Denialism, Informational Conformity and New Coke

Do you remember when the Coca-Cola recipe was updated, and how that almost destroyed the Coca-Cola company's flagship product? There's a brilliant retrospective of that story; go read it, I promise it's relevant.

The article compellingly argues that the bizarre sequence of effects was due to a phenomenon that can be called informational conformity.

Informational conformity was first formally documented by Dr Muzafer Sherif in 1935, when he placed a group of subjects in a dark room with a single point of light in the distance. He asked them to estimate how much the light moved around, and although each person perceived a different amount of movement, most of them relinquished their own estimates to conform to the predominant guesses within the group. In reality, the light had not been moving at all; it only appeared to move because of the autokinetic effect, a quirk in visual perception where a bright point of light in complete darkness will appear to wander. It is thought that this imagined movement occurs due to the lack of a fixed visual reference point, and it may be the cause of many nighttime UFO sightings.

In short, groupthink. In a way it's the "nobody ever got fired for buying from IBM" phenomenon: holding an opinion in isolation is much harder than holding an opinion in common with a social group with which you identify. The PR industry learned its lesson from the Coke debacle; whether mainstream journalism understands it or not is another matter.

I think David Brin's focus on Mr. Murdoch is apropos. I think Mr. Murdoch understands informational conformity, whether the rest of the press understands or not. Whether Glenn Beck or Sarah Palin are harmless buffoons or real dangers to civilization depend on whether they reach a critical mass of people who think they are something other than harmless buffoons. People like Mr. Murdoch, who hold a large fraction of people's attentions, are the people who make that decision.

This sheds some light on the Sarewitz quandary:

A dangerous idea has taken hold in modern politics, and the sooner it is discredited, the better. The idea is that political disagreements can be resolved by science. Its basic logic seems sensible: As good children of the Enlightenment, we should turn to science to establish the facts about problems such as climate change before deciding what policies to implement. Yet the types of things that scientists are good at figuring out don't have much to do with the types of things that politicians need to decide.

I objected (see link above) but John Fleck backed him up:

When I entered the profession of journalism nearly three decades ago, it was with the idea that it gave me a chance to help civic processes by helping the body politic better understand hard or complex issues, so political/policy decisions could be based on the best available information.

At every city council meeting, the training ground of many young reporters, technical experts deliver to decision makers their best available data on issues such as traffic engineering. Week after week, I saw political actors seek out their own alternatives to what I reasonably viewed as the best available data when that data conflicted with their values. In the years since, I have seen this happen across scales, from issues as local as whether to install stop signs or speed humps, to regional and state issues like the water supply in New Mexico, to national issues like the appropriate disposal path for various types of nuclear waste, to the current global discussion we’re all so engaged in regarding greenhouse gas emissions and climate change.

I have seen liberals side with what I regarded as the best available data on some issues, conservatives on others. In some cases, environmentalists have had what seemed to me the best available data on their side, while in other cases industry has. At the local scale, I saw many issues that didn’t break down on any sort of liberal-conservative spectrum, but instead fell along geographic lines (rural/urban, this neighborhood v. that one, etc.).

My experience with the pattern is sufficiently consistent that I believe Sarewitz has correctly described not a specific problem found in specific situations, but a general principle.

Michael might wish it were not so, but my decades of experience in the midst of political fights large and small suggests otherwise.

I suggested it was cultural and contingent. Perhaps Canadians are more accessible to reason than Americans, and perhaps Swedes and Dutchmen more than Canadians. And I still think that is true. Certainly, scientists and engineers and doctors are more accessible to reason than, hmm, bankers and accountants and real estate brokers. (No offense if you happen to be in the latter category; there are exceptions both ways.) Cultures can change. So the job, it seems, is to change the culture.

In the light of informational conformity, though, a new theory arises. While scientists and engineers and doctors embrace skepticism as part of their identity, others embrace this or that belief. ("oil companies are bad", "Obama is Kenyan", "cell phones give you cancer", "vaccines cause autism", "climategate reveals awful things about climate science"...) These are not generally evidence-based decisions, but cultural cohesion decisions, like identifying with the Packers or the Bears. There's no rational reason for it; one team winning makes your friends happy, the other makes your friends unhappy, that's all there is to it.

The New Coke piece also explains how this is achieved: enthusiastic allegiance by someone perceived as part of your peer group is worth far more than mere agreement.

