Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We provide new evidence on motivations for voluntary contributions to public goods in a natural setting. Using natural language processing on users’ Twitter posts, we measure revealed sentiment changes before and after making a donation to Wikipedia. We find strong evidence that sentiment improves in the hour and minutes directly before contributing, which we call ‘preheating’. Results are robust to alternative fixed effects and approaches to inference and supported by a complementary online experiment with randomised mood inducement among Twitter users. Preheating suggests that affective states influence giving in addition to reward-seeking, utility-maximising behaviour that has been documented in other contexts.