Saturday, June 30, 2007

Lott versus Freakonomics

Economist John Lott has been blasting Freakonomics author Steven Levitt for a while now because of the shoddy and questionable scholarship (see these archived posts). Here is a review of Lott's new book rebutting Levitt from The American:

John Lott, Loaded

Thursday, June 28, 2007

The economist’s new book fires back at Steven Levitt and other critics.

freedomnomicsFreedomnomics: Why the Free Market Works and Other Half-Baked Theories Don’t, by John R. Lott, Jr. (Regnery, June 2007) 275 pages, $27.95.

For years now, economists John R. Lott and Steven Levitt have been at each other's throats. They can’t seem to agree on anything, from guns to abortion. They’ve been collecting and analyzing data like mad, striving to prove their respective worldviews correct.

Ten years ago, Lott (then an adjunct scholar at the American Enterprise Institute, which publishes The American) authored a study showing that “right-to-carry” laws—which allow all non-criminal adult citizens to carry guns—reduce crime. In his bestselling Freakonomics, Levitt wrote that “[w]hen other scholars have tried to replicate his results, they found that right-to-carry laws simply don’t bring down crime.” Lott sued for libel; the judge threw out the complaint about the Freakonomics passage, but a trial about a related e-mail Levitt sent is set for October 1.

Now, with Freedomnomics, Lott aims to discredit Levitt—and in doing so, he takes on Levitt’s style. Like Levitt, Lott bases his writing heavily on his own academic research. And like Freakonomics, Freedomnomics is a collection of essays on a smattering of topics. Both books rely on econometric methods.

Freakonomics was one of the worst book titles in memory, and ripping it off probably isn’t the best idea (though the similarity in covers might help Lott sell books). And Freedomnomics’s positioning as a response to Freakonomics means it will always sit in the latter’s shadow.

What’s more, the book shares Freakonomics’s lack of focus. In part it’s a response to Levitt, but in other ways it’s a defense of the free market, even on topics Levitt didn’t touch. And some essays are neither anti-Levitt nor pro-market; Freedomnomics could more accurately have been called The Brief Book of Everything John R. Lott, Jr., has Ever Written About.

Nonetheless, most of what John R. Lott, Jr., has ever written about has been fascinating. He’s a great writer, especially for the general public, and the book renders lots of charts, graphs and statistical analysis into clear, uncomplicated conversation. In particular, those who haven’t followed Lott’s research and op-eds should read it.

As a response to Freakonomics, the book is often persuasive. The right-to-carry issue is near and dear to Lott’s heart, so he shows some restraint by dedicating only six of about 200 pages to the topic. He provides lists of studies finding that right-to-carry reduced, did not affect, or increased violent crime—more studies, by Lott’s count, came to the first conclusion than the second (Levitt’s). Freakonomics’s implying that no other scholars agree with Lott may fall short of libel, but it’s definitely untrue.

In saying that none of the studies have concluded right-to-carry increases crime, Lott classifies the conclusions of Mark Duggan’s “More Guns, More Crime” as equivocal. Lott explains, “after correcting four typing mistakes, sixteen of his thirty estimates actually show statistically significant drops in crime, while only one shows a significant increase.”

The “More Guns, Less Crime” thesis is certainly tenable, but Lott should have mentioned two government “study-of-the-studies” analyses—in the last few years, both the CDC and the National Academy of Sciences went through the available data on right-to-carry. Neither study saw any clear link, either way, between right-to-carry provisions and crime rates.

Lott and Levitt’s other main point of contention is abortion. In Freakonomics, Levitt claimed that 1973’s Roe v. Wade culled out potential ne’er-do-wells. The decision decreased the birth rate by 6 percent, and unwanted children tend to commit more crimes. Abortion thus, he says, helps explain the 1990s crime drop—crime fell first in states where abortion became legal before Roe.

In joining the general-public discussion, Lott’s a bit late to the party. Levitt’s original paper with John J. Donohue III came out in 2001, and Lott published an academic response at the time. It’s been two years since The Economist ran “Oops-onomics,” a description of flaws in Levitt’s abortion work that were first unearthed by two Boston Federal Reserve economists.

