Thursday, July 31, 2014

Regarding the Definition of the Word 'Unscientific'

So, yesterday the left-wing part of the internet had a bit of fun reading this essay (of sorts) by one Charles C.W. Cooke for the National Review Online. Apparently NRO hasn't heard of single-page formatting, so that link goes only to page 1. The thing I'm gonna talk about in this post is on page 2. However, you should read the whole thing, because it is staggeringly, mind-blowingly, hilariously awful. It's basically a rant against the supposed "nerd culture" of American liberalism, with a bizarre fixation on Neil deGrasse Tyson as the emblem thereof. I could probably write a post that had a one paragraph per sentence ratio of response to this essay, but that is almost certainly a poor allocation of resources so I'm mostly gonna let it speak for, and against, itself.

But I did want to make a bit of a point regarding one particular paragraph late in the article. After criticizing liberals for claiming that their worldviews are driven solely by data, he says the following:
This is nonsense. Progressives not only believe all sorts of unscientific things — that Medicaid, the VA, and Head Start work; that school choice does not; that abortion carries with it few important medical questions; that GM crops make the world worse; that one can attribute every hurricane, wildfire, and heat wave to “climate change”; that it’s feasible that renewable energy will take over from fossil fuels anytime soon...
The sentence continues, but this is the part I'm interested in.


He lists what I'll call six beliefs which he attributes to "progressives" generally and which he describes as unscientific. Right off the bat it's worth noting that several of them are ones that basically no one on the political left holds. Abortion carries with it few important medical questions? Who the hell thinks that? I mean, early-term abortion is at this point a pretty routine procedure, medically speaking; late-term abortion is a far different story. The liberal position is just that, like a lot of other somewhat risky medical procedures, it should be basically up to the individual, in consultation with her doctor, to decide whether or not she wants it. And as for the climate change thing, puhleeze. There may be members of the Democratic rank-and-file who believe this, but this is basically a hanging curveball to any member of the progressive nerd demographic he's describing: "no, you can't trace each individual storm to global warming, but the warming causes a general increase in the frequency and severity of extreme weather events." It's not hard. Literally, any of the people he's talking about would just hit that one out of the park at a moment's notice. Does he literally not know what people who believe in climate change say when confronted with this point?

As for the others, okay, the GM foods one is one where the scientific community (including very much Neil deGrasse Tyson himself, who was literally just featured in a video defending GM foods) thinks that liberals tend to be irrationally fearful of a certain new technology. Note, however, that that doesn't necessarily make those fears unscientific--scientists are wrong sometimes! Being somewhat skeptical of the claims of scientists that XYZ new technology is perfectly safe can be unscientific, but it doesn't have to be, if there's some evidence to support the skepticism. The one about fossil fuels I think it's not quite proper to describe as unscientific (never mind the fact that it's probably, y'know, true). Predicting the future isn't really a scientific enterprise. I mean, it can be, as for instance with something like baseball, where there are reasonably scientific methods for trying to project player performance, but that's because we have a large sample of repeated trials. Whether or not renewable energy becomes feasible on a massive scale sometime soon, well, we don't have any past observations to really guide our predictions there. We know what the world is like now, we know what some trends are (solar panels, for instance, are pretty clearly getting much cheaper), but we honestly just don't know, not in a scientific way, whether or not those trends will continue. Also, the liberal position is not so much that renewables will take over sometime soon as that it is imperative that they take over sometime soon, and that we must do everything in our power, including through the channels of public policy, to make it so. Which, again, is not exactly scientific--I mean, the existence of global warming is a scientific fact, but it's a political choice to decide that we must do everything within our power to fight it.

But let's turn to those first two--that Medicaid, the VA, and Head Start work, and that school choice does not. Huh? What? In what sense are these "unscientific," except in that Charles C.W. Cooke disagrees with them? This is just... kinda weird, right? And, I mean, to the best of my knowledge the evidence is that Medicaid and the VA are both pretty good; the VA in particular is the best element of the U.S. health care system in providing satisfactory care to its beneficiaries, according to long-time polling data (perhaps tied with Medicare, another single-payer system which Cooke conveniently fails to mention...). I'm not sure exactly what the evidence is regarding Head Start, or exactly what it would mean for it to "work." Similarly I'm not entirely sure what it would mean for school choice to "work," or what the state of the evidence is there--I think it's pretty mixed, and also that the role such evidence should play in policy debates is complex. So what gives? Why is he calling these things "unscientific"?

