More from one of cjbprime's sources: GiveWell does a lot of evaluation of charities, trying to find where an individual donor's dollar can make the most difference. They hold donations to GiveDirectly to a pretty high standard, comparing it against stuff like antimalarial bednets and deworming, and think it comes out looking alright. In addition to cjbprime's link, they've written:
2) The Case For Cash--posted after their recommendation of GiveDirectly, talking about how a one-time cash transfer can have long-term benefits and how giving to the third-world poor differs from giving to the U.S. poor: http://blog.givewell.org/2012/12/26/the-case-for-cash-2/
They produce a lot of thoughtful research with a lot of effort to ground their recommendations in reality--browsing around their site a bit is a pretty rewarding use of time if you're interested in this kind of thing.
> They hold donations to GiveDirectly to a pretty high standard, comparing it against stuff like antimalarial bednets and deworming, and think it comes out looking alright.
There's this line of reasoning (coming from African economists that have spoken out against the latter two forms of aid[1]) that if you give the poor money, they can spend it on antimalarial bednets and deworming themselves, which is is effectively the same as the former but with the added benefit of bootstrapping the local economy (provided the facilities are there of course, but that is more likely to develop if the local population has money...).
Of course that's a gross oversimplification of a complicated problem, but it's an interesting counterpoint to traditional forms of aid.
I think there are a couple of general reasons bednets or deworming (and vaccines) are still good things to deliver via aid program rather than sell, without taking away from the awesomeness of GiveDirectly. (And I know you're saying you know such reasons might exist, but I think it's worthwhile to go into them:)
One is logistics: a program to reach everybody unlocks some economies of scale. Much of the cost of these programs is just getting supplies and people to all of these remote villages with impassable roads, the occasional corrupt official or militia, etc.[1] Some of that is effectively fixed cost, and doesn't vary much whether you have two boxes of deworming pills in the back of your truck or 20. The cost per person is lower if you can amortize those fixed costs across everyone.
The other is that sometimes active encouragement is worth it. Even rich countries have policies to try and get folks to vaccinate and for other health goals like getting people off cigarettes or fighting obesity. In poor countries there are two further reasons a push could help: there's not enough accurate, trusted medical info (not even high school bio for everyone; that's part of how "traditional healers" survive), and the cost of a vaccine feels a lot higher.[2]
All that said, I get the frustration that there are so many big problems aid like this _doesn't_ directly help with. There's partly just a problem of scale--you need a lot more resources (whether investment or aid or what) to deal with the lack of infrastructure, etc. But I do think in the meantime, the science and to some extent plain old arithmetic indicate that basic health programs are still doing a ton of good.
[1] Digression, but if you want to read about what it's like to take the long way to a remote old tin mining town in the DRC, http://www.amazon.com/dp/1851689656/ is amazing
1) Cash transfers intervention report--talking about where recipients spend the money, GiveWell's worries about transfers, and studies on the topic: http://www.givewell.org/international/technical/programs/cas...
2) The Case For Cash--posted after their recommendation of GiveDirectly, talking about how a one-time cash transfer can have long-term benefits and how giving to the third-world poor differs from giving to the U.S. poor: http://blog.givewell.org/2012/12/26/the-case-for-cash-2/
3) Responses to objections on cash transfers--mostly about more in-the-weeds issues, like how to interpret their numerical comparisons among cash, bednets and deworming: http://blog.givewell.org/2013/07/31/responses-to-objections-...
They produce a lot of thoughtful research with a lot of effort to ground their recommendations in reality--browsing around their site a bit is a pretty rewarding use of time if you're interested in this kind of thing.