This service report was prepared by the Energy Information Administration (EIA), in response to a September 27, 2006, request from Senators Bingaman, Landrieu, Murkowski, Specter, Salazar, and Lugar (Appendix A).  The Senators requested that EIA assess the impacts of a proposal that would regulate emissions of greenhouse gases (GHGs) through an allowance cap-and-trade system.  The program would set the cap to achieve a reduction in emissions relative to economic output, or greenhouse gas intensity.⁷

The analysis presented in this report builds on previous EIA analyses addressing GHG limitation, including earlier EIA reports requested by Senator Bingaman⁸, Senator Salazar⁹, and Senators Inhofe, McCain, and Lieberman.¹⁰  All of the analysis cases incorporate the economic and technology assumptions used in the AEO2006 reference case.  While increased expenditures for research and development (R&D) resulting from the creation of the Climate Change Trust Fund are expected to lead to some technology improvements, a statistically reliable relationship between the level of R&D spending for specific technologies and the impacts of those expenditures has not been developed.  Furthermore, the impact of Federal R&D is also difficult to assess, because the levels of private sector R&D expenditures usually are unknown and often far exceed R&D spending by the Federal Government.

However, the recent reports for Senators Bingaman and Salazar include additional sensitivity analyses on the assumptions made regarding the availability of GHG emissions reductions outside the energy sector and the pace of advances in technology used to produce and consume energy.  The report for Senators Inhofe, McCain, and Lieberman also examines the economic implications of possible alternative approaches to recycling revenues collected by government under a cap-and-trade program in which significant amounts of government revenue is collected from allowance auctions.  Alternative assumptions in these areas can have a major impact on the results obtained, and the insights from those prior sensitivity cases would also be applicable to the proposals analyzed this report.  Readers interested in how the results reported below might be affected by different assumptions in these areas are encouraged to review the earlier reports. 

In this latest request, EIA was asked to analyze the impacts of a draft Congressional bill (Appendix B) that was provided with the analysis request.  A summary of the program and the reasoning behind it (Appendix C) was also provided.  The draft bill calls for an emissions cap-and-trade program similar to that recommended by NCEP, but with differences in timing, flexibility, allowance allocation, and other provisions.  The bill also establishes a program to provide incentives and fund research, development, and deployment (RD&D) of technologies to reduce greenhouse gas emissions.

The gases covered in this analysis of the proposal include energy-related carbon dioxide (CO2), methane from coal mining, nitrous oxide from nitric acid and adipic acid production, hydrofluorocarbons, perfluorocarbons, and sulfur hexafluoride.¹¹  Under the draft bill, the emissions of these gases would be regulated through an “upstream” market-based emissions allowance program, meaning fuel suppliers and producers of greenhouse gases would be regulated.  This contrasts with proposals where the entities responsible for the actual release of the emissions, such as electric power producers, manufacturers, or building owners are held responsible for emissions compliance and reporting.  An upstream approach could simplify the administration of a greenhouse gas cap and trade program by limiting the number of regulated entities. 

The bill establishes caps on annual emissions by creating a fixed number of tradable emission permits, or allowances, which provide the right to release a given amount of greenhouse gases in units of carbon dioxide equivalence, based on the global warming potential of each gas.  To comply, regulated entities would have to submit the allowances to cover the emissions that result from their products.  The number of allowances issued for each year would be determined based on a goal to reduce GHG intensity in two phases.  In the first phase, from 2012 to 2021, the GHG intensity reduction goal is 2.6 percent per year, allowing some growth in the absolute level of emissions, because the gross domestic product is projected to grow faster, about 3 percent annually, over the same period.   Beginning in 2022, the GHG intensity reduction goal is raised to 3.0 percent per year.  With an emissions cap based on projected GHG intensity, the cap is flexible, allowing a higher level of future emissions if economic growth is faster than anticipated or a lower subsequent cap if economic expansion is slower than anticipated.

