You are on page 1of 117

Chapter 4

Climate Science, Rational Basis Analysis and Endangerment

Contents
4.1

Climate Science ................................................................................................................... 2

4.2

Legal Comments Regarding Endangerment Finding and Rational Basis Analysis ........... 27

4.3

Calculating "Contribution" ................................................................................................ 30

4.4

Social Cost of Carbon ........................................................................................................ 35

4.4.1 Methodology .................................................................................................................... 35
4.4.2 Climate Science Components of the Modeling Underlying SC -CO2 .............................. 47
4.4.3 Global vs. Domestic Value .............................................................................................. 61
4.4.4 Uncertainty....................................................................................................................... 70
4.4.5 Process-related Concerns ................................................................................................. 75
Carbon dioxide fertilization .................................................................................................. 91
Discount rate ......................................................................................................................... 93
Application and aggregation ................................................................................................. 99
4.5

Miscellaneous .................................................................................................................. 112

4-1

4.1

CLIMATE SCIENCE

Comment 4.1-1: Commenters 9685 and 10966 assert that there are flaws with the major lines of
evidence that are referenced in EPA’s 2009 Endangerment Finding, and that therefore the basis
for the Endangerment Finding should be brought into question. Commenter 10966 notes that
each of the three lines of evidence are invalidated by the failure of real-world data to support
each line of evidence.
Response 4.1-1: EPA has previously responded to comments that took issue with the general
three lines of evidence that EPA referenced from the assessment literature in its 2009
Endangerment Finding. EPA, referencing the assessment literature, described the three lines of
evidence attributing observed climate change to anthropogenic activities in its 2009
Endangerment Finding as such: The first line of evidence arises from our basic physical
understanding of the effects of changing concentrations of greenhouse gases, natural factors, and
other human impacts on the climate system. The second line of evidence arises from indirect,
historical estimates of past climate changes that suggest that the changes in global surface
temperature over the last several decades are unusual. The third line of evidence arises from the
use of computer-based climate models to simulate the likely patterns of response of the climate
system to different forcing mechanisms (both natural and anthropogenic).
Comments on the first line of evidence (our basic physical understanding of the effects of
changing concentrations of greenhouse gases) were responded to in the RTC document
accompanying the 2009 Endangerment Finding (see, for example, section 3.2.2 of volume 3).
More recent evidence, such as from the 2013-14 IPCC Fifth Assessment Report (AR5),
continues to support the line of evidence that the anthropogenic buildup of greenhouse gases is a
major driver of climate change compared to natural factors such as changes in solar radiation
(see Figure SPM.5 of the Summary for Policymakers of the IPCC AR5, Working Group I).
Previous comments on the second line of evidence (that recent climate changes stand out as
unusual compared to estimates of past climates) were also addressed in the RTC, volume 2,
which accompanied the 2009 Endangerment Finding and Volume 1 of the Response to Petitions.
More recent conclusions in the IPCC AR5 state that current levels of the two most important
greenhouse gases, CO2 and CH4, are at higher levels now compared to at least the last 800,000
years, whereas this statement could only be made relative to the past 650,000 years based on
evidence available at the time of the 2009 Endangerment Finding. Furthermore, numerous
comments questioning the validity of the historic data, such as past records of global temperature
change, were addressed at length in 2009 (see, for example, Response 2-29 of the RTC where, in
response to numerous comments raising concerns about the global surface records, EPA states it
“has concluded that the three primary global surface temperature records (NOAA, NASA, and
HadCRUT) are reliable and credible.”). The more recent IPCC AR5 stated that, “Warming of the
climate system is unequivocal, and since the 1950s, many of the observed changes are
unprecedented over decades to millennia.”
Comments regarding the third line of evidence (the capability of climate models to simulate
expected patterns of human-induced climate change) were also addressed in the RTC in 2009
(volumes 3 and 4). More recently, IPCC AR5 stated, “Climate models have improved since the
AR4. Models reproduce observed continental-scale surface temperature patterns and trends over
4-2

many decades, including the more rapid warming since the mid-20th century and the cooling
immediately following large volcanic eruptions.”
EPA’s 2010 Denial of Petitions and accompanying response to petitions document addressed
similar critiques of the three lines of evidence. To note one part of EPA’s 2010 Denial (74 FR
49572): “Petitioners’ evidence does not materially change or warrant any less reliance on the
other important lines of evidence linking greenhouse gases and climate change: Our basic
physical understanding of the effects of changing greenhouse gas concentrations and other
factors; the broad, qualitative consistency between observed changes in climate and the computer
model simulations of how climate would be expected to change in response to anthropogenic
emissions of greenhouse gases (and aerosols); as well as other important evidence of an
anthropogenic fingerprint in the observed warming.” With several more years of research, the
IPCC AR5 actually strengthened their conclusion that humans have contributed to most of the
recent warming, finding that it is extremely likely (>95 percent confidence) that “more than half
of the observed increase in global average surface temperature from 1951 to 2010 was caused by
the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings
together.”
In sum, the general three lines of evidence attributing human activities to observed climate
change, as originally described in EPA’s 2009 Endangerment Finding, were sufficiently
supported in that original record, and remain well supported by more recent evidence.
Comment 4.1-2: Commenter 10091 argued that climate change does not result in negative
human health impacts, and may result in positive health outcomes. This commenter cites
improvements in general health statistics to argue that climate change has not resulted in
negative health outcomes and to suggest that continued climate changes could result in continued
improvements in these health metrics. For example, the commenter suggests that human health
in the United States has improved due to climate change, as evidenced by an increase in the life
span of Americans. The commenter also argues that climate change does not result in negative
health impacts from changes in the frequency or severity of extreme events, as evidenced by a
decrease in deaths from extreme events. The commenter contends that heat-related mortality risk
has declined over time in the US despite a rise in temperature and an increase in the magnitude
and frequency of heat events, suggesting that adaptation measures to decrease the US
population’s sensitivity to extreme heat are a health benefit that will continue under continued
climate change. In addition, commenter 10662 states that EPA does not directly demonstrate
adverse health effects of greenhouse gas emissions, relying instead on “conjectures” regarding
the relationship of ozone and PM2.5 with heat waves and drought. The commenter argues that,
because these pollutants are already required to meet health-based standards through the
NAAQS, that any further reductions in these pollutants cannot be considered to be justifications
for regulation. The commenter also argues that the EPA would have to regulate GHGs as ozone
pollutant precursors under NAAQS if they do directly affect ground-level ozone or PM2.5.
Finally, the commenter claims that the EPA has not identified any concentration or emissions
level of CO2 that would significantly decrease purported climate change effects.
Response 4.1-2: The EPA continues to rely on the best available science in order to make its
determinations. See also Coalition for Responsible Regulation v. EPA, 684 F. 3d at 120 holding
both that “[t]he body of scientific evidence marshaled by EPA in support of the Endangerment
4-3

Finding is substantial” and that EPA reasonably considered by major assessments in the field in
making its determination (“EPA simply did here what it and other decision-makers often must do
to make a science-based judgment: it sought out and reviewed existing scientific evidence to
determine whether a particular finding was warranted, it makes no difference that much of the
scientific evidence in large part consisted of ‘syntheses’ of individual studies and research. Even
individual studies and research papers often synthesize past work in an area and then build upon
it. This is how science works”; see also id. characterizing petitioners’ arguments regarding
consideration of the assessments as “little more than a semantic trick”). The EPA relies on major
assessment reports, including the US Global Change Research Program’s National Climate
Assessments (NCA), which consider the entire literature in context rather than selecting a small
number of individual studies. The third and most recent NCA (2014) states that:
“Climate change threatens human health and well-being in many ways, including impacts
from increased extreme weather events, wildfire, decreased air quality, and illnesses
transmitted by food, water, and diseases carriers such as mosquitoes and ticks. Some of
these health impacts are already underway in the United States.”
Improvements in some health metrics in recent decades, such as an increase in life expectancy or
a decrease in deaths due to extreme events, does not provide evidence to support a claim that
climate change is or will be beneficial to health. Improvements in broad health metrics can be a
result of changes in medical care, diet, income, or other factors. As noted in EPA’s Response to
Comments on the 2009 Endangerment Finding (RTC 5.5), the single metric of life expectancy
does not convey or provide information on the risks associated with changes in climate and
whether or not they are increasing, either in the past or in the future.
EPA does not dispute the findings of Davis et al. (2003a and 2003b) or Kalkstein et al. (2009)
that heat-related mortality in major U.S. cities declined in the second half of the 20th century and
that adaptive measures contributed to the decline. However, these findings do not support a
conclusion that climate change does not present a risk to human health. In fact, the 3rd NCA
(2014) states:
“Some of the risks of heat-related sickness and death have diminished in recent decades,
possibly due to better forecasting, heat-health early warning systems, and/or increased
access to air conditioning for the U.S. population. However, extreme heat events remain a
cause of preventable death nationwide.”
Commenter 10091 mischaracterizes the results of Bobb et al. (2014), which demonstrated a
decrease in heat-related mortality and sensitivity to extreme heat events in the United States, by
suggesting that this increase in acclimatization represents an increase in health benefits
associated with climate change. In fact, the conclusion of the Bobb et al., 2014 study states:
“The population has become more resilient to heat over time. Yet even with this
increased resilience, substantial risks of heat-related mortality remain. Based on 2005
estimates, an increase in average temperatures by 5°F (central climate projection) would
lead to an additional 1,907 deaths per summer across all cities.”

4-4

As noted in EPA’s Response to Comments on the 2009 Endangerment Finding (RTC 5-5, 5-7
and 5-36), risks to human health can be reduced by technology and adaptive measures, but heat
presently causes substantial mortality in the Unites States and adaptive measures do not
eliminate risk and are rather a response to that risk. In fact, decreased sensitivity or decline in
heat-related mortality may not be as pronounced in the future since access to air conditioning is
already high in most regions (Ebi et al., 2008). (See also RTC 5-45, 5-48, 5-51) The major
scientific assessments that EPA relies on, including the NCA, IPCC AR5, and the IPCC SREX,
all project negative impacts of climate change, and the subsequent increase in frequency or
severity of extreme events, on human health.
Commenter 10662 criticized EPA’s demonstration of the linkage between climate change and
health. First, it is important to note that the only facet of this linkage that the commenter
addressed was the connection between climate change and air quality. The commenter did not
present any arguments regarding the impacts of climate change on human health through heat
waves, fires, extreme weather events, or food, water, or vectorborne illnesses. Regarding the
criticism of the linkage between air quality and climate change, it is also important to note that
the commenter did not provide any evidence or citations to research that contradict such a
linkage. In the Endangerment Finding, EPA cited EPA’s Interim Assessment stating that there
“is now consistent evidence from models and observations that 21st century climate change will
worsen summertime surface ozone in polluted regions of North America compared to a future
with no climate change.” The 2014 National Climate Assessment cited estimates that climatechange induced increases in ozone and particulate matter, absent changes in human emissions,
are projected to lead to “1,000 to 4,300 additional premature deaths nationally per year by
2050.” The commenter did not provide any evidence counter to these conclusions. The EPA
specifically addressed the argument by the commenter that the NAAQS are protective in the
Endangerment Finding, (Federal Register Vol 74, No 239, pg. 66530). EPA stated first that
despite the existence of a NAAQS, not all areas are in attainment, and climate change can
worsen air quality in those areas. EPA stated second that even in areas that are in attainment,
because ozone and PM have no threshold below with no effects can be observed, increases in
concentration in those areas would still likely result in additional adverse health e ffects for
some individuals. The EPA concluded that, “[t]he clear risk to the public from ozone increases
in nonattainment areas, in combination with the risk to some individuals in attainment areas,
supports the finding that overall the public health is endangered by increases in ozone resulting
from climate change.” Despite the connection between climate change and ozone levels,
greenhouse gases do not qualify as “ozone precursors” which is a term for compounds that react
in the atmosphere to form ozone locally, whereas greenhouse gases cause increased ozone
concentration through the mechanism of changing the climate globally. Finally, the EPA has
stated that any decrease of greenhouse gas emissions or concentrations will reduce climate
change.
Therefore, EPA has considered these comments, and finds that they do not provide credible
evidence of flaws in the EPA’s approach of relying upon the best estimates of the major
assessments, nor do they demonstrate that climate change does not represent a threat to human
health.
Comment 4.1-3: Commenter 10951 submitted a review of the so-called Climategate 2.0 e-mails
claiming that “the second installment of Climategate e-mails provides a strong basis for renewed
4-5

scrutiny of the Endangerment Finding.” The commenter identifies 5 themes within the e-mails
that they claim demonstrate that the IPCC produced flawed scientific work, and that therefore the
endangerment finding is similarly flawed. The five themes identified by the commenter include:




The Divergence Problem with historical temperature proxies based on tree-ring data.
Leading climate scientists’ evasion of freedom of information act requests and
destruction of responsive material.
Internal expressions of doubt among the leading climate scientists as to the validity
and accuracy of their work.
Concerted efforts to downplay the Medieval Warming Period as natural analogue to
20th Century warming.
Shortcomings in the peer review process.

The commenter claims, “Considered together, the foregoing themes significantly undermine the
prevailing theory of climate change and unprecedented warmth in the late 20th Century as
reported in the Fourth Assessment Report of United Nations Intergovernmental Panel on Climate
Change, and as reflected in the motor vehicle endangerment finding. The Climategate 2.0
material makes abundantly clear that the leading proponents of this theory, who were key IPCC
authors, were biased and that their approach to science was, in fact, highly unscientific and
results-oriented. In short, climate science must be rigorously re-evaluated in light of these
systemic problems.”
Response 4.1-3: This comment should be considered in the context of an extensive record
including the Endangerment Finding, the Response to Comments document, the Denial and
Response to Petitions, and the recent Court ruling. The Administrator considered the full body of
scientific evidence in making the Endangerment Finding, and this body of evidence is
voluminous and well documented.
The e-mails that Commenter 10951 has highlighted are a second batch of e-mails from the
incident in 2009 when a server at the University of East Anglia was hacked and a number of e mail communications between scientists at the Climatic Research Unit (CRU) from 1999 to 2009
were copied from that server. The first batch of e-mails from the 2009 incident was analyzed and
responded to in great detail in the EPA Denial and Response to Petitions. The conclusion of the
EPA, upheld by the Court, was that the contents of the original set of e-mails did not undermine
the Endangerment Finding, as stated in the Denial:
“As EPA’s review and analysis shows, the petitioners routinely take these private e -mail
communications out of context and assert they are ‘‘smoking gun’’ evidence of
wrongdoing and scientific manipulation of data. EPA’s careful examination of the emails and their context shows that the petitioners’ claims are exaggerated, are often
contradicted by other evidence, and are not a material or reliable basis to question the
validity and credibility of the body of science underlying the Administrator’s
Endangerment Finding or the Administrator’s decision process articulated in the Findings
themselves. Petitioners’ assumptions and subjective assertions regarding what the e-mails
purport to show about the state of climate change science are clearly inadequate pieces of
4-6

evidence to challenge the voluminous and well documented body of science that is the
technical foundation of the Administrator’s Endangerment Finding.1”
Similar conclusions have been reached independently by a number of other bodies, including the
UK House of Commons, Science and Technology Committee, the University of East Anglia,
Oxburgh Panel, the Pennsylvania State University, and the University of East Anglia, Russell
Panel. The outcomes of each of these independent panels are also in the existing record.
The commenters’ arguments regarding the hacked e mails have also been held by the D.C.
Circuit to be of insufficient import to justify reconsidering the endangerment finding because
they do not provide a basis for questioning the correctness of that determination. See Coalition
for Responsible Regulation v. EPA, 684 F. 3d at 124-25 (e mails not of “’central relevance to the
outcome of [the Endangerment Finding]… and do not “’provide substantial support for the
argument that the regulation should be revised’”).
EPA has reviewed all of the new e-mails provided by Commenter 10951. There is nothing in the
e-mails that would cause EPA to come to a different conclusion than it came to in 2009 or 2010.
Commenter 10951 repeats all the same errors, misunderstandings, faulty conclusions, acts of
hyperbole, and “cherry-picking” that the Petitioners did in 2010. The 2010 Denial of Petitions
and its more technical supplemental Response to Petitions already addressed the bulk of the
Commenter’s arguments, and the Commenter asserts but does not demonstrate that EPA’s Denial
or Response to Petitions mischaracterized the original e-mails in any way.
Regarding the five themes identified by the commenter, the EPA has previously addressed each
of these themes, and the commenter does not demonstrate that they are raising any new issues.
Regarding the so-called “divergence problem” and the related issue of downplaying the medieval
warm period, the EPA addressed these issues in Volume 1.1 of the Response to Petitions, finding
that all the relevant uncertainties in attempts to determine historical temperatures through
indirect measures such as tree rings were appropriately discussed in the assessment reports, and
thus already considered in the Endangerment determination. The commenter does not
demonstrate that these uncertainties were mischaracterized, or how these uncertainties would
undermine the conclusions of the Endangerment Finding. Regarding freedom of information
(FOI) issues, the commenter fails to link the e-mails regarding FOI to any substantive scientific
errors or to any data that has not been made publically available in the period since the emails
were written. Regarding internal expressions of doubt among climate researchers, the commenter
does not recognize that these discussions of uncertainty parallel discussions of uncertainty in the
published literature and the assessment reports. Finally, regarding assertions of shortcomings in
the peer review process, the commenter does not show why the emails they provide, which
highlight scientists criticizing the work of other scientists, are improper. As stated in Volume 3.3
of the Response to Petitions, “the petitioners make broad leaps in logic to impugn the overall
integrity of the peer-reviewed climate change literature based on very limited examples and
largely speculative assertions.”
1

EPA’s Denial of the Petitions to Reconsider the Endangerment and Cause or Contribute Findings for Greenhouse
Gases, http://www.epa.gov/climatechange/Downloads/endangerment/response-decision.pdf, 49557.

4-7

An identical comment was submitted to the docket for the 2012 proposal. The EPA also
reviewed the comment at that time, and discussed the emails in questions in the preamble of the
2013 proposal:
“One commenter submitted emails between climate change researchers from the period
1999 to 2009 that were surreptitiously obtained from a University of East Anglia server
in 2009 and publicly released in 2011. According to the commenter, these emails showed
that the climatologists distorted their research results to prove that climate change causes
adverse effects. The EPA reviewed these emails and found that they raised no issues that
Petitioners had not already raised concerning other emails from the same incident,
released in 2009. The commenter's unsubstantiated assumptions and subjective assertions
regarding what the emails purport to show about the state of climate change science is not
adequate evidence to challenge the voluminous and well-documented body of science
that underpins the Administrator's Endangerment Finding.”
The commenter did not acknowledge EPA’s discussion of this issue, and the EPA’s conclusion
still holds that these emails do not provide a reliable basis to question the validity and credibility
of the body of climate science.
Comment 4.1-4: Commenters 9733 and 10045 argue that EPA should exclude biogenic CO2
emissions from this rule because they are carbon neutral. The commenters begin by arguing that
EPA’s 2009 Endangerment Finding only identified as the cause of endangerment emissions from
fossil fuels, and specifically omitted biomass from national GHG emission totals upon which it
relied. Therefore, the commenters argue that the rational basis upon which the EPA relies for
regulating CO2 cannot be relied upon for biogenic carbon, and that it would therefore be arbitrary
and capricious for EPA to regulate biogenic CO2 before it has finalized its scientific review and
Accounting Framework which will determine whether these sources should be regulated under
the Clean Air Act. Commenter 9733 argues that language in the EPA 2009 Inventory of
Greenhouse Gas Emissions and Sinks admits that carbon “release during the consumption of
biomass is recycled as U.S. forests and crops regenerated, causing no net addition of CO2 to the
atmosphere.” Commenter 10045 also notes that the 2006 IPCC Guidelines recommend that
inventories place emissions of CO2 from combustion of biomass in the section devoted to
forestry and land-use changes as opposed to the section devoted to energy production, and states
the RFS2 program excludes emissions of crop-derived CO2 from combustion and fermentation in
comparisons of the lifecycle GHG emissions of newly-proposed biofuels. Commenter 9733
provides additional data regarding carbon sequestration from forests. Commenter 9733 argues
that the use of forest product manufacturing residuals for combined heat and power facilities is
preferable to letting the residuals decay, potentially leading to release of methane, citing a study
which calculated that the lifecycle emissions from decay will exceed that from burning the
residuals, on average, for any time period greater than 2.4 years. Commenter 10045 argues
specifically that indirect emissions of CO2 associated with production and gathering of cropderived fuels are also harmless. Commenter 10045 also disagrees with EPA’s 2009
Endangerment Finding Response to Comments response 9-9, stating that the author of the
response failed to recognize that the definitional foundation of the 2009 Findings contained a
science-based determination that crop-derived CO2 is inconsequential to global warming.
Commenter 10045 concludes by arguing that, even if EPA did have the authority to regulate
4-8

crop-derived CO2, it should exclude it on de minimis grounds, and that when EPA finishes its
Biogenic Accounting Framework it should set a default of zero for crop-derived CO2 emissions.
Response 4.1-4: Regarding the 2009 Finding, the commenters have pointed to the fact that some
of the inventories relied upon by the EPA have excluded biogenic emissions. However, the
inventories, and the Endangerment TSD, have been clear that land use, land-use change, and
forestry emissions are not included in the gross national totals but rather in net emission totals.
This is not a statement on whether these emissions endanger. The Endangerment arises from the
elevated concentrations of greenhouse gases: while fossil fuel emissions may be the cause of the
majority of the increase in concentrations of carbon dioxide, land use, land-use change, and
forestry have also contributed to that increase (the “air pollution” at issue), as noted in the
Endangerment TSD, and therefore there was no intent to exclude biogenic sources from the
Endangerment decision.
EPA developed the revised draft report, Framework for Assessing Biogenic CO2 Emissions from
Stationary Sources, to continue advancing our understanding of the role the use of biomass can
play in strategies to address greenhouse gas emissions. Specifically, the Framework was
developed as a policy-neutral framework for assessing biogenic CO2 emissions from stationary
sources—it was not developed as technical guidance in conjunction with any specific policy or
program. The EPA is engaging in a second round of targeted peer review of the revised draft
Framework with the SAB in 2015. In November 2014, the agency released a second draft of the
technical report, Framework for Assessing Biogenic Carbon Dioxide for Stationary Sources. The
revised Framework, and the EPA’s Science Advisory Board (SAB) peer review of the 2011
Draft Framework, finds that it is not scientifically valid to assume that all biogenic feedstocks
are “carbon neutral” and that the net biogenic CO2 atmospheric contribution of different biogenic
feedstocks generally depends on various factors related to feedstock characteristics, production,
processing and combustion practices, and, in some cases, what would happen to that feedstock
and the related biogenic emissions if not used for energy production. For more information
regarding the revised Framework.2
Comment 4.1-5: Commenter 3593 notes that the EPA endangerment finding for greenhouse
gases did not rely on any analysis done within EPA and instead erred in using science
assessments developed by others. Commenter 10951 states the assessments produced by the
Intergovernmental Panel on Climate Change (IPCC) are scientifically flawed due to how the
reports are developed. One commenter argues that the NRC assessment report process is not
transparent, and that the reports are not independent because they rely on previous assessments
developed by IPCC and USGCRP. Commenter 10046 argues that by placing lesser weight on
individual studies than on major scientific assessments, EPA introduces timing and data-bias
issues. Commenter 10046 first points out that the next partial IPCC assessment was due in
March, with the final document due October of 2014, and further states that EPA timed its
proposal by not waiting for the full and formal review of that assessment, and that EPA is
therefore systematically refusing to consider the most recent evidence on these issues.

2

see: http://epa.gov/climatechange/ghgemissions/biogenic-emissions.html

4-9

Response 4.1-5: Commenters repeat the argument already rejected by the D.C. Circuit as “little
more than a semantic trick”. See Coalition for Responsible Regulation, 684 F. 3d at 120. As
outlined in Section VIII.A. of the 2009 Endangerment and Cause or Contribute Findings for
Greenhouse Gases under Section 202(a) of the Clean Air Act (Endangerment Finding), EPA’s
approach to providing the technical and scientific information to inform the Administrator’s
judgment regarding the endangerment question was to rely primarily upon the recent, major
assessments by the U.S. Global Change Research Program (USGCRP), the Intergovernmental
Panel on Climate Change (IPCC), and the National Research Council (NRC) of the National
Academies. In brief, these assessments addressed the scientific issues that the EPA was required
to examine, were comprehensive in their coverage of the greenhouse gas (GHG) and climate
change problem, and underwent rigorous and exacting peer review by the expert community, as
well as rigorous levels of U.S. government review and acceptance, of which EPA took part.
Primary reliance on the major assessments provided EPA greater assurance that it was basing its
judgment on the breadth of well-vetted science that reflected the consensus of the scientific
community. The major findings of the USGCRP, IPCC, and NRC assessments supported the
finding that GHGs threaten the public health and welfare of current and future generations. EPA
demonstrated this scientific support at length in the Endangerment Finding itself, in its Technical
Support Document (which summarized the findings of USGCRP, IPCC and NRC), and in its
RTC document. Use of IPCC, CCSP, USGCRP, and NRC Reports as the primary scientific
basis is addressed in Response to Comment (RTC) Volume 1 Section 1, with specific comments
on EPA’s use of IPCC reports addressed in Section 1.1.1. The IPCC Principles and Procedures
are outlined in EPA’s Response to Comments Volume I Appendix A.
On June 26, 2012, the United States Court of Appeals for the District of Columbia Circuit )
confirmed the appropriateness of using scientific assessment literature, stating that the EPA had
based its decision on “substantial scientific evidence” 3 and noting that EPA’s reliance on
assessments was consistent with what decision-makers often must do to make a science-based
judgment4.
In addition to the above description of EPA’s rationale for relying on the major assessments as
the primary scientific basis for the endangerment determination, Section III.A of the
Endangerment Finding, “The Science on Which the Decisions Are Based,” describes that it was
not necessary nor logical for EPA to conduct an additional and separate review of the underlying
climate data and research. Section III.A states: “EPA has no reason to believe that the
assessment reports do not represent the best source material to determine the state of science and
the consensus view of the world’s scientific experts on the issues central to making an
endangerment decision with respect to greenhouse gases. EPA also has no reason to believe that
putting this significant body of work aside and attempting to develop a new and separate
assessment would provide any better basis for making the endangerment decision, especially
because any such new assessment by EPA would still have to give proper weight to these same
consensus assessment reports.”

3 Coalition for Responsible Regulation, Inc. v. EPA, 684 F.3d 102, 120 (D.C. Cir. 2012).
4 Id. at 120; see also id. (“EPA is not required to re-prove the existence of the atom every time it approaches a
scientific question”).

4-10

In response to the comment regarding IPCC’s assessment report development process, we note
that this comment has been previously raised during the public comment and reconsideration
processes for the Endangerment Finding (see Section 1.1.1 of Volume 1 the Response to
Comment document, and Section 2.2 of Volume 2 of the Response to Petitions document).
Response to Comment 1-14 in Volume 1 of the Response to Comment document addresses IPCC
processes for ensuring information quality, objectivity, and transparency, and clearly states that
EPA and the U.S. Government approve the rigorous IPCC review process and the legitimacy of
its scientific content:
“EPA has both evaluated and participated in the development and review of IPCC
reports, and the IPCC process is transparent and rigorous. The IPCC ensures scientific
credibility and legitimacy of its reports by fairly representing the range of scientific
opinions on climate change, and the IPCC provides multiple opportunities for input from
experts along the entire spectrum of opinions…
…The evidence is clear that the IPCC’s procedures are sufficient and effective for
ensuring quality, transparency, and consideration of multiple and diverse perspectives.
Because the assessment reports EPA used in developing the TSD represent the best
available science, and because supporting studies were conducted in accordance with
sound and objective scientific practices, were peer reviewed, and adhered to standards of
quality based on objectivity, utility, and integrity, we find that IPCC’s information
quality process is consistent with EPA’s Guidelines for Ensuring and Maximizing the
Quality, Objectivity, Utility and Integrity of Information Disseminated by the
Environmental Protection Agency.
The U.S. Government participated fully in the development, review, and ultimate
acceptance and approval of IPCC (2007). As stated on the USGCRP’s Web site: “When
governments accept the IPCC reports and approve their Summary for Policymakers, they
acknowledge the legitimacy of their scientific content”.”
In response to the comment regarding the NRC assessments and the procedures employed during
their development, we note that this comment was already responded to following the public
comment process for the Endangerment Finding5. In short, we disagree with the commenter that
the NRC reports are not transparent. The NRC assessment reports cited in the Endangerment
Finding record were prepared following rigorous and transparent processes addressing such
issues as the nomination and selection of authors, the caliber of literature reflected in the
assessment, and the processes for review and revision of reports. We also disagree with the
comment that the reports are not independent. Just because the NRC reports reference findings
from previous scientific assessments produced by others does not signify that the conclusions of
the NRC report were not independently reached. Instead, previous findings that are reassessed in
light of the most recent scientific literature and subsequently reaffirmed, helps to demonstrate
increased confidence in those findings. In fact, in its ruling, the DC Circuit Court stated “EPA is
not required to re-prove the existence of the atom every time it approaches a scientific question:”
similarly, scientific assessments do not need to ignore the existence of all previous assessments.
5

See Section 1.1 of Volume 1 of the Response to Comments document “Use of IPCC, CCSP, USGCRP, and NRC
Reports as the Primary Scientific Basis” and Appendix C of Volume 1 “NRC Report Development Procedures”

4-11

In response to the comment regarding the use of individual studies and the timing of new
assessments, we note that EPA gave careful consideration to all of the scientific and technical
information received during the public comment processes. As stated above, the Agency
primarily relied on the major assessments because these documents provided EPA with greater
assurance that it was basing its judgment on well-vetted science that reflected the consensus of
the scientific community. On the other hand, less weight was placed on the much smaller
number of individual studies that were not considered or reflected in the major assessments—
often the case when the study was published after some cut-off date to be incorporated into the
larger assessments. EPA reviewed the individual studies submitted to the Agency largely to see
if they would lead EPA to change its interpretation of, or place less weight on, the major findings
reflected in the assessment reports. From its review of individual studies submitted to the
Agency, EPA has concluded that these individual studies do not change the various conclusions
or judgments EPA would draw based on the more comprehensive assessment reports. EPA also
reviewed the most recent science assessments, such as IPCC’s Fifth Assessment Report, that
have been published since the proposed action was released. The Agency found that the findings
of these most recent assessments are consistent with the conclusions previously relied upon by
the Agency, and in some cases strengthen the scientific understanding of how GHGs endanger
public health and welfare. One of the primary goals of reviewing the individual studies
submitted to the Agency and the latest scientific assessments was to address the data and timing
biases raised by the commenters. EPA plans to continue relying primarily on the major
assessments by the IPCC, the USGCRP, and the NRC. Studies from these bodies address the
scientific issues that the Administrator must examine, represent the current state of knowledge on
the key elements for the endangerment analysis, comprehensively cover and synthesize
thousands of individual studies to obtain the majority conclusions fro m the body of scientific
literature and undergo a rigorous and exacting standard of review by the peer expert community
and U.S. government.
Therefore, EPA has considered these comments, and finds that they do not provide credible
evidence of flaws in EPA’s approach of relying and placing primary weight on the major
scientific assessments.
Comment 4.1-6: A number of commenters use three general, somewhat related, arguments to
claim that recent science suggests that the sensitivity of the climate system to increased
concentrations of greenhouse gases is lower than was previously assumed. The first is that
commenters claim that there has been a recent pause in observed global warming that has lasted
since 1998. One commenter claims that this pause makes the causal connection between
greenhouse gases and climate change uncertain (commenter 10046). Another commenter
suggests that estimates of future warming should be lowered because climate models did not
predict this pause and a historical over-prediction of warming implies faulty models (commenter
10951). The second argument for a lower climate sensitivity estimate is based on studies by
Loehle (2014), Spencer and Braswell (2013), or on a review of 18 recent studies that all show a
best estimate of climate sensitivity lower than the mean of the IPCC models (commenter 3953).
The third argument is that the EPA should take into account changes in how the 5 th IPCC
Assessment (AR5) characterized climate sensitivity. In this regard, one commenter claims that
the IPCC decision not to present a “best estimate” of climate sensitivity completely undermines
4-12

its findings (10966) while another commenter claims that the IPCC presented evidence from
several studies showing a climate sensitivity range of 1.5 to 2 degrees.
Response 4.1-6: The EPA continues to rely on the best available science. In the case of climate
sensitivity, the EPA relies on the major assessment reports, including the recent IPCC 5 th
Assessment Report. These assessment reports consider the entire literature in context, rather than
selecting a small number of individual studies. The likely range for the climate system response
to a doubling of carbon dioxide concentrations (i.e., the climate sensitivity) described by the
IPCC AR5 assessment of 1.5 to 4.5 degrees represents a sound judgment of the literature, and the
appropriate range on which to base EPA decisions. The commenters compare reported climate
sensitivity from a set of studies to the reported climate sensitivity of the climate models assessed
by the IPCC. The studies selected by the commenters may present a lower climate sensitivity
than included in most climate models, but the estimated likely (>90%) climate sensitivity range
made by the IPCC assessment itself is broader than the range found in climate models, and
encompasses all but one of the studies presented by the commenters.
The commenters’ characterization of the available evidence is limited. For example, commenters
discuss a recent 16 year pause in the rate of global surface temperature rise, but mischaracterize
or do not consider a number of important factors that put this “pause” in context within observed,
long term climatological trends. The first factor is that this is not actually a pause but rather a
reduction in the rate of surface temperature rise. The second is that observations of the slower
rate during this pause depends on which dataset is chosen, but, according to the IPCC, all global
surface temperature datasets do exhibit a warming trend over the period 1998-2012;. The third is
that this reduction in rate is mostly evident only in atmospheric temperatures, and that ocean heat
content and sea level rise (which can be a proxy for ocean heat content) both continue to increase
rapidly. Other measures of climate change, such as Arctic sea ice retreat and glacial melt,
similarly show no pause or even demonstrate an acceleration during this time period. A final
important factor is that trends based on short time periods such as a 16 year period often do not
reflect long-term climate trends.
As with the issue of climate sensitivity, the EPA relies on the major assessment reports,
including the recent IPCC 5 th Assessment Report when considering the characterization of the
recent slowdown in the rate of rise of global surface temperatures. The IPCC AR5 assessment
discussed the recent reduction in surface temperature rate at length, noting several of the above
issues. They also note that models generally did not capture the reduction in rate in large part
because they do not reflect a recent increase in observed heat uptake by the oceans which has
been a result of natural variability.
Estimating the climate sensitivity is an area of ongoing research. Many of the studies cited by
commenters are based on a method of using simple models and recent observed temperatures. It
is important to also give appropriate weight to other methodologies, such as paleoclimate and
theoretical approaches, as in done in the IPCC assessment process. Many of the studies cited by
the commenters were considered by the IPCC, and a number of authors of the cited studies were
also IPCC authors. Therefore, EPA finds the IPCC process to have appropriately accounted for
not only the research highlighted by the commenters but the full spectrum of the literature that
relates to climate sensitivity estimation. The IPCC range does include the possibility that the
4-13

sensitivity is below 3 degrees, but also shows that there is a possibility that the climate sensitivity
is higher than 3 degrees.
The IPCC AR5 provided a best estimate of a likely range of climate sensitivity of 1.5 to 4.5
degrees of warming for a doubling of CO2, compared to a range of 2 to 4.5 degrees in AR4.
While such an expansion in the range of best estimates of climate sensitivity to include lower
sensitivity values would likely lower estimates of future damages, the reduction in the best
estimate would have to be extreme in order to reduce the damages to the point where they would
no longer endanger public health or welfare. In contrast, this is a relatively small change to the
previous uncertainty range.
EPA previously responded to a number of similar comments in 2009, in Section 4.3 of the EPA’s
Response to Comments on the 2009 Endangerment Finding. In those responses, EPA also
showed how the approach of the major assessments to determine the Earth’s climate sensitivity
was sound, and considered the entire body of science in context, rather than relying on individual
studies that were often sensitive to a choice of specific temperature dataset or single
methodology.
Therefore, EPA has considered these comments, and finds that they do not provide credible
evidence of flaws in the EPA’s approach of relying upon the best estimates of the major
assessments, or that the assessments did not take into account the new literature, nor do they
demonstrate that a lower best estimate of climate sensitivity would eliminate endangerment
rather than merely slightly reducing the danger.
Comment 4.1-7: Commenter 10951 states that the IPCC Fifth Assessment Report (IPCC AR5)
includes greater doubts about surface temperature data quality than was admitted in previous
IPCC assessments, and that the IPCC acknowledges that its earlier dismissal of the problem of
surface data contamination was made without supporting evidence. The commenter provides as
support a document by Ross McKitrick which provides several quotes from the IPCC AR4 and
AR5 assessments regarding the urban heat island and a discussion of a 2004 paper by McKitrick
and Michaels along with several subsequent papers. The commenter claims that the
endangerment finding relied on claims of the previous IPCC report, and that since, according to
the commenter, this new language overturns that claim, the endangerment finding should be
revised.
Response 4.1-7: The EPA continues to rely on the best technical and scientific information,
primarily based upon the recent, major assessments by the U.S. Global Change Research
Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National
Research Council (NRC) of the National Academies. While the final version of the IPCC AR5
assessment was not available at the time of proposal, the EPA has since reviewed the AR5
assessment, along with other new assessments of the USGCRP and the NAS, and found that the
improved understanding of the climate system they present strengthens the case that GHGs
endanger public health and welfare. Particularly relevant for discussions of surface temperature,
confidence in attributing recent warming to human causes has increased since the publication of
the 4th Assessment Report: the IPCC stated that it is extremely likely (>95 percent confidence)
that human influences have been the dominant cause of recent warming, compared to the
previous Fourth Assessment Report conclusion that most of the recent warming was very likely
4-14

(>90 percent confidence) due to the observed increase in human-produced greenhouse gases.
Moreover, the IPCC found that the last 30 years were likely (>66 percent confidence) the
warmest 30 year period in the Northern Hemisphere of the past 1400 years.
The evidence is clear that the Earth has warmed. A number of metrics are clear indicators of a
warming planet, including increases in ocean heat content, sea level rise, glacial melt, and
satellite observations of a warming atmosphere. These indicators add to the evidence provided by
the surface temperature record. However, we also disagree that there are major flaws in the
surface temperature record. Similar comments were submitted in 2009, and were extensively
discussed in the Response to Petitions, Volume 1. The Response to Petitions described how a
number of different methods by different research groups all conclude that there has been recent
warming, regardless of whether they accounted for heat island effects by using nightlights, or
more stations, or, in the US, a Climate Reference Network designed specifically to be used for
monitoring climate change. As in most of climate science, there is ongoing research on
methodology improvements (whether to adjust for urban heat islands or for lack of coverage of
rapidly warming regions such as the Arctic), but the major conclusions remain robust.
Regarding the commenter’s discussion of the change in IPCC reports, we find that the
commenter has mischaracterized the conclusions of the 5th Assessment report. The commenter
claims that “[t]he IPCC now acknowledges that its earlier dismissal of the problem of surface
data contamination was made without supporting evidence”. While the IPCC did note that
“explicit” evidence had not been included in the original assessment regarding a lack of
statistically significant correlation between socioeconomic development and warming (where
such a correlation would be potential evidence of temperature station bias), this does not imply
that the previous conclusions were wrong. In fact, the IPCC cites a newer publication (Schmidt,
2009) that “supported AR4 conclusions”, though the IPCC also cites a pair of publications by
McKitrick that “still yielded significant evidence for such contamination of the record”. While
this paragraph in isolation does not provide a clear determination about the accuracy of the
surface temperature record overall, in context of Section 2.4.1.3, it is clear that the overall
conclusion of the IPCC was that the urban heat island effects are small compared to the global
trend. This conclusion was based on a variety of methodologies, ranging from using satellites to
classify stations and determine that there is little difference between warming at urban and rural
sites (Wickham et al. 2013, Hansen et al. 2010), to reanalysis methods that compare surface
warming from models constrained by observed sea surface temperatures (such as Parker, 2011).
The IPCC also found that regression-based results such as the McKitrick papers were less
convincing than other approaches, stating that they were “inconsistent with many other
components of the global observing system.” So when the IPCC considered all the available
approaches to assessing surface temperature biases and urban heat island effects, it determined
with high confidence that based on the totality of the evidence, that it was unlikely that
uncorrected effects would have causes an error or more than 10%.
The commenter also mischaracterizes the change between IPCC AR5 assessment and the
previous Fourth Assessment, stating that “Rather than saying categorically that no such bias
exists [resulting from urban heat islands], they say only that it is unlikely the effects are greater
than 10% of the observed global average trend”. However, the IPCC never categorically claimed
that the urban heat island effect did not exist: in fact, the AR4 report (Trenberth et al., 2007)
states: “Studies that have looked at hemispheric and global scales conclude that any urban4-15

related trend is an order of magnitude smaller than decadal and long-scale trends evident in the
series.” The “order of magnitude smaller” conclusion from AR4 is consistent with the “10%”
conclusion in AR5.
Therefore, EPA has considered these comments, and finds that they do not provide credible
evidence that the surface temperature record has major flaws, or that any flaws that may exist
would materially change the conclusions that greenhouse gases endanger public health and
welfare.
Comment 4.1-8: Comments 9664 and 10961 describe that greenhouse gas emissions, including
CO2, seriously endanger public health and welfare. Comment 1871 states that the expected
negative health impacts of climate change have been well studied and include: increased
malnutrition and infectious disease, injury, heat stress, respiratory disease, reproductive and
developmental disorders, cancer, physical and psychological trauma from extreme weather
disasters, and increases in insect-borne disease. Commenter 9514 states that scientific and
analytical assessments issued since the 2009 Endangerment Finding indicate that climate change
is a more serious threat than previously realized, and that CO2 emissions are more, not less,
likely to endanger the public health and welfare.
Commenter 10090 states the standards to be set by EPA should ensure that new power plants use
the most efficient, lowest-emitting technology available, strongly supports EPA's limits on new
coal power plants and supports requirements that would mandate that any new coal plants use
technology to capture carbon emissions to ensure these plants meet emission limits set by EPA
because addressing carbon pollution from coal fired power plants tackles one of America's
dirtiest fuel sources and is an immense step forward Commenter 9514 similarly states that a
strong NSPS for fossil fuel-fired power plants is critical to achieving the emissions reductions
necessary to mitigate the dangers of climate change.
Response 4.1-8: In 2009, the EPA Administrator found that the current and projected
concentrations of the six key well-mixed greenhouse gases — carbon dioxide (CO2), methane
(CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur
hexafluoride (SF6) — in the atmosphere threaten the public health and welfare of current and
future generations. The full scientific record supporting the Endangerment Finding, in particular
the scientific assessment literature as summarized in the TSD, support the Administrator’s
determination. We therefore agree with the comments that greenhouse gases, including CO2,
endanger public health and welfare.
Regarding the adverse health effects of climate change, we note that the EPA evaluated and
considered scientific information on observed and projected health impacts during the
development of the 2009 Endangerment Finding. The specific health effects mentioned by the
commenter have been described in the scientific assessments that the Agency relied upon in
making the Endangerment Finding and include direct (temperature effects) and indirect (e.g.,
changes in climate-sensitive diseases) impacts on human health. For more specific information
regarding EPA’s consideration of these health effects, see Volume 5 of the Response to
Comments document supporting the 2009 Endangerment Finding.

