with regard to welfare-theoretic criteria, whether of the Paretian or Kaldorian variety. As will be shown in the following chapters, different modes of interaction will indeed differ significantly in both of these dimensions, and it is also possible to specify the conditions that would increase the chances that a particular mode of interaction will in fact produce outcomes that increase aggregate welfare and that reflect criteria of distributive justice.
In doing so, however, we must keep two things in mind: First, actor constellations differ greatly in the degree of difficulty that must be overcome in order to arrive at normatively acceptable outcomes. Games of pure coordination are easy to resolve in satisfactory fashion through any mode of coordination. By contrast, in a Chicken game there are much greater obstacles that stand in the way of a satisfactory resolution, and the difficulties would again increase under conditions corresponding to the asymmetrical conflict game of Figure 4.5. Thus when considering the implications of a particular mode of interaction, the actor constellations that it is required to deal with, and their characteristic difficulties, must be considered as well. As long as the European Community, for instance, was primarily engaged in realizing the common interest of member states in gaining access to a common market, the fact that its mode of interaction depended on the unanimous agreement among member-state governments was relatively innocuous -- but it becomes deficient when issues requiring the resolution of major conflicts of interest among member states become more prevalent.
The second reminder concerns the fact that we will always be dealing with interactions among policy actors rather than with the underlying policy problems. Thus when we are comparing the welfare efficiency and distributive justice produced by different modes of interaction, we always should remain aware of the need to retranslate these outcomes into the terms of the original policy problems. With this in mind, in the following chapters I will examine the specific implications of four different modes of interaction -- unilateral action, negotiated agreement, majoritarian voting, and hierarchical direction.