This document is spread across three files.  This is the first file.


SR/OIAF/98-03
Distribution Category UC-950

Impacts of the Kyoto Protocol
on U.S. Energy Markets
and Economic Activity


October 1998

Energy Information Administration

Office of Integrated Analysis and Forecasting
U.S. Department of Energy
Washington, DC 20585


This report was prepared by the Energy Information Administration, the independent statistical and analytical
agency within the Department of Energy. The information contained herein should be attributed to the Energy
Information Administration and should not be construed as advocating or reflecting any policy position of the
Department of Energy or of any other organization. Service Reports are prepared by the Energy Information
Administration upon special request and are based on assumptions specified by the requester.



Contents


ExecutiveSummary................................................................................. xi


1.ScopeandMethodologyoftheStudy............................................................... 1
Background..................................................................................... 1
MethodologyoftheAnalysis...................................................................... 5
UseofModelsforAnalysis........................................................................ 16


2.SummaryofEnergyMarketResults................................................................ 19
CarbonReductionCases.......................................................................... 19
SensitivityCases................................................................................. 29


3.End-UseEnergyDemand.......................................................................... 33
Background..................................................................................... 33
ResidentialDemand ............................................................................. 34
CommercialDemand ............................................................................ 42
IndustrialDemand............................................................................... 50
TransportationDemand.......................................................................... 59


4.ElectricitySupply ................................................................................ 71
Introduction .................................................................................... 71
TrendsinFuelUseandGeneratingCapacity........................................................ 73
ElectricityPrices................................................................................. 88
SensitivityCases................................................................................. 91


5.FossilFuelSupply................................................................................ 95
NaturalGasIndustry............................................................................. 95
OilIndustry..................................................................................... 103
Coal ........................................................................................... 110


6.AssessmentofEconomicImpacts .................................................................. 119
ObjectivesoftheMacroeconomicAnalysis.......................................................... 119
TheU.S.PermitSystemandInternationalTradingofPermits.......................................... 120
SummaryofMacroeconomicImpacts .............................................................. 120
EstimatingTheUnavoidableImpactontheEconomy................................................. 123
EnergyPricesandtheRoleofMonetaryandFiscalPolicy............................................. 124


7.ComparingCostEstimatesfortheKyotoProtocol.................................................... 137
Introduction .................................................................................... 137
SummaryofComparisons........................................................................ 137
TheÒFive-LabStudyÓ............................................................................ 146


Appendixes

A.ModificationstotheReferenceCase............................................................. 153


B. ResultsfortheCarbonReductionCases.......................................................... 159
C.SummaryComparisonofAnalyses.............................................................. 213
D.LettersfromtheCommitteeonScience.......................................................... 223


Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity v



Tables

ES1. SelectedVariablesintheCarbonReductionCases,1996and2010 ................................... xv
ES2. SelectedVariablesintheCarbonReductionCases,1996and2020 ................................... xvi
ES3. Energy Market Assumptions for the Macroeconomic Analysis of Three Carbon Reduction Cases,
AverageAnnualValues,2008through2012 ...................................................... xxi
ES4. Macroeconomic Impacts in Three Carbon Reduction Cases, Average Annual Values, 2008-2012 ......... xxii
ES5. ProjectedImpactsonGrossDomesticProduct,2005and2010.......................................xxiii
ES6. ProjectedImpactsonGrossDomesticProduct,2005and2020.......................................xxiii
ES7. Projected Losses in Potential and Actual GDP per Capita, Average Annual Values, 2008-2012 ........... xxv


1. Carbon Emissions Factors for Major Energy Fuels and Calculated 1996 Delivered Energy Prices
WithaCarbonPriceof$100perMetricTon ...................................................... 12

2. Summary Comparison: Reference, 1990+24%, 1990+9%, and 1990-3% Cases, 2010 and 2020 ............. 21


3.PrimaryandEnd-UseEnergyConsumptionbySector,1996........................................ 33


4. Change in Projected Average Efficiencies of Newly Purchased Residential Equipment
inCarbonReductionCasesRelativetotheReferenceCase,2010..................................... 38

5. Cost and Efficiency Indexes of Best Available Technologies for Selected Residential Appliances, 2015 .... 40

6. Change in Projected Penetration Rates for Selected Technologies in the Commercial Sector
RelativetotheReferenceCase,2010............................................................. 46

7. Projected Carbon Prices and Average Fuel Prices for the Commercial Sector in Technology Sensitivity
Cases,2010................................................................................... 49

8. Projected Highest Available and Average Efficiencies for Newly Purchased Equipment
intheCommercialSector,2015 ................................................................. 49

9. ProjectedEnergyIntensitiesforIndustrialProcessStepsandEndUses............................... 55


10. Projected Average Transportation Energy Intensities by Mode of Travel, 2010 ........................ 60


11. Projected Penetration of Selected Technologies for Domestic Compact Cars, 2010...................... 63


12. Projected Penetration for Selected Advanced Technologies for Aircraft, 2010.......................... 65


13. ProjectedPenetrationofSelectedTechnologiesforFreightTrucks,2010.............................. 66


14. Projected Fuel Consumption Shares in the Transportation Sector by Fuel and Travel Mode, 2010 ........ 67


15. Projected Alternative-Fuel Vehicle Shares of New Light-Duty Vehicle Sales by Type
intheHighTechnologyCases,2010............................................................. 70

16. Cost and Performance Characteristics of New Fossil, Renewable, and Nuclear Generating Technologies. . 73

17. CarbonEmissionsFromFossilFuelGeneratingTechnologies....................................... 75


18. HypotheticalExamplesofLevelizedPlantCostsatVariousCarbonPrices............................ 76


19. ProjectedU.S.ElectricityGenerationFromRenewableFuels........................................ 80


20. ProjectedU.S.ElectricityGenerationCapacityFromRenewableFuels................................ 81


21. U.S.BiomassResources........................................................................ 85


22. Components of Differential Petroleum Product Prices Relative to the Reference Case, 2010 ............. 108


23. ProjectedNumberofCoalMiningJobsbyRegion,2010............................................ 114


24. CoalIndustryWagesandEmployment,1996..................................................... 115


25. Energy Market Assumptions for the Macroeconomic Analysis of Three Carbon Reduction Cases,
AverageAnnualValues,2008through2012 ...................................................... 121

26. Macroeconomic Impacts in Three Carbon Reduction Cases, Average Annual Values, 2008-2012 ......... 122


27. Projected Losses in Potential and Actual GDP per Capita, Average Annual Values, 2008-2012 ........... 123


28. AverageProjectedAnnualLossesinEconomicOutput,2008-2012................................... 124


29. Projected Economic Impacts of Carbon Reduction Cases Assuming Personal Income Tax Rebate ........ 131


30. Comparison of Results for Reducing Carbon Emissions to 7 Percent Below 1990 Levels
WithoutTrading,Sinks,Offsets,orCleanDevelopmentMechanism................................. 140

31. Comparison of Results for Reducing Carbon Emissions to 7 Percent Below 1990 Levels
WithAnnexITrading,Sinks,andOffsets ........................................................ 141

32. Comparison of Energy Consumption, Gross Domestic Product, and Energy Intensity Results
forEIAandFive-LabStudyAnalyses............................................................ 147

33. Comparison of Carbon Emissions Results for EIA and Five-Lab Study Analyses....................... 147


Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Figures

ES1. ProjectionsofCarbonEmissions,1990-2020....................................................... xiii
ES2. ProjectionsofCarbonPrices, 1996-2020.......................................................... xvii
ES3. Average Projected Carbon Prices and Annual Carbon Emission Reductions, 2008-2010 ................. xvii
ES4. ProjectionsofU.S.ElectricityGeneration,1990-2020............................................... xvii
ES5. Projected Reductions in Carbon Emissions From the Electricity Supply Sector, 1990-3% Case, 1996-2020 . . xvii
ES6. Projected Reductions in Carbon Emissions by End-Use Sector Relative to the Reference Case, 2010 ...... xviii
ES7. Projected Changes in Average Delivered Prices for Energy Fuels in the 1990+9% Case
RelativetotheReferenceCase,1996-2020 ........................................................xviii
ES8. ProjectionsofFuelSharesofTotalU.S.EnergyConsumption,2010.................................. xix
ES9. ProjectionsofU.S.CoalConsumption,1970-2020.................................................. xix
ES10. ProjectionsofU.S.PetroleumConsumption,1970-2020............................................. xix
ES11. ProjectionsofU.S.NaturalGasConsumption,1970-2020........................................... xx
ES12. ProjectionsofU.S.NuclearEnergyConsumption,1970-2020........................................ xx
ES13.ProjectionsofU.S.RenewableEnergyConsumption,1990-2020..................................... xx
ES14. Projected Changes in Consumer Price Index Relative to the Reference Case, 1998-2020 ................. xxii
ES15. Projected Annual Costs of Carbon Reductions to the U.S. Economy, 2008-2012 ........................ xxiii
ES16. Projected Dollar Losses in Potential GDP Relative to the Reference Case, 1998-2020 .................... xxiv
ES17. Projected Changes in Potential and Actual GDP in the 1990+9% Case Relative to the Reference Case
UnderDifferentFiscalPolicies,1998-2020........................................................ xxiv
ES18.ProjectedAnnualGrowthRatesinPotentialandActualGDP,2005-2010............................. xxv
ES19.ProjectedAnnualGrowthRatesinPotentialandActualGDP,2005-2020............................. xxv
ES20. Projected Carbon Prices in the 1990+9% High and Low Economic Growth and
HighandLowTechnologySensitivityCases,2010................................................. xxvi


1. ProjectionsofCarbonEmissions,1990-2020....................................................... 19


2.ProjectionsofCarbonPrices,1996-2020.......................................................... 20


3. Average Annual Carbon Emission Reductions and Projected Carbon Prices, 2008-2012 ................. 22


4. AverageDeliveredPricesforEnergyFuelsinthe1990+24%Case,1996-2020.......................... 23


5. AverageDeliveredPricesforEnergyFuelsinthe1990+9%Case,1996-2020........................... 23


6. AverageDeliveredPricesforEnergyFuelsinthe1990-3%Case,1996-2020 ........................... 23


7. Projected Changes in Average Delivered Prices for Energy Fuels in the 1990+9% Case
RelativetotheReferenceCase,1996-2020 ........................................................ 23

8. ProjectionsofFuelSharesofTotalU.S.EnergyConsumption,2010.................................. 24


9. ProjectionsofU.S.CoalConsumption,1970-2020.................................................. 24


10. ProjectionsofU.S.NaturalGasConsumption,1970-2020........................................... 24


11. ProjectionsofU.S.PetroleumConsumption,1970-2020............................................. 25


12. ProjectionsofU.S.NuclearEnergyConsumption,1970-2020........................................ 25


13.ProjectionsofU.S.RenewableEnergyConsumption,1990-2020..................................... 25


14. ProjectionsofU.S.ElectricityGeneration,1990-2020............................................... 26


15. Projections of U.S. Carbon Emissions per Unit of Primary Energy Consumption, 1990-2020 ............. 26


16. Projected Reductions in Carbon Emissions by End-Use Sector Relative to the Reference Case, 2010 ...... 27

17. ProjectionsofU.S.IndustrialEnergyIntensity,1996-2020........................................... 27


18. ProjectionsofU.S.Light-DutyVehicleTravel,1996-2020 ........................................... 27


19. Projections of Average Fuel Efficiency for the Light-Duty Vehicle Fleet, 1996-2020 ..................... 28


20. ProjectionsofU.S.MotorGasolineConsumption,1996-2020........................................ 28


21. Projected Fuel Use for Electricity Generation by Fuel in the 1990+24% Case, 1996-2020 ................. 29


22. Projected Fuel Use for Electricity Generation by Fuel in the 1990+9% Case, 1996-2020 .................. 29


23. Projected Fuel Use for Electricity Generation by Fuel in the 1990-3% Case, 1996-2020................... 29


24. Projected Carbon Prices in the 1990+9% High and Low Economic Growth and
HighandLowTechnologySensitivityCases,2010................................................. 30

25. ProjectionsofPrimaryEnergyConsumption,1990-2020............................................ 33


26. IndexofResidentialSectorDeliveredEnergyConsumption,1970-2020............................... 35


27. IndexofResidentialSectorDeliveredEnergyIntensity,1970-2020................................... 36


28. ResidentialSectorCarbonEmissions,1990,1996,and2010 ......................................... 36


29. Delivered Energy Consumption in the Residential Sector by Major Fuel, 1970, 1980, 1996, and 2010 ...... 37

30. ResidentialSectorEnergyUseperHousehold,1996 ............................................... 37


31. Average Projected Annual Growth in Residential Sector Energy Consumption by End Use, 1996-2010 . . . 37

32. IndexofResidentialSectorEnergyPrices,1970,1980,1996,and2010................................. 38


Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity vii



Figures (Continued)

33. ProjectedStocksofGround-SourceHeatPumps,1995-2020......................................... 40


34. AverageResidentialSectorEnergyPrices,1995-2020............................................... 40


35. ProjectedEnergyExpendituresintheResidentialSector,1995-2020.................................. 41


36. Changes From Reference Case Projections of Energy Intensity for Residential Water Heating
inThreeSensitivityCases,1995-2020 ............................................................ 42

37. Changes From Reference Case Projections of Residential Energy Consumption in Three Sensitivity Cases,
1995-2020.................................................................................... 42

38. IndexofCommercialSectorDeliveredEnergyConsumption,1970-2010.............................. 45


39. CommercialSectorCarbonEmissions,1990,1996,and2010......................................... 45


40. Real Prices for Delivered Energy in the Commercial Sector by Fuel, 1970, 1980, 1996, and 2010 .......... 45


41. IndexofDeliveredEnergyIntensityintheCommercialSector,1970-2020............................. 46


42. Delivered Energy Use and Electricity-Related Losses in the Commercial Sector,
1970,1980,1996,and2010...................................................................... 46

43. Projected Fuel Expenditures in the Commercial Sector in Low and High Technology Cases, 1996-2020 . . . 49

44. IndexofIndustrialSectorEnergyPrices,2000-2020................................................ 51


45. Index of Delivered Energy Consumption in the Industrial Sector, 1970-2020 .......................... 52


46. IndustrialSectorCarbonEmissions,1990,1996,and2010........................................... 53


47. IndustrialSectorEnergyConsumptionbyFuel,1970,1980,1996,and2010............................ 53


48. ProjectedEnergyIntensityintheIndustrialSector,1995-2020....................................... 53


49. ProjectedChangeinIndustrialSectorEnergyIntensity,1996-2010................................... 54


50. Structural and Efficiency/Other Effects on Industrial Energy Intensity, 1980-1985, 1980-1996,
and1996-2010................................................................................ 54

51. Change From Projected Reference Case Energy Expenditures in the Industrial Sector
forAlternativeCarbonReductionCases,2010 .................................................... 54

52. Natural-Gas-Fired Cogeneration and Biomass Consumption in the Industrial Sector
inAlternativeCarbonReductionCases,2010..................................................... 57

53.Light-DutyVehicleEnergyIntensity,1996and2010............................................... 60


54. CarbonEmissionsintheTransportationSector,1990,1996,and2010................................. 60


55. FuelConsumptionintheTransportationSector,1970-2020 ......................................... 60


56. Light-DutyVehicleTravel,1970-2020............................................................ 61


57. ProjectedNewCarandLightTruckFuelEconomy,2010........................................... 62


58. ProjectedSharesofAutomobileSalesbySizeClass,2010........................................... 63


59. Projected Reductions From Reference Case Projections of Car and Light Truck Horsepower
intheCarbonReductionCases,2010and2020.................................................... 64

60. Projected Fuel Consumption in the Transportation Sector by Mode in the Reference Case, 2010 ......... 64


61.ProjectedFuelConsumptionintheTransportationSectorbyFuelType,2010......................... 64


62. ProjectedNewandStockAircraftFuelEfficiency,2010 ............................................ 65


63. ProjectedNewandStockFreightTruckFuelEfficiency,2010 ....................................... 66


64. Projected Reductions From Reference Case Projections of Transportation Sector Fuel Consumption
inHighandLowTechnologySensitivityCases,2010 .............................................. 69

65. ElectricityGenerationbyFuelinthe ReferenceCase,1949-2020..................................... 71


66. Projections of Electricity Sales, Carbon Emissions, Fossil Fuel Use,
andFossil-FiredGeneration,1997-2020 .......................................................... 72

67. Projections of Carbon Emissions From the Electricity Supply Sector, 1996-2020 ........................ 74


68. Projected Reductions in Carbon Emissions From the Electricity Supply Sector, 1990-3% Case, 1996-2020 . . 74

69. ElectricityGenerationbyFuel,1990+9%Case,1949-2020........................................... 74


70. ElectricityGenerationbyFuel,2010 ............................................................. 74


71. ProjectionsofCoal-FiredElectricityGeneration,2000-2020 ......................................... 75


72. OperatingCostsforCoal-FiredElectricityGenerationPlants,1981-1995.............................. 76


73. ProjectionsofCoal-FiredGeneratingCapacity,2000-2020 .......................................... 76


74. ElectricityGenerationCapacitybyFuel,2010..................................................... 76


75. ProjectionsofNatural-Gas-FiredElectricityGeneration,2000-2020 .................................. 77


76. Natural-Gas-FiredElectricityGeneration,1990-3%Case,1996-2020.................................. 77


77. Projections of Natural-Gas-Fired Electricity Generation Capacity, 2010 ............................... 77


78. Projections of Nonhydroelectric Renewable Electricity Generation, 2000-2020 ......................... 79


79. Projections of Wind-Powered Electricity Generation Capacity, 2000-2020 ............................. 80


80. Projected Shares of Most Economical Wind Resources Developed by Region, 1990-7% Case, 1996-2020 . . . 82

viii Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Figures (Continued)

81. Estimated Biomass Resource Availability and Projected Generating Capacity in 2020 by Region ......... 83


82. ProjectionsofNuclearElectricityGeneration,2000-2020............................................ 88


83. ProjectionsofNuclearElectricityGenerationCapacity,2000-2020 ................................... 88


84. Projected Changes in Electricity Sales Relative to the Reference Case, 2000-2020 ....................... 88


85. ProjectionsofElectricityPrices,1996-2020........................................................ 89


86. Projected Electricity Prices in Regulated and Competitive Electricity Markets, 2000-2020 ............... 90


87. Projected Carbon Prices in Regulated and Competitive Electricity Markets, 2000-2020.................. 90


88. Projected Percentage of Time for Different Plant Types Setting National Marginal Electricity Prices,
2010and2020 ................................................................................ 90

89. Projected Percentage of Time for Interregional Trade Setting Marginal Electricity Prices, 2020........... 91


90. Projections of Average Heat Rates for Natural-Gas-Fired Power Plants in High and Low
TechnologyCases,1996-2020................................................................... 92

91. ProjectedElectricityPricesinHighandLowTechnologyCases,1996-2020............................ 92


92. Projections of Nuclear Generating Capacity in the 1990-3% Nuclear Sensitivity Case, 2000-2020 ......... 92


93. NaturalGasConsumption,1996-2020............................................................ 96


94. IncreasesinNaturalGasProduction,1983-1984and2005-2006...................................... 97


95. IndexofNaturalGasReserve-to-ProductionRatios,1990-2020...................................... 98


96. NaturalGasWellheadPrices,1970-2020.......................................................... 102


97. DeliveredNaturalGasPricesintheResidentialSector,1970-2020 ................................... 102


98.PetroleumConsumption,1970-2020............................................................. 104


99. Lower48CrudeOilReserveAdditions,1990-2020................................................. 104


100. Net Expenditures for Imported Crude Oil and Petroleum Products, 1974-2020 ........................ 105


101. ConsumptionofEthanolintheTransportationSector,1992-2020.................................... 106


102. GasolinePricesintheTransportationSector,1990-2020 ............................................ 107


103. RetailGasolinePricesbyRegion,AverageofAllGrades,1996and2010.............................. 108


104. ProjectedWholesaleGasolineMargins,1996-2020................................................. 109


105. U.S.CoalProduction,1970-2020 ................................................................ 111


106. WesternShareofU.S.CoalProduction,1990-2020................................................. 112


107. AverageU.S.MinemouthCoalPrices,1970-2020.................................................. 113


108.CoalPricestoElectricityGenerators,1970-2020................................................... 113


109. CoalMineEmployment,1970-2020.............................................................. 114


110. Projected Annual Costs of Carbon Reductions to the U.S. Economy, 2008-2012........................ 122


111.ProjectedAnnualGrowthRatesinPotentialandActualGDP,2005-2010............................. 122


112.ProjectedAnnualGrowthRatesinPotentialandActualGDP,2005-2020............................. 122


113. Projected Dollar Losses in Potential GDP Relative to the Reference Case, 1998-2020 .................... 123


114. AverageCarbonReductionsandProjectedCarbonPrices,2008-2012................................. 123


115. Comparison of Average U.S. Economic Losses Projected by the NEMS and DRI Models, 2008-2012 ...... 124

116. Projected Changes in Wholesale Price Index for Fuel and Power
RelativetotheReferenceCase,1998-2020 ........................................................ 125

117. Projected Changes in Producer Price Index Relative to the Reference Case, 1998-2020 .................. 125


118. Projected Changes in Consumer Price Index Relative to the Reference Case, 1998-2020 ................. 126


119. Total Projected U.S. Payments for Domestic and International Carbon Emissions Permits, 1998-2020 ..... 126

120. Projected Destinations of Funds Paid for Carbon Emissions Permits, 2010 and 2020 .................... 127


121. Projected Changes in U.S. Inflation Rate Relative to the Reference Case, 1998-2020..................... 128


122. Projected Changes in U.S. Unemployment Rate Relative to the Reference Case, 1998-2020 .............. 128


123. Projected Changes in U.S. Federal Funds Rate Relative to the Reference Case, 1998-2020................ 128


124. Projected Changes in Potential and Actual U.S. Gross Domestic Product in the 1990+9% Case
RelativetotheReferenceCase,1998-2020 ........................................................ 129

125. Projected Changes in Potential and Actual U.S. Gross Domestic Product in the 1990-3% Case
RelativetotheReferenceCase,1998-2020 ........................................................ 130

126. Projected Changes in Potential and Actual U.S. Gross Domestic Product in the 1990+24% Case
RelativetotheReferenceCase,1998-2020 ........................................................ 130

127. Projected Changes in Real Consumption in the U.S. Economy Relative to the Reference Case, 1998-2020. . 130

128. Projected Changes in Real Investment in the U.S. Economy Relative to the Reference Case, 1998-2020 .... 130

129. ConsumptionandInvestmentGrowthRates ..................................................... 132


130. Projected Changes in U.S. Federal Funds Rate in the 1990-3% Case Relative to the Reference Case
UnderDifferentFiscalPolicies,1998-2020........................................................ 133

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity ix



Figures (Continued)

131. Projected Changes in U.S. Federal Funds Rate in the 1990+9% Case Relative to the Reference Case
UnderDifferentFiscalPolicies,1998-2020........................................................ 133

132. Projected Changes in U.S. Federal Funds Rate in the 1990+24% Case Relative to the Reference Case
UnderDifferentFiscalPolicies,1998-2020........................................................ 133

133. Projected Changes in Potential and Actual U.S. Gross Domestic Product in the 1990+9% Case
RelativetotheReferenceCaseUnderDifferentFiscalPolicies,1998-2020............................. 133

134. Projected Changes in Real Consumption in the U.S. Economy Relative to the Reference Case, 1998-2020,
AssumingaSocialSecurityTaxRebate .......................................................... 134

135. Projected Changes in Real Investment in the U.S. Economy Relative to the Reference Case, 1998-2020,
AssumingaSocialSecurityTaxRebate .......................................................... 134

136. Projected Sectoral Growth Rates in Real Economic Output in the 1990+9% Case, 2005-2010 ............. 135


137. Projected Sectoral Growth Rates in Real Economic Output in the 1990-3% Case, 2005-2010.............. 136


138. Projected Sectoral Growth Rates in Real Economic Output in the 1990+24% Case, 2005-2010 ............ 136


Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Preface


From December 1 through 11, 1997, more than 160
nations met in Kyoto, Japan, to negotiate binding limitations on greenhouse gases for the developed nations,
pursuant to the objectives of the Framework Convention
on Climate Change of 1992. The outcome of the meeting
was the Kyoto Protocol, in which the developed nations
agreed to limit their greenhouse gas emissions, relative
to the levels emitted in 1990. The United States agreed to
reduce emissions from 1990 levels by 7 percent during
the period 2008 to 2012.

