Structural Estimation of Labor Adjustment Costs With Temporally Disaggregated Data

Abstract

I estimate a dynamic optimization model of labor adjustment of establishments based on data that permit: specifying any desired adjustment frequency; estimating the model based on net and on gross employment flows; and allowing for simultaneous hirings and separations. The unit of observation is an establishment. Results for adjustment costs depend crucially on the model specification. Only a monthly adjustment model yields cost parameters in a reasonable range, while estimates from quarterly and annual adjustment models imply negative or excessive adjustment costs. Estimating the model on net employment changes implies hiring and separation costs of four annual median salaries, while the model on gross changes implies costs on the order of 1.7 annual median salaries. (JEL C25, D22, J23)

Labor Demand on a Tight Leash

“…The reason is that traditional models of labor demand assume that employers adjust input factors at no cost (Addison, Portugal, and Varejão, 2014). However, given the specific nature of the labor input, cost of adjusting labor are substantial (Oi, 1962;Muehlemann and Pfeifer, 2016;Yaman, 2019). Importantly, higher labor market tightness amplifies search frictions for firms that render hiring more costly (Muehlemann and Strupler Leiser, 2018).…”