A new Discussion Paper out yesterday from Dallas Burtraw and Matt Woerman at Resources For the Future (RFF) confirms the main conclusion from NRDC’s December 2012 report on curbing carbon pollution from power plant: The Clean Air Act can be used to achieve big reductions, with big health and environmental benefits, at a relatively low cost. This new analysis, using a different economic model and a somewhat different approach, is very timely since the president has directed EPA to set flexible carbon pollution standards for power plants under the Clean Air Act.

The key result from Burtraw and Woerman’s analysis (B&W) is that a CO2 emission rate standard that allows averaging across all electricity generators could reduce electric sector CO2 emissions in 2020 by 380 million tons compared to business as usual, or 16 percent, without significantly increasing electricity prices (average retail electricity prices in 2020 are projected to be 1.3 percent higher than they would be otherwise). The total cost of achieving these emission reductions would be $9.7 billion according to RFF’s Haiku model, but the benefits would be $34.8 billion, a 3.5-to-1 return.

These results are quite consistent with the conclusions of NRDC’s analysis. We analyzed a specific proposal for how EPA could establish power plant carbon pollution standards that tailors pollution limits to the energy mix of each state, and gives electric utilities the flexibility to hit their targets in the most cost effective way.  We found that our proposal would reduce emissions by 560 million tons, with estimated benefits of $25-$60 billion in 2020. We found that this could be achieved at a considerably lower cost of only about $4 billion by relying heavily on cost-effective energy efficiency programs, which were not considered by B&W. We found a slight reduction in wholesale electricity prices on average in the regions we examined, but did not estimate retail prices, so B&W have contributed an important new finding by showing that retail prices would increase so little even in a scenario that is substantially more expensive overall than the one NRDC modeled.

B&W also demonstrate the value of allowing broad compliance flexibility (albeit not as broad as in NRDC’s proposal). This is seen most clearly by comparing their “Coal” scenario to their “AllGen-MC” scenario described above. These cases were constructed by holding the marginal cost of the carbon standard constant (at $33 per ton), but limiting compliance flexibility in the “Coal” scenario to averaging emission rates only within the coal-fired power plant fleet. Not surprisingly, total carbon pollution is reduced less under the narrower standard. Perhaps more surprising is how much less: The coal only standard reduces emissions by only 93 million tons, or only one-quarter as much as in a “AllGen” standard with the same marginal cost for CO2 emission reductions. The total costs of the “Coal” standard are also less ($1.4 billion) but so are the benefits (only $6.4 billion). That means that the net benefits are far less—$5 billion compared to $25 billion in the “AllGen-MC” case.  

Both B&W and NRDC’s analyses are probably too conservative. The reference cases for both studies were benchmarked to the Energy Information Administration’s 2011 Annual Energy Outlook. A lot has changed in the electricity market since then. Many older coal plants have retired or announced that they will retire over the next few years; expected natural gas prices are lower; and expected growth in electricity generated by wind and solar is higher. As a result, expected carbon dioxide emissions in 2020 without federal carbon emission standards are 6 percent lower in the 2013 forecast than they were in the 2011 forecast. Given this revised baseline the cost of hitting any given carbon emissions level will be lower than estimated using the 2011 reference case.

While B&W’s analysis is extremely valuable, I don’t agree with all of their policy recommendations. Both NRDC and B&W agree that EPA should offer states the opportunity to submit plans that differ from the approach specified in the guideline EPA issues provided that the state plan is as “stringent” as EPA’s model. B&W suggest that the concept of stringency is ambiguous and that EPA should consider multiple criteria in evaluating state plans. The problem with this concept is that EPA would not have a clear basis for determining whether any given state plan is acceptable or not. Multiple evaluation criteria may be valuable intellectually, but EPA needs a single metric to decide whether or not to approve a state plan. Given the goal of the program the metric that matters is total tons of carbon pollution from the state. As long as a state can show that its plan will result in power sector carbon emissions equal to, or lower than, what would occur under EPA’s model rule then EPA should approve the plan; if a state can’t make this showing their plan should be rejected.

The Clean Air Act requires EPA to consider costs in establishing its guideline, but cost can’t be the basis for determining the environmental equivalency of state plans because states must be allowed to achieve emission reductions in a more expensive way if they want to. In other words, EPA can’t outlaw stupidity, but it shouldn’t reward it either. For example, if EPA’s model rule would reduce emissions in State X by 5 million tons and cost $20 million, EPA shouldn’t approve an alternative plan that also costs $20 million but only reduces emissions by 1 million tons. On the other hand, if State X wants to spend $100 million to get 5 million tons of emission reductions that’s up to it (and its citizens).

This is an important policy issue and I look forward to a fruitful dialogue with Burtraw and Woerman about it, but it should not detract from the crucial bottom line from their analysis: The Clean Air Act can deliver big reductions in carbon pollution at low cost. Fortunately the president has directed EPA to do just that.