4-16

In response to the comment regarding assessments released since the 2009 Endangerment
Finding, we agree that these reports, in general, show stronger evidence that climate change
poses risks to human health and welfare. The preamble summarizes many of these findings. For
example, recent assessments provide new evidence that reaffirms the linkages between climate
change and certain extreme weather events such as increases in heavy rainfall in some areas and
greater extent of wildfire in others; estimates of greater sea level rise than previously projected;
more evidence of the threats of ocean acidification; impacts of climate change on air quality;
more confidence in the role of human emissions of greenhouse gases in recent global climate
change; and many other examples.
Comment 4.1-9: Some commenters (9685, 10091, 10951) argue that the EPA has not
adequately considered the benefits of CO2 fertilization on plants, especially agricultural crops.
Commenters rely primarily on NIPCC (2009) and Idso (2013) when stating that increased
atmospheric CO2 concentrations can increase crop yields, reduce the amount of water plants
need, and improve plants’ ability to respond to environmental stresses. Another commenter states
that, “more than 40 recent peer-reviewed academic journals articles have appeared, supporting
the benefits of increased atmospheric CO2 on plant life.” Further, they note that higher crop
yields can increase food availability and lower food prices. In contrast, a different commenter
(9681) expressed concern that CO2 fertilization benefits have been overstated and may be
cancelled out by negative climate change impacts on agriculture, such as extreme heat, pests, and
weeds.
Response 4.1-9: The EPA has never disputed that CO2 is a required ingredient for plant growth
and survival, or that CO2 can stimulate plant growth. This comment was put forth previously and
responded to in the Response to Public Comments Document of the Endangerment Finding. The
2009 Endangerment Finding Technical Support Document (TSD) summarizes the scientific
assessment literature on the stimulatory, or fertilization effect of CO2 on plant growth and
productivity. For example, Section 3 states that CO2, “can have a stimulatory or fertilization
effect on plant growth” and “studies have demonstrated increases in CO2 affects water use and
water use efficiency in plants” (EPA 2009). Section 9 focuses specifically on U.S. agriculture
and shows that for grain and oilseed crops, increased CO2 and temperature will likely speed plant
growth in the near-term future (USDA CCSP 2008). The IPCC (2007) also confirms that
experimental studies of elevated CO2 on crops increases yields.
However, the EPA considers the stimulatory effects or beneficial influences of elevated CO2 on
plants alongside other climate change impacts. Many aspects of climate change can negatively
affect plant growth and crop yields and can outweigh the benefits of CO2 fertilization, as
suggested by commenter 9681. For instance, the USCGRP found that high temperatures, extreme
weather events (i.e., heavy downpours, droughts), enhanced pest and weed growth, and ozone
exposure can significantly limit the benefits of the direct stimulatory CO2 response on plant
growth. The UGCRP also concluded that as temperature rises beyond a certain threshold in the
future, grain and oilseed crops will increasingly begin to experience failure, especially if climate
variability increases and precipitation lessens or becomes more variable. The IPCC (2007)
concluded that the global price of cereals may decrease with warming below 2 oC, but above this
temperature, prices could increase by 30%. Further, USGCRP concludes that the marketable
yield of many horticultural crops (e.g., tomatoes, onions, fruits) is very likely to be more
sensitive to climate change than grain and oilseed crops. More recent assessments (not available
4-17

at the time of the 111(b) proposal) released by the IPCC (AR5) in 2013, and the USGCRP in
2014, reaffirm these findings. For instance, the IPCC states, “based on many studies covering a
wide range of regions and crops, negative impacts of climate change on crop yields have been
more common than positive impacts”.
EPA has considered the comments regarding the effects of elevated CO2 concentrations on
plants, and determined that they do not provide credible evidence of flaws in the EPA’s approach
of relying upon the characterization of CO2 fertilization in the assessment literature in making its
determination. In particular, the commenters did not recognize that EPA had already considered
the effects of CO2 fertilization when originally finding that greenhouse gases endanger public
health and welfare, and had responded to a number of related comments. The commenters
generally rely on individual studies showing benefits of elevated CO2 in the absence of any other
climate changes: therefore, the commenters neither present a comprehensive picture of impacts
of CO2 on the agricultural sector in specific, or on the broader issues of public health and
welfare. In contrast, EPA relies on the assessment literature, which demonstrates that the net
impacts of elevated GHG concentrations on agriculture will likely be negative, when considering
relevant factors such as temperature, precipitation, weather variability, and ground level ozone,
and that in the broader context of all climate impacts, it is even more clear that climate change
represents a threat to public health and welfare.
Comment 4.1-10: Multiple commenters highlight the need to protect children from the negative
impacts of climate change (commenters 10025, 1871, 6358, 6318, 6351, 3984, 6292, 5608).
Several commenters note that climate change have already negatively affected children’s health
and are projected to further exacerbate health impacts, such as asthma and other respiratory
illness, infectious disease, malnutrition, heat stress, and mental health disorders (commenters
1871, 6318, 6351, 3984). These commenters state that there exists an obligation to protect future
generations from the effects of climate change.
Response 4.1-10: The EPA continues to rely on the best technical and scientific information,
primarily based upon the recent, major assessments by the U.S. Global Change Research
Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National
Research Council (NRC) of the National Academies. These assessment reports consider the
entire peer-reviewed literature in context to draw conclusions based on the entire body of
findings from the literature.
The preamble summarizes many new assessment reports published since the 2009 Endangerment
Finding, including the recent USGCRP’s 2014 Third National Climate Assessment (NCA3), the
IPCC’s 2013-4 Fifth Assessment Report (AR5) and IPCC’s 2012 “Special Report on Managing
the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation” (SREX).
These new assessments provide further evidence supporting the 2009 Endangerment Finding that
climate change will have a negative impact on human health and welfare, including children’s
health.
EPA has considered the information provided by the commenters and finds that it is generally
consistent with the findings of the major scientific assessments. We agree with commenters that
children are a population of concern with respect to health impacts from climate change. As
noted in EPA’s Response to Comments on the 2009 Endangerment Finding (RTC 5-21), children
4-18

are particularly vulnerable to illness and death from heat waves and acute and chronic illnesses
of the respiratory system from particulate matter emissions during forest fires. The most recent
National Climate Assessment (2014) found that children, primarily because of physiological and
developmental factors, will disproportionately suffer from the effects of heat waves, air
pollution, infectious illness, and trauma resulting from extreme weather events.
Overall, the EPA finds that the information provided by the commenter is generally consistent
with and does not alter the 2009 Endangerment Finding that climate change will have a negative
impact on human health and welfare.
Comment 4.1-11: Commenter 3593 refers to a document written by Alan Carlin at the EPA
criticizing the science used in the Endangerment Finding. The commenter argues that EPA
violated due process by ordering Alan Carlin not to discuss the report. The commenter alleges
that after Alan Carlin’s report was forwarded to the endangerment finding workgroup that it was
dismissed without consideration or response due to a predetermined decision by the EPA, and
that the EPA never analyzed the underlying scientific issues raised by Dr. Carlin. The commenter
states that the appropriate remedy for the due process violations under the Fifth Amendment,
including violation of the rights of protected classes of individuals, that EPA cannot promulgate
the rule as written. The commenter alleges that the Agency cannot appropriately address the
mischief involved, and that therefore it must withdraw the proposal in its entirety.
Response 4.1-11: The document written by Alan Carlin was not ignored. EPA reviewed and
addressed the comments submitted by Alan Carlin in the record for the Endangerment Finding in
2009. At the time, Alan Carlin’s document was a collection of various statements; it was not a
publication nor a new analysis. In Volume 1 of EPA’s responses to public comments (RTC) in
support of the Endangerment Finding, we stated: “EPA has reviewed new and additional
scientific literature [in addition to the major scientific assessments on which EPA placed primary
weight] received through public comment. This review includes the Carlin document.” 6 In other
volumes of the RTC there are numerous examples where EPA responds directly to statements
made in the Carlin document. In volume 2 of the RTC 7, numerous commenters used statements
from the Carlin document and these were responded to in detail. For example, several
commenters submitted a comment from Alan Carlin claiming that the most reliable sets of global
temperature data available, using satellite microwave sounding units, show no appreciable
temperature increases during the time from 1978 to 1997. EPA provided a detailed response
(Response 2-48), concluding it was “unclear how the Carlin comment reached this conclusion,
because our review of the two satellite temperature datasets…show temperature increases for the
period from 1979 (when annual data became available) to 1997.” In volume 4 of the RTC
(available at: http://epa.gov/climatechange/endangerment/comments/volume4.html), several
commenters referred to the Carlin document to state that the IPCC ignored the possibility that
there may be other significant natural effects on global temperatures that we do not yet
understand, and that therefore this possibility invalidates IPCC statements implying that one
must assume anthropogenic sources in order to duplicate the temperature record. EPA’s response
(4-14) addresses the comment in detail, including the statements that “EPA has reviewed Alan
6 Response 1-13; available at: http://epa.gov/climatechange/endangerment/comments/volume1.html.
7 http://epa.gov/climatechange/endangerment/comments/volume2.html.

4-19

Carlin’s document”, that “the assessment literature clearly demonstrates that the historical
pattern of global temperature change is due to a combination of natural and anthropogenic
factors,” and that “we find that the evidence that the Carlin document relies on for this analysis
of recent temperature trends is very weak.” In sum, Alan Carlin’s comments were considered,
reviewed, and responded to by EPA, and these comments did not warrant calling into question
the fundamental science relied upon that supported EPA’s 2009 Endangerment Finding.
Comment 4.1-12: Some commenters have argued that the EPA relies too heavily on the 2009
Endangerment Finding and that the agency has inadequately considered the most up -to-date
science in moving forward with the 111(b) rulemaking. One commenter (9201) asserts that the
fact that the Endangerment Finding (and subsequent petition denials) has been upheld in the
courts does not preclude the agency from reconsidering the endangerment determination in light
of the most up-to-date climate science. The commenter also notes that the science relied upon
was current in 2006 and will be nearly a decade old when states submit plants for the upcoming
111(d) rulemaking. Another commenter (10951) notes that the 2007 IPCC report (AR4) and the
2009 USGCRP report, both of which the EPA relied upon in making the Endangerment
determination, provide the scientific knowledge only up to 2006 and 2008, respectively. This
commenter asserts that current scientific knowledge does not fully support the scientific
conclusions that underpin the Endangerment finding, focusing on their interpretation of the most
recent IPCC report (AR5), NRC reports, and on the so-called “Climategate.”
Response 4.1-12: The EPA recognizes that climate change is a very active area of scientific
research and that multiple assessments have been released since the 2009 Endangerment Finding.
Accordingly, the EPA has and continues to review all new relevant assessments and use this
updated information to inform rule development. This continued scientific review began
following the Endangerment determination and is ongoing. As discussed in the preamble, new
assessments since the 2009 Endangerment Finding were considered and evaluated, and informed
this 111(b) rule. These recent assessments include several NRC reports of the National Academy
of Sciences, the IPCC Fifth Assessment Report (AR5), the 2014 National Climate Assessment ,
and the IPCC’s 2012 “Special Report on Managing the Risks of Extreme Events and Disasters to
Advance Climate Change Adaptation” (SREX).
Based on our review of the more recent scientific assessments, we disagree with the assertion
that recent reports summarizing climate change research contradict the body of science that
underpinned EPA’s 2009 Endangerment Finding. In fact, these assessments show stronger
evidence to support the determination that GHGs are reasonably anticipated to endanger public
health and welfare. The preamble summarizes the findings of many of these reports. For
example, recent assessments provide even stronger evidence attributing the human-induced
buildup of GHGs in the atmosphere to recently observed climate change, as well as new
evidence reaffirming the linkages between human-induced climate change and extreme weather,
increases in heavy rainfall in some areas and greater extent of wildfire in others, increased loss of
sea ice in the arctic, and greater projected sea level rise than previously documented.
Recently released assessments that were not available prior to the 111(b) rulemaking further
strengthen the determination that GHGs threaten human health and welfare. Both the IPCC
(2013) and the National Climate Assessment (USGCRP 2014) document widespread impacts on
heath and numerous sectors important to human well-being. For instance, the IPCC Fifth
4-20

Assessment Report states that, “Impacts from recent climate-related extremes, such as heat
waves, droughts, floods, cyclones, and wildfires, reveal significant vulnerability and exposure of
some ecosystems and many human systems to current climate variability”. These include
disruption of food production and water supply, damage to infrastructure, morbidity and
mortality, and consequences for human well-being, among others. Additional impacts include
increased air pollution and exposure to allergens, as well as greater emergence of diseases such
as lyme (USGCRP 2014). In sum, it is our view that consideration of more up-to-date climate
science would only serve to reaffirm EPA’s judgment that led to the original Endangerment
Finding.
Comment 4.1-13: Commenter 10091 states that there has been an overall long-term decline in
mortality from extreme weather events (hurricanes, floods, lightning, tornadoes) in the United
States, adjusted for population changes. The commenter cites Goklany, I.M. (2011). Wealth and
Safety: The Amazing Decline in Deaths from Extreme Weather in an Era of Global Warming,
1900–2010. Policy Study 393, Reason Foundation. Based on this study, the commenter
concludes that anthropogenic climate change will not cause adverse human health impacts from
extreme weather events.
Response 4.1-13: The EPA continues to rely on the best technical and scientific information,
primarily based upon the recent, major assessments by the U.S. Global Change Research
Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National
Research Council (NRC) of the National Academies. These assessment reports consider the
entire peer-reviewed literature in context to draw synthesis conclusions based on the body of
findings from the literature. In contrast, the commenter draws a conclusion from one individual,
non-peer-reviewed study.
The preamble summarizes many new assessment reports published since the 2009 Endangerment
Finding, including the recent USGCRP’s 2014 Third National Climate Assessment (NCA3), the
IPCC’s 2013-4 Fifth Assessment Report (AR5) and IPCC’s 2012 “Special Report on Managing
the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation” (SREX).
These new assessments provide additional evidence to support the 2009 Endangerment Finding
that climate change will have a negative impact on human health and welfare, including
reaffirming the linkages between climate change, extreme weather, and human health. The
assessment literature comprehensively discusses mortality, morbidity, and future projected
impacts via various climate-health exposure pathways.
The commenter, on the other hand, focuses solely on whether there is an observed national
mortality trend from hurricanes, floods, lightning, and tornadoes. The commenter ignores three
key aspects of the assessment literature’s synthesis of the current state of scientific
understanding: (1) that the discussion of health impacts focuses on those types of extreme
weather events with appropriate links to climate change; (2) that in the U.S., the health impacts
of extreme weather will manifest differently according to region and population affected; and
(3) that health impacts of extreme weather include morbidity, not just mortality.
Regarding these three key aspects, the NCA3 concludes the following:

4-21

Certain types of extreme weather events with links to climate change have
become more frequent and/or intense, including prolonged periods of heat,
heavy downpours, and, in some regions, floods and droughts (NCA3
Highlights pg x).
Extreme weather events often lead to fatalities and a variety of health impacts
on vulnerable populations, including impacts on mental health, such as
anxiety and post-traumatic stress disorder. Large-scale changes in the
environment due to climate change and extreme weather events are increasing
the risk of the emergence or reemergence of health threats that are currently
uncommon in the United States, such as dengue fever [NCA3, pg 34].
The effects of temperature extremes on human health have been well
documented for increased heat waves, which cause more deaths, hospital
admissions and population vulnerability. […] The effects of weather extremes
on human health have been well documented, particularly for increased heavy
precipitation, which has contributed to increases in severe flooding events in
certain regions. Floods are the second deadliest of all weather-related hazards
in the United States. Elevated waterborne disease outbreaks have been
reported in the weeks following heavy rainfall, although other variables may
affect these associations. Populations living in damp indoor environments
experience increased prevalence of asthma and other upper respiratory tract
symptoms. […]The effects of extreme weather on mental health have been
extensively studied. Studies have shown the impacts of mental health
problems after disasters, with extreme events like Hurricane Katrina, floods,
heat waves, and wildfires having led to mental health problems. Further work
has shown that some people with mental illnesses are especially vulnerable to
heat. Suicide rates vary with weather, dementia is a risk factor for
hospitalization and death during heat waves, and medications for
schizophrenia may interfere with temperature regulation or even directly
cause hyperthermia. Additional potential mental health impacts include
distress associated with environmental degradation, displacement, and the
knowledge of climate change [NCA3, pgs 252-253]
In addition, the IPCC AR5 states that
Extreme heat events currently result in increases in mortality and morbidity in
North America (very high confidence), with impacts that vary by age, location,
and socioeconomic factors (high confidence). […] Extreme coastal storm events
can cause excess mortality and morbidity, particularly along the East Coast of the
USA, and the Gulf Coast of both Mexico and the USA […] Further climate
warming in North America will impose stresses on the health sector through more
severe extreme events such as heat waves and coastal storms […]
In light of these conclusions, EPA has considered the information provided by the commenter
and finds that it does not provide credible evidence of flaws in the EPA’s approach of relying
4-22

upon the synthesis conclusions of the major assessments, nor does it demonstrate that there are
no health impacts associated with extreme weather events related to climate change.
Comment 4.1-14: Commenter 10091 states that there is no definitive scientific evidence
regarding the direction or magnitude of changes in extreme weather events due to climate
change. Commenter 10046-6983 states that no observational evidence exists to support a global
increase in extreme weather events.
Response 4.1-14: The EPA continues to rely on the best technical and scientific information,
primarily based upon the recent, major assessments by the U.S. Global Change Research
Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National
Research Council (NRC) of the National Academies. These assessment reports consider the
entire peer-reviewed literature in context to draw synthesis conclusions based on the body of
findings from the literature. In contrast, the commenters do not cite the basis of their conclusion.
The preamble summarizes many new assessment reports published since the 2009 Endangerment
Finding, including the recent USGCRP’s 2014 Third National Climate Assessment (NCA3), the
IPCC’s 2013-4 Fifth Assessment Report (AR5) and IPCC’s 2012 “Special Report on Managing
the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation” (SREX).
These new assessments provide additional evidence to support the 2009 Endangerment Finding
and reaffirm the linkages between climate change, extreme weather, and human health.
The NCA3 concludes that “Certain types of extreme weather events with links to climate change
have become more frequent and/or intense, including prolonged periods of heat, heavy
downpours, and, in some regions, floods and droughts” (NCA3 Highlights pg x). In contrast to
the commenter’s argument regarding lack of observational evidence of extreme weather trends,
additional findings of the NCA3 state:
Some extreme weather and climate events have increased in recent decades, and new and
stronger evidence confirms that some of these increases are related to human activities.
Changes in extreme weather events are the primary way that most people experience
climate change. Human-induced climate change has already increased the number and
strength of some of these extreme events. Over the last 50 years, much of the United
States has seen an increase in prolonged periods of excessively high temperatures, more
heavy downpours, and in some regions, more severe droughts (NCA3 pg 24).
The IPCC also finds observed changes in climate extremes related to temperature, precipitation,
tropical cyclones, and sea level. As noted in the TSD for the 2009 Endangerment Finding, these
changes “apply generally to all parts of the globe, including the United States, although there are
some regional and local exceptions due to patterns of natural climate variability” [4(k) pg43].
The IPCC SREX assessment report (IPCC 2012) further supports the Administrator’s finding
regarding extreme weather and climate events:
A changing climate leads to changes in the frequency, intensity, spatial extent, duration,
and timing of extreme weather and climate events, and can result in unprecedented
extreme weather and climate events. […] There is evidence that some extremes have
4-23

changed as a result of anthropogenic influences, including increases in atmospheric
concentrations of greenhouse gases. […] Extreme events will have greater impacts on
sectors with closer links to climate, such as water, agriculture and food security, fores try,
health, and tourism.
In light of these conclusions, EPA has considered the information provided by the commenter
and finds that it does not provide credible evidence of flaws in the EPA’s approach of relying
upon the synthesis conclusions of the major assessments, nor does it demonstrate that no
observational evidence exists to support a global increase in extreme weather events.
Comment 4.1-15: Commenter 1871 concludes that there are many studies regarding the
expected negative health impacts of climate change, which includes “physical and psychological
trauma from extreme weather disasters.” Commenter 9681-5964 states that climate change is
making coastal flooding, drought, and impacts from extreme weather worse, citing Peterson, T.
C., M. P. Hoerling, P. A. Stott and S. Herring, Eds., 2013: Explaining Extreme Events of 2012
from a Climate Perspective. Bull. Amer. Meteor. Soc., 94 (9), S1–S74.
Response 4.1-15: The EPA continues to rely on the best technical and scientific information,
primarily based upon the recent, major assessments by the U.S. Global Change Research
Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National
Research Council (NRC) of the National Academies. These assessment reports consider the
entire peer-reviewed literature in context to draw synthesis conclusions based on the body of
findings from the literature.
The preamble Section II.A.3 summarizes many new assessment reports published since the 2009
Endangerment Finding, including the recent USGCRP’s 2014 Third National Climate
Assessment (NCA3), the IPCC’s 2013-4 Fifth Assessment Report (AR5) and IPCC’s 2012
“Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate
Change Adaptation” (SREX). These new assessments provide additional evidence to support the
2009 Endangerment Finding that climate change will have a negative impact on human health
and welfare, including reaffirming the linkages between climate change, extreme weather, and
human health.
EPA has considered the information provided by the commenters and finds that it is generally
consistent with the conclusions of the assessment literature. The NCA3 concluded that “certain
types of extreme weather events with links to climate change have become more frequent and/or
intense, including prolonged periods of heat, heavy downpours, and, in some regions, floods and
droughts” (NCA3 Highlights pg x). The Peterson (2013) study was not available at the time
when NCA3 was being developed, but an earlier edition was assessed and cited in NCA3.8
The NCA3 also discusses various health impacts of extreme weather associated with climate
change—including deaths, hospital admissions, elevated waterborne disease outbreaks, increased
prevalence of asthma and other upper respiratory tract symptoms, and mental h ealth problems
8

Peterson, T. C., P. A. Stott, and S. Herring, 2012: Explaining extreme events of 2011 from a climate perspective.
Bulletin of the American Meteorological Society, 93, 1041-1067, doi:10.1175/BAMS-D-12-00021.1.

4-24

after extreme weather disasters—that are consistent with the commenter’s description of
“physical and psychological trauma from extreme weather disasters.”
Overall, the EPA finds that the information provided by the commenter is generally consistent
with and does not alter the 2009 Endangerment Finding that climate change will have a negative
impact on human health and welfare.
Comment 4.1-16: Commenter 10618 states EPA presents calculations using a benefit per ton
reduced for NOx and SO2 that suffer from additional flaws. National Ambient Air Quality
Standards are designed to protect the human and ecological health with an adequate margin of
safety. Additional reductions of these pollutants well below these standard levels should not have
any quantifiable health benefit. Additionally, the modeling and calculations presented by EPA
ignore the projected impacts of the MATS Rule on air quality, which in many cases will drive
emissions below the Lowest Measured Levels in the studies used to support the health claims,
thus invalidating the applicability of the studies to a benefit calculation. Furthermore, the
Krewski et al. and Lepeule et al. studies which underpin the health benefit calculations that EPA
estimates (see, e.g., RIA at 5-42) use data from the 1980s and 1990s that do not take into account
current air quality emission levels or trends. For example, air emissions post-MATS
implementation are going to be significantly lower than the years used in these studies.
Additionally, these studies fail to differentiate health response between various components of
particulate matter even though more recent studies show associations between locally produced
carbonaceous compounds but NO associations between utility produced SO 2 and NOx emissions.
There is also concern with EPA's continued use of the Value of a Statistical Life calculated from
willingness to pay survey results as these results do not appropriately value premature mortality
that could be measured in days. A more robust measure needs to be developed that appropriately
values premature mortality with consideration to temporality.
Commenter 10618 states EPA should not include any discussion of health benefits within the
RIA as any such benefits are indirect and speculative. Furthermore, EPA should update its
calculations of health benefits to include use of a broader range of scientific literature, including
peer-reviewed studies showing no association with PM and mortality and significant issues with
health benefit calculation methodologies.
Commenter 9687 states EPA's own data shows that PM 2.5 concentrations are well below the
national standard and are continuing to decrease, yet EPA claims a co-benefit for PM reductions
under the proposed rule. Air Quality Trends: EPA creates air quality trends using measurements
from monitors located across the country. The table below shows that air quality based on
concentrations of the common pollutants has improved nationally since 1980. EPA data shows
that annual PM2.5 concentrations have been reduced by 33% from 2000 to 2012. The 24-hour
PM2.5 concentration has been reduced even more: 37%. The PM trends graphic shows that the
national average concentration of PM 2.5 fell below the EPA standard in 2007. By 2012 the
average was roughly 20% below the national standard and the band showed that ninety percent
of sites have concentrations below the national standard. Commenter concludes that EPA is
claiming harm from concentrations that are BELOW the national standard.
Response 4.1-16: The benefit-cost analysis included in the RIA accompanying the proposed and
final rules was conducted in compliance with Executive Order 12866. While the primary
4-25

conclusion of the analysis was that there would be no costs or benefits associated with these
standards, because all new generating units built in the period of analysis would be compliant
with these standards in the baseline. However, the EPA conducted two illustrative analyses that
considered the costs and benefits if key assumptions were to change. Consistent with OMB9 and
EPA guidance10, when conducting a cost-benefit analysis to meet the requirements of EO 12866,
the EPA estimates all of the anticipated costs and benefits associated with a regulatory action to
the extent feasible, including benefits anticipated to occur from reducing air pollution to below
the NAAQS levels. As EPA has consistently stated, the NAAQS are not risk free, and as a
result, consistent with scientific evidence and CASAC review.
The EPA disagrees that it has double-counted the benefits for this rule, as all of the estimated
health benefits are attributable to the emissions reductions associated with this proposed rule.
The EPA notes that NAAQS are set at a level deemed by the EPA Administrator to be protective
of public health within an adequate margin of safety and are not set at a level of zero risk.
Therefore, there would still be health benefits that occur at the lower concentrations, even if we
have less confidence in the magnitude of those benefits. The EPA believes that the best estimate
of benefits includes benefits both above and below the levels of the NAAQS and maintains it is
not double-counting benefits simply because the magnitude of the health benefits that occur at
lower concentrations are more uncertain.
The EPA’s standard practice for its rules is to estimate, to the extent data and time allow, all
benefits of the emissions reductions achieved by a rule beyond control requirements for other
rules, i.e., establish a baseline. While it can be difficult to account for concurrent rulemakings in
a baseline, the EPA clearly identifies what is and what is not in the baseline for each analysis. If
this proposed rule was duplicative of other rules, then there would be no additional costs or
benefits attributable to this proposed rule. Prior to estimating the health benefits of this proposed
rule (and any other rule), we simulated what PM2.5 concentrations would be in the future to
account for the air quality benefits that would occur due to other regulations (e.g., MATS) or
economic factors in this baseline. Any emissions changes expected as a result of this proposed
rule are additional emissions reductions beyond the other regulations included in the baseline
(e.g., MATS). Therefore, the benefits from particle reductions are not double-counted – they are
real health benefits from emissions reductions anticipated to be achieved by this proposed rule.
Further, the PM2.5 and ozone health co-benefits expected from this proposed rule are not doublecounted with benefits estimated in the NAAQS RIAs. NAAQS RIAs illustrate, but do not
predict, the emissions reductions strategies that States may choose to enact when implementing a
revised NAAQS. Subsequent Federal and State implementation rules will be reflected in future
baselines for PM and ozone NAAQS reviews. Also, because it is not possible to accurately
account for rules that have not yet been promulgated, RIAs prepared for a future rulemaking will
9

Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis. Washington, DC. Available
on the Internet at http://www.whitehouse.gov/omb/circulars/a004/a-4.html.
10
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Guidelines for Preparing Economic Analyses. EPA
240-R-10-001. National Center for Environmental Economics, Office of Policy Economics and Innovation.
Washington, DC. December. Available on the Internet at <http://yosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-056850.pdf/$file/EE-0568- 50.pdf>.

4-26

likely include any additional rulemakings in the baseline. For example, the baseline in this RIA
reflects many recently promulgated rulemakings, including MATS, CSAPR, and CCR.
The U.S. EPA follows a highly structured, transparent and peer-reviewed process for estimating
the human health impacts attributable to poor air quality. As a first step , the Agency identifies
the health endpoints to quantify by consulting the Integrated Science Assessment, a systematic
review of the clinical, toxicological and epidemiological literature. This document is reviewed by
the independent Clean Air Scientific Advisory Committee. The Agency quantifies those
endpoints classified by the ISA as being causal or likely causal, selecting concentration -response
relationships from studies that satisfy a host of criteria—most notably whether the population
characteristics match those affected by the regulation, thus excluding studies performed outside
of the U.S. and Canada. The Agency’s approach for characterizing the benefits of air quality
changes has been extensively reviewed by independent scientific bodies including the National
Academies of Sciences and the U.S. EPA Science Advisory Board.
Consistent with the findings of the PM ISA, which found that the log-linear no-threshold model
best estimates PM-related premature death, the Agency quantifies PM impacts resulting from
changes in ambient concentrations at all levels.

4.2

LEGAL COMMENTS REGARDING ENDANGERMENT FINDING
AND RATIONAL BASIS ANALYSIS

Comment 4.2-1: Multiple commenters submitted comments on whether EPA is required to do
an endangerment finding for this rule. Some commenters supported EPA’s view that section
111(b) only requires an endangerment finding when a source category is listed and that no
additional endangerment finding is needed to set standards for additional pollutants emitted from
a listed source category. Other commenters argued that EPA must make a pollutant-specific
endangerment finding for any pollutant that it regulates under section 111(b), and so here must
make a pollutant-specific endangerment finding for CO2.
Response 4.2-1: An endangerment finding is only required when EPA lists a source category
under section 111(b)(1)(A). Nothing in section 111 requires that EPA make further
endangerment findings with respect to each pollutant that it regulates under section 111(b)(1)(B).
EPA’s views on these issues are discussed in detail in Section III.A. of the preamble.
Comment 4.2-2: One commenter argued that EPA could not base this rule on the 1971 listing
decision of EGUs because the statutory listing criteria in effect at that time required only a
“cause or contribute” finding and not that the source category “contribute significantly,” as is
currently required for listing under section 111(b)(1)(A).
Response 4.2-2: EPA disagrees that the the 1971 listing decision must be redone – with a new
endangerment finding -- in light of the revision to the listing provision in section 111(b)(1)(A).
Nothing in the statute requires that. Source categories that were listed under the previous listing
provisions continue to be listed.
Comment 4.2-3: Some commenters state that Section 111(b)(1)(A) establishes that the purpose
of the NSPS program is to regulate and reduce emissions of significant "air pollution" that
4-27

"endanger public health or welfare" and that EPA cannot reasonably interpret Section 111 such
that the endangerment determination is separated from the subject of regulation. These
commenters state that EPA’s interpretation that it does not need to do an endangerment finding
means there is no assurance that regulations would serve to mitigate an actual and significant
endangerment of public health or welfare, and thus this interpretation of Section 111(b)(1)(A)
would allow EPA to impose standards for air pollutants that do not contribute significantly to an
endangerment of public health or welfare. Similarly, other commenters state that EPA’s
interpretation would allow it to regulate any air pollutant, no matter how insignificant, from a
source category that was listed based on significant contributions of other pollutants.
Response 4.2-3: These comments ignore the role that EPA’s rationale basis analysis plays in
deciding what pollutants will be regulated under section 111(b)(1)(B). As discussed in Section
III.A of the preamble, EPA has applied the rational basis analysis in the past to conclude that it
would not regulate certain pollutants that are emitted from the source category. By considering
whether there is a rational basis to regulate a given pollutant from a listed source category, EPA
ensures that it regulates pollutants that warrant regulation.
Comment 4.2-4: Commenters stated that EPA’s interpretation that it should apply a rational
basis analysis when it decides what pollutants to regulate under section 111(b)(1)(B) is not a
reasonable interpretation because EPA’s rational basis analysis considers the same issues as an
endangerment finding under section 111(b)(1)(A): the health and welfare impacts of the pollutant
and the amount of the pollutant emitted by the source category. EPA should instead conduct an
endangerment finding.
Response 4.2-4: EPA disagrees with this comment. EPA’s interpretation to look to the same
factors as an endangerment finding is both true to the statute and common sense. First, EPA
believes that looking to the same considerations that Congress identified for the endangerment
finding in section 111(b)(1)(A) when it applies the rational basis analysis under section
111(b)(1)(B) supports, rather than contradicts, the reasonableness of EPA’s approach. Second,
EPA’s view is that the harm imposed by a pollutant and the amount of emissions of a pollutant
are common-sense considerations in determining which pollutants should be regulated. Finally,
commenters’ position here is internally inconsistent in that their criticism of EPA’s approach is
that it is too similar to the analysis that commenters contend is required.
Comment 4.2-5: Multiple commenters submitted comments on whether EPA has a rational basis
for regulating CO2 emissions from power plants under section 111(b). Some commenters
supported the conclusion that the information and discussion in the record provided a rational
basis for EPA to regulate CO2 emissions from power plants under section 111(b). Other
commenters state that EPA has not shown that it has a rational basis.
Response 4.2-5: As discussed in detail in Section III.A. of the preamble, EPA has concluded that
it has a rational basis for regulating CO2 emissions from EGUs under section 111. The
information and analysis in Sections II.A and B. of the preamble, as well as EPA’s responses to
comments and elsewhere in the record, demonstrate that CO2 endangers human health and
welfare and that EGU CO2 emissions contribute significantly to CO2 levels in the atmosphere.

4-28

Comment 4.2-6: Some commenters stated that EPA cannot rationally find that CO2 emissions
from the new power plants that will be covered by the new source rule contributes significantly
to emissions of CO2 that endanger public health and welfare because EPA expects few, if any,
such facilities to be built. Similarly, some commenters stated that EPA cannot consider the CO2
emissions from existing sources that would be regulated under section 111(d) in deciding
whether to regulate new sources under section 111(b). They contend that, even if it is rational to
regulate existing sources, EPA must show that it is rational to regulate new sources. Finally,
some commenters stated that any endangerment finding has to be based on the significant
contribution of sources that will be regulated under section 111(b) and cannot include the
contribution of existing sources that would be regulated under section 111(d).
Response 4.2-6: These comments mistakenly focus on the emissions and resulting
endangerment arising only from sources that will become subject to the new source standard.
EPA disagrees that the proper analysis under either a rational basis analysis or for an
endangerment finding is based on subsets of the sources in the source category. First, the
rational basis analysis is appropriately done with respect to the source category as a whole,
because the issue is whether it is rational to regulate the emissions of a given pollutant from that
source category and that encompasses both regulation of new sources under 111(b) and of
existing sources under 111(d). It is important to remember that the pollutants regulated under
section 111(b) limits the scope of what pollutants can be regulated under 111(d), see section
111(d)(1)(A)(ii), and thus the consideration of what pollutants should be regulated under section
111(b)(1)(B) also implicates what pollutants can be regulated under section 111(d). Further,
where (as here, with CO2) the pollutant at issue is not a criteria pollutant or a HAP, regulation
under section 111(d) is where emissions from existing sources will be covered. For these
reasons, EPA reasonably considers emissions from both new and existing sources in determining
what pollutants should be regulated under section 111 from listed source categories. Since
endangerment determinations under section 111 (b) must be made on the basis of the source
category (“[h]e shall include a category of sources in such list if in his judgment it causes, or
contributes to, air pollution which may reasonably be anticipated to endanger public health or
welfare”), it is rational to do so.
With respect to an endangerment finding (if one were required here), Section 111(b)(1)(A)
clearly requires that the finding will be made with respect to the source category as a whole and
says nothing that would support the argument that limits the analysis to new sources.
Comment 4.2-7: Multiple commenters submitted comments on whether, if an endangerment
finding is necessary, the information and discussion in the record demonstrate that power plants
contribute significantly to emissions of CO2 that endangers public health and welfare. Some
commenters supported EPA’s view that, if an endangerment finding is necessary, the information
and discussion in the record provide a basis for EPA to conclude that emissions of CO2 endanger
public health and welfare. Other commenters assert that EPA’s 2009 endangerment finding
cannot be used as the endangerment finding here and that EPA’s information and analysis here is
not adequate to support an endangerment finding. Further, some commenters argue that EPA
cannot rely on the 2009 endangerment finding for the cause-or-contributes-significantly
component of the endangerment finding because it was made with respect to motor vehicle
emissions and because the relevant CAA provision at issue there (CAA section 202(a)) requires
no finding of significance.
4-29

Response 4.2-7: EPA’s explanation of the basis for an endangerment finding, if required, is in
Section III of the preamble. The information and analysis supporting the conclusions that CO2
endangers public health and welfare, and that EGUs significantly contribute to that
endangerment, are discussed in Section II of the preamble. The arguments that, if an
endangerment finding is necessary here, EPA cannot use the 2009 endangerment finding as the
necessary finding are based on a misunderstanding of the role of the 2009 endangerment finding
in EPA’s analysis in this rule. EPA is not saying that the 2009 finding is the requisite finding
here. Instead, if an endangerment finding is required, then the information and analysis in that
finding, coupled with the more recent information that confirms and expands on that finding,
support a finding that CO2 endangers human health and welfare. Finally, the comments that state
that the 2009 endangerment finding cannot support the “cause-or-contribute significantly”
finding here attack a position that EPA is not taking. EPA’s discussion of the 2009
endangerment finding relates to the issue of whether CO2 endangers public health and welfare,
not whether EGUs contribute significantly to CO2 emissions. In short, as discussed in detail in
Section III.A. of the preamble, if EPA is required to make an endangerment finding here, the
information and analysis support both that CO2 endangers human health and welfare and EGU
CO2 emissions contribute significantly to CO2 levels in the atmosphere.
Comment 4.2-8: Some commenters stated that EPA cannot determine whether a source category
meets the “significantly contributes” component of an endangerment finding without
determining a minimum threshold for what is “significant.” Other commentors agreed with EPA
that the agency can determine that a given contribution is significant without determining the
smallest contribution that would be considered significant.
Response 4.2-8: EPA’s view is that it does not need to set a minimum threshold for what is
“significant” when it considers the “contributes significantly” component in making an
endangerment finding. Here, if an endangerment finding is required, the information
demonstrates that the emissions from EGUs constitutes a “significant contribution.”
Comment 4.2-9: Commenters state that EPA’s logic is flawed in that the agency assumes that
because a listed source category has been found to contribute significantly to emissions o f one
pollutant, that source category must cause or contribute significantly to an endangerment of
public health and welfare for a different pollutants.
Response 4.2-9: These comments misstate EPA’s position. EPA’s view is not that a source
category that was listed under section 111(b)(1)(A) for one pollutant necessarily causes or
significantly contributes to an endangerment of public health and welfare for a different
pollutant. EPA’s view is that once a source category is listed under section 111(b)(1)(A), the
CAA requires no further endangerment finding for that source category when EPA sets standards
for the pollutants that it emits under section 111. Instead, EPA determines which pollutants
emitted by the listed source category should be regulated based on a rational basis analysis.

