Your thoughts on Steven Levitt's book?

[softwareNerd]

The Wall Street Journal recently ran a short article about Steven Levitt, a professor of Economics at the University of Chicago. Mr. Levitt seems to specialize in the statistical analysis of some interesting topics.

Some of his papers: "Catching Cheating Teachers", "The Political Economy of an American Street Gang", "...Corruption in Sumo Wrestling."

I have not read any of the papers. He has now published a book, titled "Freakonomics : A Rogue Economist Explores the Hidden Side of Everything".

Has anyone read the book? If so, was it interesting? Would you recommend it?

Edited July 12, 2005 by softwareNerd

[scottkursk]

The Wall Street Journal recently ran a short article about Steven Levitt, a professor of Economics at the University of Chicago. Mr. Levitt seems to specialize in the statistical analysis of some interesting topics. 

Some of his papers: "Catching Cheating Teachers", "The Political Economy of an American Street Gang", "...Corruption in Sumo Wrestling." 

I have not read any of the papers. He has now published a book, titled "Freakonomics : A Rogue Economist Explores the Hidden Side of Everything".

Has anyone read the book? If so, was it interesting? Would you recommend it?

It's really less a book than more a collection of essays. I must say it was a pretty fun and light read but I was left feeling a bit empty. Some of his ideas were very compelling and well thought out but in so many cases it was like he was inches away from completely sewing up his ideas and then he completely misses the point entirely.

I think the chapter on Roe v. Wade having an effect on crime rates is probably one of the better sections. It's one of the best documented as well as the one with the most supporting evidence from third parties.

When it comes to the sumo wrestling, I think he completely misses the point about honor. He spends a great deal of time discussing the role of honor and its especially important role in sumo. I think he forgets that a successful sumo would be willing to give up a little face so that a man doesn't have to quite literally wipe someone's rear for six months. He applies western standards to the system and assumes that the only reason someone would deliberately loose a match would be for money. I could easily see someone losing a match that would cost me no loss of "face" or prize money but allow a fellow sumo to save face. I think he is very much wrong about the graft issue. Not to say that it doesn't happen, I just think it isn't the issue he makes it out to be.

The cheating teachers is a mixed bag. It seems to argue against standardized testing in as much as he says that it encourages cheating by its very nature. Or more precisely the reward systems for good versus bad almost always turns out bad. He doesn't seem to offer any solutions, just to say that it's a problem. I found his methodology for testing for cheating and then proving the cheating as well as testing again for good teachers particulary interesting from a numbers geek point of view very interesting.

Overall, I'd give the book a B+. He has some really good ideas but he so fails to deliver on some of his ideas. I forget how old he is but I seem to remember but I think he's not even 30 yet. So hopefully given time he'll snap to it and be able to refine his methodology. It's worth a quick light read. Especially if you are a person that is a numbers freak.

[softwareNerd]

I echo Scott's mild recommendation of the book, and also his comment about feeling a bit "empty". I think the reason for the emptiness is that the authors present data and information, but make very little normative commentary, nor do they make any attempt to tie the various chapters together. Levitt proudly claims that he has no theme! Indeed, an honest title could be: "Summaries of various unrelated Statistical Studies".

The findings in the studies range from the obvious to the interesting. The obvious ones: real estate agents -- on the average -- do not always act in the best interest of their clients, some teachers help their students cheat on standardized tests, people lie in personal-ads when they're looking for a date.

The study about drug-gangs was mildly interesting. The author demonstrates how illogical a choice of profession it is to be a gang-member. The title of that chapter sums it up: "Why do Drug Dealers still live with their moms?"

The other study that was interesting (not so much the study, but the final conclusion) was: Does the name you give your child impact their future? While the author concludes that a name does not matter, I find it curious that many people read the book and do not get it. Reading the book actually caused them to be careful about selecting a more "upper-class" name. (This is from examples I have seen on forums/blogs and also from one person I actually know.)

