Are All Market-Based Environmental Regulations Equal?

Economists have long advocated using the market to achieve environmental objectives. Possible policies include taxes on waste emissions and programs through which the government limits pollution by issuing emission permits that can be traded among companies. Unlike “command and control” regulations that specify which technology companies must use, market approaches such as these allow companies flexibility to choose their own ways to reduce pollution at lowest cost.

But which market-based instruments are best? Research suggests that the economic costs of tradable permits can be much larger than those of emissions taxes if the permits are given out free or grandfathered to firms. Emissions taxes provide revenues that can be recycled in tax reductions that increase employment. In addition, tradable permits can result in disproportionately high costs for the poor, whereas the revenues from emissions taxes can be converted into tax cuts that help the poor.

Policymakers are currently debating proposals to implement nationwide tradable permit programs for nitrogen oxide and mercury and to strengthen the existing allowance trading program for sulfur dioxide. The tradable permits approach pioneered by the United States is also receiving a great deal of attention throughout the world as a possible tool in managing greenhouse gas emissions. Before proceeding further down this path, policymakers should acquire a better understanding of the full implications of each of their policy options.

Measuring economic effects

To understand how environmental policies interact with the broader fiscal system, start with the so-called “double dividend” argument. Revenues generated from environmental taxes (or auctioned emissions permits) can be used to pay for cuts in labor taxes, such as income and payroll taxes. At first glance, these measures appear to increase employment while simultaneously reducing pollution, thereby producing a double dividend. The argument is particularly appealing for climate change policies, where the revenue potential is so large.

Suppose, for example, that the United States were to recommit to its Kyoto agreement pledge to reduce annual carbon emissions to 7 percent below 1990 levels by 2010. This might be achieved by a tax on the carbon content of fossil fuels of anywhere between $50 and $150 per ton of carbon, depending on whether the United States could buy credits for carbon reductions overseas. Setting a hypothetical carbon tax at $75, revenues would be around $90 billion per year–about one-sixth of the entire federal receipts from personal income taxes.

But the effect of a tax would not end there. Environmental taxes raise the cost of producing products, acting as a tax on economic activity. This means that (before tax revenue recycling) the levels of production and employment are lower than they would be in the absence of the emissions tax. The increase in government revenue would make it possible to reduce income taxes for workers, which would boost employment somewhat. Although this would soften the effect of the carbon tax, most studies find that the net result would be a decrease in jobs. This doesn’t make the carbon tax bad policy, but it does undermine the argument of those who claim that a carbon tax with revenues recycled into income tax reductions would increase employment.

The effects of tradable permits are similar to those of emissions taxes. They raise production costs and reduce economic activity. If a polluting company increases production, it must either buy permits to cover the extra emissions or forego sales of its own permits to other companies. Either way, it pays a financial penalty for producing emissions. Permits that are distributed to firms for free have adverse effects on employment in the same way that emissions taxes do, but without the potential offsetting benefits of recycling government revenue in other tax reductions. This has two important policy implications.

First, if permits were auctioned off by the government rather than given away for free, then society would be better off, as long as revenue from permit sales would be recycled into other tax reductions. Tax economists have estimated that for each dollar of revenue used to reduce income taxes, there will be a gain in economic efficiency of around 20 to 50 cents. Lower income taxes increase employment, and they also reduce distortions in the pattern of expenditure between ordinary spending and “tax-favored” spending such as owner-occupied housing and employer-provided medical insurance. In the carbon example above, the United States might be better off to the tune of $20 billion to $45 billion per year if it had a carbon tax or auctioned permits policy rather than grandfathered permits.

Second, the economic cost of grandfathered permits can be substantially higher than previously thought. According to an estimate by Roberton Williams, Lawrence Goulder, and myself, the cost to the United States of meeting the initial Kyoto target by a system of grandfathered permits imposed on fossil fuel producers rises from roughly $25 billion per year (in current dollars) to around $55 billion when their effect on reducing employment and compounding labor tax distortions is included.

In fact, taking account of fiscal interactions might compromise the ability of grandfathered permits to generate overall net benefits for society. Suppose that global environmental damage from carbon emissions (for example, economic damage to world agriculture from climate change and the costs of protecting valuable land against sea level rises) was $70 per ton. (This is actually on the high side compared with most damage studies, although the estimates are subject to much dispute.) The initial Kyoto target for the United States would have reduced carbon emissions by around 630 million tons in 2010, implying annual environmental benefits of around $45 billion in current dollars. Using our cost figures and ignoring fiscal interactions, the grandfathered permit scheme would produce an estimated $20 billion in environmental benefits. But include fiscal interactions, and the policy fails the cost/benefit test, because costs exceed environmental benefits by $10 billion. Only the emissions tax/auctioned permit policies pass the cost/benefit test, producing net benefits of between $10 billion and $35 billion.

Fiscal interactions do not always have such striking implications for the economic performance of environmental policies. Consider the sulfur dioxide program of grandfathered permits, which has reduced power plant emissions by about 50 percent, or 10 million tons. Annual benefits from the program, mainly from reduced mortality, have been measured at more than $10 billion per year. Estimates of the annual cost of the program, including fiscal interactions, are only $1.7 billion under the existing grandfathered scheme or $1.2 billion if the permits are auctioned rather than grandfathered. Regardless of whether permits are auctioned or not, estimated benefits swamp the costs of the sulfur dioxide program.

Who benefits?

Of course, there is no guarantee that the revenue from environmental taxes or permit auctions will be used wisely, but evidence from Europe indicates that it can be. Denmark recently introduced a package of taxes on sulfur dioxide, carbon dioxide, fossil fuels, and electricity. Revenues, which amount to about 3 percent of gross domestic product, have mainly been used to lower personal and payroll taxes.

One potential obstacle to green tax shifts, however, is the conflicting objectives of different government agencies. An environmental agency would typically be concerned about setting rates to meet a particular environmental goal without regard to revenue. The treasury department might be more concerned about revenue, regardless of whether the taxes lead to under- or overshooting environmental objectives.

But suppose tax cuts don’t happen? Suppose that in the United States revenue from environmental taxes or auctioned permits were used to reduce the federal budget deficit instead? This implies that taxes could be lower in the future and still cover debt interest and repayment of principal. There would still be an economic gain from lower taxes, although one that is deferred to the future.

What if the revenue were used to finance additional public spending? The bulk of federal spending consists of transfer payments, such as Social Security, or expenditures that effectively substitute for private spending, such as medical care and education. Loosely speaking, the private benefit to people from a billion dollars of this type of spending is a billion dollars. But the benefits to society might be greater if the spending is achieving some distributional objective such as a safety net for the poor. If instead the revenue financed cuts in distortionary taxes, households would receive the billion dollars back, and on top of this there is a gain in economic efficiency as the distortionary effect of taxes on employment and so on are reduced. Thus, the social benefits from extra transfer spending could be larger or smaller than the benefits from cutting taxes, depending on the particular spending program.

Governments also provide public goods (such as defense, crime prevention, and transport infrastructure) that private companies usually do not. People may value a billion dollars of this spending at more than a billion dollars. If they do, the benefits from this form of revenue recycling may also be as large as (or larger than) benefits from reducing taxes.

Policymakers might also be concerned about the effects of environmental policies on different income groups. Unfortunately, environmental and distributional objectives often appear to be in conflict. A number of studies suggest that the burden on households from environmental regulation imposed on power plants, refineries, and vehicle manufacturers is moderately regressive. The increase in prices as producers pass on the costs of regulations tends to hurt lower-income groups disproportionately, because they spend more of their income on polluting products than better-off households do.

But distributional concerns should not be an excuse for avoiding action on serious environmental problems. Pollution control measures should be evaluated mainly by weighing their environmental benefits against their economic costs for society as a whole. Distributional objectives are much better addressed by altering the income tax system or providing a safety net through the benefit system.

It still makes sense, however, to avoid environmental policies that increase income inequality. That is a major drawback of grandfathered emissions permits. When the government gives away rights to pollute for free, companies acquire an asset with market value. This enhances their net worth. The increase in company equity values in turn leads to more profits for shareholders, either directly through higher dividends and capital gains or indirectly though their holdings in retirement accounts. Stock ownership is highly skewed toward the rich; the top income quintile owns about 60 percent of stocks, whereas the bottom income quintile owns less than 2 percent. Using annually allocated grandfathered permits to meet the original U.S. carbon pledge under Kyoto could transfer more than $50 billion each year in pretax income or larger retirement assets to the top income quintile. Thus, higher-income groups can benefit greatly from grandfathered permits, with their windfall gains easily outweighing their income losses from higher product prices. Poor households, by contrast, are worse off. According to a study by the Congressional Budget Office, grandfathered permits to reduce U.S. carbon emissions by 15 percent would cut the annual real spending power of the lowest-income quintile by around $500 per household, while increasing that for the top income quintile by around $1,500 per household.

Auctioned emissions permits and emissions taxes do not create windfall gains to shareholders. Instead, the government obtains revenue that can be returned to households in a distributionally neutral manner, such as proportional reductions in all marginal income tax rates, or in ways that disproportionately benefit the poor, such as increasing personal allowances.

Making choices

One reason why emissions permits (whether auctioned or not) might be preferable to an emissions tax is because the emissions reduction that a given tax rate will induce may not be known in advance. Suppose, for example, that an environmental agency wishes to prevent a lake from being polluted beyond a certain threshold that will harm aquatic life or make the lake unsuitable for recreation. If companies must obtain permits for each unit of emissions they put into the lake, the agency can limit pollution below the threshold with certainty simply by limiting the number of permits. Under an emissions tax, where the amount of pollution abatement that will be achieved is initially uncertain, the agency will be in the uncomfortable position of having to adjust and readjust the emissions tax to ensure that the pollution target is attained.

On the other hand, an emissions tax puts a ceiling on program costs. If abatement is very costly, companies will avoid it and instead pay a larger tax bill. Under a permit scheme, companies are forced to reduce pollution by the limit on permits, no matter how costly the required abatement turns out to be. These considerations have led some economists to recommend a hybrid policy, or “safety valve.” Under this scheme, a limited number of permits would be issued with the aim of hitting a desired environmental goal, but the government would sell additional permits if the permit price reached an unacceptably high level. That way, the stringency of the environmental goal is relaxed somewhat if abatement costs turn out to be particularly large.

Although environmental policies that raise government revenue are appealing (so long as the revenues are not used for pork-barrel spending projects), the political reality is that interest groups in the United States do not eagerly hand money over to the government. It is no accident that grandfathered permits have been more common than taxes or auctions. We can expect a battle over how much if anything should be extracted from the regulated industries. Political compromise will often lead to lower tax rates or a tax that applies to only a limited number of activities. That’s politics.

But what if the choice is between grandfathered emissions permits or nothing? In the case of the sulfur dioxide trading program, grandfathered permits appear to be a good idea in spite of the drawbacks emphasized here. But even in the context of climate policies, I believe that a system of grandfathered carbon permits, appropriately scaled, would be preferable to doing nothing. For one thing, tradable permits provide incentives for the development of cleaner production methods. This offers hope that the next generation of Americans will be substantially less dependent on fossil fuels than the present one. Moreover, as people become more receptive to the idea of tradable permits, it is conceivable that the government may hold back an increasing portion of the annual carbon allowances for auctioning. This is a gradual approach that could at least put us on the path away from the idea that pollution permits should be bestowed freely, with the nation getting nothing in return.