Feynman's Question

Don't forget that Feynman was the one who ignored the politics on the Rogers Commission and brought the O-Ring problem with Challenger's reusable booster rockets to light. He didn't discover it -- he was a physicist, not a rocket engineer. He just asked enough questions of the right people until he unearthed an answer that NASA management buried. It's a real shame he wasn't around for the Columbia disaster, but perhaps a greater shame is that science could only produce one Richard Feynman in the last 50 years.

I do not think that development economists were asking these sorts of questions properly in 1983, much less in 1963. I am not sure that they are asking them properly today.

I do not think Arnold has been following the development literature well. You probably came to this conclusion from only reading Easterly's books.

Of course Arnold is right. Feynman was genius, but he was too optimistic in implying that the scientific method was being applied in development policy. In fact, it took 50 years after Feynman's comment for the development establishment (the World Bank, the Interamerican Bank of Development, FAO, USAID, etc.) to take notice that their policies were not working, the technologies being transferred were not taking root and that their monies were being stolen. So they are now into institution building and promoting transparency. Of course no preliminary scientific testing was done to verify if it is actually working or workable. But then it would be necessary to recognize that there are no new ideas, that nothing seems to be working and close the shop.

Certainly Arnold Kling (via Cato) is aware of Peter Bauer. Does his work not bear mention?

In his book http://www.grameenfoundation.org/resource_center/books_and_publications/ Muhammad Yunus describes more methods than microfinance that have helped many desperately poor people raise themselves out of poverty.
He describes how he and his co-workers at the Grameen Bank have been unendingly experimental in developing and refining those methods.

(Yunus got his Ph.D. in economics from Vanderbilt University. He was a professor of economics at Chittagong University in Bangladesh when he began his field research and field trials.)

It's development economics that works. However, the Grameen Bank is not part of any government.

Sorry about that.
The book is Banker to the Poor

There's two essential elements of the scientific method that are missing from almost all foreign aid efforts, whether governmental or NGO: measurement and feedback. No-one measures whether project A is having the desired (or any) effect, and even if they did, there is no mechanism for communicating that measurement back into the process. It's like doing an experiment in the lab over and over and not noting down the results. Another blind spot that the aid agencies have is their inability to cherry pick what worked from previous effort and apply it to new projects.

In short, even the alchemists had more of the scientific principle in their work than these guys.