Cultural cognition of weather: a cool (or warm) guest post!

And now for something completely different– a guest post from someone who knows what he’s talking about! (And is this just my politics speaking or was July really friggin hot?!)

Cultural Cognition of Weather
by Larry Hamilton
Carsey School of Public Policy, Univ. New Hampshire

December of 2015 was the warmest ever recorded in New Hampshire, by far. Indeed, in temperature anomaly terms (degrees above or below average) it was the warmest of any month for at least 121 years. January, February and March of 2016 were less extreme but each still ranked among the top 15, making winter 2015–2016 overall the state’s warmest on record — eclipsing previous records set successively in 1998, 2002 and 2012 (Figure 1).

Seeing in this record a research opportunity, colleagues and I added a question to a statewide telephone survey conducted in February 2016, to ask whether respondents thought that temperatures in the recent December had been warmer, cooler, or about average for the state. Two months later (April), we asked a similar question about the past winter as a whole. Physical signs of the warm winter had been unmistakable, including mostly bare ground, little shoveling or plowing needed, poor skiing, spring-like temperatures on Christmas day, and early blooming in a state where winters often are snowy and springs late. Not surprisingly, a majority of respondents correctly recalled the warm season. Their accuracy displayed mild but statistically significant political differences, however. Tea Party supporters, and people who do not think that humans are changing the climate, less often recalled recent warmth (Hamilton & Lemcke-Stampone 2016). Although percentage differences were not large, these patterns echoed greater differences seen in studies that asked about longer-term changes. Our February and April surveys had found counterparts on a much more immediate, tangible scale.

Although the February and April 2016 results fit with broader patterns, they were not overwhelming by themselves. Believing in the value of replication, we asked the question one more time on a July 2016 survey, with winter several months behind. Most people still recalled the unseasonable warmth. Our July wording and results are as follows:

Thinking back to earlier this year, would you say that THIS PAST WINTER, the weather where you live was generally colder, warmer, or about average for winter in your area? ROTATE 1–3

1          Colder than average winter for your area (4%)
2          Warmer than average winter for your area
3          About average winter for your area
98        DK/NA (4%)

Political and climate-belief gaps now appeared wider than they had been earlier in the year. Figure 2 shows one striking example: a 21-point gap between supporters of Clinton and Trump (this was, after all, primarily a political poll).

Figure 3 breaks down the percentage of “warmer” responses on the July survey by other respondent characteristics, including their beliefs about climate change. P-values summarize tests from probability-weighted logit regression.

One notable pattern in Figure 3 involves political identification; we see a 17-point gradient from Tea Party supporters to Democrats in recollections about the winter they had all just experienced. Climate-change beliefs produce wider differences: respondents who don’t believe that climate is changing, or that climate is changing but for natural reasons, were much less likely to recall the warm winter.

Figure 4 places this July poll in context with political gradients (using the same 4-party scheme) from five previous surveys that asked longer-term climate/weather questions. Panels (a) and (b) involve atmospheric CO2 levels and Arctic sea ice (Hamilton 2012, 2015). Panels (c) and (d) depict results from two Northeast Oregon surveys that asked whether summers there had become warmer in the past two decades (Hamilton et al. 2016). Panel (e) charts responses to a question about whether flooding in New Hampshire had increased over the past decade (Hamilton et al. in press). Panel (f) repeats the unpublished July survey results described earlier, on whether New Hampshire’s recent winter had been warm.

What underlies this replicable pattern? Atmospheric CO2 levels and Arctic sea ice are not directly experienced by most people. They are measured and communicated mainly by scientists, so public resistance to these well-observed realities might be conceived as a problem of science communication, highlighting the need for ideologically-tailored methods. But science communication on these topics already involves many different organizations, research teams, and individual scientists taking diverse and ofttimes innovative approaches. An alternative hypothesis is that the partisan gradients reflect not shortcomings of science communication but the efficacy of counter-science communication, convincing ideologically receptive audiences that undisputed facts are false. The sociological literature about such counter-messaging has recently been summarized by Dunlap and McCright (2015).

Science communication seems distant, moreover, from panels c–f, which involve phenomena that can be directly experienced. Warmer, dryer summers in Northeast Oregon have exacerbated insect and disease threats to forests, both directly and indirectly contributing to the frequency of large wildfires. Such changes are visible, and in isolation seem equally compatible with individual beliefs that climate is change is happening either for natural or anthropogenic reasons — which together comprise 85% of the respondents in both Oregon surveys. Nevertheless, we find steep political gradients. Similar observations apply to flooding in New Hampshire, which has caused significant damage, and is most salient not through scientific reports but through news coverage if not personal experience. Again, most news coverage made no explicit connections with climate change, and most people (89% on these surveys) agreed anyway that climate is changing, whether from human or natural causes.

Although wildfires and floods might not impact everyone, or impress them with decadal change, the snowiness or un-snowiness of a winter affects daily life for just about everyone living in New Hampshire. Panel f depicts ideology-influenced perceptions at the mundane scale of recent weather.


Dunlap, R.E. & A.M. McCright. 2015. “Challenging climate change: The denial countermovement.” Pp. 300–332 in R.E. Dunlap & R.J. Brulle (eds), Climate Change and Society: Sociological Perspectives. New York: Oxford University Press.

Hamilton, L.C. 2012. “Did the Arctic ice recover? Demographics of true and false climate facts.” Weather, Climate, and Society 4(4):236–249. doi: 10.1175/WCAS-D-12-00008.1

Hamilton, L.C. 2015. “Polar facts in the age of polarization.” Polar Geography 38(2):89–106. doi: 10.1080/1088937X.2015.1051158

Hamilton, L.C., J. Hartter, B.D. Keim, A.E. Boag, M.W. Palace, F.R. Stevens & M.J. Ducey. 2016. “Wildfire, climate, and perceptions in northeast Oregon.” Regional Environmental Change doi: 10.1007/s10113-015-0914-y

Hamilton, L.C. & M. Lemcke-Stampone. 2016. “Was December warm? Family, politics, and recollections of weather.” Durham, NH: Carsey School of Public Policy.

Hamilton, L.C., C.P. Wake, J. Hartter, T.G. Safford & A. Puchlopek. in press. “Flood realities, perceptions, and the depth of divisions on climate.” Sociology doi: 10.1177/0038038516648547

Leave a Comment