This is a book whose primary contribution is the elaboration and empirical verification of a new mechanism by which democratic institutional design affects one particular manifestation of bad government, namely, political corruption. As such, it is a work that is animated by the conviction that the development of a rigorous understanding of causal mechanisms ought to be a key goal of institutional analysis in the social sciences. Although this conviction is held by a large (and growing) community of scholars in the field of political science, it is far from universally held, and indeed much debate exists about the intellectual and practical payoff of placing mechanisms front and center in studies of the consequences of the structure of democratic institutional design.

In this chapter I present a general defense of the broader enterprise this book is engaged in: mechanism-based institutional analysis. My central claim is that if carefully conducted, analyses in this vein hold the promise of providing broad guidance to policy makers about the types of behavioral changes likely to be unleashed by institutional reforms and may perhaps even permit limited, qualitative statements about the magnitude of said changes. This is, of course, a very provisional type of guidance. Mechanism-based institutional analyses of the sort conducted in this book do not, in general, allow researchers to present policy makers with a point estimate that describes the amount by which a given institutional reform will change some aggregate outcome of interest in a particular polity.