I think the key lies earlier in the piece. Here's the paragraph two above the one I quoted:
It’s useful, too. For all of the hype, much of the fadlike fetishization of “Big Data” is merely the latest repackaging of old and tired progressive ideas about who in our society should enjoy the most political power. Outside of our laboratories, “it’s just science!” is typically a dodge — a bullying tactic designed to hide a crushingly boring orthodox progressivism behind the veil of dispassionate empiricism and to pretend that Hayek’s observation that even the smartest of central planners can never have the information they would need to centrally plan was obviated by the invention of the computer. If politics should be determined by pragmatism, and the pragmatists are all on the left . . . well, you do the math.
Okay, first of all, I know plenty of liberals who are awfully uncomfortable with the whole Big Data social management/optimization stuff. See, e.g., Kevin Drum, who has written multiple blog posts expressing his discomfort with grocery store loyalty cards that track all your purchases in return for savings. But let's focus in on that line about Hayek, and pair it with the stuff about Medicaid and school choice. This is why he asserts that these liberal beliefs are "unscientific": because they favor a relatively greater degree of central planning, and as we all know, Hayek proved that all forms of central planning are doomed to fail because information. Except, here's the thing: Hayek's works aren't science. His "observation" is not an observation in the scientific sense, but an observation in the philosophical sense. It's not even really political science, it's political theory. Hayek has an argument for why central planning won't work. Honestly, it's a pretty good argument, explains a fair amount about why communism-as-practiced doesn't work, and is a good reason to favor market-based solutions (like, say, carbon taxes) when available. But it's not a law of nature. Believing that a certain centrally-planned program might work isn't unscientific the way believing in a perpetual motion machine is unscientific. But that's clearly what Cooke believes. Medicaid is for him the equivalent of a perpetual motion machine, or time travel: it's just theoretically impossible, so if it looks like it's working that has to be an illusion somehow, or unsustainable, or something. Believing in Medicaid is automatically unscientific because Medicaid is impossible.

Except, y'know, it isn't, because the word of Hayek is not the gospel truth. Hayek gives us a reason to think that, most of the time (and he concedes there are exceptions! like pollution!), central planning is inferior to aggregated individual choice. But in any given context, that argument may or may not be applicable. There's a ton of evidence, empirical and theoretical, that health care is an area where it's just flatly inapplicable. Like, that every country with more centralized systems than ours gets far better results much more cheaply. The evidence is not ambiguous on this point. Moreover, the Hayekian point applies most clearly to cases where we don't know what the utility function is, i.e. where we don't have any good way to know what all our social goals should be. But we know what the goal of a health care system is: a healthy population. So the informational issues are far more limited. Plus, individuals just plain do not have the expertise to make all of their own health-care decisions correctly; that's what doctors are for. As for education, I dunno exactly what the theoretical arguments are, not in as much detail, but I do know that theory isn't necessarily controlling. Our goals for an educational system are matters of principle, but which institutions will lead to those results is a matter of empirical reality, which our theories might or might not adequately predict. If they fail, we revise our theories; that is literally the definition of science. Believing something that contradicts a theory isn't unscientific; clinging to a theory in the face of all the evidence is. Cooke seems not to understand that--in fact, his whole column just oozes anti-empiricism, heaping derision on spreadsheets and data like one of the old-school baseball guys Billy Beane supplanted in Moneyball.

So, to recap: he identifies six beliefs which he says liberals hold and which he says are unscientific. Two of them, no liberals really hold, certainly not any liberal nerds. One of them is about predicting the future of society, an area intrinsically beyond the purview of the scientific method. One of them is an area where lots of liberals are skeptical of the general scientific community's views, and the general scientific community is somewhat scornful of said skepticism. And the other two are places where the evidence is at worst mixed but where the "unscientific" charge is made on the basis of a certain piece of political theory. That's... a pretty poor batting average. And it's particularly odd given his sort of schizophrenic insistence that science is specifically defined by the peer-reviewed results generated by expert scientists and also that reliance on data is a terrible thing. (Does he think that peer-reviewed scientific articles don't use spreadsheets?)

I believe there's some pop culture reference about awarding no points and may god have mercy on your soul that's applicable here.

No comments:

Post a Comment