Labor-market volatility in a matching model with worker heterogeneity and endogenous separations

Abstract

This paper shows that introducing worker heterogeneity into a standard search and matching model can help increase the volatility of unemployment without violating the tight negative correlation between vacancies and unemployment, i.e., the Beveridge curve. In the model, periods of high job destruction and unemployment correspond with periods of more severe mismatch between the demands of firms and the qualifications of job seekers. A more severe mismatch translates into fewer successful employment matches conditional on the number of contacts per firm and, as a result, into a higher expected recruitment cost per worker hired, with adverse effects on incentives to open vacancies. JEL classification: E24; E32; J63; J64 Keywords: search and matching, endogenous separations, worker heterogeneity, unemployment and vacancies volatility

Loss of skill and labor market fluctuations

“…Unemployment is the Bureau of Labor Statistics data series LNS14000000 (http://www.bls.gov/data/); vacancies are the composite Help-Wanted Index computed byBarnichon [2010]; productivity is the Federal Reserve Bank of St. Louis data series PRS85006163 (https://research.stlouisfed.org/fred/).14 For the same reason, instead of the value of 20 used inShimer [2005]'s analysis, Pissarides[2009] retains a value of 7.56 for the elasticity of labor-market tightness with respect to productivity. The latter is very close to the elasticity of 7.82 reported inTable 2.15 SeeFujita and Ramey [2012] andChassamboulli [2013] for further discussions of this issue. Intuitively, the problem with endogenous separations is that a negative productivity shock generates a higher inflow of newly unemployed workers and thus a lower expected duration to fill a vacancy.…”

“…Let us note for completeness that the theme of (ex ante) worker heterogeneity is also carefully analyzed in a paper by Chassamboulli [2013]. We only briefly mention this study because its focus is on match formation decisions (conditional on meeting) and job separation decisions, which are two margins of the DMP model that we abstract from in the model.…”

The Cyclical Behavior of Unemployment and Vacancies With Loss of Skills During Unemployment

“…Using the results derived earlier, the elasticity of the job-finding rate is 3.94, and the elasticity of unemployment is −3.64. Chassamboulli (2013) reports, using evidence from Shimer (2005), that the empirical counterparts for these elasticities are 2.34 and −3.88. Therefore, the model is also able to generate sizable fluctuations in both the job-finding rate and unemployment.…”

“…One could overcome this issue by simulating the model with a finite number of skills. However, one runs into the curse of dimensionality very quickly unless the number of types is very small [for example, Pries (2008) and Chassamboulli (2013) consider only two types of workers; in Bils et al (2012) there are four types, but the model is simplified by assuming separate labor markets]. Approximating the distribution of unemployment histories, which is exponential, with two types would be a poor approximation of the distribution.…”

Vacancy chains and the business cycle. Stringing together job-to-job transitions in micro data

“…The number of poorly qualified workers increases in recessions, but so does the number of qualified workers, and, as firms adjust their employment in recessions, they are able more easily to hire highly skilled workers than in expansions. See also Chassamboulli (2010).…”