One comment in particular merits a response: that Botswana, one of the best-governed countries in sub-Saharan Africa, and one of the most prosperous (its GDP per capita is about $10,000, and thus close to the Republic of South Africa's $12,200), has one of the highest AIDS rates in Africa (24 percent), and therefore I am wrong to suggest that bad government and poverty are the roots of Africa's disproportionate incidence of HIV-AIDS. I had pointed out, however, that the factors that influence a country's AIDS rate are multiple. One of them is migrant labor, which facilitates prostitution and casual sex. Botswana sends many of its workers to South Africa, and also sits astride major north-south traffic arteries. In addition, its relative prosperity has led to rapid urbanization, and cities provide greater opportunities for casual sex than rural areas. Another factor that has undoubtedly influenced the AIDS rate in Botswana is that the AIDS epidemic started in southern Africa, which is probably why the highest African AIDS rates are in the countries of southern Africa, which include Botswana. The epidemic had already spread widely before preventive measures, including education in the danger of the disease, were taken. But that of course cannot explain why Botswana's AIDS rate is higher even than South Africa's.
A further point about Botswana is that although its average income is high, because of its diamond mines and tourist attractions, the income is very unevenly allocated. Most of the population is very poor (30 percent are below the poverty level) and 20 percent are illiterate. However, a wealthy country with many poor can pay for public health, and in fact Botswana has an advanced program for combating AIDS by education, free condoms, etc., and has had it for a number of years, so it is a puzzle why the overall incidence of HIV-AIDS remains so high.
I want to emphasize two other factors that did not receive adequate attention in my original post. One is that short life expectancies reduce the cost of risky behavior: a 25 percent risk of dying from AIDS is very high, but the expected cost of death that is generated by that risk is lower the higher the probability of dying young from some other disease, a probability much higher in sub-Saharan Africa than in the United States. Another factor is the prevalence of other sexually transmitted diseases, such as syphilis, which increase susceptibility to infection by the AIDS virus. Both of these points are discussed in an early article by Tomas J. Philipson and me on AIDS in Africa--‚ÄúThe Microeconomics of the AIDS Epidemic in Africa,‚Äù 21 Population and Development Review 835 (1995)--and in a more recent article by Emily Oster, "Sexually Transmitted Infections, Sexual Behavior and the HIV/AIDS Epidemic," 120 Quarterly Journal of Economics 467 (2005). We note in our article the curious positive correlation in Africa of AIDS with education, and suggest that educated Africans are likely to be urban and therefore have more opportunities for casual sex. On the economics of AIDS generally, see also Philipson's and my book Private Choices and Public Health: The AIDS Epidemic in an Economic Perspective (1993).