UC Berkeley

This manuscript presents three essays on development and labor economics. Each chapter examines a factor that affects the educational or labor market outcomes of individuals in low income environments. All three papers take advantage of detailed secondary data sources ranging from detailed household surveys, to data from a computerized homework platform. This data is then used to conduct careful examinations of the educational outcomes and labor market choices made by individuals in low-income environments. In Chapter 1, I show that students of lower socio-economic status (SES) under perform on mathematic exams that feature a higher share of monetary themed questions due to an attention capture effect of poverty. In Chapter 2, we conduct a detailed characterization of the role that the seasonality of labor demand in agriculture plays in the persistence of rural poverty in Malawi. Finally in Chapter 3, I evaluate whether nominal wage rigidity created smaller distortionairy effects on rural labor markets in irrigated areas of India that experience less production volatility.

In my first manuscript, I examine how students with lower SES indicators perform worse when randomly given an exam or assignment that features a larger share of mathematics questions in which money is salient. I examine this effect and find a similar pattern in three different data sets: a homework platform in the United States, a national assessment exam in Mexico and an international cross-country standardized assessment examination. I find that this pattern begins as early as in the fourth grade, is largest for the most disadvantaged and is responsive to income shocks. For students with SES indicators below the national median, a 10 percentage point increase in the share of monetary themed questions depresses exam performance by 0.026 to 0.038 standard deviations. The magnitude of the effect represents about 6% of the overall performance gap for below median SES students. Evidence from a homework platform shows that acquiring a mathematical skill takes differentially more time and effort for low SES students when it is practiced using monetary prompts. Using question-level data, I confirm the role of financial salience by comparing performance on monetary and highly similar non-monetary questions. Furthermore, by leveraging the randomized ordering of questions, I identify an attention capture effect on directly subsequent questions, providing evidence that the attention capture effects of poverty affect policy relevant outcomes outside of experimental settings as performance on examinations is a significant determinant of educational and economic opportunities.

In my second manuscript, using data from Malawi, we explore how the seasonality of agriculture contributes to the persistence of rural poverty in Sub-Saharan Africa. We show that labor calendars for rural households offer similar employment opportunities as for urban households in terms of time worked at peak planting time, but much lower opportunities throughout the rest of the year. Given the high level of urban unemployment in Malawi, an urban-based structural transformation does not offer a clear alternative to rural poverty through rural-urban migration. By contrast, we show that elements of both an agricultural and a rural transformation can help fill-in and smooth-out labor calendars, providing a pathway to rural poverty reduction. Importantly, in this paper, we develop a methodology to map labor time use calendars across an agricultural season that can be applied broadly to agricultural household surveys in other contexts.

Finally, in the third manuscript, I investigate whether downward nominal wage rigidity is more binding and has larger distortionary effects in Indian districts that face greater production volatility. I use the instrument developed for dam construction in Duflo and Pande (2007), and apply the identification strategy used in Kaur (2019) to estimate whether dam presence influences the magnitude of the distortions created by nominal wage rigidities. Despite suggestive evidence, results are inconclusive. Results are likely confounded by non-linearity in the effect of dams, the effects of lagged rainfall on contemporaneous production as well as substitution towards water intensive farming strategies in irrigated areas that exposes producers to significant production drops in severe drought years.