In the past 15 years, cap-and-trade programs have become the preferred means of regulating air pollutants. A cap-and-trade program establishes the annual number of allowable emission permits (the "cap"), which is set below the existing emissions level. Each regulated entity must cash in one permit for each unit of air pollution it emits. The cost savings come from allowing firms to trade permits, so that a firm that finds it costly to reduce its marginal unit of pollution can instead purchase a permit from another firm that can reduce a unit of pollution for less cost. Because the overall cap is binding, the result is a reduction of pollution to the target level at costs much lower than the more rigid command-and-control regulations.

The cap-and-trade program often cited as a success story is the sulfur dioxide permit market created under Title IV of the Clean Air Act Amendments of 1990. The market has cut acid rain sulfur dioxide emissions from the electric power industry by 40 percent over 1980 levels, at a savings of about $1 billion per year compared to the conventional command-and-control approach. In recent years, the Environmental Protection Agency has initiated a regional cap-and-trade market for nitrogen oxides and is currently considering a proposal to further reduce sulfur dioxide and nitrogen oxides using cap-and-trade approaches. The EPA is also considering reducing mercury emissions from power plants using a national cap-and-trade program.

Allocation

Though much has been written and discussed about the advantages and disadvantages of cap-and-trade programs, one issue of great importance is frequently neglected and can instead only be heard occasionally in conversations among academics. The issue is how to allocate the permits to the regulated entities once the cap is established.

That negligence is understandable because of the heavy attention that is given to the choice of the level at which the cap is established. The cap level imposes costs on the regulated entities (costs borne by shareholders, workers, and consumers), and it yields environmental benefits. The goal is to establish a cap level that maximizes the difference between benefits and costs. Given the difficulty in determining all the components of benefits and costs, and given the vested interests of stakeholders on all sides, it is no wonder that most of the policy debate focuses on the appropriate level for the cap.

Nonetheless, the question of how to distribute the permits is itself important. There are two general options: The permits can be given out for free each year or they can be auctioned off. Under the giveaway approach, characteristics of the firms could be used to rank which firms get priority for the permits. For example, electric utilities' permit allocation could be based on measures of heat input or electricity output in a baseline year. Under the auction approach, the highest bidders would receive the permits.

Not surprisingly, firms prefer to receive the permits for free. When asked the reason for that preference, they often argue that auctions hurt the economy. But their argument does not withstand scrutiny. Though the cap-and-trade system does impose costs, those costs result from the establishment of a cap below existing emission levels. The decision to allocate the permits for free or to auction them has only distributional impacts, not economic impacts. Unlike a tax on output or labor, an auction does not change any of the economic decisions of a firm. If a firm is given the permits for free, each unit of pollution it emits means that it must forgo selling a permit at the market price. Similarly, if a firm is in the auction setting, each unit of pollution it emits means that it must pay the market price for a permit. In other words, whether you are charged $5 or you must forgo earning $5 for every unit of emissions, your economic incentives are the same. The only difference between an auction and a giveaway is who gets the pot of cash.

It might be hard to believe that the government can raise revenues (through an auction) without imposing economic costs. The important point is that it is the permit cap itself that imposes costs by restricting the economic activities of the regulated firms. It is therefore important that the government strive to set the permit caps at the level that maximizes the difference between the environmental benefits and the costs that the permits impose on society. However, once the permit cap is established, there are no additional economic costs associated with auctioning the permits rather than giving them away for free. In economic parlance, the auction is a non-distortionary tax; i.e., a tax that does not inhibit economic activity and thus does not impose additional costs. Of course, it does no good if the government squanders the auction revenue on unproductive activities. But if the auction revenue is instead earmarked to reduce other distorting taxes, then auctioning permits is preferable to giving away permits.

