How Reality-Based Is the Community?

by PM

7 February 2013, 1107 EST

Community-cast John Quiggin at Crooked Timber discusses the American right’s quick shift to admitting a decline in U.S. income mobility. He then asserts that this is part of a process by which “objective truth, rather than political acceptability, should be the criterion against which factual claims are tested.” (There’s also a long discussion of The Overton Window, although I suppose he meant this one.

Quiggin goes further:

If this view is right, then the most important single development was probably Nate Silver’s successful prediction of the 2012 election. Silver was up against both the pseudo-science of the Republican “unskewers” and the faith of centrist pundits (historically exemplified by Broder) that their deep connection with the American psyche was worth more than any number of least-squares regressions. Given the centrality of horse-race journalism to the pundit class, their defeat by relatively straightforward statistical analysis of opinion polls was a huge blow.

My response to this is somewhat more tempered than Quiggin’s–although probably warmer than the average CT reader’s. First, I’m skeptical of the notion that Science and Progressive Politics will go through life merrily holding hands. There’s no particular reason to think that liberal values are anything but orthogonal to the findings of most research, lab-scientific or observational-scientific. There are some nicely convenient findings for liberal values–the democratic peace hypothesis, for one–but would anyone be willing to give up democracy if we found incontrovertible evidence that democracies (not just Mansfield-and-Snyder-style democratizing countries) are more bellicose? Or, alternatively, would we give up democracy if the field coalesced around a consensus that democracies are less bellicose because they are more successful at using social pressure or other nonviolent forms of coercion to eradicate dissenting views? Social scientific findings rarely provide evidence that prompts us to revise our value systems.

The problem that liberals will increasingly face is close to the dilemma that the brilliant, late BBC sketch programme That Mitchell And Webb Look satirized by asking whether, if “the computer” said that killing the poor would raise economic growth, you would actually do it.[1] It is true that at the moment Republicans are pretty far from occupying the intellectual high ground, but it strikes me as a little suspect that a great many liberals (and here I’m thinking less of the named folks who write blogs and more of the mass who comment on blogs) are now convinced that Science Is On Their Side.

Nate Silver is a visible part of the debate, possibly because he happened to tell readers what they wanted to hear and partly because (largely coincidentally, from the readers’ point of view) he was “right.” Silver predicted a Democratic win, and conservatives obligingly vilified him. But if, counterfactually, the state of the world had changed just enough that Silver’s models and the underlying data-generating process were pointing toward a Romney victory, would liberals have adopted him as a prophet or would there have been an unskewed polls movement created by the left? [2]

Quiggin is too comfortable in his analysis. The notion that “objective truth” sits much more easily with leftist politics now than it does with conservatives is, ahem, objectively true. But that is not because of an essential affinity with leftism. Forty years ago, the Soviet Union was comparatively inhospitable to objective truth. And basing one’s value arguments so firmly on an evidence-based foundation is clearly self-defeating. Liberals’ response to the question “Should the Boy Scouts admit gays?” is not to commission a randomized field experiment to compare outcomes (merit badges per scout?) across gay-friendly and traditional clubs. Nor, when conservatives raise arguments gay marriage (which I must say are almost always prima facie stupid), liberals rarely react with the measured caution of a peer-reviewed article. Instead, they react with scorn (which, I have to say, I pretty much share, even if I think the T-shirt is a bit over the top).

It’s pretty easy to conceive of a political issue in which the right is friendlier to the objective evidence than the left: genetically modified organisms spring readily to mind. [3] Of course, given that the right is openly hostile to science, that set is unlikely to get larger any time soon. But we shouldn’t delude ourselves into thinking that politics is any less tribal now because our tribe happens to be allied with the small faction of nerds in lab coats.


[1]Explaining why we can’t kill all the poor, one policy analyst sputters, “Well, ’cause they do all the — they clean — they do all the things that we don’t fancy!” Second analyst: “Aren’t you thinking of immigrants?” Also notable: after the model suggests that killing the poor wouldn’t help, the minister responds “Have you tried raise VAT and kill the poor?” I can only dream of a day in which U.S. cabinet ministers are so familiar with interaction terms.

[2] Almost certainly to have been named “Occupy Statistics.”

[3] Indeed, if I were a conspiracy theorist, I’d suggest that GMOs are a distraction for liberals to keep them from taking up antitrust actions against Monsanto, ConAgra, and AGM.