Freakin Awesome

OK, first let's dispense with the pleasantries. Loved the book, wanted more, etc. With regard to the ideas found in the book, I would say that if there was a "unifying theme" it was that economic tools can be used in interesting and insightful ways, however, this book removed the economic analyses to such a large degree that one couldn't really check the underpinnings for mistakes. That is not to say that I assume that there are any, but what about another respectable economist who comes to precisely opposite conclusions, having claimed to have controlled for roughly similar variables? It seems like this book could have fruitfully given a way to distinguish between good economics and bad economics. It seems as though there are so many "well-respected" economists out there, and yet it's almost impossible for most to tell whether or not they overstep their "good economist" capacity on a regular level (Paul Krugman, perhaps?). In any event, I think the strategies are well-described in the book, but they seem anecdotal more than formulaic; as the authors note, "It should be said that regression analysis is more art than science" (163). If that's the way we should look at it, what aesthetic philosophy can we use to separate good regression analysis from the bad? Though the thought is all the more horrifying if politicians are involved, it seems as though a regression could claim to "prove" anything if its purveyor is previously "well-regarded" by other economists who have come to opposite conclusions.

The other issue I take with the style of problem-solving presented in the book is that it exclusively represents hindsight. There is no expertise to be offerred for nascent fields or emerging issues that do not have exhaustive data to analyze. Nowhere do the authors offer policy ideas or outlooks because they skeptically wonder if such an attitude is even appropriate. Granted, understanding the past is the key to understanding the future, but economists themselves may be subject to the trappings of the conventional wisdom; they usually exclude the possibility of exogenous influences in their determinations, which may inappropriately rule out the possibility of some unconsidered or unexaminable cause. However, I would still argue that these objections apply to a very small number of issues at best; the analysis clearly works in a number of useful areas. I only wonder at the fallibility of economics, but I suppose that at the moment there is well more than enough public and political skepticism of economics to keep it in line for a while.

Regarding potential freakonomics questions, I too had thought of Chris' gas stations on the corner question. Another possible question to examine would be the effect of the presence of police cars on accident rates (or more generally examining the influence exerted on driving behavior by the mere presence of police cars), though such a question would involve lots of controls on whether or not accidents get reported at all, how serious they were and so on. Perhaps comparing standard accident rates to rates during police balls or something like that would work; I bet Steven Levitt could work it out.

One other possible experiment would be to look at steroid use in professional sports and compare the costs and benefits (in terms of salary) of using steroids and being discovered, and so on. Though it would be nearly impossible to find out what athletes used or are using illegal performance-enhancing drugs, I bet it would have similar results as the crack dealer living with their moms analysis.