Human Capital and the Latin American Growth Puzzle

Eric A. Hanushek
Published Date
The importance of economic growth for the economic well-being of countries is well-recognized throughout Latin America, but the long term history of development in the region has been almost uniformly disappointing. Moreover, while the causes of this lagging growth have been the subject of dispute with various economists pointing to such factors as political instability, international trade policy, monetary instability, and the like, recent analysis by Ludger Woessmann and me suggests a very straightforward explanation. The quality of schools across the region is not up to world standards and is putting the brakes on potential growth. Our analysis looks around the world at what determines differences in growth rates and arrives at a simple answer: Long run growth is very closely linked to the human capital of the population as measured by international mathematics and science scores. On this count, the nations of Latin America have done very poorly compared to those in all other regions except Sub-Saharan Africa. The relationship between growth rates between 1960-2000 and test scores, allowing for differences in GDP in 1960. SSAFR is Sub-Saharan Africa; LATAM is Latin America; COMM is the old Commonwealth countries of Australia, Canada, New Zealand, and United States; MENA is Middle East and North Africa; EURO is Europe, and ASIA is East Asia. For data and description, see Eric A. Hanushek and Ludger Woessmann, “Schooling, Educational Achievement, and the Latin American Growth Puzzle,” Journal of Development Economics (forthcoming). This is readily seen from the attached figure. The figure compares test scores on the international tests (PISA and TIMSS) to growth rates in GDP per capital from 1960-2000 after making adjustments for the initial level of GDP. (It is common to control for starting income level because nation’s that start behind simply have to copy what other nations are doing instead of inventing new ways of doing things. This adjustment explains why the graph is labeled as “conditional” growth rates and “conditional” test scores.) The remarkable thing is that all of the regions fall almost perfectly on a simple line. Latin America has always had considerable school attainment. In 2001 expected school attainment in Latin America was 13 years, compared to 7.1 in Sub-Saharan Africa and 9.5 across all developing countries. But the years of schooling have not translated well into learning, and Latin American countries have consistently performed at or near the bottom of the world rankings on international assessments. Only recently have Chile and Brazil begun to catch up to the more developed world – but there is a long way to go for all Latin American countries. The simple point is that growth is highly dependent on the quality of the labor force. And the quality of the labor force depends on high levels of cognitive skills developed in schools. The challenge to Latin America is to bring schools up to be world competitive.