Must Read: Walton Francis on the Social Cost of Carbon

by Marlo Lewis on April 10, 2014

in Blog

Post image for Must Read: Walton Francis on the Social Cost of Carbon

Walton Francis, formerly director of regulatory analysis at the Department of Health and Human Services, submitted a sharply critical comment letter to the Office of Management and Budget (OMB) on the Obama administration’s social cost of carbon (SCC) estimates.

Yes, I realize, this isn’t breaking news, but the SCC debate will be with us for years, and, as explained previously on this site, SCC analysis has become a menace to society.

SCC analysis is computer-aided sophistry. By “sophistry,” I mean what Socrates meant by it in Plato’s dialogues. Sophistry is sham wisdom raised to the level of a τέχνη (technê), the Greek word for art, craft, or trade.

The leading rap against the sophists in classical literature is that they “make the weaker argument [defeat/appear to be] the stronger.” That is, sophists use specious arguments to win in courts and public assemblies regardless of the merits of the case or issue in dispute. Turning common sense upside down, they make wrong look right and base look noble.

SCC analysis, similarly, uses sophisticated modeling to make renewable energy look like a bargain at any price and carbon energy look unaffordable no matter how cheap. OMB Circular A-4 on cost-benefit analysis admonishes agencies that “you cannot conduct a good regulatory analysis according to a formula.” SCC analysis is a license to regulate by formula. Grant the premise that carbon has a social cost, and presto, climate activists conclude that taxing and regulating away reliable, plentiful, affordable energy will make the economy more “efficient.” For further discussion, see the free market organizations’ comment letter to OMB on the social cost of carbon.

For those unfamiliar with this debate, the administration’s Interagency Working Group (IWG) has put out two technical support documents on the social cost of carbon (2010 TSD, 2013 TSD). The documents define the social cost of carbon as “an estimate of the monetized damages associated with an incremental increase in carbon dioxide [CO2] emissions in a given year.” Implicit in that definition is another one: The SCC “is the carbon tax that would be imposed by a benevolent social planner.”

The IWG’s 2013 SCC estimates stirred up controversy because they were roughly 60% higher than the 2010 estimates, yet nothing had happened either to the climate system or in climate science to warrant the revision. What did happen is the warming “pause” continued for another three years, the divergence between climate model projections and observed temperatures increased, and scientific research further discredited Al Gore’s doomsday scenarios.

With that as background, let’s review Francis’s comment letter.

Francis begins by noting how incomplete current knowledge is of the causes of climate variability and change:

We are also unable to explain what caused the Minoan, Roman, or Medieval warm periods, the Little Ice Age, or the emergence from the Little Ice Age, either as to causes, duration, or timing of change. As for CO2, even though we know a great deal about the theoretical basis for estimating the effects of anthropogenic global warming, its real-world “magnitude is highly uncertain” (testimony of climatologist Judith Curry before the U.S. Senate, January 16, 2014). Indeed, as has been widely reported, the observed level of warming over the past 15-20 years is so far below the predictions of current climate models that recent years fall below even the lower 95th percentile of their simulation runs.

Models v Observations Christy McKnider, WSJ Feb 20, 2014

Models exhibiting so little skill in predicting climate change are not up to the task of estimating either climate change damages or the benefits of CO2 reductions.

In the remainder of the comment letter, Francis discusses “major biases” that “strongly exaggerate” carbon’s social cost in the administration’s SCC analysis. The 2010 and 2013 TSDs:

  • Fail to explain that the underlying climate models are highly flawed, are not a valid basis for projecting future costs and benefits, are based on unverified assumptions, and substantially overstate warming compared to observed temperatures.
  • Fail to consider two plausible alternative scenarios: (1) climate sensitivity (the amount of warming expected from a doubling of CO2 concentrations) is closer to 1ºC than to 3ºC; (2) human CO2 emissions mitigate dangerous global cooling or even prevent the onset of the next ice age.
  • Fail to incorporate damage estimates that are limited by prevention measures — notably geo-engineering — that might cost-effectively mitigate global warming, and thereby, the associated climate damages.
  • Inflate CO2-related costs by adopting “incredibly pessimistic” assumptions about the costs of adaptation or “hugely exaggerated” assumptions about the scope and magnitude of climate impacts.
  • Assume unrealistically low economic growth rates — and, thus, unrealistically low adaptive capabilities — in developing countries in coming decades and centuries.
  • Fail to incorporate carbon “leakage” and “displacement” effects, which limit the benefits of domestic CO2 reduction measures.
  • Flout OMB guidance by using “only low discount rates” that inflate SCC estimates.

I will now condense, quote, and excerpt passages illustrating the points above.

1. Unverified and Biased Modeling Assumptions

The climate models underpinning the IWG’s SCC estimates “have now, by the normal standards of scientific inquiry, been falsified by actual temperatures observed. . . .Hence, current climate models, including specifically those used in the 2013 SCC revision, do not provide a credible or reliable basis for estimating either physical effects or subsequent economic effects.”

