When asking people why they think HIV prevalence is so high among the Luo population of Nyanza province in the West of Kenya, they give all sorts of explanations that have been hypothesized over the years. Wife inheritance is one of the most common, although some say this doesn't happen so much except in rural areas and is on the decline. Some blame it on poverty and lack of economic opportunity, even though HIV prevalence has often been more closely correlated with wealth than poverty. And even still, HIV prevalence tends to be more closely associated with people who are formally employed than those who are unemployed or who subsist in some kind of informal economy. Which suggests that 'being idle', a commonly heard explanation of high rates of HIV transmission, is not really an explanation of high HIV prevalence.
The latest Ugandan Aids Indicator Survey (AIS) results have not been released yet, even though they were collected in 2011. A preliminary summary has been published, but other bits of data are being drip-fed to journalists, it seems, and released in newspaper articles every now and again. Certain groups that are favorites among those forming hypotheses and speculating about how the virus is spread have sometimes claimed that house girls are infected in large numbers and that they can transmit the virus to the father, who can transmit to his wife, and then the house girl can also infect sons in the house. I remember seeing a HIV drama enacting this very scenario in 2002. I have never seen any evidence that it could account for a substantial proportion of transmissions. But the results from Uganda find that those in domestic employment are one of the lowest prevalence occupational groups, at just under 7%.
Many early articles on HIV, and a lot of the policy documentation still going around, point the finger at long distance lorry drivers and others involved in the transport sector. But the Ugandan AIS estimates HIV prevalence to be 7% in this group, similar to prevalence in the country as a whole. It's also similar to prevalence among those working in real estate, which is estimated at 8.6%. Prevalence among those working in the hotel and restaurant industry is estimated to be 10.3%, which is a bit higher than the national figure. But some of the earlier literature claims that a lot of women working as waitresses or 'bar girls' are generally sex workers or engage in sex work some of the time. Perhaps they do, but it sounds like they don't do so as often as the official view of HIV would have it.
But one of the top explanations given for high HIV prevalence among Luos is the fishing industry: one can often read of a practice whereby women who sell fish have to sleep with the fishermen before they get the first choice of which fish to sell. While this practice may exist and may be deplorable to some, it seems it may not account for as much HIV transmission as we have been led to believe. The Ugandan AIS estimates that prevalence among those in the fishing industry is only 7.1%, which is not extraordinarily high. Looking at all these figures, it is tempting to suggest that the sexual behavior explanation of HIV transmission may sometimes lack explanatory power, that a significant proportion of HIV transmission is non-sexual.
The highest figure by occupational group that I have seen is for those working in public administration, which stands at 16.3%, well over double the national figure; also around the prevalence figure for Kenyan Luos in Nyanza province. If unsafe sexual behavior is the explanation for high prevalence among Kenyan Luos, is unsafe sexual behavior also the explanation for people working in public administration? And what sort of sexual practices exist in the Ugandan public administration sector that could account for this very high figure? Or is there another explanation entirely, one that is not purely about sexual behavior? As I haven't seen the complete AIS, I don't know how these figures are to be reconciled with the received view of HIV transmission. But I'm sure those collecting the figures have a whole slew of epidemiologists at their disposal to explain away anything that looks anomalous, anomalous meaning in any way inimical to the theory of the African who does little but engage in unsafe sex with as many people as possible, for as much of the time as possible.
Having followed newspaper articles and other documentation about HIV over the years, I'm not surprised that, as I go around interviewing people, the same explanations of HIV transmission appear with alarming regularity. People are mystified and if it wasn't for these explanations, which often have a certain exotic quality, they might just resort to equally exotic, but equally implausible, explanations of their own. But each of the much loved explanations, that depend on the belief that most HIV is transmitted sexually and that Africans have lots of sex, collapses if you start to look for evidence. And like other stereotypes, biases, myths and various kinds of pseudo explanation, they then continue under their own inertia, no less loved for being entirely without foundation.