June 14, 2015

How High Would a Carbon Tax Need to Be to Hit CPP Emission Reductions?

Many of those who oppose the EPA’s Clean Power Plan (CPP) regulations offer no realistic alternative to replace them (nor have offered any plausible way to stop them from being enacted).

First there is the “just say no” crowd, which advocates that states refuse to produce the operative State Implementation Plans (SIPs); we suspect there will be few enthusiasts for this approach, since it simply hands control of their state electricity sector over to the EPA. And there will be even fewer when they see the soon-to-be published “model” Federal Implementation Plan (FIP) that EPA will impose (unless the courts stay the CPP pending the Supreme Court’s final decision on its fate in 2019 or so) on such states. Then there are those who believe in an alternate reality where Congress not only votes to repeal the CPP, but that President Obama would not veto a repudiation of one of his signature legacies. More realistic are those who predict a Republican administration in 2017 that will repeal the CPP and then take its time crafting a much more lenient replacement, but even that has at best a 50-50 chance.

So, presuming that states are stuck with having to deal with the CPP’s regulatory morass, what would it take to replace the CPP with a carbon tax? How much would such a tax—applied only to power plants—have to be to produce the same emission reductions as EPA proposes, and thereby be a viable alternative?

Fortunately, EPA has done the work for us, and even better, it is shown in the map below, courtesy of Stanford’s Michael Wara. The data (technically the “CO2 Constraint shadow price”) is carefully buried in a model output file on EPA’s website. The “shadow price” is technically defined as the allowance price a cap and trade system would require for a cap equaling the emissions the CPP requires in each year for the rate-based (as opposed to mass-based) option. Or, more simply, the price EPA believes would generate the same results as the CPP for that state.

CPP Tax Equivalents


As you can see, the answer varies widely by states. It ranges from zero in Vermont—where the CPP sets no target because there is no coal fired power generation—to a high of $101/ton CO2 in West Virginia, with Utah ($63) and Colorado, Massachusetts, and Connecticut (all $47) next highest.

EPA’s figures do not tell us what a single, national carbon tax rate on these plants would have to be to achieve these same reductions, as that would depend on whether lower reductions from states where the price is higher than the average are offset by higher reductions from states where the price is lower than average. That said, after weighting for emissions, we can say that the national average of these state figures is about $29/ ton in 2030, and (not shown on this chart) would start at about $27 in 2020. All of these amounts are in 2011$. That is probably a good approximation of the kind of national rate that would be required to meet EPA’s goals … were the tax applied just to the electricity sector.

Two things leap out from these state numbers.

First, EPA’s numbers suggest that a state carbon tax would be a compelling compliance option. Assuming you believe they are reasonably accurate (and after all, a SIP only has to convince EPA), the 2030 tax required to comply would be $2/ton in New Mexico, $3 in Washington, $6 in Illinois, $7 in Montana, $11 in Ohio. Such a tax would:

  • Generate new state government resources
  • Allow all compliance costs to remain in the state (rather than as under CPP, going to equipment manufacturers out of state, or even abroad)
  • Allow a managed phase down of coal usage—and resources to address social impacts on coal communities
  • Allow reliability concerns to be addressed in a locally managed way, rather than needing waivers or policy exceptions.

Understaffed state environmental departments have no obvious way of avoiding the nightmare task of producing what will be huge and politically controversial SIP document; but at least in states where EPA says a relatively small carbon tax would do the trick, they have a far simpler compliance option worthy of serious consideration.

Second, EPA’s numbers allow us to see what will happen under the option most favored by CPP advocates; regional cap and trade systems. For the nine North-East RGGI states, the 2030 average is also just over $27/ton, which is approximately five times the CO2 price ($5.50) in the latest RGGI auction. And because the existing RGGI program is assumed to continue in EPA’s base case, the actual average across the RGGI states is at least $32/ton.

The fact that RGGI states would be paying six times what they are paying now may be uncomfortable news for the RGGI enthusiasts, but it is not actually all that surprising. Far more importantly, the increased RGGI average cost is held down by the very low compliance costs in New York, Delaware, Rhode Island, and Vermont (which has no obligations under the CPP but still must buy allowances under RGGI rules for other CO2 sources). Since RGGI has one allowance price, those states will end up subsidizing the other RGGI states, especially Massachusetts, Connecticut, and New Hampshire.

Thus, ironically, instead of RGGI being a low-cost means of complying with the CPP, the CPP may spell the end of RGGI, if ratepayers and lawmakers in Delaware, New York, Rhode Island, and Vermont decide not to hand significant amounts of voters’ cash to their neighbors. We believe that this rather critical issue is also worthy of serious consideration . . . at least in those four states.