You are hereConferences / Past Conferences / How Access to Firm-Level Data Informs Public Policy / Program / Thursday Session 2 / Finance and Misallocation: Evidence from Plant-level Data
Finance and Misallocation: Evidence from Plant-level Data
We study a model of industry dynamics in which idiosyncratic risk is uninsurable and establishments are subject to a nancing constraint. We ask: does the model, when parameterized to match salient characteristics of plant-level data (Colombia and South Korea), predict large aggregate TFP losses from misallocation of factors across productive units? Our answer is: no. We estimate financing frictions that are fairly large: one-half of the establishments in both countries are constrained and face an external finance premium of 5% on average. Efficient establishments are, nonetheless, able to accumulate internal funds and quickly grow out of their borrowing constraints. Parameterizations of the model that hinder this process of internal accumulation can, in principle, cause very large TFP losses. Such parameterizations are, however, at odds with important features of plant-level data, most notably the difference in returns to factors across establishments that expand/contract (young vs. old) and the variability and persistence of plant-level sales.