Evaluating Emissions Trading Using a Nearest (Polluting) Neighbor Estimator

This paper uses a nearest neighbor matching estimator to examine the effects of an emissions trading program. An important perceived advantage of “cap-and-trade” programs over more traditional, more prescriptive forms of regulation is that enhanced compliance flexibility and cost effectiveness can make more stringent emissions reductions politically feasible. A potential disadvantage is that a reliance on markets, versus prescriptive regulations, to coordinate emissions abatement can result in problematic price volatility, non-compliance, and environmental injustice. All of these issues have been raised with the RECLAIM program, which caps NOx and SO2 emissions in Southern California. We match facilities in the RECLAIM program with facilities in other California non-attainment counties that have similar operating characteristics and industry affiliation. Using a non-parametric differences-in-differences estimator, we investigate the relative effect of the RECLAIM program on facility emissions. Our results indicate that (1) emissions at facilities complying with RECLAIM fell approximately 40 percent, on average, relative to the counterfactual we construct; (2) emissions fell significantly more at large facilities, as is consistent with scale economies in abatement technology; and (3) neighborhood demographic characteristics were insignificant determinants of the observed changes in emissions. JEL Classification: H23, Q25, D63