4.3

CALCULATING "CONTRIBUTION"

Comment 4.3-1: Commenter 9682 states the NSPS provisions must be interpreted in pari
materia with the NAAQS standard setting provisions and, in particular, the NSPS provisions,
like the NAAQS provisions, must be construed as limited to preventing exposures to localized
4-30

pollutants-that is, pollutants whose concentrations vary by locality as opposed to dispersing
throughout the globe as do GHG emissions.
Response 4.3-1: The comment is misplaced. There is no reference to the NAAQS in section
111 (a), section 111 (a) and (b) set technology-based NSPS unlike the purely health-based
ambient standards, there is no language in the text of section 111 (a) limiting consideration to
localized air pollution effects (and of course, PM, PM precursors, and ozone precursors are all
pollutants with regional as well as local effects, and are routinely the subject of NSPS), and
section 111 (a) and (b) do not limit standards under these provisions to criteria pollutants. EPA,
in fact, has routinely issued section 111 (b) standards of performance for non-criteria pollutants
since the inception of implementation of these provisions. See, e.g. 40 CFR part 60.192
(standards for fluorides emitted by primary aluminum facilities, first issued in June 1980).
Comment 4.3-2: Commenter 9514 states the courts (including Coal. for Responsible Regulation,
Inc. v. EPA, 684 F.3d 102, 116-27 (D.C. Cir. 2012) and Util. Air Regulatory Grp. v. EPA, 134 S.
Ct. 418, 2013 U.S. LEXIS 7380 (Oct. 15, 2013)) found that the Endangerment Finding was
procedurally sound, consistent with Supreme Court case law, and amply supported by the
administrative record, observing that "[t]he body of scientific evidence marshaled by EPA in
support of the Endangerment Finding is substantial."
Response 4.3-2: EPA agrees with this comment, and notes that this decisive judicial precedent
has not dissuaded commenters from presenting identical or nearly identical arguments to EPA
again in this rulemaking. See section 4.1 above.
Comment 4.3-3: Commenter 10025 supplies documents that address the effects of climate
change in humans, especially in children.
Commenter 9589 states it is our obligation to put people before polluters and protect our children
and future generations from the effects of climate change. This historic standard is a game
changer for the climate and health in America and around the world.
Response 4.3-3: EPA acknowledges these commenters’ general support for the proposed
standards.
Comment 4.3-4: Commenter 9514 states the Regulatory Impact Analysis (RIA) provides an
overview of the pressing threats posed by greenhouse gas emissions and ably canvasses the
dangers that the New Source Performance Review (NSPS) must combat. Commenter provided
extended discussion of the harms associated with climate change, and provided extensive
citations; discussed new research, reports and assessments showing the increasing severity of
harm.
Response 4.3-4: This comment is consistent with EPA’s own reading and interpretation of the
scientific evidence. See section 4.2 above.
Comment 4.3-5: Commenter 10025 states EPA should explain the specific goal of the proposed
rule, what it means for our children and posterity in terms of global heating, and how much the
economic interests of the fossil fuel industry have played a role in this proposed r ule versus
atmospheric health. Commenter calls for EPA to disclose to the public how this rule compares to
4-31

the science-based emission reduction levels needed from power plants. Commenter adds nothing
in this proposed rule says that the rule is aimed at adequately reducing carbon emissions from
these sources at levels that are in line with a national plan to achieve atmospheric health and
climate recovery according to the best available science.
Commenter 5605 states agreement that carbon dioxide and other global warming pollutants that
endanger public health and welfare should be limited and that we have an obligation to protect
our children and future generations from these effects by addressing these causes and impacts,
including carbon pollution.
Response 4.3-5: EPA is addressing GHG emissions. The agency has issued standards for GHG
emissions from all cars, SUVs, light trucks, pickups and vans, vocational vehicles, and
combination tractors, in some instances up to model year 2025. The agency recently proposed
further GHG standards for heavy duty vehicles and engines (including trailers attached to
combination tractors) through the 2027 model year. In this proceeding EPA is addressing CO2
emissions from the largest stationary source emitting source category. In doing s o, of course,
EPA can only act within the limits of its delegated statutory authority, and so lacks legal
authority to adopt some of the more wide-ranging actions urged by this and other of the
commenters on this issue.
Comment 4.3-6: Commenter 10119- states EPA must "tak[e] into account . . . any nonair quality
health and environmental impact and energy requirements" of the standard11 stating that EPA did
not undertake such an analysis in the 2012 Proposed Rule and in the new Proposed Rule.
Commenter adds hydraulic fracturing and other unconventional gas drilling and production
techniques have a wide array of adverse effects on public health and the environment, citing
additional studies have been published since the 2012 Proposed Rule comments. Commenters
10087, 10087 state using CCS for EOR would result in a net increase in CO2 emissions.
Commenter 10087 states that using the CO2 captured from CCS for EOR will likely lead to an
increase in net CO2 emissions because the CO2 from the power plant may be stored, but only by
aiding in the production of oil.12 Much of the oil will be turned into fuel with resulting CO2
emissions, and thus, the CO2 emissions from the oil would be greater than the stored CO2
emissions.
Commenter 9396 states EPA's assertion that non-air impacts of Section 111(b) NSPS standards
are not within the scope of the rulemaking seems at odds with legal standards. In Sierra Club, the
court noted that "crucial issues such as waste disposal...which may continue to limit overall
acceptability of this technology remain to be answered."13 This, coupled with the direct
description of Congressional intent that the Administrator must "consider non-air quality health
and environmental impacts in making [a 111 NSPS] determination" 14 suggests that waste stream
analysis is indeed a required component of an NSPS determination. As part of the waste stream
11

CAA Section 111(a)(1); 42 U.S.C. Section 7411(a)(1).
12 William Yeatman, EPA's Carbon Pollution Standard May INCREASE CO2 Emissions!, GlobalWarming.org,
Sept. 25, 2013, http://www.globalwarming.org/2013/09/25/epas-carbon-pollution-standard-may-increase-co2emissions.
13 Sierra Club v. Costle, 657 F.2d at 341, fn157.
14 H.R. 95-294. 95th Cong., 2 Leg. His.2478 (1977).

4-32

evaluation, EPA must consider where disposal is appropriate and not cost prohibitive. The record
does not support that this evaluation was adequately undertaken. Based on the SAB letter to the
Administrator, while the SAB has deferred to EPA's erroneous assertion that waste streams are
not subject to analysis, the SAB has concluded that the concern regarding "unintended multimedia consequences" warrant a review of the underlying science by the National Research
Council or the SAB.15 This assertion should serve as a clear warning to EPA that it has failed to
sufficiently review and establish the efficacy of sequestration for the purposes of establishing
CCS as NSPS.
Commenters 9201 states there are many other environmental effects related to the proposed
NSPS that EPA fails to examine. Several include: Groundwater quantity and quality from
increased shale gas development; Air Quality impacts from shale development: Impacts from
building out natural gas infrastructure to meet power sector demand for natural gas; Impacts of
new CO2 pipelines; Water Use; and Impacts on Wildlife.
Response 4.3-6: The EPA agrees, emphatically, that it must consider nonair quality health and
environmental impact as part of the process of assessing whether a system of emission reduction
is “best” and is “adequately demonstrated”. EPA’s consideration of these issues with respect to
the standard of performance for new coal-fired steam electric generating units is set out in
preamble section V.O. We further disagree with the commenters’ statement that EPA failed to
consider either the efficacy or cost of disposal of captured CO2. See preamble section V.I.5 and
V.N. The comment urging the EPA to account from the rule’s impacts on hydraulic fracturing
are misplaced. Natural gas use is increasing, but is doing so for reasons unrelated to this
rulemaking. See RIA chapter 4. The commenter also evidently assumes CH4 emissions
associated with natural gas production and transport are uncontrollable, which the EPA does not
believe to be the case.
The commenters also do not fully represent the facts regarding SAB consideration of issues
relating to geologic sequestration. In developing the rules for injection of CO2 for permanent
disposal – the Class VI rules -- EPA engaged with the SAB, providing detailed information on
key issues relating to geologic sequestration – including monitoring schemes; methods to predict
and verify capacity, injectivity, and effectiveness of subsurface CO2 storage; and characterization
and management of risks associated with plume migration and pressure increases in the
subsurface.16 In addition, EPA developed a peer reviewed Vulnerability Evaluation Framework,
which served as a technical support document for both the Class VI and Subpart RR rules. 17 This
extensive interaction preceded the present rulemaking.
In the section 111 (b) rulemaking here, the SAB Work Group, in a letter endorsed by the full
SAB Committee, found that “while the scientific and technical basis for carbon storage

15

Letter to Regina McCarthy from EPA Science Advisory Board. January 29, 2014. EPA-SAB-14-03 p. 3.
16 See: http://yosemite.epa.gov/sab/sabproduct.nsf/0/AD09B42B75D9E36D85257704004882CF?OpenDocument.
17 See: http://www.epa.gov/climatechange/Downloads/ghgemissions/VEF-Technical_Document_072408.pdf.

4-33

provisions is new and emerging science, the agency is using the best available science and has
conducted peer review at a level required by agency guidance.” 18
Comment 4.3-7: Commenter 10951 states the policies EPA is promoting have been proven to
drive up electricity prices everywhere those policies have been tested. These economic impacts
will in turn impair public health and welfare because the effect will be felt by those least able to
afford them. Dollars spent on higher energy bills will in turn crowd out dollars that would
otherwise be available to pay for good nutrition and health care. Jobs lost because of higher
electricity costs means less money for health insurance. It is a truism that wealth equals health,
and it is equally true that health will deteriorate as electricity costs rise, particularly for lower
income people and those living on fixed incomes. Commenter continues That in a report
prepared for the American Coalition for Clean Coal Electricity entitled Energy Cost Impacts on
American Families, 2001-2012, attorney and economist Eugene Trisko reported that electricity
cost increases fall disproportionately on those least able to afford them. These effects were als o
described in the attached report, The Social Costs of Carbon No, the Social Benefits of Carbon
and are graphically displayed in the following chart. Further, Commenter 10951-6171 states
energy poverty is the great global challenge of the future. 3.5 billion people, including 1.2 billion
children, do not have access to an adequate supply of energy.8 There can be no improvement in
living standards in developing societies unless energy poverty is solved. Every single one of the
United Nations Millennium Development Goals requires access to electricity as a necessary
perquisite.9 (Global Energy Institute, 2008). The U.N. Secretary General has called energy the
golden thread that connects economic growth, increased social equity and an environment that
allows the world to thrive.19
As discussed below, every ten-fold increase in electricity is linked to a stunning 10-year increase
in life spans, and coal is the only sustainable fuel with the scale to meet the primary energy needs
of the world's rising populations and economies. As a result, coal is the world's fastest growing
fuel, and coal use expanded nearly 50 percent this past decade. The example the United States
sets for the rest of the world thus cannot be that electricity should be made scarcer and more
expensive. The world needs more, not less, electricity, and coal remains the leading global and
domestic electricity resource.
Response 4.3-7: The EPA is required to consider energy requirements as part of its
determinations under section 111 (a), and has done so here. See preamble section V.O.3. As
found there, and in RIA chapter 4, the standards of performance here are not anticipated to have
significant impact on energy prices because .in the baseline, all projected unplanned capacity -projected capacity additions that are not under construction. -- affected by this standard during
the analysis period would already be compliant with the rule’s requirements (e.g., natural gas
combined cycle units, low capacity factor natural gas combustion turbines, and small amounts o f
coal-fired units with carbon capture and storage (CCS) supported by Federal and State funding).

18 Memorandum of Jan. 7, 2014, from SAB Work Group Chair to Members of the Chartered SAB and SAB
Liaisons, p. 3. The letter was subsequently endorsed by the full SAB. Work Group Letter of Jan. 24, 2014, as edited
by the full Committee.
19
http://www.un.org/News/Press/docs/2012/sgsm14242.doc.htm

4-34

4.4

SOCIAL COST OF CARBON

4.4.1 Methodology
Comment 4.4-1: Commenter 9681 states the IWG's use of Integrated Assessment Models
(IAMs) reflects the best available, peer-reviewed science to tally the benefits and costs of
specific regulations with impacts on carbon dioxide emissions. The IAMs include benefits and
costs that have been quantified to date. Other modeling assumptions represent reasonable choices
on the part of the IWG. For example, cutting off the analysis after around 300 years simply
reflects the fact that any benefits and costs thereafter are de minimis given the strong impact
discounting has on reducing the value of distant effects in the SCC. The IWG should indeed
periodically update its choice of socio-economic scenarios to reflect major changes in economic
growth projections and other factors influencing emissions trajectories. However, the long-term
emissions impact of any one particular energy efficiency or other carbon reduction standard is
only marginal to the trajectory of socio-economic scenarios, minimizing any risk of double
counting the regulatory benefits.
Commenter 9194 states multiple experts indicate that the modeling used does not provide a
sufficient range of accuracy for policy-making purposes. OMB and the IWG failed to use the
latest science which indicates that the equilibrium climate sensitivity (ECS) distributions used by
the modelers caused the SCC to be over-stated. The models employed and the assumptions used
were not subjected to peer review nor were the 2010 and 2013 SCC estimates developed by the
IWG.
Commenter 10501 states that furthermore, the SCC calculation assumes there is only one, and as
yet unproven, solution to all potential problems that a warming planet might cause, i.e. CO2
emissions control. Clearly there may be other high-confidence approaches to mitigating specific
problems that global warming may cause, such as building sea walls in the US to prevent coastal
flooding from sea level rise. Such alternatives may be much less costly than the impact of
unilateral USA CO2 emissions control regulations that have very little chance of success, given
the current world political environment and lack of a world-wide CO2 emissions control
agreement. China, India and other rapidly developing countries will dominate world -wide CO2
emissions for the next several hundred years, rendering unilateral USA CO2 emissions control
totally ineffective in avoiding the damages purportedly represented by the SCC. Therefore, the
SCC cannot in any rational, credible argument be considered to be an appropriate economic
metric that can be used to justify the proposed new EPA rule. The rational decision process
normally used in US government and industry to weigh cost and benefits of several competing
alternative solutions to anticipated problems appears to be totally absent from the SCC
methodology. In our typical research team member's more than 50 years of experience in
addressing complex problems and critical decisions with life or death consequences for manned
space mission crew members, we can only view the IWG's rationale and process for computing
SCC as totally lacking in common sense or any true scientific value. It is unthinkable in our
experience to expose the entire population of the USA to such economic risks from electrical
power cost increases expected from this new EPA rule, without a more certain outcome
regarding avoidance of SCC damages used to justify the regulation.

4-35

Commenter 9401-4769 continues history provides an important lesson on how increased energy
costs have both a direct and indirect impact on reduced employment and GHG leakage. Figure 2
details the relationship between rising natural gas prices and falling employment in the
manufacturing sector. According to EIA, electricity prices rose 210 percent from 1999 to 2008.
During the time period of 2000 to 2008, the manufacturing sector shut down an estimated 54,905
facilities, decimating the economy of thousands of communities across the country. During that
same time period, net manufactured goods imports grew by 46 percent, a clear indication of loss
of US manufacturing competitiveness. A key point that needs to be emphasized is that if
electricity prices rise because of the EPA GHG rule, manufacturing employment and GHG
leakage would have the same effect as higher natural gas prices.
Commenter 3593-7453 states EPA fully recognizes that increased regulatory costs fall more
heavily on minorities, on women and especially mothers who are single parents, and on the
elderly, and if they did not before receiving this comment, they do now. EPA acknowledges its
responsibilities with regard to equal protection of citizens. These considerations were ignored for
the reason that impacting climate which occurs under no scenario of this rule was not the
objective. Indeed, research shows that such regulations will create massive job losses and a
major loss of GDP. These impacts -- both those flowing directly from the higher electricity rates
and those flowing indirectly therefrom, and flowing directly or indirectly instead from the
industrial policy which the proposed rule represents -- have human and environmental
consequences, also not considered. The only rational reason for declining to consider this is that
this outcome is the objective, which is to say, it was not attained by consideration but
predetermined. This is particularly troublesome given the failure of the regulation to achieve its
stated goal, by the admission of senior EPA officials.
Commenter 9650 states EPA uses a flawed methodology in calculating the social cost of carbon
("SCC") valuation that inflates the costs associated with climate change. Commenter 502 states
there are too many uncertainties in both the integrated assessment models used to derive SCC
values to derive values to be useful in regulatory decision making.
Commenter 9650 states the integrated assessment models ("IAMs") constructed to estimate the
SCC have flaws (e.g. the discount rate is arbitrary; the model's descriptions of climate change
impact are ad hoc, and no theoretical or empirical foundation).
Commenter 9685 states emissions input errors: concentrations generated for the IAMs are
seriously flawed. The models are based on emissions scenarios that have not been validated. The
SCC report then raises a major transparency issue: "we decided to rely on the recent Stanford
Energy Modeling Forum exercise, EMF-22. EMF-22 uses ten well-recognized models to
evaluate substantial, coordinated global action to meet specific stabilization targets." A key
advantage of relying on these data is that GDP, population, and emission trajectories are
internally consistent for each model and scenario evaluated. The problem with this approach: the
Stanford Energy Modeling Forum is not a public process. EPA needs to assure that any
participation by government employees conforms to the DQA requirements. EPA needs to: a.
Identify the EPA employees who participated in the EMF 22 exercise, b. Post the information
provided to the EMF 22 exercise, and c. Provide the results of the exercise related to EPA input,
especially for EPA analysis.

4-36

Commenter 9685 states SCC recognizes the problems with the IPCC standard SRES scenarios as
"extreme outliers" but then the SCC inexplicably chooses to use even less reliable scenarios but,
there is no detailed justification for the selection of the models. As can be seen in the table
above, the divergence in the emissions begins in 2030. And by 2050 the divergence is substantial
and it grows even larger by 2100. The SCC makes the assumption that the scenarios have equal
probability. EPA needs to explain this rationale.
Commenter 9685 states that a major question is: why were the four scenarios selected? Since the
EMF is a secret process, there is no documentation that describes the rationale for the "modelers'
judgment". Nor do we know who the "modelers" were or whether their "judgment" has ever been
validated by comparing their projections with data. The IAMs use these scenarios under the
assumption that the scenarios are equally likely, a highly suspect assumption.
Commenter 9685 states the IWG attempted to justify its selection of emissions scenarios by a
casual "comparison" to the EIA AEO 2009. However, the selected emissions scenarios are well
above EIA AEO2013/IEO2013 projections. For some reason the IWG chose to ignore the most
recent EIA analysis that has projections to 2040. The average of the emissions scenarios is
roughly 20% above the EIA forecast. The cumulative difference for the 10 years 2010-2020 is 6
Gigatons CO2 for the U.S. alone and the cumulative difference grows to over 50 Gigatons CO2
by 2040. Since the "damages" are attributed to the atmospheric concentrations (related to
cumulative emissions) the distortion is egregious. The same problem can be found in comparing
the scenarios for global emissions. EPA must reject the selected emissions scenarios as "extreme
outliers" (the same standard that was used to reject the IPCC SRES scenarios). The scenarios
lead to an significant overestimate of the emissions that are input to the model that lead to a
significant and growing error in the atmospheric concentration of greenhouse gases. Commenter
concludes that EPA should discard the SCC 2010 and SCC2013 results on this basis alone.
Commenter 9685-7213 states damage function errors: catastrophic estimates are qualitative
speculation based on "expert elicitation". The SCC attributes quantitative certainty to qualitative
speculation that has a major influence on the Damage Function results. As is obvious in the table
below the range of impacts is highly questionable in terms of timing and impact. OMB should
have rejected this approach. EPA should expose the details of "expert elicitation" so that the
public can provide comment. The public needs to know: 1) Who were the "experts" "elicited"?
and 2) What was the specific data provided? 3. Was all data given the same weight?
Commenter 9685 states model errors: if the AR4 climate models are wrong, the IAMs are wrong.
Recent House testimony shows the failure of the models used to define the IAM equilibrium
climate sensitivity "So it's clear that the IPCC modeling community has known about the
modeling failure (divergence between the model predictions and real data) for over 20 years. The
SCC IAMs were constructed on the backs of these same flawed models. EPA must take
immediate action to correct this abuse of the taxpayer: not one more dime should be spent on the
US models used in the IPCC AR5. Those responsible should be held accountable." The
independent analysis conducted by Dr. Christy would have been difficult, if not impossible, to
document without the contribution to transparency by the Royal Netherlands Meteorological
Institute (KNMI) Climate Explorer (climexp.knmi.nl). The KNMI provided a comprehensive
collection of climatic data sets and analysis tools. All of the IPCC AR5 model runs are available
to the public with a user-friendly interface. Why didn't our own government do the same? What
4-37

did the modelers have to hide? EPA needs to follow the example of the Dutch government and
implement a similar requirement for all analysis involved in climate policy, including posting on
an archive website: 1. Documentation of all models used by EPA, 2. An archive of all versions of
models used for specific policy or regulations, and 3. An archive of all model runs used for any
policy analysis including the input variables. The IQA guidelines provide for this, but agencies
appear to routinely disregard their responsibility.
Commenter 9685 stats the IAMs are nested and not transparent and compound major errors at
every stage. Each IAM progresses through a series of cumulative, compounding errors that result
in "costs" that bear no resemblance to reality. The emissions scenarios are outliers compared to
actual data and the EIA projections. The scenarios should have been rejected based on the same
rationale that the IPCC SRES scenarios were rejected by the IWG.
Commenter 9685 continues these flawed scenarios are then used in a Monte Carlo analysis that
cloaks the errors in an opaque process. The pseudo sophistication implied by using a Monte
Carlo analysis is at best misleading. If the input data is wrong, the output distributions are wrong
(this is Garbage In, Garbage Out). As discussed below, the greenhouse gas emissions data are
wrong. Therefore the resulting greenhouse gas concentrations are seriously flawed. The climate
sensitivity related to the flawed concentrations based on flawed emissions is based on the failed
IPCC AR4 models that are demonstrated failures. The Equilibrium Climate Sensitivity (ECS)
probability density functions (PDFs) developed using the AR4 models are fatally flawed. The
resulting temperatures based on the flawed emissions driving flawed driving flawed
concentrations driving flawed climate sensitivity are then input to "damage" calculations based
on fictional un-validated formulations. These results are even more egregiously flawed.
Commenter 10091 states there is a misleading disconnect between climate change and the SCC
in the IAMs. It is generally acknowledged, the results from IAMs are highly sensitive not only to
the model input parameters but also to how the models have been developed and what processes
they try to include. One prominent economist, Robert Pindyck of M.I.T. recently wrote (Pindyck,
2013) that the sensitivity of the IAMs to these factors renders them useless in a policymaking
environment: Given all of the effort that has gone into developing and using IAMs, have they
helped us resolve the wide disagreement over the size of the SCC? Is the U.S. government
estimate of $21 per ton (or the updated estimate of $33 per ton) a reliable or otherwise useful
number? What have these IAMs (and related models) told us? I will argue that the answer is very
little. As I discuss below, the models are so deeply flawed as to be close to useless as tools for
policy analysis. Worse yet, precision that is simply illusory, and can be highly misleading. [A]n
IAM-based analysis suggests a level of knowledge and precision that is nonexistent, and allows
the modeler to obtain almost any desired result because key inputs can be chosen arbitrarily.
Nevertheless, the EPA incorporates the IWG2013 determinations of the SCC into this proposed
regulation ill-advisedly so in our opinion.
Commenter 10091 continues stating if one considers the following: the social cost of carbon
should reflect the relative impact on future society that human-induced climate change from
greenhouse gas emissions would impose. In this way, we (policymakers and regular citizens) can
decide how much (if at all) we are willing to pay currently to reduce the costs to future society. It
would seem logical that we would probably be more willing to sacrifice more now if we knew
4-38

that future society would be impoverished and suffer from extreme climate change than we
would be willing to sacrifice if we knew that future society would be very well off and be subject
to more moderate climate change. We would expect that the value of the social cost of carbon
would reflect the difference between these two hypothetical future worlds - the SCC should be
far greater in an impoverished future facing a high degree of climate change than an affluent
future with less climate change. But if you thought this, you would be logically correct, but
wrong. Instead, the IAMs as run by the IWG2013 produce nearly the opposite result the SCC is
far lower in the less affluent/high climate change future than it is in the more affluent/low
climate change future. Such a result is not only counterintuitive but misleading. We illustrate this
illogical and impractical result using the DICE 2010 model (hereafter just DICE) used by the
IWG2013 (although the PAGE and the FUND models generally show the same behavior). The
DICE model was installed and run at the Heritage Foundation by Kevin Dayaratna and David
Kreutzer using the same model set up and emissions scenarios as prescribed by the IWG2013.
The projections of future temperature change (and sea level rise, used later in the Comment)
were provided to us by the Heritage Foundation. The figure below shows the projections of the
future change in the earth's average surface temperature for the years 2000-2300 produced by
DICE from the five emissions scenarios employed by the IWG2013. The numerical values on the
right-hand side of the illustration are the values for the social cost of carbon associated with the
temperature change resulting from each emissions scenario (the SCC is reported for the year
2020 using constant $2007 and assuming a 3% discount rate numbers taken directly from Table
A3 of the IWG2013 report). The temperature change can be considered a good proxy for the
magnitude of the overall climate change impacts. Notice in the figure above that the value for the
SCC shows little (if any) correspondence to the magnitude of climate change. The MERGE
scenario produces the greatest climate change and yet has the smallest SCC associated with it.
The "5th Scenario" is a scenario that attempts to keep the effective concentration of atmospheric
carbon dioxide at 550 ppm (far lower than the other scenarios) has a SCC that is more than 20%
greater than the MERGE scenario. The global temperature change by the year 2300 in the
MERGE scenario is 9dC while in the "5th Scenario" it is only 3dC. The highest SCC is from the
IMAGE scenario a scenario with a mid-range climate change. All of this makes absolutely no
logical sense - and confuses the user.
Commenter 10091 states if the SCC bears little correspondence to the magnitude of future
human-caused climate change, then what does it represent? The figure below provides some
insight. When comparing the future GDP to the SCC, we see, generally, that the scenarios with
the higher future GDP (most affluent future society) have the higher SCC values, while the
futures with lower GDP (less affluent society) have, generally, lower SCC values. Combining the
results from the two figures above thus illustrates the absurdities in the IWG's use of the DICE
model. The scenario with the richest future society and a modest amount of climate change
(IMAGE) has the highest value of the SCC associated with it, while the scenario with the poorest
future society and the greatest degree of climate change (MERGE) has the lowest value of the
SCC. A logical, thinking person would assume the opposite. While we only directly analyzed
output data from the DICE model, by comparing Tables 2 and Tables 3 from the IWG2010
report, it can be ascertained that the FUND and the PAGE models behave in a similar fashion.
Commenter 10091 states this counterintuitive result occurs because the damage functions in the
IAMs produce output in terms of a percentage decline in the GDP which is then translated into a
4-39

dollar amount (which is divided by the total carbon emissions) to produce the SCC. Thus, even a
small climate change-induced percentage decline in a high GDP future yields greater dollar
damages (i.e., higher SCC) than a much greater climate change-induced GDP percentage decline
in a low GDP future. In principle, the way to handle this situation is by allowing the discount
rate to change over time. In other words, the richer we think people will be in the future (say the
year 2100), the higher the discount rate we should apply to damages (measured in 2100 dollars)
they suffer from climate change, in order to decide how much we should be prepared to sacrifice
today on their behalf. Until (if ever) the current situation is properly rectified, the IWG's
determination of the SCC is not fit for use in the federal regulatory process, such as this EPA
regulation, as it is deceitful and misleading.
Commenter 9681 sated that updates to the SCC should also consider other models that are
similarly peer reviewed and based on the state of the art of climate-economic modeling. One
such model is Climate and Regional Economics of Development (CRED); another is the World
Bank's ENVironmental Impact and Sustainability Applied General Equilibrium (ENVISAGE)
model. CRED borrows its fundamental structure from William Nordhaus's DICE and RICE
models but also offers significant changes. For one, it uses updated damage functions and
Marginal Abatement Cost Curves (MACC). Moreover, it uses different global equity weights,
and uses additional state-of-the-art methodologies. ENVISAGE represents a broader modeling
effort by the World Bank, where perhaps the largest contribution is a more detailed sectoral
breakdown, using 57 different sectors. This level of detail allows for a more detailed view of
agriculture as well as food and energy sectors that are particularly important to any climateeconomy modeling. No model fully captures the costs of climate impacts to society. In fact,
virtually all uncertainties and current omissions point to a higher SCC value. That makes it
essential to use the established IWG process, which provides for updating the SCC estimates
every two to three years in order to capture the advances in physical and social sciences that have
been incorporated into the models during the intervening period, in order to revisit both the
choice of models and the key inputs used.
Commenter 9681 continues the IWG's choice of three IAMs was fully justified but should still be
revisited in its next iteration. In its calculations of the SCC, the IWG relied on the three
Integrated Assessment Models (IAMs) available at the time, all with a long record of peerreviewed publications that link physical and economic effects: the Dynamic Integrated Model of
Climate and the Economy (DICE), the Climate Framework for Uncertainty, Negotiation, and
Distribution (FUND), and Policy Analysis of the Greenhouse Effect (PAGE). The government's
first SCC estimates, published in 2010, used the then-current versions of the models; the recent
update employed revised, peer-reviewed versions of the models but maintained the underlying
assumptions of the 2010 IWG analysis. As stated by the 2010 IWG, the main objective of [the
2010 IWG modeling] process was to develop a range of SCC values using a defensible set of
input assumptions grounded in the existing scientific and economic literatures. DICE, FUND,
and PAGE are well-established, peer-reviewed models. They represent the state-of-the-art IAMs.
Each of these models has been developed over decades of research, and has been subject to
rigorous peer review, documented in the published literature. However, updates to the SCC
should also consider other models that are similarly peer reviewed and based on the state of the
art of climate-economic modeling. One such model is Climate and Regional Economics of
Development (CRED); another is the World Bank's ENVironmental Impact and Sustainability
4-40

Applied General Equilibrium (ENVISAGE) model. CRED borrows its fundamental structure
from William Nordhaus's DICE and RICE models but also offers significant changes. For one, it
uses updated damage functions and Marginal Abatement Cost Curves (MACC). Moreover, it
uses different global equity weights, and uses additional state-of-the-art methodologies.
ENVISAGE represents a broader modeling effort by the World Bank, where perhaps the largest
contribution is a more detailed sectoral breakdown, using 57 different sectors. This level of detail
allows for a more detailed view of agriculture as well as food and energy sectors that are
particularly important to any climate-economy modeling. No model fully captures the costs of
climate impacts to society. In fact, virtually all uncertainties and current omissions point to a
higher SCC value. That makes it essential to use the established IWG process, which provides
for updating the SCC estimates every two to three years in order to capture the advances in
physical and social sciences that have been incorporated into the models during the intervening
period, in order to revisit both the choice of models and the key inputs used.
Commenter 9681 states the IWG should update its socio-economic assumptions to reflect the
latest Shared Socio-economic Pathways (SSPs). One key input is the use of socio-economic
scenarios reflected in the choice of economic growth rates and emissions trajectories. Current
IWG socio-economic and emissions scenarios were chosen from the Stanford Energy Modeling
Forum exercise, EMF-22, and consist of projections for income/consumption, population, and
emissions (CO2 and non-CO2). The IWG selected five sets of trajectories, four of which
represent business as usual (BAU) trajectories (MiniCAM, MESSAGE, IMAGE, and MERGE
models) and a fifth that represents a CO2 emissions pathway with CO2 concentrations stabilizing
at 550 ppm. Given the possibility of increases in emissions above those expressed by Business
As Usual Scenarios, a high-CO2 emissions pathway should also be considered. The assumptions
used in calculating the SCC should be updated regularly to reflect the latest thinking around
possible scenarios, reflecting the latest Shared Socio-economic Pathways (SSPs). These SSPs
represent the latest, consistent pathways, feeding, for example, into the latest IPCC report.
Commenter 9514 states the cost of a partial CCS standard, or, indeed, a full CCS standard, is
small when considering a broad view of costs that accounts for the social costs of carbon. The
Joint Environmental Commenters maintain that the social cost of carbon is much higher than the
values in the interagency rule. Commenter 10525 states the financial benefit attributed to avoided
CO2 pollution is actually greater than estimated by the task force, explaining that other elements
that should have been considered by the task force but which were not include the effects on
drought, energy, and water supply; forest fires; temperature extremes on crops and increasing
food prices; food scarcity; impacts of air pollution from enhanced smog; and the consequences of
total ecosystem losses.
Commenter 8925 states the 2300 extensions are inconsistent across variables, explaining
specifically, the land-use CO2 emissions and non-CO2 radiative forcing extensions have no
relationship to the population, GDP, and fossil and industrial CO2 emissions extensions, as
well as each other.
Commenter 8925 states the use of a 550 ppm policy emissions/socioeconomic projection is
inconsistent with the other scenarios. The current averaging of the results across
socioeconomic/emissions scenarios requires that (a) the values being averaged are from the same
4-41

statistical population, and (b) that the values are all equally likely. As for (a), given the 550 ppm
scenario's assumption of a global climate policy (targeting 550 ppm in 2100), it is difficult to
claim that it is from the same statistical population of potential futures as scenarios with no
climate policy. As for (b), the specific 550 ppm scenarios used in the analysis assume immediate
global cooperation in decarbonizing the global economy in achieving the 550 ppm target in
2100.20 This is inconsistent with current global mitigation efforts and is therefore not equally
likely to business-as-usual (BAU) futures. A similar assessment of likelihood would be
appropriate for all the socioeconomic/emissions scenarios to avoid including scenarios that can
be reasonably ruled out, or at least to recognize that some scenarios may be less likely. One
commenter 8925 states the creation of the "average" 550 ppm emissions/socioeconomic scenario
is problematic because internal consistency is lost when averaging the variables, and the
resulting set of average socioeconomic and emissions projections do not represent an internally
consistent scenario.
Commenter 9396 The SCC is based on three Integrated Assessment Models (IAMs). The SCC
ascribes a level of certainty that far exceeds what the IAMs actually deliver. Despite the
admonition that the level of uncertainty associated with the development of the SCC reduced its
relevance as a policymaking tool for the purposes of establishing cost-benefit analysis where
GHG reductions present the predominant benefit, the US EPA is proceeding with regulatory
actions that base significant regulatory burdens on industry based on GHG reductions calculated
using the SCC.
Response 4.4-1: The comments mirror those submitted to the Office of Management and
Budget’s separate comment solicitation on the SCC (78 FR 70586; November 26, 2013). As a
member of the interagency working group (IWG) on SC-CO2, EPA has carefully examined and
evaluated comments submitted to OMB’s separate solicitation. EPA has also carefully examined
and evaluated all comments received regarding SC-CO2 through this rulemaking process and
determined that the IWG responses to the comments on the OMB solicitation address the
comments on the SC-CO2 methodology, including the models used to estimate the SC-CO2.
Specifically, EPA concurs with the IWG’s response to these comments and hereby incorporates
them by reference. 21 The remainder of this section elaborates on these comments in the context
of this rulemaking.
EPA appreciates the commenters’ recommendations for potential opportunities to improve the
SC-CO2 estimates and has considered each one in the context of this rulemaking. EPA and other
members of the IWG have reconfirmed their commitment to periodic review and update of the
methodology and estimates to ensure that they continue to reflect the best available science and
economics. Furthermore, EPA has determined that the SC-CO2 methodological
recommendations require additional research, review, and public comment before it can apply
them to a rulemaking context.

20

Clarke, L., J. Edmonds, V. Krey, R. Richels, S. Rose, M. Tavoni, 2009. International climate policy architectures:
Overview of the EMF 22 International Scenarios. Energy Economics 31 (Supplement 2): S64-S81.
21
Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 6 to 11 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf .

4-42

To help synthesize the technical information and input reflected in the comments, and to add
additional rigor to the next update of the SC-CO2, the EPA and all of the other IWG members
plan to seek independent expert advice on technical opportunities to improve the SC-CO2
estimates, including the approaches suggested by commenters and summarized in this document.
Specifically, EPA and all of the other IWG members plan to ask the National Academies of
Sciences, Engineering, and Medicine (Academies) to examine the technical merits and
challenges of potential approaches to improving the SC-CO2 estimates in future updates. Input
from the Academies, informed by the comments submitted to this rulemaking and the peerreviewed literature, will help to ensure that the SC-CO2 estimates used by EPA continue to
reflect the best available science and methodologies.
After careful evaluation of the full range of comments and associated technical issues described
in section 4.4, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SCCO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in RIA
Chapter 5.22 EPA will continue to consider these comments and will share the recommendations
with the IWG as it moves forward with the Academies process.
The remainder of this section provides more detailed responses to the comments and
recommendations.
Choice of IAMs used to estimate the SCC.
EPA agrees with the commenters who regard the selection of the three IAMs—DICE, FUND,
and PAGE—as the most appropriate for the purpose of estimating the SCC. EPA and all of the
other IWG members made this determination when they began developing the SCC estimates in
2009-2010. DICE, FUND, and PAGE are the most widely used and widely cited models in the
economic literature that link physical impacts to economic damages for the purposes of
estimating the SC-CO2. As stated in the 2010 TSD:
These models are frequently cited in the peer-reviewed literature and used in the IPCC
assessment. … These models are useful because they combine climate processes,
economic growth, and feedbacks between the climate and the global economy into a
single modeling framework. … Other IAMs may better reflect the complexity of the
science in their modeling frameworks but do not link physical impacts to economic
damages.
In addition, the National Academies of Science (NAS) identified these three models as "the most
widely used impact assessment models" in a 2010 report (NAS, 2010). Furthermore, in a
comprehensive literature review and meta-analysis conducted in 2008, the vast majority of the

22

See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.

4-43

independent impact estimates that appeared in the peer-reviewed literature were derived from
FUND, DICE, or PAGE (Tol, 2008).
While the development of the DICE, FUND and PAGE models necessarily involved
assumptions and judgments on the part of the modelers, the damage functions are not simply
arbitrary representations of the modelers’ opinions about climate damages. Rather they are based
on a review by the modelers of the currently available literature on the effects of climate change
on society. The conclusions that the modelers draw from the literature, and the bases for these
conclusions are documented, and all three models are continually updated as new information
becomes available. While EPA recognizes that there are limitations with these models, including
some of those discussed in Pindyck (2013), nonetheless IAMs provide valuable information for
regulatory impact analysis. In a recent article in the peer-reviewed literature, Weyant (2014)
addressed this issue as follows:
While Pindyck’s observations about the empirical weaknesses of IAMs or calculations of
the SCC are worthy of careful study, the conclusion that IAMs are therefore useless
fundamentally misconceives the enterprise. IAMs and the SCC are conceptual
frameworks for dealing with highly complex, non-linear, dynamic, and uncertain
systems. The human mind is incapable of solving all the equations simultaneously, and
modeling allows making “If…, then…” analyses of the impacts of different factors. The
models have provided important insights into many aspects of climate-change policy.
EPA thus believes that it was appropriate to base the SC-CO2 estimates on the DICE, FUND and
PAGE models. Moving forward, EPA will continue to follow and evaluate the latest peer
reviewed literature applying IAMs. As previously noted, EPA and all of the other IWG members
will seek external expert advice on the technical merits and challenges of using additional
models (e.g., CRED, ENVISAGE) to estimate the SC-CO2 and/or removing existing models
from the ensemble (DICE, FUND, and PAGE) used to estimate the SC-CO2.
Improving the IAMs or damage functions
To date, EPA has accepted the models as currently constituted, and omitted any damages or
beneficial effects that the model developers themselves do not include. EPA recognizes that none
of the three IAMs fully incorporates all climate change impacts, either positive or negative.
Some of the effects referenced by commenters (e.g., “catastrophic” effects, disease, and CO2
fertilization) are explicitly modeled in the damage functions of one or more of the current models
(although the treatment may not be complete), and the model developers continue to update their
models as new research becomes available. In fact, EPA and all of the other IWG members
undertook the 2013 revision because of updates to the models, which include new or enhanced
representation of certain impacts, such as sea level rise damages. In addition, some of the
categories mentioned by commenters are currently speculative or cannot be incorporated into the
damage function for lack of appropriate data. Using an ensemble of three different models was
intended to, at least partially, address the fact that no single model includes all of the impacts.
EPA recognizes that there may be effects that none of the three selected models addresses (e.g.,
impacts from ocean acidification) or that are likely not fully captured (e.g. catastrophic effects).

4-44

EPA agrees that it is important to update the SC-CO2 periodically to incorporate improvements
in the understanding of greenhouse gas emissions impacts and will continue to follow and
evaluate the latest science on impact categories that are omitted or not fully addressed in the
IAMs. Also, EPA and all of the other IWG members will seek external expert advice on the
technical merits and challenges of potential approaches to update the damage functions in future
revisions to the SCC estimates.
For EPA’s more detailed response to comments on climate sensitivity and other climate science
components of the IAMs, see 4.4-2, and for treatment of CO2 fertilization in the IAMs, see 4.4-6.
For EPA’s more detailed response to comments on consistency with the OMB Circular A-4, the
Information Quality Act, and peer review, see 4.4-5.
For EPA’s more detailed response to comments on aggregation issues, see 4.4-8.
Recommendations and comments regarding socioeconomic scenarios.
The rationale for using the EMF-22 scenarios is explained in the 2010 TSD. In addition to the
fact that they were recent, peer-reviewed, and publicly available, they had the key advantage that
GDP, population, and emissions trajectories are internally consistent for each model and scenario
evaluated. As noted in the 2010 TSD, the scenarios used “span a wide range, from the more
optimistic (e.g. abundant low-cost, low-carbon energy) to more pessimistic (e.g. constraints on
the availability of nuclear and renewables).”
OMB guidance in Circular A-4 requires benefits and costs to be computed relative to a baseline
that represents “the best assessment of the way the world would look absent the proposed
action.” EPA and the other members of the IWG determined that BAU socioeconomic scenarios
best reflect this approach. While the IWG agrees that, all else equal, the baseline used for the
SCC calculation should be updated periodically to reflect the latest projections of BAU
scenarios, the RCP/SSP scenarios used in support of AR5 may not be easily adaptable for use in
SCC modeling.
To understand why, it is important to note some key differences between the criteria used by the
IWG to select socioeconomic-emissions scenarios and what the RCP/SSP scenarios represent
and how they were developed. Each scenario in the EMF-22 exercise used by the IWG was
clearly identified as reflecting a reference case BAU or a specific CO2 stabilization scenario.
Lacking data on the probabilities of specific scenarios, the IWG chose four BAU scenarios and
one scenario representing stabilization at 550 ppm CO2-e and weighted them equally in its
analysis. The IWG acknowledges that this is not a precise characterization of the baseline but
believes it is a reasonable approach at present, in light of data limitations.
Some of the AR5 scenarios differ from these scenarios in ways that make them difficult to adapt
to the SCC context. An ad-hoc group was formed in anticipation of AR5 to develop new
scenarios to support climate impacts, adaptation, and vulnerability research. The initial work
focused on the development of RCPs, which represent a wide range of potential radiative forcing
pathways over the 21st century. The SSPs are a series of human development pathways that
could lead to radiative forcing pathways described by the RCPs. The RCPs are not based on a
fixed future path of key socioeconomic parameters, which means that multiple SSPs are
4-45

associated with an RCP. No assessment was made as to the likelihood of any particular RCP
being realized, but some of the scenarios clearly would require large-scale global mitigation
efforts to be achieved and U.S. involvement in these efforts may be necessary for their success.
In other words, the developers did not assign likelihoods that the various scenarios would
achieve the RCPs in the absence of policy interventions, which means that it is not clear which
scenarios could reasonably serve as BAU projections. Therefore, using the RCP/SSP scenarios
would require the IWG to conduct an assessment of which RCP/SSPs represent BAU scenarios,
or of the probability that each scenario occurs in a BAU case. The IWG believes that data are
currently lacking to conduct such an assessment.
EPA will continue to follow and evaluate the latest science on socioeconomic-emissions
scenarios and, along with the members of the IWG, will seek external expert advice on the
technical merits and challenges of potential approaches to update these scenarios in future
revisions to the SC-CO2 estimates.
Regarding the inclusion of a scenario associated with stabilization of atmospheric concentrations
of GHGs at 550 ppm CO2-e, the 2010 TSD clearly notes that this is “not derived from an
assessment of what policy is optimal from a benefit-cost standpoint. Rather, it is indicative of
one possible future outcome.” As noted above, OMB guidance in Circular A-4 states that the
correct baseline for regulatory impact analysis is an agency’s best assessment of the state of the
world without the regulation, and specifically states that this may include the potential for new
domestic and foreign policies that would occur absent the regulation. Including a scenario
representative of future mitigation actions is consistent with this guidance, as long as those future
policy actions are expected to occur with or without the regulation under examination. As
explained in the 2010 TSD, the IWG aimed to select scenarios that span most of the plausible
range of outcomes for the socioeconomic variables. Given the level of uncertainty in these
trajectories, the IWG felt that it was appropriate to consider a trajectory with significant global
mitigation, assuming that this is a distinct possibility even in the absence of U.S. actions.
Because there were five scenarios, and each received equal weighting, the stabilization scenario
received 20% of the total probability weight.
Regarding potential inconsistencies between scenarios and IAMs, given the nature of estimating
the SC-CO2 and available data/resources, a full harmonization along all possible dimensions of
the three IAMs used to estimate the SC-CO2 with the four models used to develop the scenarios
was not possible. Therefore, the IWG chose to harmonize the models with respect to the scenario
variables to which SCC estimates are most sensitive (GDP, population, and emissions) using
common techniques in the literature. The scenarios used were developed by highly respected
international modeling groups and published in the peer-reviewed literature. In terms of potential
inconsistencies across scenario variables past 2100, an effort was made to account for some basic
correlations among scenario variables in the post-2100 extrapolation. For example,
extrapolations were based on GDP per capita growth, which implicitly correlates population and
GDP growth, rather than GDP levels or growth alone. Similarly, extrapolations were based on
CO2 emissions intensity with respect to GDP, which correlates emissions and GDP growth,
rather than CO2 emissions levels or growth alone.

4-46

Consistent with historical observations, it is expected that growth rates of rapidly developing
economies will exceed those of already developed economies in the near term. Scenarios with
projections of global economic growth that exceed recent trends in developed economies are
consistent with this expectation.
The chosen scenarios capture a wide range of potential future states of the world, but were not
intended to represent a comprehensive accounting of the full range of uncertainty, and therefore
it is possible that future outcomes will fall outside of this range. The IWG acknowledges that the
projection of the scenarios beyond 2100 has greater uncertainty than shorter-term projections and
will continue to monitor the literature, including the development of extended RCP/SSP
scenarios, for ways to improve the estimated trajectories and improve internal consistency.
Other
In addition, EPA disagrees with Commenter 10502’s characterization of the SC-CO2 as a
rulemaking. The commenter has misunderstood what the SC-CO2 is and how it is used. The
SC-CO2 itself is not a rulemaking, it is an estimate of the net economic damages resulting from
CO2 emissions, and therefore is used to estimate the benefit of reducing those emissions. The
SC-CO2 calculation does not make any assumptions about a “solution to all potential problems
that a warming planet might cause, i.e. CO2 emissions control.” The SC-CO2 is just one
component of a larger analysis that includes consideration of many other potential impacts. See
EPA’s response to comments 4.4-3 for explanation as to why EPA uses global estimates of the
SC-CO2, response to comments 4.4-6 for details about the treatment of CO2 fertilization in the
IAMs, response to comments 4.4-8 for details about the application of the SC-CO2 to the
illustrative cost and benefit scenarios in the RIA.

4.4.2 Climate Science Components of the Modeling Underlying SC-CO2
Comment 4.4-2: Commenter 9685 states the following climate sensitivity errors: the methane
contribution to the radiative forcing is severely underestimated resulting in an over -estimate of
the relative contribution by CO2. Dynamic analysis is required to estimate the radiative forcing
for atmospheric concentrations of methane. Even worse, EDF claims that using the long term
GWP100 (lower) for methane is incorrect, the true methane GWP requires a dynamic analysis
since the emissions occur 24/7. This has the potential for quadrupling the impact of methane
concentrations and reducing the relative importance of CO2 even more. This is another correction
that the SCC has not addressed. The impact is substantial and will dramatically affect the model
results in calculations for CO2 only. The possible adjustment required for the non-CO2 gases
contribution to radiative forcing: 0.48(86/21) + 0.16 + 0.34 = 2.47(v 0.98) W/m2. The result is a
major shift in the relative importance on CO2 and a major reduction in the "climate sensitivity"
for increased CO2. Surely EPA advised the IWG of this major error. Commenter 9685-7210
continues stating the methane contribution to the radiative forcing is severely underestimated
resulting in an over-estimate of the relative contribution by CO2. The IPCC has revised upward
its estimates for the global warming potential (GWP) of methane. The GWP compares the
amount of heat methane adds to the climate system against the amount of heat that carbon
dioxide adds. The GWP of CO2 is assumed to be constant, while the GWP of methane varies
substantially over time. Over a 20-year time frame, methane has a global warming potential
(GWP20) of 86, up from 72, in the IPCC AR5. Methane is 34 times more potent a greenhouse
4-47

gas than carbon dioxide on a 100-year time scale (GWP100). It was previously assumed to be 25
times more potent (EPA uses an out-dated value of 21). The SCC models do not reflect this
correction. The net effect of this adjustment is lowering the "climate sensitivity" of CO2 because
the contribution of methane would correspondingly increase, reducing the relative importance of
CO2. Dynamic analysis is required to estimate the radiative forcing for atmospheric
concentrations of methane. Also, Commenter 9685-7212 states that the methane contribution to
the radiative forcing is severely underestimated resulting in an over-estimate of the relative
contribution by CO2. Also, the AR5 indicates another correction is required: Emissions of CH4
alone have caused an RF of 0.97 [0.74 to 1.20] W m-2. This is very likely much larger than the
concentration-based estimate of 0.48 [0.38 to 0.58] Wm-2 (unchanged from AR4). This
difference in estimates is caused by concentration changes in ozone and stratospheric water
vapor due to CH4 emissions and other emissions indirectly affecting CH4 (see Figure SPM.4).
These corrections of the methane GWP are another major flaw in the IAMs.
Commenter 10091 states the IWG2010 notes (p.14) that the calibrated Roe and Baker
distribution better reflects the IPCC judgment that "values substantially higher than 4.5degrees C
still cannot be excluded." The IWG2010 further notes that "Although the IPCC made no
quantitative judgment, the 95th percentile of the calibrated Roe & Baker distribution (7.1 dC) is
much closer to the mean and the median (7.2 dC) of the 95th percentiles of 21 previous studies
summarized by Newbold and Daigneault (2009). It is also closer to the mean (7.5 dC) and
median (7.9 dC) of the nine truncated distributions examined by the IPCC (Hegerl, et al., 2006)
than are the 95th percentiles of the three other calibrated distributions (5.2-6.0 dC)." In other
words, the IWG2010 turned towards surveys of the scientific literature to determine its
assessment of an appropriate value for the 95th percentile of the ECS distribution. Now, more
than three years hence, the scientific literature tells a completely different story. Instead of a
95th percentile value of 7.14 degreesC, as used by the IWG2010, a survey of the recent scientific
literature suggests a value of 3.5 degreesC-more than 50% lower. And this is very significant and
important difference because the high end of the ECS distribution has a large impact on the SCC
determination -a fact frequently commented on by the IWG2010. For example, from IWG2010
(p.26): "As previously discussed, low probability, high impact events are incorporated into the
SCC values through explicit consideration of their effects in two of the three models as well as
the use of a probability density function for equilibrium climate sensitivity. Treating climate
sensitivity probabilistically results in more high temperature outcomes, which in turn lead to
higher projections of damages. Although FUND does not include catastrophic damages (in
contrast to the other two models), its probabilistic treatment of the equilibrium climate sensitivity
parameter will directly affect the non-catastrophic damages that are a function of the rate of
temperature change." And further (p.30): Uncertainty in extrapolation of damages to high
temperatures: The damage functions in these IAMs are typically calibrated by estimating
damages at moderate temperature increases (e.g., DICE was calibrated at 2.5 degreesC) and
extrapolated to far higher temperatures by assuming that damages increase as some power of the
temperature change. Hence, estimated damages are far more uncertain under more extreme
climate change scenarios. And the entirety of Section V [sic] "A Further Discussion of
Catastrophic Impacts and Damage Functions" of the IWG 2010 report describes "tipping points"
and "damage functions" that are probabilities assigned to different values of global temperature
change. Table 6 from the IWG2010 indicated the probabilities of various tipping points. The
likelihood of occurrence of these low probability, high impact, events ("tipping points") is
greatly diminished under the new ECS findings. The average 95th percentile value of the new
4-48

literature survey is only 3.5degrees C indicating a very low probability of a warming reaching 35 degrees C by 2100 as indicated in the 3rd column of the above Table and thus a significantly
lower probability that such tipping points will be reached. This new information will have a large
impact on the final SCC determination using the IWG's methodology. The size of this impact has
been directly investigated.
Commenter 10091 continues in their Comment on the Landmark Legal Foundation Petition for
Reconsideration of Final Rule Standards for Standby Mode and Off Mode Microwave Ovens,
Dayaratna and Kreutzer (2013) ran the DICE model using the distribution of the ECS as
described by Otto et al. (2013) a paper published in the recent scientific literature which includes
17 authors, 15 of which were lead authors of chapters in the recent Intergovernmental Panel on
Climate Change's Fifth Assessment Report. The most likely value of the ECS reported by Otto et
al. (2013) was described as "2.0degreesC, with a 5-95% confidence interval of 1.2-3.9degreesC."
Using the Otto et al. (2013) ECS distribution in lieu of the distribution employed by the IWG
(2013), dropped the SCC by 42 percent, 41 percent, and 35 percent (for the 2.5%, 3.0%, 5.0%
discount rates, respectively). This is a significant decline. In subsequent research, Dayaratna and
Kreutzer (2014) examined the performance of the FUND model, and found that it too, produced
a greatly diminished value for the SCC when run with the Otto et al. distribution of the
equilibrium climate sensitivity. Using the Otto et al. (2013) ECS distribution in lieu of the
distribution employed by the IWG (2013), dropped the SCC produced by the FUND model to
$11, $6, $0 compared with the original $30, $17, $2 (for the 2.5%, 3.0%, 5.0% discount rates,
respectively). Again, this is a significant decline. The Dayaratna and Kreutzer (2014) results
using FUND were in line with alternative estimates of the impact of a lower climate sensitivity
on the FUND model SCC determination. Waldhoff et al. (2011) investigated the sensitivity of
the FUND model to changes in the ECS. Waldhoff et al. (2011) found that changing the ECS
distribution such that the mean of the distribution was lowered from 3.0 degrees C to 2.0degrees
C had the effect of lowering the SCC by 60 percent (from a 2010 SCC estimate of $8/ton of CO2
to $3/ton in $1995). While Waldhoff et al. (2011) examined FUNDv3.5, the response of the
current version (v3.8) of the FUND model should be similar. Additionally, the developer of the
PAGES model affirmed that the SCC from the PAGES model, too drops by 35% when the Otto
et al. (2013) climate sensitivity distribution is employed (Hope, 2013).
Further, Commenter 10091 states these studies make clear that the strong dependence of the
social cost of carbon on the distribution of the estimates of the equilibrium climate sensitivity
(including the median, and the upper and lower certainty bounds) requires that the periodic
updates to the IWG SCC determination must include an examination of the scientific literature
on the topic of the equilibrium climate sensitivity. There is no indication that the IWG undertook
such an examination. But what is clear, is that the IWG did not alter its probability distribution of
the ECS between its 2010 and 2013 SCC determination, despite a large and growing body of
scientific literature that substantially alters and better defines the scientific understanding of the
earth's ECS. It is unacceptable that a supposed "updated" social cost of carbon does not include
updates to the science underlying a critical and key aspect of the SCC. Commenter note that
there has been one prominent scientific study in the recent literature which has argued, on the
basis of recent observations of lower tropospheric mixing in the tropics, for a rather high climate
sensitivity (Sherwood et al., 2014). This research, however, suffers from too narrow a focus.
While noting that climate models which best match the apparent observed behavior of the
vertical mixing characteristics of the tropical troposphere tend to be the models with high climate
4-49

sensitivity estimates, the authors fail to make note that these same models are the ones whose
projections make the worst match to observations of the evolution of global temperature during
the past several decades. The figure below shows the observed global surface temperature history
from 1951-2013 compared with the temperature evolution projected by the collection of models
used in the new IPCC 2013 report. We broke the climate models down into two groups-those
which have a climate sensitivity greater than 3.0 degreesC (as suggested by Sherwood et al.,
2014) and those with a climate sensitivity less than 3.0 degreesC. The Figure shows that while
neither model subset does a very good job is capturing evolution of global temperature during
the past 15-20 years (the period with the highest human carbon dioxide emissions), the high
sensitivity models do substantially worse than the lower sensitivity models.
Commenter 10091 adds while Sherwood et al. (2014) prefer models that better match their
observations in one variable, the same models actually do worse in the big picture than do
models which lack the apparent accuracy in the processes that Sherwood et al. (2014) describe.
The result can only mean that there must still be even bigger problems with other model
processes which must more than counteract the effects of the processes described by Sherwood
et al. After all, the overall model collective is still warming the world much faster than it actually
is (see Figure below). In fact, for the observed global average temperature evolution for the past
30 years largely lies below the range which encompasses 95% of all climate model runs an
indication that the observed trend is statistically different from the trend simulated by climate
models. And for periods approaching 40 years in length, the observed trend lies outside of
(below) the range that includes 90% of all climate model simulations and indication that the
observed trend is marginally inconsistent with climate model simulations. These results argue
strongly against the reliability of the Sherwood et al. (2014) conclusion and instead provide
robust observational evidence that the climate sensitivity has been overestimated by both climate
models, and the IWG alike.
Commenter 10091 states regarding equilibrium climate sensitivity that in May 2013, the
Interagency Working Group (IWG) produced an updated SCC value by applying updates to the
underlying three Integrated Assessment Models (IAMs) used in its initial 2010 SCC
determination, but did not update the equilibrium climate sensitivity (ECS) employed in the
IAMs. This was not done, despite there having been, since January 1, 2011, at least 11 new
studies and 17 experiments (involving more than 44 researchers) examining the ECS, each
lowering the best estimate and tightening the error distribution about that estimate. Instead, the
IWG wrote in its 2013 report: "It does not revisit other interagency modeling decisions (e.g.,
with regard to the discount rate, reference case socioeconomic and emission scenarios, or
equilibrium climate sensitivity)." The earth's equilibrium climate sensitivity is defined in the
Interagency Working Group on Social Cost of Carbon 2010 (hereafter, IWG2010) report as "the
long-term increase in the annual global-average surface temperature from a doubling of
atmospheric CO2 concentration relative to pre-industrial levels (or stabilization at a concentration
of approximately 550 parts per million (ppm))" and is recognized as "a key input parameter" for
the integrated assessment models used to determine the social cost of carbon.
Commenter 10091 continues that the IWG2010 report has an entire section (Section III.D)
dedicated to describing how an estimate of the equilibrium climate sensitivity and the scientific
uncertainties surrounding its actual value are developed and incorporated in the IWG's analysis.
The IWG2010, in fact, developed its own probability density function (pdf) for the ECS and used
4-50

it in each of the three IAMs, superseding the ECS pdfs used by the original IAMs developers.
The IWG's intent was to develop an ECS pdf which most closely matched the description of the
ECS as given in the Fourth Assessment Report of the United Nation's Intergovernmental panel
on Climate Change which was published in 2007. The functional form adopted by the IWG2010
was a calibrated version of Roe and Baker (2007) distribution. It was described in the IWG2010
report in the following Table and Figure (from the IWG2010 report): The calibrated Roe and
Baker functional form used by the IWG2010 is no longer scientifically defensible; neither was it
at the time of the publication of the IWG 2013 SCC update. The figure below vividly illustrates
this fact, as it compares the best estimate and 90% confidence range of the earth's ECS as used
by the IWG2010/2013 (calibrated Roe and Baker) against findings in the scientific literature
published since January 1, 2011. Whereas the IWG2010/2013 ECS distribution has a median
value of 3.0 dC and 5th and 95th percentile values of 1.72 dC and 7.14 dC, respectively, the
corresponding values averaged from the recent scientific literature are 2.0 dC (median), 1.1 dC
(5th percentile), and 3.5 dC (95th percentile). These differences will have large and significant
impacts on the SCC determination.
Commenter adds (10091) the IWG2010 report noted that, concerning the low end of the ECS
distribution, its determination reflected a greater degree of certainty that a low ECS value could
be excluded than did the IPCC. From the IWG2010 (p. 14): "Finally, we note the IPCC judgment
that the equilibrium climate sensitivity "is very likely larger than 1.5dC." Although the calibrated
Roe & Baker distribution, for which the probability of equilibrium climate sensitivity being
greater than 1.5 degrees C is almost 99 percent, is not inconsistent with the IPCC definition of
"very likely" as "greater than 90 percent probability," it reflects a greater degree of certainty
about very low values of ECS than was expressed by the IPCC." In other words, the IWG used
its judgment that the lower bound of the ECS distribution was higher than the IPCC 2007
assessment indicated. However, the collection of the recent literature on the ECS shows the
IWG's judgment to be in error. As can be seen in the chart above, the large majority of the
findings on ECS in the recent literature indicate that the lower bound (i.e., 5th percentile) of the
ECS distribution is lower than the IPCC 2007 assessment. And, the average value of the 5th
percentile in the recent literature (1.1dC) is 0.62dC less than that used by the IWG-a sizeable and
important difference which will influence the SCC determination. In fact, the abundance of
literature supporting a lower climate sensitivity was at least partially reflected in the new IPCC
Fifth Assessment Report issued in 2013. In that report, the IPCC reported: Equilibrium climate
sensitivity is likely in the range 1.5dC to 4.5dC (high confidence), extremely unlikely less than
1dC (high confidence), and very unlikely greater than 6dC (medium confidence). The lower
temperature limit of the assessed likely range is thus less than the 2dC in the AR4.... Clearly, the
IWG's assessment of the low end of the probability density function that best describes the
current level of scientific understanding of the climate sensitivity is incorrect and indefensible.
But even more influential in the SCC determination is the upper bound (i.e., 95th percentile) of
the ECS probability distribution.
Commenter 10502 included a report (BOUNDING GHG CLIMATE SENSITIVITY FOR USE
IN REGULATORY DECISIONS) published in February 2014 to argue that the computing
Social Cost of Carbon (SCC) is scientifically flawed. In the Purpose section of the report they
also state that: "We believe that a metric, the Equilibrium Climate Sensitivity (ECS), is being
misused by the Interagency Working Group (IWG) that developed the SCC calculation process.
The IWG included participants from a number of US Government agencies in their attempt to
4-51

justify the economics of CO2 and other "Green House Gas" (GHG) emissions control
regulations". They explain that: "This report provides rigorous scientific documentation of how
the ECS metric is being misused in a number of ways as described in the TSDs and proposes a
more scientifically based metric to replace ECS use in regulatory activity." In the 86 pages report
they provided the following tables and figures: Table 1.1 (page 15) describing all the metrics
used and derived in the report. Table 1.2 (page 29) presents the Climate Model TCR/ECR Ratio.
Table 5.0 (page 72) shows Statistical Distributions for ECS. Figure 1.1 (page 18) which is the
Dr. John Christy Study - Climate Model Comparison to Actual Physical Data of the midTroposphere Temperature Anomalies. Figure 1.2 (page 30) show the IPCC AR4 Climate Model
TCR/ECS Trends vs. ECS. Figure 1.3 (page 30) showing the IPCC AR4 Climate Model
ECS/TCR Trends VS. TCR. Figure 2.1 (page 34) showing the Simple Spring-Mass-Damper
Dynamic System. Figure 3.1 (page 38) presents Atmospheric CO2 Concentration 1832-2013 and
Projections to 2100. Figure 4.1 (page 46) shows Comparison of Several Popular Global Average
Temperature Databases. Figure 4.2 (page 49) presents Ljungqvist Temperature Reconstruction
for the Northern Hemisphere. Figure 4.3 (page 51) shows HadCRUT4 Global Yearly Average
Temperature Anomaly. Figure 4.4 (page 53) presents Case 1 HadCRUT4 Data Curve Fit,
TCSTRF = 1.5 deg. C. Figure 4.5 (page 56) presents HadCRUT4 Data Fit with TCSTRF = 0.8
deg. C and 1000 Year Climate Cycle. Figure 4.6 (page 58) shows determining an Upper Bound
for Transient Climate Sensitivity, TCSTRF. Figure 4.7 (page 60) presents Historical
Reconstruction of TSI by Kopp - (1610 - 2010). Figure 4.8 (page 61) presents Total Solar
Irradiance (TSI) Variation: 1650 - Present and Beyond. Figure 4.9 (page 65) presents Recent
Published Trends in ECS Uncertainty Range. Figure 4.10 (page 68) shows Bounding Maximum
Possible AGW Temperature Rise. Figure 5.1 (page 74) shows the Comparison of ECS
Probability Density Functions.
Commenter 10502 states the SCC attempts to weigh possible future damage to the world-wide
society from US CO2 emissions, against the immediate and certain increases to US energy and
related economic cost created by issuance of each new regulation. Such unilateral US CO2
emissions regulations imposed on US citizens, without an enforceable world-wide agreement to
limit CO2 emissions, will have insignificant effect on global warming, and will almost certainly
cause a migration of US energy intensive manufacturing jobs to other countries that do not
impose CO2 emission regulations on themselves. Total curtailment of US-only CO2 emissions
could not prevent the joint statistical probability of high climate sensitivity to CO2 and high
atmospheric CO2 levels that lead to high computed values of SCC. Such conditions using the
IWG's scientifically unsupported ECS distribution of Figure 6 <page 11> are related to a
significant computed probability of high temperatures and high CO2 levels that could melt
permanent ice sheets on Greenland and West Antarctica and cause rapid sea level rise. There is
not much scientific agreement on the probability that significant portions of these permanent ice
sheets will be melted by rising GHG levels. Therefore, the assumptions in the SCC process that
lead to high values of SCC must be carefully scrutinized. These are key issues that need to be
reviewed in detail by an independent scientific review team to ensure that no unnecessary harm
to the US population results from unilateral US CO2 emission regulations. Commenter states
10502 that the ECS method used to determine the Social Cost of Carbon (SCC) is flawed and
compared it with other metrics that in their opinion are more appropriate. In the current SCC
calculation process, temperature rise estimates from present until the year 2300, a key driver of
SCC damage estimates, are highly corrupted by the inappropriate use of the ECS metric for
4-52

computing temperature changes due to atmospheric CO2 changes over the next 300 years. The
"best available science" that the IWG is required to employ in computing SCC should be based
on a less uncertain transient climate sensitivity metric, such as the IPCC's Transient Climate
Response (TCR) metric or our research team's Transient Climate Sensitivity (TCS) metric,
defined and very accurately determined and bounded from actual climate data as documented in
Ref. 1 (Attached document ending with Docket ID ending with 10502-A1). Commenter says that
these transient climate sensitivity metrics are much more scientifically appropriate than ECS for
computing temperature changes over the next 300 years and do not include highly disputed and
uncertain effects that have been postulated to occur much more than 300 years from now, and
that have led to so much uncertainty in establishing a confident value for ECS. Both the IPCC
AR4 report and AR5 report that compute global average temperature rise from slowly increasing
levels of GHG, inherently use the transient climate sensitivity aspects of climate simulation
models, as opposed to the much larger effects that develop over 1000's of years in the totally
inappropriate ECS simulations. However, as we have demonstrated in Ref.1 (Docket ID ending
with 10502-A1) , it is not necessary to use complex and un-validated climate simulation models
to accurately estimate and bound global average surface temperature rise due to CO2 emissions
over the next 300 years. There is ample climate data to accurately determine the TCS metric and
to use it in an algebraic equation that captures the direct and climate feedback warming effects of
GHG forcing of the climate as a function of various proposed Representative Concentration
Pathways (RCP). The IWG's SCC is computed for climate changes they expect only from the
present until the year 2300. At present, global atmospheric CO2 levels are only about 397 ppm
and increasing at a rate of about 0.5 percent per year. But, ECS is defined by the IPCC to be the
climate response that would occur after the atmospheric CO2 level is suddenly doubled from preindustrial levels of about 280 ppm to about 560 ppm and held at that level until the climate
stabilizes at a new equilibrium condition. Because of long feedback responses of the oceans to
give up even more CO2 as they gradually warm, and cause amplified GHG warming, the final
equilibrium state imbedded in the ECS value is not reached until 1000 or more years later after
the sudden doubling of CO2 levels. It is not clear how, or if, the IWG recognizes this fact in how
it predicts global temperature rise and related damages for each ton of CO2 emitted between now
and 2300. Moreover, at the present time, with an actual slowly increasing atmospheric CO2 level,
not at all like the sudden doubling of atmospheric CO2 required to compute the ECS value in
climate models, our oceans are cool enough to absorb about half of the CO2 emitted each year,
leaving only about half of the yearly emissions available to increase atmospheric CO2 levels.
Since ECS, TCR and TCS are metrics that attempt to describe temperature sensitivity to
atmospheric CO2 levels, not CO2 emissions, the TSDs have not clearly described how the ECS
value selected from the assumed statistical distribution for ECS can be used to accurately
compute global temperature rise due to the slowly rising CO2 levels over the next few hundred
years. This critical aspect of the SCC calculation process is not based on the best science
available to the IWG and its member organizations, and indicates the need for an in-depth and
independent scientific review of the entire SCC process. The commenter included a number of
Figures, some of which are also included in the attached report document (Docket ID ending
with 10502-A1). The list of Figures in this document is as follows: Figure 1 (page 6): Dr. John
Christy Study - IPCC AR5 Report Climate Model Comparison to Actual Physical Data of the
mid- Troposphere Temperature Anomalies. Figure 2 (page 8): Merged Law Dome and Mauna
Loa Datasets for Atmospheric CO2 Concentration. Figure 3 (page 9): Extracting Transient
Climate Sensitivity (TCS) from the HadCRUT4 Data. Figure 4 (page 10): Extracting an Upper
4-53

Bound for TCS from the HadCRUT4 Data. Figure 5 (page 10): Max Possible GHG Warming
Forecast, Present - 2300. Figure 6 (page 11): Comparison of TRCS and IWG probability
distributions for ECS.
Commenter 9681 counters, stating IWG's treatment of climate sensitivity is appropriate and
could go further in highlighting the costs of fundamental uncertainties. In addition to choosing an
appropriate discount rate and sensitivity analyses around different SSPs, another important
parameter to which the SCC estimates are sensitive is Equilibrium Climate Sensitivity (ECS)
how the climate system responds to a constant radiative forcing, which is typically expressed as
the temperature response to a doubling of CO2 concentration in the atmosphere. In its current
iteration, the IWG conducted extensive sensitivity analyses over a range of equilibrium climate
sensitivity estimates. (Specifying the climate sensitivity parameter as a random variable has a
basis in PAGE02, which species a probability distribution function for the parameter. The IWG
calibrated the Roe and Baker distribution, a right-skewed distribution, to characterize the
probability distribution function of this parameter. The 2010 TSD explains the IWG's choice of
the Roe and Baker distribution. The right-skewed nature of the climate sensitivity parameter's
probability distribution function is independent of the IWG's choice of the Roe and Baker
distribution. Rather, this skewness results from the IPCC's finding that values of the climate
sensitivity parameter above 4.5 degree Celsius cannot be excluded. As a result, all of the
probability distribution functions fit by the IWG for the climate sensitivity parameter were
skewed to the right (see Figure 2 in the 2010 TSD), including Roe and Baker. See 2010 TSD,
supra note 5, at 14, fig. 2.) The assumptions are clearly stated in the TSD. In addition to its
sensitivity analysis, the IWG conducted a Monte Carlo simulation over the climate sensitivity
parameter and the other random variables specified within the three IAMs.97 (A Monte Carlo
simulation will run an integrated assessment model thousands of times, each time randomly
picking the value of uncertain parameters from a probability distribution function, i.e. a function
that assigns a probability to each possible parameter value. In the case of the SCC, the IWG ran
10,000 Monte Carlo simulations for each of the three IAMs and five socio-economic scenarios,
randomizing the value of climate sensitivity, i.e., the change in average global temperature
associated with a doubling of CO2, and all other uncertain parameters in the IAMs by the original
authors. For each randomly drawn set of values, the IAM estimated the associated damages, with
the final SCC estimate equaling the average value across all 10,000 runs, five socio-economic
scenarios, and then across all three models. Therefore, each SCC estimate is calculated using
150,000 runs.) The range for the Equilibrium Climate Sensitivity (ECS) is derived from a
combination of methods that constrain the values from measurements in addition to models.
These include measured ranges from paleoclimate records, observed comparisons with current
climate, as well as responses to recent climate forcings. The currently agreed likely range for the
ECS (from both the IPCC TAR and AR5) is 1.5-4.5 degrees Celsius, with a best estimate of 3
degrees Celsius. Physical constraints make it extremely unlikely that the ECS is less than 1
degree Celsius and very unlikely greater than 6 degrees Celsius.98 (IPCC, supra note 95, at 14.)
A host of analyses points to the costs of such uncertainty both for values that go outside the
likely range and for uncertainty within it: in short, the optimal SCC tends to increase with
increased uncertainty, sometimes dramatically so. While the current treatment of uncertainty
around climate sensitivity by the IWG is sufficient to highlight the range of possible
uncertainties, a reconsideration of the assumptions feeding into the SCC ought to take the latest
advances highlighting the potentially higher costs of deep uncertainty into account.
4-54

Commenter 9685 continues that Senate testimony comparing model predictions with actual
observations. The data presented in testimony confirms that all of the models used by the IPCC,
including the 19 U.S. models, are fatally flawed and cannot be used in any policy analysis,
including the ECS input to the IAMs used in the SCC.
Commenter 9685 also states estimates of climate sensitivity should be based on data, not flawed
models. Although most of the IPCC AR5 models incorporate a mean sensitivity of 3.4 degreesC
(range of 2.1 to 4.7 degreesC), recent studies either ignored by the IPCC or papers available after
the cut-off date for the AR5 indicate the true sensitivity is much lower as suggested by Dr.
Spencer. Rather than using the demonstrably flawed IPCC models, these studies use a
combination of simple climate models bounded by data or use only data. These papers indicate
the ECS is substantially lower than what was used as input to the IAMs and that the "fat tail" is
constrained by real data. The consequence: the SCC is as much as 90% lower based on the ECS
correction alone. Commenter believes that EPA must withdraw SCC2010 and SCC2013 and
proceed with a wholesale revision of the ECS input to the IAMs using the recent papers that
document a substantially lower value than used in the SCC reports.
Commenter 9685 states Damage function errors: sea level rise is based on the flawed IPCC
models. The Sea Level Rise (SLR) treatment in the IAMs is highly suspect. Nordhaus
announced "changes" in DICE, but provided no data to show the projection of SLR versus time
or temperature. This is a major problem for the IAMs: there is no data to compare results.
Dependence on the flawed AR4 and AR5 models for SLR projections is another major problem
for the IAMs. The public is unable to examine the details because the Damage Function
aggregates the results. Each element has to be unbundled, this is especially important for SLR.
EPA must require that each IAM provide the details of each element of the Damage Function.
Commenter 9685-7221 continues, stating the model does not properly reflect benefits of fossil
fuel use and CO2 fertilization. The damage functions for the three IAMs are presented in
SCC2010 Figures 1A and 1B, using the modeler's default scenarios and mean input assumptions.
There are significant differences between the three models both at lower (figure 1B) and higher
(figure 1A) increases in global-average temperature. But the results displayed in Figures 1A and
1B show that the FUND model has negative damages ("benefits") through roughly 2080. Since
the detailed elements of the "damage" function are not transparent, one can only surmise that the
"benefits" are dominated by the CO2 fertilization effects. It is apparent that the other two models
either do not reflect this effect or minimize the effect and are thus seriously deficient. This is a
substantial flaw in SCC2010, SCC2013, and SCC2013r . Rising amounts of CO2 in the
atmosphere have the effect of increasing U.S. and global crop yields. Other benefits of CO2
include reducing the amount of fresh water needed to produce any given amount of crop and
improving the response to environmental stress (NIPCC2009) The amounts of increased yields
by crop have been calculated for certain elevated levels of CO2 in the atmosphere. The benefits
manifest themselves in the form of greater availability of food and lower prices which in turn
contribute to improved health and quality of life. These benefits largely flow to those who
otherwise struggle to afford an adequate supply of food, in other words, the poor. This effect is
the dominant "negative damage" from the effects of CO2 emissions. The IAMs do not properly
account for this effect.
Commenter 10091 continues the determination of the SCC used by the EPA in its cost
comparison analysis is discordant with the best scientific literature on the equilibrium climate
4-55

sensitivity and the fertilization effect of carbon dioxide two critically important parameters for
establishing the net externality of carbon dioxide emissions at odds with existing Office of
Management and Budget (OMB) guidelines for preparing regulatory analyses, and founded upon
the output of Integrated Assessment Models (IAMs) which encapsulate such large uncertainties
as to provide no reliable guidance as to the sign, much less the magnitude of the social cost of
carbon.
Commenter 10091: Additionally, as run by the Interagency Working Group (IWG) (whose
results were incorporated by the EPA in this action), the IAMs produce illogical results that
indicate a misleading disconnect between a climate change and the SCC value. And we show
that the sea level rise projections (and thus SCC) of at least one of the IAMs (DICE 2010) cannot
be supported by the mainstream climate science. Commenter states that until this situation in its
entirety can be properly rectified, the SCC should be barred from use in this and any other
federal rulemaking. It is better not to include any value for the SCC in cost/benefit analyses such
as these, than to include a value which is knowingly improper, inaccurate and misleading.
Response 4.4-2: All of the comments on the climate science components of the modeling
underlying the SC-CO2 estimates mirror those submitted to the Office of Management and
Budget’s separate comment solicitation on the SCC (78 FR 70586; November 26, 2013). As a
member of the interagency working group (IWG) on SC-CO2, EPA has carefully examined and
evaluated comments submitted to OMB’s separate solicitation. EPA has also carefully examined
and evaluated all comments received regarding SC-CO2 through this rulemaking process and
determined that the IWG responses to the comments on the OMB solicitation address the
comments on the climate science components of the modeling underlying the SC-CO2.
Specifically, EPA concurs with the IWG’s response to these comments and hereby incorporates
them by reference.23
EPA appreciates the commenters’ recommendations for potential opportunities to improve the
SC-CO2 estimates and has considered each one in the context of this rulemaking. EPA and other
members of the IWG have reconfirmed their commitment to periodic review and update of the
methodology and estimates to ensure that they continue to reflect the best available science and
economics. Furthermore, EPA has determined that the SC-CO2 methodological
recommendations require additional research, review, and public comment before it can apply
them to a rulemaking context.
To help synthesize the technical information and input reflected in the comments, and to add
additional rigor to the next update of the SC-CO2, the EPA and all of the other IWG members
plan to seek independent expert advice on technical opportunities to improve the SC-CO2
estimates, including the approaches suggested by commenters and summarized in this document.
Specifically, EPA and all of the other IWG members plan to ask the National Academies of
Sciences, Engineering, and Medicine (Academies) to examine the technical merits and
challenges of potential approaches to improving the SC-CO2 estimates in future updates. Input
from the Academies, informed by the comments submitted to this rulemaking and the peer23 Referred to as the "OMB Response to Comments on SC-CO2." See pgs 11 to 17 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf.

4-56

reviewed literature, will help to ensure that the SC-CO2 estimates used by EPA continue to
reflect the best available science and methodologies.
After careful evaluation of the full range of comments and associated technical issues described
in section 4.4, EPA has determined that it will continue to use the current SC-CO2 estimates 24
until further updates can be made to reflect forthcoming guidance from the Academies. EPA
believes that the current estimates continue to represent the best scientific information on the
impacts of climate change available in a form appropriate for incorporating the damages from
incremental CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the
current SC-CO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in
RIA Chapter 5.25
EPA will continue to consider these comments and will share the recommendations with the
IWG as it moves forward with the Academies process.
The remainder of this section provides more detailed responses to the comments and
recommendations.
Regarding comments on equilibrium climate sensitivity (ECS), the EPA is aware that this is an
active area of research and remains committed to updating the SC-CO2 estimates to incorporate
new scientific information and accurately reflect the current state of scientific uncertainty
regarding the ECS. While we agree with commenters that the ECS distribution, along with other
climate modeling inputs to the SCC calculation, should be updated periodically to reflect the
latest scientific consensus, care must be exercised in selecting an appropriate range of estimates
for this important parameter. Many studies estimating climate sensitivity have been published,
based on a variety of approaches (instrumental record, paleoclimate observations, models, etc.).
These individual studies report differing values and provide different information. Picking a
single study from the high or low end of the range, or even in the middle, will exclude relevant
information. A valid representation of uncertainty regarding climate sensitivity should be
obtained from a synthesis exercise such as that done by the IPCC that considers the full range of
relevant studies.
At the time the 2013 SC-CO2 update was released, the most authoritative statement about ECS
appeared in the IPCC’s AR4. Since that time, as several commenters noted, the IPCC issued a
Fifth Assessment Report that updated its discussion of the likely range of climate sensitivity
compared to AR4. The new assessment reduced the low end of the assessed likely range (high
confidence) from 2°C to 1.5°C, but retained the high end of the range at 4.5°C. Unlike in AR4,
the new assessment refrained from indicating a central estimate of ECS. This assessment is based
on a comprehensive review of the scientific literature and reflects improved understanding, the
extended temperature record for the atmosphere and oceans, and new estimates of radiative
forcing.

24 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.
25 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.

