PNL-SA-22977
Karen Fisher-Vanden
Jae Edmonds
Hugh Pitcher
Dave Barns
Richard Baron
Sonny Kim
Chris MacCracken
Elizabeth L. Malone
Ronald D. Sands
Marshall Wise
September 1993
Prepared for the U.S. Department of Energy under Contract DE-AC06-76RLO 1830
Presented to the Sixth Annual Federal Forecasters Conference, Crystal City,VA.
Pacific Northwest Laboratory
Washington, DC 20024
We begin our discussion with a brief description of the model, describe the derivation of reference year parameters, proceed to develop reference case assumptions, describe the resultant reference case scenario, and then explore the costs and benefits of selected policy options for stabilizing greenhouse emissions.
The analysis reported here uses SGM Version 0.0. While 20 global regions are being developed for the global implementation of the SGM, only the United States module is used in this exercise.
We note that seven of the nine sectors of the economy in SGM Version 0.0 are energy sectors. The emphasis on the energy sector in SGM Version 0.0 reflects the central role energy plays in the determination of emissions of greenhouse gases to the atmosphere.
In addition to the producing sectors of the SGM, there are four final demand sectors:
Producing sectors use goods and services produced by other producing sectors, itself, and primary factors of production to produce net output. The three primary factors of production are:
There is a market, which is cleared by the price mechanism, for each producing sector of the economy. Similarly there is a market for each of these three primary factors of production.
The interactions of the main components of the SGM are shown in Figure 1, where the seven energy sectors are included in the energy box. In general energy, agricultural, and other production sectors supply net quantities of goods and services for each other and for final consumption by households, the government, investment, and net exports. Primary factors of production -land, labor and capital- are also used by the producing sectors. Labor and land are assumed to be owned by members of the household sector who supply them to the market. Capital is associated with producing sectors. Net profits are returned to the household sector, while taxes are collected from producers and households and provide the revenues for local, regional, and national governments.
Producing sectors are assumed to make decisions regarding production and investment with the objective of maximizing expected wealth. Governments are assumed to produce government services, including education, national defense, and general services, according to prescribed production functions. Households behave as if they are maximizing utility with regard to the allocation of resources to current consumption; however, the supply of savings and labor is modeled by exogenously specified rules.
All produced goods and services, and primary factors of production pass through markets which are assumed to clear in every period via a price mechanism. Estimates of gaseous emissions are computed in the model via the application of emissions coefficients to inputs and outputs of processes. Variation in these coefficients reflect variation in the emissions characteristics of alternative human activities.
In the remainder of this section, we will briefly highlight some of the more salient features of the SGM.
Just as sectors can have subsectors, subsectors can have technologies. Technologies are alternative modes for producing the product of the subsector. For example, there may be two technologies for producing electricity with natural gas (a subsector of Electric Power Generation): Conventional combustion, and Combined Cycle combustion. The framework of the model is sufficiently flexible to allow an arbitrary number of technologies to compete.
Fossil fuels and uranium are treated as depletable natural resources. They are divided into two categories, Resources and Reserves. Reserves are those energy sources whose location is known, which are producible using present technologies under present and anticipated economic conditions, and which investment for extraction purposes has been made. (Some crude oil, discovered conventional, and some Alaskan oil are examples of crude oil reserves.) Resources include all energy sources, including those which are known by location and those whose existence is inferred, those which are producible under present technologies and economic conditions, and those which may require greater incentives to exploit. Total resources include reserves. (In the SGM, reserves are created as economically recoverable resources are discovered). In the SGM, energy production from depletable natural resources occurs only from reserves. The rate of production from reserves depends on the amount of productive capacity put in place with discovery, and prices of inputs and outputs.
Biomass, solar power, and hydroelectric power are renewable resources. Biomass feedstocks are treated as a component of the agricultural production. Total domestic production is therefore constrained both by the total available arable land, and the competition for the use of that land for other purposes. Biomass feedstocks are either transformed to liquids or gases, or consumed as a solid by electric utilities.
Electric utilities also consume solar and hydro power. Hydro power capacity is fixed for the purposes of this analysis. No resource constraint is placed on the solar power component, a source of electricity generation based on photovoltaic array technology.
Land is modeled as a separate factor of production. The identification of land reflects its potential importance as a constraint in emissions reduction strategies which utilize biomass or carbon sequestration by trees.
