Cantoni (2015) — The Economic Effects of the Protestant Reformation

Citation. Cantoni, Davide. “The Economic Effects of the Protestant Reformation: Testing the Weber Hypothesis in the German Lands.” Journal of the European Economic Association 13(4), 2015: 561–598.

Summary

The most consequential empirical refutation of the strong-form Weber thesis in the modern literature. Using population growth of 272 German cities across 1300–1900 — and exploiting the Holy Roman Empire’s religious heterogeneity, which makes the German lands a near-ideal natural setting for testing the Weber hypothesis — Cantoni finds no economic-growth effect of Protestantism. The null result is precisely estimated, robust to a range of alternative specifications and controls, and does not depend on data selection or sample-size choices.

The empirical strategy is methodologically tight. Cantoni assembles a long-run city population panel (city population is widely treated as a usable proxy for local economic prosperity in pre-modern data, given the costs of supporting urban populations from rural surpluses); he tracks the religious affiliation of each city’s territorial sovereign over time, allowing within-city variation as boundaries shift; and he tests whether becoming or being Protestant changes a city’s subsequent population trajectory relative to comparable Catholic cities. The answer is a clean no.

The paper is the field’s strongest single empirical word on the Weberian work-ethic mechanism in the German case (the Reformation’s heartland). Combined with Becker & Woessmann (2009) — which finds a Protestant correlation in 19th-century Prussian income but explains it via literacy and not work-ethic — the contemporary empirical record on Weber is: the correlation is locally present in some specific 19th-century settings; the work-ethic mechanism is unsupported; the correlation that does exist is mediated by human capital.

Key claims

• Across 272 German cities 1300–1900, there is no statistically significant effect of Protestantism on economic growth as measured by city population.
• The null result is precisely estimated — not “Protestant effect could be there but data are too noisy”; the effect is well-bounded to be small or zero.
• The result is robust to alternative specifications, alternative city samples, alternative time windows, and the inclusion of standard economic-and-political controls.
• Combined with the Becker-Woessmann human-capital reinterpretation of the Prussian data, the modern empirical record provides no support for the Weberian work-ethic mechanism as a cause of economic growth in the German Reformation lands.
• The strong-form Weber thesis (Protestant work-ethic causes economic growth) cannot be sustained on the available evidence.