Neither I personally nor the colloquium have focused as much in the past as we are likely to in the future on the carbon taxes/cap and trade set of issues. (I've never written about these issues because I don't as yet see the angle or intellectual arbitrage opportunities using my skill set.) But yesterday Mitch Kane's paper provided a welcome step in the direction of focusing on them more.
Mitch's paper takes as given that we'd be doing cap and trade not a carbon tax, this being his assigned topic for a forthcoming NYU conference in Abu Dhabi. (Not to mention that current political noises center on cap and trade.) But every time one reads the papers or thinks about the issues it becomes clearer that, even though in principle the two approaches (with suitable ongoing adjustment to each) are interchangeable, in practice it's insane to go the Rube Goldbergesque cap and trade rather than carbon tax route.
My preferred way of thinking about the interchangeability is as follows. Under a carbon tax polluters pay as they go. With cap and trade, they prepay the tax before engaging in the polluting activities, and then use it up by doing the now-permitted carbon-emitting activity. Only, they can sell the taxes-paid voucher to someone else, and, if they want to pay more to emit more, they can't unless the government is willing to sell more prepaid tax vouchers. (Whereas in the carbon tax model you automatically can pay more to pollute more.)
There's a theoretical literature suggesting that the relative merits depend on whether one is more uncertain about the social cost of emissions (which the carbon tax ideally would reflect) or about supply and demand responses (which get reined in by cap and trade if the government keeps the permit supply fixed). But if you keep adjusting the carbon tax to control output levels, or the number of outstanding permits to keep their prices relatively constant, they start to look like each other. In terms of the political frictions if there are adjustment lags, I'd tend to think it makes more sense to take a stab at the social costs and set a carbon tax than to posit something about desired emission levels when the inputs, such as cost of abatement, have such a big effect (which one may not understand) on the optimal reduction in emissions for a given period.
The political reason for cap and trade, of course, is that it isn't called a tax. Only, everyone knows it's a tax (which obviously is what you need to make polluters internalize the social costs of their activities), so one doesn't really gain very much. Plus you get the insane and apparently irresistible political incentive to shower money on polluters by giving them permits for free rather than through auction. This reflects a failure to recognize that we are giving them prepaid tax vouchers without requiring them actually to pay the taxes first. The optics of carbon taxes would be unlikely to yield this result. And, as Alan Auerbach noted at the session, their profits are likely to go up in the post-permits because the reduced output enables them to raise prices. It's as if we organized them into a cartel by charging a tax to reduce output and letting them keep all the revenue. Only mingled corruption and confusion could produce such a result, but evidently they are not in short supply.
Mitch's paper discusses several of the tax issues in handling a cap and trade system. He argues that there is an essentially arbitrary choice between "pre-regulatory" and "post-regulatory" baselines in determining tax basis for permits that were granted for free, but we argued that this is better conceived of as a standard transition issue. (For handy reading on transitions, I suppose one could take a look at this.)
Mitch also focuses on the question of how to minimize inefficiency in the choice between permits and abatement, and offers two alternative approaches that he dubs "no clienteles" and "harmonious clienteles." Alan Auerbach argued, to my view entirely persuasively, that essentially what one needs is income tax neutrality (i.e., permits and alternative abatement methods all are taxed the same for each taxpayer), with marginal rate differences between taxpayers not distorting anything other than via the general work and savings effects of an income tax. This was related to but more demanding than the requirements for satisfying Mitch's "harmonious clienteles" scenario.