The use of a feebate system to reduce emissions from power plants

The patchwork of laws designed to control air pollution from U.S. power plants has been criticized from a variety of different perspectives. Business groups argue that the laws are too complex and burdensome to industry. Environmentalists maintain that power plant emissions need to be further reduced because of their negative health effects, to combat global warming and to eliminate haze in our national parks. Economists claim that large emission reductions could be achieved rather cheaply by focusing control efforts on the decades-old power plants in the Midwest, but Midwesterners and their political representatives are understandably resistant to having to shoulder the costs. This paper describes an economic mechanism that has been used in Sweden since 1990 to control NOx emissions and indicates how, if it were modified and expanded to include other pollutants, could be made to work much like multi-pollutant cap-and-trade — but with more flexibility and efficiency. In this feebate system, each power plant would either pay a fee or collect a rebate for each ton of emissions above, or below, their assigned "breakeven point". With the per-ton fee/rebate rate set high enough to bring the needed emission reductions and breakeven points adjusted each year so that total rebates equaled total fees, the resulting system would work like a frictionless cap-and-trade: companies having high control costs would pay fees into the system, in effect paying companies with low control costs to do some of their controlling for them. At the same time, the system would bring control costs at Midwestern plants into line with those of other plants throughout the country and would also (through the assignment of breakeven points) provide a wide range of choices of how the additional control costs could be distributed. (The views expressed in this paper are those of the author and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency.