If reducing emissions really has to involve moving on many fronts, anything that looks like an administrative solution — telling, say, power companies what to do or not to do — is going to be much more costly than carbon pricing that exploits all the possibilities. But if a large part of the solution is going to involve a fairly limited set of measures — such as putting a quick end to the practice of burning coal to generate electricity — getting to broad-based carbon pricing is much less central.
And what I gather from reading various analyses of our prospects is that we’re closer to case #2 than to case #1: the problem of limiting climate change isn’t all that complex. End coal-burning and you’ve gone a significant way; a few other big things get you another substantial part of the way. Yes, comprehensive carbon pricing would be best, but it’s not the sine qua non of effective action.
Most of his points concern the status of Econ 101.
Addendum: Ashok Rao comments.