Intersecting Inequalities: Uncovering Disparities in Education and the Labor Market

Abstract

In the past decades, the world has witnessed a significant decline in global poverty and a narrowing of global inequality in income, education, health, and other well-being factors. However, this positive trend is mostly driven by decreasing between-country inequality and stands in contrast with the rise of inequalities within many countries. A key contributor to these within-country disparities is persistent inequality between social groups, such as gender, ethnicity, and migration status. Existing research tends to examine these group-based inequalities in isolation, yet this approach disregards the reality that individuals often have multiple overlapping identities, leading to the potential for compounding disadvantages. Intersectionality, which emerged as a theoretical framework from the humanities, recognizes that the overlapping of social identities can amplify inequalities in ways that are not fully captured when these identities are considered separately. However, the economic literature has so far been reluctant to adopt intersectionality as a framework to quantitatively analyze inequalities. Building on the premise that group-based inequalities may be more intricate than the sum of their parts, this dissertation focuses on three research questions to analyze economic disparities through an intersectional lens. Chapter 1 takes a global perspective to address the question of how the intersectionality framework can be integrated into the measurement of group-based inequalities in education. The other two chapters focus on group-based inequality in the Swiss unemployment system. Chapter 2 analyzes how the effects of unemployment on earnings differ between intersecting groups based on the gender and migration status of the unemployed. Drawing on the insights of Chapter 2, Chapter 4 investigates the influence of public employment service caseworkers in shaping these disparities. In the first paper (co-authored), we seek to integrate intersectionality into the measurement of horizontal or group-based inequality in education. Using Demographic and Health Surveys and census data from 40 countries, we analyze educational inequalities at the intersection of gender and ethnicity. Inequality is measured as the average education ratio between the most and least disadvantaged groups. We find high intersectional inequalities in almost half of countries, exceeding single-group inequalities. We then introduce a new measure, called "differential intersectionality," to quantify the portion of intersectional inequality that is "more than the sum of its parts". The countries with high differential intersectionality are not necessarily the ones with high intersectional inequality ratios. In a cross-country regression, we find that intersectional inequality is strongly correlated with individual education inequality, but not average education levels. Differential intersectionality is associated with neither. The results suggest that addressing education gaps based on gender and ethnicity separately might not be enough to "leave no one behind," as the 2030 United Nations Agenda envisions. This paper underscores the value of intersectionality in providing additional information beyond conventional indicators. In the second paper (single-authored), I investigate disparities in the impact of unemployment on future earnings at the intersection of gender and migration status in Switzerland. Using administrative data more than 450,000 people registered for unemployment benefits between 2010 and 2019, I adopt a stacked difference-in-differences design to address potential self-selection issues associated with unemployment. On average, the unemployed experience a 24% decrease in yearly earnings over four years after unemployment. However, there are significant disparities: while the gender gap in earnings losses is minimal, migrants suffer 85\% higher losses than non-migrants. Analyzing the effects intersectionally reveals foreign women experience 6 percentage points higher losses than foreign men. Possible explanations for these results include longer unemployment duration, recurring unemployment, motherhood, greater losses for foreigners in high-skilled roles, and time since migration. The findings suggest unemployment reinforces labor market inequalities, though it may reduce inequality within the unemployed since the effects are more pronounced for high-income workers. In the third paper (single-authored), I analyze the effect of caseworkers on the labor market outcomes of the unemployed, with a particular emphasis on gender and migration status disparities. The study utilizes data from more than 250,000 first-time job seekers to evaluate caseworkers' effectiveness based on their "value-added" --- a measure derived from the average earnings and unemployment durations of job seekers they assist. Results show that a one-quintile increase in caseworker value-added is associated with a 30% rise in the earnings difference before and after unemployment and a decrease in unemployment duration by 10 days. However, the benefits are not evenly distributed; while women with a migration background benefit slightly more from high-performing caseworkers in terms of earnings, foreign men benefit less compared to Swiss men. Additionally, application requirements and sanctions, tools caseworkers can use to enforce unemployment benefits rules, have varying effects: requirements speed up re-employment, but can reduce earnings, and sanctions improve outcomes, but not for all groups. These findings highlight the need for policy measures that guarantee caseworkers' strategies contribute to equitable and quality employment outcomes for all job seekers. The dissertation finds pronounced inequalities in education and unemployment when analyzed through an intersectional lens. Intersectionality increases complexity, but also reveals large and previously unidentified inequalities. Targeted policies that consider individuals' overlapping identities and are designed to specifically benefit disadvantaged groups are needed. Overall, the research highlights the real-world implications of intersectionality for policymaking.