The last capability can be used to explore behavior when future prices are known. We note also that different combinations of expectation formation for market prices and taxes/subsidies can be constructed. Thus, the formulation of price expectations is not restricted to the assumption of perfect knowledge about future events.
Sectors and subsectors with the highest expected profits experience the greatest realized investments, although within a sector a logit function distributes investment resources across all subsectors which are economically attractive. Investment opportunities which have zero or negative expected contributions to wealth (after allowing for taxes and subsidies) receive no investment funds. The inclusion of subsidies in the expected contribution to wealth calculation is important.
The SGM uses a "putty-clay" specification of capital. That is, once an investment occurs, capital is permanently associated with that particular application. Thereafter, capital does not move from one sector to another. This is a particularly useful assumption for modeling energy applications where capital investments are highly specialized. It is also not particularly restrictive to the rest of the economy, which is modeled as a single aggregate sector.
Using a vintage approach to modeling capital also means that capital investments become a fixed cost of production. Thus, existing vintages continue to operate as long as they can cover their operating expenses, even if the rate of profit differs from that anticipated at the time the original investment was made. We note here that other factors of production are not fixed; they can be varied to either expand or contract output from an existing facility. Other factors of production also move freely among alternative applications.
Because a production function is associated with each capital investment by application, factor productivity can be manifest as both embodied and disembodied changes in input requirements. That is, combustion efficiencies maybe embodied in the physical plant and vary depending upon the date of installation. On the other hand, agricultural productivities may be weakly associated with capital stocks, and more closely associated with changing management practices and seedstocks. The SGM can represent both types of productivity change.
A variety of statistical information was used to develop calibration parameter estimates for the SGM. Table 2 contains a summary of these data, their sources, and an indication of their use.
The methods used to transform data into model parameters is documented in detail in Fisher-Vanden, et al (1993).
The SGM United States module requires input assumptions in six different areas: Demographics, Energy Resources, Productivity Change, Nuclear Power, International Trade, Fiscal Policy, and Emissions Coefficients. We discuss each in this section.
Figure 2 shows the demographic profile of the U.S. for our benchmark year of 1985 (Department of Commerce, 1992). The bulge in the chart for the age groups of 20-40 years corresponds to the "baby boom" generation. Even by 1985, it is clear that the cohorts that follow are much smaller. By the first quarter of the next century, the population bulge will have reached retirement age, and the fraction of population that is available for the labor supply may be much smaller than it is currently.
In SGM version 0.0, fertility rates, survival rates, and migration rates are determined exogenously. We have specified initial values for these rates based on data for 1985. We also included a terminal value for each of these rates as a modeling parameter. Both the terminal values and the time to reach the terminal values can be specified by the user. Table 3 shows the initial values of the demographic rates which are assumed to persist throughout the modeled period.
In the United States coal is by far the most abundant conventional fossil fuel resource. In fact, the domestic supply of coal in the cheapest grade--Grade 1--is enough to sustain current levels of production for nearly 300 years. Natural gas is the next most abundant domestic source of energy, although much of its supplies are in the more expensive grades. Gas resources in the first three grades are sufficient to sustain current production for approximately the next 35 years; therefore, the supply of relatively inexpensive natural gas becomes a binding constraint over the year 2030 study horizon of the model. Currently about half of the oil consumed in the U.S. is imported, so the implications of consumption on domestic resources are not as straightforward. Under current production levels, the supply of oil from the first three grades is clearly a constraint over the next 40 years. However, the future price of foreign oil plays a major role in determining domestic investment in the more expensive grades. Even so, domestic supplies of the more costly grades of oil are limited.
One of the more important assumptions affecting the development of a reference description of the United States economy is the set of assumptions which govern productivity change in the variety of human activities in the SGM. Productivity changes in the SGM in either of two ways, through smooth changes in production function coefficients associated with new investment options, which we refer to as general productivity growth, or by the introduction of a new technology option at a specific point in time.1
Assumed rates of change for general, Hicks neutral, productivity in new investments are given in Table 5.
Since the Other Production Sector produces approximately 85 percent of total output, the productivity growth rate of this sector was adjusted to attain a growth forecast of the U.S. economy similar to what was assumed in the National Energy Strategy. Lack of detailed knowledge of variations in future productivity growth between sectors resulted in the assumption of no variation in productivity growth rates between the first five sectors and the Other Production Sector. Future versions of the SGM will incorporate current research in this area to better define variations in productivity growth between these sectors.
