In this essay, we outline some reasons why one should be cautious about grounding a theory of growth on institutions. We emphasize how very different institutional structures have often been found to be reasonable substitutes for each other, both in dissimilar, as well as similar, contexts. The historical record, therefore, does not seem to support the notion that any particular institution, narrowly defined, is indispensable for growth. Moreover, we discuss how the evidence that there are systematic patterns to the ways institutions evolve undercuts the idea that exogenous change in institutions is what powers growth. Institutions matter, but our thinking about how they matter should recognize that they are profoundly influenced by the political and economic environment, and that if any aspect of institutions is crucial for growth, it is that institutions change over time as circumstances change.