When participants were asked whether they would drink Coca-Cola if it were modified to use this new formula, most responded positively. However about 11% of the samplers– even some who preferred the new flavor in the blind tests– were hostile to the idea. They were astonished that the soft-drink juggernaut would have the audacity to tinker with the American-as-bald-eagle-pie beverage. This indignation was so potent that it exerted indirect peer pressure within the focus groups, thereby contaminating the results; but Coke experimenters were quick to detect and correct the effect. ... At first, Goizueta’s surly synopsis proved accurate. The company’s stock went up upon the announcement, and sales improved by 8% in the first few weeks. Surveys indicated that an impressive 75% of consumers were happy with New Coke, and would buy it again. ... Within a few weeks, however, unpleasant sentiments began to ooze from the unpredictable public. There was a segment of the population– about 11%, strangely enough– who disliked New Coke with such enthusiasm that their complaints and harsh editorials began to disintegrate public approval. New Coke became a vehicle for large-scale informational conformity, the human tendency to unconsciously adjust one’s opinions to correlate with the outspoken views of the social group.

On this model, the loss of confidence among conservatives for what they can easily interpret as extremist green evangelism as opposed to science, is going to be very persistent. Perhaps only a small proportion strongly holds this belief, but they are regarded as peers in a larger section of the society, and their belief is expressed adamantly and enthusiastically. It fits in neatly with and reinforces many of their beliefs, and it offers entertainment and delight in mocking the dour and gloomy predictions which they find far-fetched.

The conservatives aligned in this way in part because of the American press's cowardice and in part because of the British press's blind belligerence, but in any case the press firmly included the "hoax" model of climate change as a socially acceptable theory. Some people picked it up with enthusiasm, and they in turn won over their demographic. We already dropped the ball; it will be very hard to win it back.

Unless and until conservatism goes away, reversing this is both necessary and, on the New Coke model, nearly impossible.

I think that our best hope is to address conservatives who are also scientifically inclined skeptics. These people will still be amenable to reason. While we may win them over to the ranks of the reasonable, though, it will be difficult to win them over to the ranks of the enthusiastic, especially in the USA. The conservative culture in America, albeit self-proclaimedly freedom-loving, is immensely conformist and rejects anyone who steps out of line on matters of allegiance. The social pressures to keep a lid on global environmental concerns, now that those are considered disreputably leftist, will be immense.

Nevertheless, the facts are on our side. In this regard if perhaps in few others, Steve McIntyre is right. He advises us to create an "engineering-level" description of "the whole argument"; this is in fact impossible because engineering texts take an omniscient voice and simply don't stress facts they use without a full elaboration. (A and B showed (ref) that (equation 4.17) etc. etc. is not really going to fly.)

This is not an easy task, but it's something we ought to have done anyway. How to acquire the resources for such a massive task is an issue. But it's what we need to do; essentially to build an entire reference network for the many bits of science that go into our understanding of the climate system. In the end, probably nobody will read all of it, but it should provide a compelling resource for anyone digging into any corner of it.

Each of us should go digging in our souls for our inner conservative. Those of us who come up empty-handed should probably just work on something else besides outreach. We should be especially kind to people who are conservative but sane, who understand and appreciate the science, people like Tokyo Tom or even Jim Manzi, even if we disagree with their understanding of economics and politics. And we absolutely need to cherish people like Katharine Hayhoe who speak up for climate science within conservative communities.

What is crucial is to get people to understand that this is real, that it is interesting, that it raises difficult questions. Even convincing them that it is a big deal is secondary. We have to make it permissible to be conservative and to respect the climate sciences.

Our main advantage is that the science is absolutely fascinating. Thank goodness for that.

What we can't do is just shallow, preachy outreach. That has reached everybody it's going to reach, and offends the people we most need to reach. The choir is already singing.

We have to really spread the science out on the table, give people knobs and levers, and let them get the picture in detail for themselves.

Of course, that does not mean capitulating to harassment. Scientists have enough burdens already without being subjected to legal wrangling for their lab notes. Demands for openness at the level of research can reach impractical extremes, but openness about established science is both possible and needed. And we have a long way to go.

The only alternative I see to this sort of detailed outreach, which of course is both difficult to do and not guaranteed to succeed even if done well, is to just wait until the climate system starts causing major damage. Maybe the conservative culture will turn around at that point, not to admit that they were wrong in the past, but to admit that finally, the evidence is in and it's time to act.

Unfortunately, at that point, it will probably be too late. All bets will be off. Predicting the climate becomes difficult at that point, but predicting the way history will twist around is simply impossible. I think we're better off not finding out. Call me conservative.