Friday, 21 March 2008

Africa's overachievers

It's common knowledge that many high growth countries are in East Asia. So there's no surprise in the table by two academics (Barreto and Hughes, in a 2004 paper "Under Performers and Over Achievers", in The Economic Record) that four of the top ten outperformers in the world from 1960-90 according to a rough classification were in the region. By outperformer, I mean that their growth was higher than predicted by a simple model derived from all the world's data. Some countries will be underachievers, some overachievers, but an "average" country will fit the model perfectly. The classification is rough because the model is not very sophisticated (OLS, if you are interested in regression analysis).

More surprising is the presence of three African countries in the top ten. Botswana heads the list, Burundi is fifth, and Gambia is seventh. Botswana's growth may be explained by diamond sales (although many other resource-rich countries have had dire performances), but neither Burundi nor Gambia can explain the performance by resource sales. They would have been even higher if Saudi Arabia (growth fuelled by oil sales) was excluded from its fourth position, as is normal for oil producers in growth regressions. Burundi's performance is all the more remarkable given the hundreds of thousands of people who died in conflicts during the period.

The model predicts performance based on starting income, investment, population growth, and education. A country's under or overperformance is due to other factors, which lie outside classical economic analysis, and are sought constantly by analysts looking for ways to boost growth. The top four countries (Saudi Arabia excepted) are Botswana, Hong Kong, Korea, and Burundi. The countries share a reputation for valuing education (over-and-above actually receiving education) and ordered societies (Burundian political violence notwithstanding), so it is possible that these features had an influence on growth performance. They are difficult to measure in growth regressions, though.

The data suggests that if Burundi and Gambia had had high savings and education rates, and controlled population growth, then they would have been overperformers on the scale of the South East Asian tigers, and as rich as them.

No comments: