Timo Laine, 15 September 2004
In this essay I will consider the prospects of extending traditional liberalism to international relations. The ultimate aim of such a project is an international law that can be justified with liberal arguments. I will take John Rawls's work The Law of Peoples as my starting point, because it is a major contribution to the topic. After explaining Rawls's theory, I will make an attempt of assessing its plausibility by reviewing some of the criticism it has attracted.
A theory can be challenged in at least two ways: either point-by-point by criticizing some of its specific aspects, or by presenting an alternative theory that aims to give better answers to similar questions. There are alternative theories of global liberalism as well. I will try to present a summary of a prominent "cosmopolitan" alternative to the Rawlsian approach.
Rawls extends his Theory of Justice with The Law of Peoples. In the earlier work he was concerned with domestic justice, and the later one with international (or rather "interpopular", as we shall see) relations. Many of the conclusions of the theory of justice are taken for granted, and go virtually unquestioned: whatever the principles of foreign policy may turn out to be, the liberal domestic policies stay more or less the same. What we are interested in are how liberal peoples, and all other decent peoples, should conduct themselves towards other members of the "Society of Peoples". In the end, Rawls claims that the proper principles to guide this conduct require respect of the freedom and equality of other peoples, observance of treaties, observance of human rights, nonaggression except in self-defense, certain restrictions in the conduct of war, and assistance of peoples not living under a decent regime (Rawls 1999b, 37). I will give a summary of the reasoning that leads to these principles.
According to Rawls, we can choose our principles of justice ourselves. We just need the right, "fair" method to do that. The method for choosing these principles for international relations is very similar to the method of choosing such principles for the liberal society: hypothetical representatives are led to what is called the original position, behind a veil of ignorance (Rawls 1999a, 11). But unlike the domestic case, in the global case those represented are not individuals but peoples. Thus, it may be more correct to talk about interpopular rather than international relations. (I will talk more about what Rawls means by "peoples" in 3.1.)
In the domestic case, the veil of ignorance is meant to separate you from all information that is not morally relevant (Rawls 1999a, 118-119). This means that you do not know your age, wealthiness, religious beliefs, or anything like that. You only have general, more or less objective knowledge on how the world works. You do not know your conception of the good, but we need some kind of an idea of what is good to be able to decide how the good things should be distributed. Rawls gives us a thin theory of the good, which is roughly speaking a list of goods it is rational to want, if you want anything at all (Rawls 1999a, 348-349). This list of primary goods may be taken to include social goods like rights, liberties, wealth and other such things, in addition to natural goods, like health, intelligence and other such things (Rawls 1999a, 54). Equipped with all this information, a representative would choose principles that require equal and as extensive liberties as possible for everyone, and that social and economic inequalities are to the benefit to the least advantaged and attached to positions open to all, with fair equality of opportunity (Rawls 1999a, 266).
Like I said, the method of choosing global principles of justice is very similar to this. Again, the hypothetical representatives are behind a veil of ignorance, but this time as representatives of peoples. A people has one, and only one, representative. A difference is that behind this second-level veil, more comprehensive conceptions of the good are not excluded: the representatives know the "fundamental interests" of their peoples, and these interests can and do vary (Rawls 1999b, 33). Another difference in the international original position is that it requires two "runs". The members of the society of peoples are different from each other mostly because of their different conceptions of justice. Rawls divides peoples into two categories, well-ordered peoples and peoples that are not well-ordered (Rawls 1999b, 4). The well-ordered peoples include both liberal and decent peoples. Liberal peoples follow the principles of justice very similar to the ones in the original theory of justice. Decent nonliberal societies fall short of the liberal standards, but satisfy two criteria (Rawls 1999b, 64-67, 88). First, they are peaceful, and do not have aggressive aims. Second, they respect human rights, have a "decent consultation hierarchy" (the elected bodies represent all the subgroups), and their officials operate in the honest belief that their law is guided by a "common good idea of justice". The three not well-ordered kinds of peoples are outlaw states (I am not sure if Rawls believed in the possibility of an outlaw people), societies burdened by unfavorable conditions (cannot peoples be burdened by such conditions just as well as societies?), and benevolent absolutisms. The international original position has representatives from both kinds of well-ordered peoples, which is why it requires two runs, one for each. Rawls starts with the liberal run.
