Global Health Policy

I also enjoyed this article by Oster in Esquire a while back, Three Things You Don't Know About AIDS in Africa.

The concurrency hypothesis has an important policy implication which is little discussed, and did not come up at Dr. Epstein's seminar.

Many people I've spoken with appear to believe that it means nothing for policy, since it is practically impossible to change societal attitudes about the acceptability of long-term concurrent partners. Even if that's true, the concurrency hypothesis has a very different and very important policy implication: Simply urging people to be less promiscuous can actually hasten the spread of the disease.

How? Because concurrency and the number of partners can be inversely related. Suppose a man has nine sexual partners over the course of a year, only one or two of them concurrently. Upon hearing a public service message that promiscuity is bad, he limits his annual number of partners to three -- the same three, all year. But if he has roughly the same number of sexual contacts over the course of the year, then his concurrency rate has probably gone *up*, even though his number of partners went down. He believes he is safer. But if everyone in the society does the same thing, then, according to the concurrency hypothesis, everyone is more at risk. This is not just a possibility; in general, if people reduce their number of partners, my prior would be that people don't much reduce their frequency of sexual contacts, meaning that concurrency would probably go up.

*If* the hypothesis is right, then, there is indeed a policy implication: the message that governments and activists give to the public need to be much more nuanced than simply "be less promiscuous". It can't be ruled out that such a message could cause deaths.

Please also check out FRONTLINE/World's new story that mirrors Epstein's conclusions about Aids and condom use in Uganda.

The piece is visual and engaging and free to watch on the web at:
http://www.pbs.org/frontlineworld/rough/2007/07/uganda_the_cond.html

Thanks,
Charlotte Buchen
FRONTLINE/World

The Washington Post article that Mead cites offers a compelling story of how economic activity can affect behavior and HIV prevalence. That story and Mead's interpretation of Oster's results, focusing on the income effects of changes in exports, are more convincing to me than the story that Oster herself tells. She focuses quite narrowly on the potential effects of exports on the volume of trucking, which increases the opportunities for men to have sex with partners that are more likely to be infected than when they are home. There is almost surely something to that hypothesis, but the magnitude of the effect that Oster finds seems likely to be overstated for several reasons.

First, the same volume or value change in exports would not be expected to have the same effect in an economy, such as Uganda, where exports are 10 percent of the economy, as in one, such as Namibia, where exports account for nearly half of GDP. Second, the hypothesis, as Oster notes, is based on changes in the volume of trade. But for reasons of data availability, Oster also uses dollar values of exports and the results using the value measures turn out to be more robust than those using the volume measure. It should be noted that the results for Uganda using the volume measure are at the low end of her range, 30 percent, rather than 50-70 percent. Third, Oster uses an average of the current and prior year's exports "to capture the possibility of a lag in the relationship between exports and HIV," but she does not explain why such a lag might exist and it seems inconsistent with her theory. Oster reports the results for each export year separately in an appendix and the export volume measure in the year when HIV incidence is measured is much smaller than when the average is used and only marginally statistically significant. There might be more of a lag effect between HIV and higher incomes due to increased values of exports and that could be what she is capturing, rather than the effect of an increased volume of trucking.

You might be interested in this piece about cooperative work on an AIDS test for CD4 count by two companies, one university and a research institute.

http://insideinnovation.wordpress.com/2007/07/26/aids-fighters-in-the-lab/

All the best,

Colin Stewart
Columnist
O.C. Register

I am not an infectious disease expert nor have I done any research on AIDS/HIV. Having said that this confirms my long held belief that concurrency of sex partners contributes significantly to transmission rates. My oldest brother died from AIDS and I remember my mother chiding him that his behavior with all these women would lead to his demise. When my brother died in 1992, I vividly recall my mom lamenting with frustration that if only my brother had listened he would still be alive. I was in Zimbabwe for 2 years (5 years ago) and I was frustrated to learn that behavior education programs did not attempt to address this problem. However, I would be the first to admit that I also observed a fatalistic attitude “what’s not to kill you”. In the face of hopelessness, it looks like sex is the one thing people can still control. Sadly, this scenario is repeated many times over in Southern Africa and there doesnt seem to be an end in sight.

