The U.S. Environmental Protection Agency (EPA) has proposed a finding that greenhouse gases contribute to air pollution that may endanger public health or welfare. The agency's press release says "EPA’s proposed endangerment finding is based on
rigorous, peer-reviewed scientific analysis of six gases – carbon
dioxide, methane, nitrous oxide, hydrofluorocarbons, perfluorocarbons
and sulfur hexafluoride – that have been the subject of intensive
analysis by scientists around the world.
"The science clearly shows that
concentrations of these gases are at unprecedented levels as a result
of human emissions, and these high levels are very likely the cause of
the increase in average temperatures and other changes in our climate. The scientific analysis also confirms that climate change impacts human health in several ways."
How does EPA present evidence? I'm interested in how the agency presents evidence on this complex issue, because it can significantly impact people's ability to interpret it and respond. Specifics appear in Proposed Endangerment and Cause or Contribute Findings for Greenhouse Gases under the Clean Air Act. The proposed regulations will be published in the Federal Register in the near future, and written comments will be accepted for 60 days after that. (But don't hold your breath for federal regulations to take effect anytime soon; when I worked in air quality regulatory compliance in the '90s, controversial things moved slower than molasses.)
"Fact" sheet? First, I looked over the agency's Summary of the Science Supporting EPA’s Finding That Greenhouse Gases Threaten Public Health and Welfare (pdf here). It's pretty light, which is understandable, since it attempts in two pages to summarize key points distilled from a massive amount of research and analysis. But I believe the agency could have done better -- for example, they could have hyperlinked each statement to a more detailed information source (a task that would have been much easier if the scientists and analysts authoring EPA's source material had presented their findings in a more reader-friendly way). But what really bugs me is that EPA says "All of the points in this fact sheet come from the published scientific literature...." However, many of their statements aren't "facts" at all -- they're statements of probabilities about expected future events: For instance, the agency says "Climate change is expected to worsen regional ozone pollution, with associated risks in respiratory infection, aggravation of asthma, and premature death." and "Storm impacts are likely to be
more severe, especially along the Gulf and Atlantic coasts. Heavy
rainfall events are expected to increase, increasing the risk of
flooding, greater runoff and erosion, and thus the potential for
adverse water quality effects."
I wish they'd give us more credit, and not imply that facts and findings are the same thing -- it cheapens the whole process. There are plenty of facts to go around, but lots of the important evidence is based on scientific modeling and prediction techniques. Basically, what EPA will be doing over the next few years is managing dialogue and argumentation over public policy concerning greenhouse gas (GHG) emissions; EPA and various stakeholders will be stating claims, making controversial value judgments, and issuing recommendations. And I agree with experts on logic and decision-making that "as starting points for argumentation, facts are empirical knowledge derived from observation or experience over which there is no controversy."* A statement of probability (such as "The changes to our climate may increase the likelihood of extreme and high-impact events such as more intense hurricanes.") is also a legitimate starting point for argumentation, especially when supported with evidence and qualified with estimates about uncertainty. EPA should be more explicit in identifying what types of statements are being made.
Transparency. EPA's evidence is presented in something called a Technical Support Document (TSD) (download a 171-page pdf here). The TSD
"relies most heavily on existing, and in most cases very recent,
synthesis reports of climate change science and potential impacts,
which have gone through their own peer-review processes including
review by the U.S. Government. The information in this document has
been developed and prepared in a manner that is consistent with EPA's Guidelines
for Ensuring and Maximizing the Quality, Objectivity, Utility and
Integrity of Information Disseminated by the Environmental Protection
Agency. In addition to its reliance on existing and primarily
recent synthesis reports from the peer reviewed literature, it also
underwent a technical review by 12 federal climate change experts,
internal EPA review, and interagency review." Kudos to EPA for listing the core
references most heavily relied upon in preparing the TSD (see page 4).
Gauging confidence in the evidence, and likelihood of events. EPA does a decent job of explaining how they handle uncertainty about the evidence: "Uncertainties and confidence levels associated with the scientific conclusions and findings in this document are reported, to the extent that such information was provided in the original scientific reports upon which this document is based." EPA includes a summary of how they treat uncertainty. First, describing confidence: "On the basis of a comprehensive reading of the literature and their expert judgment, authors have assigned a confidence level to major statements on the basis of their assessment of current knowledge, as follows:
- Very high confidence: At least 9 out of 10 chance of being correct
- High confidence: About 8 out of 10 chance
- Medium confidence: About 5 out of 10 chance" and so forth.
This is a good way to qualify claims being made about future environmental impacts. EPA also explicitly identifies a scheme for assessing likelihood: "Likelihood refers to a probabilistic assessment of some well defined outcome having occurred or occurring in the future, and may be based on quantitative analysis or an elicitation of expert views. When authors evaluate the likelihood of certain outcomes, the associated meanings are: Virtually certain: >99% probability of occurrence; Very likely: 90 to 99% probability; Likely: 66 to 90% probability" and so forth (TSD pages 3, 6).
More graphics, please. The TSD
would be easier to absorb if EPA included graphics identifying
specific causes and their presumed effects. What they've given us are
sets of raw data, presented in traditional ways, accompanied by lengthy
narrative. (I can't resist a quick plug for the Tiny Soapbox methodology, something I developed for displaying ideas and supporting evidence. EPA could use something like this to display the specific claims they're making, and then link to the supporting evidence.)
More links, please. The bulk of EPA's 171-page TSD consists of charts, statements of findings, and references. The bibliography itself is 11 pages long. Overall this is a traditional document, not optimized for online consumption. Only a handful of the bibliographic references include a hyperlink to the source.
Plain English, please. One
of EPA's key sources, the U.S. Climate Change Science Program (CCSP),
calls its reports "synthesis and assessment products" (SAPs); these
deliverables are intended to "address the highest priorities for U.S.
climate change research, observation and decision-support needs." The
TSD provides a URL for information about the clearance and review
procedures for the CCSP SAPs... alas, it's a non-working URL (TSD page
3).
FYI: If you want just a quick summary of EPA's proposal, here are links to the New York Times story, the Yahoo! news story, and the EPA press release.
*R. Rieke and M. Sillars, Argumentation and Critical Decision Making (Addison Wesley, 4th Edition, 1997), page 23. Now in its 7th Edition; available from amazon.