Why does Rawls not want the representatives to ignore their peoples' fundamental interests? According to Charles Beitz, this decision is related to a larger question: Rawls does not actually even want to find principles for international institutions, but principles to guide the foreign policy of liberal societies (Beitz 2000, 675). This policy is meant to be reasonable enough to be accepted by societies that are not liberal, because Rawls is prepared to accept a "reasonable" pluralism as an inescapable fact (Rawls 1999b, 31-32). He also sees toleration of nonliberal peoples in the society of peoples as analogous to toleration of comprehensive doctrines of the good in a liberal society (Rawls 1999b, 59-60), which of course implies that he believes the scope of many of the liberal arguments for toleration can be extended to international relations.
We can think of what would happen if the fundamental interests were excluded. The result would in all likelihood be that we would have more principles, or that the principles would be more demanding, or both. This would probably make them unacceptable for many (all?) decent nonliberal peoples, which is exactly what Rawls wants to avoid. He believes that even if we admit that we have different conceptions of the good, we can still find enough common ground to make peaceful coexistence possible. Also, this peace would not be accidental, but guided by principles acceptable to (and accepted by) all decent peoples, liberal or not. Still, he has something to say about the good as well. Liberal societies can use Rawls's thin theory, while hierarchical ones have their own, more comprehensive conceptions of the good.
After the liberal run, Rawls turns to the representatives of decent nonliberal peoples. The aim is to show that the principles would be acceptable to them as well, and that the liberals' belief that the decent peoples would accept them is not unreasonable. Using the two criteria that the decent peoples satisfy, Rawls argues that since they do not have aggressive aims, and want to protect their human rights and their good, there is no reason for them not to accept his principles. Indeed, they would have a reason to endorse those principles as "specifying fair terms of cooperation with other peoples". (Rawls 1999b, 68-69.)
This is not an exhaustive list of the things to criticize in Rawls's theory, and Rawls would probably be able to give reasonable answers to at least some of the criticism. Nevertheless, the theory has been criticized.
Perhaps the most significant question is why Rawls has chosen peoples as the entities to be represented behind the veil of ignorance, and what really is a people anyway. He could have chosen something else instead, like states or individuals. Rawls sees two important differences: first, a people's lack of sovereignty as it is traditionally attributed to a state, and a state's lack of a moral character Rawls attributes to a people. In addition to this, a people is united by "common sympathies" (a term borrowed from Mill) and a desire to be under the same democratic government. (Rawls 1999b, 23-27.)
State sovereignty means two different things. First, it means the "right to go to war in pursuit of state policies". Second, it means a "certain autonomy [...] in dealing with its own people". Rawls thinks both of these are wrong, and that these powers of sovereignty must be reformulated. Apparently he hopes that displacing states in this way as the primary agents of the theory also enables him to move from one kind of agency to another, that is from the egoistic state to the moral people. States and peoples are both rational agents, which means that they have reasons for their actions, and that the reasons in some way make them act. But a state and a people are moved by different reasons. A state can only be moved by the so-called "reasons of the state", which are defined as its "prudent or rational pursuit interests". According to Rawls, these reasons are not the right ones when it comes to justice. On the other hand, a people can be moved by the right reasons, that is moral reasons. The law of peoples is meant to provide these reasons.
Allen Buchanan argues that Rawls is still committed to talking about something closely related to states. He points out that the principles of the law of peoples presuppose most of the traditional powers of sovereignty. In addition to this, he also notes that Rawls appears to use the term "society" interchangeably with "people". This suggests that although Rawls's peoples are not sovereign in the traditional sense, they are organized: "Their organizing structures include independent, territorially based political institutions--what one would normally call states." So, peoples are "groups with their own states". (Buchanan 2000, 698.)
This interpretation seems to make sense. Rawls does not want to do away with states, but he believes that unlike peoples (or people), states they cannot be moved for the right reasons. Therefore, his solution is apparently to subordinate the state to the people, to preserve a moral character without losing much of traditional state sovereignty. Although we may call the principles of the law of peoples "interstate" principles as Buchanan does (Buchanan 2000, 699), it is still the case that states are only the peoples' means of internal organization and interaction with other peoples.
