This post was written with my colleague Ben Longstreth.
Opponents of EPA’s effort to cut carbon pollution from existing power plants under Clean Air Act Section 111(d) have recently been peddling a nonsensical argument that EPA has no authority to set those standards because of a drafting glitch in the 1990 Clean Air Act Amendments. The claim is that EPA may not regulate carbon from any industrial source, such as power plants, if the Agency has regulated hazardous air pollutants from that source under a different section of the Act (Section 112). There is no basis for this claim, which would convert two technical housekeeping provisions adopted on different pages of the 1990 amendments into a sweeping repeal of section 111(d) of the Act. It’s time to put this claim to rest.
Some background. Congress included Section 111(d) in the 1970 Clean Air Act Amendments to serve a gap-filling function: it provides authority to set standards for existing stationary sources of dangerous air pollutants that are not regulated pursuant to two other provisions -- Section 110 (covering criteria air pollutants, such as those that form smog) and Section 112 (covering hazardous air pollutants). Section 111(d) directed the Administrator to “establish emission standards for any existing source for any air pollutant (i) . . . which is not included on a list published under section 108(a) or 112(b)(1)(A) but (ii) to which a standard of performance under subsection (b) would apply if such existing source were a new source.” 43 Pub. L. 91-604 § 4(a); 84 Stat. 1684.
The 1990 Clean Air Act Amendments made extensive changes to the hazardous air pollution section and, in the process, deleted the specific subsection that the pre-1990 Section 111(d) had referenced. To fix this, Congress made a conforming amendment to Section 111(d) so that it no longer referred to the deleted portion of the hazardous air pollutant section.
The House and Senate bills each included different language correcting the reference to the deleted portion of Section 112. A legislative drafting glitch occurred when both of these duplicative amendments were included in the final statute, one on page 2467 of the Statutes at Large, and the other on page 2574. Pub. L. 101-549, 104 Stat. 2467. The House version appeared under the heading “Miscellaneous Guidance.” It amended Section 111(d) as follows: The Administrator can “establish standards of performance for any existing source for any air pollutant . . . not included on a list published under section 7408(a) or emitted from a source category which is regulated under section 112.” 104 Stat. 2467.
The Senate version appeared under the heading “Conforming Amendments” and replaced “112(b)(1)(a)” with “112(b).” The Administrator can “establish standards of performance for any existing source for any air pollutant . . . not included on a list published under section 7408(a) or 112(b).” 104 Stat. 2574. Although only the House version was included in the codification found in U.S. Code, Supreme Court caselaw makes clear that it is the Statutes at Large, which contains both versions, that governs. U.S. Nat’l Bank of Oregon v. Independent Ins. Agents of Am., 508 U.S. 439, 488 (1993).
The Supreme Court’s decision in Chevron v. NRDC provides a familiar two-step framework for considering this statutory language. The first step is to ask whether the intent of Congress is clear from the language. If it is, then that meaning governs. If the language is ambiguous, Chevron step two provides that courts must defer to the implementing agency’s interpretation unless it is unreasonable.
Under this framework, the first question is whether a consistent Congressional intent can be drawn from the House and Senate language. It can. The most straightforward way to interpret the combined language is simply that, as was the case before the 1990 Clean Air Act amendments, Congress intended EPA to use Section 111(d) only for pollutants that are neither hazardous air pollutants regulated by Section 112 nor criteria pollutants regulated under Sections 109 and 110. It is undisputed that this is the meaning of the Senate amendment. And, as EPA recognized in the Clean Air Mercury Rule, the House language can be read to the same effectâEPA can use Section 111(d) to regulate a pollutant except if the pollutant is emitted by a source category that is regulated under Section 112, i.e., unless that pollutant is a hazardous air pollutant.
This is how the Supreme Court interpreted the provision in American Electric Power v. Connecticut. There the Court held that plaintiffs could not bring a federal tort action against major electric power companies for their emissions of carbon dioxide, because the Clean Air Act authorized EPA to regulate power plants’ carbon dioxide pollution under Section 111(d). The Court also noted that Section 111(d) applied to pollutants not regulated under Sections 110 or 112. “EPA may not employ [Section 111(d)] if existing stationary sources of the pollutant in question are regulated under the national ambient air quality standard program, [Section 110], or the “hazardous air pollutants” program, [Section 112].” The Court could not have found federal nuisance law displaced by Section 111(d) if it had interpreted the Act to eliminate EPA authority to regulate greenhouse gas pollutants from power plants under Section 111(d) as soon as the agency finalized hazardous air pollutant standards for power plants.
The alternative interpretation proposed by opponents of EPA’s carbon pollution standards is nonsensical and untenable. It would turn the House language into a dramatic revision of Section 111(d) intended to immunize power plants’ carbon pollution from regulation under Section 111(d) simply because another power plant pollutant was regulated under Section 112. Indeed, it would virtually eliminate Section 111(d)’s reach over carbon pollution, because nearly all large CO2–emitting industrial sources are also subject to standards under Section 112 for one or more hazardous air pollutants.
There is absolutely no indication in the legislative history that the House had any intent to make such a sweeping change, and the text cannot reasonably be read to have done so. As the Supreme Court said in Whitman v. American Trucking Association, Congress “does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisionsâit does not, one might say, hide elephants in mouseholes.” 531 U.S. 452, 468 (2001).
In short, applying basic principles of statutory interpretation, the duplicative amendments can and should be harmonized by interpreting the House language to have the same meaning as the undisputed meaning of the Senate language. So doing knocks the legs out from under the carbon pollution opponents’ specious argument.
Even if the duplicative provisions actually conflicted, there is still no reason the radical interpretation of the House provision would prevail. Under another longstanding judicial doctrine, the Senate language is controlling because it comes after the House language in the Statutes at Large. As the D.C. Circuit held in Lodge 1858, American Fed. of Gov. Employees v. Webb, “[t]he established rule is that if there exists a conflict in the provisions of the same act, the last provision in point of arrangement must control.” 580 F.2d 496, 511 (D.C. Cir. 1978).
Finally, addressing the legal question under Chevron step two produces the same outcome. If a court found the meaning of the two duplicative provisions ambiguous, EPA’s interpretation would govern unless a court found it unreasonable and impermissible. In the 2005 Clean Air Mercury Rule, EPA has previously interpreted Section 111(d) to allow regulation of pollutants not regulated under Section 112. Given the clarity of the Senate language and the statutory context, there can be little doubt that EPA’s interpretation of the House language, and of the two provisions together, would be found reasonable and upheld.
So while there is indeed a drafting glitch in the Act, it in no way blocks EPA from setting carbon pollution standards for existing power plants.