Do Climate Change Consensus Messages Cause Reactance? A Comment on Chinn and Hart (2021)

Chinn and Hart (2021) conclude that communicating the scientific consensus on climate change causes psychological reactance. Here we identify several issues that cast doubt on these findings. First and foremost, the authors replicate the finding from van der Linden, Maibach, et al. (2019) that consensus messages do not increase perceptions of manipulation relative to control, neither in general nor amongst partisans. The authors also fail to replicate the three-way interaction from Ma et al. (2019) between the consensus message, partisanship, and prior attitudes. The significant reactance findings only appear for the post-only condition, and the effect size of the partisan interaction appears minimal. Moreover, as no other dependent measures were reported, it remains unclear whether respondents still updated their beliefs despite any reactance. In light of these issues and the elusive nature of backfire effects more generally, we strongly caution against the conclusion that scientific consensus messages cause psychological reactance.