Ruben Durante (PhD candidate from Brown Univ. in political economy, May 2010) has an interesting job-market paper exploring the origins of social trust by examining variability in precipitation and temperature 1000-1500 years ago.
He posits that norms of trust developed as a result of collective action and mutual insurance triggered by farmers coping with dramatic climatic change. While most of these areas have now become industrialized, these medieval norms lived on.
For a copy of his paper (c. Nov. 2009) see here. He compares contemporary social trust (using European Social Survey data) with reconstructed paleoclimatic temperatures from 1500-1900. (For the statistical junkies out there, one standard deviation in precipitation variability corresponds with a .17 standard deviation increase in social trust and this is robust to controlling for average temperature, terrain ruggedness, soil quality, standard deviation in soil quality, etc.). The effect, while statistically significant, is not that large, but it is still amazing that weather 1000-1500 years ago convincingly predicts social trust today.
While Durante hasn’t run his data on Italy specifically, his results might help explain the puzzling finding of my colleague Robert Putnam who found that high-trust Italian regions in the north were also the same regions that were high trusting in 1500. My hunch is that there is much greater temperature and precipitation in the high-trusting northern regions of Italy than in the southern regions of Italy so Durante’s results would likely hold up in Italy specifically. Durante hasn’t run this same analysis within Italy, or worldwide for that matter.
I’m not sure that I see how his study explains another anomaly in social trust. There is a remarkable similarity if you rank the average trust levels of third generation immigrants to the U.S. sorted by their grandparents’ home country with contemporary levels of social trust in those countries. In other words, if one looks at third generation immigrants to America, you find, for example, that the third generation Swedes or Norweigians are at the top of the list in social trust and the third generation Brazilians, for example, are among the least trusting American third generation immigrants. These rankings look remarkably like the rankings of those countries today by social trust. In other words, for immigrants whose ancestors came over 70-100 years ago and who are doing little to affect social trust back in their origin countries, both their offsprings’ level of social trust and the level of trust of natives who remained in those origin countries 70-100 years later are remarkably similar. That’s quite a mystery, and doesn’t seem explained by Durante’s paper on temperature and precipitation variability since the immigrants have moved to a new country whose social norms were presumably influenced (according to Durante) by what precipitation and temperature variability were in the “new country” (America), not their old country.
Anyway, food for thought…