The growth of climate change misinformation in US philanthropy: evidence from natural language processing

Abstract

Two of the most consequential developments affecting US politics are (1) the growing influence of private philanthropy, and (2) the large-scale production and diffusion of misinformation. Despite their importance, the links between these two trends have not been scientifically examined. This study employs a sophisticated research design on a large collection of new data, utilizing natural language processing and approximate string matching to examine the relationship between the large-scale climate misinformation movement and US philanthropy. The study finds that over a twenty year period, networks of actors promulgating scientific misinformation about climate change were increasingly integrated into the institution of US philanthropy. The degree of integration is predicted by funding ties to prominent corporate donors. These findings reveal new knowledge about large-scale efforts to distort public understanding of science and sow polarization. The study also contributes a unique computational approach to be applied at this increasingly important, yet methodologically fraught, area of research.