Congressional Budget Office
Director's Blog

Recently the Department of Energy (DOE) announced that, subject to certain conditions being met, it would guarantee $8.3 billion in loans for the construction of two nuclear reactors in Georgia.  Those guarantees would be made under title 17 of the Energy Policy Act of 2005. As with any loan guarantee, the government would run the risk of losing money if the borrowers were to default on those loans.

In 2003, CBO prepared a cost estimate for legislation (S. 14) that would have established a similar loan guarantee program. The Congress subsequently enacted a loan guarantee program for advanced nuclear energy facilities and other innovative energy technologies in title 17 of the Energy Policy Act of 2005. Since that time, CBO has estimated the cost of appropriation acts and two other pieces of legislation—S. 1321 and S. 1462—related to DOE’s title 17 loan guarantee program.  A number of people have inquired as to whether the information in those estimates is relevant to estimates of the credit risk of the announced loan guarantees.  The short answer is: not necessarily.

CBO’s 2003 Cost Estimate.  S. 14 would have authorized loan guarantees for up to 50 percent of the construction costs of new nuclear power plants and would have authorized the government to enter into long-term contracts to purchase power from those plants. At that time, CBO concluded that the risk of default would be well above 50 percent, primarily because high construction costs would make such plants uneconomic to operate given the economic outlook at that time. CBO also expressed concern about the technical risks associated with a new generation of nuclear plants and the possibility of delays and interruptions that might result from licensing and regulatory proceedings. After taking into account the amounts the government would recover from continued plant operations after a default, CBO estimated a net subsidy cost to the government equal to about 30 percent of the amount guaranteed. S. 14 was not enacted into law.

CBO’s Recent Cost Estimates Related to the Title 17 Program.  DOE’s authority to guarantee loans under the title 17 program is subject to annual appropriation action. Under those appropriation laws, the subsidy cost of the guarantees must be paid by the borrower.  For a number of reasons, CBO has concluded that it would be difficult to set the fee so as to entirely cover the estimated cost to the government and has therefore estimated that the fees charged to borrowers would be at least 1 percent lower than the likely cost of the guarantees.  As explained in CBO’s 2007 cost estimate for S. 1321, setting the fee accurately is difficult because there is a large degree of uncertainty about the cost and performance of innovative projects. In addition, requiring the borrower to pay the subsidy cost shifts most of the risk back to the project, which may limit how large the fee can be. Borrowers also may turn down a guarantee if they believe DOE’s fee is too high but may go forward if they consider it too low, increasing the likelihood that DOE’s portfolio will include more projects for which the subsidy fee has been underestimated than overestimated. Consequently, CBO assigned a cost of $470 million to the provision of the Omnibus Appropriations Act, 2009 (Public Law 111-8) that authorized DOE to guarantee debt totaling $47 billion under the title 17 program.

Applicability of Those Estimates to Specific Projects. CBO’s 2003 estimate was intended to represent the possible costs of loan guarantees to build the first units of the next generation of nuclear power plants.  The estimate reflected the average financing costs, construction costs, and other project characteristics that were anticipated at that time. The assumptions and analyses supporting that estimate reflected information about the technical, economic, and regulatory environment as it existed in 2003, almost seven years ago.  Such generalized estimates of credit risk may not apply to a guarantee for any particular power plant because of variations in the technical, economic, regulatory, and contractual characteristics of each project. Without such information, much of which would be proprietary, CBO has no basis for estimating the cost to the government of any specific loan guarantee of this type.

CBO’s recent estimates address a wide variety of possible projects involving both nuclear and non-nuclear energy sources. Whether the government will incur subsidy costs for the program as a whole, or for individual projects,  will depend not only on the characteristics of the projects, but also—critically—on what fees the Department of Energy will charge to cover the government’s potential costs. From a practical perspective, there is probably a limit on how large a subsidy fee can be charged without jeopardizing the project’s financial prospects.  What fees will be charged for the guarantees for the two reactors in Georgia is not yet known.

