Please read Arnold Kling.
My sympathies in economics lie with the so-called “new institutionalists.” I think institutionalists are going to see more clearly than neoclassicals the rather big difference between a carbon tax and a whole new market institution for trade in government-created and government-rationed permits.
But let's back up a little, to the pre-applied political economy context. It is a fundamental finding of Vernon Smith-style experimental economics that, say, theoretically equivalent auction institutions in practice lead to different results. So you ought to be pretty unimpressed by the fact that there is a possible institution of tradeable carbon permits that is equivalent to some carbon tax on the blackboard. Even were a cap-and-trade permit market or a carbon tax implemented without a hitch according to the blackboard assumptions, there would be little reason to expect them to actually produce equivalent results, since the way actual people interface with an institutions depends on the way actual humans process the rules of the game. Mathematical equivalence doesn't imply cognitive or motivational equivalence. Therefore it does not imply policy equivalence.
Moving back to the applied political economy context, the fact that cap and trade creates a new set of institutions in a way tax doesn't is pretty important, I think. For one thing, there is a question of reversibility. I think Arnold gets it right when he writes:
In practice, cap-and-trade creates an irresistable opportunity for politicians to use carbon permits as political favors to be handed out to special interests. This in turn means that there will be special interests with a large stake in keeping cap-and-trade policies in place, regardless of what transpires in terms of global temperature.
What is the probability that the United Nations derivation of climate model consensus overstates the problem of man-made global warming? Unless that probability is zero, it seems to me that we should prefer a climate change policy that is reversible to one that it irreversible.
Because I think that the probability that the UN is misleading us is significantly greater than zero, I think that the issue of irreversibility is quite important. Therefore, I think that a carbon tax is far preferable to cap-and-trade. Ten years from now the global warming scare might be completely debunked, and yet we will still be unable to unwind cap-and-trade. As with farm subsidies and many other policies, the problem will be long gone but the solution will be with us in perpetuity.
I think the probability that global warming is a significant problem is low. I think the probability that carbon externality pricing policies will meaningfully reduce global warming is even lower. If we are going to have global warming policies at all, I want them to be minimally damaging, which means maximally reversible. Like Arnold, I think it would be a lot easier to simply keep cutting a carbon tax until it hits zero than it would be to dismantle the institutional framework of a carbon permit trading system and all the subsidiary industry that grows up around it. Better yet would be to do nothing. If global warming does turn out to be a problem in a few decades, then unleash the carbon sequestering trees, or what have you. The idea that doing something now is better than doing nothing has, as far as I can tell, a surpassingly thin empirical rationale.