The Power of Words: Does language matter for attitudes toward private and public debt?

Title: The Power of Words: Does language matter for attitudes toward private and public debt?

Abstract: We aim to understand whether language can causally affect people's attitudes toward private and public debt. To do so, we provide different sets of empirical results: (i) descriptive evidence from various settings and (ii) causal evidence from survey experiments in five countries. The experimental component of our research design involves the randomization of wording in outcome questions about private and public debt. Exploiting the fact that the commonly used word for ``debt'' also means ``guilt'' in German, Dutch, and Swedish, we analyze how respondents' attitudes toward private and public debt are affected when these negatively connoted words are replaced with neutral alternatives. Our results confirm our main hypothesis: the use of guilt-connoted wording leads to lower approval of private and public debt compared to neutral wording. Finally, we provide evidence that language is strategically used in the political arena by applying natural language processing and machine learning techniques to parliamentary speeches in the German Bundestag.