Predicting public attitudes on climate change: the ‘Holy Grail’
This is a guest article by Andy A. West. He can be found on Twitter at @andywest_tweets.
There is a long-standing and extensive literature that searches for ‘social predictors’ of public attitudes on climate change: the social or demographic factors – such as political attitude, age, education or relative prosperity, among many other candidates – that can explain how people view the issue. As predictive power is sought at national as well as individual level, further candidates might include, for example, the relative civil liberties of a nation or its exposure to climate impacts.
The Holy Grail of this field is a single variable that is potent enough to explain, in broad terms, public attitudes on climate change, for instance whether it is man-made, whether they’ll be harmed by it, or what priority they might give to fighting it (relative to other issues or threats); in other words, a variable that can predict a large part of the level of such attitudes in a population; if not at the dizzy heights of 50%, then maybe 30%, say, where this is at least the largest single factor among (perhaps many) others. The general idea behind this search, sometimes made explicit, is that knowledge of the drivers of attitudes would enable publics to be nudged in the ‘right’ direction, which of course is the direction of supporting Net Zero policies.
However, outside the USA[1] the Holy Grail has slipped further and further away. A dozen years or more back, the search still featured attempts to find a single powerful predictor. But as the years passed without success, efforts moved to multivariate analysis; it was hoped that groups of say three or four variables might yield greater predictive power when considered together. Unfortunately, these efforts did not deliver dramatic improvements, and introduced disagreements about which groupings were best. So, more and more variables were then pulled into the search, in a seemingly desperate attempt to find the missing magic that would make prediction possible. This required more complex models and increasingly sophisticated statistics.
As a result, the field now features webs of related variables, some of which themselves are high-level concepts such as ‘New ecological paradigm’, or ‘Class’ (engaged, pessimistic, indifferent, doubtful). A higher-level concept at the national level is for example ‘attitudes to environmental care’. Ridiculous variables have been dragged in too, such as the number of climate scientists residing in each nation!
Figure 1 is representative of the current predictor literature. It depicts fiendish complexity!
Figure 1: Public attitudes on climate change: a typical model from the literature (Source: Ruiz et al.)
The top half of Figure 2 shows a range of typical predictor values from this literature; some are older, some newer, some from simpler models and some more complex ones, some concern individual attitudes while others are national in scope. However, ultimately, they all result from a chronic lack of underlying theory about the attitudes being examined. Only one paper exceeds the 30% level.[2]
Figure 2: Power of predictors in the literature (top) and using cultural causation (bottom)
This situation has now changed dramatically. As my book ‘The Grip of Culture’ sets out, there is an underlying predictor that can explain public attitudes on climate change. In other words, at the national level, we have found the Holy Grail: a single variable that predicts a high proportion of the public attitudes on climate change (33–87% depending on the attitude), and across a wide range of attitudes; it is national religiosity. Figure 3 represents the theory in the same manner as the example diagram from the literature, and the bottom half of Figure 2 shows the dramatically increased predictive power from this very straightforward model. Attitude data comes from many independent sources – mainstream pollsters such as YouGov, Ipsos, Pew and more, plus the UN and EU.
Figure 3: Public attitudes on climate change: the cultural causation model
The underlying theory is that overall attitudes are shaped by cultures and their interactions. There is a cultural entity, a secular religion of climate catastrophism, which dominates public attitudes (and policy too) across nations. It also interacts with the older culture of religion (any faith), which is why the predictor works.
While my book includes a list of technical reasons for why the Holy Grail predictor was missed, it can be summarised as being the result of entrenched bias in social science, which labours under a misapprehension that global climate catastrophe is an output of ‘hard science’, rather than recognising its true nature – an emotive cultural narrative that contradicts mainstream science. Of course, if you aren’t looking for a culture, you probably won’t find one!
Cultures can have very unintuitive effects; for instance many countries appear to be simultaneously very concerned about climate change and very keen that nothing should be done about it. Wanna know why? Read the book!
Notes
[1] One of the few things upon which I agree with the literature, is that the situation in the US is different, due to the very high degree of political polarisation there on climate change and many other issues. The theory of cultural causation still holds, but mapping this to the US is more complex. Chapter 12 in the book covers this (political stance is a great predictor, but more subtly 4 cultures are operating, of which one is still climate catastrophism).
[2] Lo and Chow’s result was due to a near miss of the correct theory, but unfortunately this wasn’t followed up. Another study (T1) exceeds 30% for some nations but not others.