Junk Law: The CO2 Litigation of the State Attorneys General


On October 23, 2003, 12 states, three cities, and 14 advocacy groups sued the U.S. Environmental Protection Agency (EPA) for rejecting an October 1999 petition by the International Center for Technology Assessment (ICTA) and other eco-activists to regulate carbon dioxide (CO2) from new motor vehicles, and for withdrawing—as no longer representing the agency’s views—statements by Clinton-Gore EPA officials asserting the “potential” applicability of several Clean Air Act (CAA) regulatory provisions to CO2.

The D.C. Court of Appeals has scheduled oral argument in the CO2 case for April 8, 2005. Plaintiffs’ briefs are due on June 22, 2004. EPA’s brief is due on October 12, 2004. Usurpation by Litigation

To see why these suits are without merit, it suffices to ask two simple questions. First, why was the McCain-Lieberman Climate Stewardship Act, which seeks to impose mandatory controls on CO2 emissions, arguably the most controversial piece of legislation to be debated and voted on in the current Congress? Second, why is the Kyoto Protocol on climate change, which would require more stringent CO2 emission reductions, arguably the most controversial treaty to be considered by U.S. policymakers in the past six years?

The answer is that both Kyoto and McCain-Lieberman would fundamentally alter U.S. law and regulatory policy on energy production and use. The federal government has never regulated CO2 emissions. That is hardly surprising. Carbon dioxide is the inescapable byproduct of the carbon-based fuels—coal, oil, and natural gas—that supply 86 percent of all the energy Americans use. U.S. energy consumption is expected to increase by 34 percent over the next 20 years, and carbon-based fuels are expected to supply the lion’s share of that increase. The power to regulate CO2 is literally the power to cripple U.S. productivity, competitiveness, and growth.

The Senate preemptively rejected the Kyoto Protocol as too costly and unfair to the United States when, in July 1997, it passed the Byrd-Hagel resolution by a 95-0 vote. The Senate similarly rejected the McCain-Lieberman bill by 55-43 on October 30, 2003. Yet the recent CO2 lawsuits imply that Kyoto and McCain-Lieberman, in substance if not detail, are current law—a preposterous position. What the litigants are really trying to do is usurp Congress’ lawmaking power. They are attempting, through not-very-clever legalisms, to impose on the American people an energy-rationing system that Congress never approved. Worse yet, many of the litigants are attorneys general, sworn to uphold the Constitution of the United States. They deserve to have their bottoms paddled.

Although the litigants have not yet submitted their briefs to the U.S. Court of Appeals, the gist of their argument is well known from previous filings by ICTA and the attorneys general of Maine, Massachusetts, and Connecticut.2 In a nutshell, the AGs contend that EPA has a “mandatory duty” under the CAA to regulate CO2, and that failure to regulate is a “violation of that Act.” However, far from it being EPA’s duty to regulate CO2, EPA has no authority to do so. The plain language, structure, and legislative history of the Clean Air Act demonstrate that Congress never delegated such power to EPA .3

The AGs and the other anti-energy litigants somehow miss the obvious. The CAA provides distinct grants of authority to administer specific programs for specific purposes. It authorizes EPA to administer a national ambient air quality standards program, a hazardous air pollutant program, a stratospheric ozone protection program, and so on. Nowhere does it even hint at establishing a climate change prevention program. There is no subchapter, section, or even subsection on global climate change. The terms “greenhouse gas” and “greenhouse effect” do not appear anywhere in the Act.

Begging the Key Question

Lacking even vague statutory language to point to, the state attorneys general build their case on “definitional possibilities” of words taken out of context—a notoriously poor guide to congressional intent.

The AGs argue as follows:

1. CAA Section 302(g) defines “air pollutant” as “any … substance or matter which is emitted into or otherwise enters the ambient air.” Carbon dioxide fits that definition, and is, moreover, identified as an “air pollutant” in Section 103(g).

2. Section 108 requires EPA to “list” an air pollutant for regulatory action if the Administrator determines that it “may reasonably be anticipated to endanger public health and welfare.”

