Freakonomics APR 11 2005
Two years ago, Stephen Dubner wrote an article for the NY Times Magazine on Steven Levitt, an economist with a knack for tackling odd sorts of problems. Last year, Dubner and Levitt collaborated on an article called What the Bagel Man Saw about the economic lessons gleaned from a man who's been successfully selling bagels on the honor system in offices for more than 20 years. Now Levitt and Dubner are out with a new book called Freakonomics: A Rogue Economist Explores the Hidden Side of Nearly Everything, an overview of Levitt's work and collaborations with other economists.
Dr. Levitt was kind enough to answer a few questions I had about the book:
jkottke: In Freakonomics, you state that you're interested in applying economic tools to "more interesting" subjects than what one may have learned about in my high school economics class. What's your definition of economics? Is it a tool set or a science or what?
Steven Levitt: I think of economics as a worldview, not a set of topics. This worldview has a few different pieces. First, incentives are paramount. If you understand someone's incentives, you can do a pretty good job of predicting their behavior. Second, the appropriate data, analyzed the right way are key to understanding a problem. Finally, political correctness is irrelevant. Whatever the answer happens to be, whether you think it will be popular or not, that is the answer you put forth.
jkottke: Your talent for ignoring seemingly applicable but ultimately irrelevant information (not that different from a professional-grade batter taking cues from certain aspects of a pitcher's mechanics and ignoring the extraneous ones in order to hit well), where does that come from? Good genes or was it all the books in your childhood home?
Levitt: If nothing else, I had an unusual home environment. My father is a medical researcher whose claim to fame is that he is the world's expert on intestinal gas (he's known as the King of Farts). My mother is a psychic who channels books. From an early age, my life was different from that of other kids. For instance, when I was in junior high, my father would wake me up at night to drill me with questions in hopes that I would be the star of the local high school quiz show.
jkottke: In looking at the world through data, you've investigated cheating schoolteachers, falling crime rates due to abortion, and the parallels between McDonald's corporate structure and the inner workings of a crack-dealing gang. What's the oddest or most surprising thing you've uncovered with this approach? Maybe something you still can't quite believe or explain?
Levitt: It's not the oddest result I've ever come up with, but there is one finding I have always puzzled over: when cities hire lots of Black cops, the arrest rates of Whites go up, but no more Blacks get arrested. When cities hire White cops, the opposite happens (more Black arrests, no more White arrests). It was an amazingly stark result, but I'm not quite sure what the right story is.
jkottke: In the chapter on the effect of abortion on crime rates, you and Stephen take care emphasizing what the data says and the strong views that people in the US hold on the issue of abortion. Still, if someone wants to twist your observations into something like "abortion is good because it lowers crime", it's not that difficult. Have your observations in this area caused any problems for you? Any extreme reactions?
Levitt: I have gotten a whole lot of hate mail on the abortion issue (as much from the left as from the right, amazingly). What I try to tell anyone who will listen -- few people will listen when the subject is abortion -- is that our findings on abortion and crime have almost nothing to say about public policy on abortion. If abortion is murder as pro-life advocates say, then a few thousand less homicides is nothing compared to abortion itself. If a woman's right to choose is sacrosanct, then utilitarian arguments are inconsequential. Mainly, I think the results on abortion imply that we should do the best we can to try to make sure kids who are born are wanted and loved. And it turns out that is something just about everyone can agree on.
jkottke: In the book, you say "a slight tweak [in incentives] can produce drastic and unforseen results". If you were the omnipotent leader of the US for a short time, what little tweak might you make to our political, cultural, or economic frameworks to make America better (if you can forgive the subjectivity of that word)?
Levitt: I would start by increasing the IRS budget ten-fold and doing a lot more tax audits. If everyone paid their taxes, tax rates could be much lower and otherwise honest people wouldn't be tempted to cheat. For some reason, everyone hates the idea. But we can't all be cheating more than average on our taxes. I think it would be for the better. And after I got done with that, I'd legalize sports betting, and I would also do away with most of the nonsense and hassle that currently goes into airport security.
jkottke: In the war between the film and music industries and their customers, there's an argument over how much the explosive increase in Internet piracy affects sales of CDs, movie tickets, and DVDs. Using the same data, the music/movie industry argues that sales are down because of piracy (or at least diminished from what they "should" be in a piracy-free marketplace) while the other side argues that sales are up and that piracy may actually have a beneficial effect. The question of "how does piracy affect record/movie sales?" seems well suited to your particular application of economic tools. Have you looked at this question? And if not, do you have sense of which special view of the data might reveal an answer?
Levitt: I have not myself studied the issue. I have a former student who has studied this issue. Alejandro Zentner. He argues that music sales are way down as a consequence of downloading. He uses the availability/price of high-speed internet across areas and relates that to patterns of self-reported music buying.
But on the other hand, I have a good friend Koleman Strumpf who has also written on this and comes to the opposite conclusion using a whole bunch of clever arguments.
This is a great issue - an important one and a tough one. Having studied both of these papers, I don't know which one to believe.
Thanks, Steven. For more information on Freakonomics, check out the book's web site -- which includes a weblog written, in part, by the authors -- or buy the book on Amazon. Check out also this email conversation between Levitt and Steve Sailer on the connection between legalized abortion and reduced crime in the 1990s, a short profile in Wired, and this profile in Esquire (free subscription required).
Update: Here's a Freakonomics excerpt from Slate on how distinctively black or white names affect a child's course in life.