![]() |
![]() |
![]() |
![]() |
Expected Modes of Policy Change in Comparative Institutional Settings WORKING PAPER Working Paper 97-08 with the Political Institutions and Public Choice (PIPC) program, Michigan State University. 1997. Abstract Elsewhere, we have argued (Hammond and Butler 1996) that merely focusing on institutional rules—as the multiple-veto-points argument does—does not completely specify the underlying situation. In particular, we argue that policy change is produced by an interaction of institutional rules, preference profiles, and the location of the initial status quo. As such, there are conditions under which two institutional systems should be expected to exhibit the same "mode" of policy change, whether it is incremental change, oscillating change, random change, or policy stability. While these conditions do explain why presidential systems sometimes undergo great policy changes and why parliamentary systems sometimes exhibit great policy stability, they do not explain the empirical regularity advanced in the literature. The purpose of this paper is to investigate the extent to which we should expect to see greater long-run stability in presidential systems than in parliamentary systems. We assume a basic unidimensional spatial model of legislative voting under several different institutional rules. Rather than focusing solely on the number of veto points, we also concern ourselves with preference profiles and the location of the initial status quo. Holding preference profiles constant across institutions, we can determine whether the institutions will produce similar or dissimilar policies. In our previous paper (Hammond and Butler 1996), we made one-period static comparisons and two-period dynamic comparisons. Our examination demonstrated that different systems can produce similar policies and exhibit similar modes of policy change. In this paper, we make multi-period dynamic comparisons among institutional systems. By constructing a common history of preference-profile changes and using these profiles in computer simulations, we approximate the frequency of each mode of policy change for each institutional system examined. While these simulation results demonstrate greater stability for presidential systems than for parliamentary systems in the long-run, they also show that different institutional systems not infrequently exhibit the same mode of policy change. This paper emphasizes the shortcomings of purely institutional-rule arguments while proposing an alternative argument that explains both the long-run empirical regularity of differences in policy stability as well as the occasional short-term similarities. Thus, our argument represents an advance over the existing argument since it explains everything that the existing argument explains but also explains things that the existing argument cannot. You may download this paper here. |