To limit the potential cost of the program, a “safety-valve” provision allows regulated entities to pay a pre-established emissions fee in lieu of submitting an allowance.  As a result, actual aggregate emissions could exceed the cap if compliance costs reach the safety-valve level, as some entities would pay the fee to meet their obligations.  The alternative emissions fee, or safety-valve price, is initially set at $7 per metric ton of carbon dioxide equivalent for 2012 and increases each year by 5 percent over the projected rate of inflation, as measured by the projected increase in the implicit price deflator for gross domestic product (GDP).  In 2004 dollars, the safety valve is set to $5.89 in 2012 and rises to $14.18 in 2030.

Several provisions offer additional flexibility and extend the effective scope of the program.  Under the emission offset provision, covered entities can submit “offset credits” that reflect certified reductions in greenhouse gases from uncovered (or exempted) sources in place of allowances.  To provide further flexibility, emission allowances can be saved, or banked, for use in subsequent years if they are left unused in the year for which they were issued.  This provides an incentive to over-comply, or reduce emissions further than that required, thus saving the allowances for future years in which marginal compliance costs might be higher.  Another provision provides incentives to reward entities that reduce emissions before the program starts.  

A portion of the emission allowances allocated each year is auctioned by the Federal government, with the percentage set at 10 percent in the first 5 years, then increasing by 2 percentage points each year beginning in 2017.  Table 1 shows the percentage of allowances distributed by year to individual sectors for 2012 through 2030.  The methodology for allocating allowances to individual entities in each of the sectors varies, but, in general, a “grandfathered” approach where the allocation is based on historical output is used.  For example, the allocation to individual coal mines is based on the each coal mine’s share of the total carbon of coal produced over the 3-year period beginning January 1, 2004.  Most of the allowances not auctioned each year are allocated to fossil fuel suppliers and producers of non-CO2 GHGs, electric power generators and other fuel users and GHG emitters in the industrial sector, and State governments.  Some allowances (up to 1 percent) would be provided as incentives to reward entities that have voluntarily reduced emissions in advance of the 2012 start date, while others (up to 5 percent) would be provided as incentive for activities that sequester carbon, such as certain agricultural practices, increases in forestation, and carbon capture and storage.  Increases in carbon sequestration do not substitute for allowance requirements, so any net emissions impact from this provision is incremental.   Insofar as the timing of auctions is concerned, half the emission allowances are auctioned 4 years in advance of the year for which they are issued, and half are auctioned in the year of issue.¹²

In the original analysis request, EIA was asked to analyze two variants of the proposal:  one in which the allowances are mostly allocated freely as indicated in the draft bill, with some allowances auctioned, and another in which no allowances are allocated freely (all are auctioned).  These two cases, referred to as the Phased Auction¹³ case and the Full Auction case, will be the primary focus of the analysis.  Both cases incorporate the $7 per metric ton of carbon dioxide equivalent safety-valve price for 2012, which then increases each year by 5 percent over the projected rate of inflation.  The Phased Auction case incorporates the allowance allocation plan outlined above, while the Full Auction case provides no free allocation of allowances to fossil fuel suppliers and producers of non-CO2 GHGs, electric power generators and other fuel users and GHG emitters in the industrial sector, or State governments. Nor are allowances used as an incentive for carbon sequestration or early action.  Instead, the Full Auction case assumes all of the auction revenue collected goes to the U.S. Treasury.  The RD&D spending under the bill’s Climate Trust Fund provision, capped at a cumulative $50 billion dollars, is treated similarly in the two cases.  The difference, however, is that under the Full Auction case, the auction revenue collected is substantially more than under the Phased Auction case.

In a follow-up letter (Appendix D), EIA was asked to address several additional issues. EIA was asked to examine the impact of higher and lower starting prices for the program’s safety valve price. To address this request, two sensitivity cases with the safety valve prices starting at $5 and $9, referred to as the $5 Phased Auction and $9 Phased Auction cases, were prepared.