The analysis in this report was undertaken at the request
of the Committee on Science of the U.S. House of Representatives. In its request, the Committee asked the
Energy Information Administration (EIA) to analyze the
Kyoto Protocol, Òfocusing on U.S. energy use and prices
and the economy in the 2008-2012 time frame,Ó as noted
in the first letter in Appendix D. The Committee specified that EIA consider several cases for energy-related
carbon reductions in its analysis, with sensitivities
evaluating some key uncertainties: U.S. economic
growth, the cost and performance of energy-using technologies, and the possible construction of new nuclear
power plants.

The energy projections and analysis in this report were
conducted using the National Energy Modeling System
(NEMS), an energy-economy model of U.S. energy
markets designed, developed, and maintained by EIA.
NEMS is used each year to provide the projections in the
Annual Energy Outlook (AEO). In its second letter, in
Appendix D, the Committee requested that the analysis
use the same general methodologies and assumptions
underlying the Annual Energy Outlook 1998 (AEO98),
published in December 1997; however, some minor
modifications were made to allow greater flexibility in
NEMS in response to higher energy prices and to
incorporate some methodologies that were formerly
represented offline. These differences are outlined in
Appendix A. The macroeconomic analysis used the Data
Resources, Inc. (DRI) Macroeconomic Model of the U.S.
Economy, which is also used for the economic analysis
in the AEO.

Chapter 1 of this report provides background discussion
of the Kyoto Protocol and the framework and methodology of the analysis. Chapter 2 summarizes the energy
market results from the various carbon reduction cases.
Chapters 3, 4, and 5 analyze in more detail the issues and

results for the end-use demand sectors, the electricity
generation sector, and the fossil fuel supply markets,
respectively. Chapter 6 provides the results of EIA's
analysis of the macroeconomic impacts of carbon reduction under different monetary and fiscal policy assumptions. Chapter 7 compares the results of this study with
those from other studies of the costs of carbon reduction,
with accompanying tables in Appendix C. Appendix B
includes the detailed energy market results from the
carbon reduction cases.

Within its Independent Expert Review Program, EIA
arranged for leading experts in the fields of energy and
economic analysis to review earlier versions of this
analysis and provide comment. The assistance of the following reviewers in preparing the report is gratefully
acknowledged:

Joseph Boyer
Yale University


Lorna Greening
Consultant to Hagler Bailly Services, Inc.

William Hogan
Harvard University


William Nordhaus
Yale University


Dallas Burtraw
Resources for the Future


Richard Newell
Resources for the Future


William Pizer
Resources for the Future


Michael Toman
Resources for the Future


John Weyant
Stanford University Energy Modeling Forum.

The legislation that established EIA in 1977 vested the
organization with an element of statutory independence. EIA does not take positions on policy questions. It
is the responsibility of EIA to provide timely, high-
quality information and to perform objective, credible
analyses in support of the deliberations of both public
and private decisionmakers. This report does not purport to represent the official position of the U.S. Department of Energy or the Administration.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Other EIA reports on the topic of greenhouse gases
include the following annual reports:

¥ Annual Energy Outlook 1998, published in December
1997, with projections of domestic energy carbon
emissions through 2020
¥ International Energy Outlook 1998, published in April
1998, with projections of international energy carbon
emissions through 2020
¥ Emissions of Greenhouse Gases in the United States 1996,
published in October 1997, with an inventory of all
domestic greenhouse gas emissions
¥ Mitigating Greenhouse Gas Emissions: Voluntary
Reporting, published in October 1997, reporting voluntary actions in 1995 to reduce greenhouse gases in
the United States
¥ Greenhouse Gases, Global Climate Change, and Energy,
an information brochure on greenhouse gases.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Executive Summary


Greenhouse Gases
and the Kyoto Protocol


Over the past several decades, rising concentrations of
greenhouse gases have been detected in the EarthÕs
atmosphere. It has been hypothesized that the continued
accumulation of greenhouse gases could lead to an
increase in the average temperature of the EarthÕs surface and cause a variety of changes in the global climate,
sea level, agricultural patterns, and ecosystems that
could be, on net, detrimental.

The Intergovernmental Panel on Climate Change (IPCC)
was established by the World Meteorological Organization and the United Nations Environment Programme
in 1988 to assess the available scientific, technical, and
socioeconomic information in the field of climate
change. The most recent report of the IPCC concluded
that: ÒOur ability to quantify the human influence on
global climate is currently limited because the expected
signal is still emerging from the noise of natural variability, and because there are uncertainties in key factors.
These include the magnitudes and patterns of long-term
variability and the time-evolving pattern of forcing by,
and response to, changes in concentrations of greenhouse gases and aerosols, and land surface changes.
Nevertheless, the balance of evidence suggests that
there is a discernable human influence on global climate.Ó1

The text of the Framework Convention on Climate
Change was adopted at the United Nations on May 9,
1992, and opened for signature at Rio de Janeiro on June

4. The objective of the Framework Convention was to
Ò. . . achieve... stabilization of the greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the
climate system.Ó The signatories agreed to formulate
programs to mitigate climate change, and the developed
country signatories agreed to adopt national policies to
return anthropogenic emissions of greenhouse gases to
their 1990 levels.
The first and second Conference of the Parties in 1995
and 1996 agreed to address the issue of greenhouse gas

emissions for the period beyond 2000, and to negotiate
quantified emission limitations and reductions for the
third Conference of the Parties. On December 1 through
11, 1997, representatives from more than 160 countries
met in Kyoto, Japan, to negotiate binding limits on
greenhouse gas emissions for developed nations. The
resulting Kyoto Protocol established emissions targets
for each of the participating developed countriesÑthe
Annex I countries2Ñrelative to their 1990 emissions levels. The targets range from an 8-percent reduction for the
European Union (or its individual member states) to a
10-percent increase allowed for Iceland. The target for
the United States is 7 percent below 1990 levels.

Although atmospheric concentrations of greenhouse
gases are thought to have the potential to affect the
global climate, the Protocol establishes targets in terms
of annual emissions. Non-Annex I countries have no targets under the Protocol, but the Protocol reaffirms the
commitments of the Framework Convention by all parties to formulate and implement climate change mitigation and adaptation programs.

Should the Protocol enter into force, the emissions targets for the developed countries would have to be
achieved on average over the commitment period 2008
to 2012. The greenhouse gases covered by the Protocol
are carbon dioxide, methane, nitrous oxide, hydro-
fluorocarbons, perfluorocarbons, and sulfur hexafluoride. The aggregate target is based on the carbon dioxide
equivalent of each of the greenhouse gases. For the three
synthetic greenhouse gases, countries have the option of
using 1995 as the base year.

Several provisions of the Protocol allow for some flexibility in meeting the emissions targets. Net changes in
emissions by direct anthropogenic land-use changes
and forestry activities may be used to meet the commitment, but they are limited to afforestation, reforestation,
and deforestation since 1990. Emissions trading among
the Annex I countries is also allowed. No rules for trading were established, however, and the Conference
of the Parties is required to establish principles, rules,
and guidelines for trading at a future date. According to estimates presented by the Energy Information

1Intergovernmental Panel on Climate Change, Climate Change 1995: The Science of Climate Change (Cambridge, UK: Cambridge University
Press, 1996).

2Australia, Austria, Belgium, Bulgaria, Canada, Croatia, Czech Republic, Denmark, Estonia, European Community, Finland, France,
Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Latvia, Liechtenstein, Lithuania, Luxembourg, Monaco, Netherlands, New
Zealand, Norway, Poland, Portugal, Romania, Russian Federation, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine, United
Kingdom of Great Britain and Northern Ireland, and United States of America. Turkey and Belarus are Annex I nations that have not ratified
the Convention and did not commit to quantifiable emissions targets.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Administration (EIA) in its International Energy Outlook
1998,3 there may be 165 million metric tons of carbon
permits available from the Annex I countries of the
former Soviet Union in 2010. Greenhouse gas emissions
for those countries as a group are expected to be 165 million metric tons below 1990 levels in 2010 as a result of
the economic decline that has occurred in the region
during the 1990s. Additional carbon permits may also be
available, depending on the Òcarbon priceÓ that is established in international trading.

Joint implementation projects are permitted among the
Annex I countries, allowing a nation to take emissions
credits for projects that reduce emissions or enhance
emissions-absorbing sinks, such as forests and other
vegetation, in other Annex I countries. The Protocol also
establishes a Clean Development Mechanism (CDM),
under which Annex I countries can take credits for projects that reduce emissions in non-Annex I countries. In
addition, any group of Annex I countries may create a
bubble or umbrella to meet the total commitment of all
the member nations. In a bubble, countries would agree
to meet their total commitment jointly by allocating a
share to each member. In an umbrella arrangement, the
total reduction of all member nations would be met collectively through the trading of emissions rights. There
is potential interest in the United States in entering into
an umbrella trading arrangement with Annex I countries outside the European Union.

In 1990, total greenhouse gas emissions in the United
States were 1,618 million metric tons carbon equivalent.4
Of this total, 1,346 million metric tons, or 83 percent, consisted of carbon emissions from the combustion of
energy fuels. By 1996, total U.S. greenhouse gas emissions had risen to 1,753 million metric tons carbon
equivalent, including 1,463 million metric tons of carbon
emissions from energy combustion. EIAÕs Annual Energy
Outlook 1998 (AEO98)5 projects that energy-related carbon emissions will reach 1,803 million metric tons in
2010, 34 percent above the 1990 level. Because energy-
related carbon emissions constitute such a large percentage of the NationÕs total greenhouse gas emissions, any
action or policy to reduce emissions will have significant
implications for U.S. energy markets.

At the request of the U.S. House of Representatives
Committee on Science, EIA performed an analysis of the
Kyoto Protocol, focusing on the potential impacts of
the Protocol on U.S. energy prices, energy use, and the
economy in the 2008 to 2012 time frame. The request

specified that the analysis use the same methodologies
and assumptions employed in the AEO98, with no
changes in assumptions about policy, regulatory
actions, or funding for energy and environmental programs.

Methodology

The international provisions of the Kyoto Protocol,
including international emissions trading between
Annex I countries, joint implementation projects, and
the CDM, may reduce the cost of compliance in the
United States. Guidelines for those provisions, however,
remain to be resolved at future negotiating meetings,
and rules and guidelines for the accounting of emissions
and sinks from activities related to agriculture, land use,
and forestry activities must be developed. The specific
guidelines may have a significant impact on the level of
reductions from other sources that a country must
undertake. Reductions in the other greenhouse gases
may also offset the reductions required from carbon
dioxide. A fact sheet issued by the U.S. Department of
State on January 15, 1998, estimated that the method of
accounting for sinks and the flexibility to use 1995 as the
base year for the synthetic greenhouse gases may reduce
the target to 3 percent below 1990 levels.6 A similar
estimate was cited by Dr. Janet Yellen, Chair, Council of
Economic Advisers, in her testimony before the House
Committee on Commerce, Energy and Power Subcommittee, on March 4, 1998.7

Because the exact rules that would govern the final
implementation of the Protocol are not known with certainty, the specific reduction in energy-related emissions
cannot be established. This analysis includes cases that
assume a range of reductions in energy-related carbon
emissions in the United States. Each case was analyzed
to estimate the energy and economic impacts of achieving an assumed level of reductions.

A reference case and six carbon emissions reduction
cases were examined in this report. The cases are
defined as follows:

¥ Reference Case (33 Percent Above 1990 Levels).
This case represents the reference projections of
energy markets and carbon emissions without any
enforced reductions and is presented as a baseline
for comparisons of the energy market impacts in the
reduction cases. Although this reference case is
3Energy Information Administration, International Energy Outlook 1998, DOE/EIA-0484(98) (Washington, DC, April 1998).
4Energy Information Administration, Emissions of Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington, DC,

October 1997).
5Energy Information Administration, Annual Energy Outlook 1998, DOE/EIA-0383(98) (Washington, DC, December 1997).
6See web site www.state.gov/www/global/oes/fs_kyoto_climate_980115.html.
7See web site www.house.gov/commerce/database.htm.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



based on the reference case from AEO98, there are
small differences between this case and AEO98,in
order to permit additional flexibility in response to
higher energy prices or to include certain analyses
previously done offline directly within the modeling
framework, such as nuclear plant life extension and
generating plant retirements. Also, some assumptions were modified to reflect more recent assessments of technological improvements and costs. As a
result of these modifications, the projection of
energy-related carbon emissions in 2010 is slightly
reduced from the AEO98 reference case level of 1,803
million metric tons to 1,791 million metric tons.

¥ 24 Percent Above 1990 Levels (1990+24%). This case
assumes that carbon emissions can increase to an
average of 1,670 million metric tons between 2008
and 2012, 24 percent above the 1990 levels. Compared to the average emissions in the reference case,
carbon emissions are reduced by an average of 122
million metric tons each year during the commitment period.
¥ 14 Percent Above 1990 Levels (1990+14%). This case
assumes that carbon emissions average 1,539
between 2008 and 2012, approximately at the level
estimated for 1998 in AEO98, 1,533 million metric
tons. This target is 14 percent above 1990 levels and
represents an average annual reduction of 253 million metric tons from the reference case.
¥ 9 Percent Above 1990 Levels (1990+9%). This case
assumes that energy-related carbon emissions can
increase to an average of 1,467 million metric tons
between 2008 and 2012, 9 percent above 1990 levels,
an average annual reduction of 325 million metric
tons from the reference case projections.
¥ Stabilization at 1990 Levels (1990). This case
assumes that carbon emissions reach an average of
1,345 million metric tons during the commitment period of 2008 through 2012, stabilizing approximately
at the 1990 level of 1,346 million metric tons. This is
an average annual reduction of 447 million metric
tons from the reference case.
¥ 3 Percent Below 1990 Levels (1990-3%). This case
assumes that energy-related carbon emissions are
reduced to an average of 1,307 million metric tons
between 2008 and 2012, an average annual reduction
of 485 million metric tons from the reference case
projections.
¥ 7 Percent Below 1990 Levels (1990-7%). In this case,
energy-related carbon emissions are reduced from
the level of 1,346 million metric tons in 1990 to an
average of 1,250 million metric tons in the commitment period, 2008 to 2012. Compared to the reference case, this is an average annual reduction of
542 million metric tons of energy-related carbon
emissions during that period. This case essentially
assumes that the 7-percent target in the Kyoto Protocol must be met entirely by reducing energy-related
carbon emissions, with no net offsets from sinks,
other greenhouse gases, or international activities.

In each of the carbon reduction cases, the target is
achieved on average for each of the years in the first
commitment period, 2008 through 2012 (Figure ES1).
Because the Protocol does not specify any targets
beyond the first commitment period, the target is
assumed to hold constant from 2013 through 2020, the
end of the forecast horizon (although more or less stringent requirements may be set by future Conferences of
the Parties). The target is assumed to be phased in over a
3-year period, beginning in 2005, because the Protocol
indicates that demonstrable progress toward reducing
emissions must be shown by 2005. The phase-in allows
energy markets to begin adjustments to meet the targets
in the absence of complete foresight; however, a longer
or more delayed phase-in could lower the adjustment
costsÑan option that is not considered here. In this
analysis, some carbon reductions are expected to occur
before 2005 as the result of capacity expansion decisions
by electricity generators that incorporate their expectations of future increases in energy prices.

There are three ways to reduce energy-related carbon
emissions: reducing the demand for energy services,
adopting more energy-efficient equipment, and switch-
ing to less carbon-intensive or noncarbon fuels. To
reduce emissions, a carbon price is applied to the cost of
energy. The carbon price is applied to each of the energy
199019952000200520102015202005001,0001,5002,000MillionMetricTons1990+24%
1990-3%
Reference1990+9%
1990+14%
19901990-7%
HistoryProjectionsFigure ES1. Projections of Carbon Emissions,
1990-2020
Sources: History: Energy Information Administration, Emissions of
Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington,
DC, October 1997). Projections: Office of Integrated Analysis and Forecasting,
National Energy Modeling System runs KYBASE.D080398A, FD24ABV
.D080398B, FD1998.D080398B, FD09ABV.D080398B, FD1990.D080398B,
FD03BLW.D080398B, and FD07BLW.D080398B.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



fuels relative to its carbon content at its point of consumption. Electricity does not directly receive a carbon
fee; however, the fossil fuels used for generation receive
the fee, and this cost, as well as the increased cost of
investment in generation plants, is reflected in the delivered price of electricity. In practice, these carbon prices
could be imposed through a carbon emissions permit
system.

In this analysis, the carbon prices represent the marginal
cost of reducing carbon emissions to the specified level,
reflecting the price the United States would be willing to
pay in order to purchase carbon permits from other
countries or to induce carbon reductions in other countries. In the absence of a complete analysis of trade and
other flexible mechanisms to reduce carbon emissions
internationally, the projected carbon prices do not necessarily represent the international market-clearing price
of carbon permits or the price at which other countries
would be willing to offer permits.

The projections in AEO98 and in this analysis were
developed using the National Energy Modeling System
(NEMS), an energy-economy modeling system of U.S.
energy markets, which is designed, implemented, and
maintained by EIA.8 The production, imports, conversion, consumption, and prices of energy are projected
for each year through 2020, subject to assumptions on
macroeconomic and financial factors, world energy
markets, resource availability and costs, behavioral and
technological choice criteria, costs and performance
characteristics of energy technologies, and demographics. NEMS is a fully integrated framework, capturing the
interactions of energy supply, demand, and prices
across all fuels and all sectors of U.S. energy markets.
NEMS provides annual projections, allowing the representation of the transitional effects of proposed energy
policy and regulation.

NEMS includes a detailed representation of capital stock
vintaging and technology characteristics, capturing the
most significant factors that influence the turnover of
energy-using and producing equipment and the choice
of new technologies. The residential, commercial, transportation, electricity generation, and refining sectors of
NEMS include explicit treatments of individual known
technologies and their characteristics, such as initial
cost, operating cost, date of commercial availability, efficiency, and other characteristics specific to the sector.
Unknown technologies are not likely to be developed in
time to achieve significant market penetration within
the time frame of this analysis. Higher energy prices, as a
result of carbon prices, for example, do not alter the
characteristics or availability of energy-using technologies. However, higher prices induce more rapid adoption of more efficient or advanced technologies, because

consumers would have more incentive to purchase
them.

In addition, for new generating technologies, the electricity sector accounts for technological optimism in the
capital costs of first-of-a-kind plants and for a decline in
the costs as experience with the technologies is gained
both domestically and internationally. In each of these
sectors, equipment choices are made for individual technologies as new equipment is needed to meet growing
demand for energy services or to replace retired equipment. In the other sectorsÑindustrial, oil and gas supply, and coal supplyÑthe treatment of technologies is
somewhat more limited due to limitations on the availability of data for individual technologies; however,
technology progress is represented by efficiency
improvements in the industrial sector, technological
progress in oil and gas exploration and production
activities, and productivity improvements in coal production.

Carbon Reduction Cases
Carbon Prices

In 2010, the carbon prices projected to be necessary to
achieve the carbon emissions reduction targets range
from $67 per metric ton (1996 dollars) in the 1990+24%
case to $348 per metric ton in the 1990-7% case (Table
ES1 and Figure ES2). In the 1990+24% case, carbon prices
generally increase from 2005 through 2020 (Table ES2
and Figure ES2). In the 1990+14% and 1990+9% cases,
the carbon prices increase through 2013 and then
essentially flatten.

In the three other carbon reduction cases, the carbon
price escalates more rapidly in order to achieve the more
stringent carbon reductions in the commitment period.
The carbon price then declines as cumulative investments in more energy-efficient and lower-carbon equipment, particularly in the electricity generation sector,
reduce the marginal cost of compliance in the later years
of the forecast. These investments reduce the demand
for carbon permits over an extended period of time,
offsetting growth in energy demand and moderating the carbon prices. Figure ES3 shows the average
carbon prices required to achieve the average carbon
reductions.

