Coordination in the Absence of Sovereign Intervention

entities and it is somewhat difficult to speak of preferences of the collective or, in this case, the state. The state cannot have preferences, and within it different and even conflicting preferences may be present. However, at least in economics and political science, we are used to working with collective actorsthe firm, the party, the church, and many others in our theories. It is reasonable to assume that most of these actors have some specific interest they pursue: for example, to maximise profit, votes, or some collective good. In classical public-choice theory we find either the idea that there are "collective interests" (OLSON [ 1965]) or the idea that the representatives' personal interests are tied to the organisations' interest in a way so that the representative pursues what is in the interest of the collective (DOWNS [ 1957], NISKANEN [ 1971 ]). Some rational-choice theorists say that the rationality assumption might even hold better for collective actors than for individuals (ZINTL [ 1994]). However, there is in fact still a. problem of the internal aggregation of preferences in every organisation to which I tum below. Second, there are surely inequalities between states in many respects. l cannot see, however, why these inequalities should be viewed as greater than those between individuals. How could we measure and compare such differences? Apart from that, I do not understand why inequalities pose a problem for analysis in a lawand-economics fashion. Rational-choice models can deal with inequalities among their actors. These asymmetries can explain different preferences of states and thus different goals in treaty bargaining. What is the problem of the approach? Third, and similarly, why should it be impossible to analyse also nonstate actors in a law-and-economics fashion? Why should we be restricted to states? Whenever nonstate actors play a significant role in treaty preparation, we can incorporate