CSEM WP 136 Is Real-Time Pricing Green?: The Environmental Impacts of Electricity Demand Variance

Economists have long advocated for electricity pricing that accurately reflects time-varying production costs. In particular, real-time pricing (RTP) would improve the efficiency of electricity consumption and investment and would lessen the potential harm from market power. Conventional wisdom has claimed that RTP has an additional benefit, namely, reduced emissions from reduced peak demand. We argue that RTP will reduce variance of electricity load and estimate the short-run impacts of a reduction in variance on emissions of SO2, NOx, and CO2. According to our estimates, a reduction in within-day load variance would decrease emissions of some pollutants in some NERC regions (FRCC in Florida, MAAC in the MidAtlantic, and MAIN in the Illinois area). However, a reduction in within-day variance would actually increase emissions of all three pollutants in the Eastern Mid-West (ECAR) and the Southeast (SERC) and of two of three pollutants in Texas (ERCOT), the Great Plains (MAPP), and the West (WSCC). The effects are relatively small as the elasticity greatest in magnitude is 0.042. We further analyze reductions in across-day variance and find similar results. The results are also robust to alternate measures of within-day variance and to a nonparametric specification of the model. To understand our results, we note that changes in emissions are similar to changes in generation from power plants using fossil fuels. Furthermore, we observe that emissions reductions occur in regions with more hydroelectric capacity than oil-fired capacity. This supports the hypothesis that the environmental benefits of RTP come from reducing peak demand, but only if peak capacity is oil fired rather than hydroelectric. ∗Holland: Department of Economics, University of North Carolina, Greensboro, NC 27402-6165. Email: [email protected]. Mansur: School of Management and School of Forestry and Environmental Studies, Yale University, P.O. Box 208200, New Haven, CT 06520-8200. Email: [email protected]. We would like to thank Severin Borenstein, Jim Bushnell, Kevin Forbes, Jun Ishii, Nat Keohane, Al Klevorick, Frank Wolak, and seminar participants at the University of California Energy Institute and Yale University for comments. Thanks also to Meredith Fowlie and Nalin Sahni for excellent research assistance.