But Lott’s arguments are powerful, and since so many people have bought into the Levitt view that abortion cuts crime, they remain worthwhile. Illegitimacy went up, not down, after Roe v. Wade. In the 1990s, crime fell among older criminals before it fell among the younger, “culled” ones. While crime fell first in states that legalized abortion first, other states had permitted abortion quite liberally for medical reasons. Some of these “banned” states had higher abortion rates than the “legal” states did, yet saw no comparable drop in crime. Children born in the four years before Roe had markedly lower crime rates than those born in the first four post­-Roe years. Canada had the same 1990s crime drop the United States did, though it hadn’t legalized abortion until 1988.

Lott takes it a step further, though—citing two studies by other researchers in addition to his own work—and asserts that abortion’s availability actually causes illegitimacy. If this is true, abortion’s effect on crime is the opposite of what Levitt’s famous theory contends.

The idea is that legal abortions lead to more casual sex, which obviously causes more pregnancies (conceptions increased 30 percent after Roe). However, some women can’t go through with abortions when they thought they would, and some women who don’t believe in abortion have more sex to compete for mates and keep up with social mores. Additionally, men are less likely to marry women they’ve accidentally impregnated. To make things worse, adoption rates for unwanted children plummeted after Roe.

Abortion may have culled out 6 percent of the would-be population, but by Lott’s calculations, the accompanying disintegration of two-parent families more than made up for the “gain.” At the very least, many factors contributed to the breakdown in nuclear families (most notably the welfare state), and it’s difficult to sort them out.

Freedomnomics’s second function is to provide a wide-ranging defense of the free market. The book is dedicated to the memory of Milton Friedman. Many experts have made careers out of finding “market failures,” or instances where freedom has bad consequences. Lott concedes markets aren’t perfect, but he demonstrates how some common criticisms don’t really indicate failures at all.

His approach, unfortunately, is like trying to swat a roomful of flies: Lott can hit a few, but there are simply too many allegations of market failure for one book, especially a short one, to cover.

His picks do tend to be enlightening. For example, Levitt has argued that real estate agents capitalize on people’s naivette. They sell their own homes at higher prices than they sell their trusting clients’ homes—they’re in a hurry to get to the next commission, so they sell too fast. But Lott points out that the difference is only 2 to 3.3 percent, and that agents know what kinds of improvements will help a house sell. Even when they suggest these improvements to clients, many clients don’t take the advice.

Besides, he adds, anyone in any profession knows how to get the most value. Doctors probably find better health care than other people do. That doesn’t mean the market isn’t working.

Another Levitt idea is that a car’s value drops precipitously as soon as it leaves a dealer’s lot. If an owner sells soon after buying it, buyers wonder why—there must be something wrong with it. Thus the market takes advantage of sellers of barely used cars. But Lott provides data indicating that’s simply not the case. Most cars’ values decline slowly and steadily, with no plummet right after purchase. In cases where manufacturers set maximum prices on new cars, creating a shortage, sometimes barely used cars even sell for more. This is because the market created mechanisms for evaluating cars, including transferable warranties and used car certification.

Lott also suggests why alcohol is so expensive in restaurants—when a customer buys a glass of wine, he tends to spend a lot of time sipping it. The waiters can’t clear the table to make room for other customers, so the owners lose money. Drinkers rent tables, rather than forking over extra money for greedy profiteers.

Outside of the Levitt versus Lott controversy, one of the book’s most fascinating chapters argues that female suffrage has led to government growth. Though many designate the New Deal as the dawn of Big Government, spending increases began before that. States that gave women the vote before the federal government experienced higher spending at the state level.

Lott provides a variety of reasons that women support spending, and have done so increasingly. One interesting factor is no-fault divorce. Women tend to make less than men do, and sometimes married women stay home altogether—so when they can’t bargain with their husbands over divorce terms, they’re more likely to end up poor. Thus, they have strong reason to value an economic support system.

All said, Freedomnomics entertains, educates and argues forcefully. Scholar wars may seem petty at times, but this book shows they lead to some worthwhile endeavors.

Robert VerBruggen is Assistant Book Editor at The Washington Times.

No comments:


Blog Archive