More on ideological symmetry of motivated reasoning (but is that really what’s important?)

I have posted a couple times (here & here) on the “symmetry” question — whether dynamics of motivated reasoning generate biased information processing uniformly (more or less) across cultural or ideological styles or are instead confined to one (conservativism or hierarchy-individualism), as proponents of the “asymmetry thesis” argue.

Chris Mooney has applied himself to the symmetry question with incredible intensity and has an important book coming out that marshals all the evidence he can find (on both sides) and concludes  that the asymmetry thesis is right. But Mooney now has concluded that he sees the latest CCP study on “geoengineering and the science communication environment” as evidence against his position (not a reason to abandon it, of course; that’s now how science works– one simply adds what one determines to be valid study findings to the appropriate side of the scale, which continues to weigh the competing considerations in perpetuity).

Mooney’s assessment — and his public announcement of it — speak well of his own open-mindedness and ability to check the influence of his own ideological commitments on his assessments of evidence. But still, I think he has far less reason than he makes out to be disappointed by our results.

In our study, we tested the hypothesis that exposing subjects (from US & UK) to information on geoengineering would reduce cultural polarization over the validity of a climate change study (one that was in fact based on real studies published in Nature and PNAS).

We predicted that polarization would be reduced among such subjects relative to ones exposed to a frame that emphasized stricter carbon-emission controls. Restricting emissions accentuates the conflicting cultural resonances of climate change, which gratify the egalitarian communitarian hostility to commerce & industry and threaten hierarchical individualist commitment to the same. Geoengineering, in contrast, offers a solution that affirms the latter’s pro-technology sensibilities and thus mitigates defensive pressure on them to resist considering evidence that climate change is happening & is a serious risk.

The experiment corroborated the hypothesis: in the geoengineering group, cultural polarization was significantly less than in the emission-control group.

The reason that Mooney sees this result as evidence against the “asymmetry” thesis is that assignment to the geoengineering condition in the experiment affected the views of both egalitarian communitarians and hierarchical individualists. The latter viewed the study as more valid than their counterparts and the latter less than their counterparts in the emission-control condition. In other words, there was less polarization because both groups moved toward the mean — not because hierarchical individualists alone moderated their views.

Okay. I guess that’s right. But for reasons stated in one of my earlier posts, I don’t think that the study really adds much weight to either side of the scale being used to evaluate the symmetry question.

As I explained, to test the asymmetry thesis, studies need to be carefully designed to reflect the various competing theories that give us reason to expect either symmetry or asymmetry in motivated reasoning.  Those sorts of studies (if the studies are designed properly) will yield evidence that is unambiguously consistent with one inference (symmetry) or the other (asymmetry).

Our study wasn’t designed to do that; it was designed to test a theory that predicted that appropriately crafting the cultural meaning conveyed by sound science could mitigate cultural polarization over it. The study generated evidence in support of that theory. But because the design didn’t reflect competing predictions about how the effect of the experimental treatment would be distributed across the range of our culture measures, the way that the effect happened to be distributed (more or less uniformly) doesn’t rule out the possibility that there really is an important asymmetry in motivated reasoning.

I think the same is true, moreover, for the vast majority of studies on ideology and motivated reasoning (maybe all; but Mooney, who has done an exhaustive survey, no doubt knows better than I if this is so): their designs aren’t really geared to generating results that would unambiguously support only one inference in the asymmetry debate.

In the case of our (CCP) studies, at least, there’s a reason for this: we don’t really see “who is more biased” to be the point of studying these processes.

Rather, the point is to understand why democratic deliberations over policy-relevant science sometimes (not always!) generate cultural division and what can be done to mitigate this state of affairs, which is clearly inimical, in itself, to the interest of a democratic society in making the best use it can of the best available evidence on how to promote its citizens’ wellbeing.

That was the point of the geoengineering study. What it showed — much more clearly than anything that bears on the ideological symmetry of motivated reasoning — is that there are ways to improve the quality of the science communication environment so that citizens of diverse values are less likely to end up impelled in opposing directions when they consider common evidence.

For reasons I have stated, I am in fact skeptical about the asymmetry thesis. Of course, I’m open to whatever the evidence might show, and am eager in particular to consider carefully the case Mooney makes in his forthcoming book.

But at the end of the day, I myself am much more interested in the question of how to improve the quality of science communication in democracy.  When there is evidence that appears to speak to that question, then I think it is more important to figure out exactly what answer it is giving, and how much weight we should afford it, than to try to figure out what it might have to say about “who is more biased.”