Worker flows, reallocation dynamics, and firm productivity: new evidence from longitudinal matched employer–employee data

Abstract This article investigates the impact of the worker flows of a firm on productivity by using unique longitudinal matched employer–employee data. The analysis has split a firm’s total worker flows into three components: workers’ replacements (excess worker flows), hirings introduced to increase the firm’s employment level (net hirings), and separations of workers intended to decrease the firm’s workforce (net separations). This has allowed the impact of workers’ replacements, which represent the most prominent and compelling feature of worker mobility, to be isolated from the other two components. Endogeneity has been dealt with by using a modified version of Ackerberg et al.'s (2015, Econometrica, 83(6), 2411–2451) control function method, which explicitly accounts for firm-fixed effects. The main findings are that (i) excess flows have an inverted U-shape impact on productivity, (ii) net hirings foster firm productivity, and (iii) net separations damage it. The impacts are heterogeneous and vary widely on the basis of the types of replacements, the categories of workers involved, and the types of firms experiencing such flows. Overall, the findings of this article highlight the importance of reallocation dynamics to obtain better employer–employee matches, and call for a reconsideration of policies concerning the flexibility of the labor market.