Productivity growth rates in two of the three energy transformation sectors are set to 0.0% per year to maintain energy balances between inputs and outputs. These three sectors--Electric Power Generation, Oil Refining, and Natural Gas Transformation--differ from the other producing sectors because they use large quantities of energy as inputs; most of which is passed through as output instead of being consumed. For example, the amount of energy leaving the gas distribution sector is nearly equal to the energy coming in, with almost no opportunity for improvement in the ratio of energy output to energy input. In these sectors changes in productivity are modeled as discrete changes in available production functions over time.
Although productivity improvements in the Electric Power Sector are handled mainly through the incorporation of new technologies in the SGM, small productivity improvements resulting from changes in management practices warrant the inclusion of a small neutral productivity growth assumption. This value (0.2) was chosen in order to attain results within the Electric Power Sector similar to those found in AE093.
This analysis allows five new technologies to be introduced after 1990: liquids from biomass, gases from biomass, solar electric power, clean coal, and new gas turbine. Cost assumptions used to estimate production function parameters for these five technologies are given in Table 6.
The two biomass technologies purchase their feedstock (wood) from the SGM's agricultural sector, thereby competing with other agricultural products for land. Future versions of the SGM will split the Agriculture sector into various crop sectors as well as a biomass feedstock sector.
The SGM allows for electricity sector investments in solar photovoltaics, a clean coal technology (Atmospheric Fluidized Bed Combustion), and a new gas turbine technology (Natural Gas Combined Cycle) beginning in the year 1990. New capital equipment is required for both generation and distribution. Operation and maintenance costs are very small relative to capital costs. Cost assumptions for solar are somewhat optimistic for the near term.
In order to understand the economic forces at work within the SGM, it is useful to examine why GNP growth falls from 1995 to 2030. The decline is steady from 1995 to 2025 when growth rates appear to stabilize at about a half percent per year. This drop occurs primarily because of the fall in the size of working age population (Figure 4), from 182.5 million in 2010 to 167.8 million by 2030, an 8.0 percent decline. This leads to only a 2.9 percent reduction in the number of workers (149.2 million to 144.9 million) during the period 2010 to 2030 as labor force participation increases from 81.8 percent to 86.4 percent, and wage income per worker per year rises from $33,190 per year to $44,380 per year. As might be expected given the scarcity of labor, the capital/labor ratio increases over the same time frame from $136.9 to $165 million per 1000 workers, continuing the trend seen earlier (Table 8).
Since labor is scarce, the price of labor rises over time, leading to a substitution away from labor. This is seen in the much larger decline in the ratio of labor to output than in the ratio of capital to output. Figure 5 presents the path of prices for factors of production capital, labor, and land. Not only does the price of labor rise, but the fraction of the working age population at work rises as seen in Table 8.
The decline in the work age population has another route for affecting growth rates. Part of the determinant of the desired investment is the change in the rate of potential growth of the economy, namely the change in worker productivity and the change in working age population .
While productivity, measured as GNP/Worker, continues to grow rapidly (Table 9), working age population peaks in 2010 and then declines, thus depressing growth in investment. Examination of the components of GNP makes this clear, as investment becomes static after 2015. This leads to an increase in the average age of capital, slowing down the rate at which technical change in realized in the economy. (By assumption, neutral technical change occurs at a constant pace in each industry).
Oil remains the dominant fuel over this period (despite the fact that its price is assumed to rise by more than 50 percent between the years 1990 and 2010) rising to almost 50 exajoules per year by the year 2015. Consumption is relatively stable throughout the remainder of the period of analysis. The escalation in oil consumption leads to a substantial increase in oil imports beginning in the year 1990.
The only primary energy carrier whose price remains relatively stable is coal. We note that the price of electricity rises by about 50 percent between the years 1990 and 2010, but the introduction of improved technologies and a decline in coal prices lead to a reduction of cost by more than one third between 2010 and 2030.
The relative stability of the price of coal compared to other fuels leads to an increase of domestic coal production and consumption which rises by more than 50 percent between 1990 and 2030. Natural gas consumption increases by 40 percent between 1990 and 2015, but depletion of the least expensive grades of domestic natural gas resources after 2015 causes a sharp increase in the price of natural gas and thus stabilizes consumption.
The assumption that no new nuclear powerplants are built leads to a stable pattern of nuclear electricity production in the near term, but after the year 2010 production declines steadily until all nuclear facilities are retired in the year 2025.
