In previous posts I have commented on the controversy surrounding "Why Nations Fail," the book by Acemoglu and Robinson. I have recently come across three new review articles. One of them is a negative comment by Microsoft's founder, Bill Gates, which has already been replied by the authors of the book. A positive comment is provided by Peter Coy. The deepest review so far (in my opinion) has been provided by W. Bentley MacLeod in the Journal of Economic Literature. MacLeod argues that previous criticisms of the book focus on the fact that the book promotes a uni-causal view of why nations fail to develop (based on the need for inclusive institutions) and that social life is too complex to be left to only one cause. But these other reviews do not seriously question that inclusive institutions are important. MacLeod instead argues that institutions sometimes have to be inclusive and sometimes have to be exclusive, and gives as example the literature on the economics of firms and organizations, where it is argued that participation by everybody at the same hierarchical level may fail to alleviate information asymmetries or solve problems of specific investments. The reviewer also argues that the book (and also the book Pillars of Prosperity, by Besley and Persson, reviewed in the same essay) shows that despite the usual homily that correlation is not causation, the authors establish correlations by showing historical examples (too many to keep track), but no causation. For example, for poor countries it may be optimal NOT to secure property rights, in the absence of complete insurance markets, while for rich countries it may be optimal to secure property rights. Then we may observe a positive correlation between secure property rights and national wealth, although it would be sub-optimal (and therefore bad policy advice) to recommend to poor countries that they should adopt secure property rights.