ERUDITE laboratory et UFR Economie et Gestion

This paper explores the causal relationship between education and fertility. It examines whether education reduces fertility at the intensive and extensive margins. It also investigates how education impacts age at first birth. We exploit an exogenous variation in education induced by an extension of compulsory schooling in Indonesia in 1994. As this law increased education mainly in regions that were initially lagging behind, a difference-in-differences variable based on women’s year and region of birth is used to instrument education. Our results suggest that additional schooling leads to a decrease in childlessness and to a delay in first birth, but no effect is observed on achieved or desired fertility. With regard to the mechanisms, better-educated women are more attractive on the marriage market, which explains why they are less often childless. While no effect is found on the labor market, education increases contraceptive use and women’s decisionmaking authority.

A large body of research has well documented the growing contribution of private schools, including low-fee private providers, to education in underprivileged areas. Using a unique database from rural Pakistan, this paper determines the drivers of schooling behavior using a Heckman probit and a household fixed effects models that take into account non-random enrollment choice. The results suggest that gender and socioeconomic barriers still prevent certain parts of the population from accessing education and especially private schools. Both the lack of public schools and the perceived low quality of these schools explain why parents choose private schools even if free public schools are available. The household fixed effects model confirms the results and also shows that there is a significant intra-household gender gap in private school enrollment.

Cette these se propose d’etudier les choix d’education en adoptant une approche originale integrant la notion de qualite de l’enseignement. Cette etude se concentre sur deux pays : l’Indonesie et le Pakistan. Dans un premier temps, nous montrons que la scolarisation obligatoire a eu des effets sur les taux de scolarisation qui se sont repercutes sur les comportements de fecondite. Ces effets ne sont cependant pas automatiques et peuvent cacher une grande heterogeneite. Nous nous interessons ensuite au concept de qualite de l’education, une notion qui a de multiples facettes. Si on l’apprehende par le prisme des resultats scolaires, les enseignants semblent jouer un role majeur. Cependant, cette definition n’est pas entierement satisfaisante quand on s’interesse aux choix de scolarisation. En effet, les performances academiques des ecoles ne refletent qu’une partie de l’opinion des parents. Les parents semblent egalement rationaliser leurs choix a posteriori et considerent que les ec...

This paper assesses the factors driving parental perceptions about school quality in rural Pakistan. Using a three-period longitudinal database on rural households in three districts, this study estimates different models to answer this question: ordered and generalised ordered logit specifications as well as linear probability models with household and school fixed effects. The results are preoccupying as we observe strong differences between privileged and disadvantaged households. Only the richest households take into account test scores when forming their perceptions. This finding suggests that access to information is unequal and could penalize the poorest children. Both rich and poor households tend to overestimate the quality of private schools which can explain the recent growth in private enrollment. Other school characteristics such as the size of the school, the medium of instruction (English teaching) or school infrastructure are also valued by parents.