3. The Bush Administration’s Climate Action Report 2002 projects adverse health and welfare impacts from CO2-induced global warming, and EPA contributed to that report.


4. Hence, EPA must initiate a rulemaking for CO2.

The AGs’ argument may seem like a tight chain of reasoning. In reality, it is mere wordplay, a sophomoric attempt to turn statutory construction into a game of “gotcha.” The CO2 litigants beg the decisive question: Has Congress delegated to EPA the power to regulate CO2? When Congress enacted and amended the CAA, did it intend for EPA to administer a regulatory climate program? The delegation question is paramount, because agency rules are to implement laws, and the U.S. Constitution vests “all legislative powers … in a Congress of the United States.”4 Accordingly, “It is axiomatic that an administrative agency’s power to promulgate legislative regulations is limited to the authority delegated by Congress.”5 Indeed, “an agency literally has no power to act … unless and until Congress confers power upon it.” 6

No delegation of regulatory authority can be inferred from the fact that CO2 meets an abstract definition of “air pollutant” that applies equally well to oxygen and water vapor. Indeed, the very text cited by the AGs—Section 103(g)—admonishes EPA not to infer such authority. That provision concludes: “Nothing in this subsection shall be construed to authorize the imposition on any person of air pollution control requirements.” If nothing in Section 103(g) can authorize the imposition of control requirements, then the passing reference therein to CO2 as an “air pollutant” cannot do so.

As to the phrase “endanger public health and welfare,” it proves too much. It also applies to many substances that EPA does not—and may not—regulate under Section 108.

Section 108 gives EPA authority to set national ambient air quality standards (NAAQS), which determine allowable emission concentrations for certain pollutants, such as sulfur dioxide, nitrogen oxides, ozone, and particulates. EPA also regulates 53 ozone-depleting substances under Title VI. Such substances are emitted into the ambient air, and are believed to endanger public health and welfare. By the AGs’ “definitional” logic, EPA could dispense with Title VI and just use Section 108—a ridiculous proposition plainly at odds with congressional intent. 


It would make no sense for EPA to set NAAQS (allowable ambient levels) for ozone-depleting substances, because it is aggregate releases worldwide—not ambient concentrations in particular areas—that determine the degree of potential harm to public health and welfare. The only effective way to reduce the risk of ozone depletion is to ban—or severely restrict—the production and use of ozone-depleting chemicals. Section 108 provides no authority to ban chemicals of any kind.

To combat stratospheric ozone depletion, Congress twice added distinct new provisions—Title I Part B in 1977 and Title VI in 1990—precisely because existing authorities, including Section 108, were unsuited to the task of addressing global atmospheric issues. EPA General Counsel Robert E. Fabricant’s August 28, 2003 memorandum is worth quoting on this point:

In the CAA prior to its amendment in 1990, Congress specifically addressed the problem [of stratospheric ozone depletion] in a separate portion of the statute (part B of title I) that recognized the global nature of the issue and called for negotiation of international agreements to ensure world-wide participation in research and any control of stratospheric ozone-depleting substances. In the 1990 CAA amendments, Congress again addressed the issue in a discrete portion of the statute (title VI) that similarly provides for coordination with the international

community. Moreover, both incarnations of the CAA’s stratospheric ozone provisions contain express authorization for EPA to regulate as scientific

information warrants. In light of this CAA treatment of stratospheric ozone depletion, it would be anomalous to conclude that Congress intended EPA to address global climate change under the CAA’s general regulatory provisions, with no provision recognizing the international dimension of the issue and any solution, and no express authorization to regulate.7  Congress would have to amend the Act again, and expressly authorize EPA to regulate greenhouse gases in coordination with the international community, before the agency could lawfully develop and adopt mandatory controls on CO2 emissions.

