The Underlying Role of Education, Gender Inequality and Health in Economic Development: Sub-Saharan Africa, 1920 – 2013

Abstract

In the first chapter, titled “Educational Gender Inequality in Sub-Saharan Africa: A Long-Term Perspective”, I investigate together with my co-authors Joerg Baten, Michiel de Haas, and Felix Meier zu Selhausen the evolution of educational gender inequality patterns across sub-Saharan African countries and districts as well as in comparison with other world regions over the 20th century. Based on our descriptive statistics, we find that sub-Saharan Africa, relative to other developing regions including South Asia, Southeast Asia, and the Middle East and North Africa (MENA), displays the lowest educational gender inequality gaps in the beginning of the 20th century, however, became the most gender unequal region by the 1980s. which can be associated with its comparatively low educational level. During the colonial period, however, we observe a rise in educational gender disparities in all world regions. Sub-Saharan Africa’s tilting point in the 1960s was later than in other regions, and by the 1980s sub-Saharan Africa had the highest gender disparities. Within Africa, we observe a similar trend across most countries, with a widening of the gender gaps in education during the colonial era (c.1880-1960), peaking mid-century, fol-lowed by a consequent decline during independence in the post-colonial period. We find that men’s educational progress is correlated with an initial increase and subsequent decline of the educational gender gaps. We refer to this pattern as the educational gender Kuznets curve. Based on six decadal cohorts and 1,177 birth districts, we employ a Least Squares Dummy Vari-able model and investigate correlates of educational gender inequality at a sub-national level. Our regression results support the inverted U-shaped relationship between the gender gap and male education. In addition, we find that gender gaps in educational attainment were lower in districts that had early access to railroads, more urbanization, coastal location, and early 20th century mission stations.

The second, single-authored chapter, titled “Educated Girls, a Force for Development? Gender Inequality in Education and Economic Performance in Sub-Saharan Africa: A Path-Dependency Analysis”, explores the association between educational gender inequality and economic growth. This persistence study applies a repeated cross-sectional regression design and examines the effects of path dependency in education on current per capita economic output. Based on a sample of 19 sub-Saharan African countries, 1,107 sub-national regions, six decadal birth cohorts, and a total of 5,322 observations, results show a strong negative correlation be-tween gender disparities during the past century and current economic activity, proxied by nighttime light intensity per capita. Moreover, they reveal that increased investment in female education would likely have benefited sub-Saharan Africa’s economic output per capita. Investigating indirect effects, findings of the applied mediation analysis show that part of the negative association between gender disparities in education and economic development is mediated through fertility. Finally, a number of control variables, including educational expansion, trading location, access to railroads, the endowment of minerals, labor migration, and to some extent democracy and life expectancy at birth, are positively correlated with economic development.

In the third and final chapter, titled “Uncovering the Role of Education in the Uptake of Preventive Measures against Malaria in the African Population”, which I co-authored with Neha Upadhayay, we look at the role of education, in particular numeracy and literacy, in malaria pre-vention and treatment-seeking behavior adopted among the African population. We base our pooled OLS analysis on a sample of 33 African countries, 407 regions, and a total of 1,960 ob-servations covering six birth decades (1940s-1990s). Findings show that being numerate and literate are both positively correlated with our measures for malaria control, including the usage of insecticide-treated bednets, the intake of antimalarial drugs and treatment-seeking in a medical facility, in case of malaria symptoms. Standardized coefficients suggest that numeric skills are at least as important as being literate. Results suggest that, besides education, the involvement of women in healthcare decision-making and the exposure to media are positively associated with preventive attitudes and treatment practices. On the opposite, we find that having a low socio-economic status is negatively correlated with malaria control. Similarly, in higher elevated regions and areas with lower precipitation, where malaria is less prevalent, the usage of malaria protec-tion measures is low. However, in urban regions, antimalarial drugs are frequently used as a measure of protection although the disease is spread more severely in rural regions.