In a provocative analogy which appears on the first page of her stimulating Chapter 8, Faezeh Foroutan compares Sub-Saharan Africa (SSA) with Belgium, whose 1989 GNP she claims to be approximately equal to the summed GNPs of the SSA countries. ‘Imagine subdividing Belgium’, she writes, ‘into forty-something independent countries, each with its own isolated goods and factor markets’. The force of this extravagant comparison, and its inappropriateness in some respects, almost define the problem which we confront when we think about regional integration in SSA.

A notable difference between the two cases is in size. Belgium is tiny on a world map; Africa is vast. One can drive between any two points in Belgium in a few hours, even if traffic is congested. There are points on the SSA map more accessible to Brussels by air or sea than to populated towns only a few hundred miles away. Africa is poorly integrated as a geographical region, leaving aside the considerable man-created economic barriers. Equally, while Belgium may have more than its share of linguistic and religious divisions by European standards, many African countries, not to speak of African regions, would be happy to swap their fractious tendencies for those of Belgium.

Similarly, Belgium's regions have few comparative advantages of the textbook type vis-à-vis other Belgian regions. This is not to say that splitting the country into forty units would be without cost. The costs would be huge – but they would consist predominantly of the unnecessary duplication of activities with scale economies of various kinds. SSA, by contrast, encompasses great diversity, with numerous examples of textbook comparative advantage.