From Mike Moffat at econonomics.about.com:
John is a friend, but I think I agree more with Prof. Mankiw on this one. Economists need to realize that we are never going to get a textbook perfect cap-and-trade system vs. a textbook perfect carbon tax system. Becuase of that there is not a great deal of point arguing the merits between the two, beyond as a blackboard exercise. What we will get is a cap-and-trade or a carbon tax system which is flawed thanks to the compromises, sausage making and horse-trading that go on when any political decision is made. So we will either get a carbon tax system that has hundreds of little exemptions, or a cap-and-trade system that looks an awful lot like Waxman-Markey.
Now to be fair to Prof. Whitehead, the Waxman-Markey act is likely to have a great deal more flaws than, say, the SO2 trading system in the Clean Air Act (1990). So while any real-world cap-and-trade system is likely to be imperfect, it does not necessarily have to be this imperfect. Thus he has a point by suggesting that Waxman-Markey does not invalidate the benefits of cap-and-trade.
I agree with all that. And Mankiw suggests that the problem is cap-and-trade. Jeez.