Skip to content

Was education a driver of economic development in Africa? Inequality and income in the twentieth century

Literature › Works

Bas van Leeuwen; Jieli Van Leeuwen-Li; Peter Foldvari (2012). Was education a driver of economic development in Africa? Inequality and income in the twentieth century. XVIth World Economic History Congress, Stellenbosch University, 9-13 July 2012.
Type conference paper
Scope Comparative and methods
Method econometric
Period studied 1890–2010
Regions Africa (continental, five UN sub-regions), Southern Africa (includes South Africa, Swaziland, Lesotho, Namibia, Botswana)
Themes education human capital, inequality poverty unemployment, comparative theory, methods data sources
Citation key leeuwen2012

Argument

The paper asks whether education drove economic development in Africa across the twentieth century. Using a new annual series of average years of education, panel error-correction models and a panel VAR, it finds that neither education nor physical capital has a long-run effect on per capita income growth. Income instead drives education and capital, and growth is attributed mainly to general productivity (TFP) rather than factor accumulation. Education does raise technical efficiency, producing a small positive effect on TFP. Because productivity growth combined with falling technical efficiency widens gaps between countries, inequality rises. Within-country income inequality is strongly and non-linearly related to educational inequality.

Evidence and method

Per capita GDP from Maddison (2007) and the Conference Board TED, with pre-1950 estimates built from Amin, Yousef, Szereszewski, Deane, Krogh and Willers (Southern Africa), Frankema and Van Waijenburg real wages, and Prados de la Escosura. Annual average-years-of-education series built from census benchmarks (Cohen and Soto), UNESCO enrolment data, Mitchell and parliamentary papers, then compared with Barro-Lee, Morrisson-Murtin and others via reliability ratios. Physical capital from the Extended Penn World Tables, Easterly-Levine and Miketa. Income Ginis from Van Zanden et al. (2011). Educational Gini computed after Thomas, Wang and Fan.

Findings

  • Africa's per capita GDP rose roughly threefold over the twentieth century (1990 GK dollars). (p. 4)

    We can see that per capita GDP increases from 683 GK dollars in 1900 to 2023 GK dollars around 2010.

  • Southern Africa, the region that includes South Africa, outperformed other regions in income growth. (p. 4)

    it is obvious that especially Southern Africa grew faster over the twentieth century than the other regions.

  • Southern Africa also led the continent in years of schooling throughout the century. (p. 6)

    we find that especially Southern Africa did well in terms of years of education.

  • The long-run error-correction estimate shows a negative association between education and income. (p. 11)

    More education seems to be paired with lower per capita income in the long-run.

  • No long-run link between physical capital and per capita GDP is found. (p. 11)

    Another shocking result is that from the above specification we cannot find evidence in favour of a long-run relationship between physical capital and per capita GDP.

  • Panel VAR impulse responses show no effect of education on income, implying no social returns. (p. 14)

    education does not seem to have affected per capita GDP neither in the short- nor in the long-run.

  • Education has a small negative effect on general productivity, offset by a strongly positive effect on technical efficiency. (p. 18)

    only education has a small negative effect on general productivity. However, this effect is offset by the strongly positive effect education has on technical efficiency.

  • Africa is the only continent where within-country inequality rises after 1890. (p. 19)

    Interestingly, Africa turns out to the only continent where within country inequality actually rises after 1890.

  • The rise in within-country inequality is sharpest in Southern Africa, the richest region. (p. 19)

    it turns out to be particularly pronounced in Southern Africa, which is also the region with the highest per capita income.

  • Eliminating educational inequality would sharply reduce income inequality (the body text gives 36.9 Gini points; the summary and conclusion say 81%). (p. 21)

    The mean of the educational Gini in 2010 was 45.35. If this were reduced to zero, based on the coefficients in table 3, this would reduce income inequality by 36.9, which is a quite big effect