Rawls's decision to talk about peoples, while solving one problem, seems to leave us with a can of worms. Not all peoples have states, and what is more, the populations of actual states "are, in almost every case, collections of groups, and in some cases the groups are distinguished by quite different conceptions of justice or the good" (Buchanan 2000, 716). They do not necessarily have what Rawls expects, that is, a political unity. It is of course true that Rawls does not expect a very deep unity: specifically, he leaves out requirements of "a common language, history, and political culture, with a shared historical consciousness" as too demanding (Rawls 1999b, 24). Nevertheless, if we still assume that peoples are groups with their own states, we are left with the question of intra-state conflicts. The only answer Rawls is able to give to the question is that as long as human rights are respected, a people is justified in managing its own internal affairs. Buchanan maintains that this conclusion would require an actual argument, since the claim is controversial, and the challenge "formidable" (Buchanan 2000, 718).
What about human rights, then? They are something every liberal can support. But it is also important to understand that here the conception of such rights is somewhat limited: for example, both Rawls and Thomas Pogge demand only basic rights, and shun the more liberal aspirations (Rawls 1999b, 68; Pogge 1992, 49). But again, there seems to be a satisfactory consensus on what such rights are. The difference is rather with the role they play in different theories. Rawls includes them as a condition of decency of a society's domestic institutions, as both limiting the sovereignty of a people (the people must respect human rights), and securing it from intervention (as long as a people respects human rights, intervention is unjustified) (Rawls 1999b, 80). There seems to be a bit of a problem. It looks like Rawls is able to have peoples as the unit of his theory only because of a trick. He does not take up the task of showing that all decent peoples would choose to respect human rights, but simply considers such rights as so fundamental that they can be included in the very definition of a decent people. Why do this? It seems to be about right, but we cannot be satisfied with that. The definition plays a major role in setting the limits of toleration, and it requires more support. Are there reasons why we should not tolerate societies that violate human rights? Yes: "people are entitled to be treated in certain ways (partly) in virtue of their characteristics as human beings" (Beitz 2000, 686). But Rawls cannot use this argument in the second-level original position, because it is individualistic, and therefore unacceptable to at least some decent nonliberal peoples. This would not matter, because different peoples do not need to be in moral agreement on the argumentation behind the principles, but just on the principles themselves. Rawls would therefore have one option left. For every society he considers tolerable, he would have to present a moral argument for human rights that is reasonable to that society and compatible with its conception of the good. Rawls does not want to take up this difficult and maybe impossible task.
When we look at the list of Rawls's principles, we notice a curious omission: they say little about distributive justice. Yet many people consider global poverty a considerable injustice. What Rawls's principles do include is a duty of assistance. The well-ordered peoples have a duty to assist burdened societies, and only burdened societies. Such societies lack "the political and cultural traditions, the human capital and know-how, and, often, the material and technological resources needed to be well-ordered". However, they do not have aggressive aims. (Rawls 1999b, 106.)
A significant background assumption is that of self-sufficiency. Rawls believes that almost every society in the world has enough resources to become well-ordered. If a society is not well-ordered, the main reasons for this most likely lie in its traditions, and that "merely dispensing funds will not suffice to rectify basic political and social injustices", but an "emphasis on human rights" may have an effect on the conduct of negligent rulers. What the duty of assistance actually means is that the well-ordered peoples should provide enough aid to enable the burdened society to establish just or decent institutions, and once this goal has been achieved, the society will be able to take care of itself and no further aid is necessary. Rawls can conclude that the well-ordered peoples have done their duty, and that to do more is supererogatory. (Rawls 1999b, 106-109.)
Beitz and Buchanan voice their disagreement. Both attack the self-sufficiency assumption, and ask the question whether international trade might not play a role as well. Beitz is not even sure if the question is intelligible, as it may not be possible to "distinguish between domestic and international influences on a society's economic condition" (Beitz 2000, 690).
