02 December, 2009

Contra a "protestant ethic"

A long time ago, in a country far, far away, the incomparable Max Weber developed what would become a very influential theory regarding the impact of religion on economic performance. The idea, for those not familiar with it, came to be known as the "protestant ethic" and centered on the contention that the values of protestantism - puritanicalism, a focus on saving, a fetish for work and a phobia of leisure time, an emphasis on individual autonomy and a disrespect for hierarchies - provided a more receptive environment for the success of capitalism than other religions. He argued that this ethic explained the predominance of protestants in the wealthy elites, as well as the shift of global economic dominance away from Catholic France and Spain and towards protestant Britain and Germany.

While this has always seemed more or less plausible, I am pretty skeptical of cultural explanations for sharp divergences in human behavior across societies. All this is to say that Tyler Cowen has linked to a paper that finds that, contra Weber, protestantism has had no effect on economic development. Discuss!

1 comment:

spencer said...

Interestingly, another paper by Becker and Woessman was published earlier this year and is very similar to Cantoni's paper--except that it finds that Protestants ARE richer than Catholics. But it doesn't attribute this to the "ethic". Rather, it argues that Protestants' focus on individually reading the Bible lead to higher literacy and levels of education. This additional human capital made Protestants richer than Catholics, not purely cultural differences.

Cantoni's paper is very similar to Becker and Woessman since it essentially uses the same empirical strategy .

I haven't read Cantoni that carefully, but he appears to use the same empirical strategy (using distance to Wittenberg as an instrument) as Becker and Woessman with different data. Whereas Becker and Woessman have data on counties from 1871, Cantoni has a panel of cities from 1300-1900. And while Becker and Woessman use share of workers employed in manufacturing as the outcome variable, Cantoni uses city sizes.

On the whole, I think I buy the Becker and Woessman paper more for a couple reasons. First, share of manufacturing strikes me as a better proxy for income than city size. Second, when Cantoni uses a legitimate empirical strategy (instrumental variables) his coefficients do point in the right direction, though they aren't significant. This could be due as much to low sample size as the null hypothesis of no effect actually being true. So I'd say Cantoni's paper is inconclusive, while Becker and Woessman's concludes positively that Protestantism does have an effect through literacy.