The AIDS/malaria puzzle: bleg
A few months ago, I posted to Facebook a puzzle about the difference in funding for AIDS and malaria. Here is the puzzle:
DAH spending for AIDS is much more than malaria: Development assistance for health (DAH) numbers from the Institute for Health Metrics and Evaluation (IHME) show that development assistance spending on AIDS significantly exceeds spending on malaria. For instance, if you click on the link and switch to “Trends” and “Health focus” you’ll see that AIDS DAH spending in 2016 was estimated as $9.5 billion whereas malaria DAH spending was estimated as $2.5 billion. Most of this difference comes from government spending (breakdown by source also at the link). In fact, total annual spending by the US government on HIV/AIDS is around $33 billion, more than the National Institutes of Health (NIH) budget, though most of it is domestic spending.
Effective altruists and allied groups have focused much more on malaria than AIDS throughout their history: Malaria has been identified by GiveWell as a promising area since 2006, and the Against Malaria Foundation has been a GiveWell top-rated charity since 2011, excepting one year. Giving What We Can has also recommended the Against Malaria Foundation since before it became a GiveWell top charity, and it has also been the poster boy of effective altruism for fundraising groups like The Life You Can Save and Charity Science. In contrast, HIV/AIDS hasn’t been a major focus, with GiveWell getting around to reviewing a HIV/AIDS-related intervention only in 2017.
Crude estimates of the toll of the two diseases paints a picture of fairly comparable impact: Malaria affects five or more times as many people as AIDS. But on the other hand, once you get AIDS, you are stuck with it, whereas you can usually recover from malaria in a few weeks. On the third hand, the agony per unit time of having AIDS is lower than that of malaria. The annual death toll of AIDS is about double that of malaria (a million versus 400,000), though estimates for both have huge error bars.
Interestingly the Gates Foundation, which can be considered intermediate between a government donor and an “effective altruist”, has an AIDS/malaria spending split in between the two: it spends roughly equally on the two; see breakdown of funds for grants in developing countries.
The tension between (1) and (2) is an interesting puzzle. It could be that:
DAH spenders are wrong about their focus on AIDS, and in an ideal world would be directing more resources toward malaria.
Effective altruists are wrong about their focus on malaria, and in an ideal world would be directing more resources toward HIV/AIDS.
They are both right “in their own way”; HIV/AIDS spending is the right call to make for DAH spenders whereas malaria spending is the right thing to do for effective altruists. While the most conciliatory to all sides, this also demands the most explanation, since relativism challenges some of the implicit and explicit ideas of effective altruism.
I have explored some more specific hypotheses in a comment on my Facebook post, which I shall not repeat here for brevity.
I’ve already spent a fair amount of effort collating the history of malaria, including funding a bunch of malaria-related timelines such as timeline of malaria, timeline of mosquito net distribution, timeline of the Global Fund, timeline of malaria vaccine, timeline of Against Malaria Foundation, and timeline of malaria in 2014, 2015, 2016, and 2017. I intend to spend similar effort on HIV/AIDS, and return to the puzzle after that. However, I’m curious about any thoughts readers here have on the puzzle, including whether you find it interesting, potential resolutions or directions to explore, or refutations of the premises of the puzzle.
Thanks to Sebastian Sanchez and Issa Rice for working on the linked timelines. Thanks to Howie Lempel for commenting with thoughts on my original Facebook post. And thanks to IHME for collating Development Assistance for Health (DAH) spending, looking at which inspired this post.
The vast majority of large institutional spending is somewhat static. When there have been major shifts, it is usually in response to the combination of highly successful marketing campaigns and new events.
Malaria has been largely ongoing, without much newsworthiness (to regular media outlets) or specific press. It’s funding therefore is likely to have stayed at a somewhat static level in most organizations.
In contrast, HIV/AIDS was emergent in previous decades. It went from nothing to being highly prominent in a short time period. Relatively large budgets were allocated against it because:
It showed a pattern of significant growth, and there was significant fear that not containing it could lead to runaway growth.
It emerged from 0 cases to being prominent, which was highly newsworthy.
There was a strong coordinated marketing campaign to get governments and IGOs to strongly address it.
HIV/AIDS funding came at a relatively high level as a result, and because funding is largely static and the problem remains, it has stayed that way.
The reason historically is that bed nets seem to deliver more improved health per dollar (one QALY <$100) than scaling up delivery of ARVs (which seem to cost a few hundred dollars per QALY depending on what study you look at). ARVs have gotten cheaper, but not yet enough to beat bed nets as far as I’m aware.
If you were funding biomedical research HIV may well look better, because the damage done by HIV is increasing, while malaria is gradually being beaten back. But most EA funding hasn’t gone to medical research due to the major challenges that individual donors trying to support that kind of work.
For some of the research prior to starting Charity Science Health, I recall looking at two HIV interventions and ending up not that impressed. We summarized some of the research onto this grid.
Antiretroviral therapy ended up noticeably less cost-effective than our other interventions. That might favor an interpretation for DAH spenders being wrong.
On the other hand, prevention of mother-to-child transmission of HIV seemed pretty cost-effective, but the field was quite crowded already with a lot of pre-existing organizations working in the area and seeming to do quite well. This might favor a “both right” interpretation, if we assume that DAH funders has already used up all the room for more funding that Givewell / OpenPhil / EA would have used.
I think, in this type of analysis, for an infectious disease, it’s really important to look at potential for spread as well.
Malaria is region-constricted (only places with the right mosquitoes), whereas HIV is not. Therefore, there’s a natural cap at the amount of malaria we can have if malaria control ceased to exist, whereas HIV’s ‘natural cap’ is potentially “all susceptible humans”.
If you include “all future infections” into the analysis, how much suffering due to HIV can be avoided due to current efforts to control HIV? I mean, you can sort of see this in a natural experiment created by South Africa’s HIV denialism − 18.5% of the population there is infected, compared to 6% of Kenya and 3% of Nigeria, despite both Kenya and Nigeria having lower GDP/capita than South Africa. There’s an article on the costs of HIV denialism in SA here. Obviously, societal dynamics are different in SA than other places, but 3x the amount of HIV is a pretty significant number.
Anyway, the CBAs on interventions like promoting condom use, testing services, education campaigns, and such are (obviously) difficult to do, but that… really doesn’t mean we shouldn’t be funding them.
As for agony of HIV over malaria—are you sure? Does this include the ‘psychic’ cost of HIV (mental stress, stigma, constrained social mobility, shunning from society/friends/family) along with the physical cost?