Search

We describe how ethnic disparities in the labor market between prime aged Hispanic and non-Hispanic white men have evolved over the last 50 years. Using data from the March CPS, the Census, and the ACS, we examine several employment and earning outcomes. Hispanics have experienced sizable gains to employment: from a negative 2% prior to 1990 to a positive 4% after 2010 compared to non-Hispanics. In terms of earnings, Hispanics face a substantial negative disparity between 20% and 30% with some improvement after 2000. Most of the employment gain is driven by those with less than a high school degree, while the earnings disparity increases with education. Comparing Hispanic immigrants with natives reveals much of the employment and earnings gains are attributable to Hispanic immigrants, particularly immigrants not fluent in English.

In this dissertation, I build macroeconomic models to answer questions of empirical relevance for the study of labor markets. The dissertation consists of an introductory overview and three research essays. The first essay is devoted to duration dependence in unemployment, namely the fact that recently unemployed workers have a signicantly better chance of finding a job than the long-term unemployed. I build a directed search model to quantify the importance of three common explanations for this fact: (i) unobserved worker differences, (ii) skill loss, and (iii) job-search effort decline. Two novel results emerge: first, the bulk of the effect of unobserved heterogeneity is concentrated in the first six months of the unemployment spell; the drop in job-finding rates observed at longer spellsis mostly a result of skill loss and lower search effort. Second, skill loss has a vastly greater impact on job-finding than the decline in search effort. These results have two clear implications for labor market policy: (i) the impact of active labor market programs is expected to be larger for the long-term unemployed; (ii) job-training programs are expected to be more effective than job-search assistance policies at reducing long-term unemployment.In the second essay I study how information obtained by a worker while trying to find a job affects her job-search effort. Specically, I analyze how a worker, who is uncertain about her labor market traits and learns about them while looking for a job, allocates her search effort over the unemployment spell. The main result is that search effort is increasing over time when the worker is optimistic abouther traits but decreasing when the worker is pessimistic about her traits. This result can explain discrepant empirical findings from previous literature on search effort. The final essay is devoted to job-search effort as an insurance channel. I build a model in which workers face substantial risk in the labor market but they have two means of self-insurance against this risk: increase their savings and their search effort. The main result is that when labor market risk becomes more severe workers increase both their savings and search effort but the increase in savings is twice as large as the increase in search effort. That is, workers make use of search effort as an insurance channel but much less than the savings channel.

In an attempt to mitigate the negative effects of clientelism, many governments around the world have adopted meritocratic hiring of public employees. This paper challenges the effectiveness of this common practice by showing that meritocratic government hiring can have unintended negative consequences on macroeconomic aggregates. In many countries, public employees enjoy considerable job security and generous compensation schemes; as a result, many talented workers choose to work for the public sector, which deprives the private sector of productive potential employees. This, in turn, reduces firms' incentives to create jobs, increases unemployment, and lowers GDP. To quantify the effects of this novel channel, we extend the standard Diamond-Mortensen-Pissarides model to incorporate workers of heterogeneous productivity and a government that fills public sector jobs based on merit. We calibrate the model to aggregate data from Greece and perform a series of counterfactual exercises. We find that the adverse effects of our mechanism on the economy's TFP, GDP, and unemployment are sizable.