A commenter writes:

With a carbon tax, there would be a number of second order rebound effects(People buy more fuel efficient cars, and then proceed to drive more) that could actually increase emissions in the long term.

Because of this, the effects of a carbon tax on carbon production are uncertain, and the tax would have to be adjusted frequently if we were to meet any sort of international targets. Megan herself frequently posts about the wisdom of frequently changing taxes.

With a Carbon credit scheme, we not only make international carbon trading and hedging strategies far simpler, but we also have complete predictability about the effects.

1) There are no rebound effects with a tax. Rebound effects come from fuel efficiency standards. Those standards raise the price of the car, while lowering the price per mile driven. The result is that people drive more, which claws back some of the gains from fuel efficiency--estimates range from about 10% to 30%. Meanwhile, the added expense of new cars keeps older, less efficient cars on the road longer. If I recall correctly, it took about a decade for cars built under CAFE to reach half the national fleet.

2) True, but as things go, ratcheting up excise tax rates on a commodity of which we are trying to discourage consumption is pretty anodyne.

3) We don't have complete predictability about the effects of international carbon credits. If we did, they'd be a fabulous idea. As things stand, it is very unclear that they do more good than net environmental harm.

The benefit of cap and trade is that you know where you're going. The problem with cap and trade is that it is more vulnerable to gaming, and there are threshold effects. The problem with a tax is that you don't know the price you need to get to the goal you want. But the benefit is that it starts working from dollar one, and it's harder to evade by, say, purchasing dodgy offsets.

Furthermore, the commenter assumes that we actually know the optimal level of carbon. I don't think we do.