The Center's Assessment of EIA’s Analysis of the Amended (SA.2028) Lieberman-McCain Climate Stewardship Act (S.139)
On June 8, 2004, EIA released its economic analysis of SA.2028, the amended version of S.139: the Climate Stewardship Act introduced by Senators John McCain and Joseph Lieberman. (See evaluation  of the original bill.) This amended version, which would hold U.S. greenhouse gas (GHG) emissions at year 2000 levels by 2010, was considered by the U.S. Senate in October 2003. For more information, see Summary of Act .
The original EIA analysis of S.139 was undertaken at the request of Sen. James M. Inhofe (R-OK), with additional analyses requested by the bill’s sponsors. This recent release – additional work analyzing the amended bill (SA.2028) – follows a request from Sen. Mary Landrieu (D-LA). A separate analysis of SA.2028 was previously released by researchers at the Massachusetts Institute of Technology. See our Summary of MIT’s Analysis. 
The Center has examined EIA’s analysis of SA.2028 and finds it to be consistent with its earlier approach. However, it is still primarily driven by the underlying key assumptions that result in unrealistically high cost projections.
These key underlying assumptions are:
1. High growth of emissions in the baseline case: The high baseline case is the single most important element in explaining EIA’s high cost projections. EIA’s baseline case assumes high growth in:
Furthermore, EIA assumes no relevant policies will be enacted over the next twenty years:
EIA’s assumptions result in an high baseline case which widens the apparent gap that must be closed to comply with SA 2028, and thereby increases EIA’s cost projection.
2. Limited opportunities to increase efficiency and reduce consumption: EIA still assumes that covered entities in the transportation, building and industrial sectors will do very little to increase their efficiency or reduce consumption under SA 2028. For example, EIA forecasts only a very modest increase in automobile efficiency, from 26.4 mpg to 27.2 mpg by 2025. These assumptions are made despite the fact that the GHG cap-and-trade program established by SA 2028 would create a sustained price signal, which should spur technology improvements, and the diffusion of these technologies into the marketplace.
3. Limited opportunities to reduce non-CO2 GHG emissions: EIA still assumes there are limited opportunities for covered entities to reduce their emissions of the non-CO2 GHGs (methane [CH4], nitrous oxide [N2O], and the industrial high GWP gases [HFCs, PFCs, SF6]). Under this assumption, EIA projects that overall emissions of non-CO2 GHGs from covered sectors would actually increase by around 70% under SA 2028 (note that EIA projects an increase of 230% in the baseline case). In contrast, the MIT analysis (see below) finds that overall non-CO2 GHG emissions would be reduced by around 45%. Particular examples that the MIT model finds for cost-effective reduction opportunities in the non-CO2 GHGs include the industrial high-GWP GHGs (HFCs, PFCs, and SF6) in the semiconductor, magnesium and aluminum sectors, and reducing methane emissions from coal and oil production facilities, landfills and natural gas pipelines.
4. Limited available offsets: EIA assumes that there is only a limited amount of cost-effective offsets available. For example, on domestic sequestration, EIA projects only 112 MTC at under $100/TC will be available; however, a forthcoming report from the Center by Robert Stavins of Harvard University  indicates an availability of 300 MTC at under $50/TC. EIA also takes a restrictive view of available international offsets, assuming that trading will only take place with the EU-15 through 2025. This ignores the fact that access to mitigation opportunities in other developed countries and, particularly, in developing countries would greatly increase the supply of inexpensive international credits.
5. Tight natural gas supply: EIA still assumes a very tight supply of natural gas. This assumption would limit one key option for a smooth transition away from highly carbon intensive energy use. However, as shown by EIA’s own evaluation of past reference cases, natural gas has been the fuel with the least accurate forecasts of production, consumption and prices.
Pathway towards emissions reductions
Because EIA assumes limited availability of cost-effective efficiency improvements and limited reduction opportunities for non-CO2 gases, the great majority (88%) of required emissions reductions would come from fuel switching in the electricity sector. As any sizeable transition to natural gas would be limited according to EIA’s predicted tight supply, the EIA model predicts the electricity sector would comply by using expensive renewable and nuclear technologies together with premature retirement of some existing coal plants. Such a large-scale and unplanned shift of capital assets would be expensive. This overall story is the same for the amended SA.2028 as for the original S.139 bill, with the only real change from the relaxed cap being a lower requirement for new nuclear units and some greater retention of existing coal plants (in 2010 coal use is up 7% from 2001 levels, falling to 79% of current levels by 2020).
In addition, EIA's analysis is very conservative in its assumptions regarding the diffusion of those high efficiency technologies that already exist. Among other things, EIA assumes there will be less use of combined heat and power and distributed generation technologies under SA 2028 than in the baseline case. EIA is also pessimistic on the market penetration of new technologies – assuming, for example, that by 2025 no hydrogen and no coal-fired IGCC with sequestration plants would be operational despite these being principal objectives of federal energy R&D.
The actual experience of companies on the Center’s Business Environmental Leadership Council that have elected to take on a GHG reduction target is that low-cost, or cost-saving opportunities are often available – even for much more ambitious targets than proposed in the bill. For example, BP set a target of reducing GHG emissions by 10% from 1990 levels by 2010. By instituting an emissions trading program, BP met its target 8 years ahead of time and achieved $650 million of savings over three years for an estimated outlay of $20 million. DuPont has met its target to reduce GHG emissions by 65% between 1990 and 2010 and managed to use 9% less total energy in 2002 than it did in 1990, despite an almost 30% increase in production. Compared to a linear increase in energy with production, this achievement resulted in $2 billion in cumulative energy savings. EIA’s model shares the weakness of many computable general equilibrium (CGE) or “top-down” economic models – it tends not to recognize these opportunities and, as a result, overstates projected costs.
Comparison with MIT Analysis
MIT developed two cases that focused on Phase 1 reductions as entailed in SA 2028. However, the first case is actually a CO2-only case. This is much less flexible than SA 2028, but the model still projects the same allowance prices as EIA. (Both studies estimate an allowance price of $125/TC [$34/TCO2] in 2020.) The second MIT case assumes unlimited non-CO2 credits in 2020. (Here MIT projects allowance costs much lower than EIA’s, at $52/TC [$14/TCO2] in 2020.) The MIT analysis allows use of all available cost-effective non-CO2 GHG emission reductions.
In contrast to EIA’s emissions reduction pathway through fuel switching, in the MIT analysis, greater efficiency and reductions in non-CO2 GHGs lead to far less pressure on the electricity sector. With these assumptions, MIT predicts a reduction in natural gas prices compared to the baseline case, and coal use remaining stable at current levels.
Translating these emission reduction pathways to the wider U.S. economy, a fundamental point is that for a similar allowance price, EIA projects macro-economic impacts of 0.22% of GDP, while MIT finds a consumer welfare loss of only 0.02% by 2020. Consumer welfare in this case measures lost consumption (or income) by consumers (as leisure effects are not included), and consumption is the major component of GDP (the other components being investment, government expenditures and imports/exports balance). Consumer welfare is a good measure of the actual impact on the population.