Opinion: Feds fall short in pricing climate disruption

2 December 2010

As the Obama Administration prices climate change impacts to help justify emissions limits, it encounters a fundamental problem: Many of the biggest disruptions are impossible to value. 

What's the price of giving Sweden the climate of Spain?

By Michael MacCracken 

for the Daily Climate

What's the cost to society of each additional ton of carbon dioxide we emit? Last February, a federal interagency working group tried to identify this so-called "social cost of carbon." 

The panel assessed the United States' share of the global impacts of intensifying climate change [pdf] and settled on $5, $21, $35, and $65 (in 2007 dollars) per ton of carbon dioxide. The first three values reflect differing discount rates; the fourth shows the cost of worst-case impacts. The intended application of these numbers, particularly the suggested central value of $21, was to bring consistency across agencies in the required cost-benefit analyses of prospective regulatory actions aimed at limiting emissions of greenhouse gases.

As a physical scientist - unschooled in economics - who has worked on both estimating the changes in climate and the nature and intensity of impacts, I have a few concerns.

First, the phrase "social cost of carbon" misleads in an important way. The calculation attempts to provide an integrated, global estimate of the economic cost of all of the environmental and societal impacts - both positive and negative - of climate change, now and into the future, transformed to generate the present economic value of the full time history of anticipated impacts. 

That's a tall order. 

There are too many unknowns to effectively quantify climate change impacts, and many cannot be expressed in terms of economic cost: How does one value the potential loss of a significant fraction of the world's biodiversity? How does one value the cultural loss of a low lying island culture being inundated and the inhabitants having to relocate to a totally foreign society? What is the chance that the stresses of climate change impacts - such as drought and food shortages - might cause civil or regional war? 

For the United States, how might one value the impacts from the climate of the northern tier states becoming like the central tier and the central tier like the southern tier, and so on - or of the climate of a polar country like Sweden becoming something like that of, say, Spain? How does one value the inundation of Miami or New York - or at least the significant parts for which levee construction is not possible? What of the loss of social structure, character, and culture? 

2ºC upper bound

To my mind, the social cost of carbon should come in two parts: An economic estimate of the costs that need to be borne and a listing of the non-quantifiable, non-market impacts that will just have to be suffered. An economic cost of only what can be quantified is not a sufficient portrayal of the significance of climate change.

Second, what happens if global carbon emissions put us on a path beyond the 2ºC increase from pre-industrial temperatures that the world's major nations set as the limit in the Copenhagen Accord?

I raised this question last month at the invitation-only gathering, organized by the EPA, aimed at bettering federal efforts to value the social cost of carbon. We have already seen about 0.8ºC worth of warming and, given time lags in the system, are committed to more even if industrial emissions stopped today. Last December at the United Nations climate talks in Copenhagen, world leaders couldn't agree on how or when to start limiting emissions. But they did agree that going beyond the 2ºC global increase should be considered "dangerous" and unacceptable. 

That 2ºC upper bound is based on consideration of the types of impacts that cannot be monetized. It simply is unacceptable to inundate southern Florida or New York; it would be unacceptable if Sweden had the climate of Spain or Boston had the climate of Washington, D.C., or Atlanta; it is not clear that ecosystems can shift in a coherent manner that would retain their inherent value and ecological services. 

Destabilizing events

Scientists are not saying that climate change will cause humans to go extinct - only that civilization will be very different if emissions are not controlled. The actual costs of that transition, absent the impacts that cannot be monetized, are potentially calculable. But that value should not determine emissions limits. 

The Obama Administration's tally is incomplete - and likely to remain so for quite some time, given the gaps in our understanding. In Washington last month, I think we all agreed that improving the damage estimates for many of the potential impacts will require substantial research and analysis. 

Impact studies underway in California show that costs are several times larger when examined at the local scale rather than the continental to global scales presently used in available model simulations. 

And how should we account for the impacts of threshold events? Should destabilizing the Greenland ice sheet count against the present generation, or should those impacts be distributed over the time they are projected occur and then discounted to the present?

Challenging intergenerational issues abound, and at least some at the EPA conference felt that this whole approach of trying to add up all potential costs might be inappropriate - or impossible - due to the large uncertainties that simply cannot be reduced. 

Unprovable and unverifiable

Maybe it makes more sense to recognize that 2ºC is a line not to be exceeded and to set a value for carbon dioxide that would prod the marketplace to change the global energy system fast enough to avoid exceeding that temperature ceiling. 

If we were to do that, rough estimates peg the cost of carbon dioxide necessary to provoke a sufficiently deep change in the global energy system at $75 to $100 per ton. Interestingly, this is roughly the amount that updates of the models used by the interagency working group suggest are closer to the real, but ultimately unprovable and unverifiable, economic costs of future impacts.

Finally, I'm concerned about how this number - whatever it ultimately may be - will be used. 

If this "social cost of carbon" is meant to help determine the most cost effective ways to transition away from greenhouse emissions, having a perfectly accurate number may not matter. Many energy efficiency efforts more than pay for themselves in the absence of their climate change benefit and should be undertaken now even with a zero value for the cost of carbon. 

But I'd be loathe to use these cost estimates to evaluate whether or not to move forward with sharply limiting greenhouse gas emissions.

Why? Because in addition to all the uncertainties, there is no way to capture many aspects of the cultural, social, historical and ecological characteristics that we all value and that will be torn apart by global climate disruption.

Michael MacCracken is the chief scientist for climate change programs at the Climate Institute. 

Reservoir photo courtesy California Department of Water Resources.

DailyClimate.org is a nonprofit news service covering climate change.

Find more Daily Climate stories in the TDC Newsroom

 This work by The Daily Climate is licensed under a Creative Commons Attribution-No Derivative Works 3.0 United States License.