The study that I found most interesting was: What factors caused crime rates to drop from the late 70s. I had always bought into the Guliani 'broken window' idea that if one is strict about small crimes, that will help reduce the larger crimes. The author says there is no relationship. The decrease is explained better by the increase in the number of police. On the other hand, "the evidence linking increased punishment to lower crime rates is very strong".

The conclusion that comes completely out of left field (and the one that had some conservative reviewers riled up): the change in demographics from legalizing of abortion was one of the three major contributors to the decrease in the crime rate!

The author had a web site and a blog for anyone who'd like to get a better flavor of the content before reading the book.

[softwareNerd]

I least expected William Bennett to pick up (link) on the causal relationship between abortion and crime made in "Freakonomics". I think it shows a certain degree of honesty.

[scottkursk]

I least expected William Bennett to pick up (link) on the causal relationship between abortion and crime made in "Freakonomics". I think it shows a certain degree of honesty.

Bill Bennett will occasionally surprise you. He's not the died in the wool necon that infests the Republican party today. He's more of an old school Russell Kirkish conservative. I disagree with A LOT of what he has to say but I still do have respect for him. But now that he's not in the game so to speak he has a right to speak his mind. Kind of like Goldwater after he got out of politics. My respect for him jumped immensly.

[MightierPen]

I really enjoyed the book, but also felt it was a bit empty. I particularly enjoyed the part where he proved that money doesn't win elections.

[softwareNerd]

I really enjoyed the book, but also felt it was a bit empty.

Interesting, because I had the same feeling, and Scott said the same thing in his post above. For me, the emptiness was a feeling like: what's the point of all this? I think it comes from the authors's very intentional attempt to avoid a theme, and to avoid normative conclusions. I could have stopped at each chapter and chewed on the point of it for myself, but many of the topics didn't seem to be worth the time: too light.

[scottkursk]

I got the audible.com version of the book to listen to while I worked out. Something struck me that I didn't read before. In the introduction (the audio one at least since I don't believe I read the written one) he actually brags about the fact that the book is not a book per se but a collection of papers and essays. He earned his chops at Harvard sitting at a table full of subjectivists and nihlists and felt in good company. So that may something about his view on logic and why the book is disjointed.

Still, I maintain my B grade for the book. Though I'm glad I have a better understanding of why the book is the way it is.

[mrocktor]

I'll further reinforce the opinions given: the subject matter is interesting for the most part, the data is good (better in some sections), the presentation is apalling.

The fact that the author does not attempt to derive a theory, nevermind any normative consequence, and is proud of it (for some reason) takes the "book" out of the book. It is a statistical report, even calling it an essay is stretching it.

On the other hand, every issue - when analyzed from the Objectivist point of view - shows exactly the result you would expect if Objectivist metaphysics, epistemology, ethics and politics are valid. Obviously. The author's lack of interest in analysis at least has left the data free of the manipulation typical of someone with an ingrained altruist mentality or trying to prove something that is wrong (that welfare is good, for instance).

mrocktor

[softwareNerd]

The author's lack of interest in analysis at least has left the data free of the manipulation typical of someone with an ingrained altruist mentality or trying to prove something that is wrong (that welfare is good, for instance).

Yes, that was the saving grace; better to have raw data than faulty conclusions that have to be undone.

[ds1973]

It's been a few months since I read it, but it struck me that there are economists with a philosophy and Economists without one. It's not clear from the book that Levitt has a philosophy (or is willing to publicize it).

Don't expect Levitt to objectively identify behavior as "good" or "bad". As an economist, he is only concerned with how various "incentives" result in a particular outcome. Overall mrocktor's summary is right on. Levitt presents data and draws some logical conclusions from his overall analysis but don't expect him to apply any objective judgement to this work.

Just prior to the abortion chapter, there is a NY times excerpt in which it is written about Levitt: "He has little taste for politics and even less for moralizing. He is genial, low-key and unflappable, confident but not cocky. He speaks with a considerable lisp. His appearance is High Nerd..... He was a good golfer in high school but has so physically atrophied that he calls himself 'the weakest human being alive' and asks Jeannett to open jars around the house." The whole excerpt seems to try and elevate him by degrading him. It's disgusting..

