Freakonomics has caused a sensation.  With a stroke of the pen, economics went from dismal to sexy.  A book was talking about regression analysis and people were buying it: empirical social science had become hot!  Yet for all its success, in the two years since Freakonomics was published, economist Steven Levitt has been variously lionized, demonized, and sued over what he and his coauthor, journalist Stephen Dubner, wrote in the best-seller.  As Levitt has seen first-hand, the ivory tower is an illusion; academic life in the 21st century is thoroughly public.  As scholarship move from journals to the press, popular books, and blogs, the rules of academic debate are changing.  The public forum is not peer-reviewed, and it has become increasingly important that academics and the general populace recognize this.  Now in its revised and expanded second edition, Freakonomics continues to fascinate readers.  But there's a deeper story to tell: a tale of guns, abortion, and the perils of academic celebrity.

It takes a plucky outsider to prod a complacent intellectual class and benighted public in the direction of truth.  This is the central theme of Freakonomics despite Levitt and Dubner's protestations that their book in fact has no central theme.  That Levitt is meant to be just such an outsider is clear from even the subtitle: "A Rogue Economist Explores the Hidden Side of Everything."  It's a nice idea, the academic hero up against the world, but Levitt is no outsider.  In fact, he's about as establishment as you can get.  If the PhD in economics from MIT doesn't convince, his John Bates Clark Medal should.  In light of his sterling pedigree one might suspect that Levitt is meant to be a rogue in that his views are especially heterodox.  There's no question that Levitt is uniquely talented, but he's no revolutionary.  Levitt has brought economics to Everyman, but he hasn't changed the way it's practiced.  Freakonomics is simply a vehicle for Levitt to present his research to a popular audience.

Of course there's nothing wrong with that.  Levitt has every right to share his ideas with the world.  His establishment credentials are neither a reason to summarily dismiss nor accept his conclusions; the work should speak for itself.  And revolutionary or not, Levitt's research has done much to promote economics.  Despite its somewhat dubious theme, Freakonomics is a good read: well-written and entertaining.  Levitt's fascinating research has uncovered corruption in Sumo Wrestling and the Chicago Public Schools, the inner-workings of a crack gang, and a link between crime and legalized abortion.  Although the final chapter, a discussion of the economic impact of names, drags a bit, the pace is generally lively and the prose clear.  Levitt has remarkable ideas, and Dubner clearly knows how to make them accessible to a wider audience.

More than a million copies later, Freakonomics has become something of a franchise, spawning a blog [www.freakonomics.com] and a series of occasional pieces written for the New York Times.  The revised/expanded edition is as much a testament to this success as to the difficulty of writing a topical book.  The only significant difference between editions is the added "Bonus Material," a selection of pieces from the Freakonomics blog and newspaper columns presented alongside the original 2003 article that introduced Dubner to Levitt, and Levitt to mainstream America.  The appended columns, and especially blog entries, allow Levitt to respond to a changing world without interrupting the flow of the book, but at the same time, their position at the end of the volume makes them feel like something of an afterthought.  If this information is important enough to include in the revised edition, why isn't it in the body of the text?  Moreover, the bonus material represents an uneasy blurring of media: blog entries in their original form are not subject to editorial scrutiny.  This is the key to understanding the problems of Freakonomics: examining what crept into the bonus material via www.freakonomics.com, and what remained in the virtually unchanged body of the text.

In April 2006, economist John Lott sued Levitt for defamation (Glenn).  Among the items at issue was a particularly flippant passage in Freakonomics criticizing Lott's contentious "more guns, less crime" hypothesis, the idea that liberal right-to-carry laws reduce violent crime.  Criminals, the argument goes, are less likely to attack someone who might be armed.  Hence, allowing people to arm themselves will lower crime.  Levitt has his own ideas about what lowers crime, but before making his case in chapter four of Freakonomics, he criticizes numerous other explanations for the vaunted crime drop of the 1990s.  None of these critiques, however, is as pointed as the one he directs at Lott.  There are indications that this dispute is more than academic.  The two men, it seems, have nursed a private enmity for years (Glenn).

It all came down to the word “replicate”: "Lott's admittedly intriguing hypothesis doesn't seem to be true.  When other scholars have tried to replicate his results, they found that right-to-carry laws simply don't bring down crime" (Levitt and Dubner, 122).  The casual reader likely took this passage to mean that Lott's result was erroneous, or more precisely that when other researchers subjected Lott's data to the same statistical analysis, they failed to discover a link between right-to-carry laws and violent crime.  But the truth is far more complicated.  In the guns and crime debate, data and models are always contentious.  What Levitt and Dubner should have written was that other scholars, using different data, different statistical models, or both, have concluded that there is no relationship between guns and lower crime.  This would have been technically accurate, if a tad unfair.  According Professor Carlisle Moody of the College of William and Mary, an expert in this area, approximately as many papers support Lott's hypothesis as call it into question (as quoted in Glenn).

