April 16, 2005

Tierney on Levitt vs. Gladwell

John Tierney's second NYT op-ed column "The Miracle That Wasn't" reports on the debate Thursday between Malcolm Gladwell and Steven D. Levitt on what really caused crime in NY to go down in the 1990s: The Tipping Point (Gladwell) or Pre-Natal Capital Punishment (Levitt).

The essential problem with this type of debate is that it frames the issue too narrowly as: what caused murder to go down in the later 1990s? Instead, the full question should be: What were the causes of the murder rate going up in the late 1980s and early 1990s and its subsequent fall? When we look at the bigger picture, it's easier to get a more realistic sense of history than to simply assume that the early 1990s were the norm and thus we need some bestselling author to give us his unique theory:

Tierney writes:

It is an inspirational urban lesson from the 1990's: take back the streets from squeegee men and drug dealers, and violent crime will plummet. But on Thursday evening, the tipping-point theory was looking pretty wobbly itself.

The occasion was a debate in Manhattan before an audience thrilled to be present for a historic occasion: the first showdown between two social-science wonks with books that were ranked second and third on Amazon.com (outsold only by "Harry Potter"). It pitted Malcolm Gladwell, author of "Blink" and "The Tipping Point," against Steven D. Levitt, an economist at the University of Chicago with the new second-place book, "Freakonomics."

Professor Levitt considers the New York crime story to be an urban legend. Yes, he acknowledges, there are tipping points when people suddenly start acting differently, but why did crime drop in so many other cities that weren't using New York's policing techniques? His new book, written with Stephen J. Dubner, concludes that one big reason was simply the longer prison sentences that kept criminals off the streets of New York and other cities. [Undoubtedly right.]

The prison terms don't explain why crime fell sooner and more sharply in New York than elsewhere, but Professor Levitt accounts for that, too. One reason he cites is that the crack epidemic eased earlier in New York than in other cities. [Right, but why did the crack wars begin earlier in NYC?] Another, more important, reason is that New York added lots of cops in the early 90's. [Likely, although motivation matters too, as Tierney notes later.]

But the single most important cause, he says, was an event two decades earlier: the legalization of abortion in New York State in 1970, three years before it was legalized nationally by the Supreme Court.

The result, he maintains, was a huge reduction in the number of children who would have been at greater than average risk of becoming criminals during the 1990's.

No, there was not a big reduction in the number of children. Instead, there was a big increase in the number of conceptions. Let me quote Levitt's own book on what actually happened after the legalization of abortion: "Conceptions rose by nearly 30 percent, but births actually fell by 6 percent …"

Growing up as an unwanted child is itself a risk factor, he says, and the women who had abortions were disproportionately likely to be unmarried teenagers with low incomes and poor education - factors that also increase the risk.

This sounds plausible until you look at the illegitimacy rate, which continued to skyrocket. Instead, what happened was that more women got pregnant outside of marriage, and more boyfriends refused to marry them on the grounds that they could get an abortion instead. Some got abortions, some didn't, and the percentage of babies unwanted by their fathers went up.

But he says the correlations are clear: crime declined earlier in the states that had legalized abortion before Roe v. Wade, and it declined more in places with high abortion rates, like New York.

This shows how Levitt wins minds by framing the debate as why the murder rate went down instead of why did it first go up and then go down? But why not ask why the teen murder rate went up first in the places like New York that had lots of abortions in the early 1970s? It's logically bizarre to focus on purported later effects of abortion and resolutely ignore potential earlier effects.

Some criminologists have quarreled with his statistics, but the theory was looking robust at the end of the debate in Manhattan. Mr. Gladwell, while raising what he called a few minor quibbles, seemed mostly persuaded by the numbers.

"My first inclination," he joked at the beginning of his rebuttal, "is to say that everything you just heard from Steven Levitt, even though it contradicts things I have written, is true."

Gladwell appears a little out of his league in dealing with Levitt.

That's my inclination, too, as a less successful exponent of the same theory. (In 1995 I explained the crime decline with my version of the tipping point, the Squeegee Watershed, which became neither a buzzword nor a best seller.) In retrospect, the New York crime story looks like a classic bit of conventional wisdom, as the term was originally defined by John Kenneth Galbraith: an idea that becomes commonly accepted because it is "what the community as a whole or particular audiences find acceptable."

Well, a lot of people find the abortion-cuts-crime theory very comforting, even if they won't say it in public. But the issue is hardly whether it's comforting or not, but whether it's true.

My published articles are archived at iSteve.com -- Steve Sailer

No comments: