Freakonomics

Working at Google comes with many unique benefits, the most notable being the three free meals a day, but perhaps the coolest is the series of guest speakers that we get on campus. In the past year or so, I've had the opportunity to attend talks by Madeleine Albright, Jimmy Carter (got to shake his hand), Colin Powell (shook his hand, too), Thomas Friedman (with a surprise appearance by...Robin Williams?), and Peter Diamandis (Chairman of the X Prize Foundation), among others. Today we had Steven Levitt and Stephen Dubner, the authors of Freakonomics: A Rogue Economist Explores the Hidden Side of Everything, a book that Hubert turned me onto while we were in Hawaii.

The subtitle says it all. Levitt is a renowned maverick economist who uses the science to tackle unconventional and seemingly random topics, such as, why do real estate agents act in a way that is contrary to their clients' goals, is sumo corrupt, why did Seinfeld never crack the top 50 among black viewers, what impact does a name have on a person's life, what's more dangerous, a gun or a pool, why do drug dealers live with their moms, and do good schools make any difference? But the topics aren't as random as you'd think. They are all connected by the same methodology: analyzing the incentives, information gaps, and conventional wisdoms that are in play in any given situation, and crunching the numbers to identify their implications, which are usually quite broad. Levitt just happens to apply them to specific questions that interest him, even if they are sometimes obscure or seemingly frivolous.

But Levitt doesn't shy away from the big issues either. One of my favorite chapters of the book puts forth his take on the crime drop in the 90's, and the conclusion he presents is inflammatory but compelling. Leading up to the 90's, crime was uncontrollably on the rise, and dire predictions were being made by experts about the future blood bath facing America. Then, seemingly out of nowhere, the crime rate plummeted, and those same experts scrambled to come up with answers, most of which were as wrong as their initial doomsday predictions. Two theories to which he does lend credence are the increased reliance on prisons, and the increased number of cops. Prisons have a deterring effect on crime, but they also act as a prophylactic, containing those most likely to commit crimes, and more cops means greater enforcement. But the factor that may have had the greatest impact, Roe v. Wade in 1973, wasn't even conceived (no pun intended) as a possibility until Levitt first posited it in 2001. Seventeen years after the landmark case, as the unborn children of the generation (most with a background correlated to the highest indications of criminal behavior) would otherwise have been reaching their peak criminal years, the crime rate that had been growing uncontrollably for years had suddenly reversed itself completely.

The deeper social implications of his argument are disturbing, and they predictably created a furor on both sides of the political dividing line, but it was not his intention to make a social statement, nor is that where his interests lie. He's had the opportunity to work for both the Clinton and Bush White Houses, and both times he has declined. His thinking is innovative (out-of-the-box, if you will), and he wants the freedom to use it to explore the questions that interest him, without the constraints of political interests. His research is purely for the sake of exploration.

But whether he likes it or not, the potential applications of his research to more serious issues are obvious, even when he goes on wild tangents (sumo?). The later chapters on parenting, schools, and other principles of child-rearing drew lots of questions from the parents in today's crowd. They see him as an expert with the ability to find the answers to any question, and what could be more important than parenting? (according to him, a lot)

Occasionally, his research comes across as weak, but I think it's largely due to the fact that he wrote this book to make it accessible to a broad audience. So while it doesn't have the detail of your typical research paper, it is much more palatable for the mainstream. I personally would have enjoyed more detail. He often prefaces a statement with the vague "Studies have shown...", and then, as if predicting the rebuttals, follows up with the even more vague "Even when controlled for X, Y, and Z..." I'm prone to accepting his conclusions, because he presents them in such a logical manner, and he never fails to acknowledge the obvious counter-arguments, but the lack of details does leave me wanting more information about methodology. But I'm guessing the book wouldn't be the fixture of the non-fiction best seller list that it is were it as dense as I'd want.

As it is, the book is a very quick and easy read, and because it is just a collection of snippets, you can read it in parts without losing anything. And if you get to the end, you'll be rewarded with a list of baby names that are most highly correlated with life-long success (though I don't put much credence into it...Andrew made the top five). Beats the ending to Harry Potter.