Why do some countries emerging from autocratic rule feature competition between strong programmatic parties while others become preserves of clientelism? The present chapter contributes to an answer in several ways. First, it urges social scientists to think of clientelistic electoral competition not only in terms of clientelistic parties but also in terms of important non-party forms of political organization (party substitutes) that can constitute extreme manifestations of clientelistic linkage between voters and politicians in some new democracies but that are typically overlooked in studies that focus on parties alone. Second, it stresses that the strength of clientelistic politics can vary widely within a single state and that we can learn much by studying such variation, holding constant country-level variables. Third, it combines these approaches to test key elements of the general theory of clientelism developed by Kitschelt and Wilkinson in this volume's Introduction. Specifically, it takes advantage of a quasi-experimental opportunity presented by the Russian Federation, applying statistical analysis to an original database so as to understand why highly clientelistic provincial political machines are more powerful in some of Russia's eighty-nine regions than in others.

With minor exceptions, the results broadly support the volume's theoretical approach, indicating that on the whole the strongest degrees of regional clientelism are found where the attributes of economic development and political competition are lowest, where the state can most easily monitor and single out for punishment particular economic sectors, and where ethnocultural networks are politicized.