Essays on Discrimination and the Role of Earnings Expectations and Personality in Sorting into Career Paths

This dissertation consists of three chapters that study gender discrimination (Chapter 1) as well as the role that earnings expectations and personality play in sorting into career paths (Chapters 2 and 3, respectively). They do so using experimental, survey, and administrative data. In Chapters 2 and 3, data from a nationally representative sample in Germany are used to support some of the results. The first chapter investigates gender discrimination in negotiations using an online field experiment. Although gender differences in negotiations have been studied extensively, also whether discrimination plays a role for differences in outcomes, some questions such as whether women are discriminated against when they take on the role of an employer remain unanswered. In my experiment, I post home improvement jobs on an online platform for classified advertisements and negotiate with users who offer to sell the services. I experimentally vary buyer gender and whether or not the negotiation inquiry includes a justification for why the buyer asks for a discount. My findings show that sellers engage in statistical discrimination against female buyers by not granting them a counteroffer equally often as male buyers unless the buyer justifies their negotiation. There are no differences by buyer gender in the final offers and the discounts that sellers provide. I also provide (albeit limited) evidence for sellers communicating differently with female than with male buyers, and that these differences support the hypothesis of sellers’ statistical discrimination being belief-based. The results amend existing literature on gender and negotiations, correspondence studies, and statistical discrimination by using a novel setting and exploiting text data to study differences in communication. The second chapter studies earnings expectations of first-in-family (FiF) students and their role in college major choice. There are well-established differences by FiF status in human capital formation during several stages in life as well as labor market outcomes, but less is known about whether expectations about these outcomes differ. To shed light on this, we use unique data on two cohorts of incoming students at a large public university in Germany which we collected in 2022 and 2023. The data contain information on expectations about earnings, own ability, and non-wage amenities. FiF status is elicited in the survey by asking for the highest degrees of the subjects’ parents. The survey data are then linked to administrative university data to obtain information on subjects’ majors. Our results show that FiF students expect to earn considerably less than non-FiF students at the age of 30 and 40, respectively and that our data can explain the majority of this gap. Since we elicited earnings expectations for eight counterfactual study fields, we can examine the role of these expectations for students’ sorting into fields. We see that earnings expectations matter less for FiF than for non-FiF students, which is driven by FiF students believing their relative ability to be lower in fields where they expect higher earnings. Comparing our results to representative data from the German Socioeconomic Panel (SOEP), we see that the gap in our data is in line with actual earnings gaps in the labor market (and that the actual gap is larger than the gap in students’ expectations). Chapter 2 extends the literature on social mobility, earnings expectations in general, and earnings expectations and sorting decisions by being the first study to the best of our knowledge to study earnings expectations of FiF students. The results suggest that policy interventions targeted at FiF students’ expectations about earnings and ability can have the potential to reduce FiF gaps in expectations and enrollment. The third chapter discusses personality and its role in career path choice. While the literature has documented the importance of incentives like wages or job security for occupational sorting, the role of personality has only recently been found to play a non-negligible role in it as well. Since occupational choice is often preceded by choosing a path in university education leading up to it, we leverage data from the unique survey setting mentioned above to comprehensively study sorting into eight career paths along 16 dimensions of personality. Our results go beyond findings from previous literature as we not only replicate known sorting patterns such as risk aversion for teacher training students, but additionally show that sorting is relevant across all career paths and personality items in our data. We also show that subjects sort into study fields where their personality profile is closer to the average peer’s. If subjects’ personality does not fit the field, i.e., if it is further away from the average, this implies a higher likelihood of leaving the field over a time span of three years. Re-surveying subjects close to graduation shows that initial differences prevail and that they are comparable to differences in similar occupations in the workforce. We conclude that initial sorting plays a decisive role for personality differences that can be observed in the labor market. In summary, this thesis provides evidence on discrimination, expectations, and personality influencing outcomes which are important for labor market success as well as prior human capital formation. The results may contribute to the public discourse on how to foster success in education and life.