Cash for clunkers
Posted by Daniel Hall on January 30, 2009
One of the proposals being floated in the stimulus package involves a so-called “cash for clunkers” program where the federal government would purchase older, more polluting, less fuel efficient vehicles and retire them. RFF Senior Fellow Winston Harrington has done some previous research on these types of programs, which have been employed by various states for many years. One of the upshots of his previous research — here is a (gated) paper from The Review of Eonomics and Statistics and here is an ungated discussion paper — is that these state- and local-level programs are unlikely to produce very much in the way of emissions reductions for local pollutants. I was curious if his results would also apply in the context of a national policy for scrapping gas guzzlers. Winston was kind enough to offer a long and interesting reply, which with his permission I’ve posted in its entirety here.
Scrappage programs for gas guzzlers wouldn’t be as bad as they were in a local emissions context. In the latter case, the programs were reasonably cost effective (at least compared to many of the local alternatives available) as long as you had a short-term program, although the emissions reductions weren’t very large in any event and very difficult to quantify (it would depend on how vehicle replacement affected the fleet). In a continuous program there were some incentive problems that would be hard to overcome:
- All cars of the same vintage had the same emission standards, as did all light trucks (which were slightly higher). So you couldn’t target vehicles based on their new-vehicle standards very efficiently. You could, of course, base acceptance on the program on current emission rate, as determined by an emission test on the used vehicle, but you would risk creating incentives to tamper with the vehicle to increase its emissions, and therefore its eligibility.
- Another way around the targeting problem was to base eligibility on age, and most programs did this. But again, there is a perverse incentive to hold a dirty vehicle that you otherwise might get rid of until it’s old enough to be eligible.
- You had to be careful that the program wasn’t “retiring” vehicles that weren’t being used anyway. There were safeguards against this in most programs (e.g. vehicle had to be registered to be retired)
- Finally, all programs were limited in geographic extent, because they were used to help meet local SIP [DH: state implementation plan] requirements. But then, there was nothing to prevent owners from going outside the area to replace the vehicle with one as dirty or worse. For that matter, if vehicles have a higher scrap value inside a region than outside, it might suck dirty vehicles in from the outside.
I don’t think a gas guzzler program would suffer from these incentive problems, with the possible exception of 3 above. You really can target gas guzzlers. The eligible vehicles would still probably be pretty old, not worth very much and therefore not driven very much. I guess this raises the question of whether owners of old gas guzzlers that aren’t driven very much would replace them with newer gas sippers that they would drive more. You really might not be that much better off. Still, it’s probably better than corn ethanol.
Here is a recent discussion paper from Winston, suggesting a (large) VMT tax for the DC area would be highly welfare improving.