I have not read Prof. Collier's book yet, and am not looking forward to doing so. "The Bottom Billion" must be one of the most over-rated books of this century (i.e. since 2000), and the new book can only be an improvement. I spent quite a bit of time looking into basic data that went into Collier's statistical manipulations. Every time I looked at something, it turned out to be either questionable or downright wrong. Collier's method, pretending to have useful statistics about someting (e.g. "governance"), doing a series of complex statistical analyses, and then stating the obvious is a travesty of what real economics and social science should be about. He reduces everything to simplified or plain useless numbers (e.g. "length of a civil war"), and then tries to weed out the cross-correlations so as to get to some sort of pure analysis. Sorry, the cross correlations are what real social science is about. It is making sense of the magnificent complexity of human economies and societies that is our job as real social scientists. Obviously statistics have their place, indeed they are very important, provided that they are treated with appropriate skepticism. Unfortunately, Collier is simply the most prominent of what appears to be a whole generation of economists who have substitued statistical analysis for judgement and insight.