VoxEU Column Education

Income and schooling

Average income per capita is strongly correlated with more schooling, but this relationship is more complex than it appears. This column presents new research showing that a large part of the correlation is attributed to the causal effect of economic prosperity on the formation of human capital via schooling.

Countries’ average income per capita is strongly correlated with more schooling. This can be seen both by looking at the relationship between them across countries (Figure 1), and by considering their evolution over time in particular countries. For example, the percentage of the population in the US with at least a college degree rose from around 10% in the early 1960s to almost 30% in the early 2000s, while annual real GDP per capita in the same period grew from under $20,000 to over $40,0001.

Figure 1. Income and schooling

These observations suggest several possible interpretations:

  • One, favoured in earlier works, is that human capital generated through schooling spurs economic growth and creates prosperity (see Barro 1991, Barro and Sala-i-Martin 1995, Benhabib and Spiegel 1994, De la Fuente and Domenech 2006, Hanushek and Kimko 2000);
  • An alternative interpretation, suggested by Bils and Klenow (2000), maintains that increased schooling is mainly caused by growth in per capita income.

It is, of course, also possible that the correlation is driven by a third factor, such as changes in social norms or in the institutional environment.

In our research, we explore the hypothesis that national income growth drives schooling – by allocating more resources to education or by nurturing an institutional environment that is conducive to schooling (Brückner and Gradstein 2013).

Recent research

To properly identify this link one needs to extract the part of national income growth that accrues externally and is thus not driven by the accumulation of human capital via schooling. To this end, we focus on variations in countries' national income that are due to variations in international oil prices.

Specifically, using a panel data set comprising over 138 countries during the period 1970-2009, we construct instrumental-variable estimates that exploit the significant effect of changes in international oil prices weighted by countries' average net export shares of oil in GDP, which may be positive or negative (for net importers of oil). Variations in international oil prices are strong predictors of countries' GDP per capita growth. Column 1 in Figure 2 illustrates how the two are synchronised for a select group of major oil-exporting countries, and column 2 shows the reduced form effect of oil-price variations on these countries' secondary-school enrolment rates. Note that since variations in international oil prices are highly persistent, our instrumental variables estimates should be interpreted as capturing the effect of variation in permanent national income on schooling attainment.

Figure 2. Oilprice shocks and schooling (reduced form)


Our analysis reveals that national income growth has a significant positive effect on a variety of schooling attainment measures. For example, using the proportion of the relevant cohort of students enrolled in secondary schools as such a measure, we find that a 1% increase in GDP per capita increases the secondary school enrolment rate by about 0.2% on average. This effect is much larger in poor countries than in rich countries, and more so for female students. Alternative measures of schooling attainment, such as the total number of years of schooling, or literacy rates, are similarly affected by growth in per capita income. For example, an increase of 1% in GDP per capita leads, over a five-year period, to an increase of 0.3% in average number of years of schooling.

Policy conclusions

Our results suggest that a large part of the correlation between national incomes and schooling attainment should be attributed to the causal effect of economic prosperity on the formation of human capital via schooling. While not entirely refuting a reverse causal link, they indicate that it is secondary to the effect of incomes on schooling.


Barro, R (1991), "Economic Growth in a Cross Section of Countries." Quarterly Journal of Economics, 106(2), 407-44.

Barro, R and X Sala-i-Martin (1995), Economic Growth, New York, McGraw Hill.

Benhabib, J and M Spiegel (1994), "The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data," Journal of Monetary Economics 32, 143-174.

Bils, M and P Klenow (2000), “Does schooling cause growth?”, The American Economic Review 90, 1160-1183.

Brückner, Markus and Mark Gradstein (2013), “Income and schooling”, CEPR discussion paper, Centre for Economic Policy Research.

De la Fuente, A and Domenech, R (2006), “Human capital in growth regression: How much difference does quality data make?”, Journal of the European Economic Association, 4(1), 1–36.

Hanushek, Eric A, and Dennis D Kimko (2000), "Schooling, labor force quality, and the growth of nations", The American Economic Review 90, 1184-1208.

1 The former data are from the US Census Bureau, various years, and the latter – from the Penn World Tables, Heston A, R Summers, and B Aten, Penn World Table Version 7.0. Center for International Comparisons of Production, Income and Prices, University of Pennsylvania, 2011.

4,409 Reads