Today's New York Times has an interesting story (at least for policy wonks) on changing and differing valuations of the Value of a Statistical Life (VSL) for purposes of regulatory cost-benefit analysis. Vanderbilt's Kip Viscusi, who has been working on this issue longer than anyone, is quoted in the article as supporting the increased valuations, as do I. However, I also agree with John Graham, Dean of the School of Public and Environmental Affairs at IU-Bloomington (and former head of the Office of Information and Regulatory Affairs at OMB under President Bush) that there is little justification for allowing different agencies to use different VSLs. Under current OMB rules, agencies are free to use any value between $1 million and $10 million, with $5 million as the recommended central estimate. That central estimate is clearly too low, and the range is too broad. It would be better if OMB would establish a figure for all Executive Branch agencies to use.
UPDATE: Dan Farber also agrees with the increase in the VSL (see here).