Abolish the Federal Gasoline Tax!

Earlier this month, Peter Van Doren and I published a study calling for the total elimination of the federal gasoline tax. Well, the first wave of commentary is in and, thus far, we are greatly underwhelmed. Gas tax proponents are going to have to do a lot better than this to hold the intellectual fort.

Over at the Economist, we are accused of misrepresenting citations when we argue that a doubling of the gasoline tax would only reduce tailpipe pollution by about 6 percent over the long run; an accusation also levied by some commenters at Marginal Revolution. The Economist writes:

Consultation of the cited source seems to show not that an increase in efficiency leads to a 20 percent jump in vehicle miles traveled, but that roughly 20 percent of total energy savings from efficiency gains are lost to increased travel. That’s quite a different point, implying that efficiency gains could have a significant impact on emissions.

We happily accept the clarification. The Economist has crisply stated what we in fact meant to say. But clarifying the point does not undercut our argument.

The “rebound effect” discussed above is one reason why aggregate tailpipe emissions will not be reduced as much as gas tax proponent think. But the other – and more important – reason can be found in J. Daniel Khazzoom’s userID=d8b90bfe [at] cato [dot] org/01cce4405b00501c7188b&dpi=3&config=jstor">paper which we cited. To wit, current law regulates tailpipe emissions per mile traveled, not per gallon of fuel consumed. A gasoline tax will induce a consumer shift towards more fuel efficient vehicles, and that shift will lead to only modest reductions in vehicle miles traveled. The upshot is found on page 438 of Khazzoom’s paper: Just as we said, a doubling of the gasoline tax would only reduce tailpipe pollution by about 6 percent over the long run.

Now, a careful critic might point out that this problem could be remedied by regulating tailpipe emissions per unit of fuel consumed rather than by vehicles miles traveled. But we haven’t run into that careful critic as of yet. We do anticipate and acknowledge the point, however, in endnote 32 of our paper.

Another large batch of commenters score us for not conceding that gasoline taxes are an efficient means of addressing greenhouse gas emissions from cars. But it never occurred to us to state the obvious – that if society wants to reduce greenhouse gas emissions, the most efficient means of doing this isn’t with a gasoline tax. It’s with a carbon tax.

Greg Mankiw asks us: “If Congress were considering repeal of the gasoline tax together with an income tax increase to make up the lost revenue, would you favor this revenue-neutral change in the tax mix?” Answer – no. The best way to make up for the revenue loss associated with repeal of the federal gasoline tax would be to eliminate the federal spending associated with the tax. Transportation infrastructure should be a state or local undertaking – not a federal undertaking.

Lurking behind that question, however, is the belief that raising revenue via a gasoline tax imposes less efficiency losses on the economy than raising revenue via an income tax. We don’t think much of that argument, but we discuss it at length in our paper so we won’t go through it again here. Perhaps when Prof. Mankiw gets around to reading our paper, he’ll have something further to say on that score.

There is little else of substance for us to deal with after Round 1.

For instance, many commenters have argued that our paper does not properly take into consideration the underlying literature, which supposedly cuts strongly against our arguments. That literature, we are told, was most recently surveyed by Ian Parry et al. in the June issue of the Journal of Economic Literature.

But this simply tells us that most of the opinions being expressed on these blogs are uninformed by any actual reading. Even a casual look at our paper demonstrates that we review and discuss the same literature discussed in Parry et al and, in fact, we cite Parry’s work extensively throughout. Moreover, Parry et al.’s paper in the JEL makes the same point we make in our study – that a gasoline tax is a deeply problematic means of addressing the externalities associated with driving and that there are far better policy tools available to get the job done. Parry et al. suggest that federal gasoline taxes might be a reasonable “second-best” policy, but we anticipate and counter those arguments in our study, so I won’t go into them here.

Remarkably, no one has yet taken up the most radical challenge offered to the common wisdom in our paper: that even a perfectly efficient gasoline tax would do more harm than good because it would induce more mass transit use, and mass transit use imposes even more costs on society than passenger vehicle use. For this argument, we rely on work done by Mark Delucchi at the Institute for Transportation Studies at the University of California and Cliff Winston at Brookings. Is anyone up to the task?