Editors' Summary: CERCLA § 107(e) recognizes that private parties may agree to limit or to insure against their CERCLA liability by using common law contract tools, including indemnification and hold harmless agreements. But what role did Congress intend such private party agreements to play in allocating CERCLA liability? The plainly contradictory language of CERCLA § 107(e) does not give a final answer to this question, because the first sentence of CERCLA § 107(e) appears to invalidate indemnities and the like, and the second sentence appears to save them. Recently, two federal district courts in AM International v. International Forging Equipment Corp. and CPC International v. Aerojet-General Corp. issued the first decisions to analyze CERCLA § 107(e)'s legislative history. In a dramatic departure from precedent first articulated by the Ninth Circuit in Mardan Corp. v. C.G.C. Music, Ltd., the courts in AM International and CPC concluded that § 107(e) should be read only to give effect to contracts binding parties otherwise not liable, and not to contracts between potentially liable parties. Before AM International and CPC, every court that had addressed CERCLA § 107(e) interpreted it as preventing a party liable under CERCLA from completely evading its liability, but allowing it contractually to allocate the ultimate financial burden of that liability among itself and others.

This Article argues that the Mardan interpretation of CERCLA § 107(e) is correct. The Article first compares the Mardan interpretation with the contrary reasoning of AM International and CPC. It then explicates its preference for the Mardan view in light of CERCLA § 107(e)'s plain meaning and legislative history, as well as the language and legislative history of a comparable provision of the Oil Pollution Act of 1990. Finally, the Article explains why the Mardan view better advances the policies underlying CERCLA, particularly the goal of promoting private, voluntary cleanup, and why the AM International and CPC view has the opposite effect. The author concludes that the Mardan rule does not shift cleanup costs from private parties to the government, but only between private parties, and that equitable allocation of response costs does not require the wholesale invalidation of private contractual arrangements.