4

Although all fossil fuels contain carbon, coal contains about 1.75 times as much carbon per unit of heat energy as natural gas and about 1.25 times that of oil. Low-cost substitutes for fossil fuels used in electricity generation, transportation, heating and cooling, and manufacturing processes are not currently available or likely to be developed soon. Of course to the extent that pollution prevention and energy efficiency measures can be economically implemented, substitutes would not be needed. In contrast to sulfur dioxide, no commercially feasible technology for scrubbing carbon dioxide from combustion waste gasses is available. Thus, for the foreseeable future, only lower energy consumption or fuel switching could reduce carbon dioxide emissions from fossil fuel combustion.

Energy efficiency in the United States has improved greatly since the oil shocks of the 1970's. The ratio of energy use per dollar of real gross national product declined by 28 percent between 1973 and 1990. Nevertheless the U.S. economy remains more energy intensive than those of most of our major foreign competitors. In 1989 energy per unit of GNP in the United States ranged from about 8 to 16 percent higher than that of West Germany, France, and the United Kingdom, and was 65 percent above that of Japan. Because the United States depends more heavily on coal it is also more carbon intensive than our major trading partners who rely more on nuclear power, hydropower, or natural gas. However u.s. carbon intensity has fallen as fast as in any major country over 1973-1990. In 1989 coal accounted for 57 percent of u.s. electricity generation, compared with 20 percent in Canada, 15 percent in Japan, 16 percent in Italy, 52 percent in Germany, 26 percent in Belgium, 36 percent in the Netherlands, and less than 3 percent in Sweden, Switzerland, and Norway. We believe it is likely that in the long run the reduction of carbon dioxide emissions would be more costly for the United States than for many other industrialized countries. We believe this is the case even though some gains can be made on energy efficiency and are being sought by Administration efforts.

The studies completed so far by DRI using their econometric and energy models for the Department bear this out. The first study, released last April, simulated the effects of restricting the combustion of coal and oil in stationary uses throughout the OECD at 1988 levels beginning in 1990 and continuing through the year 2010, and comparing the results with a base case scenario assuming no restrictions were in effect. The restrictions did not apply outside the OECD countries. The study found that despite the restrictions, total carbon dioxide emissions in the countries studied would be 17 percent higher in 2010 than in 1988, although lower than in the base case, and that economic output in all countries covered would be reduced.

Real cumulative losses in GNP/GDP among all the countries in this study during the 20 years would range from 0.1 percent to 2.7 percent. Among the Group of Seven (G-7) countries, the United States would be among the hardest hit with a cumulative reduction of 0.9 percent in GNP over the 20 year period as

5

compared with the base case, with the biggest losses coming between 2000-2010. Japan would also suffer a loss of GDP of 0.9 percent. The United Kingdom would be the most adversely affected European country among the seven with cumulative GDP losses of 0.8 percent. Because of its heavy reliance on nuclear power for electricity, France would be least affected among the seven, with a 0.2 percent loss over the 20 years.

The study found that unemployment caused by the coal and oil restrictions would be limited, but that workers would instead suffer declines in real wages and a resulting drop in living standards.

Relative changes in global competitiveness would be limited. While most countries would suffer a modest deterioration in trade balances, there would be relatively little change in their competitive position.

The second study, released at the end of January, involved econometric simulations to determine the level of carbon tax which would be required to achieve in the United States and 11 other OECD countries stabilization at 1988 levels by the year 2000, a 10 percent reduction below 1988 levels by 2010, and a 20 percent reduction below 1988 levels by 2020. Again, the taxes were to apply only in the OECD countries. We stipulated that the taxes should be revenue neutral in all countries through reductions in the personal income tax, and also specified the level and timing of the emissions reductions. Otherwise, all assumptions and findings regrading the variables were determined by DRI/McGraw Hill. Further analyses need to be done to determine the sensitivity of the results to alternative modelling assumptions. Many of the assumptions are subject to much uncertainty and hence the results are also uncertain.

