Thursday, February 10, 2011


After buying this book due to a few glowing recommendations, I opened it one evening expecting to read just the first few pages -- and instead I sat for a full hour, unable to put it down. The writing style was so engaging and conversational, a kind of storytelling in nonfiction, and every paragraph held interesting observations and factoids.

The book centers around a fairly simple premise: what can numerical data tell us about some obvious (or not so obvious) question? In come cases, the answers seem reasonable and straightforward. For instance, a child is statistically far more likely to die by drowning in a backyard swimming pool than in an accident involving a gun in the home -- and yet the latter is generally regarded as a far more dangerous, frightening thing to have around.

Other topics were much more controversial. An entire chapter is devoted to exploring the connection between Roe v. Wade and the dramatic crime drop in the 1990’s. The argument went like this: children who are born into neglect and poverty are more likely to become involved in crime as adults, but with legalized abortion, many of these unwanted children simply weren’t being born. That the dramatic drop in crime rates came roughly twenty years later -- right around the time those born during this period would have been reaching maturity -- was no coincidence.

It was a fascinating argument, though not one that I’m particularly qualified to analyze, what with my not being a statistician or economist or sociologist or anything similarly relevant. On the one hand, there’s an understandable, common-sense kind of cause-and-effect relationship in what Levitt and Dubner suggest -- but on the other hand I worry that this kind of answer is too simplistic to explain what’s undoubtedly a very complex web of social phenomena. There’s a Wikipedia article here that summarizes their theory in some detail and links to several other sources on the topic.

Another chapter explored the question of “good schools” vs. poorer quality schools. How much influence does the quality of the school really have on a child’s academic performance? According to Levitt and Dubner, the answer is little to none -- but this was another conclusion I questioned as I was reading.

In their analysis, outcomes were measured only by the grades each child achieved, and as far as I could tell, there was no data regarding how many of those children went on to college, how well they performed there, whether they eventually graduated, etc. Is an “A” in the overcrowded inner-city school worth the same in the long run as an “A” in the upper-class suburban school? I wished the authors had taken more time to explore these kinds of questions.

One other minor annoyance was the “padding” in the book: it was only 315 pages to begin with, but the book itself ended on page 211. The remainder contained “Bonus Material” -- short articles and essays that had apparently been published elsewhere prior to the book’s release. Though these materials were for the most part interesting, some of them seemed very redundant, duplicating content that had already been addressed adequately in the book itself.

Overall, I had mixed feelings about this book; despite the various concerns and frustrations described in this review, it was still a really enjoyable book to read. On balance I think I would recommend it with caveats.

By Steven D. Levitt and Stephen J. Dubner
My Rating: 3 out of 5
3 out of 5

No comments:

Post a Comment

Feel free to add a comment, even if this post is weeks or months old by the time you find it. I always love hearing what people think!