People I Know: Workplace Networks and Job Search Outcomes

Abstract

We examine the role of information networks in job-search outcomes of displaced individuals. We draw on longitudinal Social Security records covering the universe of worker-firm matches in a tight labor market in Northern Italy. Unlike previous research, we focus on workplace networks whose labor market attributes we are able to describe extensively. A workplace network is defined as all coworkers a displaced individual worked with prior to displacement. Estimates of network effects are thus affected by omitted variable bias if the labor market sorts workers across firms along relevant determinants of search outcomes and network characteristics or if past coworkers are exposed to the same shocks. The empirical strategy accounts for these possibilities by comparing subsequent outcomes of workers displaced by the same firm; in addition, we exploit the longitudinal dimension to develop controls for potential residual within-firm heterogeneity. In particular, we control for pre-displacement wages and employment status as well as descriptions of pre-displacement firms and their workforce. Contacts' labor market attributes have a significant effect on a variety of job search outcomes. Employed contacts significantly increase the probability of re-employment. They are more effective if they experienced a recent job change and when geographically and technologically closer to the displaced. Stronger ties and lower competition for the available information also speed up re-employment. While largely irrelevant for unemployment duration, contacts' quality is a significant determinant of entry wages and subsequent job stability.