Ignoring Context

To interpret a statute, one must not only read the words, but also pay attention to where they occur—their context.8 If Congress intended for EPA to regulate CO2, we would expect to find “carbon dioxide” mentioned in one or more of the CAA’s regulatory provisions. The AGs note that Section 103(g) describes CO2 as an “air pollutant”—but they fail to note that 103(g), which contains the CAA’s sole reference to “carbon dioxide,” is a non-regulatory provision. It directs the Administrator to develop “non-regulatory strategies and technologies” for preventing or reducing emissions of “multiple air pollutants,” including, among others mentioned, carbon dioxide.

The Supreme Court has held that, “Where Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.”9

Carbon dioxide’s “disparate exclusion” from the CAA’s regulatory provisions cries out for explanation. After all, CO2 is not some arcane or newly discovered compound, but a gas emitted in vastly greater quantities than any of those listed for regulation in, for example, Sections 107-109, Section 112, or Title VI. Moreover, the potential of CO2 emissions to enhance the natural greenhouse effect has been known to scientists since the 19th Century, and Congress has taken an interest in the subject since the late 1970s. It is difficult to avoid the conclusion that Congress acted “intentionally and purposely” when it did not mention “carbon dioxide” in the CAA’s regulatory provisions.

The AGs make no reference at all to Section 602(e), which contains the CAA’s sole reference to “global warming.” It, too, is a non-regulatory provision. It directs the Administrator to “publish” (i.e., research) the “global warming potential” of ozone-depleting substances. Section 602(e) also ends with a caveat: “The preceding sentence [referring to “global warming potential”] shall not be construed to be the basis of any additional regulation under this chapter [i.e., the CAA].”

The two caveats against inferring regulatory authority—one following the CAA’s sole mention of “carbon dioxide,” the other following the sole mention of “global warming”—are a matched pair. Since Congress adopted both provisions in 1990, we may presume that the pairing is deliberate. In any event, the CAA mentions carbon dioxide and global warming only in the context of non-regulatory provisions, and in each instance admonishes EPA not to construe the law as the AGs now profess to construe it.

Absurd Exercise in Futility

The AGs of Connecticut, Massachusetts, and Maine contend that EPA must begin the process of setting national ambient air quality standards for carbon dioxide. However, the NAAQS program, with its state-by-state implementation plans and county-by-county attainment and non-attainment designations, targets pollutants that vary regionally and even locally in their ambient concentrations. The NAAQS program has no rational application to a gas such as CO2, which, because of its long residence time (50-200 years), is well mixed throughout the global atmosphere.

Consider the possibilities. If EPA set a NAAQS for CO2 above current atmospheric levels, then the entire country would be in attainment, even if U.S. hydrocarbon fuel consumption were to suddenly double. Conversely, if EPA set a NAAQS for CO2 below current levels, the entire country would be out of attainment, even if all power plants, factories, and cars were to shut down. If EPA set a NAAQS for CO2 at current levels, the entire country would be in attainment—but only temporarily. As soon as global concentrations increased, the whole country would be out of attainment, even if U.S. emissions magically fell to zero.

Moreover, since even a multilateral regime like the Kyoto Protocol would only barely slow the increase in atmospheric CO2 concentrations, it is inconceivable how any state implementation plan (SIP) could pass muster under CAA Section 107(a), which requires each SIP to “specify the manner in which national primary and secondary air quality standards will be achieved and maintained within each air quality control region in each State” (emphasis added). The Fabricant memorandum gets this issue exactly right:

The long-lived nature of the CO2 global pool … would mean that any CO2 standard that might be established would in effect be a worldwide ambient air quality standard, not a national standard—the entire world would be either in compliance or out of compliance. Such a situation would be inconsistent with a basic underlying premise of the CAA regime for implementation of a NAAQS— that actions taken by individual states and by EPA can generally bring all areas of the U.S. into attainment with the NAAQS … The globally pervasive nature of CO2 emissions presents a unique problem that fundamentally differs from the kind of environmental problem that the NAAQS system was intended to address and is capable of solving.

When certain words in a statute lead to results that are “absurd or futile,” or “plainly at variance with the policy of the legislation as a whole,” the Supreme Court follows the Act’s “policy” rather than the “literal words.”10 Attempting to fit CO2 into the NAAQS regulatory structure would be an absurd exercise in futility, and plainly at variance with the Act’s policy of devising state-level remedies for local pollution problems—powerful evidence that when Congress enacted Section 108, it did not intend for EPA to regulate CO2.

