全球变化知识资源中心

A growing body of research shows that the beliefs and concerns many Americans have about catastrophic climate change are susceptible to variations in the weather itself. These are disconcerting findings for anyone desiring an informed public discussion about global warming. Writing in Nature Climate Change, Lisa Zaval and colleagues¹ report a series of Web-based experimental studies that help explain the responsiveness of climate change attitudes to weather patterns. Their findings show that the complexity of the climate change problem combined with the salience of our experiences of local weather produce — yes — a 'perfect storm' in which individuals' attitudes about global warming are unduly influenced by the weather outside their windows.

The evidence for the effect of weather on public opinion regarding climate change is now overwhelming. Belief in climate change and concern about its consequences is more likely among those who perceive the weather to be warmer^{2, 3, 4} and among those who, according to weather data, have actually experienced unusually hot weather prior to being interviewed^{5, 6, 7, 8, 9}.

Whatever the explanation for these findings may be, it is not because Americans lack information. Since James Hansen's ground-breaking testimony on human activity and planetary warming before the US Senate in 1988, the public has experienced a quarter-century of sustained attention to the climate change problem by experts and elites, including, by our count, no fewer than 180 congressional hearings, 744 news stories on broadcast TV networks and 940 articles in the New York Times in addition to approval by the House of Representatives of major legislation to address the problem in 2009.

So, do individuals forget this kind of information when responding to survey questions about global warming? Zaval and co-authors reject this explanation. In one of their experiments, participants were randomly assigned to read a short passage about how weather variability is a poor indicator of climate change. The association between these subjects' perceptions of local temperatures and their attitudes about global warming was, nevertheless, just as strong when compared with those in a control group. Another experiment ruled out the explanation that the weather's effect comes down to the wording of the questions. Perceptions of local temperature play the same strong role in affecting attitudes, regardless of whether the problem is called 'global warming' or 'climate change'.

Our experience of weather seems to short-circuit the pathways through which we reach good judgments. Previous research found that merely placing individuals in a warm room makes them more likely to report their belief in global warming¹⁰. Zaval and colleagues show that an even simpler prime is to parse sentences with heat-related words like 'boils', 'sunburn' and 'roasted'. This leads experimental subjects to be more likely to report that they believe in global warming and are concerned about its effects.

This process is known as attribute substitution, and is an aspect of human fallibility in decision making that can apply to our thinking about difficult public policy issues¹¹. Instead of engaging in a careful assessment of the complex attributes of the issue at hand, we substitute similar ones that come more easily to mind. We then proceed to render our judgment based on our evaluation of these substituted attributes. The vulnerability of our reasoning about global warming to attribute substitution stems from three intersecting factors: climate change is hard to understand, local weather is highly salient and it has attributes that seem to be relevant to the question of global warming's existence.

If attribute substitution is at work, the perceptions of temperature that should have the most influence on climate change judgments are those that are immediately accessible, which should also influence the recall of other temperature events. Zaval and colleagues find support for both these implications: priming attention to temperature amplifies its effect only for the current day's weather, not for the preceding day, and perceptions of warmth for the current day increase the belief that recent weather has been unusually warm.

What do these findings imply for the public's understanding of climate change? To be sure, they will disappoint those who wish that the long-standing public debate on climate change were actually closing the gap between the scientific consensus and the public's divided beliefs.

But as troubling as the 'local warming' effect may be, it is not unusual: individuals regularly make assessments about society-wide conditions (like unemployment or crime) from their own experiences (of, say, losing a job or being mugged on the street). Considering that one of the chief effects of global warming is to raise the prevalence of unusually hot days, drawing conclusions about the existence of global warming from local weather is not entirely irrational. Zaval and colleagues show us that such reasoning stems from placing undue weight on this highly salient information compared with other information — like scientific research — that is much more relevant. The challenge facing scientists and climate change communicators is to raise the awareness of this latter information so that Americans might be encouraged to focus on large-scale patterns rather than the weather outside.

1. Zaval, L., Keenan, E. A., Johnson, E. J. & Weber, E. U. Nature Clim. Change 4, 143–147 (2014).
   • Article
2. Krosnick, J., Holbrook, A., Lowe, L. & Visser, P. Clim. Change 77, 7–43 (2006).
   • Article
3. Li, Y., Johnson, E. J. & Zaval, L. Psychol. Sci. 22, 454–459 (2011).
   • PubMed
   • Article
4. Semenza, J. C. et al. Am. J. Prev. Med. 35, 479–487 (2008).
   • PubMed
   • Article
5. Deryugina, T. Clim. Change 118, 1–20 (2013).
   • Article
6. Egan, P. J. & Mullin, M. J. Pol. 74, 769–809 (2012).
   • Article
7. Hamilton, L. C. & Stampone, M. D. Weather Clim. Soc. 5, 112–119 (2013).
   • Article
8. Howe, P. D., Markowitz, E. M., Lee, T. M., Ko, C.-Y. & Leiserowitz, A. Nature Clim. Change 3, 352–356 (2013).
   • Article
9. Joireman, J. A., Barnes Truelove, H. & Duell, B. J. Environ. Psychol. 30, 358–367 (2010).
   • Article
10. Risen, J. L. & Critcher, C. R. J. Pers. Soc. Psychol. 100, 777–793 (2011).
    • PubMed
    • Article
11. Kahneman, D. & Frederick, S. in Heuristics and Biases (eds Gilovich, D. T., Griffin, D. & Kahneman, D.) 49–81 (Cambridge Univ. Press, 2002).

Download references