In my challenge to read anything this year, my Father's Day gift of Freakonomics by Steven Levitt and Stephen Dubner was so good I could barely put it down. For someone who enjoys numbers and math, the book doesn't give me enough of the details, but the results and the convention-bashing are great fun. I couldn't help wanting more: data, questions answered, and toys to do the analyses myself. Do the trends that hold in the United States also hold in other countries?
The authors claim no unifying theme to their discussions, other than it is fun to get the facts (numbers) and look for interesting correlations, as opposed to believing what we are told. In the case of this book, however, I think the unifying theme is that knowledge is power. The knowledge that information assymetry is important in transactions (financial and social). The knowledge that with the right data and questions, one can test conventional wisdom. One can even see how people respond to incentives, assuming you have the data and ask the right questions. Another theme is that you also have to be diligent or lucky in some cases to get the data to do these kinds of socio-economic analyses.
The chapter headings are enough to either force you to sit and read, or to throw the book in disgust: What do schoolteachers and sumo wrestlers have in common? Some of them cheat, possible due to the incentive systems. How is the Ku Klux Klan like a group of real-estate agents? They use the power of information assymetry to their advantage. Why do drug dealers still live with their moms? Because, just like in corporations, the biggest money makers sit at the top of the pyramid. Where have all the criminals gone? Could it really be that fewer of the most likely criminal children have been born since Roe vs. Wade? What makes a perfect parent? It's what the parents are before they ever have the child, not what they do once the child is born.
And then there is that chapter on baby names and the black-white gap in names and the interesting phenomenon that names used by families on the high end of the socioeconomic scale tend to move down that scale over time. It would be fun to play with this data or display it graphically, maybe with a tool like the baby NameVoyager (the author of which happens to disagree with some of the findings in Freakonomics).
Levitt and Dubner note that name popularity has changed. The most popular names represent fewer children overall, or there is a proliferation of names. Parents are tending to look for less common names for their children - that was certainly one of our criteria in naming our child last year. There are some numbers one can play with at the Social Security Administration's Popular Baby Names online database. Looking at the SSA database of baby names, you find that "name proliferation" held pretty steady until the 1960's, where it jumped significantly. The top 20 baby boy names represented ~42 percent of all males in 1894-1964. It has steadily dropped to 19% in 2004. For girls that same number has gone from ~32 percent of all females from 1894-1954 down to 14% of girls in 2004. (Click the thumbnail for the full image.)
The authors have a blog where they discuss some of the finer points of the research and point to others who are using Freakonomics as inspiration for their own analyses.