How to make causal inferences using texts

Text as data techniques offer a great promise: the ability to inductively discover measures that are useful for testing social science theories with large collections of text. Nearly all text-based causal inferences depend on a latent representation of the text, but we show that estimating this latent representation from the data creates underacknowledged risks: we may introduce an identification problem or overfit. To address these risks, we introduce a split-sample workflow for making rigorous causal inferences with discovered measures as treatments or outcomes. We then apply it to estimate causal effects from an experiment on immigration attitudes and a study on bureaucratic responsiveness.

Automated summary

Text as data techniques have the potential to test social science theories by using large collections of text. However, estimating the latent representation of the text may introduce risks. To address these risks, a split-sample workflow is introduced for rigorous causal inferences.