I'm experiencing a similar frustration with this lack of philosophy in "Seeing what's next; Using the theories of innovation to predict industry change". In a section of this book, the authors anaylze how govt regulation in the telecommunications industry either led to more or less innovation. The whole time during this book (I'm only 1/2 way through), the Objectivist in me is screaming "Laissez-faire is the only proper system" while the authors dispassionatly give the government advice on how to implement "successful intervention" in various industries.. It makes me sick to my stomach that so many people have taken it as axiomatic that government intervention is a legitimate, and effective means to improve the economy.

More and more, I see how dangerous intelligent, productive people with no philosophy can be. It matters little how their work is used, only that they are allowed to do their work; and if the government funds it, all the better. Rand chose an appropriate title for "Philosophy: Who needs it". We all need it or else we wander around not giving a damn how we fund the fruits of our labor or how they are used.

Demetrius

[Scott_Connery]

I can't reccomend it. It was perhaps intersting light reading, but there were constant assertions witouth anything to back them up. The whole thing felt like I was being preached down to by someone who was very lazy academically.

[katherine19]

The fact that the author does not attempt to derive a theory, nevermind any normative consequence, and is proud of it (for some reason) takes the "book" out of the book. It is a statistical report, even calling it an essay is stretching it.

The book is not intended to extend existing microeconomic theory. Rather, I view it as an introduction, for novices, to the economic approach, particularly as expounded by Gary Becker and the Second Chicago School of Thought.

Furthermore, modern economics is frequently more interested in establishing causal relationships than moralizing. That this end is largely accomplished by means of statistical analysis shows the economic discipline's dedication to testing deductive theory with inductive methods. One should not summarily dismiss Levitt's ability to find meaning in numbers.

Obviously. The author's lack of interest in analysis at least has left the data free of the manipulation typical of someone with an ingrained altruist mentality or trying to prove something that is wrong (that welfare is good, for instance).

This statement contradicts the earlier one, calling Levitt's work a "statistical report." Statistics do not spontaneously arise of their own volition from any data set. Rather, the analyst must apply rigorous statistical methodology to understand how numerous variables affect an outcome. However, Levitt spares the reader his econometric analyses because such data would not interest most people. Furthermore, causality cannot be established by the suspension of analytical thought. Although some readers may not be satisfied with Levitt's refusal to make normative statements, I would not call his book lacking in analysis. Levitt is not out to make unequivocal, authoritarian statements about the nature of reality; his book is really intended to be an exposition of what economic reasoning can accomplish. If you prefer your nonfiction with a heavy-handed dose of moralizing, look elsewhere.

Edited September 3, 2006 by katherine19

[mrocktor]

The book is not intended to extend existing microeconomic theory. Rather, I view it as an introduction, for novices, to the economic approach, particularly as expounded by Gary Becker and the Second Chicago School of Thought.

The book contains no economics at all. It is a exposition of compiled data. The author claims in the introduction that he is not establishing causation merely identifying correlation. If the "Second Chicago School of Thought" prides itself on looking at lots of data and drawing no conclusions - you could be right.

This statement contradicts the earlier one, calling Levitt's work a "statistical report." Statistics do not spontaneously arise of their own volition from any data set. (...)

How does: "the author does not derive a theory or normative conclusion" contradict "the book is a statistical report"? Thats right, it doesn't. Sure, it takes work to glean information from a data set. Sure, the author may have implicit theories he was trying to prove. Irrelevant. If he wants to propose that the right to abort reduces crime - he should do so and support the argument with the statistics. Reporting the statistics and not drawing the conclusion makes him a coward or an idiot (or just a complete subjectivist, which is a subset of the second).

I'll just publish a study on the statistical correlation of the number of vegetarians with deforestation - you can draw your own conclusions on that too

mrocktor

Edited September 4, 2006 by mrocktor

[DarkWaters]

Has anybody read the Revised and Expanded edition of this book? How do the new essays compare to the old? I remember reading at least one of them in the New York Times, but I forget which one.