Short but compelling case for bottom-up development, meaning letting poor countries decide for themselves how to allocate development resources.
Make development results-based rather than consultant-based. I'll leave it to Jeffrey Sachs to explain far better than I could:
Rather than have Washington (Penguin: or anyone else) decide the kind of aid each country will receive, the recipient countries should be invited to prepare plans and budgets that would be reviewed by independent experts. These plans would describe the inputs needed by the farmers, the expected increase in production, how the strategy would be put into place and how much money would be required. Such plans, if described with care, could then be closely monitored by the United States and other donors to gauge results and avoid corruption.
Two international programs during the last decade, championed jointly by the United States, other governments and the Gates Foundation, have demonstrated the benefits of such a scientific, results-based aid approach: the Global Alliance for Vaccines and Immunization, and the Global Fund to Fight AIDS, Tuberculosis and Malaria. These programs have saved millions of lives and protected hundreds of millions more from disease and infection. Here’s how they work: Low-income countries submit national action plans to the two programs, which then scrutinize the plans on their scientific, financial and management merits. If the plans are properly put into effect, recipients get more financing.
I would even take it one step farther and in certain cases, only where is makes sense, have outside investors come in to provide either loans or risk capital. When done properly (ie, this is NOT a peanut butter solution that can be spread over every problem) but when done properly, local for-profit entities working on local social problems could benefit from insights and knowledge of outside investors.