This dissertation studies the importance of networks and other institutional factors on workers' labor market outcomes. I particularly focus on manager-employee networks formed within the workplace, for two main reasons: these networks play a critical role in determining the productivity of individual employees, and also affect the equitability of a given working environment, which in turn influences workers' outcomes. Because social networks are more likely to form among those who share similar backgrounds (such as gender or ethnicity), vertical co-worker connections may worsen existing intra-institutional gaps between majority and minority groups, as there is a higher chance of such bonds emerging among the majority. Hence understanding the characteristics and mechanisms of manager-employee connections may yield significant implications for policymakers in empowering a diverse workforce and redressing disparities.
Despite having consequential ramifications for an employee's career outcomes, little attention has been paid to manager-employee networks in the workplace, mainly due to the limited data at hand. This, in turn, limits causal evidence in the existing literature. In the first two chapters below, using web scraping techniques, I construct unique datasets that allow me to identify co-worker connections in specific professions to provide causal evidence of the effects of manager-employee connections.
In Chapter 1, I inquire whether the gender of academic leaders, i.e., college deans and department chairs, affects outcomes of faculty members in terms of (i) wages and (ii) share of female faculty in an academic unit. Exploiting data allowing for a year-by-year identification of any changes in individual departments/colleges such as chair/dean transitions, I adopt an event study design which compares female and male faculty who are exposed to a gender-constant head transition (e.g., male-to-male department chair transition) and those who are exposed to a transition that also involves a change in the leader's gender (e.g., male-to-female department chair transition). I find that managers can improve or worsen female outcomes relative to male outcomes, but the effect of managers does not depend on their gender. This finding is contrary to the common expectation that promoting female managers will have positive spillover effects on other female workers: my findings suggest that merely appointing female managers is not sufficient to reduce gender disparities and improve women's representation in universities.
In Chapter 2, I investigate whether and to what extent connections with "successful" senior colleagues (i.e., senior colleagues who rise to high-ranking positions during the course of their career paths) affect a junior prosecutor's chances of promotion. This study focuses on a professional organization that is marked by its bureaucratic hierarchy where managers train, supervise, and assess juniors as well as hold the influence to recommend them for promotion. To identify a causal network effect of successful seniors, I exploit exogenous variation in networks arising from personnel transfer assignments, an organization-specific attribute unique to the Korean prosecution service. I find that connections to successful seniors have a positive spillover effect on junior prosecutors: a one standard deviation increase in the number of connections with successful seniors increases the probability of being promoted for a junior by 10 percentage points. I further provide empirical evidence that there are at least three potential mechanisms behind the network effect: (i) skill spillovers from a senior to a junior, (ii) transmission of information on a junior's performance between seniors, and (iii) nepotism based on alma-mater connections. I also find that social networks arising within workplaces can reinforce the disparity between the minority and majority groups: the alumni of a major university. My findings thus propose that matching a successful senior with a junior within the same minority group of a given institution is an effective way of supporting the minority group within the workplace.
In Chapter 3, we study the gender gap for academic economists across a wide range of departments and institutions. Extending the faculty salary data used in Chapter 1, we quantify how much of the gender pay gap arises within versus between departments (and institutions), and explore potential explanations for the variations in the magnitude of gender disparity across different departments and universities, focusing on institutional factors such as gender composition and the overall level of dispersion in salaries at an institution and in a department.