It is an excellent book for understanding why African economic development is so slow. I used it in our study abroad program in Burkina Faso.... Here's some comments I had for the students in the program after reading their "book reviews"....
A number of very good points were made by various persons and I expand on these a bit in what follows:
• Be alert to over-generalization. Collier is explicitly writing a book about a general problem- the plight of a billion people, living in 50+ countries. He boils down the source of the problem to the four (interrelated) traps. So of course he is vulnerable to the charge that he has over-generalized. A number of you pointed out the immediate example of over-generalization: Burkina Faso itself. BF is landlocked with bad neighbors, but still has grown steadily (albeit slowly); BF is not in conflict trap; BF is not in natural resource trap; BF is not really in bad governance trap by conventional measures. So what gives? (See Kevane and Englebert article, where we argue that entrpreneurial/managerial talent may be the constraining factor to rapid growth).
• Many of you noted the problem with the landlocked condition as a "trap", noting (as discussed in class session) that the comparison of landlocked countries with non-landlocked countries was misleading. The income gap between "landlocked" regions of the United States and the coastal cities of the US is, for example, quite large. Also, it is perhaps not useful to think of being landlocked as a "trap" in the sense that there is no probability of "escaping" landlocked-dom or no mechanism by which escaping generates a tightening of the trap (i.e. no Chinese finger trap).
• Many of you noted that the natural resource trap is really more like bad governance and civil conflict trap; the Dutch Disease itself is a more technical economic policy issue, and even then it is not clear that shifting country spending pattern towards urban construction and services sector is actually a bad thing for long-term economic growth in largely rural African countries.
• Some of you noted that the bad governance trap might be more accurately described as the corruption trap. Indeed it isn't clear how Collier distinguishes bad governance from corruption.
• As discussed in class, notably absent from Collier's analysis is a sense that maybe the poverty of the bottom billion results from the acts of the powerful G-8 (though he does nod to this here and there). And yet there is the slave trade legacy, colonial legacy, the Cold War legacy, the corrupting aid legacy, etc. Is it right to make the argument that the problems are all internal? And yet, Collier's "solutions" are largely directed at G-8 actions.
• Many of you took issue with Collier's civil conflict chapter, noting that some wars seem more complex than his greed hypothesis. Collier himself I think has retracted somewhat from his bald assertion, which was largely based on the statistical result that presence of natural resources explain some civil wars, while ethnic cleavages and measures of political oppression did not. But these statistical results have been strongly criticized (the same things that cause political repression presumably cause civil conflict, so statistical causality is hard to disentangle).
• What does it mean to be falsifiable? In social science, a hypothesis is falsifiable if there is evidence that would show the hypothesis to be false. In Collier's data- the performance and condition of countries over time- there is no "experiment" possible (i.e. one cannot randomly apply traps to countries, to see if they can get out of them) so in some sense the hypotheses are not falsifiable....