Sectoral Impacts

As a result of the carbon prices and higher delivered
energy prices, the overall intensity of energy use
declines in the carbon reduction cases. Energy intensity,
measured in energy consumed per dollar of gross

8Energy Information Administration, The National Energy Modeling System: An Overview 1998, DOE/EIA-0581(98) (Washington, DC,
February 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Table ES1. Selected Variables in the Carbon Reduction Cases, 1996 and 2010

2010
1990 1990 1990 1990 1990
Variable 1996 Reference +24% +14% +9% 1990 -3% -7%

U.S. Carbon Emissions
(Million Metric Tons) ........................ 1,463 1,791 1,668 1,535 1,462 1,340 1,300 1,243
Emissions Reductions
(Percent Change From Reference Case) ........ Ñ Ñ 6.9 14.3 18.4 25.2 27.4 30.6
Total Energy Consumption
(Quadrillion Btu)............................ 93.8 111.2 106.5 101.9 99.6 95.2 93.9 91.7
(Percent Change From Reference Case) ........ Ñ Ñ -4.2 -8.4 -10.4 -14.4 -15.6 -17.5
Carbon Price
(1996 Dollars per Metric Ton) ................. Ñ Ñ 67 129 163 254 294 348
Carbon Revenuea
(Billion 1996 Dollars) ........................ Ñ Ñ 110 195 233 333 374 424
Gasoline Price
(1996 Dollars per Gallon) .................... 1.23 1.25 1.39 1.50 1.55 1.72 1.80 1.91
(Percent Change From Reference Case) ........ Ñ Ñ 11.2 20.0 24.0 37.6 44.0 52.8
Average Electricity Price
(1996 Cents per Kilowatthour)................. 6.8 5.9 7.1 8.2 8.8 10.0 10.5 11.0
(Percent Change From Reference Case) ........
Actual Gross Domestic Productb
Ñ Ñ 20.3 39.0 49.2 69.5 78.0 86.4
(Billion 1992 Dollars) ........................ 6,928 9,429 9,333 9,268 9,241 9,137 9,102 9,032
(Percent Change From Reference Case) ........ Ñ Ñ -1.0 -1.7 -2.0 -3.1 -3.5 -4.2
(Annual Percentage Growth Rate, 2005-2010) .... Ñ 2.0 1.8 1.7 1.6 1.4 1.3 1.2
Potential Gross Domestic Product
(Billion 1992 Dollars) ........................ 6,930 9,482 9,469 9,455 9,448 9,429 9,420 9,410
(Percent Change From Reference Case) ........ Ñ Ñ -0.1 -0.3 -0.4 -0.6 -0.7 -0.8
(Annual Percentage Growth Rate, 2005-2010) .... Ñ 2.0 2.0 1.9 1.9 1.9 1.9 1.9
Change in Energy Intensity
(Annual Percent Change, 2005-2010)........... Ñ -1.0 -1.6 -2.0 -2.1 -2.7 -2.8 -3.0
(Percent Change From Reference Case) ........ Ñ Ñ 55.6 96.4 108.2 161.8 177.0 199.0

aThe carbon revenues do not include fees on the nonsequestered portion of petrochemical feedstocks, nonpurchased refinery fuels, or industrial
other petroleum.

bCarbon permit revenues are assumed to be returned to households through personal income tax rebates.

Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B,
FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B, FD07BLW. D080398B.

domestic product (GDP), declines (i.e., improves) at an
average annual rate of 1 percent between 2005 and 2010
in the reference case due to the availability and adoption
of more efficient equipment. In the carbon reduction
cases, higher rates of improvement are projectedÑfrom

1.6 percent a year in the 1990+24% case to triple the reference case rate at 3.0 percent a year in the 1990-7% case.
In 2010, reductions in carbon emissions from electricity
generation account for between 68 and 75 percent of the
total carbon reductions across the cases. Electricity consumption is projected to be lower than in the reference
case, with more efficient, less carbon-intensive technologies used for electricity generation. In all the carbon
reduction cases except the 1990+24% case, carbon emissions from electricity generation in 2010 are lower than
the actual 1990 level of 477 million metric tons of carbon
emissions from the electricity supply sector. Electricity
generators are expected to respond more strongly than
end-use consumers to higher prices because this industry has traditionally been cost-minimizing, factoring
future energy price increases into investment decisions.
In contrast, the end-use consumers are assumed to consider only current prices in making their investment

decisions and to consider additional factors, not only
price, in their decisions. In addition, there are a number
of more efficient and lower-carbon technologies for electricity generation that become economically available as
the cost of generating electricity from fossil fuels
increases.

Total electricity generation is lower in the carbon reduction cases because electricity sales range from 4 to 17 percent below the reference case in 2010 (Figure ES4).
Reduction in electricity demand in response to higher
electricity prices is somewhat mitigated by the change in
relative prices. In 2010, electricity prices are between
20 and 86 percent above the reference case across the carbon reduction cases; however, delivered natural gas
prices are higher by between 25 and 147 percent. With a
smaller percentage price increase, electricity becomes
more attractive in those end uses where it competes with
natural gas, such as home heating.

Although reduced demand for electricity and efficiency
improvements in the generation of electricity contribute
to the total reductions in carbon emissions from electricity generation, fuel switching accounts for most

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Table ES2. Selected Variables in the Carbon Reduction Cases, 1996 and 2020

2020
1990 1990 1990 1990 1990
Variable 1996 Reference +24% +14% +9% 1990 -3% -7%

U.S. Carbon Emissions
(Million Metric Tons) ......................... 1,463 1,929 1,668 1,535 1,468 1,347 1,303 1,251
Emissions Reductions
(Percent Change From Reference Case) ......... Ñ Ñ 13.5 20.4 23.9 30.2 32.5 35.1
Total Energy Consumption
(Quadrillion Btu) ............................ 93.8 117.0 108.6 105.6 103.8 100.9 99.9 98.8
(Percent Change From Reference Case) ......... Ñ Ñ -7.2 -9.7 -11.3 -13.8 -14.6 -15.6
Carbon Price
(1996 Dollars per Metric Ton) .................. Ñ Ñ 99 123 141 200 240 305
Carbon Revenuea
(Billion 1996 Dollars) ......................... Ñ Ñ 162 184 202 263 306 372
Gasoline Price
(1996 Dollars per Gallon) ..................... 1.23 1.24 1.42 1.45 1.49 1.60 1.67 1.80
(Percent Change From Reference Case) ......... Ñ Ñ 14.5 16.9 20.2 29.0 34.7 45.2
Average Electricity Price
(1996 Cents per Kilowatthour) ................. 6.8 5.6 7.3 7.8 8.1 8.7 8.9 9.3
(Percent Change From Reference Case) .........
Actual Gross Domestic Productb
Ñ Ñ 30.4 39.3 44.6 55.4 58.9 66.1
(Billion 1992 Dollars) ......................... 6,928 10,865 10,815 10,808 10,796 10,799 10,793 10,782
(Percent Change From Reference Case) ......... Ñ Ñ -0.5 -0.5 -0.6 -0.6 -0.7 -0.8
(Annual Percentage Growth Rate, 2005-2020)..... Ñ 1.6 1.6 1.6 1.6 1.6 1.6 1.6
Potential Gross Domestic Product
(Billion 1992 Dollars) ......................... 6,930 10,994 10,968 10,961 10,954 10,940 10,933 10.925
(Percent Change From Reference Case) ......... Ñ Ñ -0.2 -0.3 -0.4 -0.5 -0.6 -0.6
(Annual Percentage Growth Rate, 2005-2020)..... Ñ 1.7 1.6 1.6 1.6 1.6 1.6 1.6
Change in Energy Intensity
(Annual Percent Change, 2005-2020) ........... Ñ -0.9 -1.4 -1.4 -1.5 -1.6 -1.7 -1.7
(Percent Change From Reference Case) ......... Ñ Ñ 46.3 54.0 55.7 72.1 76.9 80.9

aThe carbon revenues do not include fees on the nonsequestered portion of petrochemical feedstocks, nonpurchased refinery fuels, or industrial
other petroleum.

bCarbon permit revenues are assumed to be returned to households through personal income tax rebates.

Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B,
FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B, FD07BLW. D080398B.

of the reductions (Figure ES5). The delivered price of
coal to generators in 2010 is higher by between 153 and
nearly 800 percent in the carbon reduction cases relative
to the reference case. As a result, coal-fired generation,
which accounts for about half of all generation in 2010 in
the reference case, has a share between 42 percent and 12
percent in 2010 in the carbon reduction cases. To replace
coal plants, generators build more natural gas plants,
extend the life of existing nuclear plants, and
dramatically increase the use of renewables in the more
stringent reduction cases, particularly biomass and
wind energy systems, which become more economical
with higher carbon prices.

Assuming that carbon emissions from the generation of
electricity are shared to each of the end-use demand
sectors, based upon their consumption of electricity, the
industrial and residential end-use demand sectors
account for most of the carbon reductions, and the
transportation sector accounts for the least (Figure ES6).
In response to higher energy prices, consumers have an
incentive to reduce demand for energy services, switch
to lower-carbon energy sources, and invest in more
energy-efficient technologies.

Because coal is the most carbon-intensive of the fossil
fuels, delivered coal prices are most affected by the
carbon prices (Figure ES7). Higher electricity prices
reflect the increased costs of fossil fuels for generation
and the incremental cost of additional investments,
although the increase is mitigated by generation from
renewables and nuclear power, because their fuel prices
are not affected by carbon prices. Although the average
carbon content of petroleum products is higher than that
of natural gas, the percentage increase in the price of
natural gas is higher than that of petroleum. Higher
prices for petroleum are partially offset by lower world
oil prices, and Federal and State taxes on gasoline also
serve to mitigate the percentage increase.

Total carbon emissions from the industrial sector are
lower by between 7 and 28 percent in 2010 in the carbon
reduction cases, relative to the reference case. Total
industrial output is lower because of the impact of
higher energy prices on the economy. As energy prices
increase, industrial consumers accelerate the replacement of productive capacity, invest in more efficient
technology, and switch to less carbon-intensive fuels.
In 2010, industrial energy intensity is reduced from

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



199019952000200520102015202008001,6002,4003,2004,0004,800BillionKilowatthoursReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
HistoryProjectionsFigure ES4. Projections of U.S. Electricity
Generation, 1990-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV
.D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
1995200020052010201520200100200300400500600MillionMetricTonsFuelSwitchingGenerationEfficiencyDemandReductionsFigure ES5. Projected Reductions in Carbon
Emissions From the Electricity Supply
Sector, 1990-3% Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD03BLW.D080398B.
1995200020052010201520200501001502002503003504001996DollarsperMetricTon1990+24%
1990-3%
1990+9%
1990+14%
19901990-7%
Figure ES2. Projections of Carbon Prices,
1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
0100200300400500600AverageCarbonReductions(MillionMetricTons)
050100150200250300350AverageCarbonPrice(1996DollarsperMetricTon)
1990+24%
1990-3%
1990+9%
1990+14%
19901990-7%
ReferenceFigure ES3. Average Projected Carbon Prices and
Annual Carbon Emission Reductions,
2008-2010
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
7.6 thousand British thermal units (Btu) per dollar of
output in the reference case to between 7.4 and 7.1
thousand Btu in the carbon reduction cases.
In both the residential and commercial sectors, higher
energy prices encourage investments in more efficient
equipment and building shells and reduce the demand
for energy services. Total carbon emissions in the residential sector are reduced by 11 percent in the 1990+24%
case and by 45 percent in the 1990-7% case, relative to the
reference case. Because of reduced demand for energy
and improved end-use efficiencies, total energy use in
2010 ranges from 145 to 173 million Btu per household in

the carbon reduction cases, compared with 184 million
Btu per household in the reference case. Space heating
and cooling account for the largest share of the change in
energy demand; however, energy demand for a variety
of miscellaneous appliances, such as computers, televisions, and VCRs, is also reduced.

In the commercial sector, total carbon emissions are
lower by between 12 and 51 percent in the carbon reduction cases, compared to the reference case. Total energy
use per square foot of commercial floorspace, which is
206 thousand Btu in 2010 in the reference case, is
reduced to between 148 and 192 thousand Btu across the

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



ResidentialCommercialIndustrialTransportationResidentialCommercialIndustrialTransportationResidentialCommercialIndustrialTransprtation04080120160MillionMetricTonsElectricNon-Electric1990+24%1990+9%1990-3%
Figure ES6. Projected Reductions in Carbon
Emissions by End-Use Sector Relative
to the Reference Case, 2010
Note: Electricity emissions are from the fuel used to generate elec-
tricity and are attributed to the sectors relative to their shares of elec-
tricity consumption.
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW. D080398B.
1995200020052010201520200100200300400PercentOilNaturalGasElectricityCoalFigure ES7. Projected Changes in Average
Delivered Prices for Energy Fuels in
the 1990+9% Case Relative to the
Reference Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A and FD09ABV.D080398B.
cases. Similar to the residential sector, most of the reduction occurs for space conditioningÑheating, cooling,
and ventilation; however, more efficient lighting and
office equipment and reduced miscellaneous electricity
useÑfor example, for vending machines and telecommunications equipmentÑalso contribute to lower
energy consumption.

The average price of gasoline in 2010 across the carbon
reduction cases is between 11 and 53 percent higher than
the projected reference case price. Carbon reductions in
the transportation sector in 2010 range from 2 to 16
percent, primarily as the result of reduced travel and the
purchase of more efficient vehicles. The relatively low
carbon reductions for transportation result from the
continued dominance of petroleum, although some
increase in market share is projected for alternative-fuel
vehicles. Improvements in average fuel efficiency are
slowed by vehicle turnover rates. Although new car
efficiency in 2010 improves from 30.6 miles per gallon in
the reference case to between 32.0 and 36.4 miles per
gallon in the carbon reduction cases, total light-duty
fleet efficiency rises only from 20.5 miles per gallon to
between 20.7 and 21.7 miles per gallon. The impact of
carbon prices on the economy lowers light-duty vehicle
and airline travel and freight requirements while
inducing some efficiency improvements.

Impacts by Fuel

In order to achieve carbon emission reductions, the slate
of energy fuels used in the United States is projected to
change from that in the reference case (Figure ES8).

Because of the higher relative carbon content of coal and
petroleum products, the use of both fuels is reduced,
and there is a greater reliance on natural gas, renewable
energy, and nuclear power. Although the use of petroleum declines relative to the reference case, it increases
slightly as a share because most petroleum is used in the
transportation sector, where fewer fuel substitutes are
available.

Because of the high carbon content of coal, total
domestic coal consumption is significantly reduced in
the carbon reduction cases, by between 18 and 77 percent relative to the reference case in 2010 (Figure ES9).
Most of the reductions are for electricity generation,
where coal is replaced by natural gas, renewable fuels,
and nuclear power; however, demand for industrial
steam coal and metallurgical coal is also reduced
because of a shift to natural gas in industrial boilers and
a reduction in industrial output. Coal exports are also
lower in the carbon reduction cases, by between 21 and
32 percent, due to lower demand for coal in the Annex I
nations.

Although total U.S. coal production is reduced, the
average minemouth coal price rises in the carbon
reduction cases, by between 3 and 28 percent in 2010,
because a larger share of production is from higher-cost
eastern coal mines that tend to serve the remaining
markets. Production of western coal is further discouraged by the higher cost of fuels used for rail
transportation and by reduced incentive for investment
in new mines, which are primarily in the West. Because
of lower coal production, coal mine employment in 2010
is projected to be 15 to 63 percent lower than in the

xviii Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



reference case; however, employment in the energy
industry related to the production of natural gas and
renewable fuels is likely to increase.
Petroleum consumption is lower in all the carbon reduction
cases than in the reference case, by between 2 and 13
percent (Figure ES10). Because most of the petroleum is
used for transportation, between 68 and 82 percent of
the total reduction is in the transportation sector, as
travel and freight requirements are reduced and higherefficiency
vehicles are used. Because of lower petroleum
demand in the United States and in other developed
countries that are committed to reducing emissions
under the Kyoto Protocol, world oil prices are lower by
between 4 and 16 percent in 2010, relative to the reference
case price of $20.77 per barrel. In 2010, net crude oil
and petroleum product imports are lower by a range of 3
to 22 percent relative to the reference case. Consequently,
the dependency of the United States on
imported petroleum is reduced from the reference case
level of 59 percent to as little as 53 percent in 2010.
In 2010, natural gas consumption is higher than in the
reference case, by a range of 2 to 12 percent across the
carbon reduction cases (Figure ES11). Increased use of
natural gas in the generation sector is only partially
offset by reductions in the end-use sectors. Later in
the forecast period, continued growth in natural gas
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity xix
1970 1980 1990 2000 2010 2020
0
5
10
15
20
25
30
Quadrillion Btu
1990-3%
1990
1990+24%
Reference
1990+14%
1990+9%
1990-7%
History Projections
Figure ES9. Projections of U.S. Coal Consumption,
1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV
.D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
1970 1980 1990 2000 2010 2020
0
5
10
15
20
25
30
35
40
45
50
Quadrillion Btu
Reference
1990+24%
1990+14%
1990+9%
1990
1990-3%
1990-7%
Series 8
History Projections
Figure ES10. Projections of U.S. Petroleum
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
Reference
(111.2 Quadrillion Btu)
1990+24%
(106.5 Quadrillion Btu)
1990+9%
(99.6 Quadrillion Btu)
1990-3%
(93.9 Quadrillion Btu)
39.4%
26.1%
5.6%
6.6%
0.7%
21.7%
40.2%
27.8%
6.3%
6.9%
0.3%
41.4%
18.5%
34.6%
7.1%
7.9%
8.7%
0.2%
41.3%
32.0%
11.8%
7.0%
7.7%
0.2%
Other
Renewable
Nuclear
Coal
Natural Gas
Oil
Figure ES8. Projections of Fuel Shares of Total U.S. Energy Consumption, 2010
Note: ÒOtherÓ includes net electricity imports, methanol, and liquid hydrogen.
Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.D080398B, and
FD03BLW.D080398B.


1970 19801990200020102020012345678QuadrillionBtu1990-3%
19901990+24%
Reference1990+14%
1990+9%
1990-7%
HistoryProjectionsFigure ES12. Projections of U.S. Nuclear Energy
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
1970198019902000201020200510152025303540QuadrillionBtuReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
HistoryProjectionsFigure ES11. Projections of U.S. Natural Gas
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
consumption for electricity generation is mitigated by
the increasing use of renewables and nuclear power,
particularly in the more stringent carbon reduction
cases. As a result, in 2020, natural gas use does not necessarily increase with higher levels of carbon reductions.
As the result of higher demand, the average wellhead
price of natural gas in 2010 is higher in all the carbon
cases than in the reference case, by a range of 2 to 30 percent. Although meeting the levels of production that
may be required will be a challenge for the industry, sufficient natural gas resources are available. The potential
increases in both drilling and pipeline capacity are
within levels achieved historically (or about to be
achieved) and are not likely to be a constraint, given
appropriate incentives and planning.

Nuclear power, which produces no carbon emissions,
increases with carbon reduction targets by between 8
and 20 percent in 2010, relative to the reference case (Figure ES12). Although no new nuclear plants are assumed
to be built in the carbon reduction cases, extending the
lifetimes of existing plants is projected to become more
economical with higher carbon prices. In the more stringent carbon reduction cases, most existing nuclear
plants are life-extended through 2020, in contrast to the
gradual retirement of approximately half of the nuclear
plants projected in the reference case.

Consumption of renewable energy, which results in no
net carbon emissions, is projected to be significantly
higher with carbon reduction targets (Figure ES13).
Across the carbon reduction cases, renewable energy
consumption increases by between 2 and 16 percent in
2010 and by between 9 and 70 percent in 2020. Most of

199019952000200520102015202002468101214QuadrillionBtu1990-3%
19901990+24%
Reference1990+14%
1990+9%
1990-7%
HistoryProjectionsFigure ES13. Projections of U.S. Renewable
Energy Consumption, 1990-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV
.D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
this increase occurs in electricity generation, primarily
with additions to wind energy systems and an increase
in the use of biomass (wood, switchgrass, and refuse). In
the carbon reduction cases, the share of renewable
generation is as much as 14 percent in 2010, compared
with 10 percent in the reference case, increasing to as
high as 22 percent in 2020, compared with 9 percent
in the reference case. Because additional renewable
technologies become available and economical later in
the forecast period, the share of renewable generation
continues to increase through 2020.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Macroeconomic Impacts

In the energy market analyses, the projected carbon
prices reflect the prices the United States would be willing to pay to achieve the Kyoto targets, without addressing the international trade in carbon permits. The
macroeconomic analysis assumes that the carbon permit
trading system would function as an auction run by the
Federal Government, and that the United States would
be free to purchase carbon permits in an international
market at the marginal abatement cost in the United
States. The U.S. State DepartmentÕs assessment of the
accounting of carbon-absorbing sinks and offsets from
reductions in other greenhouse gases is assumed to
reduce the U.S. emissions target to 3 percent below 1990
levels. The 3-percent target is then achieved through a
combination of domestic actions and the purchase of
permits on the international market. Thus, two flows of
funds occurÑdomestic and international.

On the domestic side, U.S. permits are sold in a competitive auction run by the Federal Government, raising
large sums of funds. In the 1990-3% case, where the revenues come entirely from the domestic market, the revenue collected in 2010 is projected to total $585 billion
nominal dollars and $317 billion and $128 billion in the
1990+9% and 1990+24% cases, respectively. The collection of this money necessitates a careful consideration of
appropriate fiscal policy to accompany the permit auction. Two approaches are considered: first, returning
collected revenues to consumers through a personal
income tax lump sum rebate and, second, lowering
social security tax rates as they apply to both employers
and employees. The two policies are meant only to be
representative of a set of possible fiscal policies that
might accompany an initial carbon mitigation policy.

The second flow of funds is associated with U.S.
purchases of international carbon permits and assumes
that the carbon price determined in the U.S. energy
market analysis is the international price at which
permits would be traded. The differences between the
reduction level in the 1990-3% case and those in the other
cases are assumed to be met by purchases of permits in
international markets. Table ES3 shows average carbon

reductions, purchases of international permits, and the
carbon price for the three cases considered in the macroeconomic assessment for the 2008-2012 period.

The energy market analysis in this report does not
address the international implications of achieving a
particular target at the projected carbon price. For the
macroeconomic assessment, the simplifying assumption
is made that in each case the domestic carbon price is the
same as the international permit price when different
levels of trading are used to achieve the Kyoto target,
implying that different international supplies of permits
would be available in the alternative cases considered.
This is an important simplifying assumption, and the
value placed on the overseas transfer of funds to purchase international permits is subject to considerable
uncertainty. However, this element must be considered
a key factor in performing any assessment of the impacts
on the economy, and therefore it is explicitly factored
into the analysis.

As a direct consequence of the carbon price, aggregate
energy prices in the U.S. economy are expected to rise.
One way to measure this effect is to look at the percentage change in prices in the economy. For example, in the
1990+9% case, energy prices are 56 percent higher than
the reference case projection in 2010 and remain more
than 50 percent above the reference case over the rest of
the forecast period. The projected energy price increases
would also affect downstream prices for all goods and
services in the economy as measured by the producer
price index. The projected increase in producer prices
relative to the reference case in 2010 is 9 percent in the
1990+9% case. Final prices for goods and services in
2009, as shown by the consumer price index (CPI) series,
are about 4 percent higher in the 1990+9% case (Figure
ES14). Expressed as a rate of change, CPI inflation rises
by 0.7 percentage points between 2005 and 2010, as the
reference case CPI rises by 3.6 percent a year and the
1990+9% case rises by 4.3 percent a year. These figures
suggest the following rule of thumb for the year 2010:
each 10-percent increase in aggregate prices for energy
may lead to a 1.5-percent increase in producer prices and
a 0.7-percent increase in consumer prices.

