Indirect land use change (ILUC)—the theory that the use of cropland for biofuels raises food prices and thus increases the incentive to convert forests and grasslands to crop production, thereby releasing stored carbon and decreasing future carbon sequestration—is a particular concern with ethanol produced from corn kernels since corn is such a major food crop (and livestock feed) and international food commodity markets are relatively inelastic. This Article examines how the Environmental Protection Agency (EPA) dealt with the uncertainty of regulating biofuels policy in the context of corn ethanol, but the lessons have broader applicability to biofuels policy and more generally to decision making under conditions of great uncertainty. The first Part discusses how biofuels may contribute to the mitigation of climate change, analyzes the U.S. statutory scheme for renewable fuels, explores the issues surrounding the statute’s mandate for consideration of ILUC, and explains how EPA resolved the ILUC issue in the case of corn ethanol. The second Part probes EPA’s treatment of uncertainty regarding ILUC in its decision to approve corn ethanol as a renewable fuel—primarily considering whether the broad range of emission figures used made the decision riskier, whether the central estimate used was appropriate, and whether the irreversibility of the decision should have been considered. The third Part considers the policy issues surrounding ILUC more generally—addressing whether ILUC should be considered in biofuels policy or ignored because of concerns about the reliability of models of ILUC and whether there are policy responses that might ameliorate ILUC and narrow the range of uncertainty about its severity. The Article concludes that EPA acted responsibly and in good faith, however, the determination regarding ILUC for corn ethanol involved significant errors of judgment in the treatment of uncertainty. Specifically, EPA should be more sophisticated in its treatment of uncertainty by using the means of probability distributions rather than medians, taking into account lock-in effects and real option values, and explicitly considering attention model uncertainty. EPA, however, was correct to consider ILUC despite the admitted degree of uncertainty regarding ILUC’s magnitude.
The full text of this Article is available to download as a PDF.