Steven Levitt: Are children's carseats necessary?

Steven Levitt shares data that shows car seats are no more effective than seatbelts in protecting kids from dying in cars. However, during the Q&A, he makes one crucial caveat.

More like this

This guy's approach, where he weaves the story of the "two cures" without telling you what he's actually taking about -- to build the suspense or whatever the hell he thinks he's doing -- is annoying as hell, to the point that if I were in the audience, I would have thrown a soda at him by about 45 seconds... What an asshole.

By Woody Tanaka (not verified) on 02 Jul 2008 #permalink

I thought the cure thing was a brilliant idea executed here for the very first time and done poorly. Yes, annoying as hell. The TED audience was getting a first draft. Maybe the guy has a future in blogging.

But I went ahead an put this up because I think his point is interesting. (This does not mean I agree with him.)

the point is interesting, but the idea that he is trying to "resolve" the date into his hypothesis seems a bit off. to me that smacks of refusing to let go of a concept, despite the evidence against it, or at least of an extremely narrow view... I think the TED producers made the interesting decision to include just that one question at the end. I've watched a lot of their videos, and that is pretty unusual.

btw, in the story of the father and the placebo pills, he left out a third possible reason for fewer patients coming back, that they were actually sick and had died in the interim. his lack of interest in the serious injury outcome and the father's apparent lack of interest in follow-up care seem of a piece...

I'd have to look more into the matter to form an opinion on his conclusion, but speaking to the talk itself:

The idea was a good one, it was definitely the delivery that was lacking.