Buchanan draws our attention to the first pages of A Theory of Justice: Rawls wants to focus on "the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation"; this "basic structure of society is the primary subject of justice because its effects are so profound and present right from the start" (Rawls 1999a, 6-7). In the domestic society, this structure is "the political constitution and the principal economic and social arrangements". If the same kind of reasoning is appropriate on the global level, we should look at whether there is a global basic structure as well. Buchanan claims there is: "Its existence and major deatures are documented in a vast and growing interdisciplinary literature that goes under various headings: globalization, structural dependency, and theory of underdevelopment." But in The Law of Peoples, Rawls is not interested in talking about the global trade and issues related to it, although it may plausibly be said that in many cases, the decisions made by the biggest actors in the global economy can influence people's lives in a just as, if not more, profound way than the institutions of a single society ever could. This would suggest that there are problems with Rawls's self-sufficiency assumption. (Buchanan 2000, 704-706; Beitz 1999, 271.)
Rawls could still emphasize that a people does not need a lot of resources to become well-ordered. It may be true that with some work on the institutions, even a very poor society could survive in the global economy. Nonetheless, Beitz says that even if Rawls is correct, this does not necessarily help his case a lot (although to be fair, it does not help the case of his opponents either). What principles we choose are influenced by, but not determined by, empirical facts. And if it is true that each society is economically self-sufficient, this is an empirical fact. But such facts do not give us a single correct principle of justice, or a single set of such principles. What the empirical facts determine is just what kinds of principles it would be possible to choose, and how these principles are to be implemented in practice, that is, what kinds of policies they entail. What principles would actually be chosen would still remain to be determined. (Beitz 2000, 690-691.)
Another way to put Buchanan's point is that Rawls does recognize the existence of this global basic structure, and acknowledges its effects, but considers this structure as the "natural" one, and that the representatives would choose it by default. Thus he can say that any deviations from this structure require additional justification. Pogge offers this interpretation, and argues against it (Pogge 1994, 211-213). The global economy and its rules are a human artifact, and so there is "no prior distribution, no natural baseline or neutral way of arranging the economy". These rules are to be chosen behind the veil of ignorance, and it does not seem that Rawls's conditions are enough to make his case. Pogge made this objection to the earlier version of Rawls's work, but it does not seem that Rawls has answered it properly in the final version.
Rawls's theory is not the only one that aims to extend the scope of liberalism. Beitz calls Rawls's theory social liberalism, and mentions cosmopolitan liberalism as a prominent alternative. The main difference is that unlike the social variety, cosmopolitan liberalism sees the political world as consisting of persons, individuals, instead of states, peoples or other more or less state-like entities. The point is of course not to eliminate societies, but to demand that principles of justice have to be connected directly to the treatment of individual people, instead of that of societies. Like Rawls, the cosmopolitans take human rights to be the primary standard for assessing the justification of different institutions (Pogge 1992, 49). We saw that Rawls ran into problems with human rights, but it appears that in the cosmopolitan theories human rights fall into place very naturally. They are simply a practical elaboration of what it means to have principles that are primarily connected to individual people. (In another article (Beitz 1999, 280-286), he adds laissez-faire liberalism to the list. Robert Nozick and Hillel Steiner are some of the most important thinkers that have held this view. I cannot consider this view more thoroughly here.) (Beitz 2000, 677.)
There seems to be no cosmopolitan theory comparable to that of Rawls's. What we have is a group of thinkers that have things to say about international justice, and they have produced a diverse collection of articles and books. These are not always compatible with each other. If there is something deeply wrong with The Law of Peoples, we should be able to construct a theory that would not suffer from similar defects. Maybe the best cosmopolitan theory so far is actually Rawls's original theory of justice, as both Beitz and Pogge have made attempts to show that it can and should be applied to the whole world.
