We use linked administrative data from Uganda to track student progression from primary school through university scholarship admission, documenting large leaks in the education pipeline at every stage. Many high-performing students from poorer districts exit before upper secondary school, so the pool eligible for selective public university admission is highly socioeconomically stratified. Merit scholarships are heavily concentrated in a small number of elite secondary schools in the Kampala metropolitan area, implying that financial aid primarily rewards earlier advantages. The results highlight that tertiary merit aid alone cannot equalize access without upstream reforms to broaden basic and secondary education access. 

Salaried workers and entrepreneurs are typically treated as distinct agents in economic models. Yet, in many low-income settings, workers frequently transition between salaried and self-employment. In a stylized job-search model augmented with business investment and income, I show that workers can face a tradeoff between business investment and job search efforts. Furthermore, I show how misperceptions about the salaried job market can lead them to make sub-optimal decisions regarding this tradeoff. Collecting first-hand data from micro-entrepreneurs in Tanzania, I find systematic overestimation of the likelihood of transitioning from entrepreneurship to salaried employment and underestimation of the business survival rates. Micro-entrepreneurs in active job search invest less time and capital in their businesses and generate lower profits, suggesting that job search efforts crowd out business investment. Moreover, job search generates significant direct and indirect costs, estimated respectively at 13\% and 30\% of monthly profits. Using a randomized control trial, I test for the responsiveness of beliefs and behaviors to information about the actual transition rates from entrepreneurship to salaried work. Treated respondents revised downward their perceived returns to job search and became more willing to accept a hypothetical salaried job offering their current earnings level. However, the treatment did not significantly impact medium-term beliefs, nor investments in job-search or business.

The transition from school to the labor market presents significant challenges. This is particularly the case in low- and middle-income countries where the youth population attending primary and secondary school is expanding rapidly and over-optimism combined with limited information can lead to suboptimal decision-making regarding further education and other career preparation choices. We design and test through a cluster-randomized controlled trial a scalable low-cost intervention designed to help secondary school students in Tanzania develop hopeful yet realistic career pathways. This is done through a structured, edutainment podcast series and teacher-led classroom discussions. We show that treated students perform better academically, with a significant increase in national exam success and a higher likelihood of selection into further education. Additionally, self-employment rates and income levels increase. These outcomes are driven by enhanced hope ---characterized by improved agency and pathway clarity--- and by an increase in the likelihood of being active (thus gaining work experience) shortly after taking the exam. Our findings highlight the potential of structured guidance through edutainment in fostering better education and employment outcomes.

This paper explores overconfidence and its impact on students' expectations and behaviors within the context of secondary education in Tanzania. We find that 97% of students in the final grade of junior secondary school expect to perform very well in the national Certificate of Secondary Education Examination (CSEE), with only 35% achieving their anticipated outcomes. This overconfidence is evident in all three forms of overconfidence—overestimation, overplacement, and overprecision—and correlates with suboptimal study strategies, unrealistic aspirations, and a lack of contingency plans. We also highlight the role of limited information in shaping students’ expectations and behaviors, with a particular focus on how feedback can influence their decision-making. The paper contributes to educational economics by integrating psychological insights and examining the effects of information constraints in Sub-Saharan Africa, where students often lack guidance from parents who have limited formal education. Additionally, it emphasizes the broader implications of overconfidence for educational track misallocation and the underappreciation of vocational education in regions with high youth underemployment. Through these insights, we underscore the importance of addressing overconfidence to improve educational outcomes and career preparedness in Tanzania and similar contexts.

We study how merit-based and place-based public scholarships affect university graduation in Uganda, a low-income country where the gross tertiary enrollment ratio stands at just 7 percent and the cost of attendance is equivalent to three to five times GDP per capita. Uganda allocates government university scholarships through two mechanisms: a need-blind National Merit (NM) scheme based solely on national exam scores, and a District Quota (DQ) scheme that awards scholar- ships to the top scorers within each of over 100 geographic districts. Using regression discontinuity designs that exploit the binding exam score cutoffs for each scheme, and drawing on the universe of administrative data on scholarship applications, university admissions, and graduation lists for the 2015/16–2017/18 cohorts, we estimate the causal effects of each scholarship type on university admission and graduation. Crossing the NM cutoff raises admission to any public university by 27 percentage points and graduation by approximately 19 percentage points; crossing the DQ cutoff raises any-university admission by 45 percentage points and graduation by approximately 21 percentage points. A comparison of admission and graduation effects across schemes reveals a striking contrast: roughly 69 percent of marginal NM recipients would have enrolled without the scholarship, rendering the NM scheme largely an inframarginal transfer to students from more affluent districts. The DQ scheme, by contrast, reaches predominantly students on the true margin of access, with only 18 percent of DQ-eligible students enrolling in the counterfactual. Despite this, both scholarships substantially improve graduation rates, suggesting that financial barriers during enrollment—not only at the point of entry—are a critical constraint to degree completion in low-income settings.