4-57

Several of the post-AR4 studies highlighted by some commenters were cited in the AR5
assessment. In particular, both Aldrin et al. (2012) and Otto et al. (2013) were cited in both
Chapter 10 and Chapter 12 of the AR5 Working Group I assessment. Eight of the authors of Otto
et al. (2013), including the lead author, were authors of Chapter 12 for AR5’s Working Group I
and one was a lead author for the chapter. Hence it is clear that the IPCC considered Otto et al.
(2013) in its synthesis of literature on the ECS. More broadly, the AR5 climate sensitivity
distribution likely incorporates much of the literature identified by the commenters. EPA will
continue to follow and evaluate the latest science on the equilibrium climate sensitivity.
Furthermore, EPA and the other members of the IWG are seeking external expert advice on the
technical merits and challenges of potential approaches prior to updating the ECS distribution in
future revisions to the SCC estimates, including (but not limited to) using the AR5 climate
sensitivity distribution for the next update of the SC-CO2.
As discussed in preamble section II, RTC section 4.1, the 2009 Endangerment Finding, and in
the OMB RTC, among other resources, the links between CO2 and temperature are established
beyond question by laboratory measurements, physical theory, paleoclimate observations,
instrumental observations, and observations of other planets. Climate change and its impacts,
such as sea level rise, have been exhaustively documented, and synthesized internationally by the
IPCC and domestically by the U.S. National Climate Assessment. Based on the wide acceptance
of these conclusions in the scientific community, EPA believes that: (1) anthropogenic emissions
of greenhouse gases are causing atmospheric levels of greenhouse gases in our atmosphere to
rise to levels unprecedented in human history; (2) the accumulation of greenhouse gases in our
atmosphere is exerting a warming effect on the global climate; (3) there are multiple lines of
evidence, including increasing average global surface temperatures, rising ocean temperatures
and sea levels, and shrinking ice in glaciers, ice sheets, and the Arctic, all showing that climate
change is occurring, and that the rate of climate change in the past few decades has been unusual
in the context of the past 1000 years; (4) there is compelling evidence that anthropogenic
emissions of greenhouse gases are the primary driver of recent observed increases in average
global temperature; (5) atmospheric levels of most greenhouse gases are expected to continue to
rise for the foreseeable future; and (6) risks and impacts to public health and welfare are
expected to grow as climate change continues, and that climate change over this century is
expected to be greater compared to observed climate change over the past century.
While there are inherent uncertainties associated with modeling climate systems over long time
spans, the general circulation models (GCMs) upon which estimates of ECS and other climate
science research are based have been extensively evaluated. For example, since 1989 the DOE
has had a large program (The Program for Climate Model Diagnosis and Intercomparison)
dedicated to evaluating these models.
The ECS parameter is a useful parameter for summarizing the strength of the climate system’s
response to accumulating GHG concentrations in the atmosphere. However, it is influenced by
many highly complex and uncertain natural processes, some of which adjust over very long
periods of time. Therefore, persistent uncertainty about the ECS is not surprising. Furthermore,
persistent uncertainty does not suggest an absence of useful information. However, EPA does not
agree that progress has not been made in reducing this uncertainty. Over the last 30 years the
scientific community has elucidated many aspects of the climate system’s response to GHGs
accumulating in the atmosphere. While the AR5 “likely” range is slightly larger than that of
4-58

AR4, the assessment presented greater confidence in the tails. AR5 found that climate sensitivity
is very unlikely to be greater than 6°C, whereas AR4 stated that the “lack of strong constraints
limiting high climate sensitivities prevents the specification of a 95th percentile bound.”
Similarly, while the AR5 and the IPCC’s (2001) Third Assessment Report (TAR) bounds look
similar, the TAR bounds were presented as a range without estimated probabilities.
In response to the commenter citing Pindyck on climate feedbacks and the shape of the ECS
distribution, we agree that potential climate feedbacks and their strength are uncertain. However,
the IWG chose a distribution from the peer-reviewed literature based on its evaluation of the
scientific literature and the relationship between this distribution and IPCC range of the ECS.
Regarding the skewness of the calibrated ECS distribution, this characteristic is common among
the many approaches that have been used to study the ECS distribution and is not unique to the
theoretical approach used by Roe and Baker to define the functional form of their distribution.
Consistent with the AR4 discussion on limiting the distribution to a range considered possible by
experts, such as from 0°C to 10°C, the IWG truncated the distribution at 10oC (Hegerl 2007, p
719).
EPA disagrees with the comment that “AR5 moderates many of the key findings of AR4.” This
response has already discussed in detail EPA’s assessment of the AR5 findings on climate
sensitivity and the implications for the SC-CO2 estimates. The other examples given by the
commenter—anthropogenic warming, rate of warming, extreme temperatures, and tropical
cyclone activity—are not findings that have been moderated in AR5. In fact, AR5 strengthened
the conclusions about attribution of anthropogenic global warming by accounting for pollutants
with a cooling effect and by strengthening the likelihood of warming (“very likely” in AR4 to
“extremely likely” in AR5). Similarly, the AR5 findings on tropical cyclone activity
strengthened from “suggestions” in AR4 to “virtually certain” increases in frequency and
intensity of the strongest tropical cyclones in the North Atlantic basin. The remaining two
examples given by the commenter highlight historical information, not projections. EPA further
notes that the commenter has not identified or discussed whether and how these particular AR5
findings would affect the SC-CO2 estimates.
In addition, EPA notes that the preamble section II discusses many new assessment reports
published since the 2009 Endangerment Finding, including the recent USGCRP’s 2014 Third
National Climate Assessment (NCA3), the IPCC’s 2013-4 Fifth Assessment Report (AR5) and
IPCC’s 2012 “Special Report on Managing the Risks of Extreme Events and Disasters to
Advance Climate Change Adaptation” (SREX). Section II of the preamble highlights changes,
such as sea level rise, and cites the confidence in human attribution. The preamble concludes
that the “findings of the recent scientific assessments confirm and strengthen the conclusion that
greenhouse gases endanger public health, now and in the future.” See also section 4.2 of this
RTC for EPA’s response to comments about use of the latest climate science, including AR5
findings, in analyses separate from the IWG SCC estimates. Section 4.4-1, provides a detailed
response to comments on the methodology and the integrated assessment models, including the
damage functions, used to estimate the SC-CO2. Section 4.4-5 provides a detailed response to
comments regarding the transparency of the SC-CO2 process and compliance with OMB
guidance.
Other physical science components of the integrated assessment models.
4-59

A key objective of the IWG was to draw from the insights of the three models while respecting
the different approaches to linking GHG emissions and monetized damages taken by modelers in
the published literature. After conducting an extensive literature review, the interagency group
selected three sets of input parameters (climate sensitivity, socioeconomic and emissions
trajectories, and discount rates) to use consistently in each model. All other model features were
left unchanged, relying on the model developers’ best estimates and judgments, as informed by
the then-current literature. While the IAMs are periodically updated, the rapid pace of research in
the area of climate science means that at any given time there may be new research that has not
yet been incorporated into one or more models. Thus, while a given model may not always
reflect the most recent research regarding any given climate subsystem, the IWG concluded that,
at the time, the IAMs collectively represented the state of the science by bringing all these
systems together into a single framework.
A benefit of using an ensemble of three models is that they cover a range of potential outcomes
as expressed in the literature. For example the three models used collectively span a range of
carbon cycle and climate change responsiveness that reflects the uncertainty in the literature.
With regards to comments on the temperature response function, the past 15 years of observed
atmospheric temperatures cannot be compared directly to climate model simulations because (1)
observed temperatures were influenced by volcanic eruptions that were not included in
simulations because the timing and spatial distribution of eruptions are not known in advance;
and (2) the last 15 years of atmospheric temperatures have been strongly influenced by natural
climate variability due to oceanic fluctuations, such as the El Niño Southern-Oscillation; models
include this variability, but no attempt is made to synchronize the models’ timing of this
variability with observed variability. In other words, while the models incorporate variability
around a trend over time, they cannot predict how that variability affects measured temperatures
in a specific year.
Regarding the criticisms by commenters of the sea level rise projections in DICE, the IWG
recognizes that sea level rise projections are also an area of ongoing research. One key issue
involves projections of melt from the Greenland and West Antarctic ice sheets. The IPCC AR5
report notes there is a possibility of sea level rise “substantially above” their best estimate of a
likely range because of uncertainties regarding the response of the Antarctic ice sheet (AR5
Working Group I, Chapter 13). In AR5 the IPCC also discusses semi-empirical methods, stating
a low confidence in projections based on such methods, which calibrate a mathematical model
against observations rather than projecting individual processes. However, the IPCC did not
entirely discount these methods. Further supporting the use of semi-empirical methods, the U.S.
National Climate Assessment uses an average of the high end of semi-empirical projections in
order to define their “Intermediate-High” Scenario (Parris et al., 2012). Therefore, it is
reasonable for one out of three models used by the IWG to include some reliance upon semiempirical methods.
EPA is aware that more sophisticated yet still relatively simplified climate models, such as
MAGICC, could be used to replace the highly simplified climate science components of the
three IAMs. However, given the range of climate models available and the technical issues
associated with such a change, replacing the climate modules or other structural features of the
IAMs requires additional investigation before it can be applied to SCC estimation. EPA will
4-60

continue to follow and evaluate the latest science on climate modeling and, along with other
members of the IWG, is seeking external expert advice on the technical merits and challenges of
potential approaches to updating this component of the IAMs in future revisions to the SCC
estimates.
We agree with the commenters who suggested the IAMs do not fully capture the impacts
associated with changes in climate variability and weather extremes. For example, as discussed
in the 2010 TSD, the calibrations in FUND and DICE do not account for increases in climate
variability that may occur and would affect the agricultural sector. Similarly, we agree that the
models’ functional forms may not adequately capture potentially discontinuous “tipping p oint”
behavior in Earth systems. In fact, large-scale earth system feedback effects (e.g., Arctic sea ice
loss, melting permafrost, large scale forest dieback, changing ocean circulation patterns) are not
modeled at all in one IAM, and are imperfectly captured in the others. This limitation of the three
IAMs is discussed extensively in the 2010 TSD, and again in the 2013 update. The SCC estimate
associated with the 95th percentile of the distribution based on the 3 percent discount rate is
included in the recommended range partly to address this concern. EPA and the other members
of the IWG will continue to follow and evaluate the latest science on climate variability and
potential tipping points, and seek external expert advice on the technical merits and challenges of
potential approaches to improve the representation of these components of the modeling in future
revisions.

4.4.3 Global vs. Domestic Value
Comment 4.4-3: Commenter 10091 states the IWG only reports the global value of the SCC
which the IWG determines to accrue from continued carbon dioxide emissions in the United
States. This is in direct violation of existing Office of Management and Budget (OMB)
guidelines. OMB Circular A-4 (September 17, 2003) regarding Regulatory Analysis explicitly
states: Your analysis should focus on benefits and costs that accrue to citizens and residents of
the United States. Where you choose to evaluate a regulation that is likely to have effects beyond
the borders of the United States, these effects should be reported separately. In reporting the
SCC, the IWG argues away the need to focus on benefits and costs that accrue to citizens and
residents of the United States and instead bases its SCC solely on its determinations of effects
beyond the borders of the United States. Rather than reporting the latter separately, as
recommended by OMB guidelines, the IWG only reports the global costs and makes no
determination of the domestic costs. Considering that the majority (if not all) of the federal
regulations incorporating the SCC into cost/benefit analysis apply to rules regulating domestic
activities (including this one), reporting only the global impact the knowledge (in all areas, i.e.,
economics, social, environmental, etc.) of which is far less constrained than potential U.S.
impacts imparts a huge degree of uncertainty and is a grossly misleading. Thus, the IWG's
determination of the SCC is not appropriate for use in this or any federal regulatory analysis.
During the public comment period associated with new regulations such as this one which
incorporate the SCC, a distinction should be made between domestic costs/benefits and foreign
cost/benefits such that the public can judge for itself the value of the regulation. As it currently
stands, the public likely has no idea that the benefits of the proposed regulations on domestic
activities that supposedly accrue from incorporating the SCC are largely conferred upon foreign
nations. This is clearly not a transparent situation.
4-61

Commenter 10502 included a report stating: "The scientifically flawed Social Cost of Carbon
(SCC) computational process attempts to compute the future global warming damage to the
entire world-wide population that would be avoided by each ton of CO2 not emitted by US
sources and count that highly uncertain world-wide avoided damage as a benefit of the
regulation. But the SCC cost/benefit analysis only burdens the US economy and population (not
the entire world) with the immediate cost impacts from issuance of the US regulations. These
impacts to the US economy result in part from effective elimination of coal fired power plants,
our most inexpensive electrical power source. This will result in much higher US energy costs
and loss of jobs to other countries with lower coal-produced energy costs that do not choose to
bear the burden of CO2 emissions control. Only the Legislative and Justice Branches of our
government, duly protecting interests of US citizens, can sort out the wisdom of US citizens
paying for the GHG emission sins of other nations that choose not to restrict their emissions,
while willingly accepting the transfer from the USA of energy intensive industries to their lowercost energy economies." Commenter 10502 remarks stating the unilateral USA CO2 emissions
regulations justified by the SCC, commit the USA population to bear the increased costs
imposed by the regulations (a "hidden carbon tax" expressed by some critics) for highly
uncertain future damages to be incurred not only in the USA, but by the entire world-wide
society. This certainly inflates the cost of SCC with respect to the cost that would accrue to the
US only.
Commenter 10095 states in accordance with the White House Office of Management and
Budget's (OMB) guidance on completing regulatory impact analyses, EPA must "focus on
benefits and costs that accrue to citizens and residents of the United States." The RIA refers to
the federal government's conclusion in its 2010 SCC technical support document that a global
measure of the benefits from reducing U.S. emissions is preferable; however, solely using global
SCC values contradicts OMB's guidance and is an issue that was raised in public comments
regarding the 2013 SCC Technical Support Document ("TSD"). EPA, when applying SCC
values in the RIA, also failed to analyze a seven percent discount rate which OMB's guidance
suggests using as a base-case for regulatory analyses. EPA, to remain in accordance with explicit
OMB guidance, must reassess the proposal's impact assessment using domestic, not global, SCC
values and SCC values with a discount rate of seven percent.
Commenter 9194 adds that the IWG/OMB process is fatally flawed in that it fails to calculate the
SCC using OMB's prescribed "base case" discount rate of 7% and impermissibly calculates
global benefits instead of domestic benefits.
Commenter 10023 states because the proposed rule would govern only domestic activities, it is
inappropriate for EPA to use a global social cost of carbon value to estimate the benefits of those
activities. Cost-benefit analysis for domestic policy should examine domestic costs and benefits;
including global benefits leads to bias in favoring regulation.
Commenter 9396 International Benefits: The IWG chose to model the global benefits of GHG
reductions and then to assign those benefits to US reductions. The inclusion of costs of
international harm is unprecedented and illegal. Administration authority to regulate
international action or regulate for the international community is non-existent. By choosing to
include international benefits the IWG has mixed apples in with oranges in the development of
an appropriate SCC.
4-62

Commenter 9681 states that use of a global SCC value is already influencing other international
actors to follow suit, the International Monetary Fund (IMF) applies in its policy reviews an SCC
estimate based on the IWG number.26 Given the potential influence of the IMF on the
environmental policies of developing countries,27 the pull that the IWG's global estimate has at
the IMF could be very advantageous to the United States, by motivating industrializing countries
to use similar numbers in the future. In addition to this compelling strategic argument namely,
that it is rational for the United States and other countries to continue their reciprocal, tit-for-tat
use of a global SCC value to achieve the economically efficient outcome on climate change (and
avoid catastrophic climate impacts) legal obligations further prescribe using a global SCC value.
A basic ethical responsibility to prevent transboundary environmental harms has been enshrined
in customary international law. For the United States to knowingly set pollution levels in light of
only domestic harms, willfully ignoring that its pollution directly imposes environmental risks
including catastrophic risks on other countries, would violate norms of comity among countries.
The United States would be knowingly causing foreseeable harm to other countries, without
compensation or just cause. Given that the nations most at risk from climate change are often the
poorest countries in the world, such a policy would also violate basic and widely shared ethical
beliefs about fairness and distributive justice. Indeed, taking a global approach to measuring
climate benefits is consistent with the ideals of transboundary responsibility and justice that the
United States commits to in other foreign affairs. The unmistakable implication of the
Convention is that parties including the United States must account for global economic, public
health, and environmental effects in their impact assessments. Binding international agreements
also require consideration and mitigation of transboundary environmental harms. Notably, the
United Nations Framework Convention on Climate Change to which the United States is a party
declares that countries policies and measures to deal with climate change should be costeffective so as to ensure global benefits at the lowest possible cost. The Convention further
commits parties to evaluating global climate effects in their policy decisions, by employ[ing]
appropriate methods, for example impact assessments . . . with a view to minimizing adverse
effects on the economy, on public health and on the quality of the environment.
Commenter 9681 states obligations exist in domestic U.S. law. For example, the U.S. National
Environmental Policy Act recognizes the worldwide and long-range character of environmental
problems and requires federal agencies to include reasonably foreseeable transboundary effects
in their environmental impact statements. While some individual statutes under which federal
agencies will craft climate policies may be silent on the issue of considering extraterritorial
benefits, arguably the most important statute for U.S. climate policy the Clean Air Act requires
the control of air emissions that affect other countries and so encourages a global assessment of
greenhouse gas effects. Specifically, Section 115 of the Clean Air Act directs EPA and the states
to mitigate U.S. emissions that endanger foreign health and welfare. The global perspective on
climate costs and benefits required by that provision should inform all regulatory actions
26 E.g., Benedict Clements et al., International Monetary Fund, Energy Subsidy Reforms: Lessons and Implications
9 (IMF Policy Paper, Jan. 28, 2013).
27
See Natsu Taylor Saito, Decolonization, Development, and Denial, 6 FL. A & M U. L. REV. 1, 16 (2010)
(quoting former IMF counsel as saying today it is common to find these institutions [IMF and World Bank]
requiring their borrowing member countries to accept and adhere to prescribed policies on environmental
protection).

4-63

developed under the Clean Air Act, including these proposed performance standards under
Section 111.
Commenter 9681 states two practical considerations counsel in favor of a global SCC value.
First, unlike some other significant international environmental impacts, no methodological
limitations block the quantitative estimation of a global SCC value. In recent regulatory impact
analyses for major environmental rules, EPA has qualitatively considered important transnational
impacts that could not be quantified. For example, in the Mercury and Air Toxics Standards,
EPA concluded that a reduction of mercury emissions from U.S. power plants would generate
health benefits for foreign consumers of fish, both from U.S. exports and from fish sourced in
foreign countries. EPA did not quantify these foreign health benefits, however, due to
complexities in the scientific modeling. Similarly, in the analysis of the Cross-State Air Pollution
Rule, EPA noted though could not quantify the substantial health and environmental benefits that
are likely to occur for Canadians as U.S. states reduce their emissions of particulate matter and
ozone pollutants that can drift long distances across geographic borders. Yet where foreign costs
or benefits are important and quantifiable, other federal agencies frequently include those
calculations. Given that sophisticated models already exist to quantify the global SCC, the global
estimate is appropriate to use.
Commenter 9681 continues a global SCC value is in the national interest because harms
experienced by other countries could significantly impact the United States. Climate damages in
one country could generate large spillover effects to which the United States is especially
vulnerable. The mesh of the global economy is woven tightly, and disruptions in one place can
have consequences around the world. As seen historically, economic disruptions in one country
can cause financial crises that reverberate globally at a breakneck pace. In a similar vein,
national security analysts in government and academia increasingly emphasize that the
geopolitical instability associated with climatic disruptions abroad poses a serious threat to the
United States. Due to its unique place among countries both as the largest global economy with
trade and investment dependent links throughout the world, and as a military superpower the
United States is particularly vulnerable to international spillover effects. The 2010 TSD included
a rigorous examination of global versus domestic SCC estimates. Consistent with the above
discussion, the 2010 IWG reached the conclusion to estimate a global SCC value, citing both the
global impacts of climate change and the global action needed to mitigate climate change. The
IWG restated these arguments in the 2013 TSD, and refers back explicitly to its discussion in the
2010 TSD. EPA should continue using a global SCC estimate in its regulatory impact analyses.
Commenter 9681 states the IWG correctly used a global SCC value. To design the economically
efficient policies necessary to forestall severe and potentially catastrophic climate change, all
countries must use a global SCC value. Given that the United States and many other significant
players in the international climate negotiations have already applied a global SCC framework in
evaluating their own climate policies, the continued use of the global value in U.S. regulatory
decisions may be strategically important as the United States seeks to set an example for other
countries, harmonize regulatory systems, and take the lead in ongoing international negotiations.
Binding legal obligations, basic ethical responsibilities, and practical considerations further
counsel in favor of the United States using a global SCC value. To avoid a global tragedy of the
commons and an economically inefficient degradation of the world's climate resources, all
countries should set policy according to a global SCC value. The climate and clean air are global
4-64

common resources, meaning they are free and available to all countries, but any one country's
use i.e., pollution imposes harms on the polluting country as well as the rest of the world.
Because greenhouse gases do not stay within geographic borders but rather mix in the
atmosphere and affect climate worldwide, each ton of carbon pollution emitted by the United
States not only creates domestic harms, but also imposes additional and large externalities on the
rest of the world, including disproportionate harms to some of the least-developed nations.
Conversely, each ton of carbon pollution abated in another country will benefit the United States
along with the rest of the world. If all countries set their greenhouse gas emission levels based on
only their domestic costs and benefits, ignoring the large global externalities, the collective result
would be substantially sub-optimal climate protections and significantly increased risks of severe
harms to all nations, including to the United States. "[E]ach pursuing [only its] own best interest .
. . in a commons brings ruin to all."28 By contrast, a global SCC value would require each country
to account for the full damages of its greenhouse gas pollution and so to collectively select the
efficient level of worldwide emissions reductions needed to secure the planet's common climate
resources.
Commenter 9681 continues well-established economic principles demonstrate that the United
States stands to benefit greatly if all countries apply a global SCC value in their regulatory
decisions. A rational tactical option in the effort to secure that economically efficient outcome is
for the United States to continue using a global SCC value itself. The United States is engaged in
a repeated strategic game of international negotiations and regulatory coordination, in which
several significant players, including the United States have already adopted a global SCC
framework. For the United States to now depart from this implicit collaborative dynamic by
reverting to a domestic-only SCC estimate could undermine the country's long-term interests in
future climate negotiations and could jeopardize emissions reductions underway in other
countries, which are already benefiting the United States. A domestic-only SCC value could be
construed as a signal that the United States does not recognize or care about the effects of its
policy choices on other countries, and signal that it would be acceptable for other countries to
ignore the harms they cause the United States. Further, a sudden about-face could undermine the
United States credibility in negotiations. The United States has recently reasserted its desire to
take a lead in both bilateral and international climate negotiations. To set an example for the rest
of the world, to advance its own long-term climate interests, and to secure greater cooperation
toward reducing global emissions, strategic factors support the continued use a global SCC value
in U.S. regulatory decisions. Though the Constitution balances the delegation of foreign affairs
power between the executive and legislative branches, [t]he key to presidential leadership is the
negotiation function. Everyone agrees that the President has the exclusive power of official
communication with foreign governments. The development and analysis of U.S. climate
regulations are essential parts of the dialogue between the United States and foreign countries
about climate change. Using a global SCC value communicates a strong signal that the United
States wishes to engage in reciprocal actions to mitigate the global threat of climate change. The
President is responsible for developing and executing the negotiation strategy to achieve the
United States long-term climate interests. Currently, the President has instructed federal agencies
to use a global SCC value as one important part of an implicit tit-for-tat negotiating strategy that
28

Garrett Hardin, The Tragedy of the Commons, 162 SCIENCE 1243 (1968).

4-65

encourages other countries to take reciprocal actions that also account for global externalities.
The President's constitutional powers to negotiate international agreements would be seriously
impaired if federal agencies were forced to stop relying on a global SCC value.
Commenter 9681 continues the United States has already begun to harmonize with other
countries its policies on climate change and on the valuation of regulatory benefits. For example,
the United States has entered into a joint Regulatory Cooperation Council with Canada, which
has adopted a work plan that commits the two countries to synchronizing aggressive greenhouse
gas reductions, especially in the transportation sector. A separate Regulatory Cooperation
Council with Mexico calls generally for improving and harmonizing policy "by strengthening the
analytic basis of regulations," and its work plan acknowledges the transboundary nature of
environmental risks. Mexico and Canada have both adopted greenhouse gas standards for
vehicles that harmonize with the U.S. standards and that calculate benefits according to a global
SCC value. Canada has also used the IWG's global SCC value in developing carbon dioxide
standards for its coal-fired power plants, estimating $5.6 billion (Canadian dollars) worth of
global climate benefits. The direct U.S. share of the net benefits from that Canadian regulation
will likely total in the hundreds of millions of dollars.
Commenter 9681 states further efforts at regulatory harmonization are currently underway. For
example, the United States is now negotiating a Transatlantic Trade and Investment Partnership
with the European Union, and a key element is regulatory coordination. The European Union has
already adopted an Emissions Trading Scheme (ETS) to cap its greenhouse gas emissions, and
its Aviation Directive (which extends the ETS cap to international flights) is just one of the
climate policies that could be shaped by these negotiations. The European Commission has
indicated its willingness to further reduce its ETS cap if other major emitters make proportional
commitments a result that will only occur if countries consider more than their own domestic
costs and benefits from reducing greenhouse gas emissions. Moreover, several individual
European nations including the United Kingdom, France, Germany, and Norway - have adopted
a global SCC value for use in their regulatory analyses. Some other European countries, such as
Sweden, have adopted high carbon taxes that implicitly operate as a high SCC that accounts for
global externalities. Even countries that have proposed setting carbon prices below the global
SCC value acknowledge that, in an optimal system, prices worldwide would fully reflect the
global harms of greenhouse gas emissions.
Commenter 9666 states that RIA should not use global SCC to estimate benefits, citing UARG
on OMB's recent proposed rule to require agencies to use the SCC in rulemaking proceedings.
Commenters explain that because the proposed SCC value would govern only domestic
activities, it is inappropriate for the asserted benefits of the activities to be measured on a global
scale. Cost-benefit analysis for domestic policy should examine domestic costs and benefits.
Commenter 7977 states the use of the SCC (social cost of carbon) overinflates the social benefits
of the proposed rule. Commenter 7977-4861 states that the SCC is modeled to estimate the
global benefit of CO2 reductions, wherein the proposed rule is specific to U.S. EGUs. Therefore,
the social benefits calculated by EPA should represent those specific to the U.S. rather than
worldwide. This is also contrary to OMB Guidance as stated in the SCC Technical Support
Document.
4-66

Commenter 9423 The TCEQ, PUC, and RRC understand that the SCC estimates are based on the
global benefits of C02 reductions vs. the benefits that would occur domestically in the United
States. Given that the SCC estimates will be used to compare benefits to the domestic costs of a
regulatory action, the SCC estimates should be directly related to domestic benefits. TCEQ,
PUC, and RRC note that 0MB Circular A-4 states that "analysis of economically significant
proposed and final regulations from the domestic perspective is required, while analysis from the
international perspective is optional." In the 2010 Technical Support Document: Social Cost of
Carbon for Regulatory Impact Analysis Under Executive 0rder 12866, the Interagency Working
Group concluded that a global measure of benefits from reducing U.S. emissions is preferable in
order to address the global nature of climate change and to show that international action is
necessary since "even if the United States were to reduce its greenhouse gas emissions to zero,
that step would be far from enough to avoid substantial climate change." The TCEQ, PUC, and
RRC disagree that a global measure is preferable in that the public has the right to know the
domestic benefits vs. the domestic costs of a regulatory action. The use of a global SCC could
show that benefits exceed costs when they actually may not provide any net benefits for the
United States. In addition, the Interagency Working Group determined that the models were
currently inadequate to determine region- or country-specific estimates of the SCC, but "that a
range of values from 7 to 23 percent should be used to adjust the global SCC to calculate
domestic effects." This range was based on very generalized key parameter assumptions using a
2.5 or 3 percent discount rate (which in itself is a subjective and questionable rate) or
alternatively by proportionally determining benefit by the U.S. share of gross domestic product.
The Working Group itself acknowledged that this range was highly speculative. The TCEQ,
PUC, and RRC submit that the method of translating global benefits to domestic benefits is so
speculative at this time that its use is not appropriate. In summary, EPA should not use the SCC
estimates in the current 111(b) rulemaking project until there has been a complete and thorough
independent peer review of the process and model data.
Response 4.4-3: The comments regarding the scope of the SC-CO2 estimates mirror those
submitted to the Office of Management and Budget’s separate comment solicitation on the SCC
(78 FR 70586; November 26, 2013). As a member of the interagency working group (IWG) on
SC-CO2, EPA has carefully examined and evaluated comments submitted to OMB’s separate
solicitation. EPA has also carefully examined and evaluated all comments received regarding
SC-CO2 through this rulemaking process and determined that the IWG responses to the
comments on the OMB solicitation address the comments on the scope of SC-CO2 estimates and
use of the estimates in this RIA. Specifically, EPA concurs with the IWG’s response to these
comments and hereby incorporates them by reference. 29
After careful evaluation of the full range of comments and associated technical issues described
in section 4.4, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SC 29 Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 30 to 32 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf.

4-67

CO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in RIA
Chapter 5.30
In addition, the comments regarding consistency with OMB Circular A-4, including selection of
discount rates, are addressed in 4.4-5; response to comments about the selection of discount rates
applies to the SC-CO2 is in 4.4-7.
The remainder of this section elaborates on these comments in the context of this rulemaking.
EPA agrees with the commenters who stated that a focus on global SC-CO2 estimates in RIAs is
appropriate. As discussed in the 2010 TSD, the IWG determined that a global measure of SCCO2 is appropriate in this context because emissions of most greenhouse gases contribute to
damages around the world and the world’s economies are now highly interconnected. To reflect
the global nature of the problem, the SC-CO2 incorporates the full damages caused by CO2
emissions and other governments are expected to consider the global consequences of their
greenhouse gas emissions when setting their own domestic policies.
As stated in the OMB Response to Comments on SC-CO2, if all countries acted independently to
set policies based only on the domestic costs and benefits of carbon emissions, it would lead to
an economically inefficient level of emissions reductions which could be harmful to all
countries, including the United States, because each country would be underestimating the full
value of its own reductions. This is a classic public goods problem because each country’s
reductions benefit everyone else and no country can be excluded from enjoying the benefits of
other countries’ reductions, even if it provides no reductions itself. In this situation, the only way
to achieve an economically efficient level of emissions reductions is for countries to cooperate in
providing mutually beneficial reductions beyond the level that would be justified only by their
own domestic benefits. By adopting a global estimate of the SC-CO2, the U.S. government can
signal its leadership in this effort. In reference to the public good nature of mitigation and its role
in foreign relations, thirteen prominent academics noted that these "are compelling reasons to
focus on a global [SC-CO2]" in a recent article on the SC-CO2 (Pizer et al., 2014). In addition, as
noted by commenters, there is no bright line between domestic and global damages. Adverse
impacts on other countries can have spillover effects on the United States, particularly in the
areas of national security, international trade, public health and humanitarian concerns.
GHG emissions in the United States will have impacts abroad, some of which may, in turn,
affect the United States. For this reason, a purely domestic measure is likely to understate actual
impacts to the United States. Also, as stated above, the EPA and the other members of IWG
believes that accounting for global benefits can encourage reciprocal action by other nations,
leading ultimately to international cooperation that increases both global and U.S. net benefits
relative to what could be achieved if each nation considered only its own domestic costs and
benefits when determining its climate policies.
Further, as explained in the 2010 TSD, from a technical perspective, the development of a
domestic SC-CO2 was greatly complicated by the relatively few region- or country-specific
30 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.

4-68

estimates of the SC-CO2 in the literature, and impacts beyond our borders have spillover effects
on the United States, particularly in the areas of national security, international trade, and public
health. As a result, it was only possible to include an "approximate, provisional, and highly
speculative" range of 7 to 23 percent for the share of domestic benefits in the 2010 TSD. This
range was based on two strands of evidence: direct domestic estimates resulting from the FUND
model, and an alternative approach under which the fraction of GDP lost due to climate change
is assumed to be similar across countries.
Regarding the comments on Clean Air Act Section 115 and the implications for use of a global
SC-CO2, EPA notes that these commenters reads into section 115 limitations and requirements
that do not appear in the statutory text.
Section 115(a) of the Clean Air Act requires the Administrator to give formal notification to the
Governor of a State in a very specific and limited circumstances. Specifically, the provision
provides: "Whenever the Administrator, upon receipt of reports, surveys or studies from any duly
constituted international agency has reason to believe that any air pollutant or pollutants emitted
in the United States cause or contribute to air pollution which may reasonably be anticipated to
endanger public health or welfare in a foreign country or whenever the Secretary of State
requests [her] to do so with respect to such pollution which the Secretary of State alleges is of
such a nature, the Administrator shall give formal notification thereof to the Governor of the
State in which such emissions originate." Section 115(b) further provides that notices issued
under 115(a) are deemed to be findings under section 110(a)(2)(H)(ii) of the Clean Air Act and
thus any state receiving such a notice must revise its state implementation plan to address the
offending emissions.
Section 115(c), in turn, provides that the section only applies "to a foreign country which the
Administrator determines has given the United States essentially the same rights with respect to
the prevention or control of air pollution occurring in that country as is given that country by this
section." In other words, unless the Administrator determines a foreign country has given the
United States the same rights section 115 gives to foreign countries, the Administrator is not
required issue any notification that would otherwise be required under 115(a). Contrary to the
commenter’s assertion, section 115 does not mention or refer to "reciprocity agreements."
Moreover, section 115(c) does not in any way limit the Administrator’s discretion to act under
other statutory provisions. It simply limits the reach of 115(a) by defining which countries are
covered by the requirement for the Administrator to give formal notification under that section
regarding pollution originating in the United States that affects health and welfare in a foreign
country.
There is also no support in section 115 or any provision of the Clean Air Act for the
commenter’s assertion that EPA is forbidden from counting harms or benefits occurring in other
countries without first taking action under section 115. Section 115 requires the Administrator to
issue a formal notification in the limited set of specific circumstances outlined above and
identifies the obligations that flow from the issuance of such notifications. It does not, as the
commenter suggests, include language prohibiting the EPA from considering the impact of U.S.
emissions on foreign countries in other contexts. Indeed, the United States has a long history of
working together with other countries to address air quality issues. As explained above, the
reciprocity provision in section 115(c) merely defines the applicability of section 115. Nothing
4-69

in the text of the section addresses the extent to which EPA may consider or address the impacts
beyond its borders that are related to air pollution originating in the United States, or suggests
that action under section 115 is a prerequisite to counting international harms and benefits in any
analysis.
Finally, there is no support for the assertion that section 115 requires EPA to list CO2 and/or
other GHGs as criteria pollutants and notify state Governors to submit plans under section 110
before taking any action to mitigate the impact of U.S. emissions on public health and welfare in
foreign countries or counting international benefits in other rulemaking contexts. Not only does
the limitation commenters identify simply not appear in the statutory text, but section 115(d)
explicitly retains certain prior recommendations with respect to pollutants for which NAAQS
have not been established. It would be absurd to interpret a section that explicitly retains
recommendations relating to the impact of non-NAAQS pollutants on foreign countries as
prohibiting the Administrator from considering such impacts unless a NAAQS has been
established for the pollutant.
Regarding the comment that EPA lacks authority to regulate GHG emissions outside the
fenceline, EPA notes that application of the SC-CO2 to estimate the social benefits of emission
reductions is entirely separate from regulating emissions. Conducting an economic analysis does
not confer any legal or regulatory obligations. See the preamble, including section III, for
discussion about the scope of the final rule and the legal justification.
As stated in the OMB Response to Comments on SC-CO2, the IWG concluded that the only way
to achieve an efficient allocation of resources for emissions reduction on a global basis is for all
countries to base their policies on global estimates of damages. The IWG has therefore
continued to recommend the use of global SC-CO2 estimates in regulatory impact analyses. EPA
agrees with the IWG conclusion and has used the global SC-CO2 estimates in the analyses of
illustrative costs and benefit scenarios in the RIA.

4.4.4 Uncertainty
Comment 4.4-4: Commenter 9396 states that for the electric sector and the commenting parties
associated with this filing, EPA's use of the SCC as a valuation tool in major rulemakings is
especially important for consideration. The IWG and the National Research Council (NRC) have
flagged increased uncertainty regarding the relevance and accuracy of the SCC as developed by
the IWG for the purposes of regulatory actions that derive the bulk of their benefits from
reductions in GHGs. Despite the serious limits of both quantification and monetization, SCC
estimates can be useful in estimating the social benefits of reducing carbon dioxide emissions.
Under Executive Order 12866, agencies are required, to the extent permitted by law, "to assess
both the costs and the benefits of the intended regulation and, recognizing that some costs and
benefits are difficult to quantify, propose or adopt a regulation only upon a reasoned
determination that the benefits of the intended regulation justify its costs." The purpose of the
SCC estimates presented here is to make it possible for agencies to incorporate the social
benefits from reducing carbon dioxide emissions into cost benefit analyses of regulatory actions
that have small, or marginal, impacts on cumulative global emissions. Most federal regulatory
actions can be expected to have marginal impacts on global emissions.
4-70

Commenter 9396 states the SCC as derived by the IAMs is an extremely subjective metric that
was developed to try and quantify the benefits of reducing only domestic GHG emissions. The
degree of subjectivity is best captured in the 2010 IWG report which states, "When attempting to
assess the incremental economic impacts of carbon dioxide emissions, the analyst faces a number
of serious challenges. A recent report from the National Academies of Science (NRC 2009)
points out that any assessment will suffer from uncertainty, speculation, and lack of information
about (1) future emissions of greenhouse gases, (2) the effects of past and future emissions on
the climate system, (3) the impact of changes in climate on the physical and biological
environment, and (4) the translation of these environmental impacts into economic damages. As
a result, any effort to quantify and monetize the harms associated with climate change will raise
serious questions of science, economics, and ethics and should be viewed as provisional." In
other words, as with any economic model, the outcome is only as reliable as the input. The IWG
made a number of assumptions that are questionable and worthy of a broader policy review
inclusive of public participation and public comment prior to finalization and application in the
regulatory development process. The IWG estimated benefits of reductions hundreds of years
into the future, well beyond any reasonably predictable time horizon. Any rational assessment of
future climate activity must ascribe an increasingly discounted view of future economic harm as
the level of uncertainty regarding the probable outcome becomes correspondingly uncertain.
Commenter 9681: First, as a matter of law and economics, uncertainty in benefits estimates does
not mean they should be excluded from regulatory impact analyses. No benefit or cost estimate is
certain. Further, the courts have explicitly rejected this argument with respect to the SCC, and
executive orders dating back as far as the Reagan administration have all issued guidelines
specifying explicit consideration of benefits even if the precise size of the benefit is uncertain. In
2008, the U.S. Court of Appeals for the Ninth Circuit determined that agencies could not assign a
zero dollar value to the social costs of the impacts of climate change. It determined that failing to
count SCC benefits would be illegal. Like the Court of Appeals, executive orders dating back to
1981 have also required agencies to assess benefits and costs even when significant uncertainty
exists. Every president since (and including) Ronald Reagan has issued directives requiring that
agencies conduct cost-benefit analyses of proposed regulations where permitted by statute.
Specifically, agencies are directed to take into account benefits and costs, both quantitative and
qualitative . . . and use the best available techniques to quantify anticipated present and future
benefits and costs as accurately as possible. In this case, the National Highway Traffic Safety
Administration (NHTSA) had decided not to count any avoided climate damages in issuing fuel
economy standards. The court concluded: NHTSAs reasoning is arbitrary and capricious for
several reasons. First while the record shows that there is a range of values, the value of carbon
emission reductions is certainly not zero. The court's decision directly contradicts the petitioners
argument that uncertainty renders the SCC invalid, as well as their demand that agencies use a
value of zero.
Commenter 9681 states the IWG used solid economic tools to address uncertainty and ought to
go further in capturing the full extent of its implications. The IWG was rigorous in addressing
uncertainty. First, it conducted Monte Carlo simulations over the IAMs specifying different
possible outcomes for climate sensitivity (represented by a Roe and Baker Distribution). It also
used five different emissions growth scenarios and three discount rates. Second, the IWG
reported the various moments and percentiles of the resulting SCC estimates. Third, the IWG put
4-71

in place an updating process, e.g., the 2013 revision, which updates the models as new
information becomes available. As such, the IWG used the various tools that economists have
developed over time to address the uncertainty inherent in estimating the economic cost of
pollution: reporting various measures of uncertainty, using Monte Carlo simulations, and
updating estimates as evolving research advances our knowledge of climate change. The Monte
Carlo framework took a step toward addressing what is probably the most concerning aspect of
climate change, the potential for catastrophic damages, i.e., low probability/high damage events.
These damages come from: uncertainty in the underlying parameters in IAMs, including the
climate sensitivity parameter; climate tipping point thresholds that, when crossed, cause rapid,
often irreversible changes in ecosystem characteristics; and black swan events which refer to
unknown unknowns. The analysis used a right-skewed distribution of temperature (as captured in
the Roe Baker climate sensitivity parameter) and an increasing, strictly convex damage function;
this correctly results in right-skewed distributions of damage and SCC estimates. By using the
mean values of these estimates instead of the median, IWG estimates partially captured the
effects of small probability, higher damages from high-level warming events, and as a result
come to the wrong conclusions. An everyday analogy might be if an individual trying to watch
his weight by dieting focused on median calorie intake per meal in a given month instead of the
mean: the median calorie meal would be the meal such that half of all meals had calories below
its value, and half above; the mean calorie meal would be the total number of calories in a month
divided by the number of meals eaten. Dieters often deviate from their plan by occasionally
having normal or excessive-calorie meals. If they focused on their median calorie intake, they'd
never count the high calorie deviations and would undermine their efforts to lose weight (or,
indeed, could even gain weight). Rather than the median, the mean would be the correct metric
to use because it would capture these infrequent, high-calorie breaks with the diet. Another
analogy is airplane safety regulation: safety is protected by guarding against the low-probability
but highly dangerous events. With climate change we do not have the luxury of knowing how
damaging the extremes could be; all we know is that there is a very real possibility they could be
devastating.) To reflect uncertainty in estimates resulting from the right-skewed distribution of
SCC estimates, the IWG reported the SCC value for the 95th percentile from the central 3%
discount rate distribution. This is done to reflect the estimation uncertainty in terms of the
possibility of higher-than-expected economic impacts from climate change. While the IAMs take
different approaches to explicitly modeling tipping points, which to a great extent is lacking in
current versions of FUND and DICE, the IWG improved (but in no way fixed) the representation
of uncertain catastrophic damages with the Monte Carlo analysis. Still, black swan events go
completely unaddressed in the IWG modeling framework, and to the extent that society has a
strong desire to prevent the occurrence of catastrophic events that is not reflected in the
estimates.
Commenter (8925) states the uncertainty in SCC estimates has not yet been fully characterized,
nor the sensitivity of the SCC fully evaluated. Commenter adds that this is true for uncertainties
that were considered in the USG methods, as well as uncertainties that were not considered.
Commenter continues that for many of the assumptions, the documentation does not discuss
alternatives and their potential implications, nor establish that the sensitivities currently
evaluated are representative and therefore justified. Commenter similarly states that a number of
uncertainties were not evaluated, including emissions and energy intensities and technological
change, carbon cycle modeling , radiative forcing , damage functions , adaptation, and 2300
extensions.(See comment for reference) Commenter concludes that without more comprehensive
4-72

consideration of uncertainty, it is difficult to know whether the current estimates are biased, or if
the uncertainty reflected in current estimates is sufficient and the estimates robust to alternative
assumptions and additional uncertainties.
One commenter 8925 states the robustness of the current USG SCC estimates has not been
established, or sufficiently evaluated.
Response 4.4-4: The comments regarding the consideration of uncertainty in SC-CO2 estimates
mirror those submitted to the Office of Management and Budget’s separate comment solicitation
on the SC-CO2 (78 FR 70586; November 26, 2013). As a member of the interagency working
group (IWG) on SC-CO2, EPA has carefully examined and evaluated comments submitted to
OMB’s separate solicitation. EPA has also carefully examined and evaluated all comments
received regarding SC-CO2 through this rulemaking process and determined that the IWG
responses to the comments on the OMB solicitation address the process comments summarized
in this section. Specifically, EPA concurs with the IWG’s response to these comments and
hereby incorporates them by reference.31
After careful evaluation of the full range of comments and associated technical issues described
in section 4.5, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SCCO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in RIA
Chapter 5.32 EPA will continue to consider these comments and will share the recommendations
with the IWG as it moves forward with the Academies process.
The remainder of this section elaborates on the uncertainty comments in the context of this
rulemaking.
EPA agrees with the comments that supported the rigor of its uncertainty analysis. Uncertainty is
inherent in all regulatory impact analysis. It is especially salient in regards to the SC-CO2
estimates because of their broad spatial and temporal dimensions. In addition to conducting
Monte Carlo analysis for a subset of key parameters, the EPA and the other members of the IWG
included extensive discussion of uncertainty in the TSDs, and the documentation for the
individual IAMs themselves contains additional discussion of the assumptions and uncertainties
in the models. The 2010 TSD and the updated 2013 TSD provided visual depictions of the
distributions of the SC-CO2 estimates in addition to detailed statistics including percentiles and
higher order moments. In addition to the extensive information provided in the TSDs, the IWG
has provided the full range of model results to outside researchers upon request and will continue
to do so.