Today CBO released a brief that discusses how activities with emissions that are not subject to limits in a cap-and-trade program might lower the burden of reducing the concentration of greenhouse gases (GHGs) in the atmosphere. Both existing climate policies, such as the European Union’s Emission Trading System, and policies under consideration, such as the American Clean Energy and Security Act (ACESA) of 2009, which was recently passed by the House of Representatives, have recognized the potential for actions— such as disposing of waste in different ways, changing methods of farming, and reducing deforestation— to “offset” the extent to which the use of fossil fuels must be reduced to meet a chosen target for total GHG emissions.

If such offsets—which can be defined as reductions in GHGs from activities not subject to limits on emissions—are less expensive than reductions from limiting the use of fossil fuels, they can reduce the overall economic cost of meeting a target for emissions. But the difficulty of ensuring that offset activities result in verifiable, permanent and incremental reductions in global emissions raises concerns about whether the specified emissions target will actually be met. Those concerns may be especially acute when, as under ACESA, allowable offsets include actions taken outside of the country setting the target for emissions.

Although experience with offsets is not extensive, preliminary evidence suggests that they can significantly lower the economic cost of a cap-and-trade program, even after accounting for the costs of steps taken to increase confidence that the use of offsets represents true incremental reductions in GHGs. CBO estimates that the average annual savings from offsets could be about 70 percent under ACESA. Of course the intended environmental benefit would be fully realized only if the offsets provided the full reduction in global emissions of GHGs for which they are credited.

The 111th Congress is taking up the issue of addressing the risks associated with climate change, a task that would entail the regulation of emissions of a variety of greenhouse gases from a variety of sources. The Environmental Protection Agency estimates that, in 2006, households and businesses in the United States emitted nearly 7.1 billion metric tons of carbon dioxide equivalent of greenhouse gases. Those emissions were partially offset by the net absorption of roughly 900 million metric tons of carbon dioxide  by the nation’s forests and soils.

Several different approaches, or combinations of approaches, could be used to manage emissions, including direct regulations, cap-and-trade restrictions, and taxes that would directly raise the price of emitting gases. A program based on such approaches could be used to regulate any or all of those emissions.

CBO has previously produced several estimates of the budgetary impact of policies designed to mitigate emissions of greenhouse gases and will produce additional estimates during the current Congress. To do so, CBO undertakes a detailed analysis of the specific provisions of the legislation. In particular, it must estimate the marginal, or incremental, cost of reducing emissions of a number of different greenhouse gases at various levels of mitigation and at different points in the future.

On Friday, CBO released a paper describing the methodology that it uses to estimate the costs of mitigating emissions. In preparing its estimates, CBO uses projections of mitigation costs that, by construction, are in the middle of the range of estimates produced by current state-of-the-art energy-economy models. CBO can use its approach to calculate the amount of emissions generated at a given price or tax, or to determine the price or tax required to achieve a given emissions target.

The methodology involves several steps. CBO first projects a base case that serves as a marker against which to measure the effects of the proposed policies; that base case encompasses projections of future greenhouse-gas emissions and future prices of fossil fuels, electricity, and other products and services closely associated with such emissions—all assuming no new federal policies to control those emissions.  For its base-case projections, CBO relies primarily on projections from the Energy Information Administration of the Department of Energy.

Then, CBO estimates how firms and households will respond to the proposed regulatory program. In the case of a cap-and-trade system, CBO determines how the proposals would affect the prices of emission allowances, and estimates how those allowance prices would filter through to prices of fuels and other emission-intensive products, affecting the aggregate demand for such goods and services.

CBO draws on a variety of sources to calculate how sensitive emissions are likely to be to changes in the allowance price. That sensitivity is, in effect, an elasticity of emissions with respect to the price. (An elasticity is a measure of the response of one variable to changes in another; for example, the elasticity of household demand for electricity measures how much an increase in the price of electricity would reduce households’ electricity consumption.) To develop its measures of price sensitivity, CBO applies six different models, available from government agencies, academic institutions, and other researchers, that represent the current state of the art.

