How hot did you say it was?

In fact, the actual trends in temperatures had nothing to do with how people perceived them. If you graphed the predictive power of people’s perceptions against the actual temperatures, the resulting line was flat—it showed no trend at all. In the statistical model, the actual weather had little impact on people’s perception of recent temperatures. Education continued to have a positive impact on whether they got it right, but its magnitude was dwarfed by the influences of political affiliation and cultural beliefs.

And those cultural affiliations had about the effect you’d expect. Individualists, who often object to environmental regulations as an infringement on their freedoms, tended to think the temperatures hadn’t gone up in their area, regardless of whether they had. Strong egalitarians, in contrast, tended to believe the temperatures had gone up.

The same does not hold for perceptions of droughts and floods.  The full story is here, the underlying paper is here, and for the pointers I thank Michael Rosenwald and Scott Duke Kominers.