After a decade of study, the Environmental Protection Agency in 1989 promulgated the Asbestos Ban and Phase-Out Rule under section 6 of the Toxic Substances Control Act (TSCA), 15 U.S.C. § 2605 (pre-2016). Finding asbestos a carcinogen with no known safe level of exposure, EPA concluded that an across-the-board, phased ban on the manufacture, importation, processing, and distribution of most asbestos-containing products was necessary to address an unreasonable risk to human health. The rule grouped numerous product categories (including, among others, asbestos-cement pipe, gaskets, brake linings, and various paper products) and relied on a mix of qualitative and quantitative risk assessments, projections of disease avoided over long horizons, and assessments of economic impacts and product substitutes. Multiple industry petitioners—including Corrosion Proof Fittings (a manufacturer of asbestos-cement pipe and related products), brake and gasket makers, and others—petitioned for review in the Fifth Circuit pursuant to TSCA § 19, arguing that EPA failed to (1) establish that asbestos, as used in particular products, presented an unreasonable risk in light of existing controls; (2) demonstrate that a near-total ban was the "least burdensome" regulatory alternative adequate to protect against any unreasonable risk; (3) sufficiently consider and quantify costs and benefits, including discounting benefits over time; (4) assess the risks of substitute materials and the availability of safer alternatives on a product-by-product basis; and (5) coordinate with or defer to other agencies (e.g., OSHA) as contemplated by TSCA. EPA defended the rule as a rational response to a proven carcinogen presenting grave risks and contended that lesser measures (such as labeling or partial restrictions) would be inadequate. The administrative record was voluminous. The Fifth Circuit had jurisdiction under TSCA § 19(a), and the statute prescribed a substantial-evidence standard of review for TSCA § 6 rules.
Did EPA comply with TSCA § 6 when it imposed a near-total ban on asbestos by adequately finding an unreasonable risk and selecting, with substantial evidence, the least burdensome regulation necessary to protect against that risk?
Under pre-2016 TSCA § 6(a), EPA may regulate a chemical substance upon finding it presents an unreasonable risk of injury to health or the environment. In promulgating a § 6 rule, EPA must impose the least burdensome requirement that will adequately protect against the unreasonable risk and must consider the factors enumerated in § 6(c)(1): (A) the health and environmental effects of the substance; (B) the magnitude of human and environmental exposure; (C) the benefits of the substance for various uses; (D) the reasonably ascertainable economic consequences of the rule (including effects on the economy, small business, and innovation); and (E) the availability and risks of substitutes. TSCA § 19(c)(1)(B) prescribes a substantial-evidence standard of review for § 6 rules, requiring the court to determine whether EPA's findings are supported by substantial evidence on the record as a whole. Read together, TSCA's text requires EPA to: (1) balance costs and benefits in determining whether a risk is "unreasonable"; (2) consider less burdensome regulatory alternatives (e.g., labeling, partial product-specific restrictions, or targeted controls) before selecting a ban; (3) assess the risks of substitutes and existing controls by other agencies; and (4) quantify and compare costs and benefits to the extent feasible, including appropriate treatment of time (e.g., discounting) and product-by-product analysis.
The Fifth Circuit vacated and remanded EPA's Asbestos Ban and Phase-Out Rule, concluding that EPA failed to demonstrate, with substantial evidence, that a near-total ban was the least burdensome regulation adequate to address any unreasonable risk posed by asbestos across the numerous product categories. Some unchallenged portions of the rule—such as prohibitions on new uses of asbestos and on certain paper products and flooring felt—remained in effect; the bans on most existing asbestos-containing products were set aside.
The court began with TSCA's text. It read "unreasonable risk" in § 6 as requiring some balancing of costs and benefits, reinforced by § 6(c)(1)'s explicit factors, including economic consequences and the availability of substitutes. It also emphasized the statute's directive that EPA adopt the "least burdensome" regulation that will adequately protect against the risk, which the court treated not as a general aspiration but as a binding constraint requiring consideration of a suite of regulatory alternatives and reasoned selection among them. Because § 19(c) subjects § 6 rules to substantial-evidence review, the court demanded a robust, record-based showing that the statutory factors and least-burdensome requirement were satisfied. Applying these principles, the court found multiple, interrelated defects. First, EPA did not adequately consider and compare less burdensome alternatives, such as labeling, targeted controls, or partial product-specific restrictions. Instead, the agency largely defaulted to a comprehensive ban across a wide range of product categories without making product-by-product findings that a ban, rather than more tailored measures, was necessary and least burdensome for each category. TSCA, the court reasoned, required a more granular analysis, especially because risk, exposure, and substitutes differed substantially across uses (e.g., gaskets versus brake linings). Second, EPA's cost-benefit analysis and quantification of benefits were deficient. The agency did not consistently quantify the number of lives saved attributable to its rule at the product level or justify its treatment of time and discounting. The court criticized EPA for failing to align treatment of costs and benefits over long time horizons and for insufficient sensitivity analysis. The court also faulted EPA for not adequately evaluating the risks of substitutes—some of which could introduce their own hazards—thereby undermining a reasoned comparison of net benefits. Third, the court found that EPA did not meaningfully account for existing regulatory regimes (for example, OSHA workplace standards) and their efficacy in reducing exposure. TSCA contemplates coordination with and consideration of other agencies' controls; EPA needed to explain why those measures could not sufficiently mitigate risk or why referral to or harmonization with those regimes was not appropriate. Jumping directly to a ban without engaging this inquiry ran afoul of TSCA's structure. Finally, because TSCA demanded selection of the least burdensome alternative supported by substantial evidence, the court concluded that the agency's record did not clear that statutory bar. The decision therefore vacated and remanded the rule. The court made clear that it was not forbidding asbestos regulation per se; rather, it required EPA to rebuild the record to justify its choice with rigorous, product-specific analysis consistent with TSCA's commands. As a practical matter, some unchallenged or distinct portions of the 1989 rule (including the ban on new uses of asbestos and on certain paper products and flooring felt) remained intact, while the broader product bans were set aside.
Corrosion Proof Fittings is a staple of administrative and environmental law courses because it demonstrates how statutory text can both authorize and limit agency power. It shows that when Congress directs an agency to adopt the "least burdensome" regulation and to consider enumerated factors, courts will require a robust, product-by-product, evidence-backed analysis that meaningfully balances costs and benefits and grapples with alternatives and substitute risks. The case also underscores the impact of statute-specific standards of review: TSCA's substantial-evidence bar demanded more than generalized assertions about danger; it demanded quantified, comparative reasoning. Doctrinally and practically, the decision raised the evidentiary and analytic hurdles for regulating under old TSCA § 6 and contributed to decades of limited action against dangerous chemicals. Its perceived rigidity was one of the central catalysts for Congress's 2016 TSCA amendments (the Lautenberg Act), which eliminated the "least burdensome" language and restructured how EPA assesses risk and considers costs—thereby partially superseding the decision's most restrictive implications. Still, Corrosion Proof Fittings remains a key authority on how courts read substantive constraints into environmental statutes and how agencies must justify sweeping rules.
Corrosion Proof Fittings v. EPA stands as a powerful reminder that statutory text can sharply cabin agency discretion. By insisting that EPA demonstrate, with substantial evidence, that a sweeping asbestos ban was the least burdensome way to address any unreasonable risk, the Fifth Circuit required rigorous product-by-product analysis, meaningful cost-benefit balancing, engagement with substitutes, and attention to other regulatory regimes. The result was a vacatur that left only select portions of the asbestos rule in place and set a demanding precedent for future TSCA actions under the pre-2016 framework.