Electric power consumption as a whole expands steadily over the period of analysis. Larger increases in consumption of electric power occur after 2015, when a fall in the relative price of coal causes a shift towards electricity consumption.
Emissions from other greenhouse gases are displayed in Figure 8. Since the majority of carbon monoxide (CO) emissions are a result of oil combustion activities, we see the growth in these emissions mirroring that of oil consumption. Since we see a steady increase in oil consumption over the modeling period, we also see a steady rise in CO emissions. Other greenhouse gas emissions also experience a steady increase over the modeling period. SOx emissions stabilize after 2000 but rise again after 2020, mirroring coal consumption during this period.
Two strategies for stabilizing carbon emissions were analyzed:; (1) constant and variable taxes on fossil fuels based on each fuels carbon content; and (2) variable energy taxes on fossil fuels based on each fuel's energy (BTU) content.
Restricting the concept of greenhouse gas emissions narrowly implies that some potentially cost effective measures to reduce net carbon emissions to the atmosphere might be missed, implying higher than necessary costs for achieving any emissions reduction objective.
Carbon Emissions
The SGM was used to determine the carbon tax rate required in each period to offset carbon emissions above the 1990 emissions level. These tax rates are shown in Figure 9. Tax rates begin at $43/TC in 1995 and rise steadily until 2015 when a peak tax rate of $140/TC is reached.4 All revenue received from the carbon tax is recycled back to households as additions to personal income. Higher carbon tax rates after 2015 are not required due to a fall in the ratio of carbon (in TgC) to real GNP, implying a shift to more energy-efficient methods of production and less carbon-intensive fuel use.
Figure 10 compares the level of carbon emissions under a "business-as-usual" scenario with various "flat" carbon tax and "stabilization" carbon tax options. Assuming a carbon emissions reduction goal to maintain carbon emissions below the 1990 level, the SGM shows that a flat carbon tax of no less than $135/TC is required.
Impacts on GNP and Consumption
Figure 10 also shows that the $135/TC flat tax case results in an exceedance of the emissions reductions required in the years 1990 to 2030 to stabilize emissions. This will result in greater economic losses than the "stabilization" tax case. Figures 11 and 12 show the percent loss in GNP and consumption, respectively, resulting from a $135/TC "flat" carbon tax and "stabilization" carbon tax. As expected, the $135/TC tax case results in much lower growth rates than the "stabilization" case in the beginning years but converges with the "stabilization" case in later years.
It is interesting to note that the peak variable tax rate ($140/TC in the year 2015) is higher than the constant tax rate ($135/TC). This is attributable to the fact that the constant tax rate of $135/TC is built into earlier investment decisions (by way of expected prices) while the tax trajectory is not know in the case of the variable tax, thus not allowing producers to make the most optimal investment decisions. However, the difference between the two tax rates ($135/TC and $140/TC) is small, implying that the additional knowledge of future tax rates makes small differences in investment decisions.
The "flat" and "stabilization" carbon tax cases result in similar impacts, differing only in the size of the impacts; therefore, it would suffice to concentrate on the impacts of one. Since the "stabilization" tax case achieves stabilization of carbon emissions with the least economic cost, any future comparisons will concentrate on this carbon tax case.
Impacts on Primary Energy Consumption
Results from the SGM show a dual effect on primary energy consumption resulting from carbon taxes - a reduction in total primary energy consumption and a shift to lower carbon emitting fuels. Reductions in total primary energy consumption range from 8 percent in 1995 to 19 percent in 2030, while the share of coal consumed (out of all primary fuels) falls from 26 percent to 18 percent in 2030. This would indicate that the carbon taxes had a greater effect on reducing overall energy consumption than shifting energy consumption from higher carbon emitting fuels to lower ones. The Electricity Sector experiences similar impacts with the reduction in total fuel consumption outweighing the shift from more carbon-intensive fuels to less carbon-intensive ones. It is interesting to note, however, that although electricity fuel consumption drops by 28 percent in 2030, total electricity generated only drops by 9 percent. This can be attributed to a shift to less carbon intensive fuels, which are associated with more energy efficient methods of generation.
The variable carbon tax also has effect of making biomass fuels economically attractive in the later years. In the reference case, no biomass fuel production occurs throughout the modeling period 1990 to 2030. With the implementation of carbon taxes, we see both liquid and gaseous fuels from biomass being produced. By 2025, gaseous fuels from biomass make up 17 percent of total gas produced, while liquid fuels from biomass only make up 0.4% of total refined liquids produced.