31 Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 28 to 30 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf .
32 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.

4-73

Regarding comments on uncertainty in the context of the socioeconomic-emission scenarios, see
4.4-1.
The 2010 TSD reports SC-CO2 estimates out to one decimal place (i.e., at least two significant
digits) and the current TSD reports to the nearest dollar (i.e., two or three significant digits,
depending on the year and discount rate/statistic). EPA and the other members of the IWG chose
not to use decimal places in the 2013 TSD to avoid the impression of artificial precision but will
also explore presentation with a consistent number of significant digits. EPA and the other IWG
members welcome the recommendations to strengthen the characterization of uncertainty and
plan to seek external expert advice on the technical merits and challenges of potential approaches
to improve the characterization and analysis of uncertainty in future updates.
Regarding comments on the modeling of uncertainty in the SC-CO2 analysis, including the
uncertainty explicitly represented in the IAMs and whether it captures the full range of
uncertainty of the “true” value of the SC-CO2, see section 4.4- 8.
EPA disagrees with the comments stating that the SC-CO2 is neither useful nor appropriate
because of uncertainty, the broad range of estimates. EPA also disagrees that it is inappropriate
to try to value the impacts of reductions in CO2 emissions. EPA notes that in 2007, the Ninth
Circuit Court remanded a fuel economy rule to DOT for failing to monetize the benefits of the
CO2 emissions reductions in its regulatory impact analysis, noting that “the value of carbon
emissions reduction is certainly not zero.”
All regulatory impact analysis involves uncertainty. EPA acknowledges uncertainty in the SCCO2 estimates but disagrees that the uncertainty is so great as to undermine use of the SC-CO2
estimates in regulatory impact analysis. The uncertainty in the SC-CO2 estimates is fully
acknowledged and comprehensively discussed in the TSDs and supporting academic literature.
While uncertainty must be acknowledged and addressed in regulatory impact analyses, even an
uncertain analysis provides useful information to decision makers and the public. For example, if
an analysis shows that benefits of a policy option consistently do (or do not) justify costs even
over a broad range of estimates, this may increase confidence in the robustness of this
conclusion. Conversely, if choices among parameter estimates within a plausible range
significantly affect the conclusions of the analysis, this is an important consideration in deciding
how to weigh the analytical results in the decision making process. The presence of uncertainty
is thus not a reason to exclude the best available estimates of quantified/monetized benefits, as
long as it is appropriately characterized. Rather, good regulatory practice requires that agencies
use the best available scientific, technical and economic information to derive the best estimates
of costs and benefits that they can, and then communicate to the public the limitations and
uncertainties of the analyses. The RIA for this rulemaking and the SCC TSDs clearly discuss the
limitations and uncertainties of the analysis. As noted in the RIA and TSDs, EPA is committed to
periodic updates in the estimates to reflect ongoing developments in our understanding of the
science and economics of climate change, including the treatment of uncertainty.

4-74

4.4.5 Process-related Concerns
Comment 4.4-5: Many commenters either voiced support for or criticized particular aspects of
the interagency process used to develop the SC-CO2 estimates. Process issues raised included
consistency with applicable OMB guidance documents, peer review, transparency, opportunity
for public comment, and appropriate use of the estimates.
Numerous comments were made concerning peer review of the SC-CO2 and supporting models
(9685, 9396, 10618, 9194, 8966, 9423, 8032, 8925, 8966, 10095, 10662, 8032, 9194, 9422,
9195).
Commenter 8925 states the USG SCC modeling application has not been explicitly peer
reviewed, and two commenters (8925, 9592) state that the Administration should also consider
model-specific peer review. Commenter explains that there is a significant difference between
model peer review and peer review of an application of the model. Commenter adds that peer
review of all three models could substantiate whether or not all of the models used are of
equivalent caliber, and similar enough structurally. Commenter 9592 concludes that EPA should
withdraw the SCC from use in any rulemaking where the SCC is used to justify and regulation
until the peer review and public comment process has transpired.
Other comments regarding peer-review, transparency, and public comment included:

EPA, as a participant in the IWG, must subject the Technical Support Documents on the
Social Cost of Carbon to independent peer review as directed by the President and
mandated by the Information Quality Act (IQA) guidelines for Highly Influential
Assessments. President Obama has stressed the importance of adhering to established
scientific procedures, including independent peer review, when making policy decisions.
EPA has chosen to ignore this clear mandate in the instance of the SCC TSD's and the
use in rulemaking. In addition to general public comment, independent peer review of
scientific assessments is essential including: the models and the formulations of key
drivers of the models including emissions, CO2 concentrations, temperature response to
the natural forcings, sea level rise, and the details of the elements of "damage function" is
required.

The EPA should subject the Interagency Working Group's Social Cost of Carbon TSD to
independent peer review before it can use the SCC any rulemaking. The principles for
this process are stipulated in the Information Quality Act (signed by President Clinton in
2000) implemented through the 2005 guidelines for peer review. The OMB guidance
recognizes that OMB is responsible for overseeing compliance with its rules regarding
the quality of disseminated information, which include the peer review guidance. This is
a statutory responsibility under 44 U.S.C. Section 3504(d)(1), which is incorporated by
reference into the IQA. The OMB peer review guidance states "OIRA [of OMB], in
consultation with OSTP, shall be responsible for overseeing implementation of this
Bulletin."

OIRA could not approve any final rule using the SCC without ensuring that its peer
review guidance has been fully implemented. At this time, it would be impossible for
4-75

OIRA to do so until EPA has completed the required peer review. The SCC assessment is
a "highly influential scientific assessment" ("HISA") under the definition of a HISA in
the 2005 OMB IQA peer review guidance. The OMB guidance contains "requirements"
for the conduct of impartial outside peer reviews of highly influential scientific
assessments. The U.S. Court of Appeals for the D.C. Circuit determined in Prime Time
Int'l, Inc. v. Vilsack (599 F.3d 678, 685 (D.C. Cir. 2010)) that OMB guidance issued
pursuant to the Information Quality Act is legally binding. Other legal precedents
confirm that the peer review guidance is a legally binding and enforceable "legislative
rule."

The provisions for HISAs are mandatory. A key term in the HISA definition concerns the
meaning of the term "assessments." The OMB guidance also distinguishes between what
is discretionary and what is mandatory, based on use of the terms "requirements," "shall,"
and "must." The term "scientific assessment" is defined in the OMB guidance as
requiring a synthesis of various inputs. The OMB guidance definition states that it means
"an evaluation of a body of scientific or technical knowledge, which typically synthesizes
multiple factual inputs, data, models, assumptions, and/or applies best professional
judgment to bridge uncertainties in the available information." The definition expressly
includes "model assessments."

The OMB guidance contains numerous mandatory "requirements" for what must be done.
The OMB peer review guidance has been recognized as mandatory by the EPA Inspector
General. On September 26, 2011, the Inspector General transmitted a report focusing on
whether EPA had complied with OMB's and its own peer review requirements in issuing
its greenhouse gas endangerment finding. The Inspector General concluded that EPA had
not complied with OMB's peer review "requirements" for HISAs in certain respects. He
also recommended that EPA revise its internal PEER REVIEW HANDBOOK "to
accurately reflect OMB requirements..." Unfortunately EPA has chosen to ignore the IG's
clear directive in its application of the SCC. All HISA's (highly influential scientific
assessment(s)) must be peer reviewed independent of the sponsoring agency. Public
comment on a rulemaking proposal is not a substitute for peer review. EPA must instruct
the peer reviewers to prepare a peer review report that will be made public. In the case of
HISAs, EPA must prepare a written response to the report. The IAMS have not been
subjected to the required "Highly Influential Scientific Assessment" peer review.

Even though the new versions of the three integrated assessment models (IAM) used by
the U.S. government to estimate the SCC (DICE, FUND, and PAGE) were published in
the "peer reviewed literature", the IQA requires a higher standard of peer review as
described above. The fact that all of the models have been "peer reviewed" does not mean
that any of them have been deemed to have any skill for predicting future emissions,
temperatures, sea level rise or damages. EPA apparently has assumed that "new" means
"better". EPA surely knows this is not the case in terms of the IPCC models (AR4, AR5)
that are used to provide the critical climate sensitivity input. The "appearance" in
economics journals is not adequate since the main drivers of the impacts are related to
science and technology issues, not economics. The competence of the journal review
should have been questioned by EPA and the requirements of the IQA followed.

4-76

SCC2013 acknowledged the continued limitations of the approach taken by the
interagency group in 2010 (documented in the original 2010 TSD). SCC2013 did not
revisit other assumptions with regard to the discount rate, reference case socioeconomic
and emission scenarios, or equilibrium climate sensitivity. This was another major failure
of EPA. Many additional peer review papers were available and should have been
considered in the SCC2013 update. The developers of the models have an inherent
conflict and are not acceptable judges of the adequacy of the peer review for their
"updates".

The issue of public access to documentation and data is also a serious concern. All of the
model documentation, formulations based on specific papers or interpretations, validation
data and results, model code and cited papers should be available on a public website so
that the public can examine the material and independently confirm the calculations of
the models and use alternative input. As EPA knows, this is not the case. EPA has
established the CREM website for all of the models it uses in regulatory analysis. The
IAMs are missing. EPA claims that the IAMs were "updated" and "used" in peerreviewed literature. However there is no documentation that the IAMs used in either SCC
2010 or SCC 2013/2013rev were ever independently peer-reviewed and validated as
required by OMB's legally binding guidelines. EPA should immediately terminate this
proceeding until the current OMB process of comment, response and resolution on
Docket OMB-2013:0007-0001 has been completed.

The SCC fails in terms of process and transparency. The SCC is the product of a closed
secretive process. The SCC fails to comply with OMB's guidance for developing
influential policy-relevant information under the Information Quality Act, especially for a
Highly Influential Scientific Assessment (HISA). The Integrated Assessment Models
(IAMs) with inputs used for the SCC analysis were not subject to independent peer
review. The results of the SCC are fatally flawed as a result of significant cumulative
errors.

Some commenters stated that the modeling used does not provide a sufficient range of accuracy
for policy-making purposes. OMB and the IWG failed to use the latest science which indicates
that the equilibrium climate sensitivity (ECS) distributions used by the modelers caused the SCC
to be over-stated. The models employed and the assumptions used were not subjected to peer
review nor were the 2010 and 2013 SCC estimates developed by the IWG.
To meet to its obligations under the Information Quality Act guidelines and President Obama's
directive on Transparency, EPA should promptly post on regulations.gov the following
information: 1. Initial and all subsequent drafts of the SCC2010, SCC2013 and SCC2013r
reports, 2. All comments provided on each draft and documentation on how the comments were
resolved, 3. The names of the individuals responsible for the drafts and the resolution of the
comments, 4. Full documentation of every version of the Integrated Assessment Models (DICE,
FUND, PAGE), including: a. The versions of the model used in SCC 2010, SCC 2013 and SCC
2013r, b. The input to each run used in the analysis, c. The output of each model run, including
the emissions as a function of time, the concentrations as a function of time, the temperature as a
function of time, and the sea level rise as a function of temperature, d. Complete documentation
of the formulas programmed in the models, e. Complete user instructions that will allow the
4-77

public to run the models and replicate the results, f. All data used to validate the models, and g.
All validation run results. 5. The journal articles where each version of the IAMs used was "peer
reviewed", and a. All cites to articles in which the results of the models were used (if any articles
appear in pay-walled journals arrangements must be made for free public access), and b. All
comments by EPA and any other USG agencies, employees, or contractors made as either peer
reviewers or potential users.
Commenter 9396 states use of the Social Cost of Carbon (SCC) in EPA Rulemaking violates
statutory administrative requirements, Administration guidelines and the norms of peer review.
The SCC is a metric developed by an Inter-agency Working Group (IWG) to place a cost on CO2
emissions. The IWG defines the SCC as an estimate of the monetized damages associated with
an incremental increase in carbon emissions in a given year. The IWG issued an initial estimate
of the SCC in 2010 without notice and public comment. In May of 2013, the IWG, without
notice and public comment, adjusted the calculations used to account for the SCC. The change
implemented in the TSD 2013 revision substantially increased the cost estimate to a degree that
will have an identifiable and meaningful impact on future regulatory decisions. The IWG held no
public meetings and no public input was solicited prior to either the 2010 or 2013 publication of
the SCC.
Commenter 9396 Commenter's concerns with the SCC are outlined in detail later in these
comments. It is critical that the models, assumptions used in the models and the entire IWG
administrative process, including access to all data and analyses used, be opened to all
stakeholders for comment and review and that the entire process fully comply with Executive
Branch guidelines, the Administrative Procedures Act (APA) and the Data Quality Act (DQA.)
Commenter 9396 states that the development of the Social Cost of Carbon violates the
Administrative Procedures Act (APA) and the Information Quality Act (IQA) and cannot be
used by the EPA in this rulemaking. Uncertainties with the models and assumptions used to
develop the SCC render the SCC inappropriate for policy and rulemaking. EPA refers to the
social cost of carbon (SCC) in the Regulatory Impact Analysis (RIA) for this proposed rule. The
SCC is a metric developed by an Inter-agency Working Group (IWG) formed by the
Administration, headed by the Office of Management and Budget (OMB) and the Council of
Economic Advisors (CEA), to place a cost on CO2 emissions. According to an EPA letter to
ranking member of the Senate Environment and Public Works Committee (EPW), it appears that
much, if not all, of the analysis developing the SCC was done by EPA. The IWG defines the
SCC as an estimate of the monetized damages associated with an incremental increase in carbon
emissions in a given year. The same report goes on to state that [i]t is intended to include (but is
not limited to) changes in net agricultural productivity, human health, property damages from
increased flood risk, and the value of ecosystem services. In this rulemaking, EPA used the SCC
as one of its screens to justify the designation of partial CCS as BSER for new coal plants. In
response to an IQA petition submitted to OMB on September 4, 2013 by a group of trade
associations, OMB solicited public comment on the SCC through a public comment process that
applies to all Federal rulemakings and indicated that the SCC was intended for use by all
agencies measuring the social cost of carbon emissions as they evaluate costs and benefits of
rules. Thus OMB identified development of the SCC as a rulemaking to be used in other
rulemakings.

4-78

The Interagency Working Group's (IWG) use of the Integrated Assessment Models (IAMs) do
not meet the IQA standards for policy-making including utility, transparency, and robustness.
Commenters 9685 and 9687 raised issue with EPA's use of the Social Cost of Carbon
(SCC). The commenters believe that E.O. 12866's mandate to perform a transparent cost-benefit
analysis was not met with EPA's significant errors using the SCC in its regulatory analysis. The
commenters stated that EPA violated E.O. 12866 and offer a numbered list of requirements that
EPA must comply with prior to issuing any SCC technical support document or proposing any
rule pursuant to E.O. 12866 that include the following:
1. EPA must post a peer review plan in the peer review Agenda on its Science Inventory
and establish a mechanism whereby the public can comment on the plan. It should also
consider inviting public and/or professional society nominations for the independent peer
review group. EPA should post a draft charge to the peer reviewers and request comment
on the draft charge.
2. EPA must select a group of peer reviewers. The selected peer reviewers must be
independent from the government, be impartial experts, and be sufficient in number to
fairly and impartially represent expertise in science, technology, economics, statistics,
and modeling. The peer reviewers must be free of any conflict of interest. Selection
should be in conformance with National Academy procedures, as indicated in the OMB
guidance.
3. EPA must develop a charge for the peer reviewers that instructs them to address all the
assessment issues (including costs) reflected in the SCC, while avoiding allowing any
policy considerations that would bias their deliberations and conclusions to intrude. The
charge must also include instruction on the information quality standards in OMB's 2002
general IQA government-wide guidance. The agency (either in the charge or otherwise)
must instruct the peer reviewers to prepare a written report stating their conclusions in
response to the charge, and advise them that the report will be made public.
4. EPA must give the peer reviewers and the public access to all materials relevant to their
review, including the ability to run the models.
5. EPA must sponsor at least one public meeting of the peer review group at which
interested members of the public can make oral comments. It must also provide an
opportunity for written public comments. OMB must ensure that the peer reviewers have
access to all significant public comments on the scientific/technology/economics
evaluation issues.
6. Upon completion of the peer review, EPA must make available to the public the peer
reviewers' report, the charge, and the reviewers' names, affiliations, and areas of
expertise.
7. EPA must provide a written response to the peer review report in which it discusses
whether and why it agrees or disagrees with its conclusions.
EPA must require that any agency include in the administrative record for any final rule that
relies on the SCC a "certification" explaining how the agency has complied with the OMB peer
review guidance and the applicable (2002) general information quality guidelines. Other
commenters expressed support in a joint letter (9681) for the process used to develop the SCC
estimates:

4-79

The IWG's choice of three IAMs was fully justified but should still be revisited in its next
iteration. In its calculations of the SCC, the IWG relied on the three Integrated
Assessment Models (IAMs) available at the time, all with a long record of peer-reviewed
publications that link physical and economic effects: the Dynamic Integrated Model of
Climate and the Economy (DICE), the Climate Framework for Uncertainty, Negotiation,
and Distribution (FUND), and Policy Analysis of the Greenhouse Effect (PAGE). The
government's first SCC estimates, published in 2010, used the then-current versions of
the models; the recent update employed revised, peer-reviewed versions of the models
but maintained the underlying assumptions of the 2010 IWG analysis. As stated by the
2010 IWG, the main objective of [the 2010 IWG modeling] process was to develop a
range of SCC values using a defensible set of input assumptions grounded in the existing
scientific and economic literatures. They represent the state-of-the-art IAMs. Each of
these models has been developed over decades of research, and has been subject to
rigorous peer review, documented in the published literature.
The analytic work of the IWG has been transparent. The 2010 Technical Support Document
(TSD) set out in detail the IWG's decision-making process with respect to how it assessed and
employed the models. Because the 2013 IWG made no changes to the input assumptions and
procedures for deriving its SCC estimates, the 2013 TSD discusses only how the three Integrated
Assessment Models (IAMs) used in the analysis were updated in the academic literature over the
three-year interim period by the independent researchers who have developed these models. The
2013 TSD also establishes that the increase in the SCC estimate from 2010 to 2013 resulted
solely from updates to the three underlying IAMs.
The OMB recently created a further opportunity for public comments and transparent review
specifically on the calculation of the social cost of carbon. Additionally, the comment period on
these greenhouse gas proposed performance standards for power plants are yet another
opportunity for continued dialogue about areas requiring further study. Such repeated comment
processes demonstrate that the IWG's SCC estimates were developed and are being used
transparently. Given their strong grounding in the best science available, nothing should prevent
the current, continued use of this well-established estimate. As economic and scientific research
continues to develop, future revisions will be able to further refine existing estimates based on
the latest peer-reviewed literature and the latest updates to the quality of the overall modeling
exercise. Commenter contends that the SCC is an important and accepted tool for regulatory
policy-making based on well-established law and fundamental economics. Despite recent attacks
from fossil fuel and other industry groups, the legal and analytic basis for using the SCC is clear
and well established. Several industry groups recently submitted a petition to OMB and the
individual agencies involved in the IWG requesting the prohibition of the use of the 2010 and
2013 SCC estimates by the government and the withdrawal of the supporting documents that
describe the calculation of the SCC. In addition some of the same parties recently weighed in on
the use of the SCC in a Department of Energy rulemaking. In particular, petitioners contend: 1)
that estimates of the SCC are purportedly too uncertain and imprecise to be used for regulatory
analysis. The petitioners argue that alleged uncertainty and imprecision renders the SCC an
illegitimate tool for use in regulatory impact analyses. They discuss a number of technical
methodological approaches reflected in SCC models used by the IWG, such as the time horizon
of the analysis and the choices of which direct and indirect benefits to include, and mistakenly
conclude that the limitations of SCC estimates preclude the IWG's use of them. Petitioners
4-80

misrepresent both the view of the Intergovernmental Panel on Climate Change (IPCC) and
economists own assessments of SCC limitations. 2) That the IWG supposedly did not follow
appropriate protocols in either its estimation or use of the SCC. The petitioners claim that the
IWG failed to sufficiently describe the uncertainties in its analysis; made inappropriate technical
choices with respect to discount rates and the choice of a global rather than domestic SCC; failed
to go beyond the extensive peer review of the underlying models and have its use of the models
peer reviewed; and denied stakeholders proper notice or opportunity to comment on the IWG's
use of updated versions of the underlying models. 3) In light of these two claims, these industry
critics argue that the SCC must be banned from use while it undergoes a wholly new public
review process. They assert that the IWG's SCC estimates are arbitrary and capricious, and likely
overestimated. While the SCC would be under re-review, the petitioners request that OMB
require agencies to assign a value of zero to carbon pollution damages. The call for best available
techniques addresses all of the petitioners’ specific comments around appropriate time periods
and other modeling choices: The IWG's use of Integrated Assessment Models (IAMs) reflects
the best available, peer-reviewed science to tally the benefits and costs of specific regulations
with impacts on carbon dioxide emissions. The IAMs include benefits and costs that have been
quantified to date. Other modeling assumptions represent reasonable choices on the part of the
IWG. For example, cutting off the analysis after around 300 years simply reflects the fact that
any benefits and costs thereafter are de minimis given the strong impact discounting has on
reducing the value of distant effects in the SCC. Similarly, petitioners assert that assumptions
around socio-economic pathways do not properly reflect the impact of current regulations and,
therefore, lead to double counting. The IWG should indeed periodically update its choice of
socio-economic scenarios to reflect major changes in economic growth projections and other
factors influencing emissions trajectories. However, the long-term emissions impact of any one
particular energy efficiency or other carbon reduction standard is only marginal to the trajectory
of socio-economic scenarios, minimizing any risk of double counting the regulatory benefits.
Second, while agencies are expected to use the best available science to inform their regulatory
impact analyses, there is no legal basis for demanding that an agency's analysis itself undergo
academic peer review. What is required is that agencies undertaking rulemakings provide public
notice and an opportunity for public comment on their analyses, and respond to those comments.
The petitioners have had multiple opportunities to comment on the IWG's analysis and will have
additional opportunities to do so in future rulemakings. The IWG's cost of carbon estimates have
been referenced in more than forty rulemakings to date, and agencies have responded to relevant
comments submitted thus far. Indeed, many of the changes the IWG made in its updated
methodology were in response to such comments. This comment period from EPA provides yet
another opportunity for feedback. The bottom line is that the IWG has properly and lawfully
used the best available techniques to quantify the benefits of carbon emission reductions, basing
its analysis on the peer-reviewed literature. When agencies use the IWG's estimates of the SCC
to calculate the benefits of a rulemaking, they have taken, and will continue to take, comment on
the SCC and the process used to derive that value. That is what the law and good policy requires.
The petitioners offer no alternative estimation procedure. They simply ask OMB and rulemaking
agencies to contravene the Ninth Circuit's decision by substituting zero for the IWG's best
estimate of the costs of carbon pollution. This should be seen for the cynical tactic that it is: an
attempt to delay any action to cut carbon pollution.
Commenter 9681 continues the IWG's analytic process was science-based, open, and transparent.
To facilitate accounting for the costs of climate impacts and the benefits of reducing carbon
4-81

pollution in regulatory proceedings undertaken by different agencies, the United States
government assembled the IWG in 2010 to develop an estimate of a social cost of carbon that
can be utilized in rulemakings and other pertinent settings across the federal government.5 The
2010 estimate of carbon pollution reduction benefits has been used in several benefit-cost
analyses related to federal rulemakings. The IWG recently released an updated set of SCC
estimates, centered at approximately $37 per metric ton of CO2 for emissions in the year 2015, in
2007 dollars at a 3% discount rate. The 2013 SCC estimates are higher than those from 2010,
reflecting the growing understanding of the costs that climate impacts will impose on society.
(The 2010 central estimate for 2015 was $24/ton.) The increase in the SCC estimate is important
because it reflects the growing scientific and economic research on the risks and costs of climate
change, but is still very likely an underestimate of the economic cost of carbon emissions. The
increase is also necessary given the costs of climate change that we are already experiencing,
such as those associated with sea level rise and rising temperatures. Climate change is making
coastal flooding, drought, and impacts from extreme weather worse.
Commenter 9194 states the SCC calculation ignores requirements of OMB's Circular A-4 to
calculate both benefits and costs for regulatory purposes.
Commenter 9681 states they strongly affirm that the current Social Cost of Carbon (SCC) values
are sufficiently robust and accurate to continue to be the basis for regulatory analysis going
forward. As demonstrated below, if anything, current values are significant underestimates of the
SCC. As economic and scientific research continues to develop in the future, the value should be
revised, and we also offer recommendations for that future revision. Presidential orders on
regulatory analysis also support use of a global SCC value. In 2012, President Obama issued
Executive Order 13609 on promoting international regulatory cooperation. The Order built on
his previous Executive Order 13563, which in turn had affirmed its 1993 predecessor, Executive
Order 12866, in requiring benefit-cost analysis of significant federal regulations. Though White
House guidance published in 2003 on regulatory impact analysis under E.O. 12866 assumed that
most analyses would focus on domestic costs and benefits, it ultimately deferred to the discretion
of regulatory agencies on whether to evaluate "effects beyond the borders of the United States."
More importantly, since the publication of that guidance, President Obama has issued his own
supplemental orders on regulatory analysis, including E.O. 13609, which clarified the
importance of international cooperation to achieve U.S. regulatory goals. This 2012 order
explicitly recognizes that significant regulations can have "significant international impacts," and
it calls on federal agencies to work toward "best practices for international regulatory
cooperation with respect to regulatory development." By employing a global SCC value in U.S.
regulatory development, and by encouraging other countries to follow that best practice and
account for the significant international impacts of their own climate policies, federal agencies
will advance the mission of this presidential order on regulatory harmonization.
Commenter 9194 states the IWG/OMB process is fatally flawed in that it fails to calculate the
SCC using OMB's prescribed "base case" discount rate of 7% and impermissibly calculates
global benefits instead of domestic benefits. Commenter 9194-6571 continues the SCC
calculation ignores requirements of OMB's Circular A-4 to calculate both benefits and costs for
regulatory purposes. Also, Commenter 9194-6573 states that rulemaking is subject to the
Administrative Procedures Act (APA) which requires public notice and identification of the legal
authority for the rulemaking. OMB has provided no identification of a statutory authority or
4-82

basis for development of the SCC for use in all federal rulemaking. Use of the SCC in this rule
making will constitute a violation of the APA and the Information Quality Act (IQA).
Commenter 9396 states the SCC estimates are considered under the premise that U.S. citizens
will forego GHG generating activities (wealth creation/GDP) in the present to potentially
forestall global change for the benefit of future global generations 100-300 years into the future.
For the purposes of environmental review, the increased estimated benefit of GHG emission
reductions helps justify the cost of new regulations by increasing the presumed benefits of the
regulation. The IWG issued an initial estimate of the SCC in 2010 without notice and public
comment. In May of 2013, the IWG, without notice and public comment, adjusted the
calculations used to account for the SCC. The change implemented in the TSD 2013 revision33
substantially increased the cost estimate to a degree that will have an identifiable and meaningful
impact on future regulatory decisions. The IWG held no public meetings and no public input was
solicited prior to publication of either SCC. Efforts by non-governmental entities to secure
information regarding which officials served on the IWG were thwarted and virtually no
information regarding the meetings held by the IWG, meeting minutes, background materials,
underlying data or preliminary analyses have been made available to the public. Indeed, at this
time, the public has not even been advised who actually attended the meetings. The SCC
estimate adjustments made for 2013 are much higher than the ones reported in the 2010 TSD. By
way of comparison, the four 2020 SCC (measured as the benefit per metric ton of CO2
emissions) estimates reported in the 2010 TSD were $7, $26, $42 and $81 using 2007 dollar
values. The updated SCC estimates for 2020 now are almost double that of the 2010 TSD. The
updated numbers for 2020 are $12, $43, $65 and $129 using 2007 dollar values. Each of these
numbers ascribes a deemed damage function to each ton of CO2 emitted into the atmosphere in a
manner that ascribes a level of certainty that far exceeds what the Integrated Assessment Models
(IAM) actually deliver. This unfounded level of certainty is further exacerbated by the ascription
of a single number based on averaging these numbers to develop the single number of $39 in
2013. Commenter's major concern is how the SCC will be used to influence decisions regarding
energy generation options in the future. In the table below by using the IWG SCC median value,
a typical coal-fired power plant would have an annual social cost of over $210 million, with over
$154 million attributed to CO2 based on the SCC.75 Even with an environmentally preferred
NGCC power plant, the annual social cost is over $74 million with $73 million of that cost
coming from the SCC. That is a startling 98% of the social cost being attributed to CO2 and the
SCC. The question is, does this seem reasonable?
Commenter 9396 states the decisions rendered by the IWG regarding how best to account for
these variables result in outcomes that are significant for the regulated industry. As Resources for
the Future explained in a recent White Paper on the SCC: The impact of SCC numbers is not
theoretical and has consequences for the government regulatory process and therefore for the
strength of regulations on climate change that emerge from it. Application of this tool can be
problematic to achieving optimum outcomes for society. A growing literature indicates that
developing the SCC requires assumptions that go well beyond the usual boundaries of science or

33 TSD May 2013 found at
http://www.whitehouse.gov/sites/default/files/omb/inforeg/social_cost_of_carbon_for_ria_2013_update.pdf.

4-83

economics. It requires many judgment calls that are hidden in complex economic models and
largely invisible to policymakers and stakeholders.
Commenters 10098, 8925, 8925, 8925, 10087 state generally that the SCC should not be used in
any rulemaking procedure due to issues that include lack of transparency in formulating the SCC
estimates, reliance upon modeling with inputs that lacked peer review, the failure to disclose and
quantify key uncertainties involved in the modeling, and the use of a global benefit estimate that
severely limits the SCC's utility. Commenter 8925 states the description and discussion of the
2013 revisions is insufficient for understanding and interpreting the changes in the USG SCC
results and that additional justification for many of the revisions would be helpful (e.g., space
heating expenditure reductions, transient temperature responses, constant indirect methane
radiative forcing effect, saturation, regional scaling factors, probability of a discontinuity,
adaptation, CO2 absorption, regional climate modeling). Commenter continues that discussion of
uncertainty would be helpful, and it would be reasonable to explain differences across models for
similar components (e.g., sea level rise, transient temperature response, saturation, adaptation).
Commenter closes stating it is unclear which revised elements were peer reviewed, and that a
clear citation for each of the revisions in each model would be useful. Commenter 8925 states
the SCC TSDs methods are not fully clear, basic information is lacking, and commenter
requests additional detail and discussion regarding the following: PAGE and FUND
probabilistic methods and their differences; undiscounted damages over time from each model
and for different scenarios; 2013 vs. 2010 GDP loss curves with respect to temperature (i.e.,
similar to Fig. 1a in USG, 2010); regional 2300 extensions; temperature projections to 2100 and
2300; and key differences across models and their implications. Commenter further states it
would be valuable to make all the results available for download, that information critical to the
results should appear in the TSDs, and that the documentation should be written to facilitate peer
review.
Response 4.4-5: The comments regarding the process issues mirror those submitted to the Office
of Management and Budget’s separate comment solicitation on the SC-CO2 (78 FR 70586;
November 26, 2013). As a member of the interagency working group (IWG) on SC-CO2, EPA
has carefully examined and evaluated comments submitted to OMB’s separate solicitation. EPA
has also carefully examined and evaluated all comments received regarding SC-CO2 through this
rulemaking process and determined that the IWG responses to the comments on the OMB
solicitation address the process comments summarized in this section. Specifically, EPA concurs
with the IWG’s response to these comments and hereby incorporates them by refer ence.34
After careful evaluation of the full range of comments and associated technical issues described
in section 4.4, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SCCO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in RIA
34 Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 34 to 41 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf.

4-84

Chapter 5.35 EPA will continue to consider these comments and will share the recommendations
with the IWG as it moves forward with the Academies process.
See section 4.1 for EPA’s detailed response to comments on climate change science and section
4.4-3 EPA’s detailed response to comments on the use of a global versus domestic SC-CO2. See
section 4.4-7, for EPA’s detailed response to comments about the selection of the discount rate.
See also EPA’s detailed response to comments on the IAMs and damage functions in 4.4-1,
EPA’s response to comments on uncertainty in 4.4-4, and EPA’s response to comments
regarding the application of the SC-CO2 estimates to this rulemaking in 4.4-8.
The remainder of this section elaborates on the process comments in the context of this
rulemaking.
Legal/statutory authority for issuing the SC-CO2 estimates
EPA agrees with commenters who believe that federal agencies have legal authority to develop
these estimates, and notes that the IWG was cognizant of the Ninth Circuit decision referenced
by these commenters when it decided to do so. EPA does not agree that the IWG’s issuance of
the SC-CO2 estimates constituted an APA rulemaking. As noted in the OMB Response to
Comments on SC-CO2, the IWG did not agree with comments that issuance of the SCC estimates
constitutes an APA rule making (page 34). The APA definition of a rule is "an agency statement
of general or particular applicability and future effect designed to implement, interpret, or
prescribe law or policy." The SCC estimates are not designed to implement, interpret, or
prescribe law or policy. Rather they are intended to provide guidance to agencies on a sciencebased methodology for estimating the benefits of CO2 reductions in regulatory impact analysis.
EPA notes that OMB has long-established authority to oversee the regulatory review process,
including preparation of regulatory impact analyses. OMB’s authority in this area is contained in
Executive Orders 12866 and 13563, among others, and has been acknowledged by Congress in a
series of statutes, including the Small Business Regulatory Enforcement and Fairness Act, the
Congressional Review Act, the Information Quality Act, and the Regulatory Right to Know Act.
It is fully consistent with this authority for OMB to offer guidance to agencies on best practices
for conducting regulatory impact analysis, as it did, for example, in issuing Circular A-4. In the
case of the SC-CO2 estimates and underlying TSDs, OMB determined that it was appropriate to
exercise this authority through a consensus-based process involving a broad range of agencies
that may issue rules affecting CO2 emissions.
The TSDs explain in detail the factual and policy basis for all of the various methodological
choices involved in developing the SC-CO2 estimates. The IAMs are documented in the
scientific literature. In addition, EPA has assisted interested members of the public in obtaining
additional information on the workings of the models. It has also provided full technical details
of its own use of the models, including output of model runs, to interested parties upon request.
However, because the SC-CO2 estimates are not a rule, as a legal matter they are not subject to
the arbitrary and capricious standard of the APA.
35 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.

4-85

Consistency with applicable OMB guidance documents, transparency, opportunity for public
comment
EPA agrees with the commenters that concluded the analytic work of the IWG has been
transparent and provided clear examples in support of this conclusion. As discussed below, EPA
disagrees with the comments that (1) the process to develop the SC-CO2 was non-transparent and
(2) that this process was inconsistent with the Administration’s goals for transparency in
scientific information.
Furthermore, EPA strongly disagrees with the claim that the Agency has not provided
information necessary to understand the SC-CO2 estimates and that it "deprives the IWG, OMB
and other executive rulemaking agencies of the benefit of any public input on the critical
decisions, inputs, and assumptions that form the basis of the SCC estimates." First, EPA notes
that the estimates and supporting TSDs are developed and issued through a consensus based
process involving the entire working group. Second, EPA has assisted individual requestors in
obtaining more detailed information about the modeling. For example, one requestor noted
publically that the EPA "modelers have been very open, collegial, and helpful."36 EPA has also
responded to numerous letters from members of Congress regarding the SC-CO2 estimates as
well as requests under FOIA.37
In particular, EPA believes that the process was inclusive, transparent, and appropriately
considered public input. The TSDs fully document the methodology used to develop the
estimates and the considerations that led the IWG to adopt this methodology. While the details of
the IWG’s internal processes were not discussed in the TSDs, this is common for most
government (and non-government) documents. Such details were not considered germane to the
public’s understanding of the SC-CO2 estimates and the methodology used to produce them, but
a general overview is provided here.
In developing the 2010 estimates, the EPA and other members of the IWG met frequently in the
year preceding the release of the February 2010 TSD. For the 2013 update, only a few meetings
were needed because the group decided to make no changes beyond incorporating the most
recent versions of the IAMs. IWG members, in particular professional economic staff with
modeling expertise oversaw the primary modeling and calculations for both the 2010 and 2013
SCC estimates using the most recent versions of the three IAMs available at the time. To develop
the 2010 estimates, the staff members ran two of the three models and contracted with the
developer of the third model to perform those runs. The contractor did not participate in any of
the interagency meetings but rather received instructions for how to conduct the model runs (e.g.,
specification of the three sets of input assumptions as determined by the working group). The
staff members ran all three models to develop the 2013 estimates. Decision making for both the
2010 and 2013 processes was by consensus of IWG members. The details of internal discussions
are deliberative, but the discussions were generally technical in nature and the issues discussed
36

See http://dailysignal.com/2013/11/06/white-house-reopens-the-scc/.
For example, January 16, 2014 letter from Joel Beauvais, Associate Administrator, to Senator Vitter. EPA also
responded to FOIA requests submitted by Sidley and Austin (EPA-HQ-2013-009517, EPA-HQ-2013-009518, EPAHQ-2013-009523).
37

4-86

and conclusions reached are well documented in the TSDs. Regarding the transparency of the
underlying models themselves, the IWG notes that they are well documented in the academic
literature.
The Government Accountability Office (GAO) also completed a review of the process used to
develop the SCC estimates. In its 2014 report, GAO concluded that according to IWG
participants, all major issues discussed were documented in the TSDs, which is consistent with
Federal standards for internal control, and the processes and methods used were based on the
principles of (1) consensus-based decision-making, (2) reliance on existing academic literature
and modeling, and (3) disclosure of limitations and incorporation of new information through
consideration of public comments and revision of the estimates as updated research became
available.
Regarding the explanation in the TSD for the 2013 revisions, no changes were made to the input
assumptions developed by the IWG between the 2010 and 2013 estimates. The only changes
were those made by the model developers themselves to the underlying models, which are
documented in the academic literature. To assist the public in understanding these changes, the
2013 TSD provides a brief summary of the most important ones, as well as references to the
relevant literature where more detailed information can be found. As with the 2010 TSD, the
IWG did not attempt to evaluate the modeling choices made by the modelers. As noted in the
OMB Response to Comments, by selecting the three “most widely used impact assessment
models” (NAS, 2010), the IWG intended to reflect a reasonable range of modeling choices and
approaches that collectively reflect the current literature on the estimation of damages from CO2
emissions. As explained in the 2010 TSD:
The parameters and assumptions embedded in the three models vary widely. A key
objective of the interagency process was to enable a consistent exploration of the three
models while respecting the different approaches to quantifying damages taken by the
key modelers in the field. An extensive review of the literature was conducted to select
three sets of input parameters for these models: climate sensitivity, socio-economic and
emissions trajectories, and discount rates…All other model features were left unchanged,
relying on the model developers’ best estimates and judgments…The sensitivity of the
results to other aspects of the models (e.g. the carbon cycle or damage function) is also
important to explore in the context of future revisions to the SCC but has not been
incorporated into these estimates.
Accordingly, the IWG did not attempt to explain in detail how the damage functions in the
models were constructed or their strengths and weaknesses, either for the earlier model versions
used in the 2010 estimates or for the updated versions used for the 2013 estimates. Rather,
stakeholders who are interested in these details are encouraged to consult the model
documentation and the related academic literature referenced in the TSDs, as well as the model
developers themselves. As noted in the OMB Response to Comments on SC-CO2, the IWG
accepts the point that clearer citations to specific model versions and revisions would be helpful
and will attempt to address this in the next update of the TSD.
Regarding opportunities for public comment, EPA notes that such opportunities have been
provided on all aspects of the SCC estimates in the interim estimates (2009) selected by the IWG
4-87

and in the numerous proposed rules, including the proposal for this rulemaking, issued by EPA
since February 2010 that made use of the estimates. As a general practice, EPA requests
comments on all aspects of the regulatory impact analysis, thereby providing ample opportunity
for the public to comment on SC-CO2 estimates used in these analyses. These comments helped
inform the development of the 2013 revised estimates. In addition, OMB provided a stand-alone
comment period on the 2013 estimates.
Regarding the adequacy of the information provided to the public as a basis for comment, the
TSDs provide a complete record of the methodology and assumptions used to develop the SCC
estimates, including references to the academic literature. As previously noted, independent
analysts have sought and received information from EPA allowing them to implement the IWG’s
SC-CO2 approach and modify it further if they choose. Additional documentation is available in
the academic literature on the IAMs themselves.
With regard to peer review, the assumptions and models employed in generating the SC-CO2
estimates are all drawn from the peer-reviewed academic literature. According to the OMB Final
Information Quality Bulletin for Peer Review and guidelines in the EPA's Peer Review
Handbook, if the Agency has determined that information has already been subject to adequate
peer review, then it is not necessary to have further peer review of that information. See Part
II.2 of OMB’s Final Information Quality Bulletin for Peer Review, and EPA’s Peer Review
Handbook (3rd ed.) pages B-15 and B-26. To further strengthen the robustness of the SCC
estimate, EPA and the other members of the IWG plan to seek external expert advice on
technical opportunities to improve the SCC estimates in future updates, including many of the
approaches suggested by commenters and peer-reviewed literature, and summarized in this
document.
Regarding comments that the estimates not be used until a future update that corrects the
commenters’ perceived flaws is completed, EPA does not feel that it is appropriate to withdraw
the current estimates. As previously noted, EPA has carefully examined and evaluated the full
range of comments and associated technical issues described in section 8.7.2, and determined
that it will continue to use the current SC-CO2 estimates until further updates can be made to
reflect forthcoming guidance from the Academies. EPA believes that the current estimates
continue to represent the best scientific information on the impacts of climate change available in
a form appropriate for incorporating the damages from incremental CO2 emissions changes into
regulatory analysis. EPA will continue to consider these comments and will share the
recommendations with the IWG as it moves forward with the Academies process.
EPA disagrees that the TSDs and SC-CO2 estimates are inconsistent with OMB and EPA
guidance, including OMB Circular A-4 and guidelines under the Information Quality Act. To
ensure that the SC-CO2 methodology is transparent, the TSDs are comprehensive and technically
rigorous in explaining the sources of data, the assumptions employed, the analytic methods
applied, and the statistical assumptions employed. To ensure that the results are reproducible,
EPA and other IWG members have provided technical assistance and modeling results to
external stakeholders upon request.
Circular A-4 is a living document, which may be updated as appropriate to reflect new
developments and unforeseen issues. OMB was fully involved in the development of the SC-CO2
4-88

estimates as a working group co-chair and supports the recommendations regarding the discount
rate and the focus on global damages. The departure from the standard discount rate
recommendations in Circular A-4 is explained in detail in the TSDs and in Section 8.7.2,
comment 7, of this document. Briefly, the use of 7 percent is not considered appropriate for
intergenerational discounting. There is wide support for this view in the academic literature, and
it is recognized in Circular A-4 itself. The emphasis on global rather than domestic damages is
also explained in detail in the TSDs. Beyond the fact that good methodologies for estimating
domestic damages do not currently exist, basing decisions on only the domestic damages from
carbon emissions will lead to an inefficient allocation of resources to reducing them, especially if
all countries adopt a similarly short-sighted approach. An efficient outcome can only be achieved
if all countries consider the full costs and benefits of their actions; the United States continues to
be a leader in working to establish such a regime internationally.
In addition, consistent EPA-wide implementation of its peer review and information quality act
guidelines has been an agency priority for many years. The most recent document describing
this EPA guidance, Guidance for Evaluating and Documenting the Quality of Existing Scientific
and Technical Information, was issued in December 2012 as an addendum to a 2003 EPA
guidance document entitled, A Summary of General Assessment Factors for Evaluating the
Quality of Scientific and Technical Information. The December 2012 addendum does not create
new standards for assessing and accepting data from other organizations – the Agency’s existing
information quality guidance continues to be applied to previous and future actions. Rather, the
December 2012 addendum references EPA’s comprehensive existing information quality
guidance and creates a consistent Agency-wide approach for reviewing and documenting the use
of existing scientific and technical information from other organizations.
The 2010 TSD for the SC-CO2 estimates provides documentation of the interagency decisions
and the 2013 TSD documents the technical update. Consistent with EPA’s 2003 data quality
guidance document and the December 2012 addendum referenced above, the 2010 and 2013
TSDs for the SC-CO2 estimates describe the type and quality of information used to estimate the
social cost of carbon as well as the sources for the information used (documentation of modeling
inputs—e.g., climate sensitivity and socioeconomic scenarios—in the peer reviewed literature).
The 2010 and 2013 TSDs also provide exhaustive documentation of how the USG’s review
identified, evaluated, and adopted the data, assumptions, and analytical framework used to
develop the SCC estimates. Furthermore, the IAMs used to develop the USG SCC estimates are
documented within the peer reviewed literature and source code is available on the model
developers’ websites or upon request from the relevant developer.
Finally, commenters significantly misstate the law regarding both the rights afforded under the
Data Quality Act and the nature of guidance implementing it. First, Prime Time Int’l Inc. v.
Vilsack, 599 F.3d 678, 685 (D.C. Cir. 2010) does not stand for the proposition that IQA claims
are judicially reviewable, contrary to commenters’ assertions. The case makes no such
statement, and the D.C. Circuit has indicated otherwise by its citation, in State of Mississippi, of
Salt Institute’s holding that such claims are not reviewable (see text above). See also Mississippi
Comm’n on Environmental Quality v. EPA, No. 12-1309 at n. 29, 2015 WL 3461262, (D.C. Cir.
June 2, 2015) noting that Prime Time dismissed an IQA challenge “without addressing argument
that the statute creates no legal rights in third parties”, and, as quoted above, stated that “almost
every court that has addressed an Information Quality Act challenge has held that the statute
4-89

