The reversal of fortune thesis reconsidered

Acemoglu, Johnson, & Robinson (2002) have claimed that the world income distribution underwent a "Reversal of Fortune" from 1500 to the present, whereby formerly rich countries in what is now the developing world became poor while poor ones grew rich. We question their analysis with regard to both of their proxies for pre-modern income, namely urbanization and population density. First, an alternative measure of urbanization with more observations generates a positive (but not significant) correlation between pre-modern and contemporary income. Second, we show that their measure of population density as a proxy is highly flawed inasmuch as it does not properly measure density on arable land, and when corrected with better data the relationship is no longer robust. At best our results demonstrate a Reversal of Fortune only for the four neo-Europes of Australia, Canada, New Zealand and the United States; at worst, we show no Reversal for other former colonies.