Table ES3. Energy Market Assumptions for the Macroeconomic Analysis of Three Carbon Reduction Cases,
Average Annual Values, 2008 through 2012

Analysis Case
Binding
Carbon
Emissions
Reduction Target
(Million
Metric Tons)
Average U.S.
Carbon
Emissions
Reductions
(Million
Metric Tons)
U.S. Purchases
of International
Permits
(Million
Metric Tons)
Carbon Price
Value of
Purchased
International
Permits
(Billion
1992 Dollars)
1996 Dollars per
Metric Ton
1992 Dollars per
Metric Ton

1990-3%........ 485 485 0 290 263 0
1990+9%....... 485 325 160 159 144 23
1990+24%...... 485 122 363 65 59 21

Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



19952000200520102015202001234567PercentChangeFromReferenceCase1990+9%
1990-3%
1990+24%
19952005201501234567AssumingSocialSecurityTaxRebateFigure ES14. Projected Changes in Consumer
Price Index Relative to the Reference
Case, 1998-2020
Note: Carbon permit revenues are assumed to be returned to
households through reductions in personal income taxes.
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
One aspect of the CPI is particularly noteworthy. The
CPI measures the prices that consumers face, regardless
of the country of origin of the product. Import prices, to
the extent that they do not rise at the rate of domestic
prices because non-Annex I countries do not face carbon
constraints, would dampen the price effects as lower-
priced imports find their way into U.S. markets.

Because energy resources are used to produce most
goods and services, higher energy prices can affect the
economyÕs production potential. Long-run equilibrium
costs are associated with reducing reliance on energy in
favor of other factors of productionÑincluding labor
and capital, which become relatively cheaper as energy
costs rise. Short-run adjustment costs, or business cycle
costs, can arise when price increases disrupt capital or
employment markets. Long-run costs are considered
unavoidable. Short-run costs might be avoidable if price

changes can be accurately anticipated or if appropriate
compensatory monetary and fiscal policies can be
implemented. The economic assessment in this analysis
considers both the short-run and long-run costs to the
economy and focuses on the 1990-3%, 1990+9%, and
1990+24% carbon reduction cases.

The possible impacts on the economy are summarized in
Table ES4, which shows average changes from the reference case projections over the period from 2008 through
2012 in the three carbon reduction analysis cases. The
loss of potential GDP measures the loss in productive
capacity of the economy directly attributable to the
reduction in energy resources available to the economy.
The macroeconomic adjustment cost reflects frictions in the
economy that may result from the higher prices of the
carbon mitigation policy. It recognizes the possibility
that cyclical adjustments may occur in the short run. The
loss in actual GDP for the economy is the sum of the loss
in potential and the adjustment cost. The purchase of
international permits represents a claim on the productive
capacity of domestic U.S. resources. Essentially, as funds
flow abroad, other countries have an increased claim on

U.S. goods and services.
The loss of potential GDP plus the purchase of international permits represent the long-run, unavoidable
impact on the economy. The total cost to the economy is
represented by the loss in actual GDP plus the purchase
of international permits (Figure ES15). These costs need
to be put in perspective relative to the size of the
ecomomy, which averages $9,425 billion between 2008
and 2012. Tables ES5 and ES6 summarize the macroeconomic impacts projected for the years 2010 and 2020.

In the long run, higher energy costs would reduce
the use of energy by shifting production toward less
energy-intensive sectors, by replacing energy with labor
and capital in specific production processes, and by
encouraging energy conservation. Although reflecting a
more efficient use of higher-cost energy, the gradual

Table ES4. Macroeconomic Impacts in Three Carbon Reduction Cases, Average Annual Values, 2008-2012
(Billion 1992 Dollars)
Analysis Case
Loss in
Potential GDP
Macroeconomic
Adjustment Cost
Loss in
Actual GDP
Purchases of
International
Permits
Total Cost
to the Economy
1990-3% .......................
Personal Income Tax Rebate ...... 58 225 283 0 283
Social Security Tax Rebate ........ 58 70 128 0 128
1990+9% .......................
Personal Income Tax Rebate ...... 32 137 169 23 192
Social Security Tax Rebate ........ 32 59 91 23 114
1990+24% ......................
Personal Income Tax Rebate ...... 12 76 88 21 109
Social Security Tax Rebate ........ 12 44 56 21 77
Note: Loss in potential GDP plus the macroeconomimc adjustment cost equals the loss in actual GDP. The actual GDP loss plus purchases of
international permits equals the total cost to the economy.
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model of the U.S. Economy.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



reduction in energy use would tend to lower the productivity of other factors in the production process. The
derivation of the long-run equilibrium path of the
economy can be characterized as representing the
ÒpotentialÓ output of the economy when all resourcesÑ
labor, capital, and energyÑare fully employed. As such,
potential GDP is equivalent to the full employment concept in other analyses that focus on long-run growth
while abstracting from business cycle behavior. Figure
ES16 shows the losses in the potential economic output,
as measured by potential GDP, for the three carbon
reduction cases. The shapes of the three trajectories
mirror the carbon price trajectories.

The ultimate impacts of carbon mitigation policies on
the economy will be determined by complex interactions between elements of aggregate supply and
demand, in conjunction with monetary and fiscal policy
decisions. As such, cyclical impacts on the economy
are bound to be characterized by uncertainty and controversy. However, raising the price of energy and

050100150200250300Billion1992DollarsMacroeconomicAdjustmentCostLossinPotentialGDPPurchaseofInternationalPermitsAssumingPersonalncomeTaxRebate1990+9%1990-3%1990+24%
AssumingSocialSecurityTaxRebateFigure ES15. Projected Annual Costs of Carbon
Reductions to the U.S. Economy,
2008-2012
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
Table ES5. Projected Impacts on Gross Domestic Product, 2005 and 2010
Variable 1996
2005
Reference
2010
Refer-
ence
1990
+24%
1990
+14%
1990
+9% 1990
1990
-3%
1990
-7%
Potential GDP
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2010, Percent) ........
6,930
Ñ
Ñ
8,585
Ñ
Ñ
9,482
Ñ
2.0
9,469
-0.1
2.0
9,455
-0.3
1.9
9,448
-0.4
1.9
9,429
-0.6
1.9
9,420
-0.7
1.9
9,410
-0.8
1.9
Actual GDP, Assuming Personal Income Tax Rebate
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2010, Percent) ........
6,928
Ñ
Ñ
8,525
Ñ
Ñ
9,429
Ñ
2.0
9,333
-1.0
1.8
9,268
-1.7
1.7
9,241
-2.0
1.6
9,137
-3.1
1.4
9,102
-3.5
1.3
9,032
-4.2
1.2
Actual GDP, Assuming Social Security Tax Rebate
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2010, Percent) ........
6,928
Ñ
Ñ
8,525
Ñ
Ñ
9,429
Ñ
2.0
9,369
-0.6
1.9
9,337
-1.0
1.8
9,326
-1.1
1.8
9,291
-1.5
1.7
9,281
-1.6
1.7
9,247
-1.9
1.6
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model of the U.S. Economy.
Table ES6. Projected Impacts on Gross Domestic Product, 2005 and 2020
Potential GDP
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2020, Percent) ........
6,930
Ñ
Ñ
8,585
Ñ
Ñ
10,994
Ñ
1.7
10,968
-0.2
1.6
10,961
-0.3
1.6
10,954
-0.4
1.6
10,940
-0.5
1.6
10,933
-0.6
1.6
10,925
-0.6
1.6
Actual GDP, Assuming Personal Income Tax Rebate
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2020, Percent) ........
6,928
Ñ
Ñ
8,525
Ñ
Ñ
10,865
Ñ
1.6
10,815
-0.5
1.6
10,808
-0.5
1.6
10,796
-0.6
1.6
10,799
-0.6
1.6
10,793
-0.7
1.6
10,782
-0.8
1.6
Actual GDP, Assuming Social Security Tax Rebate
(Billion 1992 Dollars)..........................
(Percent Change From Reference Case) ..........
(Annual Growth Rate, 2005-2020, Percent) ........
6,928
Ñ
Ñ
8,525
Ñ
Ñ
10,865
Ñ
1.6
10,840
-0.2
1.6
10,832
-0.3
1.6
10,828
-0.3
1.6
10,833
-0.3
1.6
10,835
-0.3
1.6
10,842
-0.2
1.6
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model of the U.S. Economy.
Variable 1996
2005
Reference
2020
Refer-
ence
1990
+24%
1990
+14%
1990
+9% 1990
1990
-3%
1990
-7%
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity xxiii



199520002005201020152020010-10-20-30-40-50-60-70Billion1992Dollars1990+9%
1990-3%
1990+24%
ReferenceCasePotentialGDP:
1996=$6,9302010=$9,4822020=$10,994Figure ES16. Projected Dollar Losses in Potential
GDP Relative to the Reference Case,
1998-2020
Note: Carbon permit revenues are assumed to be returned to
households through reductions in personal income taxes.
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
199520002005201020152020050-50-100-150-200-250Billion1992DollarsPotentialGDPActualGDP,
SocialSecurityTaxRebateActualGDP,
PersonalIncomeTaxRebateReferenceCasePotentialGDP:
1996=$6,9302010=$9,4822020=$10,994Figure ES17. Projected Changes in Potential and
Actual GDP in the 1990+9% Case
Relative to the Reference Case Under
Different Fiscal Policies, 1998-2020
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
downstream prices in the rest of the economy could
introduce cyclical behavior in the economy, resulting in
employment and output losses in the short run. The
measurement of losses in actual output for the economy,
or actual GDP, represents the transitional cost to the
aggregate economy as it adjusts to its long-run path.
Resources may be less than fully employed, and
the economy may move in a cyclical fashion as the
initial cause of the disturbanceÑthe increase in energy
pricesÑplays out over time.

Collection of money from a permit auction system
necessitates a careful consideration of appropriate fiscal
policy to accompany the carbon reduction policy. Two
alternative fiscal policies are analyzed, both returning
collected revenue back to agents in the economy: a cut in
personal income taxes and a cut in social security taxes
as they apply to both employers and employees. In both
cases, the Federal deficit is maintained at reference case
levels. The personal income tax cut essentially returns
collected revenues to consumers, helping to maintain
personal disposable income. Like the personal income
tax cut, the social security tax cut returns collected funds
to the private sector of the economy, ameliorating the
near-term impacts of higher energy prices. Although
consumers and businesses still would face much higher
relative prices for energy than for other goods and services, disposable income is maintained near reference
case values to the extent that funds flow back to consumers.

In the fiscal policy settings, higher prices in the economy
place upward pressure on interest rates. The Federal
Reserve Board seeks to balance the consequences of
higher energy prices on the economy and possible

adverse effects on output and employment by making
adjustments to the Federal funds rate. The adjustments
would be designed to moderate the possible impacts on
both inflation and unemployment, and to return the
economy to its long-run growth path.

Figure ES17 shows the projected impacts on both actual
and potential GDP for the two hypothetical fiscal policies (income tax and social security tax cuts) in the
1990+9% case. The figure indicates that, in the 2008 to
2012 period, the short-run cyclical impact on actual GDP
is larger than the long-run impact on potential GDP;
however, the two output concepts begin to converge by
2015, and by 2020 they have merged into a steady-state
path reflected by potential GDP. Monetary policy is
instrumental in balancing inflation and unemployment
impacts through the adjustment period, acting in a manner to bring the economy back to its long-run growth
path.

The choice of the accommodating fiscal policy is also key
to the assessment of the ultimate impacts on the economy. While the personal income tax option moderates
the impacts through a return of funds to consumers, the
social security tax option has cost-cutting aspects of lowering the employer portion of the tax, which serves to
reduce inflationary pressures in the aggregate economy.
On the employer side, the reduction in employer contributions to the social security system would lower costs
to the firm and, thereby, moderate the near-term price
consequences to the economy. Since it is the price effect
that produces the predominately negative effect on the
economy, any steps to reduce inflationary pressures
would serve to moderate adverse impacts on the aggregate economy.

xxiv Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Another way to view the macroeconomic effects is by
looking at the effects of the carbon reduction cases on the
growth rate of the economy, both during the period of
implementation from 2005 through 2010 and then over
the entire period from 2005 through 2020 (Figures ES18
and ES19). In the reference case, potential and actual
GDP grow at 2.0 percent per year from 2005 through
2010. In the 1990+9% case, the growth rate in potential
GDP slows to 1.9 percent per year, and the growth rate
in actual GDP slows to 1.6 percent per year when the
personal income tax rebate is assumed or 1.8 percent per
year when the social security tax rebate is assumed.
However, through 2020, with the economy rebounding
back to the reference case path, there is no appreciable
change in the projected long-term growth rate. The
results for the 1990+24% and 1990-3% cases are similar.

Aggregate impacts on the economy, as measured by
potential and actual GDP, are shown in Table ES7 in
terms of losses in GDP per capita. In the 1990+9% case,
the loss in potential GDP per capita is $106; however, the
loss in actual GDP for in the 1990+9% case is $567 assuming the personal income tax rebate and $305 assuming
the social security tax rebate. Again, the lower value
(loss in potential GDP) represents an unavoidable loss
per person, and the higher values (loss in actual GDP)
reflect the highly uncertain, but significant, impacts that

individuals could experience as the result of frictions
within the economy. To provide perspective, actual
GDP per capita averages $31,528 in the reference case
between 2008 and 2012.

Sensitivity Cases

This analysis includes several sensitivity cases designed
to examine alternative assumptions that may have significant impacts on energy demand and carbon emissions over the next 20 years, including higher and lower
economic growth, faster and slower availability and
rates of improvement in technology, and the construction of new nuclear power plants. The sensitivity cases
illustrate how such factors influence the results of the
carbon reduction cases. With the exception of the
nuclear power case, the sensitivity cases are analyzed
relative to the 1990+9% case.

Because each sensitivity case is constrained to the same
level of carbon emissions as the case to which it is
compared, the primary impact is not on the carbon
emissions levels, or even on aggregate energy consumption, but rather on the carbon price required to
meet the emissions target. For example, in the high
technology case, projected carbon emissions during the

PotentialGDPActualGDP,
AssumingPersonalIncomeTaxRebateActualGDP,
AssumingSocialSecurityTaxRebate0.00.51.01.52.02.5PercentperYearReference1990+24%1990+9%1990-3%
Figure ES18. Projected Annual Growth Rates in
Potential and Actual GDP, 2005-2010
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
PotentialGDPActualGDP,
AssumingPersonalIncomeTaxRebateActualGDP,
AssumingSocialSecurityTaxRebate0.00.51.01.52.02.5PercentperYearReference1990+24%1990+9%1990-3%
Figure ES19. Projected Annual Growth Rates in
Potential and Actual GDP, 2005-2020
Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model
of the U.S. Economy.
Table ES7. Projected Losses in Potential and Actual GDP per Capita, Average Annual Values, 2008-2012
(1992 Dollars per Person)

Loss in Actual GDP Loss in Actual GDP
Loss in Potential GDP per Capita, per Capita,
Analysis Case per Capita Personal Income Tax Rebate Social Security Tax Rebate

1990-3%.................... 193 947 428
1990+9% ................... 106 567 305
1990+24.................... 40 294 187


Source: Simulations of the Data Resources, Inc. (DRI) Macroeconomic Model of the U.S. Economy.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



compliance period are the same as in the corresponding
reference technology case. What differs is the cost of
meeting the target, as reflected in the required carbon
price.

Macroeconomic Growth

The assumed rate of economic growth has a strong
impact on the projection of energy consumption and,
therefore, on the projected levels of carbon emissions.
Two sensitivity cases explore the effects of higher
and lower economic growth on the cost of reducing carbon emissions to the 1990+9% level. Higher economic growth results from higher assumed growth in
population, the labor force, and labor productivity,
resulting in higher industrial output, lower inflation,
and lower interest rates. As a result, GDP increases at an
average rate of 2.4 percent a year through 2020, compared with a growth rate of 1.9 percent a year in the reference case. With higher macroeconomic growth,
energy demand grows faster, as higher manufacturing
output and higher income increase the demand for
energy services, resulting in higher carbon emissions.
Assumptions of lower growth in population, the labor
force, and labor productivity result in an average annual
growth rate of 1.3 percent in the low economic growth
case, resulting in lower carbon emissions.

With higher economic growth, both industrial output
and energy service demand are higher. As a result,
carbon prices must be correspondingly higher to attain
a given carbon emissions target. In the high macroeconomic growth case, the carbon price in 2010 is $215
per metric ton, $52 per metric ton higher than the carbon
price of $163 per metric ton in the 1990+9% case with
reference growth assumptions (Figure ES20). In the low

128163215243163121LowGrowthReferenceGrowthHighGrowthLowTech-
nologyReferenceTech-
nologyHighTech-
nology0501001502002501996DollarsperMetricTonFigure ES20. Projected Carbon Prices in the
1990+9% High and Low Economic
Growth and High and Low
Technology Sensitivity Cases, 2010
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs FD09ABV.D080398B, LMAC09.D080698A, HMAC09.
D080598A, FREEZE09. D080798A, and HITECH09.D080698A.
macroeconomic growth case, the carbon price in 2010 is
$128 per metric ton. The higher carbon prices necessary
to achieve the carbon reductions with higher economic
growth have a negative impact on the economy and the
energy system. Nevertheless, total energy consumption
in 2010 is higher with higher economic growth, by 2.2
quadrillion Btu relative to the 1990+9% case, which
assumes the same economic growth rate as the reference
case. In the low economic growth case, total energy
consumption is lower by 2.2 quadrillion Btu in 2010.

In order to meet the carbon reduction targets with
higher economic growth, there is a shift to less carbon-
intensive fuels and higher energy efficiency. On a sectoral basis, higher economic growth affects total energy
consumption in the industrial and transportation sectors
more significantly than in the other end-use sectors.
Total consumption of both renewables and natural gas is
higher, primarily for electricity generation but also in
the industrial sector. Coal use for generation is lower,
and the use of nuclear power is higher as a result of the
higher carbon prices. Petroleum consumption is also
higher with higher economic growth, both in the transportation and industrial sectors.

Total energy intensity is lower in the high economic
growth case, partially offsetting the increases in the
demand for energy services caused by the higher
growth assumption. With higher economic growth,
there is greater opportunity to turn over and improve
the stock of energy-using technologies. In addition, the
higher carbon price induces more efficiency improvements and some offsetting reductions in energy service
demand, moderating the impacts of higher economic
growth. With higher economic growth, aggregate
energy intensity declines at an average annual rate of 1.9
percent through 2010, compared to 1.6 percent with reference economic growth. The opposite effects on energy
intensity occur with lower economic growth, with the
decline in energy intensity slowing from 1.6 percent to

1.3 percent between 1996 and 2010.
Technological Progress

The rates of development and market penetration of
energy-using technologies have a significant impact on
projected energy consumption and energy-related
carbon emissions. Faster development of more energy-
efficient or lower-carbon-emitting technologies than
assumed in the reference case could reduce both consumption and emissions; however, because the reference case already assumes continued improvement in
both energy consumption and production technologies,
slower technological development is also possible.

To analyze the impacts of technology improvement,
high technology assumptions were developed by
experts in technology engineering for each of the
energy-consuming sectors, considering the potential

xxvi Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



impacts of increased research and development for
more advanced technologies. The revised assumptions
included earlier years of introduction, lower costs,
higher maximum market potential, and higher efficiencies than assumed in the reference case.9 Also, this
sensitivity case assumed the availability of carbon
sequestration technology for coal-and natural-gas-fired
power plants, which would remove carbon dioxide and
store it in underground aquifers; however, the technology is uneconomical relative to other technologies
because of its high operating and storage costs.

These technological improvements were developed
under the assumption of increased research and development, and they are distinct from the more rapid adoption of advanced technologies that occurs with higher
energy prices in the carbon reduction cases. It is possible
that further technology improvements could occur
beyond those in the high technology sensitivity case if a
very aggressive research and development effort were
established. The low technology sensitivity case
assumes that all future equipment choices are made
from the end-use and generation equipment available in
1998, with new building shell and industrial plant efficiencies frozen at 1998 levels. Comparing this sensitivity
case to a case with reference technology assumptions
demonstrates the importance of technology improvement in the reference case.

Because faster technology development makes advanced energy-efficient and low-carbon technologies
more economically attractive, the carbon prices required
to meet carbon reduction levels are significantly
reduced. Conversely, slower technology improvement
requires higher carbon prices (Figure ES20). With high
technology assumptions, the carbon price in 2010 is $121
per metric ton, $42 per metric ton lower than the carbon
price of $163 per metric ton in the 1990+9% case with the
reference technology assumptions. With the low technology assumptions, the carbon price increases to $243
per metric ton in 2010.

In the high technology sensitivity case, total energy
consumption in 2010 is lower by 2.1 quadrillion Btu, or
about 2 percent, than in the 1990+9% case with reference
technology. Delivered energy consumption in both the
industrial and transportation sectors is lower as
efficiency improvements in industrial processes and
most transportation modes outweigh the countervailing
effects of lower energy prices. In the residential and
commercial sectors, the effect of lower energy prices
balances the effect of advanced technology, and
consumption levels are at or near those in the reference
technology (1990+9%) case. In the generation sector, coal
use for generation is 40 percent higher than with

reference technology assumptions, due to efficiency
improvements and the lower carbon price.

In the low technology sensitivity case, the converse
trends prevail. In 2010, total energy consumption is
higher by 1.5 quadrillion Btu than in the 1990+9% case
with reference technology assumptions. Delivered
energy consumption is higher in the industrial and
transportation sectors and lower in the residential and
commercial sectors, suggesting that industry and transportation are more sensitive to technology changes than
to price changes, and the residential and commercial
sectors are more sensitive to price changes. With the
higher carbon prices in the low technology case, coal use
is further reduced in the generation sector, and more
natural gas, nuclear power, and renewables are used to
meet the carbon reduction targets.

Nuclear Power

In the reference case, nuclear electricity generation
declines significantly because 52 percent of the total
nuclear capacity available in 1996 is assumed to be
retired by 2020. A number of units are retired before the
end of their 40-year operating licenses, as suggested by
industry announcements and analysis of the age and
operating costs of the units. In the carbon reduction
cases, life extension of the plants can occur if it is
economical; and there is an increasing incentive to invest
in nuclear plant refurbishment with higher carbon
prices. However, these cases do not allow the construction of new nuclear power plants, given continuing high
capital investment costs and institutional constraints
associated with nuclear power. A nuclear power sensitivity case examines the impact of allowing new plants
to be constructed. Because nuclear plants still are not
economically competitive with fossil and renewable
plants in the 1990+9% case, the nuclear power sensitivity
case was analyzed against the 1990-3% case. In addition
to allowing new nuclear plants, the higher costs
assumed in the reference case for the first few advanced
nuclear plants were reduced in this sensitivity.

Relative to the 1990-3% case, 1 gigawatt of new nuclear
capacity is added by 2010 in the nuclear power sensitivity case, and 41 gigawatts, representing about 68 new
plants of 600 megawatts each, are added by 2020. With
most of the impact from the new nuclear plants coming
after the commitment period of 2008 through 2012, there
is little impact on carbon prices in 2010. By 2020, however, carbon prices are $199 per metric ton with the
assumption of new nuclear plants, as compared with
$240 per metric ton in the 1990-3% case with the reference nuclear assumptions. In 2010, total energy consumption is about the same in this sensitivity case as in

9The design of the high technology sensitivity case differs from the high technology cases in AEO98, which generally did not include an
analysis of improvements for specific technologies.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity xxvii



the 1990-3% case, but in 2020 it is about 1.8 quadrillion
Btu higher. Somewhat lower energy prices induce
higher consumption in all sectors, and the availability of
more carbon-free nuclear generation allows the carbon
reduction target to be met with higher end-use consumption.

