I appreciate the insightful discussion about who is right, and some more cheers for use of RCTs for all types of evaluations - good stuff for students in econometrics class. But, aren't we missing the big issue here? As development economists, don't we have the responsibility to be careful in producing results that may affect millions of lives? What were the direct and indirect implications of PK or RM's papers? Imagine a situation where most donors and policy makers really believed in these results and put them to use? It would have been a flip flop of program / policy on women focused issues. The losers would be the poor women in developing countries - not PK or RM - who have their jobs and fame.