The study's salient findings are as follows:
All the nations in the study would experience losses in
economic output to achieve the targets.
The United States would be among those nations most affected
by the carbon tax with a loss of 3.1 percent in cumulative
GNP during the period 1995 - 2020.
Greece and Australia, each with a reduction of 3.5 percent
in GDP, would suffer the greatest economic losses, followed
closely by Spain (down 3.3 percent) and the United States.
In general, countries with relatively large reserves of
natural gas, large hydroelectric resources, or extensive
nuclear electricity facilities such as France, Sweden, and
the Netherlands would fare better in economic terms than
countries with a greater dependency on coal and petroleum.
For the United States, the economic consequences would be
greatest after the turn of the century.
In order for the United States to stabilize its carbon
dioxide emissions at 1988 levels by 2000 a tax of $120 per
ton of carbon is required; in order to achieve a 10 percent
reduction by 2010 a tax of $384 per ton is required; and in
order to achieve a 20 percent reduction by 2020 a tax of
$721 per ton is required. ( A tax of $100 per metric ton of

o

6

carbon would equate to a tax of $71.04 per short ton of
coal, $11.42 per barrel of oil, $1.66 per MCF of natural
gas, and $0.27 per gallon of gasoline.)
Fuel and energy prices would increase sharply. In the United
States increases in 2020 above the base case level would be
as follows: natural gas in the industrial sector up 116
percent, industrial electricity up 157 percent, gasoline up
82 percent, and diesel fuel up 89 percent.
The tax required to achieve stabilization and reduction in
carbon emissions is different for each country.
required for stabilization ranged from $58 in Sweden to $442
in Spain; for a 10 percent reduction from $384 in the United
States to $1,312 in Spain; and for a 20 percent reduction
from $489 in Sweden to $2,428 in Japan.
The cost in lost GNP/GDP per ton of carbon removed would be
substantial, but varies among countries. Averages for the
25 years ending in the year 2020 range from a low of $320
per ton in 1989 dollars in Greece to a high of $986 in
Japan. In the United States, the average cost is estimated
at $755.
The United States would suffer a significant loss in
employment as a result of the carbon tax. Employment would
be lower by 600,000, on average, over the 25 year period but
both employment and the unemployment rate would return to
base case levels by the year 2020.
U.S. wages would be reduced. In 2020 the average wage rate
would be 12 percent below the base case level.
A principal reason why carbon dioxide reduction is more
costly for the United States is that its population is
projected to grow at a substantially higher rate than in
most of the other OECD countries. In 2020 the U.S.
population is projected to be 44 million more than in 1990, ,
an increase of 17.5 percent. of this increase net
immigration is expected to account for about 45 percent. In
contrast the combined population in the eight European
countries included in this study is projected to be only 7.5
million more, an increase of 2.5 percent.

o

The economic impact of carbon taxes would not be uniform across industries or throughout the various regions of the United States given the substantial differences in their fuel mix and energy requirements. According to a 1988 survey the regional shares of total energy consumed in the manufacturing sector were as shown in Figure 1: Northeast 11.2 percent, Midwest 23.3 percent, South 52.8 percent, and West 12.7 percent. Only about 3 percent of the electricity requirement of the Pacific coast states, and 18 percent of that of the New England states, is coal fired. On the other hand coal, which would be the fuel hardest hit by a carbon tax, accounts for about 75 percent of the electricity generated in the Mountain, West North Central, East North Central, and East South Central states, as shown in Figure

Coal use for electricity varies by state from little or none

7

in Maine, Vermont, Rhode Island, Idaho, California, and Hawaii to 90 percent or more in Indiana, Ohio, West Virginia, Kentucky, North Dakota, Colorado, New Mexico, Utah, and Wyoming.

