There is a large and growing literature on the impact of institutional quality on economic performance and the broad consensus is that "good" institutions facilitate better economic performance. The literature that provides micro-level support for the policy discourse about institutional quality does not, however, account for significant intra-country variation in reactions of firms to changes in business environments, even within the same industry, and it generally ignores the possibility that the impact of institutional quality on firm performance may not be neutral. In this paper, we analyze the impact of institutions on firm performance using an approach that enables us to overcome these problems with the stylized approach. Using cross-country firm-level data, we demonstrate that not only does the marginal impact of institutional quality vary significantly within countries, but also that the impact is economically significant only at the two extremes of the distribution. We view this as prima facie evidence that policies that tinker with institutional quality on the basis of the popular wisdom about the impact of these institutions on the average firm may not have the desired or expected impact, at least at the micro level.