Impacts on Energy Resources
Further evidence that carbon taxes have a greater impact on energy efficiency than on substitution of less carbon-intensive fuels can be seen in the depletion of energy resources. Instead of seeing a more rapid depletion of less carbon-intensive fuels (e.g., natural gas) than the reference case, production of these fuels falls with the imposition of carbon taxes, leading to less depletion of these resources. In addition, as would be expected, production of the more carbon intensive fuels is lower with the imposition of carbon taxes, leading to less depletion of these resources also.
Ratios
A shift to more energy efficient modes of production in reaction to carbon taxes is also apparent when we look at the ratio of energy use (in EJ) to GNP (in trillion $). Differences from the reference case range from 8.1 percent in 1995 to 18.3 percent in 2030. To a much lesser extent, the fossil fuel use (in EJ) to energy use (in EJ) ratio shows a shift to less carbon intensive fossil fuels with a drop in this ratio of 2.1 percent in 2030 from the reference case, proving that carbon emissions reductions were obtained mainly through energy efficiency measures. Lastly, since population projections do not differ between cases, the fall in the ratio of GNP to population resulting from carbon taxes mirrors the loss in GNP. This loss ranges from 0.4 percent in 1995 to 0.7 percent in 2030.
In this modeling exercise, we impose the general structure of Clinton's energy tax on a policy to stabilize carbon emissions to the 1990 level. That is, the basic rate of 25.7 cents per million BTUs is increased until carbon emissions are stabilized in each period at the 1990 emissions level. (The supplemental oil tax is also increased in the same proportion as the basic rate). In the SGM, the energy tax is imposed on the refined oil, refined gas, coal and uranium sectors, and the hydro electric power subsector beginning in 1995. As in the carbon tax stabilization case, all tax revenue from the energy tax is recycled to the household sector as additions to personal income.
Carbon Emissions
As in the variable carbon tax case, the SGM was used to determine the energy tax rate required in each period to offset carbon emissions above the 1990 emissions level. These tax rates are shown in Figure 13. The energy tax rate trajectory follows a similar path to that of the carbon tax with rates rising steadily from 1995 to 2015 when the peak rate is reached. Energy tax rates begin at $1.03/MBTU in 1995 and rise steadily until 2015 when a peak tax rate of $3.53/MBTU is reached.6
Figure 14 shows total tax revenue from both stabilization cases (carbon tax versus energy tax). As is expected, total tax required to stabilize carbon emissions is lower in each period with a carbon tax policy option. From a tax standpoint, carbon taxes are more effective at reducing carbon emissions since this tax targets the carbon content of fuels.
Figure 15 compares the stabilization carbon tax rates and energy tax rates (converted to $/TC) applied to the coal sector. We see a larger impact from a carbon tax on this sector in the years 1995 to 2015 because this tax has the effect of reducing use of the more carbon-intensive fuels first; notably coal. As coal is driven out in the beginning periods, carbon emission reductions in later years is obtained by reductions in other fuels. An energy tax, on the other hand, has the effect of reducing overall energy use and, due to the supplemental tax on oil, reducing refined oil consumption. Carbon emission reductions are achieved in each period by targeting reductions in the use of each energy source with a emphasis on refined oil. In figure 15 carbon taxes have a greater impact on the coal sector in the early years since reductions in coal use is targeted with this policy option while an energy tax does not specifically target use of this fuel. In the later years, carbon emission reductions are still obtainable from the coal sector with an energy tax since this policy option did not drive out coal use in the beginning years like a carbon tax. Therefore, we see a shift where a energy tax has a larger impact than a carbon tax on the coal sector in the later years.
Impacts on GNP, Consumption, and Investment
Due to the higher total tax imposed on the economy with a stabilization energy tax, the impacts on GNP and consumption are slightly greater with this tax option than with a carbon tax as shown in figure 16. Reductions in GNP from a stabilization energy tax range from 0.6 percent in 1995 to 1.7 percent in 2015 while reductions from a stabilization carbon tax range from 0.4 percent in 1995 to 1.2 percent in 2015. Reductions in consumption follow a similar track with reductions ranging from 0.1 percent in 1995 to 1.1 percent in 2015 with an energy tax while a carbon tax results in reductions of 0.02 percent in 1995 to 0.8 percent in 2015.
