Studies in macro labor markets: Flows, stability, search and matching

The Mortensen-Pissarides search matching model is the standard tool for relating the search effort of unemployed workers and job vacancies to the flow of new job matches. The standard MP model, however, ignores the search behavior of employed workers, even though job switchers account for roughly 40% of new hires in the United States. This dissertation explores the effects of job switchers on labor market dynamics, along with recent empirical trends in US job stability.;In chapter 1, I add on-the-job search to a standard real business cycle model with search-matching unemployment to address the "Shinier Paradox" of too little cyclical volatility of unemployment relative to real-world data. I find that adding on-the-job search can increase the cyclical volatility of unemployment when employers prefer to hire the previously employed. New to the literature, however, I demonstrate a countervailing crowding-out effect. I show that for a reasonable parameterization, the former effect dominates and cyclical volatility is increased.;In chapter 2, I explore empirical trends in US job stability from 1983-2012 through retention rates, the probability that an employed worker will remain employed with the same employer after one year. I apply novel synthetic cohort techniques to Current Population Survey tenure data to estimate individual retention functions controlling for multiple worker and job characteristics simultaneously. I find that retention rates increased by one to two percentage points following 2002. This increase is entirely explained by increased retention among workers in their first two years of employment.;In chapter 3, I explore one possible explanation for this increase in job retention, specifically that workers became more cautious about changing jobs because of the risk of becoming unemployed in a weaker labor market. I adapt the Jovanovic model in which job match quality is measured imperfectly prior to accepting a job. I simulate this model numerically and find that a decrease in the job finding rate increases both unemployment and job retention. The temporary increase in the retention rate is larger in a model with more noise in the measurement of job match quality, but the effect is small.