Collier and data mining

I am happy to report that my post from a few days ago on the problems of causality in Paul Collier's new book has received some attention in the econ blog sphere. MR linked to it, and Aid Watch delves much deeper into possible data mining.