WASHINGTON — IER Senior Economist Robert P. Murphy will testify today at 10:00 AM ET before the Senate Environment and Public Works Committee. Murphy will discuss the “social cost of carbon” (SCC). Murphy’s testimony will show that the SCC is a malleable and arbitrary concept dependent on modeling assumptions that can have an enormous impact on the cost-benefit analyses of federal regulations. The text of his introductory remarks, as prepared for delivery, follows:
The “social cost of carbon” is a concept developed in the academic literature of the economics of climate change. It has been used to help justify over 35 regulations, where sometimes over 20 percent of the alleged benefits of these regulations are derived from the social cost of carbon.
As I will explain, the administration’s calculation of the social cost of carbon is malleable and arbitrary and therefore not appropriate for the federal government to use to justify regulation. A large fraction of the alleged benefits from reducing carbon dioxide emissions are incredibly speculative—as they occur in 50, 100, or even 250 years in the future. As I will explain, the estimated size of the social cost of carbon is heavily dependent on the discount rate used in the analysis, and the administration has ignored OMB’s guidance on the subject. In fact, the concept is so open-ended that we can generate social costs of carbon that are very high, or close to zero, or even negative, just by adjusting some key parameters. This means that the economist can produce just about any estimate of the social cost of carbon desired.
In theory, the social cost of carbon quantifies in dollar terms the damages from emitting an additional unit of carbon dioxide, because of the presumed acceleration of future climate change. It has been used to help justify policies, such as imposing stricter fuel economy standards, by giving quantifiable benefits in dollar terms from their impact on emissions.
The social cost of carbon is in the news because in May, the administration’s Working Group, without public comment or notice, dramatically increased its headline estimate of the current social cost of carbon by 50 percent from its previous estimate made in 2010, from $22/ton to $33/ton.
Let me briefly explain how the administration’s Working Group generated such estimates, to see the pitfalls in the procedure. First, they selected three popular computer models of the global economic and climate system, and ran thousands of simulations through the year 2300.
In these computer simulations, under certain scenarios carbon dioxide emitted today can sometimes at first produce net benefits to humanity, because (for example) modest warming can boost agricultural productivity, reduce cold-related deaths in the winter, and lower heating bills. Eventually, the computer models assume that an extra ton of emissions today will start producing net damages. The social cost of carbon is then an estimate of that flow of possible upfront benefits followed by a flow of damages that kick in later this century and continue through the year 2300.
Given the setup I have just described, the discount rate used to take those projected damages centuries in the future and translate them into current terms will have a huge impact. For example, the Working Group’s latest estimate put the current social cost of carbon at a mere $11/ton using a 5 percent discount rate, but it would more than quadruple to $52 at a 2.5 percent discount rate.
This huge range in the estimate has nothing to do with climate science, but is driven entirely by altering the discount rate. In this context, it is very significant that the Working Group explicitly disregarded OMB’s clear guidance that a discount rate of 7 percent should be used to provide one of the estimates. We can’t know for sure without seeing their full data, but simply adjusting this parameter and using a 7 percent discount rate would probably produce a social cost of carbon close to zero—in which case the administration’s rationale for limiting emissions would collapse.
OMB also requires that cost/benefit analysis be reported in terms of domestic impacts, with global impacts being optional. Yet the official reports of the social cost of carbon ignored this clear guideline, and instead used global figures. If we take the headline global figure of $33/ton as reported in the May update, the corresponding domestic figure could be a mere $2/ton, using the Working Group’s own range of adjustment factors.
In summary, the “social cost of carbon” is not an objective feature of the world that is “out there” waiting for economists to measure. Rather, it is generated within computer simulations that make projections centuries into the future. By tinkering with the discount rate and other parameters, the White House can justify about any estimate the administration wants. Even more troubling, the Working Group explicitly disregarded two OMB guidelines on cost/benefit analysis, either one of which would drive the official estimate toward $0/ton. Clearly, federal policy should not be formed on the basis of such a dubious concept.