By Andrea Lanteri and Pamela Medina
This paper studies the role of capital specificity and investment irreversibility on the distribution of marginal products of capital and aggregate TFP. We use a methodology new to the misallocation literature, based on the study of “mobility” across quantiles of a distribution. In a panel of Peruvian firms, we show that persistent dispersion in marginal products is explained to an important extent by the persistence of low marginal products. That is, by unproductive firms that take a long time to downsize. Using a quantitative general-equilibrium model of firm dynamics with idiosyncratic shocks, calibrated to match key features of our data, we argue that the persistence of low marginal products suggests that irreversibility frictions are large. Moreover, it is inconsistent with theories of misallocation based only on financing constraints.
Misallocation is a big topic these days, but little is known about the underlying frictions. This paper makes progress on this front with what looks to me a case of the sunk cost fallacy. Attachment to capital that “still works” prevents significant productivity improvements.