Monday, October 1, 2012

Murphy on Carbon Taxes

Bob Murphy has a new article (HT - David Henderson) summarizing work on tax interactions associated with carbon taxes, which - the argument goes - make the optimal carbon tax lower than the social cost of carbon (and potentially zero).

The logic is straightforward enough when Bob explains it, but easy to overlook when thinking about these issues. Carbon taxes raise factor costs and if distortionary taxes exist on factors (which they do), this can reduce welfare.

I am fine with the logic, but a little unsure with what direction to take this in (and one first step is to move beyond Bob's excellent article and read the AER article he is discussing).

My first thought is to push back a little on one potential interpretation of the finding (I don't say "Bob's interpetation" because for all I know he'd find this point reasonable). Does this really suggest that carbon taxes should be lower, or does it suggest that the factor taxes are more damaging if all social cost and benefits are accounted for? It seems really weird to think about this as a problem with the carbon tax, since that's not assumed to be distortionary at all - it's the factor taxes that are influenced by the price of carbon that are distortionary.

This leads me to my second thought. Since it's the interaction with factor taxes that's the problem, is this just another good reason to shift to consumption taxes? So carbon taxes shouldn't just come with a reduction in factor taxes, they should also come with an increase in consumption taxes?

I need to think about this a little more to answer another question that comes to mind: does this tell us that under a textbook carbon tax, the price of carbon will be set too high (so that carbon is utilized below its optimal level of utilization), or does it just tell us that total welfare is lost? Because if we're thinking about climate change specifically, those are two quite different things.

More questions than answers - I know. But there are plenty of very interesting answers to other questions in Bob's post. So go read it!


  1. Thanks. FWIW, I actually didn't believe the result when I first saw it--and I'm of course a foe of carbon taxes. It took me a while to even understand what they're saying.

    Here's the surprising thing: The prior existence of distortionary taxes can actually REDUCE the case for imposing a revenue-neutral carbon tax (of a given magnitude). That is an incredibly counterinuitive claim--you would think it would be the opposite, necessariiy--and took me a few days to believe.

  2. This was one of the interesting results we went over in my environmental section. And while in the case of the carbon tax I agree with Murphy, by the same token, you could use the very same argument for a larger tax (on X) if the markets that it affects are overwhelmingly subsidized. Living in a "second-best" world can go against or for unaffected market prices.

    More than anything, my skepticism of something like a carbon tax comes from the unimaginably complex general equilibrium effects (both in the sense above and w.r.t other externalities), our lack of knowledge regarding how far/near the current tax/subsidy nexus is to an equilibrium arrangement, the complete ambiguity of the Damage Function, the ineptness and inflexibility of policy that has been demonstrated in the past (if it were anything near rational we'd have gone nuclear by now), and the increased mobility of factors (in the global sense).

    For all of these reasons, I favor less than more action. And yes, as others have recently pointed out elsewhere, and it should be obvious, there is a distinction between "doing nothing" and "doing something" in the face of ignorance. The principal is applied in computer science all of the time - if you're not entirely sure what's the best, use an evolutionary algorithm to get an idea....


All anonymous comments will be deleted. Consistent pseudonyms are fine.