First, some common sense:
Let's assume self-interest explains the formation of beliefs about climate change by ordinary members of the public (I'm very happy to do that). In that case, we should expect the economic impact of climate change & proposed climate change policies on the public's perception of climate change risks to be 0.00, and the impact of cultural identity to be [some arbitrarily large number].
What the ordinary member of the public believes about climate change won't have any impact on the threat it poses to the environment or on the policies society adopts to repel that threat. The same is true about how he or she votes in democratic elections or behaves as a consumer. As an individual, he or she just isn't consequential enough to matter.
Accordingly, there is no reason to expect much if any correlation between, say, economic class, etc., and climate change risk perception.
In contrast, what an ordinary individual believes and says about climate change can have a huge impact on her interactions with her peers. If a professor on the faculty of a liberal university in Cambridge Massachusetts starts saying "cliamte change is ridiculous," he or she can count on being ostracized and vilified by others in the academic community. If the barber in some town in South Carolina's 4th congressional district insists to his friends & neighbors that they really should believe the NAS on climate change, he will probably find himself twiddling his thumbs rather than cutting hair.
It's in people's self-interest to form beliefs that connect rather than estrange them from those whose good opinion they depend on (economically, emotionally, and otherwise). As a result, we should expect individuals' cultural outlooks to have a very substantial impact on their climate change risk perceptions.
(For elaboration of this argument, see CCP working paper No. 89, Tragedy of the Risk Perceptions Commons.)
Second, some data:
I have constructed some regression models to examine the impact of household income (hh_income) and cultural worldviews (hfac for hierarchy and ifac for individualism) on climate change risk perceptions (z_GWRISK; for explanation of that measure, see here). The data come from a nationally representative survey of 1500 US adults conducted by the Cultural Cognition Project with a grant from the National Science Foundation. To see the regression outputs, click on the thumbnail to the right.
The analyses show, first, that differences in income have a very small negative impact on climate change risk perceptions (B = -0.07, p < 0.01) when consdired on its own (model 1).
Second, the analyses show that cultural worldviews have a very large impact -- a typical egalitarian communitarian and a typical hierarchical individualist are separated by about 1.6 standard deviations on the risk perception measure -- controlling for income (model 2). When cultural worldviews are controlled for, income turns out to have an effect that is practically nil (B = -0.02, p = 0.56).
But wait: the third thing the analyses show is that income does have a modest effect -- one that is conditional on survey respondents' cultural worldviews. As they become wealthier, egalitarian communitarians become slightly more concerned about climate change, while hierarchical individualists become less (Model 3).
Bottom line: economic self-interest doesn't matter; cultural identity self-interest does.