The Dynamic Model of Choice for Public Policy Reconsidered: A Formal Analysis With an Application to US Budget Data

Abstract

It has been shown empirically across countries and political systems, and for different levels of government, that the distribution of budget changes follows a non-Gaussian distribution, a power function. This implies that budgets are very stable, yet occasionally are punctuated by very large changes. To explain this strong empirical generalization, Jones and Baumgartner (2005a) developed the Dynamic Model of Choice for Public Policy, which today is the dominant explanation of stability and change in public budgets. Based on formal analysis, this article investigates the implications and scope conditions of this model. Furthermore, using US budget data, the article reveals aspects of the model that do not closely fit the empirical pattern. The article concludes with an examination of three model revisions that may improve the fit between the model and the empirical distributions of budget decisions.