EIA was also asked to estimate the impacts of not allowing or limiting GHG offsets. To analyze this issue a case not allowing offsets, the No Offsets case, was prepared. One category of offset credits is for non-energy use of fossil fuels (primarily for chemical feedstocks). The potential carbon dioxide emissions of these fuels would be included in this upstream regulatory approach, so the need to adjust for non-energy use of the fuel arises. With the National Energy Modeling System (NEMS), however, the potential carbon dioxide emissions from these fuels are omitted in the emission accounting and are therefore not treated as an offset in this analysis, but are instead omitted from the regulated emissions directly. Therefore, the No Offsets case does not change the treatment of non-energy use of fossil fuels. The classes of emissions that are treated as offsets and for which emissions abatement cost estimates were available from the Environmental Protection Agency (EPA) are methane emissions from natural gas production and distribution and methane emissions from small landfills. Estimated emissions reductions from these two sources are excluded in the No Offsets case.

EIA was further asked to address “the impact on program costs and the distribution of those costs associated with using a different point of regulation, specifically an alternative in which the point of regulation for coal was downstream (i.e., at electric power plants and industrial sources.” EIA can not quantify the effects of a different point of regulation using NEMS, but the potential implications of different points of regulation will be discussed. The results of the $5 Phased Auction, $9 Phased Auction and No Offset sensitivity cases are only discussed in key areas to highlight important impacts including the impacts on allowance prices, program compliance, and coal use. In previous reports, EIA has prepared numerous greenhouse gas cap-and-trade sensitivity analyses, examining the impacts of alternative emission caps, alternative combinations of emission caps and safety-valve prices, alternative assumptions about potential emission reductions in non-energy related greenhouse gases, and alternative assumptions about the cost of performance of new appliances, motor vehicles, and energy production equipment.¹⁴

Methodology

The analysis of energy sector and energy-related economic impacts of the various GHG emission reduction proposals in this report is based on NEMS results. NEMS projects emissions of energy-related CO2 emissions resulting from the combustion of fossil fuels, representing about 84 percent of total GHG emissions today. For this analysis, the AEO2006 reference case emissions for energy-related CO2 were augmented with baseline emissions projections for other covered GHGs to create a baseline for total covered GHG emissions. Projections of non-CO2 GHG emissions, including the covered non-CO2 gases, are derived from an unpublished, EPA “no-measures” case, a recent update to the “business-as-usual” case cited in the White House Greenhouse Gas Policy Book Addendum¹⁵ released with the Climate Change Initiative. The projections from the Policy Book were based on several EPA-sponsored studies conducted in preparation of the U.S. Department of State’s Climate Action Report 2002.¹⁶ However, the no-measures case used in this analysis was a preliminary, unpublished projection developed by EPA in preparation for a forthcoming update of that report.¹⁷

Simulations of the emissions cap-and-trade policy in NEMS were used to estimate the price of GHG allowances over time and the resulting changes in the energy system. First, starting from the projected level of energy-related CO2 emissions in 2011 from the AEO2006 reference case and the EPA projection for emissions of other GHGs in 2011, the GHG intensity rate reduction targets for each of the analysis cases were translated into annual emissions targets for the 2012 to 2030 period.

NEMS endogenously calculates changes in energy-related CO2 emissions in the analysis cases. The cost of using each fossil fuel includes the costs associated with the GHG allowances needed to cover the emissions produced when they are used. These adjustments influence energy demand and energy-related CO2 emissions. The GHG allowance price also determines the reductions in the emissions of other GHGs based on the abatement cost relationships supplied by EPA. With emission allowance banking, NEMS solves for the time path of permit prices such that cumulative emissions match the cumulative target, provided the permit price remains below the safety-valve permit price. Once the safety-valve permit price is attained and the previously-banked permits are exhausted, actual GHG emissions can exceed the calculated annual emissions target, as covered entities can pay the safety-valve fee in place of providing the government-issued emissions allowances.

