I listened to this on NPR a few minutes ago about a study on whether aid helps reduce poverty. The researchers took a group of the poorest families in 6 developing countries and gave half of them a bunch of stuff- livestock, food, training, health care, and a saving account, and the other half NOTHING. Then they compared the two groups to see if the aid made a difference.
The families who got the aid did do better in 5 of the 6 countries, and continued to do better a year after the aid was stopped, although the difference wasn't significant- just 5% on average.
I think this is a fascinating study, but I'm not sure I can get behind randomly giving some desperately poor families nothing when you give other families a lot of help- especially the health care and food. The researchers feel that it's okay because this type of research can help improve aid efforts in the long run (something that is so very desperately needed), but these are people, not statistics. And there's so much more to measure than just whether the families were doing better financially, like children whose lives are saved because of better medical care.