I recently viewed a video of a presentation given by Esther Duflo at PopTech 2009.  (Esther Duflo is a development economist who has garnered significant attention for her work on randomized evaluation.  She is based at the Abdul Latif Jameel Poverty Action Lab at MIT, which she co-founded.)  What I found most intriguing about her commentary is the way she highlights evaluation as a way to figure out what works in development.  She starts with what we all know to be true: development problems are big, intimidating, and seemingly intractable.  But fortunately, she doesn’t stop there.