While I was testifying before the Senate Environmental Public Works Committee on the social cost of carbon last month, Howard Shelanski was testifying to the House Committee on Oversight and Government Reform on the exact same topic. As Shelanski was serving in his new role as Administrator of the Office of Information and Regulatory Affairs (part of the Office of Management on Budget), his perspective naturally differed greatly from mine. In this post, I’ll walk through some of Shelanski’s key points, and explain how they unwittingly reveal just how dubious the “social cost of carbon” is as a guide to federal rulemaking.
Sins of Omission
The most striking aspect of Shelanski’s testimony is that he completely ignored the fact that the Obama Administration’s Working Group on the social cost of carbon (SCC) disregarded two separate guidelines from the Office of Management and Budget when it comes to cost/benefit analyses. Specifically, OMB gave federal agencies clear guidelines that they should (1) always include at least one estimate computed with a 7 percent discount rate, and (2) always conduct the analysis from a domestic perspective, with a global perspective being optional. (I thoroughly explain these issues in my written testimony.)
In estimating the SCC, the Working Group ignored these guidelines, with the net result being a much higher figure. In fact, had the Working Group heeded both rules, the “social cost of carbon” would probably be a social benefit of carbon, thereby destroying the Administration’s entire rationale for restricting emissions.
As I say, it is quite odd that Shelanski failed to address these issues, since he is part of OMB. He even mentioned public comments on the choice of discount rate, but never brought up the fact—let alone explained—why the Working Group ignored the clear OMB guidelines.
Three Academic Economists Effectively Setting Federal Policy
While trying to justify the ~50 percent increase in the estimated SCC figure in just three years (from the 2010 report to the May 2013 update), Shelanski unwittingly lets the cat out of the bag about the true nature of using the SCC in federal rulemaking:
As explained in the February 2010 Technical Support Document, the SCC methodology rests on three integrated climate change assessment models: the FUND, DICE, and PAGE models….These are by far the most widely cited models that link physical impacts to economic damages for the purposes of estimating the SCC….
…[T]he 2010 SCC documentation committed to regular updates, and set a goal of updating the SCC estimates within two years or after updated versions of the underlying models became available. Since the February 2010 estimates were released, the three models that underpin the interagency social cost of carbon estimates have been all significantly updated and subsequently used in peer reviewed studies….It is important to note that the only changes made in May 2013 to the SCC estimates reflect the refinements made to the underlying models. In other words, all of the changes to the social cost of carbon values were the result of updates to the FUND, DICE, and PAGE models that were made by the model developers themselves. [Bold added.]
The words I’ve put in bold are very significant: Shelanski is admitting that the very large increase in the estimated SCC—a move that will imply many billions of dollars in additional “benefits” from federal rules that restrict carbon emissions—are the result of the subjective modeling decisions made by three academic economists.
I don’t want to overstate this fact, because it is shocking enough. Yes, these aren’t just random economists pulled from a phone book; their models were originally selected by the Working Group because they are considered the leaders in the published literature. Further, we have no reason to suppose that these three gentlemen have gone clinically insane within the last 36 months.
Even so, it is a true claim that many billions of dollars of federal agency decisions—affecting industry as well as households—rest on the estimated “social cost of carbon,” which in turn is calibrated by asking these three academic economists to update their models in the ways they see fit.
But Only Update Some of the Inputs…
Some critics might object to my analysis above, by claiming that of course the experts’ estimates of the economic damage due to climate change will vary over time, and so of course rational government policy needs to respond to those updates.
Yet wait a moment. In his testimony Shelanski goes on to explain: “The Federal Government inputs, such as the discounts rate, population growth, climate sensitivity distribution, and socioeconomic trajectories used to develop the 2010 estimates remain unchanged.”
Let’s ponder the significance of this statement. The Working Group plugged in the updated damage functions as spelled out in the latest versions of the economic models. However, they did not change the “climate sensitivity distribution,” even though the peer-reviewed climate science literature shows that the 3 degrees Celsius median figure (which the Working Group got from the IPCC’s Fourth Assessment Report) is far too high.
In summary: The three academic economists in 2013 have changed their models, such that a given amount of global warming will produce more economic damages than their models suggested back in 2010. So, the Obama Administration Working Group incorporated those changes, to make sure the estimate of the “social cost of carbon” reflected this new thinking and is now much higher.
At the same time, the academic climate scientists have changed their minds since the IPCC AR4 report, and now think that a given amount of emissions will produce less global warming than their models suggested back in 2010. The Obama Administration’s Working Group ignored these updated estimates from the scientific literature, even though they would greatly reduce the estimated “social cost of carbon.”
Trying to pooh-pooh critics of the new, higher number, Shelanski explains:
Entities outside of the Federal government are using estimates that are similar to the updated SCC values. For example, these updated estimates are consistent with the SCC values used by other governments, such as the United Kingdom and Germany. Major corporations, such as ExxonMobil and Shell, have also used similar estimates to evaluate capital investments.
This is hardly persuasive. Of course other governments will have every incentive to choose their parameters (such as the discount rate, equilibrium climate sensitivity, and damage functions) in order to produce a high SCC—they also want justification for more regulations and a carbon tax.
And the fact that major corporations are internally “pricing carbon” at an amount similar to major governments, is hardly surprising. Think about it: In the textbook treatment in which most people frame this issue, a “negative externality” is not something that a private firm cares about. It will only enter the firm’s accounting through the government’s policy responses (such as cap-and-trade or a carbon tax). Thus, major companies that use carbon-intensive processes have to forecast the future carbon tax (and similar policies that the government might impose) in order to evaluate various natural gas projects, etc. So it’s no wonder that they are preparing for a carbon tax of the same size the governments are telling us they plan on implementing: The accountants at Exxon and Shell have an Internet connection, after all.
Just because Exxon and Shell run economic analyses with a possible price on carbon emissions somewhat similar to what Obama Administration’s social cost of carbon does not validate the Obama Administration’s work. It merely affirms that companies are looking at the possible cost of regulation. They treat a prospective regulatory cost in the same way they would a prospective tax increase, since regulations are essentially taxation on their business models. If they failed to do this, they might be neglecting their fiduciary responsibility. Shelanski’s argument is nonsense: Of course corporations are going to anticipate a carbon “price” of the same level that major governments are announcing that they will impose. This fact has no bearing on whether it’s the right level.
The “social cost of carbon” is a dubious concept for use in government policy. It can be tweaked to produce any number desired. Howard Shelanski, in his recent testimony before the House, unwittingly proved this with his revealing remarks. We now have federal agencies evaluating rules by using damage functions updated by three academic economists, while ignoring clear OMB guidelines that would lower the number, and ignoring the changes in the climate science literature. This obvious cherry-picking only exacerbates what is already a dangerous procedure for making federal regulatory decisions. It also shows the tortured logic at work in an administration that touts its “transparency” to justify its ruinous regulatory policies.