted that do not pursue an aggressive foreign policy and that lack the material and technological means as well as the political culture to become members of the community of well-ordered societies. The aim of the material and technical assistance is to install just and democratic institutions, which might include educational programs as well as population policies. Rawls is not talking about a principle of distributive justice along the lines of the difference principle. The assistance for societies that are on their way to becoming democratic but are too poor will end as soon as a more or less stable order has been established, and it will probably also dry up if it turns out that a “democratization” is not to be expected.

If this principle were realized, it would require that the affluent states increase their development aid and set up food and educational programs. In the long run, this would lead to an improvement in the living situation of a huge number of people. But due to the narrowly defined assistance targets, only some of the needy peoples would benefit. In societies aspiring to liberalism and democracy, the “right to assistance” would indeed lead to the institutionalization of a guarantee of a fair value of political rights, equality of opportunity, and a difference principle. The effect would be different, however, for poorer societies that only aim at developing a hierarchical social structure and not at becoming a democracy. The institutionalization of a decent standard of living for all citizens is a necessary condition for democracy but – at least according to Rawls – not intrinsic to hierarchical societies, and hence is not the goal of the targeted assistance. So the duty to assist would improve the living standard of peoples to very different extents. That would not be a problem in itself, but it becomes a problem since one can imagine that the support would probably not be sufficient to satisfy the needs of the citizens.

This would be especially true for the populations of so-called “outlaw regimes,” which are excluded from any support since the ruling elite is not interested in establishing a fair or decent social structure. This kind of restriction seems to be either unfair – in most cases citizens of a dictatorial regime suffer but are too exhausted or oppressed to change the situation – or cynical, if tolerating the population’s suffering is meant to worsen the economic situation in the long run so as bring about a regime change.

Why doesn’t Rawls argue for a principle of global distributive justice regulating economic and social injustice among peoples (119ff)? First, he claims that a “global egalitarian principle” lacks a clear target. It focuses only on the somehow vague idea of distributing natural and economic resources as equally as possible. His principle of assistance, in contrast, aims at promoting just domestic institutions in the world’s poorest societies. Secondly, according to Rawls, such a principle does not have a determined cut-off point that defines when the distribution has reached a satisfactory level. As a consequence, this kind of distribution creates new injustices because poorer societies could claim revenue sharing from richer societies, even if they had the same starting conditions but failed to establish a fair and efficient institutional structure and ended up in a worse position (117).

Both objections are unconvincing. To begin with the latter, the argument that a global difference principle would not take into account the peoples’ collective effort can and has been used as an objection to the domestic difference principle. If any wealth increase for the better-off need also lead to an increase in the numbers of the worse-off, there is no room left to take into consideration individual responsibility for individually invested effort. It is not fair, so the argument goes, if under the same starting conditions those who have not worked as hard as others receive the same level of support. If this is convincing, then Rawls has to drop the difference principle for the domestic case as well. But there is at least one approach that escapes this line of criticism. Fair distribution depends on a much more differentiated idea of what is necessary to level the playing field. Accordingly, one must distinguish among the different typical starting conditions that have an impact on the capacity to engage in a certain activity, and distribute resources according to the effort that can be expected with regard to the typical conditions one has been socialized to accept. It would be worth working out in more detail how this idea of equal opportunity could be made applicable to the international level. The other objection, that a global egalitarian principle lacks a clear target, is only true for those proposals that do not assign a purpose to the equal distribution of resources. Several approaches, however, have one or more specific targets, such as the reduction of poverty, the development of capacities, the satisfaction of basic needs, or the appropriate use of rights, to mention only some.

But a just international order indeed calls for more than a fair distribution of resources. It requires rules that regulate international financial and cultural exchange, trade, and working conditions in a fair way. A lack of such rules has a serious impact on the living situation within a society. Examples include rules over the extension of credit, copyright on intellectual property, and the disposition of a country’s natural resources and labor standards. Rawls is thoroughly aware of these problems and suggests a new principle of fair trade among liberal societies (42). Nevertheless, this principle is first reduced to anti-cartel agreements and leaves out other aspects. Second, Rawls assumes that those measures will never win the support of hierarchical societies. The same seems to be true for any kind of a human right to welfare. Can one expect a reasonable agreement on international welfare rights and principles of fair financial, cultural, and economic exchange? The answer is probably yes, but one has to ask which of the participating parties would agree on which principle. Especially from the point of view of hierarchical societies, it makes sense to favor international conventions on welfare rights and also to agree on rules that prohibit any kind of forced contract. That means those contracts are forbidden where an economically stronger negotiating partner uses his advantage to force upon another party a contract or treaty that he would not sign himself. The liberal democratic societies might be more reluctant to agree on such rules since they can calculate in the original position that, as liberal societies, they are probably also wealthy. But they cannot be against global rules on welfare and fair financial and economic trade without betraying their own moral standards. A problem could occur with rules on working standards. Hierarchical societies would not agree on a treaty on working conditions for fear of losing their competitive advantage. It has been the developing countries that have opposed a social clause in the WTO treaties that France and the US have been proposing since 1993. But here again, this argumentative deadlock can only be resolved if the reasonableness of the parties in the original position has been transformed into rationality, if the lifting of the “veil of ignorance” has dispersed an impartial point of view.