Uncertainties in the Analysis

The reference case projections in both AEO98 and this
analysis represent business-as-usual forecasts, given
known trends in technology and demographics, current
laws and regulations, and the specific methodologies
and assumptions used by EIA. Because EIA does not
include future legislative and regulatory changes in its
reference case projections, the projections provide a
policy-neutral baseline against which the impacts of policy initiatives can be analyzed.

Results from any model or analysis are highly uncertain.
By their nature, energy models are simplified representations of complex energy markets. The results of any
analysis are highly dependent on the specific data,
assumptions, behavioral characteristics, methodologies,
and model structures included. In addition, many of the
factors that influence the future development of energy
markets are highly uncertain, including weather, political and economic disruptions, technology development,
and policy initiatives. Recognizing these uncertainties,
EIA has attempted in this study to isolate and analyze
the most important factors affecting future carbon emissions and carbon prices. The results of the various cases
and sensitivities should be considered as relative
changes to the comparative baseline cases.

In addition to the uncertainties concerning the final
interpretation and implementation of the Kyoto Protocol, specific actions that might be taken to reduce greenhouse gas emissions in the United States have not been
formulated. Actions taken by other Annex I countries to
reduce emissions, future growth in worldwide energy
consumption and emissions, and the opportunities for
reducing emissions through joint implementation and

the CDM are unknown, and they are likely to have
important impacts on the international trade of carbon
permits and the carbon permit price. This analysis
assumes that auctioned permits will constrain carbon
emissions and raise the price of fossil fuels, with revenues from the auction recycled to consumers either
through personal income tax or social security tax
rebates. Alternative carbon reduction programs and fiscal policies would be likely to change the cost of carbon
reduction from the costs in this analysis. The timing of
carbon reduction programs and the amount of adjustment time allowed could also be important in determining costs.

Future technology development also cannot be known
with certainty and may have a significant effect on the
cost of achieving carbon reductions. The technology sensitivity cases in this analysis explore some of the potential impacts, but even the high technology sensitivity
does not include possible breakthrough or speculative
technologies. On the other hand, even the reference case
technology assumptions include continued development of more energy-efficient and renewable technologies, which serve to mitigate the costs of carbon
reduction. Those technology improvements are likely,
but not certain.

Finally, consumer response to carbon initiatives is
uncertain. Because energy price changes that have
occurred in the past may not provide sufficient evidence
about the reaction of consumers to sustained high
energy prices, changes in demand as a result of the
higher carbon fees cannot be projected with confidence.
In addition to price-induced changes, consumers might
also respond to climate change initiatives and a national
commitment to reduce emissions by adopting more
energy-efficient or renewable technologies sooner than
expected. Finally, public acceptance of large-scale
renewable technologies or the continuation of nuclear
powerÑboth of which make important contributions to
the achievement of the carbon emissions reductions at
the costs projected in this analysisÑcannot be known
with certainty.

xxviii Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1. Scope and Methodology of the Study
Background
The Greenhouse Gas Effect

The greenhouse effect is a natural process by which
some of the radiant heat from the Sun is captured in the
lower atmosphere of the Earth, thus maintaining the
temperature of the Earth's surface. The gases that help
capture the heat, called Ògreenhouse gases,Ó include
water vapor, carbon dioxide, methane, nitrous oxide,
and a variety of manufactured chemicals. Some are
emitted from natural sources; others are anthropogenic,
resulting from human activities.

Over the past several decades, rising concentrations of
greenhouse gases have been detected in the Earth's
atmosphere. Although there is not universal agreement
within the scientific community on the impacts of
increasing concentrations of greenhouse gases, it has
been theorized that they may lead to an increase in the
average temperature of the Earth's surface. To date, it
has been difficult to note such an increase conclusively
because of the differences in temperature around the
Earth and throughout the year, and because of the difficulty of distinguishing permanent temperature changes
from the normal fluctuations of the Earth's climate. In
addition, there is not universal agreement among scientists and climatologists on the potential impacts of an
increase in the average temperature of the Earth,
although it has been hypothesized that it could lead to a
variety of changes in the global climate, sea level, agricultural patterns, and ecosystems that could be, on net,
detrimental.

The most recent report of the Intergovernmental Panel
on Climate Change (IPCC) concluded that: ÒOur ability
to quantify the human influence on global climate is currently limited because the expected signal is still emerging from the noise of natural variability, and because
there are uncertainties in key factors. These include the

magnitudes and patterns of long-term variability and
the time-evolving pattern of forcing by, and response to,
changes in concentrations of greenhouse gases and aerosols, and land surface changes. Nevertheless, the balance of evidence suggests that there is a discernable
human influence on global climate.Ó1

U.S. Greenhouse Gas Emissions
In 1990, total greenhouse gas emissions in the United
States were 1,618 million metric tons carbon equivalent,2
according to 1997 estimates published by the Energy
Information Administration (EIA).3 Of this total, 1,346
million metric tons, or 83 percent, was due to carbon
emissions from the combustion of energy fuelsÑthe
focus of this report. By 1996, total U.S. greenhouse gas
emissions had risen to 1,753 million metric tons carbon
equivalent, including 1,463 million metric tons of carbon
emissions from energy combustion. EIA's Annual Energy
Outlook 1998 (AEO98)4 projects that energy-related carbon emissions will reach 1,577 million metric tons in
2000, 17 percent above the 1990 level. Projected emissions continue to rise at an average annual rate of 1.5
percent a year from 1996 to 2010, reaching 1,803 million
metric tons of carbon emissions in 2010, 34 percent
above the 1990 level. Because energy-related carbon
emissions are a large portion of total greenhouse gas
emissions, any efforts to reduce greenhouse gas emissions will likely have a significant impact on the energy
sector; however, as discussed later, there are a number
of factors outside the domestic energy market that also
affect emissions levels.

To put U.S. emissions in a global perspective, the United
States produced about 24 percent of the worldwide
energy-related carbon emissions in 1996, which totaled

6.6 billion metric tons, as noted in EIA's International
Energy Outlook 1998 (IEO98).5 Although continued
increases in carbon emissions are expected for the
United States and other industrialized countries, much
1Intergovernmental Panel on Climate Change, Climate Change 1995: The Science of Climate Change (Cambridge, UK: Cambridge University
Press, 1996).

2Greenhouse gases differ in their impacts on global temperatures. For comparison of emissions from the various gases, they are often
weighted by global warming potential (GWP), established by the Intergovernmental Panel on Climate Change, which is a measure of the
impact of each gas on global warming relative to that of carbon dioxide, which is defined as having a GWP equal to 1.

3Energy Information Administration, Emissions of Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington, DC, Octo

ber 1997).

4Energy Information Administration, Annual Energy Outlook 1998, DOE/EIA-0383(98) (Washington, DC, December 1997).

5Energy Information Administration, International Energy Outlook 1998, DOE/EIA-0484(98) (Washington, DC, April 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



more rapid increases are projected for the developing
countries of Asia, the Middle East, Africa, and Central
and South America. As a result, global carbon emissions
from energy use are expected to increase at an average
annual rate of 2.4 percent from 1996 through 2010, reaching 8.3 billion metric tons, to which the United States
would contribute about 22 percent.

The Framework Convention on
Climate Change

As a result of increasing warnings by members of the climatological and scientific community about the possible
harmful effects of rising greenhouse gas concentrations,
the IPCC was established by the World Meteorological
Organization and the United Nations Environment
Programme in 1988 to assess the available scientific,
technical, and socioeconomic information in the field of
climate change. A series of international conferences
followed, and in 1990 the United Nations established the
Intergovernmental Negotiating Committee for a Framework Convention on Climate Change. After a series
of negotiating sessions, the text of the Framework Convention on Climate Change was adopted at the United
Nations on May 9, 1992, and opened for signature at Rio
de Janeiro on June 4.

The objective of the Framework Convention was to Ò. . .
achieve... stabilization of the greenhouse gas concentrations in the atmosphere at a level that would prevent
dangerous anthropogenic interference with the climate
system.Ó The signatories agreed to Òformulate, implement,...and... update... programmes containing
measures to mitigate climate change by addressing
anthropogenic emissions by sources and removals by
sinksÓ and to prepare periodic emissions inventories,
promote development and diffusion of technologies for
emissions control, and cooperate in adaptation. In addition, the developed country signatories agreed to Òadopt
national policies and take corresponding measures on
the mitigation of climate changeÓ and to Òcommunicate .
. . detailed information on its policies and measures...
with the aim of returning individually or jointly to their
1990 levels these anthropogenic emissions of carbon
dioxide and other greenhouse gases.Ó The Convention
excludes chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs), greenhouse gases that are
deemed to cause damage to the Earth's stratospheric
ozone and are controlled by the 1987 Montreal Protocol
on Substances that Deplete the Ozone Layer.

The Framework Convention established the Conference
of the Parties to Òreview the implementation of the Convention and... make, within its mandate, the decisions
necessary to promote the effective implementation.Ó In
1995, the first Conference of the Parties met in Berlin and
issued the Berlin mandate, an agreement to Òbegin a
process to enable it to take appropriate action for the period beyond 2000.Ó The second Conference of the Parties, held in Geneva in July 1996, called for negotiations
on quantified limitations and reductions of greenhouse
gas emissions and policies and measures for the third
Conference of the Parties in Kyoto, Japan, in December
1997.

The Climate Change Action Plan

Responding to the Framework Convention, on April 21,
1993, President Clinton called upon the United States to
stabilize greenhouse gas emissions by 2000 at 1990 levels. Specific steps to achieve U.S. stabilization were enumerated in the Climate Change Action Plan (CCAP),6
published in October 1993, which consists of a series of
44 actions to reduce greenhouse gas emissions. The
actions include voluntary programs, industry partnerships, government incentives, research and development, regulatory programs including energy efficiency
standards, and forestry actions. Greenhouse gases
affected by these actions include carbon dioxide, methane, nitrous oxide, hydrofluorocarbons (HFCs), and per-
fluorocarbons (PFCs). At the time CCAP was developed,
the Administration estimated that the actions it enumerated would reduce total net emissions7 of these greenhouse gases in the United States to 1990 levels by 2000.

In addition to the climate-related actions of CCAP, the
Energy Policy Act of 1992 (EPACT), Section 1605(a), provided for an annual inventory of U.S. greenhouse gas
emissions, which is contained in the EIA publication
series, Emissions of Greenhouse Gases in the United States.8
Also, Section 1605(b) of EPACT established the Voluntary Reporting Program, permitting corporations,
government agencies, households, and voluntary
organizations to report to EIA on actions that have
reduced or avoided emissions of greenhouse gases. The
results of the Voluntary Reporting Program are reported
annually by EIA, most recently in Mitigating Greenhouse
Gas Emissions: Voluntary Reporting,9 which reports 1995
activities. Entities providing data to the Voluntary
Reporting Program include some participants in
government-sponsored voluntary programs, such as the

6President William J. Clinton and Vice President Albert Gore, Jr., The Climate Change Action Plan (Washington, DC, October 1993).

7Carbon dioxide is absorbed by growing vegetation and soils. Defining the total impacts of CCAP as net reductions accounts for the
increased sequestration of carbon dioxide as a result of the forestry and land-use actions in the program.

8Energy Information Administration, Emissions of Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington, DC, October 1997).

9Energy Information Administration, Mitigating Greenhouse Gas Emissions: Voluntary Reporting, DOE/EIA-0608(96) (Washington, DC,
October 1997).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Climate Wise and Climate Challenge programs, which
are cosponsored by the U.S. Environmental Protection
Agency and the U.S. Department of Energy to foster
reductions in greenhouse gas emissions by industry and
electricity generators. Voluntary activities for 1996 and
1997 will be available in the fall of 1998.

The Kyoto Protocol

Prior to the third Conference of the Parties, at the June
26, 1997, Earth Summit+5 Conference at the United
Nations, President Clinton pledged U.S. support for
binding emissions targets and announced three initiatives: a pledge of $1 billion over 5 years by the United
States for the development of more energy-efficient and
alternative energy technologies in developing countries;
the strengthening of environmental guidelines for U.S.
companies investing overseas; and a partnership with
private industry to install solar panels on 1 million rooftops in the United States by 2010.

On October 22, 1997, President Clinton proposed that
developed countries should stabilize emissions at 1990
levels between 2008 and 2012, with reductions below
1990 levels in the following 5-year period. He also indicated his support for joint implementation projects and
international emissions trading and declared that participation by developing countries was necessary for the
United States to assume binding obligations. At the
same time, the President announced additional initiatives to address greenhouse gas emissions: a $5 billion
program of tax incentives and research and development spending for energy-efficient and lower-carbon
technologies; the establishment of an emissions trading
system with credit for early reductions; the restructuring
of the electricity industry; and reductions of emissions
from Federal sources. Funding for the program was
increased to $6.3 billion in the Administration's 1999
budget request.

Representatives from more than 160 countries met in
Kyoto on December 1 through 11, 1997. The resulting
Kyoto Protocol established binding emissions targets for
developed nations, relative to their emissions levels in
1990, for an overall reduction of about 5 percent.10
The individual targets for the Annex I countries11 range
from an 8-percent reduction for the European Union
(EU) (or its individual member states) to a 10-percent
increase allowed for Iceland. Australia and Norway also
are allowed increases of 8 and 1 percent, respectively,

while New Zealand, the Russian Federation, and the
Ukraine are held to their 1990 levels. Other Eastern
European countries undergoing transition to market
economies have reduction targets of between 5 and 8
percent. The reduction targets for Canada and Japan are
6 percent and, for the United States, 7 percent. Although
atmospheric concentrations of greenhouse gases ultimately have the potential to affect the global climate, the
Protocol establishes targets in terms of annual emissions.

The greenhouse gases included in the targets are carbon
dioxide, methane, nitrous oxide, hydrofluorocarbons,
perfluorocarbons, and sulfur hexafluoride.12 For the latter three gases, individual nations have the option of
using 1995 as the base year from which to achieve reductions, instead of 1990. The aggregate target is established
using the carbon dioxide equivalent of each of the greenhouse gases. Other greenhouse gases are not limited by
the Protocol, although CFCs and HCFCs are controlled
by the Montreal Protocol. This analysis focuses on carbon emissions from the combustion of energy fuels,
which constituted 83 percent of all U.S. greenhouse gas
emissions in 1990. Carbon dioxide emissions from
sources other than energy use are not included in the
analysis, nor are emissions of the five other gases covered by the Kyoto Protocol; however, reductions in
those gases may lessen the required reductions in
energy-related carbon emissions, as discussed below.

The established targets must be achieved over the period 2008 to 2012, the first commitment period.
Essentially, each country can average its emissions over
that 5-year period to establish compliance, smoothing
out short-term fluctuations that might result from
economic cycles or extreme weather patterns. Each
country must have made demonstrable progress by
2005. No targets are established for the period after 2012,
although lower targets may be set by future Conferences
of the Parties.

Sources of emissions include fuel combustion, fugitive
emissions from fuels, industrial processes, solvents,
agriculture, and waste management and disposal. The
Protocol does not prescribe specific actions to be taken
but lists a number of potential actions, including energy
efficiency improvements, enhancement of carbon-
absorbing sinks, such as forests and other vegetation,
research and development of sequestration technologies, phasing out of fiscal incentives and subsidies that

10The text of the Kyoto Protocol is available at web site www.unfccc.de.

11Australia, Austria, Belgium, Bulgaria, Canada, Croatia, Czech Republic, Denmark, Estonia, European Community, Finland, France,
Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Latvia, Liechtenstein, Lithuania, Luxembourg, Monaco, Netherlands, New Zealand, Norway, Poland, Portugal, Romania, Russian Federation, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine, United Kingdom
of Great Britain and Northern Ireland, and United States of America. Turkey and Belarus are Annex I nations that have not ratified the Convention and did not commit to quantifiable emissions targets.

12Hydrofluorocarbons are a non-ozone-depleting substitute for CFCs; perfluorocarbons are byproducts of aluminum production and
are also used in semiconductor manufacturing; and sulfur hexafluoride is used as an insulator in electrical equipment and in semiconductor
manufacturing.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



may inhibit the goal of emissions reductions, and reduction of methane emissions in waste management and in
energy production, distribution, and transportation.

Several provisions of the Protocol allow for some flexibility in meeting the emissions targets. Net changes in
emissions by direct anthropogenic land-use changes
and forestry activities will also be used in meeting the
commitment; however, these are limited to afforestation, reforestation, and deforestation since 1990. Emissions trading among the Annex I countries is permitted.
No rules for trading are established, however, and the
Conference of the Parties is required to establish principles, rules, and guidelines for trading at a future date.
Joint implementation projects are also allowed among
the Annex I countries, whereby a nation could take emissions credits for projects that reduce emissions or
enhance sinks in other countries. It is specifically indicated that trading and joint implementation are supplemental to domestic actions.

The Protocol also establishes a Clean Development
Mechanism (CDM), under which Annex I countries can
take emissions credits for projects that reduce emissions
in non-Annex I countries, provided that the projects lead
to measurable, long-term benefits. Reductions from
such projects undertaken from 2000 until the first commitment period can be used to assist compliance in the
commitment period. This provision calls for the establishment of an executive board to oversee the projects. In
addition, an unspecified share of the proceeds from the
project activities must be used to cover administrative
expenses and to assist with adaptation those countries
that are particularly vulnerable to climate change.

BankingÑthe carrying over of unused allowances from
one commitment period to the nextÑis allowed; however, the borrowing of emissions allowances from a
future commitment period is not permitted. Under the
Protocol, Annex I countries, such as the nations of
the European Union (EU), may create a bubble or
umbrella to meet the total commitment of all the
member nations. In a bubble, countries agree to meet the
total commitment jointly by allocating a share to each
member. In an umbrella arrangement, the total reduction of all member nations is met collectively through
the trading of emissions rights. There is potential interest in the United States in entering into an umbrella trading arrangement.

Non-Annex I countries have no targets under the Protocol, although it reaffirms the commitments of the Framework Convention by all parties to formulate and
implement climate change mitigation and adaptation
programs and to promote the development and diffusion of environmentally sound technologies and
processes. Developing countries can voluntarily enter
into the Protocol by full amendment of the Protocol.

The Protocol became open for signature on March 16,
1998, for a 1-year period. Under its provisions, it enters
into force 90 days following acceptance of at least 55
Parties, including Annex I countries accounting for at
least 55 percent of the total 1990 carbon dioxide emissions from Annex I nations. Signature by the United
States would need to be followed by Senate advise and
consent to ratification.

There are a number of uncertainties and issues to be
resolved at future Conferences of the Parties. As indicated in the Protocol, rules and guidelines for the
accounting of emissions and sinks from activities related
to agriculture, land use, and forestry activities must be
developed. The specific guidelines may have a significant impact on the level of reductions from other sources
that a country must undertake. This issue was directed
to the IPCC by subsequent climate change talks in Bonn
in June 1998. In addition, rules and guidelines must be
established for emissions trading, joint implementation
projects, and the CDM.

Other issues covered in the Protocol but deferred to subsequent sessions include flexibility for Annex I countries
undergoing transition to market economies, commitments for subsequent periods, climate change adaptation actions, sanctions for failure to meet commitments,
guidelines for the reporting and review of emissions and
sinks, and international cooperation in education,
research and development, and technology transfer.

Emissions Trading

Even before the Kyoto Protocol, many analyses of the
impacts of greenhouse gas emissions reductions have
favored emissions trading programs, including joint
implementation programs, as a means of achieving
emissions reductions. In the United States, the Clean Air
Act Amendments of 1990 (CAAA90) established a trading program for emissions of sulfur dioxide (SO2)by
electricity generators in order to reduce emissions to
fixed specified levels. Permits issued to electricity generators allow them to emit up to a specified level of SO2,
with the total number of issued permits equal to the
national limit on emissions. Generators may reduce
emissions by using lower-sulfur coals, installing scrubbers, or increasing the utilization of cleaner-generating
plants. Generators that reduce emissions below their
allowed levels can sell excess emissions permits, which
can be purchased by other generators for whom it is
more cost-effective to purchase permits at the prevailing
market price than to reduce emissions. Emissions permits can also be banked for future use. Compared with
traditional control programs that mandate specific compliance options or require uniform reductions, this SO2
trading program is credited with reducing the overall
cost of compliance by allowing reductions to be made in
the most cost-effective manner.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Unlike SO2, carbon emissions are primarily an international, rather than domestic, issue. In theory, a similar
trading scheme for carbon emissions could be formulated either internationally or within individual countries to achieve fixed emissions levels. Indeed, the Kyoto
Protocol provides for international emissions trading
but defers the determination of specific guidelines and
rules for establishing an open trading market and managing the international flow of funds for the purchase of
permits. Additional complexities may arise in establishing baseline projections against which to monitor and
verify net emissions reductions, particularly with regard
to the CDM.

Even within the United States, carbon emissions trading
may be more complicated than the current SO2 trading
plan for several reasons. The largest sources of SO2 are a
small number of large coal-burning generation plants.
This makes it relatively easy to monitor their fuel use
and emissions and to build and maintain an allowance
trading system to ensure compliance. In contrast, there
are a large number of entities that emit carbon, including
households, commercial establishments, industrial
facilities, automobiles, trucks, airplanes, ships, and
fossil-fired generating stations. The development and
operation of a monitoring and trading system for carbon
emissions would thus be much more complicated. In
addition, there were technologies available to reduce
SO2 emissions at generation plants at the time the allowance trading program was initiated, and switching to
low-sulfur coal was an option. Although research is
ongoing, there are no readily available pre-or post-
combustion technologies for removing carbon from
fossil fuels (although the high technology sensitivity
case included in this analysis assumes that carbon
sequestration technologies will become available for
electricity generators). Therefore, the options for carbon
reduction are limited to fuel switching to lower-carbon
or carbon-free fuels, efficiency improvements, and reductions in energy demand.

Methodology of the Analysis

In March 1998, the U.S. House of Representatives Committee on Science requested that the EIA perform an
analysis of the Kyoto Protocol, focusing on the impacts
of the Protocol on U.S. energy prices, energy use, and the
economy in the 2008 to 2012 time frame for a number of
emissions targets. (See letters of request in Appendix D.)
The request specified that the analysis use the same reference case assumptions as in AEO98 unless changes in
the assumptions could be justified on the basis of the
ProtocolÑthat is, there should be no changes in assumptions regarding policy, regulatory actions, or funding
of energy or environmental programs, including the
energy-related provisions of the Administration's revenue proposals of February 1998.

Each target in the analysis was to be achieved on average
between 2008 and 2012, phasing in beginning in 2005
and stabilizing at the target level after 2012, although
targets beyond 2012 have not yet been established and
may in fact be more stringent. The Committee indicated
that no new nuclear plants should be allowed, although
economical life extensions of nuclear plants should be
permitted. Construction of new nuclear plants, variations in economic growth, and different assumptions
concerning technology characteristics were all to be analyzed as sensitivities to the target cases.

