
Waxman-Markey sets up a cap-and-trade system for putting a price on the emission of carbon dioxide and other greenhouse gases. So doing will force industries, businesses, and consumers to find ways to emit less. But Friedman holds his nose and indicates that a "simple, straightforward carbon tax would have made much more sense than this Rube Goldberg contraption."
True, Waxman-Markey ended up being over 1,300 pages long, and it is certainly a Rube Goldberg contraption. Yet I think its beating heart, a cap-and-trade program, beats a carbon tax.
A carbon tax would collect money from greenhouse-gas emitters for each metric ton they emit of carbon dioxide or its climate-changing equivalents.
Cap-and-trade à la Waxman-Markey would likewise "price carbon." Below is a primer on how it would do so.
What Is Cap-and-Trade?
In a cap-and-trade system, companies that emit greenhouse gases, or make products that do, are required to possess one allowance or permit for each metric ton of emissions they are responsible for.
For example, a company that generates electric power by burning coal might need allowances to cover all X tons of its CO2 emissions. So might a company that refines petroleum into gasoline that, when used in a car, causes the car to emit carbon dioxide through its tailpipe.
The government would issue only so many allowances each year. The number of allowances issued each year would equal the total number of tons of greenhouse gases permitted for the year. That number of tons of greenhouse gases is the cap for the year.
Companies that need allowances are called covered entities. Covered entities that need allowances in a quantity greater than they have on hand would buy them on an open market from entities that have more allowances than they need. This buying and selling of allowances in an open marketplace is the trade part of cap and trade.
Economic theory says that covered entities would prefer to sell some or all of their allowances, rather than use them to cover their greenhouse gas emissions, as long as what it costs them to institute efficiency measures and reduce their own greenhouse gas output is less than the market price of allowances.
Meanwhile, covered entities would prefer to buy allowances on the open market to the extent that their own cost of reducing their greenhouse gas output is greater than the market price of the allowances.
This trading — buying and selling — of allowances in a marketplace ensures that the cost of complying with the emissions cap is as low as possible. It does so by making it financially worthwhile for covered entities that can’t eliminate emissions cheaply to buy allowances from entities that are able to reduce their own emissions more cheaply.
The latter entities wind up reducing their greenhouse emissions more than the former entities do. The marketplace for allowances thus “finds” the most efficient ways for the economy to comply with the cap.
How Does Trading Minimize Costs?
Reducing greenhouse gas emissions to meet a mandatory cap costs money. If every covered entity is simply told to reduce its emissions by (say) 30 percent, Plant A (see the left side of the illustration below, which is from Union of Concerned Scientists, Catalyst Magazine, Spring 2005) would eliminate 180 of its 600 tons of emissions, Plant B 120 of its 400 tons.

(Click on the above illustration to enlarge it.)
Suppose Plant B can reduce its emissions at lower cost than Plant A. It might cost Plant A an average of $50 per ton, or $9,000 total, and Plant B an average of $25 per ton, or $3,000, to meet the twin 30% reductions. The total cost would be $9,000 + $3,000 = $12,000. Would it reduce the total cost of complying with the overall 30% emissions reduction target for Plant B to eliminate more of its emissions than 120 tons, while Plant A eliminates less than 180 tons? Yes, it would.
This can happen with a cap-and-trade-system (see the right side of the illustration) if Plant A buys allowances (called permits in the illustration) from Plant B. Plant B sells 80 tons worth of allowances to Plant A. Now Plant A reduces its emissions by only 100 tons, rather than the original 180 tons. Plant B eliminates 200 tons of emissions, not 120 tons. The total reduction is the same: 300 tons.
But this time, Plant A is able to do its part at an average cost of (say) $34 per ton of reductions. This is less than Plant A’s original $50 per ton because, for any covered entity, the cost per ton goes up as the number of tons of reductions increases.
Likewise, Plant B is now able to do its part at an average cost of (say) $28 per ton of reductions. This is more than the earlier figure of $25 per ton because Plant B is now eliminating more emissions than before.
Plant A is now spending $3,400, which is $34 per ton of reductions times 100 tons. Plant B is now spending $5,600, which is $28 per ton of reductions times 200 tons. The total cost is $3,400 + $5,600 = $9,000. This is $3,000 less than the total cost was before, which was $12,000. Even though Plant B spends somewhat more money than before, Plant A spends so much less money than it did that the total cost of compliance shrinks.
What Determines the Price of Allowances?
The price of allowances turns out to be the cost to both Plant A and Plant B of the last ton of greenhouse gas emissions they eliminate.
