Labor Search, Inequality, and Public Policy.

Abstract

This dissertation develops and examines a model labor market. Which workers search for new jobs both while unemployed and while employed. Firms may choose to employ workers under a non-negotiable (wage posting) or re-negotiable (sequential auction) wage contract. In chapter 1, I give conditions under which only the most productive firms select the re-negotiable contract. Modeling contract choice improves the on-the-job search model's fit with respect to observed wage dispersion, (un)employment durations and transitions, labor's share of output, and plausible values of social insurance. With only inflexible wage contracts the model struggles to achieve reasonable labor share and value of social insurance at the same time as realistically large wage dispersion. With only flexible wage contracts the model fails to provide incentive to search while unemployed and suggests implausibly low entry wages. In simulations, the mixed-contract model is able to achieve wage dispersion, labor share, employment transitions, and flow value of unemployment which are all simultaneously consistent with empirical observations. Improved fit makes the new mixed-contract model a better candidate for welfare analysis of social insurance programs, particularly when the researcher seeks to value the impact on workers and on firms. In chapter 2, I explore the relation between worker skill level, composition of contracts, and these outcomes. Even when skill-specific sub-markets are ex-ante identical, differences arise ex post. Higher skilled workers are more likely to receive negotiable wage offers. Consequently, high skilled workers experience lower rates of unemployment, more wage dispersion, and higher returns to experience than low skilled workers. In Chapter 3, I provide details on how to estimate the share of renegotiable labor contracts in an on-the-job search equilibrium which features firm's choice between a wage posting and a sequential auction contract as in Chapter 1. Identification is achieved using administrative data (only) on workers' histories of job-to-job mobility and on-the-job pay gain. I estimate that 22.06 percent of German workers and 62.88 percent of German firms have a wage posting contract. Methods are applicable to a broad set of data sources, making them ideal for cross-group, cross-country, and historical analysis.