NBER Publications by Jonathan Willis

This paper studies hours, employment, vacancies and unemployment at micro and macro levels. It is built around a set of facts concerning the variability of unemployment and vacancies in the aggregate and, at the establishment level, the distribution of net employment growth and the comovement of hours and employment growth. A search model with frictions in hiring and firing is used as a framework to understand these observations. Notable features of this search model include non-convex costs of posting vacancies, establishment level profitability shocks and a contracting framework that determines the response of hours and wages to shocks. The search friction creates an endogenous, cyclical adjustment cost. We specify and estimate the parameters of the search model using simulated method of...

This paper studies the dynamics of labor demand at the plant and aggregate levels. The correlation of hours and employment growth is negative at the plant level and positive in aggregate time series. Further, hours and employment growth are about equally volatile at the plant level while hours growth is much less volatile than employment growth in the aggregate data. Given these differences, we specify and estimate the parameters of a plant-level dynamic optimization problem using simulated method of moments to match plant-level observations. Our findings indicate that non-convex adjustment costs are critical for explaining plant-level moments on hours and employment. Aggregation generates time series implications which are broadly consistent with observation. Further, we find that a model...

We study labor adjustment costs. We specify a dynamic optimization problem at the plant-level, allowing for both convex and non-convex adjustment costs. We estimate the parameters of the adjustment process using an indirect inference procedure in which simulated moments are matched with data moments. For this study we use estimates of reduced-form adjustment functions obtained by the gap methodology' reported in Caballero-Engel as data moments. Contrary to evidence at the micro level in support of non-convex adjustment costs, our findings indicate that piecewise quadratic adjustment costs are sufficient to match these aggregate moments.