Which workers suffer (or benefit) from firm-level uncertainty shocks?

Abstract
An established theoretical literature argues that uncertainty shocks are important determinants of firm decisions and aggregate fluctuation. However, few studies estimate the causal effects of these shocks on firm level employment dynamics, usually focusing on samples of public firms, with limited information on the firms’ employment decisions. In this paper, we shed light on the interplay between firm-level uncertainty shocks and heterogeneous employment decisions using a matched workers-firms dataset of the population of Swedish firms for 1997-2017, with more than 18 million worker-level observations. Using shrinkage methods on commodity prices, we construct a firm-specific expected profits shock, and the associated second moment uncertainty index. We use them to measure the causal effect of exogenous changes in uncertainty on the employment decisions of firms. We find that negative first moment profitability shocks significantly reduce hiring and increase firing uniformly across different categories of workers. Conversely, uncertainty shocks change little the overall employment levels, but have significant and quantitatively important compositional effects. In particular, when uncertainty increases, firms are less likely to fire younger and short-tenured workers, are more likely to hire workers with previous experience in the same sector, and are more likely to both hire and fire more skilled workers, relative to normal times. The results emphasize the importance of workers heterogeneity, especially along the flexibility dimension, in driving the effects of uncertainty shocks on employment dynamics. Additional robustness checks confirm this interpretation and highlight the role of disruptive high uncertainty episodes as creative destruction periods for the labour force. Joint paper with Andrea Caggese (Universitat Pompeu Fabra & CREI) and Daniel Metzger (Rotterdam School of Management).