Numerous studies have been conducted on the topic of
reducing greenhouse gas emissions. They can be clustered into several broad categories. One group of studies
are cost-benefit analyses, which seek to establish an optimal level of either emissions reductions or emissions
prices with a goal of balancing the costs and benefits of
emissions reductions, explicitly accounting for the mitigation of damage as a result of emissions controls. A second category of studies address the cost-effectiveness of
alternative paths for emissions reductions. Assuming a
level of global concentrations of greenhouse gases, these
analyses derive an optimal timing strategy for the imposition of emissions controls.

Other studies are more narrowly focused on the costs of
achieving specific emissions reductions or on the
impacts of policies and technology on emissions levels.
Before the Conference of the Parties in Kyoto, analyses
examined the costs of emissions targets under a variety
of assumptions about the possible level and timing of
the targets. Since the Conference, analyses have focused
on the levels and timing specified in the Kyoto Protocol
and studied the costs of achieving those levels under a
range of assumptions about the international provisions
and other flexibility measures in the Protocol. Some of
those analyses are included in the comparison of results
in Chapter 7. This EIA analysis is among this final category of studies, with more detail on U.S. energy markets
and the economy than other analyses but not addressing
the potential benefits of emissions reductions, optimal
timing, or international trade.

The Protocol includes a number of international provisionsÑincluding international emissions trading, joint
implementation projects, and the CDMÑthat may
reduce the cost of compliance. Because EIA cannot fully
address these aspects of the Protocol at this time, the
analysis focuses on domestic impacts and includes a
range of cases with different levels of energy-related
carbon emissions. Although any impact on the global
climate will likely be caused by atmospheric concentrations of greenhouse gases, the targets in the Kyoto Protocol are in terms of annual emissions. This analysis
addresses the annual emissions targets as specified in
the Protocol.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



The National Energy Modeling System

At the request of the Committee, this analysis uses the
same basic assumptions and methodologies that were
used for AEO98. The projections in AEO98 were developed using the National Energy Modeling System
(NEMS), an energy-economy modeling system of U.S.
energy markets, which is designed, implemented, and
maintained by EIA.13 The production, imports, conversion, consumption, and prices of energy are projected
for each year through 2020, subject to assumptions on
macroeconomic and financial factors, world energy
markets, resource availability and costs, behavioral and
technological choice criteria, costs and performance
characteristics of energy technologies, and demographics. NEMS is a fully integrated framework, capturing the
interactions of energy supply, demand, and prices
across all fuels and all sectors of U.S. energy markets.

Reference case projections are developed annually using
NEMS and published in the Annual Energy Outlook
(AEO). NEMS is also used to analyze the effects of existing and proposed laws, regulations, and standards
related to energy production and use; the impacts of
new and advanced energy technologies; the savings
from higher energy efficiency; the impacts of energy tax
policy on the U.S. economy and energy system; and the
impacts of environmental policies, such as the CAAA90
and regulations on alternative and reformulated fuels.
Special analyses of these and other topics are performed
at the request of the U.S. Congress, other offices in the

U.S. Department of Energy, and other government agencies. Because NEMS provides annual projections, it is
well suited to represent the transitional effects of proposed energy policy and regulation.
Within NEMS, four end-use demand modules represent
energy consumption in the residential, commercial,
industrial, and transportation sectors, subject to fuel
prices, macroeconomic factors, and the characteristics of
energy-using technologies in those sectors. The fuel supply and conversion modules represent the domestic production, imports, transportation, and conversion
processes to meet the domestic and export demand for
coal, petroleum products, natural gas, and electricity,
accounting for resource base characteristics, industry
infrastructure and technology, and world market conditions. The modules of NEMS interact to solve for the economic supply and demand balance for each fuel.

In order to capture regional differences in energy consumption patterns and resource availability, NEMS is a
regional model. The end-use demand for energy is represented for each of the nine Census divisions. The supply and conversion modules use the North American

Electric Reliability Council regions and subregions for
electricity generation; aggregations of the Petroleum
Administration for Defense Districts for refineries; and
production regions specific to oil, natural gas, and coal
supply and distribution.

NEMS incorporates interactions between the energy
system and the economy and between domestic and
world oil markets. Key macroeconomic variables,
including the gross domestic product (GDP), disposable
personal income, industrial output, housing starts,
employment, and interest rates, drive energy consumption and investment decisions. In turn, changes in
energy prices and energy activity affect economic activity, a feedback captured within NEMS. Also, an international energy module in NEMS represents world oil
prices, production, and demand and the interactions
between the domestic and world oil markets. Within this
module, world oil prices and supplies respond to
changes in U.S. demand and production.

Technology Representation in NEMS

A key feature of NEMS is the representation of technology and technology improvement over time. The residential, commercial, transportation, electricity
generation, and refining sectors of NEMS include
explicit treatments of individual technologies and their
characteristics, such as initial cost, operating cost, date of
commercial availability, efficiency, and other characteristics specific to the sector. In addition, for new generating technologies, the electricity sector accounts for
technological optimism in the capital costs of first-of-akind plants and for a decline in the costs as experience
with the technologies is gained both domestically and
internationally. In each of these sectors, equipment
choices are made for individual technologies as new
equipment is needed to meet growing demand for
energy services or to replace retired equipment. In addition, in the electricity generation sector, fossil-fired and
nuclear generating units can be retired before the end of
their useful lives if it is more economical to bring on a replacement unit than to continue to operate the existing
unit.

In the other sectorsÑindustrial, oil and gas supply, and
coal supplyÑthe treatment of technologies is somewhat
more limited due to limitations on the availability of
data for individual technologies. In the industrial sector,
technology improvement for the major processing steps
of the energy-intensive industries is represented by
technology possibility curves of efficiency improvements over time. In the oil and gas supply sector, technology progress for exploration and production
activities is represented by trend-based improvement in

13See Energy Information Administration, The National Energy Modeling System: An Overview 1998, DOE/EIA-0581(98) (Washington, DC,
February 1998), for a summary description. Detailed documentation is available through the National Energy Information Center at
202/586-8800 or on the EIA web site at www.eia.doe.gov.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



finding rates, success rates, costs, and the size of the
resource base. Productivity improvements over time
represent technological progress in coal production.

Because of the detailed representation of capital stock
vintaging and technology characteristics, NEMS captures the most significant factors that influence the turnover of energy-using and producing equipment and the
choice of new technologies. New, more advanced technologies for buildings and equipment are generally
characterized by the technology costs, performance, and
availability, existing standards, and energy prices.
Equipment that does not meet efficiency standards is not
available as a choice.

The relative costs of purchasing and operating different
types of equipment are factored into consumer choices,
which are represented by elasticities and discount rates
derived from the analysis of available data. Within the
residential sector, for example, housing stocks are calculated by region and housing type, using aggregate housing starts from the macroeconomic forecast and
assumed retirement rates. Stocks of energy-using equipment are also tracked, reflecting equipment retirement,
replacements, and new housing starts. Choices for new
equipment and efficiency levels for new houses are
influenced by the characteristics of available technology,
existing standards, energy prices, and consumer preferences as reflected in past decisions. In the end-use sectors, all technology choices are based on the assumption
that future energy prices will remain at the same level as
the prices for the year in which the decision is being
made, this being the most likely representation of how
customer decisions are made. However, in the generation and refining sectors, which are cost minimizers,
capacity expansion decisions include foresight of future
energy prices and demand.

In all sectors, technology improvement occurs even in a
reference case because new, more efficient technology
will be adopted as demand for energy services increases
and existing buildings and equipment are retired. The
characteristics of the technologies include initial dates of
commercial availability of more advanced technologies
as well as changes in efficiency and cost that are
assumed to occur in the future. Higher energy prices
may accelerate the adoption of more efficient technologies. Past improvements in energy efficiency have
resulted in part from efficiency standards that are
included in the analysis; future efficiency standards
assumed are those approved standards with specified
efficiency levels.

The detailed characterization of energy consumption
patterns and technology decisions in NEMS allows for
an explicit representation of the introduction of new
energy-using equipment and the improvement of the

capital stock. Because longer-term forecasting models
typically are not annual models, they tend not to capture
the gradual transition of energy markets, including the
capital stock vintaging and turnover, as NEMS does. In
addition, because of the longer time horizon, longer-
term models tend to have less detailed representations
of energy markets.

Although prices play a role in consumers' decisions on
energy-consuming equipment, there are other factors
that come into play. Consumers tend to make decisions
based on a number of personal preferences and lifestyle
choices, in which energy prices may be only a part of the
decisionmaking process. Preferences for larger televisions or higher horsepower vehicles are examples of factors that may outweigh energy costs. As another
example, in the residential sector, home rental instead of
purchase and frequent moving tend to lower the incentive to invest in more energy-efficient equipment. Information also has a major role in consumer decisions and
will likely continue to do so in the adoption of new, more
advanced technologies. Particularly when a more efficient or alternatively fueled technology carries a significantly higher cost or has different operational
characteristics than more conventional technologies,
information on the benefits of the new technology will
be key to its adoption and penetration. Ultimately, the
success of a given technology will depend not on the
behavior of the marginal consumer, who may be particularly cost-conscious or innovative, but on the behavior of the average consumer, whose decision rests on a
number of considerations.

Technology improvements, even when adopted in the
market, may not necessarily lead to reductions in energy
demand. In the transportation sector, for example, the
use of more advanced technologies that could improve
vehicle efficiency has been offset by increasing demand
for larger and higher horsepower vehicles. To the extent
that energy prices are a factor in consumer decisions,
efficiency improvements may also increase energy
demand. Efficiency gains may lower the cost of driving
or operating other equipment, perhaps encouraging
more travel, larger homes, and purchases of more equipment and increasing the demand for energy services.

New or tightened efficiency standards could also reduce
the demand for energy, but stock turnover would still
limit the speed of penetration. Standards have also been
suggested to encourage the use of renewable fuels for
electricity generation, such as those in the proposed
Electric System Public Benefits Protection Act of 1997,
the proposed Electric Consumers Protection Act of 1997,
and the Administration's proposed Comprehensive
Electricity Competition Act; however, proposed and
possible future standards, legislation, and programs are
not included in the analysis.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



The Annual Energy Outlook 1998

At the request of the Committee on Science, this study of
the impacts of the Kyoto Protocol is based on the reference case assumptions of AEO98. In accordance with the
requirement that the reference case EIA projections be
policy-neutral, the AEO98 projections assume that all
Federal, State, and local law, regulations, policies, and
standards in effect as of July 1, 1997, remain unchanged
through 2020. Potential impacts of pending or proposed
legislation, proposed standards, or sections of existing
legislation for which funds had not been appropriated
are not included in the projections. In general, the
AEO98 projections were prepared using the most current data available as of July 31, 1997.

The AEO98 projections assume continued growth in the

U.S. economy, with GDP growing at an average annual
rate of 1.9 percent through 2020. Additional key factors
underlying the projections are the assumptions concerning world oil markets. Continued technological
improvement in the production of oil and the expansion
of production capability worldwide hold the increase in
the real, inflation-adjusted world oil price to an average
growth rate of 0.4 percent a year. Domestically, with
technological advances in the exploration and production of natural gas, the average annual growth in the
average wellhead price is projected to be 0.5 percent
even with rapid growth in the demand for natural gas.
The average price of coal declines throughout the projection period due to increasing productivity in coal production and the expansion of production from lower-
cost western sources.
AEO98 represents the ongoing restructuring of the electricity industry by assuming lower operating, maintenance, and administrative costs, as noted in the trends of
recent data; early retirements of higher-cost coal-fired
and nuclear power plants; and lower capital costs and
efficiency improvements for coal-and natural-gas-fired
generation technologies. Additional assumptions
include a revised financial structure that features a
higher cost of capital in competitive markets. Specific
restructuring plans are included for those regions that
have announced plans. California, New York, and New
England are assumed to begin competitive pricing in
1998 with stranded cost recovery phased out by 2008.
The provisions of the California legislation for stranded
cost recovery and price caps are incorporated. With
these assumptions and declining coal prices, electricity
prices decline at an average annual rate of 1 percent in
the AEO98 projections.

Electricity generation from nuclear power declines significantly in the projections. About 20 percent of the
nuclear capacity available in 1996 is assumed to be
retired by 2010, with no new plants constructed. It is
assumed that nuclear units would be retired as early as
10 years before the expiration of their operating licenses,

based on utility announcements and on analysis of the
age and operating costs of the units. To offset the decline
of nuclear power and to meet the growth of electricity
demand, coal and natural gas generation expand in the
projections, particularly the gas technologies. The financial assumptions for restructuring weigh against more
capital-intensive projects, such as coal and baseload
renewable technologies.

With decreases or moderate increases in the prices of
energy and continued economic growth, total energy
consumption in AEO98 increases by 1 percent a year on
average through 2020. Consumption in all end-use sectors grows in the projections; however, demand in the
transportation sector increases most rapidly, reflecting
increased travel and slow improvement in the efficiency
of vehicles. Total energy intensity, measured as energy
use per dollar of GDP, declines in the projections at an
average annual rate of 0.9 percent. This rate is considerably less than the 2.3-percent decline in energy intensity
experienced between 1970 and 1986 when rapid price
increases and a shift to less energy-intensive industries
led to rapid energy intensity improvements. On average, energy intensity has been flat between 1986 and
1996. The projected improvement still reflects continued
improvements in energy efficiency that partially offset
increases in the demand for energy services.

As noted earlier, projected carbon emissions from
energy combustion in AEO98 reach 1,803 million metric
tons in 2010, 34 percent above the 1990 level of 1,346 million metric tons, rising to 1,956 million metric tons in
2020. Total emissions are projected to increase at an
average annual rate of 1.5 percent between 1996 and
2010 in the reference case, and per capita emissions also
increase at an average annual rate of 0.7 percent during
that period, as continued economic growth and moderate price increases encourage growth in energy services
and energy consumption. Between 2010 and 2020, efficiency improvements tend to offset continued growth in
the demand for energy services, and per capita emissions nearly flatten. During that period, total emissions
increase at an average rate of 0.8 percent a year. Over the
entire projection period, the slow growth of renewable
technologies and the decline of electricity generation
from nuclear power plants also contribute to the growth
of emissions.

Projections of carbon emissions in AEO98 include EIA's
analysis of the impacts of CCAP for the 31 of the 44
CCAP actions that relate to carbon dioxide emissions
from energy combustion. The analysis does not account
for the remaining actions related to non-energy programs, gases other than carbon dioxide, or forestry and
land use. The analysis of CCAP represents EIA's estimate of the effects of incorporating assumptions concerning behavioral change as a result of CCAP and does
not result in the reductions estimated by the developers

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



of CCAP. The initial estimates of the impacts of the
CCAP actions by the Administration projected stabilization of net greenhouse gas emissions in 2000 at 1990 levels; however, a more recent review and update of CCAP
significantly reduces the expected impact.14 In AEO98,
carbon emissions in 2010 are reduced by about 36 million metric tons as a result of CCAP, compared with the
more recent estimate by the sponsors of about 95 million
metric tons for the energy-related actions in CCAP. Differences between the CCAP impacts estimated by EIA
and by the program sponsors are due primarily to differences in the estimated impacts of voluntary programs;
some estimates by the sponsors that include ongoing
trends that would occur even in the absence of CCAP;
and regulatory actions included by the sponsors but not
included by EIA because they are not yet enacted or
finalized.

The Annual Energy Outlook 1995 (AEO95)15 was the first
AEO to include the impacts of CCAP in the projections.
Even then, the goal of stabilizing carbon emissions in
2000 at 1990 levels seemed unlikely. AEO95 projected
that energy-related carbon emissions would reach 1,471
million metric tons in 2000, a level nearly reached in 1996
when emissions were 1,463 million metric tons. Each
subsequent AEO has raised the estimate of carbon emissions, primarily because of lower price projections that
encourage energy use and reduce the penetration of
renewable sources of energy.

There are several reasons that the target specified by
CCAP for 2000 is unlikely to be realized. First, U.S. economic growth has been slightly higher than assumed at
the time the CCAP programs were formulated. Second,
energy prices have increased at a more moderate rate
than initially assumed in the early 1990s. Both these factors have contributed to higher growth in energy consumption than earlier assumed, leading to higher
emissions levels. Third, the funding for a number of the
CCAP programs is lower than initially requested.
Finally, some voluntary programs have proven less
effective than initially estimated by the Administration.

Carbon Reduction Cases

The Kyoto Protocol specifies that the U.S. target for total
greenhouse gas emissions in the first commitment period will be 7 percent below the level of emissions in 1990.
This analysis focuses on the carbon dioxide emissions
from the use of energy, which constituted 83 percent of
total U.S. greenhouse emissions in 1996 (1,463 million
metric tons of energy-related carbon emissions in the

total greenhouse gas emissions of 1,753 million metric
tons carbon equivalent).

The specific reduction in energy-related carbon emissions that will be required is highly dependent on a
number of factors outside the domestic energy sector.
Programs to reduce emissions of the other five greenhouse gases covered by the Protocol may decrease the
requirement for reductions in carbon dioxide emissions.
Similarly, forestry, agriculture, and land use programs
may also offset some carbon dioxide emissions; however, the rules to account for agriculture and forestry
emissions and sinks have yet to be developed and are
subject to considerable uncertainty. According to a fact
sheet prepared by the U.S. Department of State on January 15, 1998, discussing the Kyoto negotiations, the
method of accounting for sinks and the flexibility to use
1995 as the base year for the synthetic greenhouse gasses
may mean that the reduction would be no more than 3
percent below 1990 levels, based on the Administration's estimates.16 Similar estimates were cited by Dr.
Janet Yellen, Chair, Council of Economic Advisers, in
her testimony before the House Committee on Commerce, Energy and Power Subcommittee, on March 4,
1998.17 Finally, because this analysis does not fully represent international energy markets and other activities,
the potential role of international emissions trading and
the CDM in alleviating U.S. reductions of carbon dioxide
is not directly represented in the analysis. Even those
analyses that do include international trade must make
assumptions about the activities, because the development of guidelines and mechanisms has been deferred.

The success of programs to reduce greenhouse gases at
relatively low costs may depend on the success of international trade of carbon permits, joint implementation
projects, and the CDM. Some analyses of greenhouse gas
reductions that have low costs of compliance assume
that a number of relatively low-cost carbon permits will
be available from Annex I countries with less expensive
opportunities to reduce emissions. Based on EIA's
analysis in IEO98, there may be 165 million metric tons
of carbon permits available from the Annex I countries
in the former Soviet Union in 2010, because of the economic decline of those countries in the 1990s, and additional permits may be available as a result of carbon
reduction projects. The total estimate of such opportunities is highly uncertain, however, and it is also unclear
whether those countries would choose to sell available
permits immediately or bank them for later use as their
economies and populations grow. The potential transaction costs of international trading are also unknown.

14U.S. Departmentof State, Officeof Global Change, Climate Action Report, Department of State Publication 10496 (Washington, DC, July

1997).
15Energy Information Administration, Annual Energy Outlook 1995, DOE/EIA-0383(95) (Washington, DC, January 1995).
16See web site www.state.gov/www/global/oes/fs_kyoto_climate_980115.html.
17See web site www.house.gov/commerce/database.htm.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



The role of developing countries is another area of
uncertainty for international activities. In July 1997, the
Senate unanimously passed the Byrd-Hagel resolution,
sponsored by Senators Robert Byrd of West Virginia and
Chuck Hagel of Nebraska, resolving Òthat the United
States should not be a signatory to any protocol to, or
other agreement regarding, the United Nations Framework Convention on Climate Change... which would
mandate new commitments to limit or reduce greenhouse gas emissions for the Annex I Parties, unless the
protocol or other agreement also mandates new specific
scheduled commitments...for Developing Country
Parties within the same compliance period or would
result in serious harm to the economy of the United
States.Ó18 President Clinton has declared on several
occasions that he will not submit the Protocol for ratification without pledges of meaningful participation by
developing countries. While participation by developing countries may be key to the acceptance of the Protocol, development of specific guidelines and rules for the
international programs has been deferred, including the
means to establish baseline projections and to monitor
and verify emissions reductions.

There is also a possibility that investments to reduce carbon emissions in developing countries could be limited.
First, such bilateral ventures may be viewed as substitutes for or additions to foreign aid, a political concern to
both the United States and developing countries. Also, it
is possible that the continuing discussions about the
implementation of the Protocol will raise the topic of
trade limitsÑrestrictions on the amount of reductions
that any one country can satisfy through international
programs. The Protocol states that such activities are to
be supplemental to domestic actions. In the views of
some countries, there is a potential problem with certain
nations undertaking little internal action.

Because the potential impacts of forestry and agricultural sinks, offsets from other greenhouse gases, international trading, and other international activities are
uncertain, a single target for the required reductions in
energy-related carbon emissions in the United States
cannot be developed at present. This analysis includes a
number of cases, as requested by the Committee, assuming different levels of reductions in energy-related carbon emissions, in order to develop the energy and
economic impacts of achieving those reductions. By
establishing this range of carbon reductions, the analysis
allows others to perform their own analyses of the
impacts of sinks, offsets, and international programs,
derive their own targets for energy-related carbon emissions, and use one of the EIA target cases to assess the
energy and economic impacts of the carbon reductions
in that case.

In addition to a reference case, six targets for reductions
in energy-related carbon emissions are considered.

¥ Reference Case (33 Percent Above 1990 Levels).
This case represents the reference projections of
energy markets and carbon emissions without any
enforced reductions and is presented as a comparison for the energy market impacts in the reduction
cases. Although this reference case is based on the
reference case from AEO98, as specified by the Committee, there are small differences between this case
and AEO98. Some modifications were made in order
to permit additional flexibility in NEMS in response
to higher energy prices or to include certain analyses
previously done offline directly within the modeling
framework, such as nuclear plant life extension and
generating plant retirements. Also, some assumptions were modified to reflect more recent assessments of technological improvements and costs.
Significant changes to NEMS and its assumptions
relative to AEO98 are noted in Appendix A. As a
result of these modifications, the projections of carbon emissions in the reference case for this analysis
are slightly lower than those in the AEO98 reference
caseÑ1,791 million metric tons in 2010 compared
with 1,803 million metric tons. The carbon emissions
projections in the reference case, as well as in all the
carbon reduction cases, include EIA's estimate of the
impacts of CCAP.
¥ 24 Percent Above 1990 Levels (1990+24%). This case
assumes that carbon emissions can increase to an
average of 1,670 million metric tons in the commitment period 2008 to 2012, 24 percent above the 1990
levels, but 122 million metric tons below the average
emissions in the reference case during that period.
¥ 14 Percent Above 1990 Levels (1990+14%). This case
assumes that carbon emissions in the commitment
period average 1,539 million metric tons, which is
approximately the level estimated for 1998 in AEO98
and is 14 percent above 1990 levels. This requires the
average annual carbon emissions between 2008 and
2012 to be reduced by 253 million metric tons.
¥ 9 Percent Above 1990 Levels (1990+9%). This case
assumes that energy-related carbon emissions can
reach an average of 1,467 million metric tons in the
commitment period, 9 percent above 1990 levels, an
average reduction of 325 million metric tons from the
reference case projection.
¥ Stabilization at 1990 Levels (1990). This case
assumes that carbon emissions are stabilized
approximately at the 1990 level of 1,346 million metric tons, averaging 1,345 million metric tons during
18The discussion about the resolution can be accessed in the Congressional Record of July 25, 1997, from web site www.access.gpo.gov/
su_docs/aces/aces150.html.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



the commitment period, a reduction of 447 million
metric tons from the reference case.

