Three Essays in Labor Economics

This dissertation consists of three essays in labor economics. Two main themes run through the essays: the development of human capital and the quality of economic data. The first essay reports on the impact of mis-measured data on the point estimates found from OLS regression. The second essay reports on changes over time in quantitative human capital. The third essay investigates how absences in high school can influence long-term education outcomes. This final essay blends the themes found in the first two essays by (1) using high quality administrative high school transcripts data and (2) exploring effects on human capital accumulation. The first essay, co-authored with Gary Engelhardt, provides new evidence on the extent of measurement error in respondent-reported earnings data by exploiting detailed W-2 records matched to older workers in the Health and Retirement Study (HRS). The empirical findings are qualitatively consistent with previous studies. Mean measurement error in the 1991 earnings data for men from the Original Cohort of the HRS is somewhat larger than what has been found in other validation studies, but is still modest. There is a negative correlation between the measurement error and the true value of earnings as measured by the W-2 records, which indicates the presence of non-classical measurement error. However, for men and women, this error shows little correlation with a standard set of cross-sectional earnings determinants. The second essay examines changes over time in quantitative human capital. Historically, part of the gender gap in quantitative investments was due to a gap in pre-college quantitative skills. Using the 1979 and 1997 cohorts of the National Longitudinal Survey of Youth (NLSY) I show that the pre-college quantitative skills gap between young men and women today is smaller than in the past. Then, by semi-parametrically reweighting the NLSY79 to have the characteristics of the NLSY97 I predict that the fraction of young women in quantitative fields should double. In fact, though, the allocation is virtually unchanged across the cohorts. The third essay, co-authored with Kalena Cortes and Chris Rohlfs, is a study that measures the effects of classroom absences on high school course grades and completion. Using a unique dataset of transcript files from Chicago Public High Schools, we address two interrelated questions: Does absenteeism in specific subject areas ( i.e. , Mathematics or English) affect student course grades and dropout rates? Because attendance and absences are not randomly assigned across students, those with particularly high rates of absenteeism are likely to have less motivation and ability than the typical student. We address this problem by employing a quasi-experimental research design based on the essentially random ordering of classes over the course of the day. This research design makes it possible to disentangle the causal effect of absenteeism from motivation, ability, and other determinants of academic performance. In estimating the effects of absenteeism on dropping out, we address the nonrandom selection problem by conducting two-stage least squares analysis, where the exogenous variation in course scheduling is used to instrument for student absenteeism.

Education

21670