Plant B in the example reduces its emissions, ton by ton, at an increasing cost per ton. So does Plant A. The average cost of reducing emissions is $28 per ton for Plant B (see above), but the cost of the last ton of reductions for Plant B might be, say, $60. This is called the marginal compliance cost for Plant B. The average compliance cost is $34 per ton for Plant B, while the marginal compliance cost for Plant B is $60.
Meanwhile, Plant A likewise reduces its emissions. Its average compliance cost is $34 per ton (see above). As it reduces emissions more and more, Plant A’s marginal compliance cost increases. When Plant A’s marginal compliance cost reaches $60, it pays for Plant A to cease reducing its emissions any further and buy allowances from Plant B at a price of $60 per ton.
That is, instead of continuing to seek further emissions reductions at a marginal cost of greater than $60 per ton of reduced emissions, Plant A buys allowances at a price of $60 per ton to cover the 500 tons of emissions it continues to produce.
Plant B (which in this simple example starts out with all 700 tons worth of available allowances) is willing to sell 500 tons worth of allowances to Plant A at a price of $60 per ton because the 200 tons by which Plant B is reducing its own emissions each cost Plant B less than $60 to achieve. But the next ton of Plant B’s reductions would cost it more than $60 to achieve, so Plant B is not willing to sell Plant A any more than 500 tons worth of allowances at a price of $60 per ton.
For Plant B to sell any more allowances to Plant A, the price of the allowances would have to be higher than $60 per ton. But if the allowance price were higher, Plant A would prefer to reduce its emissions even further on its own and buy fewer allowances from Plant B.
Accordingly, the allowances wind up selling at a price which makes the marginal cost of compliance the same for both Plant A and Plant B: $60 per ton. It is when all covered entities in a cap-and-trade system have an equal marginal cost of compliance — which turns out to be equal to the price of the allowances bought and sold on the open market — that the total cost of compliance (here, $9,000) is at its lowest.
If allowances were auctioned (as some but not all would be under Waxman-Markey) the price-per-ton of winning bids at auction could well be something other than (in this example) $60 per ton of emissions. Having auction prices that are different from the eventual market prices of allowances somewhat complicates the economics of a cap-and-trade system, but it does not change the fact that the eventual market prices of allowances will reflect the (identical) marginal compliance cost arrived at independently by all covered entities.
What About a Carbon Tax?
Instead of a cap-and-trade system, a carbon tax could be used to cause a reduction in greenhouse gas emissions. For example, the government could charge emitters $60 per ton of emissions. Extending the above example, Plant A and Plant B would again voluntarily reduce their emissions, as long as the cost per ton of doing so were less than $60. Once the marginal cost of further emissions reductions reached $60 per ton, each plant would simply pay the tax on its remaining emissions.
The result would be just the same as with the cap-and-trade scenario cited above. Total emissions for the year would be reduced from 1,000 tons to 700 tons. Plant A would emit 500 tons, and Plant B would emit 200 tons.
The reason that the outcome would be the same is that the $60 rate at which carbon is being taxed happens to be the one that causes Plants A and B to voluntarily reduce their emissions by 100 tons and 200 tons, respectively, in the cap-and-trade scenario. If the tax rate were set at, say, $55 or $65 per ton, the two plants would arrive independently at other levels of reduction. A $55 rate would achieve lower emissions reductions overall. A $65 rate would achieve higher reductions. Both alternative tax rates would fail to strike an optimal balance between the two plants’ reduction levels: the ones at which the average cost per ton of making the reductions is as low as possible across the system as a whole.
Only when the tax rate happens to match the allowance price in the cap-and-trade scenario — the one that would match the two plants’ marginal costs of making reductions — would the carbon tax achieve the lowest possible total price of emissions reductions. A carbon tax puts the burden on policymakers to set the right tax rate or lose control over both the overall amount of emissions and the total cost of making reductions.
Thus, my primer on the advantages of cap-and-trade. To repeat: a cap-and-trade system beats a carbon tax because it sets a firm cap on greenhouse gases that can be emitted and lets a marketplace in emissions allowances find the lowest-cost way to comply with the cap. In so doing, the market puts the lowest possible price on carbon emissions, so the cost to the economy of a given size of cap is at a minimum.
A carbon tax puts a tax of a known rate on carbon emissions, but if policymakers set the rate wrong, greenhouse-gas emissions might not be reduced as much as hoped, or the costs of making the reductions might be unnecessarily high — or both.