Freakonomics abortion research criticized
The Wall Street Journal reports an economist has found an error in Steven Levitt's abortion/crime analysis as described in Freakonomics, although Levitt claims the error doesn't undermine his conclusion. The most famous study in Freakonomics concludes that the foremost cause of the drastic crime rate drop since the mid-90's was the federal legalization of abortion in the early 70's, postulating that unwanted children are more likely to commit crimes.
<<< Scotty McLellan M.I.A. Secret Scientology Archive >>>

That's too bad, I was a big fan of that logically jump. It was a fun theory, and it remains a fun thory, regardless of how true it is.
There wasn't much weight to that book, but it was a fun read.
I haven't read Freakonomics. Anyone recommend it (besides the Times and Journal)?
Again, if you're looking for something weighty, you aren't going to find it. It's a "fun" economics book. Some of the stories are interesting, but there's no arc and there's no real point. You'll get through it in a day or two.
agreed. I also found the epigraphs very annoying. They are a byproduct of the way this book came about - it's co-authored by levitt, the economist, and dubner, a reporter. They wrote the book after dubner wrote an article about levitt. Every chapter is headed by a quote from the article, which is of course written in the third person and somewhat glowing. Since the book itself is written in the first person, this comes off as incredibly arrogant - it looks like levitt is heaping praise upon himself. Also, there's not much in the way of numerical backup in the book, which is expected, since this would make it unreadable to the lay reader, but always bugs me. To be fair, there is an extensive reference list and apparently most of his data and methodology etc. is nicely collected and publicly available, so I shouldn't complain about that.
PS. I read it on a palm pilot so any numbers or charts that were included were of no freaking use to me. That might have affected my opinion.
None of those can be fairly interpreted as a recommendation.
it's a quick, fun read but it's more about interesting results than how statistics are applied. if it was about the latter, it wouldn't be a quick read (but it might be fun, depending on your predilections).
Post new comment