VIARENGO Martina

The main objective of this research is to empirically analyze, using a novel and robust approach, the determinants of school attendance, the role of the informal sector in school dropout, and the links between youth work and productivity in small firms. First, we are interested in determining the magnitude of private returns to education in Venezuela. To do so, we use repeated cross-section data from the household survey. The disadvantage of using pseudo-panel data is the high presence of measurement error. We address this problem through the use of the instrumental variables method based on empirical moments of order greater than two. Subsequently, we seek to understand the causes of dropout in secondary education in Venezuela. Our contribution to the dropout literature is twofold. First, the inclusion of informality (represented by the parents of youths who are moonlighting) as one of the main factors that affect the likelihood that children will graduate from high school. Second, the use of panel data in the assessment of the high school dropout problem provides new and more robust insights into Venezuela's educational bottlenecks. The econometric approach is based on the use of Mundlak corrections to eliminate endogeneity bias. Finally, we analyze the effect of young workers on the productivity of unregistered small firms in Brazil. We use the instrumental variable method to ensure that our results are not biased.

This thesis contributes to the literature on food consumption, child malnutrition and school achievement in developing countries. The first chapter aims to estimate the causal relationship between malnutrition and school achievement among a sample of Malagasy children enrolled in primary school. Rainfall shocks experienced in early childhood are used as exogenous instruments to explain long-term nutritional status. Results indicate that stunting is a significant barrier to school learning. The second chapter examines the effect of food distribution cycles on food consumption in refugee households. The results suggest that average grain consumption decreases the further away from the day of the food distribution. This effect is large enough to have an impact on the short-term nutritional status of children in our sample. The third chapter exploits panel data on a cohort of Malagasy young adults to estimate a production function of cognitive abilities through the use of a value-added model. The results highlight the essential role of schooling in the acquisition of cognitive abilities.

Uncertainty weighs heavily on rural households in developing countries. This thesis aims to better understand the consequences of this vulnerability on behavior and to analyze the implementation and impact of risk management strategies, both internal and external to the household. The first chapter focuses on the effect of exposure to land risk on savings behavior in Angola. It establishes that households without property documents are not able to accumulate precautionary savings. This self-hedging behavior only occurs when several risks are accumulated. The second chapter focuses on intra-household decision-making in the choice of diversification strategies. It examines the allocation of women's work time in Senegal between domestic work and income-generating activities. A separate spheres model with a transfer between spouses is validated. The weak role of financial incentives and the importance of bargaining power are demonstrated. The last chapter analyzes the impact on household vulnerability of an external intervention through a program aimed at access to energy services. Its impact is measured on child nutrition. The results show a positive and progressive effect of the presence of infrastructure on long-term nutrition. The intensity of operation reduces short-term malnutrition. A diffusion effect is also established. These three chapters show the inability of households to implement strategies to combat vulnerability independently. They then demonstrate the need to develop access to markets in rural areas. Finally, the role of external intervention in providing management tools to households is highlighted.