A government "emissions-trading" program aimed at controlling sulfur dioxide emissions has been a clear success from both environmental and economic standpoints, according to a report by MIT researchers.
The report is the first comprehensive empirical analysis of a program created to control sulfur dioxide (SO2) emissions under the US Acid Rain Program -- Title IV of the 1990 Clean Air Act Amendments.
The research team members, all affiliated with the Center for Energy and Environmental Policy Research (CEEPR), included Professor Richard Schmalensee, the Gordon Y Billard Professor of Management and Economics at the Sloan School and CEEPR director; Professor Paul Joskow, the James and Elizabeth Killian Professor and head of the Department of Economics; Dr. Denny Ellerman, a senior lecturer at the Sloan School and CEEPR executive director; former graduate student Juan Pablo Montero of civil and envionmental engineering (now at the Catholic University of Chile); and graduate student Elizabeth M. Bailey of economics.
SO2 emissions have been reduced by more than required to meet 1995 and 1996 emissions ceilings at a cost below expectations and significantly lower than alternative control strategies, the researchers found.
Despite this success, the authors caution that the results of this experiment must be applied with care to other environmental problems. If they are designed and implemented well, emissions-trading systems can reduce compliance costs compared to traditional command-and-control approaches. However, compliance costs may still be significant, and analysts must continue to compare these costs to the environmental benefits to determine the appropriate level of emissions reduction.
The US Acid Rain Program requires that electric utility power plants reduce SO2 emissions to 50 percent of the 1980 emission level, and that they implement this reduction by using fully tradable emission permits, called allowances. Although only a slight reduction of emissions would have been required to meet the cap imposed in theprogram's first year, electric utilities reduced emissions by 40 percent more than was required.
The 1995 emissions reduction was achieved about equally by retrofitted "scrubbers," which remove SO2 from power-plant flue gas, and by switching to lower-sulfur coal. Despite fears of "hot spots" as a result of emissions trading, emissions in all states were lower than they would have been had there been no trading.
The allowances not used in 1995 are being "banked" for use after the year 2000, when the aggregate cap is ratcheted down further and all electric-utility generating units are subject to Title IV restrictions.
An efficient, competitive allowance market has developed due to the sound initial design of the permit system, EPA's "exceptionally nonbureau-cratic" implementation of that design, and the keen interest of electric utilities and brokers in pursuing opportunities to reduce costs presented by allowance trading, according to the MIT report. The researchers noted that the mandatory EPA auctions provided an early and accurate indication of the value of allowances in the years when utilities were making decisions about how to comply.
Although important in the earliest years, the allowance auction is now dwarfed by the volume of allowance trading in the large private market, which provides good information about current and future allowance prices to all market participants. Fears that the EPA auctions would distort the development of the private market have not been realized.
The MIT report also provides an after-the-fact estimate of the cost of compliance with Title IV based on a survey of electric utilities and other sources. The total cost of the 3.9-million-ton reduction of SO2 emissions achieved in 1995 was $726 million. On the basis of the $186-per-ton average, the actual cost matched the most optimistic earlier predictions for the cost of this trading program, which ranged from a high of $300 per ton (when little trading was assumed), to slightly less than $200 a ton.
The report notes that scrubber costs were much lower than predicted, and that low-sulfur western coal delivered to the Midwest was less expensive than expected due to lower rail rates resulting from rail deregulation. Although cautioning that any estimate of the cost savings due to emissions trading depends upon what "no-trading" alternative is assumed, the authors provide a rough estimate that emissions trading reduced compliance costs in 1995 by about one-third, compared to reasonable alternative approaches producing equivalent emissions control.
Allowance prices differ from average compliance costs, the report notes. Both were lower than expected, but allowance prices were far below earlier predictions, whereas compliance costs were at the lower end but within the range of earlier predictions. Today's allowance prices are below the long-run marginal cost of compliance because of early overinvestment in emissions control, the researchers explain. However, they note that the ability to bank allowances for future use mitigates the implied extra cost by deferring the need for additional controls for a few years.
Large-scale tradable-permit programs can work more or less as textbooks describe, the researchers conclude. The flexibility inherent in tradable-permit programs facilitates cost-reducing adaptation to unexpected developments; and efficient, competitive markets for tradable permits can develop, given sound program design and effective implementation, they add.
Finally, while the lessons from this public policy experiment are many, they must be applied with care to other environmental problems, the report states. Emissions trading under the US Acid Rain Program has been a clear success and it has produced surprises; however, it has not produced any miracles. In particular, this experience doesnot relieve policy-makers from the responsibility for weighing the benefits of environmental goals against the costs, even when the compliance costs are lower because of emissions trading, according to the authors.
The research was funded by grants from the National Acid Precipitation Assessment Program, the EPA and the corporate sponsors of the CEEPR.
A version of this article appeared in MIT Tech Talk on December 17, 1997.
