The interaction of workers and firms in the labor market: Applications with integrated employer-employee data

Labor economists look at outcomes that result from the interactions of both workers and employers. Unfortunately, there exist very few data sets in the United States that integrate both sides of the market and therefore the full nature of many of these interactions is not fully understood. This dissertation addresses some of the shortcomings of traditional data by documenting the construction of an integrated employer-employee (IEE) data set, and by exploring three questions that are of interest to labor economists. Specifically, Chapter one describes the construction and characteristics of the employer-employee data set used in the chapters that follow. This dataset integrates information from administrative sources including, Unemployment Insurance records, ES-202 records and Census files. Strengths and caveats of these data are discussed. Chapters two and three examine the worker-employer relation that governs youth labor markets by describing the statistical association between firm level measures of worker turnover and firm specific compensation and employment practices. I find returns to tenure are positively correlated with turnover while worker ability is negatively correlated with turnover. This is consistent with some theories of turnover and not with others. Chapter four extends the analysis of youth labor markets by examining the impact that early job mobility has on the long-term earnings prospects of young workers. Differences across high and low-turnover sectors of employment are explored. I find evidence that employers in the low-turnover sector discount earnings of workers with a history of early market instability after controlling for worker and firm fixed effects. Scarring may be the result of employers discounting earnings of highly mobile workers or with the workers' failure to acquire skills that are particularly valuable to certain firms. Chapter five investigates the impact of technology adoption on the skill wage differential of workers employed in U.S. manufacturing plants. The chapter shows that technology adoption does not have a significant effect on the earnings of high-skill workers, but negatively affects the earnings of low-skill workers after controlling for worker-plant fixed effects. These results seem to support the skill-biased technological change hypothesis.