‘creates no legal rights in any third parties’”(quoting Salt Institute)). The commenter also fails to
acknowledge all of the other cases uniformly rejecting this claim, and indeed, fails to
acknowledge the text of the IQA itself. See Styrene Info. & Research Ctr., Inc. v. Sebelius, 944
F. Supp. 2d 71, at 82 (D.D.C. 2013) (concluding “that the IQA and OMB guidelines do not
provide judicially manageable standards); Harkonen v. U.S. Dept. of Justice, 2012 U.S. Dist.
LEXIS 171415, at **30–36 (N.D. Cal. 2012) (dismissing argument that D.C. Circuit reached the
merits of an IQA claim in Prime Time); Family Farm Alliance v. Salazar, 749 F. Supp. 2d 1083,
1096–1100 (E.D. Cal. 2010) (rejecting argument that D.C. Circuit in Prime Time found an
implicit right of action under the IQA).
The commenter also reads Prime Time as holding that the OMB Guidelines are binding
legislative rules. Guidelines are just that: they provide assistance and are not binding.
Describing these OMB Guidelines, the D.C. Circuit noted that the Guidelines provide “‘policy
and procedural guidance’”, are meant to be “flexible” and are to be implemented differently by
different agencies accounting for circumstances. State of Mississippi, 744 F.3d at 1347; see
Styrene Info. & Research Ctr, Inc., 944 F. Supp. 2d at 82. State of Mississippi antedates Prime
Time. EPA consequently believes that the OMB Guidelines are not binding legislative rules.
Nor are EPA’s Peer Review guidelines binding legislative rules. EPA first issued its Peer
Review Policy in 1993, and in 1998 issued its Peer Review Handbook to provide nonbinding guidance for implementation of the Peer Review Policy. In 2004, OMB issued its
Final Information Quality Bulletin for Peer Review (OMB Bulletin) to provide governmentwide guidance to enhance the practice of peer review. EPA subsequently updated its Peer
Review Policy and Peer Review Handbook to incorporate the provisions of the OMB
Bulletin. EPA’s IQGs, Peer Review Policy, and Peer Review Handbook do not bind the
Agency, but instead guide EPA’s exercise of discretion when making decisions about
information quality and peer review. API v. EPA, 684 F. 3d 1342, 1348 (D.C. Cir. 2012)
(chastising industry petitioners for characterizing the EPA peer review guidance as binding
(“perhaps the API should have had its brief peer-reviewed”); State of Mississippi v. EPA, 744
F.3d at 1347 (D.C. Cir. 2013). The suggestion in the comments that EPA’s Inspector
General has decreed that the Peer Review guidance is binding both mischaracterizes the IG’s
statements and that Office’s function. It does not determine the legal status of agency
documents, much less countermand the holding of the D.C. Circuit just cited.
Regarding the comment that the IAMs are not available via EPA’s website, EPA notes that the
CREM is no longer an active body at EPA. In addition to obtaining information via the peerreviewed literature and the SC-CO2 TSDs, the public can instead find information about the
models in EPA’s Science Inventory database: DICE38, FUND39, and PAGE40)
38

http://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=240426&searchAll=DICE&sear
chAll=integrated+assessment&actType=&TIMSType=+&TIMSSubTypeID=&DEID=&epaNum
ber=&ntisID=&archiveStatus=Both&ombCat=Any&dateBeginCreated=&dateEndCreated=&dat
eBeginPublishedPresented=&dateEndPublishedPresented=&dateBeginUpdated=&dateEndUpdat
4-90

Regarding comments on SC-CO2 and BSER: The SC-CO2 was not used, as one commenter
stated, as a screen to justify the BSER. RIA chapters 3 and 5 present the SC-CO2 and the
application of these estimates to consider the illustrative benefit and cost scenarios. Please see
Section V.A of the preamble for discussion about the factors used to determine the Best System
of Emission Reduction (BSER).
Carbon dioxide fertilization
Comment 4.4-6: Commenter 9685 states the benefits of CO2 fertilization and improved crop
resilience are significant and must be reflected in any cost-benefit analysis. Commenter includes
excerpt from attached documents to support point.
Commenters disagreed about whether the integrated assessment models overestimate or
underestimate CO2 fertilization effects in the agriculture and forestry sectors. For example,
commenter 9685 states the benefits of CO2 fertilization and improved crop resilience are
significant and must be reflected in any cost-benefit analysis and referenced Idso (2013).
However, commenter 9861 stated that “the models do not reflect recent research on agricultural
changes, which suggests that CO2 fertilization is overestimated, particularly in the FUND model,
and that much, if not all, of the fertilization benefits may be cancelled out by negative impacts on
agriculture.” Commenter 9681 states the current inclusion of CO2 fertilization benefits likely
overstates its effects. The models do not reflect recent research on agricultural changes, which
suggest the CO2 fertilization is overestimated, particularly in the FUND model, and that much, if
not all, of the fertilization benefits may be cancelled out by negative impacts on agriculture (e.g.,
extreme heat, pests, and weeds). If the agency is not able to adequately model all agricultural
impacts it should, at a minimum, remove CO2 fertilization benefits.
Response 4.4-6: The comments regarding CO2 fertilization effects in the agriculture and forestry
sectors and "greening effect" issues mirror those submitted to the Office of Management and
ed=&dateBeginCompleted=&dateEndCompleted=&personID=&role=Any&journalID=&publish
erID=&sortBy=title&count=25&CFID=24359216&CFTOKEN=32090976.
39

http://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=240425&searchAll=FUND&sea
rchAll=integrated+assessment&actType=&TIMSType=+&TIMSSubTypeID=&DEID=&epaNu
mber=&ntisID=&archiveStatus=Both&ombCat=Any&dateBeginCreated=&dateEndCreated=&d
ateBeginPublishedPresented=&dateEndPublishedPresented=&dateBeginUpdated=&dateEndUpd
ated=&dateBeginCompleted=&dateEndCompleted=&personID=&role=Any&journalID=&publi
sherID=&sortBy=title&count=25&CFID=24359216&CFTOKEN=32090976.
40

http://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=240711&searchAll=PAGE09&s
earchAll=integrated+assessment&actType=&TIMSType=+&TIMSSubTypeID=&DEID=&epaN
umber=&ntisID=&archiveStatus=Both&ombCat=Any&dateBeginCreated=&dateEndCreated=&
dateBeginPublishedPresented=&dateEndPublishedPresented=&dateBeginUpdated=&dateEndUp
dated=&dateBeginCompleted=&dateEndCompleted=&personID=&role=Any&journalID=&publ
isherID=&sortBy=title&count=25&CFID=24359216&CFTOKEN=32090976.
4-91

Budget’s separate comment solicitation on the SC-CO2 (78 FR 70586; November 26, 2013). As
a member of the interagency working group (IWG) on SC-CO2, EPA has carefully examined and
evaluated comments submitted to OMB’s separate solicitation. EPA has also carefully examined
and evaluated all comments received regarding SC-CO2 through this rulemaking process and
determined that the IWG responses to the comments on the OMB solicitation address the CO2
fertilization effects in the agriculture and forestry sectors and "greening effect" issues
summarized in this section. Specifically, EPA concurs with the IWG’s response to these
comments and hereby incorporates them by reference.41
After careful evaluation of the full range of comments and associated technical issues described
in section 8.7.2, EPA has determined that it will continue to use the current SC-CO2 estimates
until further updates can be made to reflect forthcoming guidance from the Academies. EPA
believes that the current estimates continue to represent the best scientific information on the
impacts of climate change available in a form appropriate for incorporating the damages from
incremental CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the
current SC-CO2 estimates in the analyses of illustrative costs and benefits in the RIA.42 EPA will
continue to consider these comments and will share the recommendations with the IWG as it
moves forward with the Academies process.
The remainder of this section elaborates on the comments in the context of this rulemaking.
As noted in the OMB Response to Comments on SC-CO2, to date, the IWG has accepted the
models as currently constituted, and omitted any damages or beneficial effects that the model
developers themselves do not include. The IWG recognizes that none of the three IAMs fully
incorporates all climate change impacts, either positive or negative. Some of the effects
referenced by commenters (e.g., "catastrophic" effects, disease, and CO2 fertilization) are
explicitly modeled in the damage functions of one or more of the current models (although the
treatment may not be complete), and the model developers continue to update their models as
new research becomes available. In fact, the IWG undertook the 2013 revision because of
updates to the models, which include new or enhanced representation of certain impacts, such as
sea level rise damages. In addition, some of the categories mentioned by commenters are
currently speculative or cannot be incorporated into the damage function for lack of appropriate
data. Using an ensemble of three different models was intended to, at least partially, address the
fact that no single model includes all of the impacts. We recognize that there may be effects that
none of the three selected models addresses (e.g., impacts from ocean acidification) or that are
likely not fully captured (e.g. catastrophic effects).
EPA also recognizes that the impacts of climate change on agriculture is an area of active
research and that methodological and data challenges persist. As a result there is uncertainty as to
the magnitude of these impacts and the role of interactions between changes in the climate and
other factors, such as CO2 fertilization, temperature, precipitation, ozone, pests, etc.
41

Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 7-11 at
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-finaljuly-2015.pdf .
42
See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.
4-92

Additionally, these effects are likely to vary widely across regions and crops. However, with
high confidence the IPCC (2013) stated in its Fifth Assessment Report (AR5) that "[b]ased on
many studies covering a wide range of regions and crops, negative impacts of climate change on
crop yields have been more common than positive impacts." As noted above, the IWG’s
approach to date has been to rely on the damage functions included in the three IAMs by their
developers.
EPA agrees that it is important to update the SC-CO2 periodically to incorporate improvements
in the understanding of greenhouse gas emissions impacts and will continue to follow and
evaluate the latest science on impact categories that are omitted or not fully addressed in the
IAMs. Also, EPA and the other members of the IWG will seek external expert advice on the
technical merits and challenges of potential approaches to update the damage functions in future
revisions to the SC-CO2 estimates.
See also 4.1-9, for EPA’s response to comments about carbon fertilization, including EPA’s
general treatment of the impact of increased CO2 on plant growth and about treatment of CO2
fertilization in models.
Discount rate
Comment 4.4-7: Commenter 8925 states constant discount rates are inconsistent with the
economic growth models being used for this analysis and that discounting should be dynamic
and a function of the consumption growth rate.
Commenter 10091 states regarding discount rates that the IWG ignores OMB guidelines in its
selection of discount rates to use in calculating the SCC. OMB Circular A-4 refers to OMB
Circular A-94 which states that "a real discount rate of 7 percent should be used as a base-case
for regulatory analysis" and to show the sensitivity of the results to the discount rate assumptions
"[f]or regulatory analysis, you should provide estimates of net benefits using both 3 percent and
7 percent." Instead, the IWG opted to determine the SCC using discount rates of 2.5, 3, and 5
percent, and did not include results for a 7 percent rate. This has ramifications throughout the
federal regulatory agencies. For example, in previous Environmental Protection Agency
discussions of a proposed new rule regulating discharges from steam electric power plants (and
how the rule may impact emissions from the plants), the EPA included a Table of the annualized
benefits of the emissions reductions from NOx, SO2, and CO2. Following OMB guidelines on
how to calculate costs and benefits-the EPA reports its findings using both a 3% and a 7%
discount rate. Yet as explained in a footnote to the Table (see below), the CO2 benefits are
calculated using the 3% and the 5% discount rate "because SCC calculations are not available for
the 7 percent discount rate." The reason they are "not available" is that the IWG ignored the
OMB guidelines and did not provide the SCC at a 7 percent discount rate for use in regulator
analyses. Regulatory agencies, like the EPA, have in the past, tried to follow the rules and
calculate cost and benefits using both the 3 percent and 7 percent discount rates. Yet, when they
express the benefits of reducing carbon dioxide emissions at the recommended 7 percent rate,
they are actually going to use a number that is incorrect and inaccurate, and explain in a footnote
why they are doing so. This is an unacceptable excuse, and is yet another example of why the
IWG's determination of the SCC is inappropriate for use in federal regulatory analyses.
Commenter states that in this proposed Rule, the EPA does not investigate the results of using a
7% discount rate on the comparison of new coal-fired power plants with and without CCS
4-93

technology in direct violation of OMB guidelines.
Commenter 9396 states that in violation of the Information Quality Act October 2002 OMB
Bulletin, the IWG failed to run discount analysis using OMB guidelines, instead choosing to
inflate estimated benefits by using a lower discount rate for calculations. The IWG has the
flexibility to run modeling at whatever rate it chooses so long as the models include both 3% and
7%. The failure to run a 7% analysis has the consequence of skewing the average SCC to a
higher level than if OMB policy was followed.
Commenter 9681 provides recommendations on further refinements to the SCC. Industry groups
have argued that the IWG analysis is flawed in its use of consumption discount rates,
consideration of scientific uncertainty, and in counting both domestic and global damages caused
by our emissions. None of these criticisms have merit. Indeed, the IWG process uses
assumptions that accord with economic and scientific theory. Economic models, and the
scientific analyses they draw from, are of course improving continuously. Future updates to the
SCC should build on these and go further. As further refinements better account for climate
change impacts not yet incorporated into the modeling, all indications are that the estimated
benefits of curbing carbon pollution will rise substantially over current estimates.
Commenter 9681 continues the IWG appropriately used consumption discount rates rather than
returns on capital. With respect to the discount rate, the IWG conducted sensitivity analysis of
the results to three constant consumption discount rates: 2.5%, 3%, and 5%; for each of the
discount rates, the TSDs reported the various moments and percentiles64 (The moments of a
distribution (of SCC estimates in this case) are, loosely speaking, the various values that describe
the distribution's shape: what value is the distribution centered around (mean); how wide is the
distribution (the variance); whether the distribution is lopsided (skewness); and whether it is tall
and skinny or short and fat (kurtosis). A percentile is a statistical measure of the value (the SCC
value in this case) below which a specified percentage of (SCC) observations falls. The 1st
percentile indicates the SCC value above which (the other) 99% of observed SCC values fall.
The 99th percentile indicates the SCC value below which 99% of all observed SCC values fall. )
of the SCC estimates. The discount rate is one of the most important inputs in models of climate
damages, with plausible assumptions easily leading to differences of an order of magnitude in
the SCC. The climate impacts of present emissions will unfold over hundreds of years. When
used over very long periods of time, discounting penalizes future generations heavily due to
compounding effects. For example, at a rate of 1%, $1 million 300 years hence equals over
$50,000 today; at 5% it equals less than 50 cents. The discount rate changed by a factor of five,
whereas the discounted value changed by more than five orders of magnitude. Depending on the
link between climate risk and economic growth risk, even a rate of 1% may be too high.
Uncertainty around the correct discount rate pushes the rate lower still. The IWG correctly
excluded a 7% discount rate, a typical private sector rate of return on capital, for several reasons.
First, typical financial decisions, such as how much to save in a bank account or invest in stocks,
focus on private decisions and utilize private rates of return. Private market participants typically
have short time horizons. However, here we are concerned with social discount rates because
emissions mitigation is a public good, where individual emissions choices affect public wellbeing broadly. Rather than evaluating an optimal outcome from the narrow perspective of
investors alone, economic theory would require that we make the optimal choices based on
societal preferences (and discount rates). Second, climate change is expected to affect primarily
4-94

consumption, not traditional capital investments. OMB guidelines note that in this circumstance,
consumption discount rates are appropriate. Third, 7% is considered much too high for reasons
of discount rate uncertainty and intergenerational concerns (further discussed below).
Commenter 9681 also states that the IWG correctly adopted as one of its discount rates a value
reflecting long-term interest rate uncertainty, and-as a primary extension to current results should
go further by directly implementing a declining discount rate. The IWG was correct in choosing
as one of its discount rates an estimate based upon declining discount rates (2.5%). Since the
IWG undertook its initial analysis, a consensus has emerged among leading climate economists
that a declining discount rate should be used for climate damages to reflect long-term uncertainty
in interest rates. Arrow et al (2013) presents several arguments that strongly support the use of
declining discount rates for long-term benefit-cost analysis. Perhaps the best reason is the simple
fact that there is considerable uncertainty around which interest rate to use: uncertainty in the
rate points directly to the need to use a declining rate, as the impact of the uncertainty grows
exponentially over time. The uncertainty about future discount rates could stem from a number
of reasons particularly salient to climate damages, including uncertainties in future economic
growth, consumption, and the interest rate reaped by investments. A possible declining interest
rate schedule for consideration by the IWG is the one proposed by Weitzman (2001). It is
derived from a broad survey of top economists and the profession at large in a climate change
context and explicitly incorporates arguments around interest rate uncertainty. Cropper et al
(forthcoming) similarly argues for a declining interest rate schedule and lays out the fundamental
logic. Moreover, the United States would not be alone in using a declining discount rate. It is
standard practice for the United Kingdom and French governments, among others. The U.K.
declining discount rate schedule that subtracts out a time preference value. France's schedule is
roughly similar to the United Kingdom's. Importantly, all of these discount rate schedules yield
lower present values than the constant 2.5% Newell-Pizer rate, suggesting that even the lowest
discount rate evaluated by the IWG is too high. The consensus of leading economists is that a
declining discount rate schedule should be used, consistent with the approach of other countries
like the United Kingdom. Adopting such a schedule would increase the SCC substantially from
the administration's central estimate, suggesting that even the high end of the range presented by
the administration is likely too low.
Commenters 9194, 7977 states the IWG/OMB process is flawed in that it fails to calculate the
SCC using OMB's prescribed "base case" discount rate of 7% and impermissibly calculates
global benefits instead of domestic benefits.
Commenter 9423 states a key variable in estimating the SCC is the discount rate. OMB Circular
A-94 states that a discount rate of 7 percent should be used as a base case for regulatory analysis
as a default position while acknowledging that in some cases, other discount rates may be
appropriate. OMB Circular A-4 states that "For regulatory analysis, one should provide estimates
of net benefits using both 3 percent and 7 percent." The impact of the use of different discount
rates can be significant. In the revised Technical Support Document, the social cost of carbon for
2050 is $97/metric ton of C02 using a 2.5 percent discount factor while it is $26/metric ton of
C02 using a 5 percent discount factor. EPA should include an analysis using a discount rate of 7
percent, as suggested by OMB.
Response 4.4-7: The comments regarding discount rates applied to the SC-CO2 estimates mirror
those submitted to the Office of Management and Budget’s separate comment solicitation on the
4-95

SCC (78 FR 70586; November 26, 2013). As a member of the interagency working group
(IWG) on SC-CO2, EPA has carefully examined and evaluated comments submitted to OMB’s
separate solicitation. EPA has also carefully examined and evaluated all comments received
regarding SC-CO2 through this rulemaking process and determined that the IWG responses to
the comments on the OMB solicitation address the comments on the discount rates applied to the
SC-CO2 estimates. Specifically, EPA concurs with the IWG’s response to these comments and
hereby incorporates them by reference.43 The remainder of this section elaborates on these
comments in the context of this rulemaking.
EPA and other members of the IWG have reconfirmed their commitment to periodic review and
update of the methodology and estimates to ensure that they continue to reflect the best available
science and economics. Furthermore, EPA has determined that the SC-CO2 methodological
recommendations require additional research, review, and public comment before it can apply
them to a rulemaking context.
To help synthesize the technical information and input reflected in the comments, and to add
additional rigor to the next update of the SC-CO2, the EPA and all of the other IWG members
plan to seek independent expert advice on technical opportunities to improve the SC-CO2
estimates, including the approaches suggested by commenters and summarized in this document.
Specifically, EPA and all of the other IWG members plan to ask the National Academies of
Sciences, Engineering, and Medicine (Academies) to examine the technical merits and
challenges of potential approaches to improving the SC-CO2 estimates in future updates. Input
from the Academies, informed by the comments submitted to this rulemaking and the peerreviewed literature, will help to ensure that the SC-CO2 estimates used by EPA continue to
reflect the best available science and methodologies.
After careful evaluation of the full range of comments and associated technical issues described
in section 4.4, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SCCO2 estimates in the analysis of illustrative benefit-cost scenarios for new sources in RIA
Chapter 5.44 EPA will continue to consider these comments and will share the recommendations
with the IWG as it moves forward with the Academies process.
The remainder of this section provides more detailed responses to the comments and
recommendations.

43

Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 20 to 25
at https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-finaljuly-2015.pdf
44
See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.
4-96

OMB guidance in Circular A-4 recommends that discount rates of 3 percent and 7 percent be
used in regulatory impact analysis. The 7 percent rate is an estimate of the average before-tax
real rate of return to private capital in the U.S. economy. It is a broad measure that reflects the
returns to real estate and small business and corporate capital and is meant to approximate the
opportunity cost of capital in the United States. The 3 percent rate is an estimate of the real rate
at which consumers discount future consumption flows to their present value, often referred to as
the social rate of time preference or the consumption rate of interest. As stated in the 2010 TSD,
in a market with no distortions, the return to savings would equal the private return on
investment, and the market rate of interest would be the appropriate choice for the social
discount rate. In the real world, however, risk, taxes, and other market imperfections drive a
wedge between the risk-free rate of return on capital and the consumption rate of interest.
While most regulatory impact analysis is conducted over a time frame in the range of 20 to 50
years, OMB guidance in Circular A-4 recognizes that special ethical considerations arise when
comparing benefits and costs across generations. Although most people demonstrate time
preference in their own consumption behavior, it may not be appropriate for society to
demonstrate a similar preference when deciding between the well-being of current and future
generations. Future citizens who are affected by such choices cannot take part in making them,
and today's society must act with some consideration of their interest. Even in an
intergenerational context, however, it would still be correct to discount future costs and benefits
generally (though perhaps at a lower rate than for intragenerational analysis), due to the
expectation that future generations will be wealthier and thus will value a marginal dollar of
benefits or costs less than the current generation. Therefore, it is appropriate to discount future
benefits and costs relative to current benefits and costs, even if the welfare of future generations
is not being discounted. Estimates of the discount rate appropriate in this case, from the 1990s,
ranged from 1 to 3 percent. After reviewing those considerations, Circular A-4 states that if a
rule will have important intergenerational benefits or costs, agencies should consider a further
sensitivity analysis using a lower but positive discount rate in addition to calculating net benefits
using discount rates of 3 and 7 percent.
EPA and the other members of the IWG examined the economics literature and concluded that
the consumption rate of interest is the correct concept to use in evaluating the net social costs of
a marginal change in CO2 emissions, as the impacts of climate change are measured in
consumption-equivalent units in the three IAMs used to estimate the SC-CO2. This is consistent
with OMB’s guidance in Circular A-4, which states that when a regulation is expected to
primarily affect private consumption—for instance, via higher prices for goods and services--it is
appropriate to use the consumption rate of interest to reflect how private individuals trade-off
current and future consumption.
As explained in the 2010 TSD, after a thorough review of the discounting literature, the IWG
chose to use three discount rates to span a plausible range of constant discount rates: 2.5, 3, and 5
percent per year. The central value, 3 percent, is consistent with estimates provided in the
economics literature and OMB’s Circular A-4 guidance for the consumption rate of interest. The
upper value of 5 percent represents the possibility that climate damages are positively correlated
with market returns, which would suggest a rate higher than the risk-free rate of 3 percent.
Additionally, this discount rate may be justified by the high interest rates that many consumers
4-97

use to smooth consumption across periods. The low value, 2.5 percent, is included to incorporate
the concern that interest rates are highly uncertain over time. It represents the average rate after
adjusting for uncertainty using a mean-reverting and random walk approach as described in
Newell and Pizer (2003), starting at a discount rate of 3 percent. Further, a rate below the riskless
rate would be justified if climate investments are negatively correlated with the overall market
rate of return. Use of this lower value also responds to the ethical concerns discussed above
regarding intergenerational discounting.
EPA recognizes that disagreement remains in the academic literature over the appropriate
discount rate to use for regulatory analysis of actions with significant intergenerational impacts,
such as CO2 emissions changes that affect the global climate on long time scales. EPA and the
other members of the IWG will continue to follow and evaluate the latest science on
intergenerational discounting and seek external expert advice on issues related to discounting in
the context of climate change.
Regarding recommendations for lower discount rates or a declining rate: As noted above, the
EPA and the other members of the IWG selected a range of discount rates from 2.5 to 5 percent;
a review of the literature and the reasoning that led to the selection of this range are discussed in
detail in the 2010 TSD. Several of the issues raised by commenters were explicitly considered
and were part of the rationale for the selected range.
The 2010 TSD discusses both descriptive and prescriptive approaches to selecting discount rates.
A descriptive approach reflects a positive (non-normative) perspective based on observations of
people’s actual choices (e.g., savings versus consumption and allocation of savings over more
and less risky investments). The prescriptive approach adds a normative component and
incorporates judgments that decision makers believe should be reflected in the policy choices
that the discount rate is intended to inform. For example, some have argued on ethical grounds
that in the Ramsey formula—which dissects market rates into three components, the pure rate of
time preference (ρ), growth rate of per-capita consumption (g), and coefficient of relative risk
aversion (η)—the ρ component should be set to zero so that the welfare of all generations is
valued equally. After considering a range of plausible values from the literature for g and η, the
IWG concluded that setting a very low ρ (e.g., 0.1 percent per year) could yield rates in the range
of 1.4 to 3.1 percent.
The 2010 TSD also discussed uncertainty and its effects on discount rates. The certainty
equivalent values of the future benefits of reducing current CO2 emissions will be lower than the
expected value if the benefits and future consumption are positively correlated, assuming people
are risk averse on average. This in turn implies that when discounting expected future benefits a
discount rate that accounts for uncertainty should exceed a riskless rate. As explained in the
TSD, this consideration was part of the logic for setting the upper end of the selected range at 5
percent.
EPA and other members of the IWG also considered the issue of “climate catastrophes.” To the
extent that such outcomes may not be adequately represented in the IAMs, the central tendency
estimates from these models may not capture the full range of potential damages from CO2
emissions. For this reason, in addition to the three mean SC-CO2 estimates using discount rates
4-98

of 2.5, 3 and 5 percent, EPA and other members of the IWG recommended including a rate based
on the 95th percentile damage estimate (with a 3 percent discount rate) for the upper end of the
range of plausible SC-CO2 estimates.
With respect to declining discount rates, EPA and other members of the IWG agrees that this is
an important area of emerging research. However, no widely-accepted declining discount rate
schedule has yet been developed. Some key technical issues warrant careful consideration before
adopting a declining discount rate schedule, such as determining how to update the discount rate
schedule as uncertainty is resolved over time and ensuring that the use of declining discount rates
does not lead to the possibility of time-inconsistent choices. A recent workshop sponsored by the
federal government resulted in a paper in Science authored by thirteen prominent economists
who concluded that a declining discount rate would be appropriate to analyze impacts that occur
far into the future (Arrow et al., 2014). However, additional research and analysis is still needed
to develop a methodology for implementing a declining discount rate and to understand the
implications of applying these theoretical lessons in practice. The EPA and other members of the
IWG will continue to follow and evaluate the latest science on the use of declining discount rates
in intergenerational contexts and seek external expert advice on issues related to discounting in
the context of climate change.
Finally, regarding the comment about the consistency of discount rates applied to the climate
benefits versus other monetized impacts: while EPA continues to find the discount rates applied
to the SC-CO2 estimates to be appropriate, it recognizes that different discount rates are applied
to other impacts. The workshop mentioned in the preceding paragraph discussed how to apply
discounting in a regulation where some costs and benefits accrue intra-generationally while
others accrue inter-generationally. The experts noted that one solution to this problem is to apply
a declining discount rate schedule to all regulations. As previously noted, additional research
and analysis is needed to develop a methodology for implementing a declining discount rate.
The EPA and other members of the IWG will continue to follow and evaluate the latest science
on the use of declining discount rates in intergenerational contexts and seek external expert
advice on issues related to discounting in the context of climate change.
Application and aggregation
Comment 4.4-8: Commenter 9666 states EPA should update its modeling to include the most
recent SC-CO2 estimates, noting that EPA has not updated its modeling since the rule was first
proposed in 2012 and that OMB's current SCC estimates were updated twice in
2013. Commenter 9666 states EPA should refrain from utilizing SCC values until OMB has
finalized its rule on the use of SCC among federal agencies.
Commenter 9423 states EPA should not use the Office of Management and Budget's (OMB)
Revised Social Cost of Carbon Estimates in the Cost/Benefit Analysis. In Section 5.7 of the RIA,
EPA uses the OMB's revised estimates for the Social Cost of Carbon (SCC) in estimating the
cost and benefits of the subject regulatory proposal for new electric generating units. The TCEQ
provided comments to 0MB regarding their Technical Support Document: Technical Update of
the Social Cost of Carbon (SCC) for Regulatory Impact Analysis under Executive 0rder No.
12866 published in the Federal Register on November 26, 2013. Given that OMB has requested
comments on the SCC estimates, TCEQ, PUC, and RRC believe it is inappropriate for EPA to
4-99

use the revised estimates in the 111(b) rulemaking as it deprives stakeholders the opportunity to
comment on the final version of the OMB guidance. Notwithstanding TCEQ's objection to the
use of the OMB guidance in the 111(b) rulemaking, TCEQ, PUC, and RRC offer the following
further comments on EPA's use of the SCC estimates in the RIA. The assumptions, both
scientific and economic, of the Interagency Working Group that developed the initial SCC
estimates are largely unknown and were not subject to independent peer review as called for by
the Information Quality Act and 0MB's own internal guidelines. The TCEQ, PUC, and RRC
strongly recommend that the SCC estimates not be used in the 111(b) rulemaking until the entire
process and data used in developing the SCC estimates are subjected to thorough, independent,
external peer review. The models used to derive SCC estimates suffer from considerable
uncertainty and speculation in critical inputs including the amount of future carbon emissions,
the impact of those emissions on the climate, and the monetization of those impacts. The
uncertainty associated with temperature increases is just one example. The models used by the
United Nations Intergovernmental Panel on Climate Change (IPCC) have not predicted the
current stoppage of global warming, as evidenced by global temperature measurements, over the
last 15 years. To the extent that IPCC model data was used in developing the SCC estimates,
their use for the short term is clearly circumspect and extrapolation of temperature increases and
its impacts over a longer modeling time horizon is of even greater uncertainty.
Commenter 10618 states EPA uses the Social Cost of Carbon (SCC) to characterize potential
carbon benefits associated with the rule, with even though it is widely acknowledged that these
cost estimates are inaccurate, uncertain, and highly speculative.45 EPA acknowledges in the RIA
that "any effort to quantify and monetize the harms associated with climate change will raise
serious questions of science, economics, and ethics and should be viewed as provisional." As
such, these calculations cannot form the basis of an adequate RIA. AEP has submitted comments
on the SCC in its development and use and they are attached as Appendix F to this document.
Commenter 9194 questions the appropriateness of using the IWG estimates of the SCC for
regulatory purposes whether for justifying BSER for CO2 for coal plants under this rulemaking
or in any other EGU rulemaking that will serve as a predicate for economy-wide CO2 regulations
in the future.
Commenter 9194 continues that EPA should not use the social cost of carbon (SCC) in this or
any other rulemaking affecting electric generating units until the IWG process headed by the
OMB has been subjected to public notice and comment and until the models, all assumptions
used in the models, all underlying data used and all analyses used by OMB and the IWG have
undergone peer review and public notice and comment.
Commenter 10091 states the social cost of carbon as determined by the Interagency Working
Group in their May 2013 Technical Support Document (updated in November 2013) and used by
the EPA in this proposed Standards of Performance for Greenhouse Gas Emissions from New
Stationary Sources: Electric Utility Generating Units is unsupported by the scientific literature,
not in accordance with OMB guidelines, fraught with uncertainty, illogical and thus completely
45

See RIA 5-36 through 5-39.
4-100

unsuitable and inappropriate for federal rulemaking. As such, use of the SCC in cost/benefit
analyses in this proposed rulemaking should be suspended.
Commenter 9396 states the use of the IWG SCC Can Produce Irrational Energy Choices: The
SCC as produced by the IAMs could also produce some irrational conclusions in selecting the
most cost-effective sources for electric energy production. As seen in the table below, by using
SCC values that could be developed well within the input assumptions of the IAM developers, it
is possible to select base load generating choices that many in the industry would consider
absurd. The values developed in the table above assume that lower carbon energy will replace
existing coal-fired generation. By using the IWG central value SCC, all generating alternatives
are more cost-effective than coal-fired generation with wind energy and biomass being more
cost-effective than an NGCC unit. By using the SCC high value, which is lower than one of the
PAGE model developer recommends, wind, biomass and solar photovoltaic energy are cheaper
than an NGCC base load unit.46 Based on the SCC, one could easily conclude that it would be
very economical to shut-down all fossil fuel electric generation and replace it with zero carbon
generating sources. This conclusion would be considered absurd by most experts in the electric
utility industry; especially since base load solar and wind powered generating plants are not yet
commercially available.
Commenter 9401 states EPA has not performed an analysis of carbon leakage which will occur
as a result of the rule, thus overstating the benefits. Because manufacturing uses 25 percent of all
U.S. electricity, we are significantly impacted by electricity prices. Figure 1 illustrates that coal
consistently provides the lowest cost energy to generate electricity. The rule would stop the use
of coal for use in new power generating units and deprive manufacturing from the lowest cost
power generation, thereby impacting competitiveness. It is important to note, that having coal as
a major component of electricity fuel diversity provides for increased reliability. Higher
electricity costs and reduced reliability will shift EITE industries to other countries along with
their jobs and GHG emissions, and accomplish nothing. The EPA must perform an analysis of
carbon leakage that will occur as a result of the rule. Not having done so, the EPA has overstated
the projected benefits of emission reductions and not accounted for the associated loss of
economic activity and jobs moving outside the U.S. Without including the GHG leakage, any
benefits are arbitrary and capricious. And, since most countries have higher power generation
GHG rates, it would result in a net increase in global GHG emissions.
Commenter 9396 Averaging Model Results: The IWG provides an estimated benefit that fails to
capture model uncertainty. For example, a $43/ton average benefit in 2020 is ascribed to model
outcomes that range from $-22/ton (i.e. $22 in societal cost) to $727 in benefits. The variability
associated with model run outcomes is sufficiently significant that an informed analytical
evaluation would conclude that determining a single number for the purpose of developing
regulations is unwarranted and imprudent.
Commenter 9681 states EPA should continue to use the current IWG estimate of the SCC. The
current SCC estimates are biased downwards: more can and should be done to improve the SCC
46

The Social Cost of Carbon in US Regulatory Impact Analysis: an Introduction and Critique,
Johnson and Hope Journal of Environmental Studies and Science, September 12, 2012.
4-101

and to ensure, through regular updates, that it reflects the latest science and economics. But the
necessary process of improving the ability of the SCC to fully reflect the costs of climate impacts
to society cannot hold up agency rulemaking efforts. The SCC provides an important, if
conservative, estimate of the costs of climate change and the benefits of reducing carbon
pollution. To ignore these costs would be detrimental to the health and well-being of Americans
and contrary to law and Presidential directives to agencies to evaluate the cost of pollution to
society when considering standards to abate that pollution. In the context of agency rulemakings,
the SCC provides the best available means to factor those costs into benefit-cost analyses.
Commenter also suggest that EPA encourage the IWG to continuously update the SCC, as new
economic and scientific consensus emerges. Such updates are in line with the stated intentions of
the IWG, which committed to updating these estimates as the science and economic
understanding of climate change . . . improves. Accordingly, EPA should continue to use the
IWG's latest SCC estimates in its regulatory impact analyses on greenhouse gas performance
standards under Section 111 of the Clean Air Act.
Commenter 9514 states EPA properly relied on the federal government's most recent estimate of
the social cost of carbon and that the agency's analysis comported with the standard guidance for
regulatory analysis of costs and benefits.
Commenter 9592 states that EPA's discussion of SCC in the RIA, as well as the sample
calculations in the supporting technical document, "Technical Support Document: Social Cost of
Carbon for Regulatory Impact Analysis, Executive Order 12866, Interagency Working Group on
Social Cost of Carbon, United States Government" provide a pathway for future economic
analyses that can justify the use of natural gas as a future existing source GHG standard of
performance.
Commenter 8925 states the 2010 USG SCCs are problematic for evaluating non-incremental
changes in global emissions and, if applied, could result in misleading conclusions; that
substantial management of future climate does not have a fixed future; and that good practice
application guidance is needed to avoid misuse.
Commenter 9396 also states the IWG's initial finding that the SCC was a useful tool in assessing
costs associated with incremental benefits or costs that are derived from federal actions is
underscored when it goes on to highlight the concern about rules which have a primary purpose
of reducing GHG emissions stating, "For policies that have a large (non-marginal) impact on
global cumulative emissions, there is a separate question of whether the SCC is an appropriate
tool for calculating the benefits of reduced emissions." Despite the admonition that the level of
uncertainty associated with the development of the SCC reduced its relevance as a policymaking
tool for the purposes of establishing cost-benefit analysis where GHG reductions present the
predominant benefit, EPA is proceeding with regulatory actions that base significant regulatory
burdens on industry based on GHG reductions calculated using the SCC. EPA and other federal
agencies use the social cost of carbon (SCC) to estimate the climate benefits of rulemakings the
projected net present value of carbon dioxide mitigation benefits over the next forty years from
three vehicle rulemakings was estimated to range from $78 billion to $1.2 trillion ($2010),
depending on which of the four SCC estimates were used (i.e., the average SCC at 5, 3, and 2.5
percent and the 95th percentile SCC at 3 percent). Finding $1.2 trillion dollars of benefits
derived from application of the SCC meaningfully and substantially deviates from the warning
4-102

by the IWG that application of the SCC may be inappropriate in circumstances where the
benefits from the SCC are large to the extent that reductions of GHG are justified by the SCC
benefits. The policy questions regarding decisions made by the IWG, such as choices concerning
discount rates or measuring domestic regulation against global benefits and the application of the
SCC have never been resolved or discussed in a public forum by the federal government but
rages in the scientific and legal community. The best case for cost-benefit analysis is that the
2010 IWG recommendations are politically neutral in the sense of drawing on widely shared
intuitions about human well-being. But cost-benefit analysis cannot cope with inherently
political questions involving contested normative issues. Policymakers will have to find
alternative tools when those questions predominate. The SCC reflects arbitrarily chosen damage
functions or, more precisely, the average of several arbitrarily chosen damage functions. It
emerges from three climate models of questionable reliability. Guesswork is not always fatal to
cost-benefit analysis. Judgment is needed to distinguish between reasonable estimates and
estimates that are excessively wide of the mark. Our judgment is that the IWG report relies on
too many unwarranted assumptions and cannot be relied on. But technical problems can be
repaired through further research. Computer models of climate change are improving and will
eventually provide reasonable estimates of the SCC. The more serious problem for cost-benefit
analysis of climate change is that climate regulation requires a series of judgments that are
political and institutional rather than technical.
Commenter 10091 states the social cost of carbon as determined by the Interagency Working
Group in their May 2013 Technical Support Document (updated in November 2013) and
included in the Regulatory Impact Analysis of this EPA action is unsupported by the scientific
literature, not in accordance with OMB guidelines, fraught with uncertainty, and thus unsuitable
and inappropriate for federal rulemaking. As such, use of the SCC in cost/benefit analyses in this
and all proposed federal regulation should be suspended.
Commenter 10087 states that SCC is an arbitrary metric that should not be used for regulatory
impact analysis, citing references.
Response 4.4-8: The comments regarding the aggregation of the SC-CO2 estimates and general
application in a rulemaking context mirror those submitted to the Office of Management and
Budget’s separate comment solicitation on the SC-CO2 (78 FR 70586; November 26, 2013). As
a member of the interagency working group (IWG) on SC-CO2, EPA has carefully examined and
evaluated comments submitted to OMB’s separate solicitation. EPA has also carefully examined
and evaluated all comments received regarding SC-CO2 through this rulemaking process and
determined that the IWG responses to the comments on the OMB solicitation address the
comments on the aggregation of SC-CO2 estimates and use of the estimates in this RIA.
Specifically, EPA concurs with the IWG’s response to these comments and hereby incorporates
them by reference.47 The remainder of this section elaborates on these comments in the context
of this rulemaking.

