Saturday, December 09, 2006

Pigovian Questions

Ted Gayer, a member of the Pigou Club, emails me some good questions to put on the agenda for our next meeting (if we ever hold one):
Does the Club support Pigouvian taxes irrespective of the use of the tax revenues. Ideally, the revenues would be used to offset distortionary taxes (such as on labor or saving). But what about other uses, such as debt reduction? More problematic is if the revenue is used strictly to increase government spending. There's clearly a time-inconsistency problem here, but the Club may want to advocate an explicit linking of the Pigouvian tax to offsetting tax reductions.

On a related point, how strongly does the Club oppose a policy that re-distributes the revenue to the regulated entities. This is a bigger issue with respect to cap-and-trade programs (a form of Pigouvian taxes), in that the common practice thus far has been to allocate the permits to the regulated entities for free, rather than to auction them (as most economists prefer).
Thanks, Ted. Here is my view on these issues:

The Pigou Club wants to move beyond the rhetorical syllogism, all too common in Republican circles, that

1. Taxes are bad.
2. Pigovian taxes are taxes.
3. Pigovian taxes are bad.

Such a simplistic mindset makes it impossible for people to discuss in a responsible way the relative merits of different tax systems. Instead, we Pigovians acknowledge:

1. There will be some government spending.
2. This spending will be funded with taxes.
3. Government should use the least bad taxes it has available.

In fact, Pigovian taxes are not only least bad--they are good. They correct market failures when transactions costs are too high to expect the forces of the Coase theorem to fix the problem. Pigovian taxes allow truly distortionary taxes, either now or (through debt reduction) in the future, to be lower than they would otherwise be. And in the off-chance that we achieve libertarian utopia and reduce government spending below the level of revenue raised by optimal Pigovian taxation, the extra revenue can always be rebated lump-sum to the public.

I am less fond of cap-and-trade programs than Pigovian taxes because they, in essence, give the revenue from a Pigovian tax lump-sum to a regulated entity. Why should an electric utility, for example, be given a valuable resource simply because it has for years polluted the environment? That does not strike me as equitable. A new firm entering the market should not have to pay for something that an incumbent gets for free. And the fact that the incumbent has for years been taking a valuable resource from the rest of society is no reason to think it deserves a free ride in the future. On equity grounds, one could just as easily argue that the incumbents should compensate society for their past misdeeds.

Cap-and-trade systems are also relatively inefficient, for two reasons. First, they encourage utilities to pollute more before the cap-and-trade system is put into effect in order to "earn" pollution rights. Second, they waste the opportunity to use the Pigovian tax revenue to reduce distortionary taxes on labor and capital. Of course, cap-and-trade systems are better than heavy-handed regulatory systems. But they are not as desirable, in my view, as Pigovian taxes coupled with reductions in other taxes. One exception: If the pollution rights are auctioned off rather than handed out, then cap-and-trade systems are almost identical to Pigovian taxes, including all the desirable efficiency properties.

Pigovians have no magic bullet to keep down government spending. Like many others, I believe that government spending is too high. But Pigovians need not be united about this. The key thing that unites us is the belief that whatever government spending is done, the tax revenue to pay for that spending should be raised in a way that does the least harm or, better yet, the most good.