Misallocation and markups: Evidence from Indian manufacturing

Abstract

This study examines the implications of variable markups for resource misallocation and for aggregate productivity. Using manufacturing data from India, I show that variable markups alone account for a small fraction of the dispersion in revenue productivity when allowing for other sources of misallocation. Meanwhile, variable markups are crucial to understanding the aggregate consequences of policies that reduce misallocation. When equalizing marginal products across firms in narrowly defined industries, my model generates a smaller total factor productivity gain than does a model that uses constant markups because more productive firms endogenously choose higher markups. Thus, the indirect costs of markups are higher than one might expect.