47

Referred to as the "OMB Response to Comments on SC-CO2." In particular, see pgs 25 to 28,
40-41 at https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-commentsfinal-july-2015.pdf.
4-103

EPA and other members of the IWG have reconfirmed their commitment to periodic review and
update of the methodology and estimates to ensure that they continue to reflect the best available
science and economics. Furthermore, EPA has determined that the SC-CO2 methodological
recommendations require additional research, review, and public comment before it can apply
them to a rulemaking context.
To help synthesize the technical information and input reflected in the comments, and to add
additional rigor to the next update of the SC-CO2, the EPA and all of the other IWG members
plan to seek independent expert advice on technical opportunities to improve the SC-CO2
estimates, including the approaches suggested by commenters and summarized in this document.
Specifically, EPA and all of the other IWG members plan to ask the National Academies of
Sciences, Engineering, and Medicine (Academies) to examine the technical merits and
challenges of potential approaches to improving the SC-CO2 estimates in future updates. Input
from the Academies, informed by the comments submitted to this rulemaking and the peerreviewed literature, will help to ensure that the SC-CO2 estimates used by EPA continue to
reflect the best available science and methodologies.
After careful evaluation of the full range of comments and associated technical issues described
in section 4.5, EPA has determined that it will continue to use the current SC-CO2 estimates until
further updates can be made to reflect forthcoming guidance from the Academies. EPA believes
that the current estimates continue to represent the best scientific information on the impacts of
climate change available in a form appropriate for incorporating the damages from incremental
CO2 emissions changes into regulatory analysis. Therefore, EPA has presented the current SCCO2 estimates48 in the analysis of illustrative benefit-cost scenarios for new sources in RIA
Chapter 5 (see RIA Chapters 3 and 5; see Response to Comments 3.2 for detailed discussion
about the conclusions of the RIA). EPA will continue to consider these comments and will
share the recommendations with the IWG as it moves forward with the Academies process.
The remainder of this section provides more detailed responses to the comments and
recommendations.
Aggregation and selection of SC-CO2 estimates
Some comments discussed the aggregation of model results and the selection of the final range
of SC-CO2 estimates. Both the 2010 TSD and the current TSD present information about the full
distribution of SC-CO2 estimates within and across possible combinations of the three models
and five socioeconomic-emissions scenarios, for each of three discount rates (45 combinations in
total) (see tables A2-A4 in Appendix A of the TSDs). Additional summary statistics for the
distributions of the SC-CO2 estimates are also provided for each of the three models (see Table
A5 in Appendix A). EPA believes that the information presented in the TSDs is sufficiently
disaggregated to reflect the variability of the SC-CO2 estimates across models, input
assumptions, and discount rates. In addition, EPA has provided the full set of Monte Carlo
modeling results (10,000 model runs for each combination, for a total of 450,000 observations
per emissions year) to outside researchers upon request and will continue to do so.

48

See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.
4-104

As discussed in the 2010 TSD, using the full distribution of the SC-CO2 estimates from the 45
scenarios would be impractical in a regulatory impact analysis. To produce a range of plausible
estimates that still reflects the uncertainty about the SCC estimates, the results from the various
model and scenario combinations (150,000 observations per emissions year for each of the three
discount rates) were pooled to produce three separate probability distributions for the SC-CO2
for emissions in a given year, one for each assumed discount rate (2.5, 3 and 5 percent). Three
point estimates were then derived from these pooled distributions representing the mean at each
discount rate. EPA considers this approach for presenting expected SC-CO2 values across a
range of discount rates to be appropriate for representing the central tendency of the SC-CO2
estimates across scenarios. The fourth value, the 95th percentile of the pooled distribution using
a 3 percent discount rate, is included to represent higher-than-expected economic impacts from
climate change further out in the tail of the SC-CO2 distribution - i.e., impacts that may have
lower probability of occurring but relatively high damages. For purposes of representing the
uncertainties involved, the TSDs emphasized the importance of considering and presenting the
full range of these four estimates in regulatory impact analysis.
EPA agrees that the modeling of uncertainty in our analysis, including the uncertainty explicitly
represented in the IAMs, may not capture the full range of uncertainty of the “true” value of the
SCC. This concern is common to most quantitative assessments of uncertainty. By definition, the
modeling of uncertainty requires a model, and therefore cannot capture the uncertainty
associated with model selection. However, EPA does not agree that pooling results across
models implies that the estimates capture the full range of uncertainty, nor is such a claim made
in the TSDs or the rulemaking’s supporting documents. Rather, EPA attempted to capture a
reasonable range of uncertainty using information available in the peer-reviewed literature and
the uncertainty analysis built into the models themselves. Using three models rather than one
helps address, but does not eliminate, uncertainty associated with model choice.
Along with the four selected SC-CO2 estimates for each emissions year, the IWG presented more
detailed information about the full distribution of SC-CO2 estimates for emissions in the year
2020. Specifically, the 2013 TSD reports information on the full distribution of the SCC
estimates for emissions in year 2020 for each model, scenario, and discount rate combination ,
including the low-end percentiles (i.e., 1st, 5th and 10th percentiles). In addition, as noted
above, complete distributions for all emissions years are available upon request. While the IWG
did not present a summary 5th percentile estimate (i.e., pooling results across the models and
scenarios) for use in regulatory impact analysis, for reasons discussed below, the model and
scenario specific statistics for 2020 provide a general sense of how the 5th percentile relates to
the mean. Furthermore, as we note above EPA has provided the full set of Monte Carlo modeling
results to outside researchers upon request and will continue to do so. This information may be
used to calculate a full set of 5th percentile summary statistics that are comparable to the 95th
percentile estimates provided in the TSDs.
As the 2010 TSD discusses, the SC-CO2 estimates derived from the three integrated assessment
models have several significant limitations that could lead to a substantial underestimation of the
SC-CO2. These limitations include the incomplete treatment and monetization of noncatastrophic damages, the incomplete treatment of potential “catastrophic” damages, and
uncertainty in extrapolation of damages to high temperatures. The IPCC Fourth Assessment
Report, which was the most current IPCC assessment available at the time of the IWG’s 20094-105

2010 review, discussed these limitations and concluded that it was “very likely that [SCC]
underestimates” climate change damages. Based on the current scientific understanding of
climate change and its impacts, and on the limitations of the IAMs in quantifying and monetizing
the full array of potential “catastrophic” and non-catastrophic damages, EPA and the other
members of the IWG concluded that the distribution of SCC estimates may be biased
downwards. Since then, the peer-reviewed literature has continued to support this conclusion.
For example, the IPCC Fifth Assessment report observed that SCC estimates continue to omit
various impacts that would likely increase damages. The 95th percentile estimate was included in
the recommended range for regulatory impact analysis to address these concerns.
In addition, as acknowledged in the 2010 TSD, the SC-CO2 estimates derived from the three
IAMs did not take into consideration the possibility of risk aversion. That is, individuals may
have a higher willingness-to-pay to reduce the likelihood of low-probability, high-impact
damages than they do to reduce the likelihood of higher-probability, lower-impact damages with
the same expected cost. The inclusion of the 95th percentile estimate in the SC-CO2 values was
also motivated by this concern. In contrast, EPA is not aware of systematic upward biases in the
estimates comparable to the downward biases discussed above. For this reason, while EPA and
other members of the IWG have been fully transparent regarding the entire range of uncertainty
reflected in the probability distributions, EPA did not include a 5th percentile estimate in the
selected range for regulatory impact analysis.
The choice of the mean or the median as a measure of central tendency depends on the context.
In skewed distributions, such as for the SC-CO2 estimates, the median will often give a more
“typical” outcome, while the mean will give full weight to the tails of the distribution. In some
cases, the typical outcome is of most interest. For example, in describing household incomes the
median is most often used because the focus is on understanding the income of the typical
household, and using the mean might distort this picture by giving undue weight to a small
number of very wealthy households. In the climate change context, however, sound decisionmaking requires consideration of not only the typical or most likely outcomes, but also less likely
outcomes that could have very large (or small, or even negative) damages (the tails of the
distribution). Use of the median to represent the SC-CO2 in a regulatory impact analysis would
not necessarily lead to the most efficient policy choice that uses resources wisely to mitigate
potential climate impacts (e.g., maximize the expected net benefits). In this case, EPA believes
that the mean is the appropriate measure of central tendency.
Application to the Analyses of Illustrative Benefit-Cost Scenarios for New Sources
EPA agrees with those supportive of the application of the estimates. Most commenters who
disagreed with use of the SC-CO2 specifically disagreed with the use of a domestic estimate and
different discount rates. See section 4.4-3 for EPA’s detailed response to comments on the use
of a global versus domestic SCC. See section 4.4-7, for EPA’s detailed response to comments
about the selection of the discount rate. See also EPA’s detailed response to comments on the
IAMs, damage functions, socioeconomic scenarios, and the time horizon in 4.4-1 and EPA’s
response to comments regarding peer review and other process-related issues for SCC, including
the documentation discussing SC-CO2 methodology, in 4.4-5.
Regarding the recommendation to detail the significant and unquantified climate effects in the
RIA, i.e. the impacts currently omitted from the IAMs used to estimate the SC-CO2, EPA has
4-106

updated the RIA discussion to reference several publications that identify and discuss some of
the important, unquantified climate effects. EPA notes, however, that it is not possible at this
time to provide a precise list of each model’s treatment (i.e., included, excluded) of climate
impacts. Instead, the TSDs present a robust discussion of this key analytical issue, e.g., how
each model estimates climate impacts, the known parameters and assumptions underlying those
models, and the implications of incomplete treatment of impacts (catastrophic and noncatastrophic) for the SC-CO2 estimates. Moreover, the discussion in the SC-CO2 TSD
underscores the difficulty in accurately distilling the models’ treatment of impacts in table-form.
Most notably, the use of aggregate damage functions—which consolidate information about
impacts from multiple studies—in two of the models poses a challenge in listing included
impacts. For example, within the broad agricultural impacts category, some of the sub-grouped
impacts are not explicitly modeled but are highly correlated to other subcategories that are
explicitly modeled. EPA agrees that it is important to update the SC-CO2 periodically to
incorporate improvements in the understanding of greenhouse gas emissions impacts and will
continue to follow and evaluate the latest science on impact categories that are omitted or not
fully addressed in the IAMs. As previously noted, EPA and the other IWG members will seek
external expert advice on the technical merits and challenges of potential approaches to update
the damage functions in future revisions to the SCC estimates. Finally, the RIA also continues to
discuss climate change impacts, specifically an overview of the 2009 Endangerment Finding and
climate science assessments released since then (see RIA, Chapter 3).
Regarding comments about application of the SC-CO2 to marginal and non-marginal emission
impacts: The SC-CO2 is an estimate of the marginal benefit of a net one-ton reduction in CO2
emissions. The SC-CO2 estimates are multiplied by estimates of net CO2 emissions changes in a
given year to calculate the value of the benefits associated with the policy action that achieves
the emissions reductions. Given that the definition of marginal in this context relates to the
change in CO2 emissions relative to the global level of CO2 emissions, the SC-CO2 is the
appropriate metric to use in estimating the benefits of emissions changes in the illustrative
analyses presented in the RIA. The applicability of the SCC to monetizing the benefits of an
emissions change is not related to rulemaking’s costs or economic impacts. These factors are
independent of the SC-CO2 estimates and are therefore not related to the applicability of the SCCO2. The estimation of the social benefits of avoiding or reducing CO2 emissions is a separate
and independent calculation from the analysis of compliance costs and impacts on industry
competitiveness. See Response to Comments, section 3.2, for more detailed response to
comments on the conclusions of the RIA, including the primary conclusion that there will be
negligible costs as a result of the proposed regulation. RIA Chapter 5 presents conclusions from
the illustrative analyses.
EPA disagrees with the comments that characterize the SC-CO2 as an illusory or otherwise
unsupportable metric. EPA acknowledges uncertainty in the SCC estimates but disagrees that the
uncertainty is so great as to undermine use of the SCC estimates in regulatory impact analysis.
The uncertainty in the SCC estimates is fully acknowledged and comprehensively discussed in
the TSDs and supporting academic literature. While uncertainty must be acknowledged and
addressed in regulatory impact analyses, even an uncertain analysis provides useful information
to decision makers and the public. For example, if an analysis shows that benefits of a policy
option consistently do (or do not) justify costs even over a broad range of estimates, this may
increase confidence in the robustness of this conclusion. Conversely, if choices among parameter
4-107

estimates within a plausible range significantly affect the conclusions of the analysis, this is an
important consideration in deciding how to weigh the analytical results in the decision making
process. The presence of uncertainty is thus not a reason to exclude the best available estimates
of quantified/monetized benefits, as long as it is appropriately characterized. Rather, good
regulatory practice requires that agencies use the best available scientific, technical and
economic information to derive the best estimates of costs and benefits that they can, and then
communicate to the public the limitations and uncertainties of the analyses. As discussed in the
OMB Response to Comments on SC-CO2, this is what EPA and all of the members of the IWG
have attempted to do in developing and discussing the SC-CO2 estimates. As noted in the SCCO2 TSDs, the EPA and the other IWG members are committed to periodic updates in the
estimates to reflect ongoing developments in our understanding of the science and economics of
climate change, including the treatment of uncertainty. Moreover, as previously stated, EPA has
determined that the current SC-CO2 estimates continue to represent the best scientific
information on the impacts of climate change available in a form appropriate for incorporating
the damages from incremental CO2 emissions changes into regulatory analyses. See also
Coalition for Responsible Regulation v. EPA, 684 F. 3d at 119-20 (“State and Industry
Petitioners assert that EPA ‘delegated’ its judgment to the IPCC, USGCRP, and NRC by relying
on these assessments of clime-change science. This argument is little more than a semantic
trick….EPA simply did here what it and other decision-makers often must do to make a sciencebased judgment: it sought out and reviewed existing scientific evidence to determine whether a
particular finding was warranted. It makes no difference that much of the scientific evidence in
large part consisted of ‘syntheses’ of individual studies and research. Even individual studies
and research papers often synthesize past work in an area and then build upon it. This is how
science works. EPA is not required to re-prove the existence of the atom every time it
approaches a scientific question.”)
EPA disagrees with the comment that SC-CO2 values are not trusted because "they have
increased dramatically in recent years" due in part to the many assumptions required for
calculating SC-CO2. Commenter also believes SC-CO2 values are easily manipulated. EPA and
other federal agencies had received multiple public comments urging them to update the
estimates. In response to these comments and consistent with the 2010 commitment to
periodically revise the SC-CO2 estimates, in 2013 the IWG released an update to the SC-CO2
estimates that maintained the same methodology underpinning the previous estimates, but
applied the most current versions of the three IAMs. The science underlying the assessment and
valuation of climate change impacts is constantly evolving. Since the publication of the initial
SC-CO2 estimates in 2010, the representation of the science and economic consequences of
climate change in the three IAMs has improved. The 2013 SC-CO2 technical update allowed the
SC-CO2estimates to reflect these improvements. Some of the model revisions tended to increase
the value of SC-CO2 while others tended decreased it. The updated values reflected the net
effect of all of those changes. None of interagency working group’s 2010 modeling decisions
were revisited as a part of the 2013 update. The 2013 update used the same approach and
assumptions as the 2010 analysis, but with the latest version of each of the three models
available. In addition, the TSDs fully discuss the sensitivities of the SC-CO2 and how the
interagency working group explored those sensitivities. See also 4.4-4 for discussion about
treatment of uncertainty.

4-108

EPA strongly disagrees with the comment that climate change is an artifact of modeling, Global
Circulation Models have no connection to the real world, and SCC is therefore a model of
models. See 4.4-1, for detailed response to comments criticizing the IAMs and section 4.1 for
response to comments climate change science.
Regarding the comments about quantification versus monetization of the climate benefits and the
comment that the Agency has not provided a single quantifiable climate benefit of the proposed
rule, EPA disagrees and notes that it has in fact provided the estimated value of climate benefits.
The climate benefits estimates have been calculated using the estimated values of marginal
climate impacts, known as the social cost of carbon (SC-CO2), presented in the Technical
Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact
Analysis under Executive Order 12866. The SC-CO2 is a metric that estimates the monetary
value of impacts associated with marginal changes in CO2 emissions in a given year. It includes
a wide range of anticipated climate impacts, such as net changes in agricultural productivity and
human health, property damage from increased flood risk, and changes in energy system costs,
such as reduced costs for heating and increased costs for air conditioning. It is typically used to
assess the avoided damages as a result of regulatory actions (i.e., benefits of rulemakings that
have an incremental impact on cumulative global CO2 emissions). In order to calculate the dollar
value for emission reductions, the SC-CO2 estimate for each emissions year is applied to changes
in CO2 emissions for that year, and then discounted back to the analysis year using the same
discount rate used to estimate the SC-CO2. While the impacts of CO2 emissions changes, such as
sea level rise, are estimated within each integrated assessment model as part of the calculation of
the SC-CO2, it is the resulting monetized damages that are relevant for conducting the benefitcost analysis. As such, it is the SC-CO2 estimates that are used in the RIA to estimate the
welfare effects of quantified changes in CO2 emissions. See, e.g. RIA chapter 5.
Regarding the comments on leakage, specifically that multiplying the SC-CO2 values by
estimated CO2 reductions within the power sector only is problematic because the SC-CO2
should only be applied to estimated net changes in global CO2 emissions, EPA notes that it has
applied the SC-CO2 estimates to the best available estimate of the net emissions impact in the
illustrative analyses in the RIA (see RIA Ch 5). EPA recognizes that this is an important issue for
analysts to consider in determining the net CO2 reductions to be valued in an RIA but notes that
it does not affect the calculation of the SC-CO2 itself, which is an estimate of the marginal
benefit of a net one-ton reduction in CO2 emissions. The SC-CO2 estimates are multiplied by
estimates of net GHG emissions changes to calculate the value of benefits associated with a
policy action in a given year. It is in the estimation of net GHG emissions, and not the SC-CO2,
that any leakage should be accounted for. Furthermore, as discussed in RTC Section 3.2 and in
the RIA Chapter 4, EPA’s modeling, conducted using the Integrated Planning Model (IPM), as
well as modeling conducted by the U.S. Energy Information Administration (EIA) showed that
new generating capacity built through the period of analysis would be in compliance with the
standard even in the baseline scenario and, as a result, the rule would not lead to changes in
behavior. This finding held true even under a number of alternative scenarios. As a result, the
EPA projected there would be negligible costs, benefits, energy impacts (including changes to
electricity prices), employment impacts, or economic impacts associated with the rule in the
period of analysis. (See Chapter 4 of the RIA.) Based on these conclusions, we would not expect
business relocation, impacts to manufacturing, or the transfer of industrial activity to other
countries.
4-109

Regarding the recommendation for USG guidance on the application of the SC-CO2, specifically
guidance that would clarify the TSD’s recommendation to use all four SC-CO2 estimates in
rulemaking analyses, EPA first notes that it has followed the current guidance to consider all four
values in regulatory impact analysis. EPA agrees that consistent and appropriate application of
the SC-CO2 estimates is important. EPA will inform OMB of this comment requesting additional
guidance of the application of the SC-CO2 to regulatory impact analysis.
Value of goods and services whose production is associated with CO2 emissions
Comment 4.4-9: Commenter 10951 states EPA's use of the Administration's Social Cost of
Carbon values is misplaced. The SCC values assume that coal usage only create damage and are
blind that the world uses coal because it is a reliable, low the benefits of coal usage, the SCC
values have no worth at all in measuring whether reducing CO2 emissions would be a positive or
a negative. The attached report The Social Costs of Carbon No, the Social Benefits of Carbon
fills in the blank that the SCC values and EPA misses. The report compares the benefits of using
coal with the SCC values. As the report concludes: Of primary importance, the successful
development and utilization of fossil fuels, which generate CO2, facilitated successive industrial
revolutions, created the modern world, and enabled the high quality of life currently taken for
granted. There is a strong causal relationship between world GDP and CO2 emissions over the
past two centuries, and this relationship is forecast to continue for the foreseeable future. We
compared these indirect CO2 benefits to the SCC estimates. Commenter 10951 discussed the
benefits of coal use and incorporated these benefits in a comparison of the CO2 costs and benefits
(on a normalized per ton basis) using the SCC estimates and assumptions found that the current
benefits clearly outweigh any hypothesized costs by, literally, orders of magnitude: The benefitcost (BC) ratios range up to more than 200-to-1 (Figure AB-1). We utilized forecast data to
estimate B-C ratios through 2040 and found that future benefits also greatly exceed hypothesized
costs by orders of magnitude: In the range of 40-to-1 to 400-to-1. To place these findings in
perspective, normally, B-C ratios in the range of 2-to-1 or 3-to-1 are considered favorable. Thus,
our main conclusion is that the benefits of CO2 overwhelmingly outweigh estimated CO2 costs
no matter which SCC estimates or assumptions are used. In fact, the SCC estimates are relatively
so small as to be in the statistical noise of the estimated CO2 benefits. Thus, on balance, the
benefits of coal usage dramatically outweigh the purported harms. EPA's view seems to be that
in order to reduce CO2 emissions, nations should restrict coal usage and move to more
expensive, less reliable energy sources. But doing so will come at a cost, and that cost, as the
report and the discussion below shows, will be a very large net negative.
Commenter 3593 states estimates of the social cost of carbon that take negative values (i.e.,
because on net carbon creates more benefits than costs) document the relative value of using coal
to generate electricity and demonstrate that the benefits of coal outweigh any reasonably
estimated harm to public health and the environment, as discussed above.
Commenter 9396 states that the IWG failed to account for various direct benefits such as
increased growing seasons in some regions as well as indirect benefits arising from the positive
attributes of GHG producing activities. These include the provision of services that result in
direct health benefits such as a safer food supply, clean water, and access to health benefits
associated with heating, cooling and electricity.
4-110

Response 4.4-9: Rigorous evaluation of benefits and costs is a core tenet of the rulemaking
process. EPA agrees that these are important issues that may be relevant to assessing the impacts
of policies that reduce CO2 emissions. However, these issues are not relevant to the SC-CO2
itself. The SC-CO2 is an estimate of the net economic damages resulting from CO2 emissions,
and therefore is used to estimate the benefit of reducing those emissions.
A rule that affects CO2 emissions may also affect the production or consumption of goods and
services, in which case it could create costs and benefits for businesses and households that
either produce or use those goods and services. These costs and benefits are important to include
in an analysis of the rule’s impacts, but are not a result of changes in CO2 emissions. The SCCO2 is not a measure of social welfare from the consumption of goods and services whose
production results in CO2 emissions, or other positive or negative externalities associated with
the production of those goods and services. In other words, the SC-CO2 is just one component of
a larger analysis that includes consideration of many other potential impacts, including labor
market changes, energy security, electricity reliability, and changes in emissions of other
pollutants, among others.
Comment 4.4-10: Commenter 3594 states while OREP appreciates the approach taken, we are
also concerned that the proposed standards are based on SCC calculations that reflect
conservative understandings of the impacts of climate change, due to the rapid changes in
climate science, and an approach which is incapable of considering impacts and feedbacks that
are just now becoming visible, and either cannot yet be fully verified, or have not yet had a
chance to be incorporated. While there is not yet a basis for fully evaluating to collec tive costs
of these impacts, recent history, including the recent revision to the SCC framework, shows that
the cost will almost certainly be a significant addition to the costs used here. We therefore ask
that this expectation be duly considered and acknowledged in so far as possible in the
development of these standards.
Response 4.4-10: Please see RIA Chapter 3 for discussion about the SC-CO2 estimates as well
as response to comments 4.4-1 for detailed response to comments on the SC-CO2 methodology.
Also, see response to comments Section 3.1 and 3.2 and RIA Chapter 4, which discusses the
analysis of benefits and costs and EPA’s modeling, conducted using the Integrated Planning
Model (IPM), as well as modeling conducted by the U.S. Energy Information Administration
(EIA), that showed that new generating capacity built through the period of analysis would be in
compliance with the standard even in the baseline scenario and, as a result, the rule would not
lead to changes in behavior. This finding held true even under a number of alternative scenarios.
As a result, the EPA projected there would be negligible costs, benefits, energy impacts
(including changes to electricity prices), employment impacts, or economic impacts associated
with the rule in the period of analysis. (See Chapter 4 of the RIA.)
Comment 4.4-11: Commenter 10036 states the EPA RIA is somewhat contradictory on the topic
of benefits, in that it concludes that there will be no benefit from the rule and then hypothesizes
circumstances in which there would be benefits. Commenter continues, stating that the proposed
EGU New Source GHG Standards are not expected to change GHG emissions for newly
constructed EGUs and are anticipated to yield no monetized benefits and impose negligible
costs, economic impacts, or energy impacts on the electricity sector or society. However, the
RIA's Chapter 5 addresses the concept of "Social Cost of Carbon," and references a multi-agency
4-111

report docketed in an earlier rulemaking 49. Commenter 10036 adds there is no evidence
presented in the RIA or the proposed rule that the Social Cost of Carbon report or analysis
underwent a peer review consistent with the previously cited OMB requirements.
Response 4.4-11: EPA disagrees that there will be no benefits from the final rulemaking. See
response to comments 3.2-1 for EPA’s detailed response to this comment, including discussion
about the analysis of costs and benefits in the rulemaking. Regarding the social cost of carbon,
see response to comments 4.4-5 for detailed discussion regarding peer review and consistency
with OMB guidance.

4.5

MISCELLANEOUS

Comment 4.5-1: Commenter 8925 states the social value of non-CO2 GHGs should be
considered (e.g., for methane, nitrous oxide, and fluorinated gases), adding that the 2010 USG
acknowledges this issue but provides no clear guidance.
Response 4.5-1: As discussed in Section III.G of the preamble to the final rulemaking, separate
emission limits for non-CO2 GHGs, such as methane, were not established because they
represent less than 1 percent of the total estimated GHG emissions, on a CO2-equivalent basis,
from fossil-fuel fired electric power generating units. Therefore, EPA did not quantify potential
changes in non-CO2 GHGs in the illustrative analyses.
Comment 4.5-2: Commenter 9514 states it is important to emphasize that section 111 does not
require a traditional cost-benefit analysis, but simply requires that an agency's determination of
BSER impose costs that are not exorbitant or too high for the industry to bear. Commenter 9514
adds that CO2 impacts should be analyzed along with the benefits from reduced quantities of cooccurring pollutants.
Response 4.5-2: The commenter correctly quotes the caselaw regarding consideration of costs
under section 111 (a). In addition, EPA’s analysis in RIA chapter 5 is consistent with the
commenter’s suggestions.
Comment 4.5-3: Commenter 9685 states the terminology "social cost of carbon" is wrong and
misleading. As EPA knows, greenhouse gases, including carbon dioxide, are not "Carbon".
There are more greenhouse gases than carbon dioxide. Methane (natural gas) is an important
greenhouse gas. The IPCC2013 Summary for Policy Makers has acknowledged that the global
warming potential (GWP) of methane has been underestimated. EPA has to address the problem
in all of its analyses, including the planned regulation of new and/or existing natural gas plants.
This major error also needs to be addressed in the IAMs. Humans exhale carbon dioxide (CO2),
not carbon.
Response 4.5-3: EPA disagrees with the commenter, who has incorrectly assumed that the
common shorthand of “carbon” means that the social cost of carbon dioxide (SC-CO2) estimates
have not been adjusted to account for the mass of carbon in carbon dioxide emissions. One of
49

http://www.epa.gov/oms/climate/regulations/scc-tsd.pdf.
4-112

the steps in the estimation of the SC-CO2 used in this rulemaking is to translate the molecular
weight of carbon to the molecular weight of carbon dioxide. Furthermore, the RIA for the
proposed rulemaking clearly defined the shorthand of “SCC” or “social cost of carbon” as
accounting for the carbon in carbon dioxide emissions: “The social cost of carbon (SCC) is a
metric that estimates the monetary value of impacts associated with marginal changes in CO2
emissions in a given year” (RIA at proposal, section 5.7.1). The RIA for this rulemaking and the
supporting documentation, including the social cost of carbon technical support documents, all
clearly show that the units are denominated in dollars per ton CO2 emissions. EPA is well aware
that carbon dioxide is not the only greenhouse gas and that methane is an important greenhouse
gas. See section III.G of the preamble to the final rulemaking for discussion about why this
action regulates only emissions of CO2. See Response to Comments section 4.5, comments 4.4-1
and 4.4-2, for EPA’s response to comments on the IAMs used to estimate the SC-CO2 and the
climate science components of these models, respectively.
Comment 4.5-4: Commenter 8925 states regulatory analysis applications should be sure to
estimate changes in net global emissions associated with a proposed rule, or justify why it is
unnecessary because a proposed rule may have emissions implications beyond the bounda ries of
what is regulated (i.e., "leakage"), and those effects should be accounted for to properly estimate
CO2 reduction benefits.
Commenter 9396 Emissions Exports: The IWG fails to account for economic migration of
emissions; as one country's efforts to reduce GHG emissions makes manufacturing more
expensive, the economic activity shifts to other countries that are more climate intensive
resulting in a net increase in emissions.
Commenter 9401 states history provides an important lesson on how increased energy costs have
both a direct and indirect impact on reduced employment and GHG leakage. Figure 2 details the
relationship between rising natural gas prices and falling employment in the manufacturing
sector. According to EIA, electricity prices rose 210 percent from 1999 to 2008. During the time
period of 2000 to 2008, the manufacturing sector shut down an estimated 54,905 facilities,
decimating the economy of thousands of communities across the country. During that same time
period, net manufactured goods imports grew by 46 percent, a clear indication of loss of US
manufacturing competitiveness. A key point that needs to be emphasized is that if electricity
prices rise because of the EPA GHG rule, manufacturing employment and GHG leakage would
have the same effect as higher natural gas prices.
Commenter 9401 states EPA has not performed an analysis of carbon leakage which will occur
as a result of the rule, thus overstating the benefits. Because manufacturing uses 25 percent of all
U.S. electricity, we are significantly impacted by electricity prices. Figure 1 illustrates that coal
consistently provides the lowest cost energy to generate electricity. The rule would stop the use
of coal for use in new power generating units and deprive manufacturing from the lowest cost
power generation, thereby impacting competitiveness. It is important to note, that having coal as
a major component of electricity fuel diversity provides for increased reliability. Higher
electricity costs and reduced reliability will shift EITE industries to other countries along with
their jobs and GHG emissions, and accomplish nothing. The EPA must perform an analysis of
carbon leakage that will occur as a result of the rule. Not having done so, the EPA has overstated
the projected benefits of emission reductions and not accounted for the associated loss of
4-113

economic activity and jobs moving outside the U.S. Without including the GHG leakage, any
benefits are arbitrary and capricious. And, since most countries have higher power generation
GHG rates, it would result in a net increase in global GHG emissions.
Commenter 9401 states history provides an important lesson on how increased energy costs have
both a direct and indirect impact on reduced employment and GHG leakage. Figure 2 details the
relationship between rising natural gas prices and falling employment in the manufacturing
sector. According to EIA, electricity prices rose 210 percent from 1999 to 2008. During the time
period of 2000 to 2008, the manufacturing sector shut down an estimated 54,905 facilities,
decimating the economy of thousands of communities across the country. During that same time
period, net manufactured goods imports grew by 46 percent, a clear indication of loss of US
manufacturing competitiveness. A key point that needs to be emphasized is that if electricity
prices rise because of the EPA GHG rule, manufacturing employment and GHG leakage would
have the same effect as higher natural gas prices.
Response 4.5-4: EPA notes that it has applied the SC-CO2 estimates to the best available
estimate of the net emissions impact in the illustrative analyses in the RIA (see RIA Ch 5). EPA
recognizes that this is an important issue for analysts to consider in determining the net CO2
reductions to be valued in an RIA but notes that it does not affect the calculation of the SC-CO2
itself, which is an estimate of the marginal benefit of a net one-ton reduction in CO2 emissions.
The SC-CO2 estimates are multiplied by estimates of net GHG emissions changes to calculate
the value of benefits associated with a policy action in a given year. It is in the estimation of net
GHG emissions, and not the SC-CO2, that any leakage should be accounted for.
Furthermore, as discussed in RTC Section 3.2 and in the RIA Chapter 4, EPA’s modeling,
conducted using the Integrated Planning Model (IPM), as well as modeling conducted by the
U.S. Energy Information Administration (EIA) showed that new generating capacity built
through the period of analysis would be in compliance with the standard even in the baseline
scenario and, as a result, the rule would not lead to changes in behavior. This finding held true
even under a number of alternative scenarios. As a result, the EPA projected there would be
negligible costs, benefits, energy impacts (including changes to electricity prices), employment
impacts, or economic impacts associated with the rule in the period of analysis. (See Chapter 4 of
the RIA.) Based on these conclusions, we would not expect business relocation, impacts to
manufacturing, or the transfer of industrial activity to other countries.
Comment 4.5-5: Commenter 9687 states EPA has continued to knowingly use incorrect PM2.5
health effects data (factor of 40 too high) that needs to be corrected in both the Endangerment
Finding and the Social Cost of Carbon used to justify this rule. Commenter includes excerpt of
other document to back up point.
Response 4.5-5: The PM2.5 health data is not used to calculate of the social cost of carbon
dioxide. The health impacts of particulate matter are treated separately. See Chapter 3 of the
final RIA for a discussion about health co-benefits and sections 4.1 and 4.2 of this response to
comments for discussion about the 2009 Endangerment Finding.
Comment 4.5-6: Commenter 9687 states serious questions have been raised by Congress about
the current standards for PM2.5 and Ozone regulations which are used to justify the
4-114

Endangerment Finding, the Social Cost of Carbon, and the proposed rule for new electric power
plants.
Response 4.5-6: The standards for PM2.5 and ozone are not used to justify the Endangerment
Finding nor are they used to justify the social cost of carbon. The attachments provided by the
commenter do not support the commenter’s assertion. Rather, the commenter has provided a
letter written by a member of Congress to an EPA employee with questions about the 1997
NAAQS for ozone and particular matter, which are not relevant to this rulemaking action, and
EPA personnel matters that are likewise not relevant.
Comment 4.5-7: Commenter 9600 calls for withdrawal of this proposal, if for no other reason
based on its own conclusion that it would have no benefit including failing to accomplishing any
reductions in CO2 emissions from coal-fired EGUs over the base case modeling period. Further,
Commenter 9600-5985 states the natural gas pipeline system will have to undergo considerable
modifications and upgrades to provide natural gas on a "firm" basis to a baseload EGUs. Natural
gas pipelines are simply not sized adequately nor are they located where baseload EGU's would
be built.
Response 4.5-7: Regarding comment that the proposal should be withdrawn and that it has no
benefit, EPA disagrees. See Response to Comments, Section 3.1.
Comment 4.5-8: Commenter 10961 states the EPA's "endangerment finding" on carbon dioxide
incorporates an enormous body of scientific information demonstrating that greenhouse gas
emissions must be reduced immediately in order to avoid catastrophic climate impacts. The
Clean Air Act requires binding limits on any pollutant that threatens human health and welfare,
and carbon dioxide threatens human welfare as no other pollutant has, to the point of existential
hazard. The EPA's program for quantifying health hazards and translating them into missions
limits is robust, well-developed, and equally applicable to the damage that excess carbon dioxide
will cause. The Clean Air Act has brought extraordinary material benefits to America from
reducing pollutants that erode our health and economy, and the same challenge now exists from
carbon dioxide. Far from being unsuited to the Clean Air Act as some political opponents
suggest, global warming's lethal threat to human survival decidedly qualifies as the th reat to
human health and welfare the Clean Air Act was meant to avert.
Response 4.5-8: EPA thanks the commenter for this comment and agrees that the Clean Air Act
has resulted in notable social benefits.
Comment 4.5-9: Commenter 9685 states that EPA's claim - based on the analysis presented in
Chapter 5 of the RIA, the EPA projects that this proposed rule will result in negligible CO2
emission changes, quantified benefits, and costs by 2022 - is not valid and any use of the SCC to
justify the proposed rule is not valid as discussed below.
Response 4.5-9: See EPA’s response to comments, sections 3.1 and 3.2, for EPA’s response to
comments regarding the emission and economic impacts of this rulemaking action. See 4.5-8 for
EPA’s response to comments regarding the application of SC-CO2 to the RIA.
Comment 4.5-10: Commenter 9685 states EPA's MAGICC is used to "calibrate" the IAMs. The
IAM models are "calibrated to match the carbon cycle in the Model for the Assessment of
4-115

Greenhouse Gas Induced Climate Change (MAGICC)". There is no documentation on the input
variables used for the MAGICC run(s) nor is there data available to confirm the "calibration",
much less the data used to validate the current version 5.3 of MAGICC. MAGICC uses AR4
IPCC models for its own "calibration". EPA is fully aware of this major fatal flaw and has had
access to the information since the publication of the AR4 in 2007. Many agency members of the
IWG had employees involved in the IPCC reports and were equally awar e of the serious flaws in
the IPCC models, including the US taxpayer funded models. Model prediction errors were
acknowledged in the comments on the USGCRP assessment in 2006 and on the EPA
Endangerment Finding in 2008. Therefore, there is no excuse other than intentional omission for
the IWG to have failed to acknowledge the epic failure of the models.
Response 4.5-10: EPA strongly disagrees with the claim that the Agency has not provided
information necessary to understand the SC-CO2 estimates. The TSDs provide a complete record
of the methodology and assumptions used to develop the SCC estimates, including references to
the academic literature. Independent analysts have sought and received information from EPA
allowing them to implement the IWG’s SC-CO2 approach and modify it further if they choose. In
fact, one requestor noted publically that the EPA “modelers have been very open, collegial, and
helpful.”50 Additional documentation is available in the academic literature on the models
themselves. See also response to comments, section 4.4-5 for discussion about transparency and
opportunities for public comment on the SC-CO2.
Comment 4.5-11: Commenter 9685 states apparently errors were found that required a "revised"
Technical Support Document (issued in November 2013). There was no detailed explanation of
what the errors were that required a revision after only six months, who identified the errors, who
made the changes to the models to reflect these changes and how the models were validated.
EPA should promptly provide this information to the public by posting on regulations.gov for
this rule. The Technical Support Documents (TSDs) summarize the derivation of the SCC
estimates using the three Integrated Assessment Models and provides updated values of the SCC
that reflect "technical corrections" to the estimates released in May 2013. Unfortunately the
Interagency Working Group (in which EPA was an active participant) convened in 2009 has
operated in secret and the TSDs were developed in secret. There is no record of any comments
provided by EPA or any other agency participants. There is no identification of the EPA
participants. Commenter posits that EPA disregarded the transparency requirements expected of
any taxpayer funded government agency and as directed by President Obama. Providing
comprehensive comments on the rule is impossible without full disclosure by EPA on
regulations.gov.
Response 4.5-11: The commenter has incorrectly stated that there was no detailed explanation
about the corrections. OMB publicly announced the correction 51 and provided a public link to
the corrected TSD. Appendix B of that TSD identified the errors with complete detail, explained
how those errors were corrected, and described the effect of the correction; the TSD also

50
51

http://dailysignal.com/2013/11/06/white-house-reopens-the-scc/.
https://www.whitehouse.gov/blog/2013/11/01/refining-estimates-social-cost-carbon

4-116

presented the corrected estimates.52 Furthermore, EPA submitted a memo to the docket for this
rulemaking, EPA-HQ-OAR-2013-0495, that identified the SC-CO2 errors with complete detail,
explained how those errors were corrected, and described the effect of the correction, both with
respect to the SC-CO2 estimates and with respect to the analysis in the RIA. 53

52

See https://www.whitehouse.gov/sites/default/files/omb/assets/inforeg/technical-update-social-cost-of-carbon-forregulator-impact-analysis.pdf.
53
See EPA-HQ-OAR-2013-0495-0037 in www.regulations.gov.

4-117