Second, our data set is quite large and contains extensive information about both borrowing firms and the banks from which they obtain financing. The richness of our data allows us to analyze the robustness of our findings on the association between credit availability and market power. In particular we analyze the robustness of our results against a variety of different measures of firm-level credit constraints found in the literature. In our analysis we are able to deploy a dynamic panel approach that accounts for potential endogeneity in the data, using standard measures of credit constraints and, as an alternative, a disequilibrium methodology found in some recent papers that estimates excess demand for external funds (Ogawa and Suzuki, 2000, Atanasova and Wilson,2004 and Shikimi, 2005).