CBO’s estimate implicitly includes the sensitivity of end-use energy demand to changes in allowance prices as well as the amount of substitution that might occur among energy sources. For example, rising prices for fossil fuels would lead electric utilities to substitute some sources for others, by using more natural gas or wind and using less coal to generate electricity, but would also lead households and firms to consume less electricity. Both types of responses are implicitly built into CBO’s estimates.

Over the past several years, spurred by both rising gasoline prices and long-standing subsidies for producing ethanol, the use of ethanol as a motor fuel in the United States has grown at an annual average rate of nearly 25 percent.  U.S. consumption of ethanol last year exceeded 9 billion gallons–a record high.  CBO released a paper today that discusses the relationship between ethanol, greenhouse-gas emissions, food prices, and federal spending on nutrition programs.

Most ethanol in the United States is produced from domestically grown corn, and the rapid rise in the fuel’s production and usage means that roughly one-quarter of all corn grown in the U.S. (nearly 3 billion bushels) is now used to produce ethanol. The demand for corn for ethanol production has exerted upward pressure on corn prices and on food prices in general. CBO estimates that the increased use of ethanol accounted for about 10 percent to 15 percent of the rise in food prices between April 2007 and April 2008.

In turn, increases in food prices will boost federal spending for mandatory nutrition programs such as the Supplemental Nutrition Assistance Program (SNAP, formerly known as Food Stamps) and the school lunch program by an estimated $600 million to $900 million in fiscal year 2009. The Special Supplemental Assistance Program for Women, Infants, and Children—better known as WIC—is a discretionary program that provides a specific basket of goods to recipients rather than a set cash benefit, so changes in food prices in 2008 had an immediate impact on costs for the program.  Under the assumption that the effects are much the same, increased production of ethanol would have added less than $75 million in fiscal year 2008 to the cost of serving the same number of WIC participants as in 2007.

Last year the use of ethanol reduced gasoline usage in the United States by about 4 percent and greenhouse-gas emissions from the transportation sector by less than 1 percent. The future impact of ethanol on greenhouse-gas emissions is unclear. Research suggests that in the short run, the production, distribution, and consumption of ethanol will create about 20 percent fewer greenhouse gas emissions than the equivalent processes for gasoline. In the long run, if increases in the production of ethanol led to a large amount of forests or grasslands being converted into new cropland, those changes in land use could more than offset any reduction in greenhouse-gas emissions—because forests and grasslands naturally absorb more carbon from the atmosphere than cropland absorbs. In the future, the use of cellulosic ethanol, which is made from wood, grasses, and agricultural plant wastes rather than corn, might reduce greenhouse-gas emissions more substantially, but current technologies for producing cellulosic ethanol are not yet commercially viable.

CBO issued a study today examining possible future private investment in new nuclear power plants. The extent of such investment depends not only on possible charges for carbon dioxide (if the Congress adopts climate change legislation) but also on existing incentives provided for such plants in the Energy Policy Act (EPAct) of 2005.

The Energy Information Administration (EIA) projects that demand for electricity in the United States will increase by 20 percent by the end of the next decade. Most of the additional demand would likely be met by conventional fossil-fuel technologies without the incentives in EPAct or the prospects of a market price on carbon emissions.

