Barriers to Economic Mobility: Experimental Evidence from Education and Labor Markets

The transition from school into labor markets for youth from low-income households is marred by several constraints, including limited exposure and scarce opportunity for income mobility. These barriers begin at home, where aspirations are first formed, get emboldened at school, where students internalize what is feasible for them, and carry into the labor market, which determines who advances to a first job. This dissertation examines this transition in four chapters. I first study labor market discrimination on a mutable, poverty-linked trait, and then compare the two dominant methodologies used to measure labor market discrimination. I then study whether short internships change beliefs of low-income public school students about labor market returns, and finally examine how siblings shape occupational choice within resource-constrained households. Chapter 1 moves beyond the existing literature on labor-market discrimination, which largely focuses on immutable traits: gender, race, ethnicity, age, to examine a different dimension: oral appearance, a trait that is itself an equilibrium object. Visible dental conditions are common in low-income populations and treatable at modest cost, yet many workers do not correct them. I develop a model in which oral appearance generates hiring disparities through four channels: taste-based discrimination, statistical discrimination through productivity beliefs, customer discrimination, and coworker discomfort. Using AI-generated stimuli and an incentivized resume-rating study with more than 800 firms in India, I find that visibly poor teeth produce a hiring penalty roughly half the size of removing a qualification certificate, consistent with firms internalizing how coworkers and customers are expected to react to a visible, low-status cue. Chapter 2 compares the two dominant methods for measuring discrimination: correspondence audits and incentivized resume-rating studies. I run them simultaneously in the same labor market and find that caste and gender discrimination emerge strongly in the correspondence study but attenuate in the incentivized setting, while appearance-based discrimination appears only in the latter. The pattern is consistent with social-desirability bias suppressing the most salient discriminatory margins when employers know they are being evaluated. Chapters 3 and 4, with Emma Zhang, turn to miscalibrated beliefs about job-search frictions among low-income public-school students in Delhi. In Chapter 3, we evaluate a randomized internship program in government secondary schools and find that roughly eighty hours of work exposure raises demand for career information and modestly recalibrates beliefs about wages and job-search times, with effects driven largely by girls. In Chapter 4, we ask whether information-based interventions can shift occupational aspirations when household gender norms are transmitted through sibling exposure. We find that sibling structure is more strongly associated with stated labor-market aspirations than the experimental treatments: having an older sister who is working is associated with substantially higher application rates, and the effect is nearly twice as large for female respondents. Taken together, these chapters show that the mobility gap facing low-income youth is not a single barrier but a combination of barriers operating at distinct points in the school-to-work transition. Closing the gap requires interventions targeted to each stage.