¥ 3 Percent Below 1990 Levels (1990-3%). This case
assumes that energy-related carbon emissions are
reduced to an average of 1,307 million metric tons in
the commitment period. A reduction of 485 million
metric tons from the reference case level is required.
¥ 7 Percent Below 1990 Levels (1990-7%). In this case,
energy-related carbon emissions are reduced to an
average of 1,250 million metric tons in the commitment period, a reduction of 542 million metric tons
from the reference case projection. This case essentially assumes that energy-related carbon emissions
must meet the 7-percent target in the Kyoto Protocol
with no net offsets from sinks, other greenhouse
gases, or international activities.
Reductions in both the 1990-3% and 1990-7% cases
would likely come from domestic actions only. The
reductions in the other carbon reduction cases imply
some international trade in carbon permits, CDM activity, or joint implementation projects, but this analysis
does not address the shares that might result from international and domestic actions.

In each of the carbon reduction cases, the target is
achieved on average for each of the years in the first
commitment period, 2008 through 2012, in accordance
with the Kyoto Protocol. The Protocol provides the flexibility for the target to be achieved on average over the 5year commitment period, to accommodate short-term
fluctuations that might occur, such as severe weather or
unanticipated economic growth. Because the Protocol
does not specify any targets beyond the first commitment period, the target is assumed to hold constant from
2013 through 2020, the end of the NEMS forecast horizon. This assumption may be optimistic in that the possibility of further reductions has been advocated.

The target is assumed to be phased in over a 3-year period, beginning in 2005; that is, one-fourth of the reduction
is imposed in 2005, one-half in 2006, and three-fourths in
2007. This analytical simplification allows energy markets to begin adjustments to meet the reduction targets
in the absence of complete foresight, although a longer
or delayed phase-in may lower the adjustment cost.
Phase-in is also consistent with the requirement in the
Protocol that countries achieve demonstrable progress
toward the reductions by 2005; however, reductions
prior to the commitment period are not credited against
the required reductions.

Given the scope and potential costs of compliance with
the reduction targets of the Protocol, there is a possibility
that consumers might react differentlyÑeither taking
more immediate action or waiting. Consumers could
begin to modify their energy decisions even before the

3-year phase-in period, either in anticipation of future
price increases or because of a national commitment to
reduce greenhouse gases. On the other hand, it is possible that consumers could delay actions either until or
beyond energy price changes, taking a cautionary
approach to the magnitude and duration of price
increases or even anticipating a reversal of policy.

Although each of the six reduction cases is modeled
using NEMS, the analysis in this report focuses on three
of the cases, the 1990+24%, 1990+9%, and 1990-3% cases.
Three cases are chosen in order to keep the subsequent
presentation and discussion of the results manageable,
particularly since many of the basic trends are the same
across the reduction cases, varying only in the magnitude of the impact. Where there are specific trends to
note in any of the other cases, they are included in the
appropriate section of this report. The full results of each
of the cases are presented in Appendix B, and results
across all cases are presented graphically, where practical. Any of the reduction targets may be plausible; however, it is likely that some mitigation of the 7-percent
target will be achieved through a combination of offsets
from forestry and agriculture, reductions in other greenhouse gases, international trading, and other flexible
international mechanisms.

Carbon Prices

Each of the carbon reduction targets is achieved by
assuming that a carbon price is applied to the cost of
energy, which could result from a carbon emissions
permit system. The carbon price is applied to each of the
energy fuels at its point of final consumption relative to
its carbon content. Imported energy products receive the
same carbon price at the point of consumption, but no
carbon price is levied on other imported products. Of the
fossil fuels, coal has the highest carbon content. Natural
gas produces about half the carbon emissions of coal per
unit of energy content. Average emissions from
petroleum products are between those for coal and
natural gas. Nuclear generation and renewable fuels
produce no net carbon emissions. As an example, the
carbon emissions factors and energy costs for a
hypothetical carbon price of $100 dollars per metric ton
are shown in Table 1.

Electricity produces no carbon emissions at the point of
use; however, its generation currently produces about
35 percent of the total carbon emissions in the United
States. The carbon price is applied to the fuels used to
generate electricity, and the higher prices are reflected in
the delivered price of electricity.

Placing a value on the carbon released during the combustion of fossil fuels affects energy consumption and
emissions in three ways. First, consumers may reduce
the demand for energy services by driving less, reducing
the use of appliances, or shifting to less energy-intensive

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Table 1. Carbon Emissions Factors for Major Energy Fuels and Calculated 1996 Delivered Energy Prices
With a Carbon Price of $100 per Metric Ton

Parameter Steam Coal Gasoline Natural Gas
Carbon Emissions Factor
(Kilograms of Carbon per Million Btu) .......................... 25.49 19.19 14.40
Average Delivered Price in 1996
(1996 Dollars per Million Btu) ................................ 1.32 9.89 4.13
(1996 Dollars per Fuel Unit)a ................................ 27.52 1.23 4.25

Average Delivered Price With Carbon Price of $100 per Metric Ton
(1996 Dollars per Million Btu) ................................
(1996 Dollars per Fuel Unit)a ................................
3.87
80.68
11.81
1.47
5.57
5.73

aFuel units are short ton (coal), gallon (gasoline), and thousand cubic feet (natural gas).
Source: Office of Integrated Analysis and Forecasting.

goods and services, as examples. Second, more energy-
efficient equipment may be chosen, reducing the
amount of energy required to meet the demand for
energy services. Finally, there may be a shift to noncarbon or less carbon-intensive fuels, reducing the carbon
released per unit of energy consumed.

In the energy market analysis in this report, the carbon
prices represent the marginal cost of reducing carbon
emissions to the specified level or, conversely, the value
of consuming the last metric ton of carbon. Although
there may be a number of easy, low-cost options for
reducing energy use and emissions, higher levels of
reductions will require more expensive investment and
changes in patterns of energy demand. The projected
carbon prices reflect the price that the United States
would be willing to pay to achieve a given emissions
reduction target. The energy market analysis does not
address the international implications of achieving a
particular target at the projected carbon price. In the
absence of modeling international trade of emissions
permits, the energy market analysis makes no link
between the U.S. carbon price and the international
market-clearing price of permits, or the price at which
other countries would be willing to offer permits for sale
in the United States.

Carbon prices, or similar mechanisms, are used by most
analysts in assessing the implementation and impacts of
the Kyoto Protocol or other emissions reduction targets,
such as carbon stabilization. Carbon prices are used
because they effect all three ways of reducing
emissionsÑdemand reduction, improved efficiency,
and fuel switchingÑand may be the most efficient
mechanism. Estimates of the carbon price necessary to
achieve reductions vary widely. Lower estimates are
suggested by those who assume that there are a number
of low-cost options to reduce energy use or to shift to
low-carbon or noncarbon fuels that are readily available
and will be quickly adopted with higher energy prices.
Higher estimates are suggested by analysts who think
that the effective price of carbon-intensive fuels will
have to be raised significantly to encourage changes in
consumer choices and the development of additional
alternative technologies.

The projected energy market costs in this study represent only the marginal cost of reducing energy-related
carbon emissions and do not reflect other costs that
could occur as a result of business cycle fluctuations,
capital constraints, or implementation of emissions
reductions through less efficient mechanisms. No costs
are included for damage or adaption to potential climate
change. In addition, no benefits for avoided damage or
other ancillary benefits are included, unlike some analyses that represent the net cost of emissions reductions,
net of benefits.

Macroeconomic Analysis

EIA analyzes the macroeconomic impacts of the carbon
reduction cases using the Data Resources, Inc. (DRI)
Macroeconomic Model of the U.S. Economy. The DRI
Model is a representation of the U.S. economy with
detailed output, price, and financial sectors, incorporating gradual adjustment of the economy to policy
changes. Macroeconomic models focus on adjustment
processes of the economy associated with changing market conditions, including economic policies. Real-world
economic behavior involves adapting to changes in conditions of supply and demand, which can lead to dislocations and less than optimal use of resources in the
short run. Short-run movements in actual income are
portrayed against projected long-run levels of potential
output.

The linkage between the DRI macroeconomic model and
NEMS is a set of energy variables. Twenty-seven energy
variables in the DRI macroeconomic model are directly
related to similar NEMS variables by ensuring that the
DRI variables show the same percentage change from
the baseline as the NEMS variables. These energy variables include energy prices, energy production, and
energy consumption by different end uses, and the
revenue from auctioned carbon permits. Energy prices
include world oil prices; residential heating oil, electricity, and natural gas prices; transportation fuel prices for
both diesel and gasoline; residual fuel oil prices; average
refined oil price; wellhead natural gas price; and industrial coal and electricity prices. Coal, natural gas, and
crude oil production from NEMS is used in the DRI

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



macroeconomic model as well as the end-use demand
for oil, natural gas, electricity, and coal.

Energy prices and end-use demands for fuels are the key
energy inputs, along with the level of auctioned carbon
revenues, because energy prices affect inflation, and the
end-use fuel demand represents energy in the DRI
aggregate production function, which describes the supply potential of the economy. The amount of auctioned
carbon revenue dictates how much energy consumers
can expect to receive as rebated revenue, which in turn
affects disposable income. Changes in the values of
these variables relative to the reference case would have
major impacts on the macroeconomy.

When a system is developed for the trading of carbon
permits within the United States, a number of initial
decisions must be made: How many permits will be
available? Will they be freely allocated or sold by competitive auction? If they are allocated, how will the initial
allocations be made? If they are sold, what will be done
with the revenues? How many permits will be bought in
international markets? If the permits are traded in a free
market, holders of permits who can reduce carbon emissions at a cost below the permit price will sell their permits, and those with higher costs of reduction will buy
permits, resulting in a transfer of funds between private
parties. If the permits are sold by competitive auction,
there will be a transfer of funds from emitters of carbon
to the Federal treasury. This analysis makes the explicit
assumption that the permits will be sold in a competitive
auction run by the Federal Government.19

The macroeconomic analysis in Chapter 6 considers the
flow of funds overseas that would be represented by
international purchases of carbon permits, explicitly
assuming that the carbon price determined in the NEMS
model is the international price at which permits would
be traded. Although the U.S. target established by the
Protocol is a 7-percent reduction in greenhouse gas
emissions relative to 1990 levels, the method of accounting for sinks and the flexibility to use 1995 as the base
year for the synthetic greenhouse gasses may mean that
the reduction would be no more than 3 percent below
1990 levels, according to the U.S. State Department. The
differences between the reduction level in the 1990-3%
case and the reductions in the cases with higher levels of
energy-related carbon emissions are assumed to be met
by permits purchased in the international market at the
carbon price calculated for each case.

Many analyses of carbon mitigation have used a class of
models that are characterized as computable general
equilibrium (CGE) models. The CGE structure focuses
on the interconnectedness of the economy and calculates
the equilibrium of the economy in the long term,
abstracting from the short-run adjustment processes.
Most often the time horizon of these models is much
longerÑ20, 50, or 100 years into the future. In contrast,
the DRI macroeconomic model used in this analysis
focuses on the adjustment of the economy over time,
allowing for dislocations within the economy that yield
less than optimal levels of economic activity. While climate change can arguably be considered a long-run phenomenon, the policies and measures to induce change
may take effect in a near-term horizon.

Chapter 7 gives a more detailed comparison of the similarities and differences in the alternative model structures and results. Models of both types can contribute to
the assessment of the possible impacts on the economy
of greenhouse gas reduction. However, past analyses of
the issue using CGE and macroeconomic models have
often disagreed with each other over the concepts of the
full employment GDP of the CGE models and the actual
GDP measure presented in the macroeconomic models.
Potential GDP is a concept calculated within the DRI
Model but rarely presented as an output measure. The
discussion in Chapter 6 considers the alternative views
and introduces the concept of potential GDP into the discussion of the economic impacts of the Protocol.

International Energy Markets

The focus of the analysis is U.S. energy markets; however, changes in international markets may have a significant influence on the United States. In particular,
crude oil and petroleum products constitute an international market, and the world price of oil has a strong
impact on consumption and production of oil in the
United States. Conversely, U.S. demand for and production of oil affects the world price of oil. The feedback of

U.S. oil markets on international markets is represented
within the NEMS framework. World oil prices are determined by means of a price reaction function, assuming
that the Organization of Petroleum Exporting Countries
will expand oil production capacity to meet world oil
demand.
For this analysis, it is assumed the other Annex I countries will reduce their consumption of oil in order to help
meet their reduction targets. Each country is assumed to

19A permit auction system is identical to a carbon tax as long as the marginal abatement reduction cost is known with certainty by the
Federal Government. If the target reduction is specified, as in this analysis, then there is one true price, which represents the marginal cost of
abatement, and this also becomes the appropriate tax rate. In the face of uncertainty, however, the actual tax rate applied may over-or undershoot the carbon reduction target. Auctioning of the permits by the Federal Government is evaluated in this report. To investigate a system of allocated permits would require an energy and macroeconomic modeling structure with a highly detailed sectoral breakout beyond
those represented in the NEMS and DRI models. For a comparison of emissions taxes and marketable permit systems, see R. Perman, Y. Ma,
and J. McGilvray, Natural Resources and Environmental Economics (New York, NY: Longman Publishing, 1996), pp. 231-233.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



reduce its oil demand by the same percent that the
United States reduces oil demand from the reference
case level. Oil consumption in non-Annex I countries is
assumed to respond to changes in the world price of oil
with no additional reactions as a result of carbon reduction policies.

Coal exports are a significant portion of U.S. coal production, with exports going to both Annex I and non-
Annex I countries. Because Annex I countries must
reduce carbon emissions, it is assumed that coal production and imports in Western Europe and coal imports in
Japan would be reduced and that coal consumption in
those countries would be reduced by more than their
emissions reductions in the Protocol. In the target cases
where U.S. carbon emissions are allowed to rise above
1990 levels in 2010, U.S. steam coal exports to Europe in
2010 are assumed to be lower by 16 million tons, and
exports to Asia are 4 million tons lower than in the reference case. In the more stringent target cases, exports to
Europe and Asia are 26 and 7 million tons lower, respectively, in 2010.

As a result of the Kyoto Protocol, energy prices in the
Annex I countries may be higher than in the non-Annex I
countries, which do not have emissions reduction targets in the Protocol. As a result, it is possible that more
energy-intensive industries could shift from those countries with higher energy costs. Energy-intensive industries also may face reduced demand as consumers shift
their consumption patterns to less energy-intensive
goods and services. Consequently, the composition of

U.S. industrial output is likely to change toward the less
energy-intensive industries. Because this analysis does
not cover international energy markets, international
trade, or the international activities of the Protocol, a
complete analysis of potential changes in U.S. industrial
output is not possible (for discussion, see the box on
ÒIndustrial CompositionÓ in Chapter 3).
Sensitivity Cases

A number of factors combine to determine the NEMS
projections of energy consumption and carbon emissions. Typically, AEO explores a wide range of cases that
vary the reference case assumptions on economic
growth, world oil markets, technology improvement,
and potential regulatory changes. In this analysis, a variety of sensitivity cases are used to examine the factors
that have the most significant impacts on energy
demand and carbon emissions. With the exception of the
nuclear power sensitivity case, all the sensitivity cases
are analyzed relative to the 1990+9% case.

Low and High Economic Growth

These cases analyze the effects of different assumptions
about U.S. economic growth. The AEO98 reference case

assumes that the output of the Nation's economy, measured by GDP, will increase by an average of 1.9 percent a
year between 1996 and 2020. The same assumption is
used in all the carbon reduction cases in this analysis,
although there is a feedback within the NEMS framework that alters the baseline economic assumptions as a
result of changes in energy prices. Therefore, as emissions reductions become more stringent and the resulting carbon prices become higher, there will be a
reduction in economic growth.

In order to reflect the uncertainty in potential economic
growth, AEO98 included high and low economic
growth cases. The same alternative assumptions are
used in this analysis. The high economic growth case
includes higher population, labor force, and labor productivity, resulting in higher industrial output, lower
inflation, and lower interest rates. As a result, the GDP
increases at an average rate of 2.4 percent a year through
2020. The opposite assumptions in the low economic
growth case lead to an average annual growth rate of 1.3
percent.

Low and High Technology

These sensitivity cases examine the effects of assumptions about the development and penetration of energy-
consuming technologies on the analysis results. The reference cases in this analysis and in AEO98 include continued improvement in technologies for both energy
consumption and productionÑfor example, improvements in building shell efficiencies for both new and
existing buildings; efficiency improvements for new
appliances; productivity improvements for coal production; and improvements in the exploration and development costs, finding rates, and success rates for oil and
gas production. Additional technology improvements
in the end-use demand sectors and in the electricity generation sector could reduce energy consumption and
energy-related carbon emissions below their projected
levels in the reference case. Conversely, slower improvement than that assumed in the reference could raise both
consumption and emissions.

AEO98 presented alternative cases that varied key
assumptions concerning technology improvement and
penetration in the end-use demand and electricity generation sectors. This analysis uses the same low technology assumptions for a low technology sensitivity. In the
residential and commercial sectors, it is assumed that all
future equipment purchases will be made only from the
equipment available in 1998 and that building shell efficiencies will be frozen at 1998 levels. Similarly, in the
transportation sector, efficiencies for new equipment are
fixed at 1998 levels for all travel modes. In the industrial
sector, plant and equipment efficiencies are fixed at 1998
levels. No new advanced generation technologies are
assumed to be available during the projection period.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Technology Improvement in the Reduction Cases and the Sensitivity Cases

In AEO98, energy intensityÑprimary energy con-who is considering the choice of more efficient
sumption per dollar of GDPÑis projected to decline equipment. Perceived consumer preferences are also
by an annual average of 0.9 percent between 1996 and a factor in technology choiceÑfor example, prefer2020. This decline is significant but considerably less ences for larger, higher horsepower vehicles and
than the decline in the 1970s and early 1980s, which larger televisions, and for purchases of new heating
averaged 2.3 percent a year between 1970 and 1986. equipment that uses the same fuel as the equipment it
Approximately half the decline in energy intensity replaces. Improvements in energy intensity can be
during that period resulted from shifts in the econ-slowed by continued growth in energy servomy to service industries and other less energy-icesÑmore travel, household appliances, and office
intensive industries; however, the other half of the equipment, larger homes, and higher industrial out-
decline was due to the use of more energy-efficient putÑsome of which are assumed to respond to
technologies, resulting, in part, from the rapid escala-energy prices.
tion in the price of energy from the mid-1970s

In the carbon reduction cases, energy prices rise with

through the mid-1980s. The decline in energy inten

increasingly stringent reduction targets. Intensity

sity slowed during the late 1980s and early 1990s as

improvements in those cases result both from reduc

the growth in energy prices slowed and growth in

tions in energy service demand and from the choice of

some energy-intensive industries resumed. In the ref-

more efficient equipment as a result of higher prices.

erence case projections, continued modest increases

These cases use the same assumptions of technology

in the price of energy and growing demand for cer

availability and characteristics. Additional research

tain energy services, such as appliances, office equip-

and development in energy-efficient or alternatively

ment, and travel, moderate further declines in energy

fueled technologies would likely expand the slate of

intensity.

choices available to consumers, leading to further
improvements in energy efficiency. The high technol-

Energy intensity improvement results from opposing

ogy case explores the impacts of improvements in the

forces of growth in energy service demand and

availability, characteristics, and costs of technology as

improvement in the stock of energy-using equip-

a result of increased research and development, thus

ment. New, more efficient technology must be devel

separating the impacts of energy prices and technol

oped and available, but it also must be adopted in

ogy development.

order to contribute to energy efficiency improvements. Energy prices play a role in the consumer's Efficiency standards have contributed to past
decision when purchasing new equipment; however, improvements in energy intensity. The Corporate
other factors also influence equipment choice. More Average Fleet Efficiency and National Appliance
advanced, energy-efficient technology is typically Energy Conservation Act of 1987 standards, among
more expensive than standard equipment. The meth-others, are included in the AEO98 reference case;
odology for technology choice accounts for the rela-however, no new efficiency standards or improvetive roles of first cost and energy cost savings over the ments in current standards are assumed. The same
life of the equipment through the use of the discount assumptions are used for all the carbon reduction and
rate, the implied payback period for the consumer sensitivity cases in this analysis.

High technology assumptions were developed specifically for this analysis by experts in technology engineering for each of the energy-consuming sectors,
considering the potential impacts of increased research
and development for more advanced technologies. The
assumptions include earlier years of introduction, lower
costs, high maximum market potential, and higher efficiencies than assumed in the reference case. In addition,
the high technology sensitivity case includes carbon
sequestration technology for coal-and natural-gas-fired
generators to remove carbon dioxide and store it in
underground aquifers. By design, the effect of the high
technology assumptions is distinct from the technology
changes that are induced by the higher energy prices in
the carbon reduction cases. Because the future costs of

the public and private investment that would be needed
to develop and deploy more advanced technologies are
not known, they are not represented in the analysis;
thus, the full economic cost may be understated. It is
possible that further technology improvements could
occur beyond those represented in the high technology
sensitivity case if a very aggressive research and development effort were established. Innovative, breakthrough technologies not foreseen in the analysis of
technology could also be developed and lead to
improvements beyond those represented in the high
technology assessment, but limited time is available for
such technologies to become economically competitive
and achieve significant market share by 2010.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



New Nuclear Capacity

The nuclear power sensitivity case examines the role of
nuclear generation in reducing carbon emissions. In
AEO98, electricity generation from nuclear plants
declines significantly over the forecast period. It is
assumed that 65 units, about 51 percent of the total
nuclear capacity available in 1996, will be retired by
2020. Twenty-four units are assumed to be retired before
the end of their 40-year operating licenses, based on
industry announcements and analysis of the age and
operating costs of the units. No new nuclear plants are
constructed by 2020.

In all the carbon reduction cases, nuclear plants are life-
extended if economical; however, in this sensitivity case,
new nuclear plants can be built if they are economically
competitive with other generating technologies. In the
1990+9% case, nuclear plants are not projected to be economically competitive with other plants. They do
become competitive, however, with the higher carbon
prices projected in the 1990-3% case. Therefore, this sensitivity case is analyzed against the 1990-3% case.

Use of Models for Analysis

The reference case projections in both AEO98 and this
analysis represent business-as-usual trend forecasts,
given known trends in technology and demographics,
current laws and regulations, and the specific methodologies and assumptions used by EIA. Because EIA does
not include future legislative and regulatory changes in
its reference case projections, the projections provide a
policy-neutral baseline against which the impacts of policy initiatives can be analyzed.