The energy tax has a greater impact on investment as compared to the carbon tax case especially in later years. Reductions in 1995 investment are 1.28 percent with an energy tax and 1.02 percent with a carbon tax while reductions in 2020 are 2.22 percent with an energy tax and 0.97 percent with a carbon tax. Since the model allows for shifts in capital inputs only in the long-term (the "putty-clay" assumption), it makes sense that investment under a carbon tax would shift to less carbon-intensive energy production. With the energy tax option, however, all energy sources are targeted (with emphasis on refined oil); therefore, the effects of this tax would be to decrease overall investment in energy production .
Impacts on Primary Energy Consumption
Figure 17 shows the impacts of both stabilization tax options on primary energy consumption. Both taxes result in similar impacts on energy consumption with the energy tax resulting in slightly larger impacts. Since an energy tax effects all sources of energy, we would expect this policy option to result in lower overall energy use when compared to a carbon tax policy option.
With both tax cases, we see a shift in electricity fuel consumption from oil and coal to natural gas and renewables. This is because the percent increase in price of oil and coal is greater than that of gas in both tax cases causing as shift to gas consumption.7 The difference between the two cases is out of which fuel most of the shifting is taking place. Due to the supplemental tax on refined oil with the energy tax, we see most of the decline in electricity fuel consumption coming from refined oil; in fact, refined oil use in the electricity sector is completely phased out by 2000 under an energy tax whereas, under a carbon tax, oil is not phased out until after 2025. With a carbon tax, most of the decline in electricity fuel consumption comes from coal. In either case, the amount of electricity generated is essentially the same implying that each tax option has the same effect on electricity generation.
As with carbon taxes, we see the emergence of gaseous fuels from biomass with an energy tax with gas from biomass making up 14 percent (compared to 17 percent with the carbon tax) of gas production in 2025. No production of liquids from biomass occurs since the energy tax does not distinguish between feedstocks of the refined liquids in its supplemental tax on this sector. (We see a larger amount of the biomass fuels being produced with the carbon tax since these fuels are exempt from any carbon tax; unlike in the energy tax case where all sources of energy are affected).
Impacts on Energy Resources
In both tax cases, we see a faster depletion of natural gas than the reference case. This is due to a shift from oil and coal use in both cases to a more energy-efficient and less carbon-intensive use of energy; namely natural gas. As explained above, this shift is due to the larger percent increases in price of oil and coal than gas in both tax cases.
Ratios
As with a carbon tax, an energy tax results in more energy-efficient and less carbon-intensive energy use. As would be expected, since an energy tax targets all energy sources and does not target those which are more carbon-intensive, we see a slightly faster decline in the ratio of energy to GNP with an energy tax than with a carbon tax. In addition, we see a slightly slower decline in the ratio of carbon to energy with an energy tax than with a carbon tax for the same reason. Lastly, due to the larger reductions in GNP with an energy tax, we see a slightly smaller decrease in the ratio of GNP to population with a energy tax than with a carbon tax.
Future versions of the SGM will further expand on the capabilities of the model. Future expansions of the model include the provision of energy services and traditional biomass fuels, and the addition of a detailed description of the agriculture and health sectors to enable the model to consider the consequences of atmosphere and climate change in detail.
Figure 12 % Reduction in Real Consumption from Reference Case.
Figure 13 Energy Tax for CO2 Stabilization to 1990 Level.
Figure 14 Tax Revenues.
Figure 15 Taxes for CO2 Stabilization to 1990 Level (Energy tax converted to carbon tax based on carbon content of coal.)
Figure 16 % Reduction in GNP from Reference Casse (Base Year Prices).
Figure 17 % Reduction in Primary Energy Consumption from Reference Case.
Edmonds, J., and Barns, D.W., 1991. Factors Affecting the Long-term Cost of Global Fossil Fuel CO2 Emission Reductions. PNL Global Studies Program Working Paper. Washington D.C.
Edmonds, J., and Reilly, J.M., 1985. Global Energy: Assessing the Future. Oxford University Press. New York.
Edmonds, J., Reilly, J.M., Gardner, R.H., and Brenkert, A., 1986. Uncertainty in Future Global Energy Use and Fossil Fuel CO2 Emissions 1975 to 2075. TR036, D03/NBB0081 Dist. Category UC-11. National Technical Information Service. U.S. Department of Commerce. Springfield, Virginia 22161.
