This comes from Yuriy Gorodnichenko and Gérard Roland, who find a correlation between how individualistic a country’s residents are and that country’s long-term economic growth. After some additional testing, they also claim there’s a causal relationship:
We find a strong causal effect of individualism on income per worker, total factor productivity, and innovation as predicted by our theory. These results hold even when we exclude the Americas and Oceania where settler colonization played an important role. They also hold when controlling for measures of geographic distance, human capital, ethnic fractionalization, and other factors affecting growth….Moreover, even after controlling for measures of institutions which were previously found (e.g., Hall and Jones, 1999, Acemoglu et al., 2001) to affect long-run growth, culture continues to play a statistically significant and quantitatively important role, implying that culture has an effect on economic development that is independent of institutions.
Summary here. I’d be interested in seeing a statistical critique of this result. (Via Free Exchange.)