Beitz wants to show that if you find Rawls's original theory of justice plausible, the facts require you to interpret the theory in a cosmopolitan way (Beitz 1979, 129). This would mean that there is no need for an additional law of peoples, but the original principles would apply globally. To begin with, Beitz tries to show that even if we accept the the idea the world consists of self-sufficient nations, the international principles would still include an element of distributive justice. This is because natural resources are not distributed equally around the world, and that this distribution is morally arbitrary (Beitz 1979, 136-143). I will not go into this argument, because it is not relevant here; Beitz's main point is that nations are not self-sufficient. We need to recall the criticism of self-sufficiency: the states Rawls considers self-sufficient have in fact many kinds of relationships with each other, and these relationships are not irrelevant. States are connected to each other in a way that makes them dependent on each other. This suggests that the world is a very similar to the domestic society, which Rawls considers a "cooperative venture for mutual advantage"; and just like the domestic society, the global interdependency is also "marked by a conflict as well as by an identity of interests" (Rawls 1999a, 4). It seems that many of the arguments that require domestic principles of distributive justice support international principles as well.
The fact of interdependence does not help if there is no hope of any kind of moral unity. Even if we depend on each other economically, it may still be true that our states are still in a Hobbesian state of nature, and that there is no way out. Pogge tries to show that a "value-based" global order is possible. This requires that there are some (at least a few) important ultimate values shared by everyone. These values have to be "significantly embodied" in the institutions. (So far this is very similar to Rawls's argument, and his definition of a decent people.) The idea is that unlike in a state of nature, in a value-based order there are some nonnegotiable matters. Pogge says that to achieve this kind of order the parties have to be committed to it, they have to be able to identify some common values, and that they have to be able to modify their values to achieve a common goal. Like Rawls, Pogge believes that a degree of international pluralism is necessary, and that cultural imperialism is neither necessary nor appropriate in establishing a value-based order. We should not and cannot wish that everyone will become like us. (Pogge 1989, 228-231.)
If we now agree that the scope of Rawls's principles could and should be extended, we need to describe what this would mean in practice. Clearly there would be a stronger case for foreign aid: we would recognize that giving aid to the world's poor is not charity but justice (Beitz 1979, 172). This is perhaps not too different from Rawls's duty of assistance, but the cosmopolitans go further. Rawls does agree that the global economic institutions may generate inequalities in wealth, but as long as all peoples are able to have just or decent domestic institutions, the duty of assistance is fulfilled. But if we interpret Rawls in a cosmopolitan way, then just like in domestic society, economic inequalities must be to the benefit of the least advantaged. This has to be reflected in the institutional structure, which means that there is a need for institutional reform (Beitz 1979, 173-174). But it must also be noted that the domestic policies of a Western democracy are not safe from reform either: since there is just a single, global original position, the whole world has to be arranged according to those principles, from the ground up. Unlike in Rawls's theory, we don't have two standards of justice, domestic and global, but the only justice is global justice (Beitz 1979, 150). Our domestic institutions and all their policies have to be justifiable not just to ourselves, but to all humankind. In short, national sovereignty isn't what it used to be.
From this perspective, according to Pogge, the current global institutional scheme is not optimal, as it tends to engender a high incidence of human rights violations (Pogge 1992, 54). Some kind of reform is necessary, but what kind? Pogge traces the problem back to sovereignty, and says that the current concentration of sovereignty at one level is no longer defensible (Pogge 1992, 58). If we take human rights seriously, we need to give up the global order of sovereign states. Instead, in the scheme Pogge proposes people would be able to "govern themselves through a number of political units of various sizes, without any one political unit being dominant". Adopting this scheme would have several positive effects.
First of all, there is again the familiar idea that the sovereign state is necessarily a warlike entity. Without international rules and institutions, a military competition is inevitable. And being able to do virtually anything they want within their own territories, there are plenty of opportunities and incentives for states to develop more powerful weapons. The international institutions for preventing such arms races would not depend on the voluntary cooperation of every national government, precisely because the whole point of having them to begin with is to set the limits to the sovereignty of national governments. Secondly, sovereignty enables a state to control its population very effectively. Political opponents can be easily silenced, because there is nobody the government has to answer to. If we had more levels of political units, they would check and balance each other: if one got out of control, there would be better organized units above, below and on the same level, and they would be able even to fight the oppressors if necessary. Thirdly, there is arguably a growing need to have some kind of international rules of distributive justice, although a "global welfare bureaucracy" would of course be too much. Fourthly, many ecological issues can only be tackled with better cooperation and coordination than the system of sovereign states allows. (Pogge 1992, 61-64.)