Edmonds, J., H.M. Pitcher, D. Barns, R. Baron, and M.A. Wise. 1993. Modeling Future Greenhouse Gas Emissions: The Second Generation Model Description. Proceedings of the United Nations University Conference on "Global Change and Modeling." Tokyo, JAPAN. October, 1991.
Elliott, Douglas C. et al. 1990. "A Technical and Economic Analysis of Direct Biomass Liquefaction," in Energy from Biomass and Wastes XII |. Donald L. Klass ed. Chicago: Institute of Gas Technology.
Fisher-Vanden, K. et al. 1993. Calibration of the Second Generation Model. Pacific Northwest Laboratory. Washington, D.C.
IPCC (Intergovernmental Panel on Climate Change). 1990.
Jorgenson, Dale W. 1990. "Productivity and Economic Growth" in Fifty Years of Economic Measurement, edited by Ernst R. Berndt and Jack E. Triplett. National Bureau of Economic Research, published by the University of Chicago Press.
OECD (Organisation for Economic Co-operation and Development). 1992. International Energy Agency. Electricity Supply in the OECD. Paris, France.
U.S. Department of Agriculture, 1992. Conversations with E. Ball.
U.S. Department of Commerce, Bureau of Economic Analysis, 1987. Fixed Reproducible Tangible Wealth in the United States, 1925-85. Washington D.C., Government Printing Office.
U.S. Department of Commerce, Bureau of Economic Analysis, 1990. "Annual Input-Output Accounts for the U.S. Economy, 1985" in Survey of Current Business. Washington D.C.
U.S. Department of Commerce, Bureau of Economic Analysis, 1991. Benchmark Input-Output Accounts for the U.S. Economy, 1982. Washington D.C., Government Printing Office.
U.S. Department of Commerce, Bureau of Census, 1992. Statistical Abstract of the United States. Washington, D.C., Government Printing Office.
U.S. Department of Energy (DOE), 1991a. Limiting Net Greenhouse Gas Emissions. Report to the Congress. Bradley, R.A., E.C. Watts, and E.R. Williams (eds.). Volumes 1 &; 2. DOE/PE0101. Washington, D.C.
U.S. Department of Energy (DOE), 1991b. Integrated Analysis Supporting the National Energy Strategy: Methodology, Assumptions and Results, Technical Annex 2. DOE/S0086P. National Technical Information Service. Springfield, Virginia.
U.S. Department of Energy, Energy Information Administration (EIA), June 1991.Financial Statistics of Selected Investor-Owned Electric Utilities, 1989. DOE/ EIA0437(89). Washington, D.C.
U.S. Department of Energy, Energy Information Administration (EIA), June 1992. Annual/ Energy Review 1991. DOE/EIA-0384(91). Washington, D.C.
Wan, Edward I. and Malcolm D. Fraser, 1990. "Economic Assessment of Advanced Biomass Gasification Systems," in Energy from Biomass and Wastes XII|. Donald L. Klass ed. Chicago: Institute of Gas Technology.
World Bank, 1990. World Tables. 1989-1990 Edition. Johns Hopkins University Press. Baltimore, New York.
2 We note that the dominant non-energy related emissions of carbon are from ruminant livestock, and are given off in the form of CH4.
3 The concept of a global warming potential (GWP) was developed by the Intergovernmental Panel on Climate Change (IPCC) and elsewhere, to provide a metric by which to compare the various gases. But problems In defining these coefficients have emerged which have been most acute with regard to those gases with the greatest indirect effects on climate change; precisely those gases for which the GWP is most important.
4 Using a 1985 price for crude oil, natural gas, and coal of $4.15,S2.26. and S1.15 per MBTU, respectively (Annual Energy Review, 1991), a $43/TC carbon tax would amount to an increase in price ranging from 22 percent for oil to 93 percent for coal while a $140/TC carbon tax would amount to an increase ranging from 71 percent for oil to 304 percent for coal.
5 The conversion of an energy tax to a carbon tax was done by dividing the $/BTU energy tax by the amount of carbon per BTU in each fuel. The amount of carbon per BTU is highest for coal and lowest for natural gas;therefore, given a basic energy tax rate of $0.257/BTU across all fuels, coal has the lowest $/TC carbon tax rate. Oil has the highest carbon tax rate since it includes a $0.342/BTU supplemental tax.