Policy change is important, and a very large number of actors and events contribute to a given policy change. The list of contributions includes government change, public opinion, public debate, lobbying, expert advice, events, and media coverage. Many of these contributing factors are interdependent, their effect on policy is variable, and their contributions are often difficult to observe. This resembles the features of a complex system, and it calls for an approach to the study of policy change that takes this complexity seriously. Understanding Policy Changes: The Mixture Model of Politics offers a new theoretical and methodological approach to the study of policy changes. The approach draws inspiration from research on physical systems such as earthquakes and waves, and it utilizes mathematics to improve the precision of the theoretical arguments and the empirical applications. Theoretically, the book develops a model of politics that explains policy changes as a function of variation in the number of contributions. Methodologically, the book presents the comparative distributional method as a statistical tool to analyse policy change. Using case studies, and data on public spending across a large number of countries and issues, the book demonstrates how the mixture model accounts for general patterns in the distributions of policy changes and how institutional differences, particularly democratic institutions, are associated with variation in these distributions.