Friday, July 4, 2014

A talent to bemuse

Robert Ryan in God's Little Acre (1958).
Today's David Brooks is one of a kind of mini-series in which he pokes about in the backyard for bones borrowed from the opinionist Kevin Lewis of the Boston Globe.

Lewis performs an odd but very valuable little public service in the form of a daily blog at the National Interest where he runs, without comment, the author-written abstracts of social science papers recently published or accepted for publication.

David Brooks is a subscriber, and I guess regularly picks up on particular abstracts and copies them into his Rainy Day Social Science Prompts file, and then every once in a while (December 2010, March 2011, December 2012, and today) assembles [jump]
800 words' worth of them into an ersatz column, rewriting the abstracts for intelligibility and a little of that Bobo quirkiness. They are among his most inoffensive efforts; he doesn't especially choose them to push a propaganda line, I don't think, or anything other than providing us with a little amuse-gueule to while away our time and not take up very much of his.

But then again, it's Brooks, so there is some propaganda there, some anti-liberal trolling about girls playing sports, for example, and some anti-neoliberal trolling about Who Lost Vietnam. And because he has no idea that an author abstract is, by definition, self-serving, and that to find out what's going on in a paper and whether it actually found the stuff it claims to have found (let alone the stuff you claim it found when you're writing up your own version) you need to, um, look at the paper itself, he digs a couple of funny traps for himself.
Title IX has produced some unintended consequences. Phoebe Clarke and Ian Ayres studied the effect of sports on social outcomes. They found that a 10 percentage point increase in state level female sports participation generated a 5 or 6 percentage point rise in the rate of female secularism, a 5 point rise in the proportion of women who are mothers and a 6 point rise in the percentage who are single mothers. It could be that sports participation is correlated with greater independence from traditional institutions, with good and bad effects.
Or it could be that the results are totally spurious, as indeed seems to be the case with this one, which had apparently shown up at Deadspin, Freakanomics, and the Journal of Socio-Economics before Professor Brooks brought it to our attention.

According to Andrew Gelman at the Washington Post writing in May this year, this foolish paper was a classic case of correlation not amounting to causation. The researchers simply plugged in the state numbers, making no attempt to learn whether the girls who played high school sports had any relationship with the women who identified themselves as secular, had babies, and failed to marry their baby daddies. Thus there is no reason to suppose that the correlations mean anything at all.

It is pretty well known, in fact, from studies that ask the questions as this one did not, that the girls who play high school sports are more likely to delay first sexual encounters and avoid early pregnancy. I have no idea whether they are "secular" or not.
Hearts and minds may be a myth. Armies fighting counterinsurgency campaigns spend a lot of effort trying to win over the hearts and minds of the local populations. But Raphael Cohen looked at polling data from Vietnam, Iraq and Afghanistan and found that public opinion is a poor predictor of strategic victory. Public opinion is not that malleable, and its swings are more an effect than a cause. That is, counterinsurgency armies get more popular as they win victories; they don’t get popular and then use that popularity to win.
That's not exactly what he found, as we learn from the unedited abstract. He found that winning tactical battles might make the invader more ostensibly popular but being popular doesn't help the invader win strategic battles; if you care about what the population thinks you just don't win, which suggests to me that all victories are pyrrhic.

Cohen defended his dissertation, from which this article is a spin-off, at Georgetown University just last month, and the dissertation abstract clarifies his findings considerably:
1) historically, most successful counterinsurgencies have not been fought this ["hearts and minds"] way; 2) when this approach has been tried, it rarely proves effective; and 3) instead, military victory comes from successful population control. Population control, in turn, employs some combination of three sets of tactics: physical measures (e.g. walls, resource controls and forced resettlement), cooption (of local elite and often the insurgents themselves) and "divide and rule" strategies.
In other words, the way to win is to be that person you always swear you aren't, who aims to win and doesn't give a shit about the people you're conquering.

Since Cohen studied three counterinsurgencies in depth (Kenya, Malaya, Vietnam), it isn't clear whether he really has a big enough sample to generalize from, or even what he means by "successful". Certainly, though, the British defeated the Kikuyu "Mau Mau" uprising in Kenya and the more-or-less local-Chinese insurgency in Malaya before withdrawing from the empire business altogether a couple of years later; while the US was unquestionably defeated by the Vietnamese. I've never understood, and I mean since around 1963, how the last qualified as an "insurgency", as if the US had some kind of legitimate imperial interest in backing up the claims of the so-called RVN. I am sure that US forces used all the population control techniques Cohen advises alongside the "hearts and minds" stuff, though I don't suppose they ever got very good at coopting the enemy.

There's definitely a lesson, anyway: If you want to hold an empire, you need to be an imperialist, like, you know, Vespasian. And Trajan. And Hadrian. And if you want to be loved, imperialism may not be the best line of work for you. Or, more succinctly, imperialists need to be psychopathic assholes, it literally goes with the territory, so just don't.

Cute postscript:

An earlier version of this column misstated the findings of a study in the journal Economics Letters about corporate success. It found that C.E.O.’s were disproportionately less likely — not disproportionately likely — to have been born in June and July.

No comments:

Post a Comment