The NEMS Macroeconomic Activity Module (MAM), which is based on the Global Insight U.S. model, interacts with the energy supply, demand, and conversion modules of NEMS to solve for an energy-economy equilibrium. In an iterative process within NEMS, MAM reacts to changes in energy prices, energy consumption, and allowance revenue, solving for the effect on macroeconomic and industry level variables such as real GDP, the unemployment rate, inflation, and real industrial output. These economic impacts, in turn, feed back into the energy sectors of NEMS. The cycle is repeated until an integrated solution is obtained. The economic impacts of the legislation stem partly from its impact on energy prices and its effects on production, imports, and exports of energy goods and services. In addition, the auction and distribution of the GHG allowances generate revenue streams to the government and private sectors. The MAM represents the revenue streams accruing to these sectors based on the allowance allocations specified in the draft bill, as well as the bill’s $50 billion in cumulative RD&D¹⁸ expenditures funded from auction revenues. Together, these energy-related price, quantity, and revenue allocation effects impact on the aggregate level of prices, output, and employment within the economy.

While NEMS is able to represent the broad energy and economic impacts of the emissions allowance program, there is little in the model to distinguish the merits of the point of regulation (downstream or upstream) and only a limited ability to represent alternate allowance allocation schemes. Depending on the distribution of allowances, some industries, firms, or localities may be partly or fully compensated for the compliance costs and economic disruption. While such free allowance allocation can offset some of the direct cost of compliance to recipients, their incentives to take action based on allowance opportunity costs are similar whether they are given the allowances or whether they buy them at auction. NEMS is not designed to evaluate the distributional impacts of whether industries are better or worse off under a given allocation scheme, but it does reflect the marginal allowance opportunity costs that give rise to emission compliance activities. In addition, NEMS simulates the broad impacts of the emissions policy on the economy as a whole, without regard to how individual industries or communities will fare under such a program.

This analysis assumes that the transaction and administrative costs of implementing and operating a GHG cap and trade program will be small when compared to the costs of the allowances themselves. The “upstream” regulatory approach in the proposal analyzed, which requires large suppliers of fossil fuel and other sources of greenhouse gases to submit government-issued allowances based on the emissions potential of their products, is designed to reduce the control program costs. Because so many sources produce GHG emissions, a “downstream” approach that focuses on monitoring each source of actual emissions and collecting the required allowances from them could be much more costly. For example, there are more than 100 million residential and commercial buildings and over 200 million personal autos that produce GHG emissions. A program that required monitoring all of these sources would likely be more expensive than the one in the proposed bill. If the program costs were to turn out to be significant when compared to the cost of allowances, the efficiency and success of the program would be impacted.

NEMS, like all models, is a simplified representation of reality. Projections are dependent on the data, methodologies, model structure, and assumptions used to develop them. Since many of the events that shape energy markets are random and cannot be anticipated (including severe weather, technological breakthroughs, and geopolitical developments), energy markets are subject to uncertainty. Moreover, future developments in technologies, demographics, and resources cannot be foreseen with certainty. Nevertheless, well-formulated models are useful in analyzing complex policies, because they ensure consistency in accounting and represent key interrelationships, albeit imperfectly, to provide insights.

EIA’s projections are not statements of what will happen, but what might happen, given technological and demographic trends and current policies and regulations. EIA’s AEO2006 reference case is based on current laws and regulations as of October 31, 2005. Thus, it provides a policy-neutral starting point that can be used to analyze energy policy initiatives. EIA does not propose, advocate, or speculate on future legislative or regulatory changes within its reference case. Laws and regulations are generally assumed to remain as currently enacted or in force (including sunset or expiration provisions); however, the impacts of scheduled regulatory changes, when clearly defined, are reflected.

This report, like other EIA analyses of energy and environmental policy proposals, focuses on the impacts of those proposals on energy choices made by consumers in all sectors and the implications of those decisions for the economy. This focus is consistent with EIA’s statutory mission and expertise. The study does not account for any possible health or environmental benefits that might be associated with curtailing GHG emissions.

Notes

Table 1