Auctioning has other advantages over giving away permits for free. In a giveaway system, the government must establish criteria for how many permits each firm gets. As mentioned earlier, this typically means basing permit allocations on firm characteristics in a pre-established baseline year. However, the government may want to update the baseline year over time. For example, for the EPA's current proposal to reduce the cap for sulfur dioxide, the public comment period heard from many regulated firms that wanted the agency to update the existing baseline period (which is based on firms' characteristics from 1985 to 1987). The problem is that updating the baseline year could lead to perverse incentives for firms because they will try to change their characteristics to get more permits. That is, if firms think that their future allocation of permits will depend on their future electricity output, then they have an incentive to overproduce electricity. This distortion hurts the economy. In order to avoid those strange incentives, permit systems in the past have based allocations on one-time characteristics, which thus establish the allocations in perpetuity. In fact, the current EPA proposal for sulfur dioxide indicates that the baseline year will remain unchanged. This makes economic sense, yet it does introduce a peculiar equity issue in that firms get a "permanent right" to pollute, whereby they still receive valuable permits even after they eventually shut down their operations.

Another advantage of auctions is that they provide greater flexibility in the design of cap-and-trade programs. For example, auctions facilitate the use of a "safety valve" mechanism. A safety valve is a promise by the government to issue and sell additional permits (above the cap) upon request to any firm at a given fixed price. This offers a hedge against any unexpectedly sharp increases in the price of pollution reduction. For example, the government might establish a cap with the expectation that the associated marginal cost of reducing pollution will be $5,000 per ton of pollution. However, in order to avoid a sharp increase in this price, the government can establish a safety valve price of $10,000 per ton. That would guarantee that the market permit price would not exceed the safety valve price. If the safety valve is triggered one year, then the extra allowances can be deducted from future years' allocations. A safety valve therefore makes policy sense because it hedges against the possibility of a short-term spike in the cost of reducing pollution, yet it maintains the environmental gains over the long term. With an auction system, reducing future years' allowances is simple. With a giveaway system in which allocations are pre-established based on historical firm characteristics, it is difficult to determine how to reduce future years' allowances. This could mean abandoning the safety valve idea, unnecessarily increasing the costs of pollution control without creating environmental gains.

Experience

Though auctioning makes good economic sense, it has been used only sparingly. On the legislative front, the sulfur dioxide market established by the Clean Air Act of 1990 contains annual auctions of 2.8 percent of the permits each year. In 2004, it raised approximately $50 million. President Bush's Clear Skies program, which was introduced in 2002 to reduce sulfur dioxide, nitrogen oxides, and mercury, contains a more ambitious auction system that would be phased in over time. From 2008 to 2030, the auctions would bring approximately $16.3 billion into federal coffers. If all the permits were auctioned, the program would bring in about $128.7 billion without harming the economy. But Sen. James M. Inhofe (R-Okla.) recently introduced an amended version of the Clear Skies bill that would eliminate the auctions altogether, so the prospects of the auction remain in doubt.

On the administrative front, the EPA's new nitrogen oxides regional cap-and-trade market allocates permits to each state and allows the states discretion on how to distribute the permits to their within-state firms. Just this summer, Virginia decided to auction five percent of its permits, which raised $10.5 million. Had Virginia auctioned all of its permits, it would have raised about $210 million, and that is just for the first year. However, the Virginia legislature subsequently voted to ban all future auctions of permits.

One can legitimately debate what the appropriate pollution reduction level is. This level should be chosen to maximize net benefits and should not be chosen in order to maximize auction revenues. Indeed, we should be wary of the government designing future cap-and-trade markets strictly to raise auction revenue. But in instances where the cap level is established independent of the auction decision, it makes economic sense to auction the permits and use the resulting revenue to offset distortionary taxes. If and when President Bush's Clear Skies bill is again considered by Congress, the administration should push for the restoration of the auctions that were contained in the original proposal, especially given the president's stated goal of keeping distorting taxes low while moving toward a balanced budget.

Similarly, over the next few months, the EPA will work toward finalizing the cap-and-trade rules for nitrogen oxide emissions and for mercury emissions. Both of those proposed rules assign permit budgets to the covered states and then allow the states discretion on how to allocate the permits within their boundaries. In the final rules, it is important that the EPA maintain this federalist approach of state flexibility, and that the states consider using auctions as a means to raise targeted revenues while protecting their economies.

Ted Gayer is an associate professor of public policy at Georgetown University and a visiting scholar at the AEI. He recently served as a senior economist on the President's Council of Economic Advisers.