Tuesday, February 28, 2012

The Economist's Delusion: Money Alone Can Stop HIV

A drug [for HIV] called money, says The Economist headline. According to the article, money is the answer to reducing HIV transmission. The reason why money is the answer? Because the article assumes that most HIV is transmitted by young girls having sex with older men in return for money. There are a number of flaws in this thesis; for example, not all HIV is transmitted sexually, not all sex is transactional sex, not all poor girls are prostitutes, not all 'older' men pay younger girls to have sex, most 'older' men are not HIV positive, etc. So why would giving money to the girls be a 'drug'?

Well, a 'scientific' paper says so. The paper concludes that "Cash transfer programmes can reduce HIV and HSV-2 infections in adolescent schoolgirls in low income settings." Girls or parents in an intervention group were given varying amounts of money; some had to show that they were attending school; others didn't have to meet any conditions. But there are also a couple of flaws with this paper; it is not known how many of the girls were already HIV positive when they joined the program and, out of the girls who became infected, it is not known whether they were infected through sex or what kind of sex or whether they were infected through some other mode of infection, such as unsafe healthcare or unsafe cosmetic practices.

The paper starts off with the claim that "Lack of education and an economic dependence on men are often suggested as important risk factors for HIV infection in women." But what are these suggestions based on? I agree that many people in high HIV prevalence countries lack education and are impoverished, and often dependent; this may be particularly true of women and girls. But does much of the research data currently available, for example, the kind of socio-demographic data collected in Demographic and Health Surveys, give reason to think that poverty and low levels of education are correlated in such a way that there may be some kind of causality involved?

In fact, the contrary is often true. The relationship between economic status and educational level on the one hand, and HIV prevalence on the other, is often very unclear. And where it is clear, HIV prevalence tends to be higher in the wealthier and better educated quintiles. Fair enough, correlation is not causation; but lack of correlation does not support the sort of research that went into the above scientific paper. Only the unsupported assumption that HIV is always or almost always transmitted through heterosexual sex in African countries allows such research to be carried out, and such conclusions to be drawn from it.

Those who received money had one third of the risk of testing positive for HIV and one quarter of the risk of testing positive for herpes simplex virus, compared to those who did not receive money; there was no difference between those who had to attend school and those who did not. Among those who had already dropped out of school, there was no difference between those who received money and those who did not. Despite the small numbers infected among school-going girls, it is concluded by The Economist that money is a HIV 'drug'. What effect did the cash and/or the conditionality have on HIV transmission? We really have no idea. And if it did have any effect, it was small.

The Economist exaggerates shamefully: the researchers "conducted a randomised clinical trial of the idea that money, and money alone, can stop the spread of HIV." This allows the author to indulge in Western-style fantasies about African sexuality and money-grabbing women, sugar daddies and HIV. And it is concluded that bribery works when it comes to reducing the risk of HIV transmission. Would they also conclude that we need a sliding scale so that we bribe wealthier people with larger sums of money and poorer people with smaller sums of money? What about if people become infected with HIV, would they then be entitled to a larger bribe in return for not risking transmitting the virus to anyone else?

A good deal of Demographic and Health Survey data shows little correlation between unsafe sexual behavior and HIV prevalence. For example, those who sometimes use condoms can be more likely to be infected than those who never do. Circumcised men, despite all we may read on the subject, are often more likely to be infected than uncircumcised men. A recent paper even showed that paying for sex was inversely associated with HIV infection. There isn't even a simplistic relationship between sexual behavior and HIV, let alone between money and HIV.

The Economist may misrepresent The Lancet paper, but the latter is crying out to be misrepresented. It plays into the hands of those who hold insultingly racist and sexist views about Africans, especially those who are HIV positive and even those who live in high HIV prevalence countries. To carry out research like this you must already believe in a causal connection between HIV and transactional sex. All those involved in these exercises have demonstrated is the institutional racism and sexism that is to be found throughout the HIV industry. Of course, this is not to say that those directly involved are themselves racist or sexist; they may not be.



Anonymous said...

Are African HIV statistics exaggerated?


Simon said...

I imagine statistics are often exaggerated or misused, not just in Africa. It's true that many of the modelled figures for HIV prevalence and death rates were too high, even 10 years ago. But estimates were revised downwards a number of years ago, since the article you link to was written. Whether a country has less than 1% prevalence or more, the virus is still very serious and has major consequences for people, not just health consequences. My argument is that HIV and other health issues will not be solved by simplistic measures such as cash transfers, whether conditional or unconditional. Especially not when the distribution of transfers is based on the view that HIV is almost always transmitted sexually.

Petit Poulet said...

You can't help but see the cultural colonialism going on here. The editors of the Lancet and the Economist feel the need to save these people from their ways by giving the girls money, genitally mutilating the boys, and swaying them from their evil sexual practices. As you have mentioned elsewhere, these interventions are a salve for those who donate the money, but do little be confuse the issues on the ground.