My potentially unpopular opinion? I think this trend has been disastrous for the overall quality of economic history.
Víctor M. Gómez Blanco
Víctor M. Gómez Blanco21.7. klo 16.18
4/ 🧪 Methods: From Time Series to Causal Inference In 2000, 70% of papers used mostly descriptive time-series analysis. Now? It's <40%. Econometrics (IVs, DiD, panel models) are dominant. Machine learning & text mining are on the rise. Qualitative work? Nearly extinct.
@sethposting @twst12612648 @_vonarchimboldi @CommenterSerial And not only that, but the myopic emphasis on methods means there’s now a kind of hollowing-out of historical expertise, leading to all sorts of absolute howlers being published because the peer-reviewers have no idea of the historical detail and just comment on method design
8,62K