LLMs will reveal many secrets of the world:
Do academic writings in economics reflect the political orientation of economists? We use machine learning to measure partisanship in academic economics articles. We predict the observed political behavior of a subset of economists using expressions from their academic articles, show good out-of-sample predictive accuracy, and then predict partisanship for all economists. We then use these predictions to examine patterns of political language in economics. We estimate journal-specific effects on predicted ideology, controlling for author and year fixed effects, which are consistent with existing survey-based measures. We show considerable distribution of economists across research fields based on predicted partisanship. We also show that partisanship is detectable even within fields, even among those that estimate the same theoretical parameter. Using policy-relevant parameters collected from previous meta-analyses, we then show that imputed partisanship is correlated with the estimated parameters, such that the implicit policy prescription is consistent with partisan orientation. For example, we find that moving from the leftmost estimate of the elasticity of top taxable income to the rightmost estimate decreases the optimal tax rate from 84% to 58%.
Underlined by TC. It’s from a new paper by Zubin Jelveh, Bruce Kogut and Suresh Naidu, recently published in Economic review.
Via the excellent Kevin Lewis.
