1. Carbon taxes need not impose higher regulatory costs than caps and standards. Moreover, high regulatory costs are not necessarily beneficial for either consumers or environmental interests.
2. The "environmental certainty" of caps and standards is a dubious advantage if emissions are not capped at a sustainable level and the regulations provide no incentive for overcompliance, even when emission prices are very low.
3. Taxes and caps are not mutually exclusive. A cap-and-trade system with a price floor combines advantages of both, in that it caps emissions while also guaranteeing at least a minimal ongoing economic incentive. To take an example, if the U.S. Acid Rain program had used a price floor to maintain emission prices at the original expectation level, SO2 emission reductions might have been doubled. The tax-versus-cap debate would become moot if this principle were broadly understood.

Focusing on the first point: Swedish regulations for stationary-source NOx emissions illustrate how free allocation can be implemented in the context of an emission tax. (The same approach is applicable to cap-and-trade with auctioned allowances, with or without a price floor.) At first glance, the Swedish program seems rather odd: It employs no caps, standards, or timetables. It is an emission tax of sorts, but virtually all the tax revenue is given right back to taxed entities. Why would such a scheme be expected to motivate industry to do anything at all?

The trick to the Swedish program is that while taxes are applied to emissions, refunds are allocated in proportion to "useful energy output" (e.g., MWh for electricity generation). You pay based on what you emit; you get money back based on your output. Say a high-emission power plant pays an emission tax equating to $2.00/MWh, while a low-emission plant pays $1.00/MWh. If the refund rate is $1.25/MWh, then the dirty plant pays a net charge of $0.75/MWh and the clean plant gets a net subsidy of $0.25/MWh; but the $1.00/MWh difference between the clean and dirty plants is unaffected. The refund does not affect their relative competitiveness, but makes it possible to raise the emission price by an order of magnitude. That greatly amplifies the technology-forcing incentive.

That's the theory, but does it work? The proof is in the pudding:

"In the two years between the approval and the activation of the Act on NOx, many companies began extensive efforts to reduce emission levels in anticipation of the charges they would face in 1992. ... Overall, the Swedish NOx feebate policy can be described as having surpassed the best expectations set when it was introduced in 1992. Emission levels have plunged much faster than was ever anticipated, with the 35% reduction target set for 1995 (from 1990 levels) achieved two years early in 1993. Thanks to the rebate system, for many firms the installation of NOx-reducing equipment has been a profitable venture." [Barg, S., Duraiappah, A., Exan, V. E., 2000. Economic Instruments for Environmental Policy Making in Ontario. International Institute for Sustainable Development (pp. 48--50)]

"Although the charge system only became official in 1992, steps to reduce emissions of nitrogen oxides had actually started to be taken two years earlier, after the passing of a bill in Parliament in June 1990. Between 1990 and 1995 specific emissions from the affected plants dropped from an average of about 160 milligrams of NOx per megajoule (mg/MJ) of useful energy to 60 mg/MJ, or by about 60 per cent." [Ågren, C., 2000. Nitrogen oxides: emissions charge works well. Acid News 2, 1--4]

"... The Swedish retrofitted unit, in contrast, demonstrates that NOx levels well below the Swedish standard (and also below the German or United States standards) are achievable. ... The Swedish regulatory system, incorporating an economic incentive, clearly motivates the Swedish plant] to achieve minimal NOx rates rather than just comply with the applicable emission standard." [(USEPA), 1997. Performance of Selective Catalytic Reduction on Coal-Fired Steam Generating Units. U.S. Environmental Protection Agency, Office of Air and Radiation. page 37]

"... the fact that polluters are more likely to accept [Refunded Emission Payments] than taxes may be a decisive factor for the political viability of the regulation." [Sterner, T., Hoglund, L., 2000. Output-Based Refunding of Emission Payments: Theory, Distribution of Costs, and International Experience. Discussion Paper 00-29. Published by Resources for the Future. page 6]

"A charge was set at 40 Swedish kronor (SEK) per kilogram of NOx emitted. This corresponds to 6000 USD/ ton, which can be compared to permit prices that are usually in the hundreds of dollars (although occasionally higher) in the US programs for NOx permits. The few other countries in Europe that have NOx fees, like France, Italy and Galicia in Spain, all have fee levels of about 150 USD/ton or less ..." [Isaksson, L., Sterner, S., 2006. Refunded emission payments theory, distribution of costs, and Swedish experience of NOx abatement, in: Ecological Economics 57 (1), 93-106.]

"... With the rebate, however, a similar calculation finds a price rise of less than $0.0004 per KwH. This is one-fifth the rise without a rebate, or an increase in price of between 0.5 and 1.0 percent if Swedish electricity sells for $0.05 to $0.10 per KwH. ... the actual price rise was not of much practical concern." [Wolff, G. H., 2000. When Will Business Want Environmental Taxes? Redefining Progress. page 6]

Output-based tax refunding would similarly be applicable to GHG regulation. From the perspective of regulated entities, the refunds would effectively equate to "free allocation" (like cap-and-trade in the EU), but with the bonus of zero price volatility (unlike cap-and-trade in the EU), so industry would logically favor such a system over cap-and-trade. Moreover, in the low-carbon energy market the combination of high competitive incentives and low regulatory cost could help maintain energy supplies and keep consumer prices low more effectively than simply giving consumers the tax revenue.

In general, any allocation formula (output-based, grandfathered, whatever) can be applied equally well to the distribution of either permit auction or tax revenue. The only difference for a regulated firm is that instead of getting a free allowance having a market value of, say, $100, it would get a $100 cash disbursement.

If cap-and-trade were implemented with auctioned allowances, a price floor could be imposed as an auction reserve price; the system would thereby operate effectively as an emission tax as long as some allowances remain unsold. With revenue refunding, regulated firms' instinctive, knee-jerk opposition to anything that looks or smells like a tax would be overcome when they realize how much money they can make from emission reductions.