Signals of Doubt: Text-Mining Climate Skepticism

Although agreement among scientists on anthropogenic climate change is clear, national surveys show that the American public’s perceptions on the science of climate change diverge significantly from the “consensus view.” Explanations for this divergence range from the influence of local weather events to the health of the economy, while many place blame on the efforts of climate change deniers to sway public opinion away from mainstream climate science. This study examines the dynamics of climate change “contrarianism” in the American context by analyzing the content of major conservative think tanks on the issue from the mid 1990s to the present. Specifically, we rely on an unsupervised text classification algorithm to produce time series measures of climate skepticism. Our results highlight how even simple uses of recent advances in natural language processing provide insight into key questions in the literature on media coverage of