Improving student learning in Uruguay: Causal estimates of the effects of educational policies and interventions

This thesis addresses key questions in the economics of education by examining how various inputs —ranging from school resources and teacher salaries to family support and non-monetary incentives— affect student outcomes across different educational levels in Uruguay. The research is structured into four chapters that investigate the formation of human capital in both primary and university settings. Each chapter focuses on a distinct policy intervention or educational input, such as a teacher salary increase aimed at attracting more experienced educators, an unconditional cash transfer program for early-life support, the provision of relative performance feedback to university students, and the impact of major soccer events on study behavior. This comprehensive approach seeks to untangle the complex interactions among various determinants of educational achievement, thereby providing insights for policymakers aiming to implement evidence-based reforms.

Methodologically, the thesis is grounded in the framework of credible causal inference, utilizing tools such as Regression Discontinuity Designs (both sharp and fuzzy), a Randomized Control Trial, and natural experiments. These methods are applied to exploit exogenous variations —whether arising from clear eligibility cutoffs, random assignment in an experimental setting, or the unexpected scheduling of major soccer matches— to isolate the causal impacts of the studied interventions. The research design emphasizes overcoming selection bias and distinguishing between correlation and causation, drawing inspiration from the “credibility revolution” in empirical economics and the pioneering work of scholars such as Angrist, Imbens, and Card.

The findings of the dissertation offer nuanced contributions to the literature on education policy. The analysis reveals that while policies such as increased teacher salaries significantly improve teacher experience and tenure in disadvantaged schools, their direct impact on student test scores remains limited. Similarly, early-life cash transfers show modest long-term benefits, particularly for low birth weight children, and relative performance feedback exhibits gender-specific effects, improving satisfaction and outcomes for men while potentially harming academic performance among women. Finally, exploiting the natural experiment of major soccer events demonstrates that increased leisure costs due to these events can negatively affect exam performance, with evidence suggesting that female students may be disproportionately impacted.