The fine art of putting words into other people's mouths, as practiced by Brad DeLong: "I think that if you want to buy Seatbelt Sam's argument and call mandatory seatbelt laws examples of harmful and counterproductive government red tape . . ."

Note DeLong doesn't use quote marks and he doesn't say Sam Peltzman actually said or implied that seatbelt laws were "harmful" and "counterproductive." He just says that if "you" buy his argument, "you" might therefore decide that the laws were harmful and counterproductive. Pretty clever, no?

(We can be sure that Professor Peltzman didn't use those words in his article, "The Effects of Automobile Safety Regulation," Journal of Political Economy, August 1975, thanks to a machine-readable version and Control-F. I submit that a fair reading of the article suggests that he didn't imply them, either.)

DeLong also accuses Peltzman of doing his econometrics in bad faith. This seems harsh given that many other studies have found evidence consistent with safety devices causing compensating behavior, now known as the Peltzman Effect. Even the most recent research I could (quickly) find on this, "The Effects of Mandatory Seat Belt Laws on Driving Behavior and Traffic Fatalities," Alma Cohen and Liran Einav, June 2002 (and forthcoming in Review of Economics and Statistics) finds that seat-belt usage raises non-occupant fatalities in a simple OLS regression. It isn't until they either 1) include state and year fixed effects, or 2) instrument for the adoption of seat-belt laws that they get different results. People reasoning in good faith could disagree over the merits of different specifications.

But since Peltzman is quoted in a textbook written by Gregory Mankiw and since Professor Mankiw is now working for the Bush Administation, might "you" think DeLong was just, once again, trying to slam almost everything connected with that administration?

Comments

[Brad DeLong]

Don't be an idiot. I didn't say that Sam Peltzman did his econometrics in bad faith. I said that his results seemed to me to be "sensitive--the kinds of results you get when you are trying too hard and have argued yourself into believing the right specification is the one that is friendliest to your prior beliefs."

Brad DeLong

[Patrick R. Sullivan]

What, he doesn't follow his politeness policy on other people's blogs either?

[King Banaian]

Dear Brad, when you show me an article you've written where you've done this kind of robustness test (think Leamer), I'll be happy to accept your (revised) critique. Otherwise, that's a stock criticism that can be leveled at the 98% of the profession that didn't read Leamer. I confess to being one of the 98%.

[spencer]

Cohen and Eivan found that their research did not support the Peltzman effect.

Personally, I think what Peltzman found was the impact of the baby bloomers starting to drive,not the impact of seat belt regulation. In the 1960s there was a massive increase in the number and share and of share of teenage drivers because of the baby bloomers entering the 15-24 age bracket that Peltzman did not properly account for. This is the reason he found that risk taking increase, not because of seat belt regulations.