Working Papers


(This version: March, 2015)
Empirical evidence shows that longer spells of unemployment are associated with lower job-finding rates, and also fewer interview opportunities. We develop a model of rational stigma that combines two explanations identified in the literature, namely dynamic sorting and ranking. We study a directed search economy with multilateral meetings, heterogeneous workers and incomplete information about a worker's type. Skilled workers are more productive and more likely to be suitable for any job. Firms observe a private signal, which perfectly identifies unsuitable matches. As longer spells signal lower expected productivity, firms rank applicants by unemployment duration in equilibrium. Exit rates from unemployment decline with duration because of both dynamic sorting and ranking.  According to our calibration, ranking by duration accounts for approximately one fourth of the fall in exit rates. Furthermore, our model  matches quite well the mild decline of re-employment wages along duration found in the data.

       Online Appendix

(This version: March, 2015)

It is well known that exit rates from unemployment and re-employment wages decline over the unemployment spell after controlling for worker observable characteristics. Although unobserved heterogeneity has long been recognized as an important explanation for the decline of the exit rates, very little theoretical work has been done to model it. We study an economy in which workers are privately informed about their skills and can direct their search.  We show that the unique equilibrium is separating, and skilled workers face more job opportunities and obtain higher wages. As the composition of the unemployment pool varies with unemployment duration, average exit rates and average wages fall with duration. The equilibrium result relies on two key ingredients: performance-pay schemes and the firms' ability to commit to renting an input complementary in production with worker skills. 

           (This version: December, 2014)

In this paper we study the efficiency of equilibrium worker turnover, both theoretically and quantitatively. We construct a competitive search model with risk-averse workers, market incompleteness, and information frictions. Job-seekers trade off wages and matching rates, taking into account that low-wage jobs are typically short-lived.  Because of market incompleteness, non-employment risks are not fully insured. Risk-averse workers factor in such uninsured risks when searching for jobs. As a result, worker turnover is inefficiently high in equilibrium. Non-employment benefits yield efficiency gains by covering such risks, and induce agents to form longer-lasting matches. Using SIPP data for the US, we document that worker turnover is large, with above 42\% of newly employed workers falling into non-employment within a year. However, according to our calibration, worker turnover in the US economy falls short of the optimal level.


Work in Progress

  • Intermediation in the Housing Market
  • Effects of Switching Occupations along the Job Search

Published Papers


        Online Appendix