As you may well know, foreign aid is fairly controversial. This isn't just because people don't want to be charitable. Unfortunately, a wide range of studies have found aid does not have a statistically significant relationship to economic growth. That said, there's two important corollaries: 1) aid may still make recipient's lives better in other ways and 2) there may be specific types of aid that, when appropriate to the country in question, work well with some consistency.
Another argument is that trying to help is generally a waste of effort because poorer nations can't be helped. There's a new book out on the topic that has gotten a lot of buzz called Farewell to Alms by Gregory Clark that reached the controversial conclusion that Britain went through the industrial revolution because for decades beforehand the rich had more surviving ancestors than the poor. The actual will study data seems pretty sound, but I tend to buy the argument by David Warsh that this book goes for the Guns, Germs and Steel level of scale without nearly enough support.
So, for those of us rejecting the idea that things are hopeless, here's two possible approaches:
So the real debate is not about whether aid works or not, but about (a) under what circumstances it actually works; (b) how it can be reformed, in principle, to become more effective; and (c) how likely is it that the requisite reforms can in fact be undertaken. The disagreements among Sachs, Easterly, Subramanian, Birdsall et al. are about these questions, but they are often left implicit in the discussion.
And Arvind Subramanian's:
The real issue is figuring out the most effective ways that the rich world can help in boosting living standards and improving the other conditions of underdevelopment (poor health, education, sanitation etc.) in developing countries.... For example, many demonstrably effective ways of helping the poorest, such as financing research to create new agricultural technologies for Africa, have been, and remain, neglected. More tellingly, even if better ways of helping have not been pursued, rich countries have flouted even the Hippocratic rule of doing no harm—and that too on an issue such as improving health and saving lives—while being ostensibly generous with providing aid. I have in mind here the WTO’s intellectual property rules (TRIPs), legislated into being by rich countries, which significantly impede access to low-cost drugs.
I tend to think this is a pretty productive trade space for figuring out to help. However, this economics debate is a bit above my level. My main current thought on aid stems from an excellent talk by Paul Collier on his new book the Bottom BIllion which discusses state-level factors that keep a billion people stuck in absolute poverty (less than $1 a day purchasing power parity). One big factor is that a lot of states, particularly in Africa, don't really have viable borders. This is also a big driver of conflict and is a direct result of European colonial policies.
So what do we do about that sort of thing? Well Peacekeeping and such can be part of the solution, but simply avoiding conflict won't suddenly make these nations viable. On the whole, since WWII, since the founding of the UN, and particularly since the end of the cold war there's fewer wars and casualties of war. But in the absence of conquest there aren't really a lot of mechanisms available for creating viable countries. Simply clarifying borders won't suddenly solve the problems of a landlocked poor country without trade able resources. Moreover, there's pressure in places like Kosovo to create more states that probably aren't economically viable.
Absent massive refugee flows, I think we're going to need some new tools. The two basic options are: ceding territory to neighbors (which states are always loath to do) or increased political integration (which helped in post-WWII Europe, but has run into snags with the EU). If we want more economically viable states without returning to the bad old days of inter-state war, we need to figure out how to facilitate one or both of these options.