Traffic deaths declined 20% nationwide during March and April, 2008 versus the same period a year ago as a result of high gasoline prices according to a study by Michael Sivak of the University of Michigan's Transportation Research Institute.  It pays to be cautious is jumping to conclusions based upon a few months of evidence, but having logged a lot of time on Washington's Beltway recently, I can assure you there's less traffic, and most of that traffic is travelling at slower speed.  The Washington Post interviewed area transportation experts who still see lots of traffic fatalities, particularly on motorbikes, but Sivak predicts we will drop below 40,000 fatalities nationwide in 2008 for the first time since 1961!

I had always thought short-run gasoline demand elasticities were quite low already, but this paper by Jonathan E. Hughes, Christopher R. Knittel, and Daniel Sperling of the University of California, Davis finds they have dropped even more in 2001-2006 as compared to those of the 1970s and 1980s.  This supports another of my recurring themes:  Now that we have hard won changes in driving behavior in hand, let's hang onto them by preventing gasoline prices from declining too much.  How much is too much?  I don't know, but pick a number, say $3.50 gallon and impose a variable tax to keep it there.  That could raise money for research and investment in alternative fuels, and it would foreclose a return to higher speeds, larger vehicles, and more traffic deaths on our highways.