In her statement of partial dissent to the Federal Energy Regulatory Commission’s March 14 order reinstating the certificate authorizations for the Southeast Market Pipelines natural gas infrastructure project (SMP Project), FERC Commissioner Cheryl A. LaFleur defended the use of the concept known as the “social cost of carbon” (SCC). Though Commissioner LaFleur wrote that she “believe(s) that the SMP Project is in the public interest after carefully balancing the need for the project and its environmental impacts,” she took umbrage with the Commission’s position that the SCC was ill-suited for guiding regulatory decisions, arguing that the SCC is a scientific computation of the dollar-value of harms from carbon dioxide emissions. LaFleur contended that the Commission’s stance would be akin to someone saying, “I guess it is fine to eat this donut, because there is simply no way to assess if it will make me fat,” comparing the SCC to a donut’s calorie count.
I am not a lawyer and will not comment on specific regulatory decisions. Yet as an economist I will explain why Commissioner LaFleur’s analogy is utterly flawed. Indeed, the Institute for Energy Research has argued for years that the concept of the SCC is dubious, and should not be used for federal cost-benefit calculations. LaFleur wants to liken the Social Cost of Carbon to a calorie count. But to make the analogy work, it would mean that two different nutritionists could make the calories in a donut today add up to either 2 or 20,000, just by adjusting their forecasts about how many donuts our grandkids will eat. This would be nonsense of course; the calories in a donut are an objective fact, and don’t have anything to do with our expectations of the future. Yet calculations of the Social Cost of Carbon do depend on our plans for the future, and they also critically depend on the discount rate we use to compare future to present dollars. Even if we stipulate for the sake of argument the physical science behind greenhouse gases and global temperatures, we can make the SCC very high, moderate, or close to zero, just by adjusting these other parameters.
LaFleur Approves of Science
Let’s first provide the broader context of Commissioner LaFleur’s remarks on the SCC:
…I cannot accept the Commission’s justifications for excluding the Social Cost of Carbon from its consideration of the SMP Project. Today’s order generally finds that the Social Cost of Carbon cannot meaningfully inform our decisions on proposed pipeline projects. Further, the order claims that the Social Cost Carbon is not an appropriate tool for evaluating the significance of downstream GHG emissions. I disagree. That is precisely the use for which the Social Cost of Carbon was developed—it is a scientifically-derived tool to translate tonnage of carbon dioxide or other GHGs to the cost of long-term climate harm. I have drawn the simplistic analogy of human food consumption and diet. It would be convenient for a person to say “I guess it is fine to eat this donut, because there is simply no way to assess if it will make me fat.” But there is such a tool, in the form of calories, which have been scientifically derived to translate the consumption of a specific food item to impact on weight gain. Similarly, we are able to estimate what the long-term consequence of a ton of carbon dioxide emissions is likely to be, by use of the Social Cost of Carbon tool. [Bold added, endnote removed.]
So it is clear that LaFleur imagines the SCC is a scientific unit of measurement, just like calories.
Yet it is precisely against such thinking that I warned in my 2013 testimony before the Senate on the concept. At the time I explained:
Because of the significant impact it could have on energy prices and other economic conditions, it is crucial that citizens and policymakers alike realize that the SCC is a very malleable figure. It is not analogous to a physical constant such as the charge on an electron or the boiling point of water, with scientists coming up with ever more precise estimates of a feature of nature that is “out there” to be measured. Instead, the estimation of the SCC relies on computer simulations of the economy and climate system for hundreds of years into the future, and furthermore depends on many subjective modeling assumptions. As I will demonstrate, these assumptions can have an enormous impact on the final number, meaning that an analyst can generate just about any SCC he or she wishes by adjusting certain parameters.
Why the SCC Isn’t Like Calories
It’s true that two nutritionists might disagree on how many calories a particular donut has, because there are always measurement errors. But they would both be in the same ballpark. Furthermore, they would both agree on all of the important parameters needed in order to calculate the calories. Specifically, the formal definition of a calorie (in this context) is “the energy needed to raise the temperature of 1 kilogram of water through 1°C,” and our nutritionists would both agree on what we mean by a “kilogram of water” and “1°C,” because these also are straightforward, objective physical units that require no subjective interpretation.
In stark contrast, the Social Cost of Carbon is defined (by the Obama Administration’s Interagency Working Group) as “an estimate of the monetized damages associated with an incremental increase in carbon emissions in a given year.” Now to parse that definition and make it operational, we have to make projections of global economic activity between now and the next 300 years. In order to know how much an additional ton of CO2 emitted in, say, the year 2020 will impact humanity, we need to know the trajectory of climate change as a baseline, upon which that additional ton in 2020 will be added.
Already we see that this is a qualitatively different program. It would be as if you couldn’t count up the calories in a donut, without first making assumptions about how much food a person’s grandchild will eat during his lifetime.
But wait, it gets even worse. So far it sounds like I’m just pointing out that the estimation problem is messier and speculative. But that’s not the true problem. The real difference between counting calories versus counting the “social cost” of more emissions is that in order to convert future climate change damages into present-dollar terms, we need to employ a “discount rate.” In other words, even if economists could agree that an additional ton of CO2 emitted in the year 2020 would mean that humans from (say) 2090 through 2200 would experience an additional $100 of global warming damages in each of those years, the economists still wouldn’t agree on what the Social Cost of Carbon in the year 2020 should be. That’s because they might use different discount rates to turn that stream of far-distant damages into 2020 dollars. An economist who gave more weight to the future (by adopting a low discount rate, like 1 percent) would come up with an enormous figure (about $3,365), while an economist who used the higher of two discount rates as laid out by OMB guidelines (namely 7 percent) would come up with an SCC of about $13.
To repeat, in the above calculation, we are making the same assumptions about the “science” of climate change, and even the human impact measured as damages in a given year. But because the Social Cost of Carbon is driven by enormous damages that might occur decades and even centuries down the road, the seemingly innocuous choice of a discount rate can mean the difference between an existential crisis and a triviality.
Policymakers have been seduced by the scientific allure of the technical-sounding “social cost of carbon.” The fact that so many different agencies collaborated on its estimation, using cutting-edge computer simulations, obscures the fact that its value is driven entirely by (largely arbitrary) non-scientific judgments. Indeed, MIT’s Robert Pindyck excoriated the climate-economic models as “close to useless” for driving policy—and Pindyck is a supporter of a carbon tax. As Pindyck argued in a peer-reviewed article, “[Integrated Assessment Model]-based analyses of climate policy create a perception of knowledge and precision, but that perception is illusory and misleading.”
As LaFleur’s donut analogy reveals, policymakers have fallen for the bluff put on by the economists in this arena. They have convinced the gullible that the “social cost of carbon” is an objective, scientifically rigorous unit by which we can guide regulatory and tax policy. Yet the Social Cost of Carbon is nothing like calories. If it were, an apple fritter would have fewer calories for people who didn’t mind looking fat in their Christmas photos.