Policy Cooperation Among Benevolent Governments May Be Undesirable

Abstract

This paper presents a simple counterexample to the belief that policy cooperation among benevolent governments is desirable. It also explains circumstances under which such counterexamples are possible and relates them to the literature on time inconsistency. Since the work of Hamada (1976), investigating the effects of increasing policy cooperation among countries has been a major topic in international economics. A standard conclusion of this work is that increasing policy cooperation among countries is desirable. In a seminal paper, Rogoff (1985) has challenged this view. Using a simple monetary model, Rogoff shows that cooperation among policy makers can lead to a lower level of welfare than noncooperation does. Rogoff's result has caused much consternation among those who advocate policy cooperation, and his work has been criticized along several dimensions. For example, some authors, including Canzoneri and Henderson (1988), have noted that a key assumption in Rogoff's model is that the objective function of each country's policy maker does not coincide with the objective function of its residents. Indeed, if in his model policy makers maximize the welfare of their country's residents, the counterexample is overturned and cooperation strictly dominates noncooperation. This feature leads some to interpret Rogoff's result as simply saying that if policy makers form a coalition against the private sector, they may be worse off than if they do not. Others, such as Neck and Dockner (1988), have claimed that Rogoff's result depends on private agents acting strategically. Under this interpretation, Rogoff's result is relevant to, say, economies with a large trade union, but not to economies with a large number of competitive private agents. In a somewhat different vein, Persson (1988) and, especially, Devereux (1986a,b) have questioned the significance of welfare comparisons across different institutional regimes in a model without a solid foundation for the behavioural relationships. This paper presents a simple model in which governments are benevolent, but cooperation is still undesirable. The model is a two-country version of Fischer's (1980) optimal tax model. In it, private agents are competitive (in that each agent takes both prices and government policies as uninfluenced by his actions) and each government maximizes the welfare of its country's residents. In the paper, the two different regimes-cooperative and noncooperative