In the manufacturing sector the heaviest use of energy occurs in the basic industries - primary metals, chemicals, rubber, plastics, building materials, paper, and coal and petroleum products. In 1988 these industries accounted for 86 percent of the energy used in the u.s. manufacturing sector. They also accounted for 23 percent of the employment and 31 percent of the value added in manufacturing. In contrast industrial machinery and equipment, electronics, and transportation equipment combined accounted for only about 4 percent of energy use. Not only do the basic industries account for a relatively large share of total energy used in manufacturing, some are also intensive users of energy. For example, in 1988 BTUs required per dollar of value added were about 65,000 for primary aluminum, 96,000 for steel, 151,000 for hydraulic cement, 192,000 for nitrogenous fertilizers, and 96,000 for paperboard. On the other hand in a number of industries energy intensity is relatively low. Energy required in the apparel, furniture, printing, electronics, and instrument industries did not exceed 3,100 BTUs per dollar of value added. (These industry characteristics are detailed in Table 1). This does not mean, however, that industries which are not themselves energy intensive would necessarily be unaffected by carbon taxes. They may use materials and components which are the products of energy intensive industries, and demand for their products could be affected by shifts in the economy caused by carbon taxes.

In order to assess the state and regional effects of the carbon tax scenario, the third DRI/McGraw Hill study for the Department will examine the impacts of the carbon taxes imposed in the second study on the various industry groups and by state. The study will determine changes in a host of economic variables such as industrial production and employment, in comparison with a base case assuming no carbon taxes. This study is currently under way and should be completed in the near future.

A number of other econometric studies to assess the economic impacts of carbon taxes in the United States and other industrialized countries have been conducted by others. These include a U.s. Department of Energy study recently completed, simulations with two models described in a 1990 Congressional Budget Office report, a 1991 OECD study, and a 1991 study by Messrs. Manne and Richels. The first three studies examine the effects of taxes imposed unilaterally by the United States. The latter two are multilateral and are thus comparable in scope to the DRI studies for the Department. This distinction is important. Unilateral action by the United States would be ineffective in reducing global emissions of carbon dioxide and thus would not address the global climate change issue. Multilateral emission reductions are necessary for meaningful reductions in global carbon dioxide emissions. And it is only in the multilateral context that the full impact of changes in

8

international trade flows on a country's economy can be assessed thoroughly.

In the OECD study, a reduction of carbon dioxide emissions to 80 percent of 1990 levels is assumed to be reached by 2010 and maintained through 2020. It found that U.S. GNP would be reduced by 0.8 percent from a no-tax baseline and a carbon tax of $230 per ton in 2020 would be required. The Manne and Richels study found that a $475 a ton tax would be needed and that U.S. GNP would be reduced by 1.5 percent in 2020 in order to achieve a 20 percent reduction in emissions by that year.

In addition to these econometric studies, there are a number of studies based on engineering evaluations of new technologies that suggest that energy efficiency efforts can achieve significant and economically efficient results at low cost. DOE, EPA and other agencies are pursuing efforts to achieve such gains. We also have the results of the interviews with executives of companies in the basic industry sector I mentioned previously. Almost without exception these executives told us that their companies would be severely disadvantaged, or even forced out of business, if high carbon taxes or substantial restraints on industrial use of coal and oil were to be imposed. This would be especially so if the United States adopted restrictions unilaterally, or if foreign countries with substantial productive capacity in the basic industries were to be exempted from such restrictions.

If heavy carbon taxes were imposed by the United States alone, or by the United States and other industrialized countries but not by the developing countries, adverse foreign trade effects in the energy intensive and other industries seem likely. Some developing countries are already substantial producers and exporters of a number of energy intensive products, and a rise in their productive and export capacity could occur if the industrialized countries tax carbon or restrict the industrial use of fossil fuels. Moreover such action by the industrialized countries alone would not reduce global greenhouse gas emissions. Carbon efficiency within the OECD as a whole is projected to rise even as economic growth occurs. Emissions in developing countries are projected to grow substantially along with their populations over the next century.

Proposals by other countries

I would like to defer to the Department of State on the positions of other countries.

Transfer of Technology to Developing countries to Address climate Change

Since developing countries are projected to account for most of future emissions growth, encouraging use of lower emitting technologies in these countries is likely to be a highly cost