• Carbon dioxide charges of about $45 per metric ton would probably make nuclear generation competitive with conventional fossil fuel technologies as a source of new capacity and could lead utilities to build new nuclear plants that would eventually replace existing coal power plants. At charges below that threshold, conventional gas technology would probably be a more economic source of baseload capacity than coal technology. Below about $5 per metric ton, conventional coal technology would probably be the lowest cost source of new capacity.
• EPAct incentives would probably make nuclear generation a competitive technology for limited additions to base-load capacity, even in the absence of carbon dioxide charges. However, because some of those incentives are backed by a fixed amount of funding, they would be diluted as the number of nuclear projects increased; consequently, CBO anticipates that only a few of the currently proposed plants would be built if utilities did not expect carbon dioxide charges to be imposed.
• Uncertainties about future construction costs or natural gas prices could deter investment in nuclear power. In particular, if construction costs for new nuclear power plants proved to be as high as the average cost of nuclear plants built in the 1970s and 1980s (adjusted for inflation), or if natural gas prices fell back to the levels seen in the 1990s, then new nuclear capacity would not be competitive, regardless of the incentives provided by EPAct. Such variations in construction or fuel costs would be less likely to deter investment in new nuclear capacity if investors anticipate a carbon dioxide charge, but those charges would probably have to exceed $80 per metric ton in order for nuclear technology to remain competitive under a scenario with high construction costs and low natural gas prices.

The study was written by Justin Falk of our Microeconomic Studies Division.

CBO released a study today on consumers’ responses to the substantial upward trend in gasoline prices that began in 2003.

Many drivers have responded to higher gasoline prices in the way that they drive, but overall the response has been very small.

• Freeway-driving motorists have adjusted to higher prices by making somewhat fewer trips and by driving somewhat more slowly.
  • CBO used data collected at a dozen metropolitan highway locations in California, along with data on gasoline prices in California, to identify changes in driving patterns: On weekdays in the study period, for every 50 cent increase in the price of gasoline, the number of freeway trips declined by about 0.7 percent in areas where rail transit is a nearby substitute for driving; transit ridership on the corresponding systems increased by a commensurate amount; and on weekends median speeds on uncongested freeways declined by about 0.75 miles per hour.

• After increasing steadily for more than 20 years, the market share of light trucks (including SUVs and minivans), relative to all new passenger vehicles, began to decline in 2004. As a result, the average fuel economy of new vehicles has increased by more than half a mile per gallon since 2004 (because light trucks tend to be less fuel efficient than cars). Stricter fuel economy standards for light trucks have also contributed to that increase

• Used-vehicle prices have shifted, reflecting changing demand, particularly with respect to fuel economy: The average prices for larger, less-fuel-efficient models have declined over the past five years as average prices for the most fuel efficient automobiles have risen.

• Total U.S. sales of midgrade and premium gasoline have declined gradually since 2000, even as consumption of the less expensive regular formulations has increased. Although consumption of the different grades of gasoline depends strongly on what kinds of vehicles consumers drive (and on manufacturers’ fuel octane recommendations for those vehicles), it also might have been influenced by the general increase in gasoline prices since 2003.

The study notes that the response of consumers to higher gasoline prices has important implications for policies that affect gasoline consumption, including CAFE (Corporate Average Fuel Economy) standards for cars and light trucks. Because higher gasoline prices increase the demand for vehicles with better fuel economy ratings, they reduce the economic costs (and fuel savings) of adopting more-stringent CAFE standards. At the same time, to the extent stricter CAFE standards improve fuel efficiency beyond what consumers would choose in the absence of such standards, they reduce the per-mile costs of driving — which would partially reverse some of the effects of higher gasoline prices discussed in this study. The federal tax on gasoline, by contrast, reinforces rather than neutralizes the behavioral and vehicle choice effects of higher gasoline prices. It also immediately affects all motorists’ incentives to reduce gasoline consumption, whereas CAFE standards primarily affect motorists only after they replace the vehicles they were driving at the time the standards were implemented.

David Austin, an economist in CBO’s Microeconomic Studies Division, wrote the report. In addition to his work on gas prices and CAFE, David has done research in the areas of liability policy and toxic emissions; Clean Air Act regulations; consumer benefits of new technologies; and allocation of emissions controls, and research and development in the pharmaceutical industry. He has been at CBO for six years; prior to that was at Resources for the Future for eight years. He received his undergraduate degree from Stanford University and his economics Ph.D. from UC Berkeley. He also has a master’s degree in statistics from Yale University. And he has an impressive track-and-field record, including a mile best of 4:19.