I find that Epstein's hypothesis and Oster's are compatible and still consistent with the notion that sex workers are a key target group. Concurrent partnerships are not risky if the network of partnerships remains closed. This is the case in many polygamous households. When the concurrent network is opened to occasional partnerships with, for example, an infected sex workers, the infection spreads more quickly through the infected partner because contacts will inevitably occur when the newly infected partner is at the stage of high viral load. The "hub" of concurrent partnership that are opened to infections is more often than not an "MMM" a mobile man with money. The number of MMM's and their propensity to travel and spend money for concurrent and occasional relationships can easily be seen to correlate with exports and economic activity. It still makes sense to target sex workers (and for that matter, young women/girls involved in transactional sex) not because they are to "blame" for the spread of HIV, but because it is probably easier to convince them to see MMM's as risky partners with whom to insist on condom use than it is to convince MMM's to reduce their relationships or increase their condom use.

I am neither a researcher nor an economist. I am a development worker, indigenous to Uganda - founder of "AIDS ORPHANS EDUCATION TRUST/ Action for Empowerment" - an organization that now has Office and operations in Uganda, Kenya, Zambia and Rwanda (www.aoet.org).

I've read with a lot of interest comments in this forum and many others concerning the decline/increase of HIV infection rates in Africa!!!

In her new book "The invisible Cure: Africa, the West, and the fight against HIV/AIDS", Helen Epstein raises a few issues that as an African, and a Ugandan in particular leave me studying with a lot of interest how conclusions are made while searching for causes of increase and declines of HIV infection rates!!!

1. For Emily Oster to say that INCREASE in mens’ incomes as a result of increased national exports translates into increase in HIV infection rates is an attempt to keep African economies in poverty. (I say this based on Dr. Helen Epstein's title of the book - "The invisible cure..." and her debate that follows.

In response to Dr. Helen and Ms. Emily's findings, I would like to make my contribution with the following debate:

Uganda being a peasant agricultural economy, VERY few Ugandans produce for export. I work in MANY districts of Uganda, Kenya and Zambia and I know that many people in these countries produce for home consumption and many live without money - and surprisingly, according to us as an organization we find A LOT of infection rates in this category of people.

2. I am not sure whether Ms. Emily is talking from an urban or rural (where the country's biggest population is) point of view!!!
Most researchers use Land Cruisers and 4X4 Jeep-type vehicles to collect data, but when you go to where these vehicles can't go, deep rural areas, you will find that infection rates are getting higher. These are not necessarily as a result of INCREASED exports, yet this is where the biggest population is.
If indeed more money meant increased infections, the West would be having LOTS of new infection rates due to Tens of thousands of tourists that come to Africa from the west and have the money to give to casual partners. A good next step for a research would be to find out what the situation is with this category that would be the same as TRUCKERS and traders in most African countries that have the money and would be admired for concurrence).

3. Also, to say that the increase in Men’s’ pocket money is an increase in concurrent partners, which leads to an increase in HIV infection rates, says to me that increasing women’s’ incomes (making them self sufficient) will reduce dependency on men for provisions which reduces infection rates!!!
Sadly, in our experience as an organization, and my experience as a development/social worker, this is not true.
PLEASE do not misunderstand me. I fully agree that concurrent partners create a high number of deaths.
My argument is whether these are economically driven!

Please also note that the BIGGEST number of commercial workers in Uganda and Kenya use condoms yet this is where the biggest display of sex for money is noted!
Generally speaking, the biggest number of people that will be after money for Sex will also be exposed and educated enough to use condoms!!
Therefore, for the case of Uganda, and indeed for most of sub-Sahara Africa, I would be careful before concluding that increase in exports means increase in HIV infections.
In my view, increased HIV infection rates have far more reaching social than economic causes.
I would attribute the biggest percentage of falling infection rates to community based HIV/AIDS education - where people come to realize the danger of concurrent partners, or casual sex... and less on having economic means to get sex or afford concurrent partners.

Uganda's example speaks for itself:
Almost everywhere you go in this tiny country - Uganda, you will find Community HIV/AIDS education!
As an organization, I remember several members of our staff taking 14" monitor TVs deep into rural areas in the late 90's - into Schools, community centers... teaching people about the spread and prevention of HIV...
MANY other organizations have been doing this for years in Uganda.

Increased exports may have an effect on HIV infection rates, but in my view would not be significant enough to be represented here as a major cause - especially taking it as far as Africa as a continent or sub-Sahara for that matter is concerned. More research is needed - especially with assistants from the very rural areas we are trying to collect data from as opposed to city assistants.