PhD defence: Adrien Gosselin-Pali

Adrien Gosselin-Pali
CERDI, Université Clermont Auvergne

Examiners

Sylvie Démurger, Center for Economic Research on Governance, Inequality and Conflict (CERGIC); Department of Economics, ENS de Lyon
Catherine Guirkinger, Université de Namur, Center for Research in Economic Development (CRED)
Élodie Maitre d'hôtel, CIRAD
Jean-Louis Combes, Université Clermont Auvergne, LEO-UCA
Bity Diene, Université Clermont Auvergne, CERDI, Supervisor
Théophile Azomahou, Université Clermont Auvergne, Supervisor
Simone Bertoli, Université Clermont Auvergne

Abstract

This dissertation is composed of three empirical essays in economics, positioned at the intersection of development economics, health economics, and migration and family economics, with a shared focus on nutrition in sub-Saharan Africa.

Chapter 1 serves as a general introduction, presenting the current context of malnutrition in developing countries, the motivations behind this thesis, and the overarching concepts such as the household, mobility, and geographic dimensions. It also provides a summary of the results and contributions of each chapter.

Chapter 2 analyzes the dynamics of the double burden of malnutrition (DBM) and overweight or obesity (OVOB) using South African longitudinal data. To investigate whether DBM (the coexistence, within the same household, of at least one overweight or obese person and one underweight person) is a transient or persistent phenomenon, a dynamic random-effects probit model is employed. The findings suggest that the double burden is a transitory phenomenon: most affected households do not retain this status in subsequent survey waves. In contrast, OVOB households (with at least one overweight or obese person but no underweight individuals) exhibit strong persistence of their status, also reflecting the long-term stability of overweight and obesity at the individual level. Conversely, the transient nature of DBM can be explained by the tendency of underweight individuals to transition out of this condition over time.

Chapter 3 investigates the effect of migration on the nutritional outcomes of the left behind—individuals who previously co-resided with a migrant. Utilizing longitudinal data from Ghana and a combination of matching and difference-in-differences, the analysis reveals that internal migration negatively impacts the nutritional status of left-behind individuals, particularly children. The primary channel driving the adverse nutritional impact appears to be the short-term disruptive effect caused by migration, likely leading to a negative income shock. While remittances do not consistently offset these negative effects, they may potentially contribute to improved outcomes for left-behind children in the long term. However, the initial shock could have lasting consequences for children's growth trajectories.

Chapter 4 examines the impact of child fostering, a widespread practice in sub-Saharan Africa in which a child is sent to live temporarily or permanently in a household different from that of their biological parents, on the nutritional status of children. Using the same longitudinal dataset as in Chapter 2, which tracks individuals in South Africa, machine learning techniques are employed to address biases related to selection into fostering and endogenous attrition. The results indicate that fostering reduces the probability of being stunted by 7 percentage points, equivalent to a 45% decrease compared to the mean prevalence. This improvement likely results from foster children moving to smaller, rural households, often including retired individuals—typically grandparents receiving a pension. Additionally, fostering not only enhances the nutritional status of foster children but also improves the nutrition of siblings who remain in the sending household, suggesting that fostering can produce mutually beneficial outcomes.

Keywords