Are high prices as good as a Pigouvian Tax?

Free Exchange takes Felix Salmon to task for writing this:

Greg Mankiw must be happy: Oil just hit a new high of $81.24 a barrel this morning. This is a Pigovian tax with the proceeds going to Saudi Arabia rather than the US Treasury, but if Mankiw is right that a carbon tax would reduce carbon emissions, then these high oil prices should be instrumental in reducing oil consumption, carbon emissions, and, ultimately, the pace of global warming.

On the other hand, if demand for energy does not fall appreciably as a result of these stratospheric prices, then the whole basis of a carbon tax is disproved, and Mankiw will have to throw his weight behind a cap-and-trade system (with auctioned rather than allocated emissions rights).

As Free Exchange writes, this is not quite correct for several reasons. First, Mankiw advocates a carbon tax, which would cut down the use of things like coal that are much worse for the environment than oil. Second, a Pigouvian tax is meant to price into the product the negative externalities of its use. It doesn't matter how high the price of oil is; a carbon tax would make it even higher, and thus have more effect. Moreover, these prices may not be high enough to produce an efficient reduction. And finally, oil demand is fairly inelastic over the short term, but much less so over the long run, as people invest in conserving technologies and rearrange their commutes and other energy-intensive activities. The long reactive lag is what lulled OPEC into a false sense of security in the early eighties, with disastrous results for their governments when the price finally plummeted. It will be some time before we can assess the full effect of these prices on consumption, if indeed they last.

But there are other reasons that I think this doesn't quite work, even though I toyed with this notion a few days ago. First of all, if the price increase is reflecting the fact that oil production has basically peaked, as some observers have argued, then in essence we have a cap and trade system on oil. Assuming, as I think is reasonable, that India and China will not accede to any sort of global regime, there will be no carbon-reducing benefits to implementing the kind of regime that Mr Salmon favors. Rather the reverse, in fact; India and China use oil much less carbon-efficiently than industrialized countries do.

But a Pigouvian tax could still be useful. As an energy correspondent of my acquaintance was fond of saying, carbon taxes make sense not just environmentally, but because they redirect money from some of the worst regimes in the world. A tax takes money not just from American consumers, but also from Hugo Chavez and assorted nasty Middle Eastern regimes.

I know what you're thinking--a chart! Why doesn't she give us a chart?

Oh, all right, my little chickadees, but it's the last time . . .



It's a simple chart, with the price on one side, and the quantity on the other. Like always, demand slopes down, with people wanting to consume less as the price rises, while suppliers want to sell more.

At time zero, the market clears at P₀, which purchases S₀ units. Then we impose a tax, which raises the price to P_Tax, and reduces the quantity consumed to S_Tax. The oil producer, however, which used to earn [P₀ X S₀] now gets the same price for a smaller quantity sold, earning only [P₀ X S_Tax]. If the taxes collected are rebated as reductions in other taxes (admittedly a big if), American consumers are not much worse off--not at all worse off, in a larger sense, as long as the tax has correctly priced the negative externality (admittedly, another big if). But Iran suddenly has a lot less money to spend on nuclear weapons.