The shift in the interest of international agencies and of the World Bank, in particular in the 1990s towards issues of governance, is on the face of it long overdue. It appears that after a long hiatus, mainstream economists are returning to the political economy of growth in general and to the problems of policy implementation in particular. Evidently, the new consensus is to be welcomed as a huge improvement over the market theology of the 1980s. The state is now recognized as important, as are investments, and how effectively states and other non-state institutions, like financial institutions, work to create conditions conducive for investment must clearly have a l ot to do with explaining growth and in directing policy attention when growth is poor. On the other hand, the new consensus is resistant to taking into account the role of political power in general, and in particular in explaining the patterns of corruption in different countries and the effects of this corruption. This is not because the importance of political corruption is not recognized. Economic theorists working on corruption as well as the Bank and IMF, have explicitly recognized the importance of political corruption. Rather, the problem is that political corruption is difficult to model and the policy implications of identifying this problem is that politics has to be targeted, a prospect that is not attractive to international institutions unless the political reform can be presented in terms familiar to their Western funding constituency. This is why when the political underpinning of corruption has been discussed, the policy suggestions have been things like greater democratization or the encouragement of civil society participation in monitoring the state. Typically, economists and international agencies have shied away from trying to identify the classes and groups involved in corruption in different contexts, the consequences of their corruption in each case and the possible ways in which the political structure can be changed to change the magnitude or effects of corruption. We will argue that although economic models have made a contribution to our understanding of corruption, by leaving out the political determinants of corruption, these models are not just deficient but in many cases may be misleading. The required direction of reform may often be quite different from that suggested by the emerging models on corruption.This chapter is structured in the following way. In the first section we discuss definitions of corruption and some of the evidence which has underpinned the recent interest in the phenomenon. In the next section we examine a number of contemporary approaches to corruption, focusing on the Shleifer-Vishny (1993) model of the effects of corruption under different institutional arrangements, as an example of the direction of research based on game theory models. Finally, in the last section we look at the importance of class and group structure in determining incentives for different types of corruption, which we ...