Flunking Legislative History

The 12 AGs, three cities, and 14 advocacy groups are suing the Environmental Protection Agency on behalf of the International Center for Technology Assessment petition to regulate CO2 from new motor vehicles. However, the Clean Air Act’s legislative history also compels the conclusion that EPA may not regulate CO2, whether from stationary or mobile sources. The Senate version of the 1990 CAA Amendments, S. 1630 introduced by Max Baucus (D-Mont.), initially included a provision requiring EPA to set performance standards for CO2 emissions from new motor vehicles. In its deliberations, the Senate deleted that language from the bill. 


Moreover, when House and Senate conferees agreed on a final version of the 1990 CAA Amendments, they discarded Senate-passed language to make “global warming potential” a basis for regulating “manufactured substances” and to establish CO2 reduction as a national goal. Thus, when Congress last amended the CAA, it considered and rejected regulatory climate change prevention strategies. As the Supreme Court has stated: “Few principles of statutory construction are more compelling than the proposition that Congress does not intend sub silentio to enact statutory language that it has earlier discarded in favor of other language.”11

Congress similarly rejected proposals to establish CO2 reduction as a national goal, and to require limits on CO2 emissions from “major” sources, when it enacted the 1992 Energy Policy Act. A review of U.S. climate policy shows that, whenever Congress has legislated in this area, it has authorized the Executive Branch to engage in research, conduct international negotiations, and administer voluntary programs, but never to develop or adopt emission control requirements. The Senate’s 55-43 vote against the McCain-Lieberman Climate Stewardship Act is just the latest example of Congress’ longstanding opposition to regulatory climate policies.

Indeed, Congress has prohibited agencies from taking the kinds of actions the state attorneys general and ICTA demand. For example, in FY 1999, Congress enacted the Knollenberg provision as part of the appropriations bill (Veterans Affairs, Housing and Urban Development, and Independent Agencies) that funds EPA. The provision, authored by Rep. Joseph Knollenberg (R-Mich.), stipulates that, “None of the funds appropriated by this Act shall be used to propose or issue rules, regulations, decrees, or orders for the purpose of implementation, or in preparation for implementation, of the Kyoto Protocol.” Congress enacted the Knollenberg provision in six appropriations bills in FY 2000, and eight appropriations bills in FY 2001.12 Kyoto is first and foremost a treaty to regulate CO2. How could EPA regulate CO2 without gutting the Knollenberg provision? How could that not conflict with congressional intent?

Ignoring FDA v. Brown & Williamson


In Food and Drug Administration v. Brown & Williamson, (529 U.S. 120, 2000), the U.S. Supreme Court rejected arguments identical in form to those offered by ICTA and the state attorneys general. The Food and Drug Administration (FDA) claimed authority to regulate cigarette sales and advertising on the grounds that nicotine was a “drug” and cigarettes were “drug delivery devices” within the meaning of the Food, Drug, and Cosmetics Act. The Court rejected this argument from “definitional possibilities” and struck down FDA’s tobacco regulations as exceeding the agency’s delegated authority.

The Court partly based its holding on “common sense as to the manner in which Congress is likely to delegate a policy decision of such economic and political magnitude to an administrative agency.” Quoting a previous decision, 13 the Court stated that, “it is highly unlikely that Congress would leave the determination of whether an industry will be entirely, or even substantially, rate-regulated to agency discretion.” It is even more unlikely that Congress would leave the determination of whether to ration or suppress carbon-based fuels, to agency discretion. Whereas regulating tobacco would affect just one industry, regulating CO2 would affect entire economic sectors. As Fabricant observes:

The production and use of fossil fuel-based energy undergirds almost every aspect of the nation’s economy. For example, approximately 75 percent of the electric power used in the U.S. is generated from fossil fuel, and the country’s transportation sector is almost entirely dependent on oil. To the extent significant reductions in U.S. CO2 emissions were mandated by EPA, power generation and transportation would have to undergo widespread and wholesale transformations, affecting every sector of the nation’s economy and threatening its overall economic health.

