OK. I've finally figured out what is going on.
Let's draw a graph, with the tax rate on labor (and capital, and other things) on the vertical axis and the tax rate on carbon on the horizontal axis, like so:
We would like, ideally, if there was nothing we wanted the government to do, to have a tax of zero on labor and a positive tax on carbon emissions to correct for the global-warming environmental externality--we would like to be at point O, if we could.
But we can't be at point O. There are very valuable things that we need the government to do, and the government needs to spend money to do them. We need to impose taxes on labor and carbon in order to raise revenue. Thus the "required revenue line," running through points S and A. Our tax rates on labor and carbon must be such as to keep the economy on or above the required revenue line.
At the moment we are at point S--the status quo. We would like to impose a carbon tax and use the revenue to lower our labor tax and thus get to point A--that is the social optimum, the point that (a) gets us as close as possible to getting prices right and thus getting allocative economic efficiency, while also (b) raising enough revenue for the government to do the things that the government ought to be doing. To first order, at least, we can get to point A either by (i) imposing the right efficient carbon tax directly, or (ii) having the government sell the right efficient number of carbon-emission permits.
The best thing to so would be to assemble a legislative coalition of:
- environmentalist Democrats
- Republicans interested in economic efficiency
and pass a carbon tax or a cap-and-trade with a permit auction system to get us to point A.
However, at the moment there are next to zero Republican legislators interested in economic efficiency. Instead, Republicans legislators appear to be exclusively interested in trying to make Obama's presidency appear a failure so that they can then benefit in the 2010 and 2012 elections--the Gingrich-Dole strategy, which they pursued with very short-term political but no other kind of success over 1993-1998, by the end of which both had been bounced out of the legislature.
Thus Obama has to go for a second best. He has to assemble a legislative coalition of:L
- environmentalist Democrats
- blue dog Democrats
where the blue dogs insist that we cannot tax but rather must pay the polluters: businesses that currently emit carbon must be given free permits that they can then sell on the open market, and thus profit from. This is the equivalent of (a) imposing a carbon tax, and (b) then giving the revenues raised away to last-year's polluters in lump sums. This gets us to point B. Point B is clearly better than the status quo of point S--prices are closer to being right. But point B is worse than point A.
So the right strategy for Obama right now is (i) pass cap-and-trade with a permit giveaway now, and (ii) work toward some future in which the Republicans interested in economic efficiency wiill cooperate in a bipartisan faction. Hope that they will someday actually be willing to be bipartisan--to give a higher priority to the welfare of the country than to trying to make Obama's presidency a failure. Then whenever that, happens assemble the bipartisan legislative coalition of environmentalist Democrats and public-spirited Republicans and move from point B to point A.
Now comes Greg Mankiw. Greg says: "Wait a minute. A tax on carbon is not just a tax on carbon--not a move parallel to the x-axis from point S to point B. Instead, whenever you tax carbon you automatically raise the tax on labor too--there is no way that the economy can reconfigure itself to emit less carbon and still employ as much labor as efficiently s it now does, and so a carbon tax is also a labor tax, and points up and to the right from point S to point C. Impose a carbon tax," Mankiw argues, "and we move not from S to B but from S to C and wind up further away from the point O that we want to be at."
Hence Mankiw's call for Obama to veto cap-and-trade-with-permit-giveaway.
Is Mankiw right that carbon is emitted and labor employed in a process of production so joint that any carbon tax will raise taxes on labor enough to move after-tax prices along a vector that makes a more than 90 degree angle with the S-O line--which is what you need for a cap-and-trade-with-permit-giveaway to be a bad idea? He doesn't present any evidence that this is the case.
Larry Goulder, Rob Stavins, and Severin Borenstein would certainly know. Let me try to call them...
Reference: Laurence Goulder (1994), "Environmental Taxation and the Double Dividend: A Reader's Guide"
brought to you by Livescribe