Moreover, the damage functions in the integrated assessment models (IAMs) that infer economic losses from global temperature increases “involve inherently arbitrary as well as judgmental assumptions or guesstimates.”

2. Bias in Omitting Alternative Outcomes

As we approach 20 years of a temperature plateau predicted by none of those models, the 3-5 degree [celsius] prediction has become all but indefensible as even an outer limit. In fact, even the IPCC (an organization with a vested interest in perpetuating the belief in “dangerous” global warming) recently admitted that an assumption of 1.5 degrees is within the plausible range of an accurate climate sensitivity projection, based on data that had not even incorporated the continued plateau of global temperatures in 2013. The IPCC also has admitted that catastrophic sea-level rise is a nonexistent threat in the next century (in addition to its previous admissions that other catastrophic scenarios are highly unlikely, and that global warming is not causing extreme weather events). In other words, even the IPCC has all but admitted that estimates of future catastrophe are either unfounded or much farther in the future than previously asserted [IPCC AR5, Chapter 12, Table 12.4, p. 1115].

But if future climate catastrophe is unlikely to occur, or will occur many decades later than previously assumed, the SCC estimates on which [the OMB requests] comment are necessarily invalid.

OMB Circular A-4 requires agencies to undertake a “sensitivity” analysis if “plausible changes” in assumptions would cause estimates of net benefits to change from positive to negative, or vice versa. “The failure of the SCC estimate to include the results of an assumption of 1- or 1.5 degree climate sensitivity equilibrium (rather than 3-5 degrees) flatly violates this standard.”

Alternatively, if the climate is highly sensitive to CO2 “forcing,” as many models assume, then rising CO2 emissions might prevent the onset of the next ice age over the next thousand years or the return of Little Ice Age conditions in the more immediate future. The IWG does not consider those potential benefits.

3. Bias in Disregarding Geo-engineering and other Mitigation Measures as a Limit on Damage Estimates

The most prudent assumption to make in cost-benefit analysis is that “limited sums will be spent to avoid losses of far larger sums.” If geo-engineering strategies are comparatively cheap and effective, governments will likely deploy them, thereby mitigating global warming and the associated cost of carbon. The IWG, however, assumes carbon taxes or regulations will always be the only game in town.

While the science of geo-engineering is still primitive, there are known methods that could substantially reduce global temperatures by changing the albedo [reflectivity] of the earth, including “painting” rooftops and highways white, broadcast dissemination of particulates or aerosols at high altitudes, and others. It seems inconceivable that the nations of the earth would stand by wringing their hands as the ice caps melted and coastal cities drowned, rather than take such simple and relatively inexpensive measures to reduce global temperatures. . . .

The SCC estimates fail to assume that such cost-reducing steps will be taken, and therefore fail to meet accepted analytic standards. . . .

. . . . Any such regulations that do not include the universe of lower cost mitigation measures in estimating benefits in their Regulatory Impact Assessments are inherently flawed.

4. Bias in Estimating Adaptation, Including Technological Advances

The administration relies on three IAMs: DICE, FUND, and PAGE. Either the cost estimates are “unduly high” or the adaptation estimates are “unduly low.” DICE projects damages in 2095 of $12 trillion, or 2.8% of global GDP. PAGE assumes adaptive measures will fail after global warming exceeds 2ºC. “The 2010 TSD says that DICE and FUND assume no technological advances in agriculture.”

In an era when genetic modification of plants is in its infancy, these assumptions seem incredibly biased towards exaggerating economic effects. Simply extrapolating changes in agricultural productivity of the last hundred years forward another hundred years, it is apparent that the area needed to feed all of humanity will likely be less than half of today’s area, and certainly far, far less than the land loss due to sea rise. Likewise, while there will be huge costs to protect coastal cities, these are essentially one-time costs. . . .Regardless of the “right” or “best” estimate, it is clear that the human guesses built into the IAM models are massively skewed towards the “worst case” end of the plausible range.

ML Comment: Francis does not exaggerate the technological pessimism of the 2010 TSD. From page 30: “Although DICE and FUND have both calibrated their agricultural sectors under the assumption that farmers will change land use practices in response to climate change (Mastrandrea, 2009), they do not take into account technological changes that lower the cost of this adaptation over time.”

5. Bias in Estimating Low Rates of Growth in Per Capita Income

The IAMs appear to assume that global per capita GDP will increase by less than 2% annually for the next two centuries. Those estimates are “very pessimistic.”

Given the population sizes involved, they imply that Brazil ($11,000 in 2012), China ($6,000 in 2012), and India ($1,500 in 2012) will have sharply reduced growth rates and not catch up to the present U.S. level ($52,000) in this century or even the next century. Previous estimates of even the Intergovernmental Panel on Climate Change have been many times higher. . . .William Nordhaus, the creator of the DICE IAM model, in The Climate Casino (2013) predicts that per capita income in India will rise “by a factor of almost 40 over the 2000-2100 period” (page 97) and that “another half century of similar growth will raise the per capita of India and China to around $50,000” in 2050 (page 145). . . .