Hence, Fabricant concludes:

In view of the unusually profound implications of global climate change regulation, it is unreasonable to believe that Congress intended “to delegate a decision of such significance in so cryptic a fashion.” Id. Junk Science Makes Bad Law

Has EPA “determined” that carbon dioxide emissions endanger public health and welfare, as the state attorneys general and ICTA claim? The Bush Administration’s Climate Action Report 2002 (CAR) is an alarmist document, forecasting that U.S. average temperatures will rise as much as 9° F in the 21st Century, and EPA was a key contributor to the report. However, the CAR’s scary climate scenarios are a rehash of the Clinton-Gore Administration’s report, U.S. National Assessment of the Potential Consequences of Climate Variability and Change, and the Bush Administration, in response to litigation by the Competitive Enterprise Institute, Senator James Inhofe (R-Okla.), and others, agreed that the National Assessment’s climate scenarios are “not policy positions or statements of the U.S. Government. ”14 More recently, the White House Office of Science and Technology Policy, to settle a related suit with CEI, acknowledged that the National Assessment was not “subjected” to federal data quality guidelines. 15

The National Assessment/CAR climate scenarios rely on two non-representative climate models—the “hottest” and “wettest” out of some 26 models available to Clinton-Gore officials. In addition, as University of Virginia climatologist Patrick Michaels discovered, and National Oceanic and Atmospheric Administration scientist Thomas Karl confirmed, the two underlying models—British and Canadian—could not reproduce past U.S. temperatures better than could a table of random numbers. 16 Models that cannot “hind-cast” past climate cannot be trusted to forecast future climate. At once biased and useless, the CAR flunks Federal Data Quality Act (FDQA) standards for utility and objectivity. Any rulemaking based upon it would be challengeable as arbitrary and capricious.


In any event, because the Clean Air Act provides no authority for regulatory climate strategies, EPA could not regulate CO2 even if the CAR scenarios were based on credible science—which they are not.

Power Grab

What drives the state attorneys general to peddle such claptrap? Partisan politics, regional economic warfare, and personal ambition all appear to be factors.

Eleven of the 12 AGs are Democrats and none is a Republican. If successful, the CO2 litigation would force the anti-Kyoto Bush administration to eat crow and implement Al Gore’s climate policy. It would effectively overturn one of the key policy consequences of the 2000 presidential elections.

Ten of the AGs are from states that derive little to none of their electricity from coal. 17 Because coal is the most carbon-intensive fuel source, CO2 regulation would make coal generation—and the industries dependent on it—less- (or non-) competitive. If successful, the AGs’ suit would tend to shift economic power to the Northeast. Just as the European Union hoped to use the Kyoto treaty to hobble the United States in global trade (Europe’s stagnant economic and population growth rates make Kyoto’s emission reduction targets comparatively easier to achieve), so some Northeast politicians view Kyoto-type policies as a way to handicap the more industrialized, energy-intensive economies of the Midwest and the faster-growing economies of the South and Southwest.

The AGs would also benefit personally if EPA were to classify CO2 as a regulated pollutant. Instantly, tens of thousands of hitherto law-abiding and environmentally responsible businesses—indeed, all fossil fuel users—would become “polluters,” and be in potential violation of the Clean Air Act. Since states have primary responsibility for enforcing the CAA, the AGs’ prosecutorial domain would grow by orders of magnitude.


The CO2 litigation of the state attorneys general is based on junk law. Boiled down to essentials, the AGs argue that the Clean Air Act compels EPA to implement the Climate Stewardship Act—a non-enacted bill—or even the Kyoto Protocol—a non-ratified treaty. It would be interesting to know how many billable hours the AGs and their staffs end up spending on this travesty of “public interest” litigation, and what it costs their states’ taxpayers. Too bad the AGs won’t have to pay back the cost out of their own pockets when they lose their suit.