Lott alleged that by using the word "replicate", which in its strict sense means "to make an exact copy of" (Oxford American Dictionary), Levitt had essentially accused him of cooking his results (Glenn).  Of course, if one defines replicate more loosely as "corroborate," Levitt is correct: other, though not all scholars disagree with Lott's result.  In January 2007, Judge Ruben Castillo dismissed the portion of Lott's complaint stemming from Freakonomics, ruling that the use of the word "replicate" was too vague to constitute defamation.  He did, however, allow the second part of the complaint, based on an email exchange, to proceed (Castillo).  What Levitt and Dubner wrote may not have been criminal, but it was certainly irresponsible.  It's perfectly legitimate for Levitt to disagree with Lott's conclusions, data, or methods.  But in a peer-reviewed article he could never have dismissed research that supported Lott's result without explanation.  The revised edition of Freakonomics gave Levitt an opportunity to tone down his language or mention that some scholars agree with Lott's result.  He chose not to.  This is especially interesting in light of something Levitt did choose to add to the revised edition, after one of his own theories came under fire.

Why did crime drop in the 1990s?  Levitt's controversial answer is legalized abortion.  Based on a paper written with John Donohue (Levitt and Donohue), the idea isn't really all that surprising. Children born into extreme poverty are more likely to become criminals.  Women whose children would otherwise be born into extreme poverty are more likely to have abortions.  The effect of legalized abortion, then, is to keep that segment of the population most likely to commit crimes from ever being born.  At the beginning of the 1990s, nearly eighteen years after Roe v. Wade legalized abortion nationwide, a wellspring of criminality had suddenly evaporated from society.  Levitt and Donohue showed that this intuitive idea finds support in U.S. crime data; legalized abortion really does appear to have lowered crime.  It comes as no surprise that this result became a major point of political contention, but far more interesting was the academic controversy that developed shortly after Freakonomics' initial publication. As it turned out, Levitt's abortion-crime paper contained a major flaw.

Soon after Levitt and Donohue's abortion paper appeared in the prestigious Quarterly Journal of Economics, Christopher Goetz and Christopher Foote, both of the Federal Reserve Bank of Boston, took a closer look at the result.  What they found surprised them: Levitt and Donohue's computer code contained an error that, once corrected, substantially weakened the abortion-crime link (Goetz and Foote).  When the two published their findings in a 2005 Federal Reserve working paper, it was a major news story, appearing in the Economist's “Economic Focus” (Economist).  Levitt quickly replied on www.freakonomics.com with a twofold strategy.  First he downplayed the importance of the section of his paper that had been called into question: "the approach they [Foote and Goetz] criticize was one of four distinct pieces of evidence we presented in that paper; they offer no criticisms of the other three approaches" (Levitt and Dubner, 256).  Second, he criticized the data Foote and Goertz used, which happens to be the same data Levitt had used throughout his original paper.  Using better data, and the proper statistical techniques, Levitt argued, the evidence in favor of the abortion-crime link is actually stronger than before.  

Levitt's response is perfectly reasonable, and may well be correct.  But even for someone trained in econometrics, the branch of statistics concerning the estimation of economic models, the details are hard to follow.  The study involves hundreds of variables, some of which are very inaccurately measured.  The statistical techniques required to compensate for this inaccuracy are complicated to say the least.  In the end, peer-review will sort out whether Levitt or his critics are correct.  As surely as Foote and Goetz are shopping for a journal to publish their critique, Levitt and Donohue are writing up a reply of their own.  But this may take years.  In the meantime, by including his response in a popular book, Levitt has effectively guaranteed that thousands more people will hear his side of the story than ever bother to read Foote and Goertz's paper.  The point is not that one is right and the other wrong, but that a member of the public has no way of telling the difference.  Levitt is an expert, and people trust experts.  But experts, like everyone else, face incentives.  With an impressive reputation at stake, is it reasonable to suppose that Levitt is entirely disinterested in the outcome of the abortion-crime debate?  And with such an attractive target as Levitt, would anyone believe that ego plays no role in Foote and Goertz's critique?  As Dubner and Levitt remind us, "it may not be surprising to learn that experts…can be self-interested to the point of deceit" (Levitt and Dubner, 81).

Evaluating Freakonomics is a difficult task. While ostensibly attempting to shatter the conventional wisdom, through his popularity Levitt only supplants it with his own ideas.  Though skeptical of experts and their motives, Freakonomics is ultimately written by one for an audience of laymen, and cannot escape this fact.  Add to this the more general problems of social science and you have a very problematic book indeed.  But to linger on these concerns is to miss the point.  Freakonomics is more than an enjoyable read - it is a chance to learn about some of the most exciting contemporary research in applied economics up close.  Though the topics of his research may seem trivial to some, Levitt will ignite the layman to think like an economist.  That at the very least makes Freakonomics well worth a look.

Levitt is not deceptive; he's simply human.  In his informal writing Levitt has said nothing that could not have been overheard in the staff room of any economics department in the country.  Yet informal though they may be, blogs and popular books are mass media aimed at non-specialists.  And because of his celebrity, Levitt reaches more people than practically any other academic of our day.  If economic science is to be thought of as anything more than a matter of opinion, scholars must keep their public comments sober and nuanced, restricting more heated debate to scholarly journals where it can be properly moderated.  The problems of Freakonomics are more generally the problems of academic celebrity: we must never allow the aura of academic respectability to obscure the fact that science is a social activity, and its practitioners all too fallible.  It's important to question the conventional wisdom.  But be sure to question Freakonomics as well.