Last but not least, the question arises as to what Rawls’s answer would be to fundamental human rights violations such as those the world witnessed recently in Bosnia and Kosovo. And, furthermore, one might be interested to know if, with his law of peoples at hand, one could formulate an appropriate response to the terrorist attacks of September 11. It may be surprising, but here it turns out that Rawls’s proposal adheres to a strikingly old-fashioned legal paradigm. This is because it is only peoples and not individuals that figure as subjects of international law. It is only in a footnote that Rawls addresses the question of whether it is legitimate to penetrate the borders of a so-called “outlaw state” to hinder further violation of fundamental human rights (93, n6). He admits that there are cases where disrespect of a state’s sovereignty is legitimate, but he does not specify what these cases are. One can say that a change in international law took place already in 1945 with the London Agreement and Charter, an influential reinterpretation of the 1928 Kellog-Briand Pact. This Agreement defined the well-known jurisdiction of the Nuremberg Tribunal, a landmark on the way to the establishment of a concept of international crime for which “individual responsibility” becomes important. This was a major step beyond customary international law, which focuses on the duties and rights of states. The Nuremberg tribunals, as well as the Tokyo military war crime tribunals and more recently the ad hoc tribunals for Bosnia-Herzegovina and Rwanda, set the stage for further measures to establish a permanent court. Whereas the International Court of Justice (ICJ) in The Hague only handles cases between states, the planned International Criminal Court (ICC) will deal with crimes committed by individuals. The Rome Statute for the ICC of July 17, 1998, which has been adopted by more than 120 countries, defines three categories of crimes as “core crimes”: genocide, crimes against humanity (including crimes committed during peacetime), and war crimes. Meanwhile 40 states have ratified the Rome Statute and deposited the instrument of ratification with the UN Secretary General. The Statute will come into force when it has been ratified by at least 60 states, a goal that is expected to be met during the next year. The US, however, plays an awkward role. Together with Israel, it ratified the Rome Statute literally minutes before the treaty closed for signing on 31 December 2000, but one cannot really expect that the current government will complete ratification; it fears that the Court would have so many competences that it might minimize the US’s own status as a leading power.

One cannot find legal principles that focus on individual war crimes committed by persons, however, on Rawls’s agenda. This is all the more surprising as, if one goes through his Collected Papers, one is impressed by how important the individual is to him. His interest in the effects social institutions have on persons, their moral development, and their well-being in pluralistic societies is so obvious that one wonders why it has been obscured in The Law of Peoples. Is it a tribute to maintaining a “down-to earth” vision of a law of peoples? Or maybe this shift to peoples is caused by his decision right at the beginning of the study not to focus on the individual to avoid criticism from those who link individuality with the western Enlightenment tradition. In this respect, his proposal falls behind developments that have taken place in international law during the last decades, and at least since 1992, when the ad hoc tribunal of Bosnia-Herzegovina was set up.

This problem also arises if one asks what – according to Rawls’s law of peoples – would be an appropriate response to the terror attacks of September 11. It is easier to say which way one cannot go with Rawls. The international community could not identify the attacks as a crime against humanity, even though crimes against humanity are by definition systematic crimes committed by individuals against civilians. Within the framework of Rawls’s law of peoples, the attacks could only be defined as “armed attacks” directed from abroad, so that the right to individual and collective self-defense could be applied (as defined in Article 51 of the UN Charter). If one pursues this route, the first task is to categorize the terror attack as an armed attack conducted by a state, and the second, to offer clear proof of the involvement of the targeted states – whereas in the case in question, the existing law would have to undergo extensive reinterpretation in order to do justice to the situation. The other route is already open: it would offer a legal path to prosecuting international criminals, including by military means. One of the aims of dealing with such an international crime would be to bring the perpetrators to the International Criminal Court. Rawls’s proposed law of peoples could only confirm the current legal measures even though other legal paths already appear as realistic options. A utopian perspective, however, no longer seems to be an option at all.