Amazingly, these IAM models project that worldwide per capita income growth declines to a rate of zero by the year 2300, based on some combination of “scarcity of natural resources” and “degradation of environmental sinks” (2010 TSD page 44), as well as an implicit assumption that technological advances end. In other words, they adopt essentially the same assumptions as used by a book called The Limits to Growth. This 1972 book predicted gloom and doom by using computer simulation models. It was widely ridiculed by economists at the time . . . and the last 40 years have demonstrated its predictions to be laughable. . . .

The poorer the country, the less its ability to adapt to climate change, and the higher the estimated SCC. Nonetheless, and barring implausible climate catastrophes, developing nations should be able to adapt to climate change even assuming the TSDs’ pessimistic growth rate projections.

Despite the remarkable pessimism of these assumed low growth rates, a [per capita income] level of $37,000 will surely mean urbanization at our levels or higher, and air conditioning in essentially 100% of housing units, in the continents of Africa, Asia, and South America. This means the entire world will use modern sanitation methods (water, sewage, waste disposal, insect control, etc.), screened windows, air conditioned houses and automobiles, and a host of other adaptations to heat that we observe today in American cities such as Miami, Houston, and Dallas. . . .

6. Bias in Adverse Health Effects Assumptions

The IPCC has hugely exaggerated potential adverse effects on human health from global warming. It has published epidemiologically absurd claims that vector borne diseases such as malaria and yellow fever might return to North America, where such diseases were endemic two hundred years ago — when temperatures were far lower than today — and are now nonexistent. . . .In an Asia or Africa with per capita income levels higher than those of the U.S. today, there is no reason to think such diseases will even exist, with or without global warming. . . .Over the last century there have been massive reductions in deaths and death rates related to extreme weather events worldwide. . . .

 With respect to deaths due directly to heat, the U.S. data demonstrate how negligible these are in a modern economy . . . .fewer than a thousand deaths a year . . . .clearly a very small proportion of total deaths of around 2 million a year. As a matter of simple adaptation, devices available today or in the very near future using current technology could prevent most of these, whether reported or not. . . .

. . .the failure of epidemiologists to find evidence of higher levels of insect-borne disease, heat stroke, etc. in hotter American cities suggests that such evidence does not exist. Indeed, the hotter cities are undoubtedly better adapted to higher temperatures, and that is precisely the point.

7. Bias in Environmental Harm Assumptions

Polar bears have survived multiple warm and cold cycles over the ice age, including thousands of years of free-ice arctic waters and tens of thousands of years of ice sheets that prevented access to any arctic waters by polar bears or their prey. Coral reefs have survived 400 feet of sea rise over the last 15,000 years and could presumably survive another 10 or 20 feet in the unlikely event that this occurred. After one hundred years of global warming, the total number of species shown to have been driven to extinction from climate change is zero. . . .

In addition, the last interglacial period had average temperatures at least 5 degrees above this interglacial, and the rising temperature rate predicted by IAMs has been exceeded many times. It is hard to believe that any substantial magnitude of environmental harms could be attributed to CO2 emissions.

8. Bias in Omitting Leakage and Displacement Effects

The IWG omits estimates of leakage — CO2-emitting investments or activities that would migrate from the U.S. to countries with less or no carbon regulation and lower energy costs. Leakage reduces the benefit of domestic CO2 regulation relative to cost. A related issue is displacement. Rather than adopt or enforce carbon reduction measures, some countries may opt to free-ride off the (alleged) climate benefits of U.S. emission reductions. If so, our emission reductions would displace rather than combine with emission reductions by other countries.

The SCC estimates, and the derivative regulatory analyses, omit estimates of both leakage and displacement effects, and thereby estimate them at zero. This is another major bias in these calculations.

9. Bias in Discount Rate Calculations

ML Comment: Other things equal, the lower the discount rate, the higher the estimated SCC. The easiest way to ensure big, scary-sounding SCC estimates is to use discount rates below the 3%-7% range recommended by OMB. For “inter-generational” issues like climate change, OMB Circular A-4 allows agencies to perform additional “sensitivity” analyses using discount rates lower than 3%. However, estimates using a 7% discount rate are still required. Flouting OMB guidance, the 2010 and 2013 TSDs calculate SCC values using 2.5%, 3%, and 5% discount rates, but not a 7% discount rate. 

The failure to use and compare the cost of carbon at discount rates of both 3% and 7% is indefensible. It is a professional failure in the cost of carbon report, in every regulatory impact analysis that has quoted that report, and a failure on the part of OIRA to maintain the integrity of the regulatory review process. It was precisely to display the potentially wide range in estimates with distant costs or benefits that differed from near term costs or benefits, and to prevent “gaming” of estimates, that OMB mandated the use of these rates in economic analysis. This omission in estimating the cost of CO2 has no conceivable rationale other than to deceive. . . .

ML Comment: As of October 2013, federal agencies had adopted nearly 30 regulations embodying this deceptive practice. Here is a list, courtesy of the Institute for Energy Research:

Social Cost of Carbon List of Regulations October 2013



Comments on this entry are closed.

Previous post:

Next post: