Economic Incentives for Mitigating Global Climate Change:
Comparing emission permits and emission taxes

4 Sep 2007 in Regulatory Economics

In a Wall Street Journal op-ed (temporarily available to nonsubscribers), City Journal contributing editor Nicole Gelinas argues that a carbon tax would be a "cleaner" (apparently meaning "more efficient") solution for mitigating global climate change than the several emission permit proposals ("cap-and-trade") that have been proposed. Gelinas provides a useful introduction to the debate, but misses crucial aspects of how these economic incentive systems are the same, and how they differ.

It has been understood among economists for more than 20 years that emission taxes and emission permits are analytically equivalent tools when diagrammed on the economist's blackboard. Thus, if the stringent conditions imposed by the mathematics actually apply, it doesn't make any difference which economic incentive instrument policy makers select.

The problem is that these conditions do not apply. Taxes and permits are both highly susceptible to a range of implementation problems and perverse incentives. Determining which approach would be more efficient requires a very complicated analysis that Gelinas does not present. (The City Journal article on which the Journal op-ed was based is here.)

The table below summarizes a few of these issues.

ISSUES COMMON TO PERMITS AND TAXES

Commodity uncertainty

Clarity is essential concerning what exactly must be done to obtain an emissions permit, or what exactly will be subject to the tax. The greenhouse warming potential of carbon dioxide and other gases has been estimated by EPA and the Intergovernmental Panel on Climate Change, and EPA has recently published a 2005 US emission inventory. Yet most of the attention has been devoted to carbon dioxide. Would all greenhouse gases be covered in proportion to their global warming potential? If not, then achieving any fixed carbon reduction target would require either too few carbon dioxide permits to be issued or too high a carbon dioxide tax be levied.

There is no obvious reason why a permit scheme or tax program would have a better defined commodity. Efficiency is best served with the widest possible base.

Measurement uncertainty

Both permits and taxes require accurate emissions data. Large measurement or estimation errors weaken the program's intended incentive effects by making the cost of the program seem random. Biases in measurement or estimation undermine public confidence in the program by at least inviting suspicions of favoritism. A successful permit or tax system requires that emissions data and estimates be both unbiased and reasonably precise, in fact and in appearance. Either incentive systems fails if the public perceives that favored interests can game the system.

Legislative uncertainty

Uncertainty about program design and implementation undermines programmatic success. Normal financial risk aversion leads regulated parties to respond to uncertainty by discounting governmental promises and commitments. Regulated parties may be reasonably concerned that the rules of the game will be changed after it gets underway, and if so they can be expected to behave strategically. Uncertainty about governmental actions is not necessarily cynical, just prudently precautionary.

Such precaution is justified because no future Congress can be bound by the actions of the current one. There is always a risk that any permit program or tax regime will be changed in the future, and this risk is always factored into market expectations. Moreover, Congress has the means, opportunity and often a motive to change a program surreptitiously, such as by adding "midnight riders" to omnibus bills enacted at the hurried end of a legislative session. In theory, this problem can be policed in part by the adoption of parliamentary rules that make it easy for Members to successfully challenge such actions. In practice, however, neither the House nor the Senate is inclined to impose on itself rules that fundamentally limit their discretion.

Legislative uncertainty also can arise if the agency tasked with implementing the program also is given significant administrative discretion. Even when an agency lacks administrative discretion, it always has enforcement discretion. The more discretion that is available to the agency, the greater will be the uncertainty about how the program will be implemented.

Legislative uncertainty is a greater problem for permit systems than tax regimes. If Congress writes tax rates into law, no agency can change them by administrative action. In contrast, cap-and-trade proposals are often ambiguous about how many permits will be issued, who is required to obtain them, what must be done to create them, and whether they are genuine property rights (more about that below.)

Administrative and transactions costs

All government programs are costly to administer, and regulated parties must bear nontrivial transactions costs whether the program is founded on permits or taxes.

In general, permit systems appear to have higher unit administrative costs. Both permit and tax regimes require regulated entities to measure or estimate emissions, maintain required records, and file relevant forms or reports. Under a permit system, however, both buyer and seller must perform due diligence before they consummate an exchange. These costs might be reduced if they are delegated to an intermediary such as a certifying agent, but however they are managed they are costs that do not arise in a tax system because no market exchange is involved. Once a regulated entity pays the tax, its obligations are fulfilled.

A tax regime can be designed to be more cumbersome than a permit system, but that requires a policy choice.

PROBLEMS PREDOMINANTLY ASSOCIATED WITH PERMITS

Baseline uncertainty

The choice of baseline is crucial for any permit system, and there is no obvious "best" choice. Using any date in the future encourages a ramp-up in emissions that makes future reductions less onerous. Using any date in the past rewards some actors and penalizes others. The Kyoto Protocol, for example, used a baseline of 1990 that benefited the United Kingdom (having previously decided to replace coal with natural gas), Germany (recently reunified), and Russia (the remnant of the imploded Soviet Union), and plausibly harmed France (which had the greatest reliance on nuclear power).

In short, the choice of baseline implicitly rewards energy inefficiency. The reward is larger if this inefficiency also is associated with reliance on carbon-based fuels. This leads to four predictions:

Regulated entities that are energy-efficient will tend to prefer emission taxes rather than permits.
Regulated entities that are energy-inefficient will tend to prefer emission permits rather than taxes. (Support for permits will be attenuated if they are allocated by auction.)
When considering alternative emission permit schemes, the preferences of regulated entities will depend on whether they can capture credit for recent (but not statutorily required) actions taken to improve energy efficiency.
The lower a regulated entity's relative "carbon footprint," the more it will prefer a system system in which permits are auctioned.

Any permit system is undermined if government can retroactively change the baseline. Increases in the baseline reduce the number of permits needed to achieve any fixed carbon reduction target, and thus lower the value of permits. Conversely, reductions in the baseline increase their value. Permit holders (and prospective permit creators) will be acutely aware of and sensitive to any indication that the government retains the authority to retroactively change the baseline.

As long as a tax cannot be imposed retroactively, tax regimes are insensitive to this gamesmanship. As in the case of permits, however, any future date for commencing a tax regime will create an incentive to expedite emissions in advance of the effective date.

Property rights uncertainty

A successful permit system requires that the property value contained in the permit be inviolable. Neither government nor any other party can be allowed to expropriate it or interfere with its sale. Otherwise, buyers will pay less than the full market value to acquire permits, with the market price approaching zero as legal uncertainty about property rights rises.

Previous federal emissions permit programs did not provide surety in property. Under the Acid Rain Program authorized by the Clean Air Act Amendments of 1990, for example, permits were not genuine property rights. States were not prohibited from interfering with their sale or purchase, and several states did so. The government also extracted a 20% charge on transactions, thereby making all exchanges less attractive (and permit prices lower) than they otherwise would have been.

Private rentseeking

A permit system is especially sensitive to allocation by means other than economic efficiency, such as politics. Properly designed auctions roughly ensure that permits go to the highest bidder, meaning the purchaser for whom they offer the greatest value. This could be a regulated entity such as an electric utility, which would use them to "pay for" carbon emissions, or to an investor interested only in their value as a financial asset. Or the highest bidder could be an environmental group or Hollywood celebrity.

Rentseeking arises if permits are allocated by any means other than an auction. Because it only reallocates wealth, rentseeking has no social or economic value. It also creates an environment in which corruption is an omnipresent temptation.

Government rentseeking

A special form of rentseeking afflicts government: The desire for additional tax revenue. Auctioning emissions permits enables the government to obtain a large amount of new revenue. Thus, private rentseeking can be avoided by auctioning permits, but auctioning permits invites governmental rentseeking. This problem of governmental rentseeking can be minimized by crafting an emission permit program so that it is revenue neutral -- that is, taxes can be reduced dollar-for-dollar elsewhere for every dollar obtained by auctioning emission permits. (A commitment to earmark auction proceeds for some worthy cause is not the same thing. First, it is explicitly not revenue neutral. Second, money is fungible; general fund appropriations to that worthy cause can be reduced dollar-for-dollar so that total resources are unchanged, and the earmark has no practical effect.)

PROBLEMS PREDOMINANTLY ASSOCIATED WITH EMISSION TAXES

Government rentseeking

The same kind of government rentseeking that afflicts auctioned emission permits is a central feature of emission taxes. As in the case of auctioned permits, the problem can be minimized by explicitly designing an emission tax regime to be revenue neutral.

Revenue dependency

Government rentseeking is a more serious problem for emission taxes. Unlike auctions, which could result in a one-time surge in revenues, a tax regime generates revenue every year that it is in place. Governments can (and often do) become dependent on these revenues, and this creates special problems when the tax is intended solely to internalize an externality.

Over time, taxpayers alter their behavior in desirable ways. In the case of an emission tax, that means reduced emissions of greenhouse gases and presumably less damage from global climate change. However, once government has become dependent on the revenues generated by an emission tax, the very behavioral change that the tax was intended to motivate now works to the government's disadvantage as a revenue source. As "bad" behavior declines, government collects less revenue. It has three options:

Reduce its spending to match the reduction in income.

This is self-explanatory, but it's hard to find an example in which government has responded this way.

Increase the tax rate on "bad" behavior.

This can work only in the short run, however, as higher tax rates motivate additional behavioral change, which results in more revenue losses. In the 1980s, many states funded their new hazardous waste management programs by levying taxes on waste disposal. Firms responded by reducing waste generation, which reduced the amount revenue collected to support the programs, which led to increases in waste disposal tax rates and further reductions in waste disposal. Note that the economic purpose of the tax -- internalizing the externality of hazardous waste -- was quickly overwhelmed by government rentseeking.

Indirectly support or encourage "bad" behavior to maintain the flow of revenue.

When governments become dependent on the revenue from taxes on "bad" behavior, they become complicit in sustaining the conditions in which "bad" behavior can flourish. This is precisely what has happened to the States after they negotiated the Master Settlement Agreement with Big Tobacco. California reports that it stands to collect $908 million in 2008, and its counties anywhere from $18,000 (Alpine) to $126 million (Los Angeles; plus $11 million to the City of Los Angeles), with rising amounts forecast through mid-century. Even rabidly anti-smoking San Francisco will obtain more than $20 million. Only the 22 counties that have securitized their future receipts are free from the temptation to indirectly encourage smoking so as to keep the revenue coming.

A similar fate awaits the federal government if Congress enacts the proposed expansion of the State Children's Health Insurance Program, which depends on smokers willingly paying much higher federal taxes and not responding by reducing tobacco consumption

IMPORTANT IMPLEMENTATION ISSUES WITH EMISSION PERMITS AND EMISSION TAXES
Issue	Emissions Permits ("Cap-and-Trade"	Emissions Taxes ("Carbon Tax")
Issues common to both permits and taxes
Commodity Uncertainty	Are emission permits well-defined commodities? What greenhouse gas emissions are excluded? (Exclusions reduce efficiency.)	Is the emission tax base well-defined? What greenhouse gas emissions are excluded? (Exclusions reduce efficiency.)
Measurement Uncertainty	Is the quantity of emissions covered by a permit easy to estimate?	Is the quantity of emissions subject to tax easy to measure?
Legislative Uncertainty	Can the number of permits (or other material program elements) be changed administratively or legislatively?	Can the tax base or tax rate be changed administratively or legislatively?
Administrative Costs	Is the cost of making and implementing the permit market low? Is the cost of enforcement low?	Is the cost of voluntary tax compliance low? Is the cost of enforcement low?
Government Rentseeking	If permits are awarded: Susceptible to criminal activity (e.g., bribery, extortion, blackmail, etc.). Susceptible to legally permitted "pay-to-play" campaign finance activities. If permits are auctioned: Generally avoided if revenues are offset by tax reductions elsewhere. Design of the auction can be crucial.	Avoided only if revenues are offset by tax reductions elsewhere. If tax regime is not revenue neutral, then susceptible to revenue dependency (see below).
Issues predominantly associated with permits
Baseline Uncertainty	What baseline year is used for measuring progress toward the goal? Is the baseline resistant to pressure that it be altered after the act?	As long as taxes are not levied on past emissions, the choice of baseline is immaterial.
Property Rights Uncertainty	Are emission permits inviolable property rights? Are they fully tradable without restriction, including restrictions sought by states or NGOs? Can regulatory agencies (including 3rd-party agencies, such as state utility commissions) indirectly punish or reward buyers or sellers?	Not applicable.
Private Rentseeking	Associated with non-price allocation (e.g., assignment based on past emissions) Not associated with auctions (but may be associated with auction format). Fertile environment for corruption (see Government Rentseeking above).	Not feasible if all emitters pay the same tax per unit of greenhouse warming potential.
Transparency	Impact of permits on consumer prices: Hard to estimate Difficult to make transparent	Impact of permits on consumer prices: Easy to estimate Can be made transparent by thoughtful program design
Issues predominantly associated with taxes
Government rentseeking	Relevant only if permits are auctioned: Revenue neutrality requires taxes elsewhere be reduced dollar-for-dollar Government will be tempted to authorize too many permits	Relevant to all tax regimes: Avoided if taxes elsewhere are reduced dollar-for-dollar (e.g., revenue neutrality) If not revenue neutral, then susceptible to revenue dependency (see below)
Government revenue dependency	Feasible only if permits are auctioned Applies only of permits are auctioned frequently	May be seriously incompatible with efficiency: To sustain future revenue, government will be tempted to raise tax rate above level need to internalize the externality To sustain future revenue, government will be tempted to indirectly encourage taxable emissions

[add a comment]