Martin Wolf  asks: is aid petrol on a fire, or water on a plant?
There are at least two negative views of aid (needless to say there are many positive ones, too). Both parallel views on the famous 'resource curse': few resource-rich developing countries have grown strongly over the past four decades, while many resource-poor countries have done so. Angola, Nigeria and Venezuela compare poorly with China, Mauritius and Thailand.
The first potential problem (of both aid and mineral resources) is that resource flows can make other sectors of the economy uncompetitive, especially the sectors that lead to learning by doing and technological progress. See What Undermines Aid's Impact on Growth?  by Rajan and Subramanian of the IMF - a paper that has created a big stir recently.
An alternative source of pessimism is that resource flows may corrode the quality of political or economic institutions, by promoting corruption and isolating governments from their responsibilities. Current or former World Bank staff have taken a leading role in this research. See The Curse of Aid , Aid Dependence and the Quality of Governance , and Is Aid Oil ?
My take? Both risks must surely be taken seriously, although hardly amount to a knockdown argument against aid. They also explain why project-level evaluations are often so positive, while macroeconomic attempts to measure the connection between aid and growth fail to convince.
It's hard to argue with Martin's proposed solutions, which include:
Fifth, [donors] should give a part of the resources directly into the hands of citizens. Finally, they should promote an environment conducive to a productive private sector.
For more, see Aid and the Resource Curse , a short note I wrote with Michael Klein.