June 21, 2016

The Social Cost of Carbon on Trial

The social cost of carbon (SCC)—how much damage a ton of CO2 or other GHGs is doing (and will do in the future) to the planet—is a concept much revered in environmental circles as supporting rational policy making. We’ve written about the series of complex assumptions that go into calculating it, and to be honest, we are not fans of its use. Now, a recent administrative law decision in Minnesota may be the nearest we’ll get to an in-depth public debate about putting the SCC to use in policy. The process—and the near-certain appeal—provides a useful airing and summary of the arguments that deserves wider attention.

More than 20 years ago, Minnesota enacted a law requiring its Public Utilities Commission to “quantify and establish a range of environmental costs associated with each method of electricity generation.” Utilities were then required to use these cost figures in their resource planning. As a result (and after much dispute), in 1997, the PUC set the “environmental cost” of CO2 at between 30 cents and $3.10/ton.

In 2013, environmental groups petitioned the PUC to update the CO2 value and to set it at the federal SCC value; between $13 and $137 per ton (as of 2020, and expressed in 2011$). The PUC set up a stakeholder process to see if a consensus would emerge on the appropriate value but—no surprise—that failed. As a result, the PUC sent the matter to an administrative law judge (ALJ) for a “contested case” procedure.

Before we get to the ALJ’s decision, we must explain why the process used in Minnesota is so unusual. When EPA uses the SCC as part of a Clean Air Act rulemaking, those regulations can only be challenged in the appropriate federal appellate court. If someone were to dispute the SCC value in that process, the court would not decide what the appropriate value of a ton of CO2 is, but only whether EPA’s use of the SCC value was “arbitrary or capricious.” In other words, showing that the SCC value is “wrong” is not enough. The challenge has to prove it to be so wildly wrong as to essentially make no sense whatsoever. Given the normal level of deference to an agency’s judgment as to factual matters within its area of expertise, we think it would be very difficult to overturn EPA’s use of the Social Cost of Carbon.[1]

In contrast, a contested case hearing is the equivalent of a bench trial—briefing, witnesses, oral argument, etc. Moreover, instead of the uphill battle to prove that the SCC is “arbitrary and capricious,” it is the environmental petitioners’ burden to prove that the SCC (referred to in Minnesota as the Federal Social Cost of Carbon or FSCC) “is reasonable and the best available measure of the environmental cost of CO2”.  In other words, the PUC put the Social Cost of Carbon itself on trial.

And so, after a deluge of paper and testimony from 19 expert witnesses (including seven testifying on behalf of Peabody Energy), on April 15, the ALJ issued her 150-page decision. Therein you will find a lot of discussion about discount rates, “leakage”, using a U.S. SCC v. a global SCC, average ton of CO2 v. marginal ton, “equilibrium climate sensitivity”, and more. All of these are significant and real sources of controversy in the calculation of, and limitations on the appropriate use of, the SCC.

We urge those who throw the numbers around (which is almost everyone in the climate debate) to read the decision in full. While the SCC is clearly not zero, the complexity and assumptions underneath the current calculation process make the FSCC of questionable usefulness in the practical policy debate.

The decision itself demonstrates the cognitive dissonance necessary to uphold the FSCC for regulatory policymaking.

The ALJ first concluded that “the FSCC underestimates the negative effects that increased warming will have on human health” and that the FSCC models “do not account for a significant number of important environmental impacts which will occur as a result of climate change.” Accordingly, “the FSCC understates the full environmental cost of CO2”.

On the other hand, and somewhat contradictorily, the ALJ then identified one of the crucial weaknesses in the FSCC, which is that it is based on climate effects projected out to the year 2300:

[Petitioners] failed to demonstrate by a preponderance of the evidence that a modeling time horizon extending to the year 2300 is reasonable. An additional two-hundred years will add increased numbers of cost values at lower interest rates and accelerating rates of damages with the passage of time and increased temperature. Therefore, the Administrative Law Judge finds that an extrapolation extending two-hundred years beyond the year that the EMF-22 scenarios were constructed to end is a degree of uncertainty that is not reasonably supported by adequate evidence.

Confronted with an FSCC that underestimates the health and environmental impacts, but extends 200 years beyond what is reasonable to model (the impacts in 2100), the ALJ indulged in a classic case of splitting the baby:

However, weighing the importance of accounting for the CO2 that will remain in the atmosphere beyond the year 2100, and the understated nature of the FSCC, the Administrative Law Judge concludes that it is reasonable to implement the [Federal Interagency Working Group’s] extrapolation for 100 years, to the year 2200. While the evidentiary underpinning is no greater for this extrapolation than it would be to extend the model to the year 2300, this approach lessens the danger of multiplication of errors within the extrapolation while providing a response to the strong evidence of damage from CO2.

In other words, the ALJ decided that, on the one hand, the three models used to predict future climate impacts failed to include certain damages, but on the other hand, there was no basis for predicting damages two centuries beyond 2100.  And so she compromised by using the “understated” damages, but only for the next one hundred years.

There were also interesting debates about the validity of using global damages in the Minnesota SCC cost-benefit calculation. The ALJ elected to accept the global value, which makes some sense for the Federal SCC because U.S. emissions are a significant part of the global total. It makes less sense, however, to require an individual state to consider possible global benefits in its policy making. It also seems like a rather open ended precedent for other environmental externalities (e.g., deforestation, acid rain, GMOs). The ALJ accepted the global number despite the contradiction in the FSCC advocates’ position: that Minnesota should use the global number as the measure of damage for each ton emitted in Minnesota and should also count emissions reductions from pricing at the FSCC as an accruing climate benefit (even though they acknowledged emissions might well be shifted to locations with a lower, or zero, carbon price).

And thus, subject to the reduced timeframe and another (minor) modification, the ALJ “recommends that the Commission adopt the Federal Social Cost of Carbon as reasonable and the best available measure to determine the environmental cost of CO2.”  Assuming that the PUC does so, doubtless the Minnesota Chamber of Commerce, Minnesota utilities, etc. (and Peabody, if that’s OK with the Bankruptcy Court) will appeal through the Minnesota courts.

We do not see this as a very rational outcome, but until someone comes along with better models and better assumptions, this is the sort of result we can expect. In the meantime, the FSCC has broadly withstood a serious challenge, and is now one step closer to being the default value.

[1] The federal SCC will finally be put to this test in the challenge to EPA’s methane standards for new oil and gas production, processing, transmission and storage facilities, which relies exclusively on greenhouse gas reduction benefits it its cost-benefit analysis. EPA’s rule uses the “social cost of methane”, as opposed to the social cost of carbon, but the disputed issues are essentially the same.



[Image Source: MPR News]