Results from any model or analysis are highly uncertain.
By their nature, energy models are simplified representations of complex energy markets. On the other hand,
models provide a structured accounting framework that
allows analysts to capture the interrelationships of a
complex system in a consistent manner. Also, the
assumptions and data underlying a model can be explicitly cited, in contrast to a more ad hoc analysis. The
results of any analysis depend on the specific data,
assumptions, behavioral characteristics, methodologies,
and model structures included. In addition, many of the
factors that will influence the future development of
energy markets are inherently uncertain, including
weather, political and economic disruptions, technology
development, and policy initiatives. Recognizing these
uncertainties, EIA has attempted in this study to isolate

and analyze the most important factors affecting future
carbon emissions and carbon prices. The results of the
various cases and sensitivities should be considered in
terms of the relative changes from the baseline cases
with which they are compared.

It has been suggested that models may be inherently
pessimistic in analyzing the potential impacts of policy
changes. For example, in the Annual Energy Outlook 1993
(AEO93),20 the first EIA analysis of CAAA90 compliance, the cost of a SO2 allowance was projected to be $423
a ton in 2000, in 1996 dollars, rising to $751 a ton in 2010.
Currently, the cost of an allowance is $95 a ton, and
AEO98 projects that the cost will be $121 a ton in 2000
and $189 in 2010. Projected coal prices in AEO98 are 34
and 48 percent lower in 2000 and 2010, respectively, than
those projected in AEO93, reflecting recent improvements in mine design and technology, economies of
scale in the mining industry, and lower transportation
costs induced by rail competition. There has been more
fuel switching to low-sulfur, low-cost Western coal than
previously anticipated (it was initially assumed that
many eastern coal-fired plants would not be able to burn
western coal without considerable loss of performance).
There has also been downward pressure on short-run
allowance costs because generators have taken actions
to comply with the SO2 limitations earlier than anticipated.21 Finally, technology improvements have lowered the costs of flue-gas desulfurization technologies,
or scrubbers, from $313 per kilowatt for scrubber retrofitting as assumed in 1993 to $191 per kilowatt in 1998.
The cost of SO2 compliance was overestimated to a large
extent because compliance relied on scrubbing, a relatively new technology with which there was little experience. On the other hand, the current analysis of carbon
reduction does not rely on a single technology but rather
on fuel switching and efficiency improvements, both
issues of long experience in energy markets.

In contrast, however, analyses of policies can also be too
optimistic. As noted earlier, reductions in greenhouse
gas emissions as a result of CCAP have been overestimated. In addition, some early analyses of the potentially beneficial impacts of price controls on oil and
natural gas proved in error because of the negative
effects on production and competition in the industry.

A number of uncertainties may affect the costs of achieving emissions reductions. As previously noted, the interpretation and implementation of many provisions of the
Kyoto Protocol are undetermined at this time. The flexibility allowed by the international activities may considerably lower the costs of the Protocol.

20Energy Information Administration, Annual Energy Outlook 1993, DOE/EIA-0383(93) (Washington, DC, January 1993).
21A.E. Smith, J. Platt, and A.D. Ellerman, ÒThe Cost of Reducing SO2,Ó Public Utilities Fortnightly (May 15, 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



The availability and costs of technology remain one of
the more significant factors in determining the cost of
emissions reductions, and this analysis seeks to quantify
that uncertainty to some degree with low and high technology sensitivity cases. Although it is sometimes
hypothesized that more cost-effective technologies are
developed once the requirements are established, it
must be noted that the cost and availability of some of
the more advanced technologies in the reference case are
not certain, and even the reference assumptions may be
optimistic.

Although the Kyoto Protocol specifies reduction targets,
signature and ratification by the United States would
need to be followed by the formulation of policies and
programs to achieve the carbon reductions. This analysis has chosen one possible mechanism, the imposition
of a carbon fee with revenue recycling by two alternative
methods. Other programsÑvoluntary initiatives, mandatory standards, or other nonmarket policiesÑcould
result in higher or lower costs. Even with a carbon fee,
other fiscal policies for recycling the revenues, including
not recycling, are likely to have different impacts on the

U.S. economy.
The timing of policy initiatives may also be an important
factor in the cost of emissions reductions. Given that the
Kyoto Protocol includes a specific timetable for reducing
emissions, policies and initiatives that begin earlier may
allow for more gradual adoption and a less costly transition, particularly if consumers react with foresight of
anticipated price increases and emissions restrictions.
Consumer response to anticipated or realized price
increases and other policy initiatives is likely to be
another significant determinant of the cost of the Kyoto
Protocol. Finally, other energy policies formulated for
purposes other than the Protocol, such as electricity
industry restructuring and other emissions controls,
may have ancillary impacts on carbon emissions.

In the next chapter, Chapter 2, the results from the
carbon emissions reduction cases and the sensitivity
cases are summarized. Chapters 3 through 6 present
more detailed analysis of the results for the end-use
demand sectors, electricity generation, fossil fuel
markets, and the macroeconomy, respectively. Chapter
7 concludes with a comparison of this analysis and
similar studies of the costs of carbon emissions
reductions.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



2. Summary of Energy Market Results
This chapter summarizes the energy market results of
the carbon reduction and sensitivity cases evaluating the
effects of the Kyoto Protocol in the National Energy
Modeling System (NEMS). The first set of cases examine
the impacts of six carbon emissions reduction targets,
relative to a reference case without the Kyoto Protocol,
as described in Chapter 1. The remaining cases examine
the sensitivity of those results to variations in key
assumptionsÑthe macroeconomic growth rate, the rate
of technological progress, and the role of nuclear power.
More detailed analyses of the energy market results are
presented in Chapters 3, 4, and 5. The macroeconomic
results are described in Chapter 6. Although the results
of the carbon reduction cases are consistent with the
assumptions made, the projected impacts are subject to
considerable uncertaintyÑparticularly with the more
stringent carbon reduction targetsÑbecause the cases
reflect significant changes in energy markets.

Carbon Reduction Cases
Carbon Prices

Under the Kyoto Protocol, the United States is committed to reducing greenhouse gas emissions to 7 percent
below 1990 levels in the period 2008 through 2012. The
reduction in energy-related carbon emissions that the
United States must achieve to comply with the greenhouse gas reduction target in the Protocol depends on
the level of emissions offsets credited for sinks, reductions in other greenhouse gases, international permit
trading, joint implementation, and the Clean Development Mechanism (CDM). A set of six cases examines a
range of carbon emissions reduction targets, ranging
from 7 percent below 1990 levels, an average of 1,250
million metric tons during the period 2008 to 2012, to 24
percent above 1990 levels, or an average of 1,670 million
metric tons. The most stringent case assumes that the
target of reducing greenhouse gases to 7 percent below
1990 levels is the domestic goal for energy-related carbon emissions, with no offsets from sinks, offsets, international trade, the CDM, or compensating changes in
other greenhouse gases.

The six carbon reduction cases are compared against a
reference case similar to the one published in the Annual
Energy Outlook 1998 (AEO98) (Figure 1). The Protocol
indicates that the greenhouse gas reductions must be

achieved on average in each of the years between 2008
and 2012, and the targets are assumed to hold on
average for that period. At the specification of the
Committee, the targets were held constant after 2012
through the forecast horizon of 2020. To provide energy
markets time to adjust, mandatory carbon reduction
targets were phased in beginning in 2005, the year when
the Protocol indicates that progress toward compliance
must be demonstrated.

199019952000200520102015202005001,0001,5002,000MillionMetricTons1990+24%
1990-3%
Reference1990+9%
1990+14%
19901990-7%
HistoryProjectionsFigure 1. Projections of Carbon Emissions,
1990-2020
Sources: History: Energy Information Administration, Emissions of
Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington,
DC, October 1997). Projections: Office of Integrated Analysis and Forecasting,
National Energy Modeling System runs KYBASE.D080398A, FD24ABV.
D080398B, FD1998.D080398B, FD09ABV.D080398B, FD1990.D080398B,
FD03BLW.D080398B, and FD07BLW.D080398B.
In order to reduce carbon emissions, demand for energy
services must be reduced, more efficient energy-
consuming technologies used, or less carbon-intensive
fuels consumed. Thus, to constrain the overall level of
carbon emissions to a given target, a price on carbon
emissions is included in the delivered price of fuels. The
carbon price is equivalent to the cost of a carbon permit
under a market-based program within the United States
to regulate the overall level of carbon emissions. In such
a program, the purchase of fossil fuels would require the
exchange of carbon permits, and a market for carbon
permits would operate to allocate the overall supply of
permits among U.S. energy consumers. More restrictive
carbon targets would lead to higher market-clearing
prices for carbon.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



In analyzing the carbon emissions reduction targets, the
carbon prices are incorporated as an added cost of consuming energy; that is, as an increase in the delivered
price of energy. The added cost is in direct proportion to
the carbon permit price and the carbon content of the
fuel consumed. As a result, energy consumers face
higher energy costsÑboth for the fossil fuels they consume directly, such as gasoline, and for the indirect use
of fossil energy used to generate electricity. The higher
energy costs also affect the cost of producing goods and
services throughout the economy and, as a result, have
macroeconomic effects beyond the impacts on the
energy sector.

As indicated in Figure 1, some carbon reductions occur
before 2005, based on anticipatory behavior, primarily
as a result of forward-looking capacity planning decisions assumed in the electricity industry. For the electricity industry, where fossil fuel purchases are a
predominant operating cost, planners are assumed to
incorporate future fuel costs in their economic evaluation of generating plant alternatives.22 As a result, some
capacity choices reflected in the reference case before
2005 are altered in the carbon reduction cases based on
carbon prices beginning in 2005, thus lowering carbon
emissions before the assumed start of carbon permit
trading.

Table 2 presents a summary of the key results in 2010
and 2020 for the reference case, the 24-percent-above1990 (1990+24%) case, the 9-percent-above-1990
(1990+9%) case, and the 3-percent-below-1990 (1990-3%)
case. Tables of the complete results for all the carbon
reduction cases are included in Appendix B.

Figure 2 depicts the estimated carbon prices, in constant
1996 dollars, necessary to achieve the carbon emissions
reduction targets. Generally, the highest permit price
occurs early on in the commitment period. The carbon
price declines over time as cumulative investments in
more energy-efficient and lower-carbon equipment,
particularly in the electricity generation industry, tend
to reduce the marginal cost of compliance in later years.

For most of the cases, the trend of carbon prices includes
some relatively minor year-to-year fluctuations. Also,
particularly in the more stringent reduction cases, the
carbon price generally peaks in 2008, the first year of the
commitment period, because of the 3-year phase-in
period. A longer adjustment period might reduce the
price; however, early reductions do not count toward
the required reductions in the commitment period. In
some cases, 1-to 2-year declines in prices occur as

1995200020052010201520200501001502002503003504001996DollarsperMetricTon1990+24%
1990-3%
1990+9%
1990+14%
19901990-7%
Figure 2. Projections of Carbon Prices, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
electricity generators complete construction of low-
carbon replacement plants. The new plants allow
generators to shift from coal to lower-carbon energy
sources, reducing their need to purchase carbon permits
and holding down carbon prices. Because the additions
of replacement capacity occur in discrete amounts, the
year-to-year changes in carbon prices can be somewhat
uneven. The short-term fluctuations in projected carbon
prices are consistent with, but probably understate, the
degree of short-term price movements that would be
expected in a market for carbon permits.

The carbon prices from 2008 to 2012 average $159 per
metric ton in the 1990+9% case, which represents a
carbon reduction averaging 325 million metric tons a
year relative to the reference case (Figure 3). In the more
stringent 1990-3% case, the average carbon price from
2008 to 2012 is $290 per metric ton, achieving an average
annual carbon reduction during that period of 485
million metric tons. In the 1990+24% case, carbon prices
average $65 per metric ton in the compliance period,
with average carbon reductions of 122 million metric
tons.

Carbon prices decline in most of the cases after
2012, despite continued growth in the demand for
energy as the carbon target is held constant. While
increased energy demand would be expected to
exert upward pressure on carbon prices over time,
downward pressure results from the cumulative effect
of investments to improve energy efficiency and
switch to lower-carbon energy sources. These long-lived

22The modeling approach assumes perfect foresight of carbon prices for capacity planning in the electricity industry. Perfect foresight, in
this context, means that the carbon prices that are anticipated during planning are later realized. An algorithm solves for the path of carbon
prices in which anticipated and realized carbon prices are approximately the same, while ensuring that the carbon prices clear the carbon
permit market each year. In the end-use demand sectors, foresight is assumed not to have a material influence on energy equipment decisions, and such decisions are modeled on the basis of prices in effect at the time of the decision.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



investments tend to reduce the demand for carbon per-Thus, although high carbon prices must be sustained
mits over an extended period of time, outweighing the over several years to induce such investments, carbon
opposing effect of moderate growth in energy demand. prices eventually moderate.

Table 2. Summary Comparison: Reference, 1990+24%, 1990+9%, and 1990-3% Cases, 2010 and 2020

2010 2020
Refer1990 1990 1990 Refer1990 1990 1990
Summary Indicators 1996 ence +24% +9% -3% ence +24% +9% -3%

Carbon Price (1996 Dollars per Metric Ton) ......... NA NA 67 163 294 NA 99 141 240
Delivered Energy Price (1996 Dollars per Million Btu)
Coal......................................... 1.32 1.12 2.82 5.24 8.57 1.01 3.50 4.57 7.18
NaturalGas................................... 4.13 3.76 4.71 6.45 8.49 3.96 5.69 6.95 8.30
MotorGasoline................................. 9.89 10.11 11.23 12.53 14.49 10.00 11.45 12.04 13.48
JetFuel ...................................... 5.52 5.62 6.69 8.15 10.24 5.76 7.32 8.01 9.66
DistillateFuel.................................. 7.84 7.81 8.91 10.50 12.71 7.67 9.21 9.79 11.49
Electricity..................................... 20.19 17.22 20.92 25.70 30.68 16.31 21.44 23.77 26.10
Primary Energy Use (Quadrillion Btu)
NaturalGas................................... 22.60 28.97 29.57 31.82 32.49 32.65 34.50 36.02 35.39
Petroleum..................................... 36.01 43.82 42.83 41.12 38.89 46.88 45.25 44.78 42.94
Coal......................................... 20.90 24.14 19.70 11.68 6.72 25.27 15.28 7.06 2.59
Nuclear....................................... 7.20 6.17 6.68 6.98 7.36 3.80 5.06 5.90 6.86
Renewable.................................... 6.91 7.27 7.44 7.72 8.23 7.59 8.29 9.77 11.91
Othera ....................................... 0.39 0.80 0.25 0.25 0.23 0.83 0.26 0.26 0.25
Total........................................ 94.01 111.18 106.48 99.57 93.93 117.02 108.64 103.79 99.94
Electricity Sales (Billion Kilowatthours) ............ 3,098 3,865 3,696 3,492 3,286 4,240 3,972 3,837 3,718
Carbon Emissions by Fuel (Million Metric Tons)
NaturalGas................................... 318 415 424 456 466 468 495 517 507
Petroleum..................................... 621 752 735 704 660 805 777 767 727
Coal......................................... 524 621 506 299 172 652 393 181 66
Total........................................ 1,463 1,791 1,668 1,462 1,300 1,929 1,668 1,468 1,303
Carbon Emissions by Sector (Million Metric Tons)
Residential.................................... 286 337 301 238 199 375 291 224 181
Commercial................................... 230 277 244 186 147 299 225 168 130
Industrial ..................................... 476 559 519 462 418 582 505 449 405
Transportation................................. 471 617 605 576 536 673 647 626 588
Total........................................ 1,463 1,791 1,668 1,462 1,300 1,929 1,668 1,468 1,303
ElectricityGeneration............................ 517 657 567 409 312 726 519 351 246
Carbon Reductions by Sector (Million Metric Tons)
Residential.................................... NA NA 37 99 139 NA 85 151 195
Commercial................................... NA NA 33 91 130 NA 73 131 169
Industrial ..................................... NA NA 41 98 141 NA 77 133 177
Transportation................................. NA NA 12 41 81 NA 26 47 85
Total........................................ NA NA 123 329 491 NA 261 461 625
ElectricityGeneration............................ NA NA 90 248 345 NA 207 375 481
ElectricityGenerationasPercentofTotal............ NA NA 74 75 70 NA 79 81 77
Energy Fuel Expenditures (Billion 1996 Dollars) ..... 560 637 726 834 952 674 807 862 945
Energy Intensity
(Thousand Btu per 1992 Dollar of GDP) ............ 13.57 11.80 11.42 10.78 10.33 10.78 10.05 9.62 9.27
Carbon Intensity
(Kilograms per Million Btu) ....................... 15.6 16.1 15.7 14.7 13.8 16.5 15.4 14.1 13.0

a

Includesnetelectricityimports,methanol,andliquidhydrogen.
NA=notapplicable.
Note:Totalsmaynotequalsumofcomponentsduetoindependentrounding.


Sources:1996: EnergyInformationAdministration,AnnualEnergyOutlook1998,DOE/EIA-0383(98)(Washington,DC,December1997).Projections: OfficeofIntegratedAnalysisandForecasting,NationalEnergyModelingSystemrunsKYBASE.D080398A,FD24ABV.D080398B,FD09ABV.D080398B,andFD03BLW.D080398B.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



0100200300400500600AverageCarbonReductions(MillionMetricTons)
050100150200250300350AverageCarbonPrice(1996DollarsperMetricTon)
1990+24%
1990-3%
1990+9%
1990+14%
19901990-7%
ReferenceFigure 3. Average Annual Carbon Emission
Reductions and Projected Carbon Prices,
2008-2012
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
Energy Prices

With the carbon prices included in the delivered cost of
energy, the prices under the various carbon targets rise
significantly above the reference case. Figures 4, 5, and 6
show the average delivered prices of coal, natural gas,
petroleum, and electricity in the 1990+24%, the 1990+9%
and the 1990-3% cases, respectively. In percentage
terms, coal prices are most affected by the carbon prices,
with the delivered price of coal in the 1990+9% case
increasing 346 to 368 percent above the reference case
price in the 2008 to 2012 period (Figure 7). Natural gas
prices in the 1990+9% case increase 64 to 74 percent
above the reference case prices, and oil prices increase
by 25 to 29 percent. Electricity prices, reflecting the
higher costs of fossil fuels used for generation, as well as
the incremental cost of additional plant investments to
reduce carbon emissions by replacing coal-fired plants,
increase to 47 to 50 percent above the reference case
level.

Compared with the changes in coal and natural gas
prices, the average increase in electricity prices is relatively low. Larger amounts of electricity would be generated from renewable and nuclear power, for which
fuel costs are unaffected by carbon prices. In addition,
cost-of-service electricity pricing is assumed for most of
the country, so that fuel costs would be only a partial
determinant of electricity prices. Nonfuel operating and

maintenance costs and capital equipment costs have a
larger role in setting electricity prices under cost-ofservice pricing. In regions where electricity prices are
assumed to be set competitively on the basis of marginal
costs (California, New York, and New England), carbon
prices would have a more significant influence on electricity prices, particularly when coal-fired plants are the
marginal generators. On the other hand, those regions
are less dependent on coal than are many other areas of
the country.

The pattern of projected delivered energy prices
matches the trend for carbon prices, especially in the
more restrictive carbon reduction cases. In these cases,
the carbon prices become a dominant component of the
delivered cost of fossil energy; however, market forces
continue to play a role in energy prices, especially for
petroleum products. The reduced demand for oil under
the various carbon reduction targets tends to reduce
world oil prices. World oil prices are projected to fall as
demand is reduced in the United States and in other
developed countries that are committed to reducing
emissions under the Kyoto Protocol. In 2010, world oil
prices are projected to be about $20.00 per barrel in the
1990+24% case, $18.70 in the 1990+9% case, and $17.80 in
the 1990-3% case, as compared with $20.80 per barrel in
the reference case. With lower world oil prices, the
change in delivered petroleum product prices with the
various carbon prices is not as high as for natural gas
prices, despite the higher carbon content of petroleum.23

In contrast to petroleum, coal prices are unlikely to be
moderated by competitive forces. Much of the demand
for coal by electricity generators is eliminated in the carbon reduction cases, particularly with the more stringent targets. Coal consumption for other uses, including
industrial steam coal and metallurgical coal, is also
reduced but on a smaller percentage basis than for electricity generation. Although coal produced for export is
also lower in the carbon reduction cases due to lower
demand in the Annex I nations, the change is relatively
small in comparison with the reductions in production
for domestic use. Coal exports, projected at 113 million
short tons in 2010 in the reference case, are 89 million
short tons in 2010 in the 1990+24% and 1990+9% cases
and 76 million short tons in the 1990-3% case. Because
the industrial and export coal markets are served primarily by eastern coal producers, eastern production
declines less in the carbon reduction cases than does production from western mines, which primarily serve the
electricity generation market. Thus, while regional
minemouth prices generally decline in the carbon reduction cases relative to the reference case, the national

23A related factor influencing the effect of carbon prices on gasoline demand is that the price of gasoline already includes Federal and
State excise taxes averaging 37 cents per gallon in 1996, equivalent to a carbon permit price of $155 per metric ton. When additional carbon
permit prices are included in the delivered price of gasoline, the percentage increase in price is not as high as it would be if gasoline were untaxed initially. In turn, the percentage change in gasoline demand due to the carbon price is not as high as it would be if gasoline were not already taxed.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



19952000200520102015202005101520251996DollarsperMillionBtuOilNaturalGasElectricityCoalFigure 4. Average Delivered Prices for Energy
Fuels in the 1990+24% Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD24ABV.D080398B.
199520002005201020152020051015202530351996DollarsperMillionBtuOilNaturalGasElectricityCoalFigure 6. Average Delivered Prices for Energy
Fuels in the 1990-3% Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD03BLW.D080398B.
1995200020052010201520200510152025301996DollarsperMillionBtuOilNaturalGasElectricityCoalFigure 5. Average Delivered Prices for Energy
Fuels in the 1990+9% Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD09ABV.D080398B.
1995200020052010201520200100200300400PercentOilNaturalGasElectricityCoalFigure 7. Projected Changes in Average Delivered
Prices for Energy Fuels in the 1990+9%
Case Relative to the Reference Case,
1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A and FD09ABV.D080398B.
average minemouth price increases because of the shift
in share to the higher-priced coal mined in the East.
Western coal production is also discouraged by higher
rail transportation costs and reduced incentive for the
development of new mines.

Natural gas demand is higher in the carbon reduction
cases relative to the reference case primarily because of
higher use in the electricity generation sector, offsetting
reductions in the end-use demand sectors. As a result,
the average wellhead price of natural gas, excluding any
carbon price, is higher relative to the reference case in all
the carbon reduction cases. The higher wellhead prices
are an indication that greater reliance on natural gas
under the Kyoto Protocol could benefit some domestic
energy producers.

Impacts by Fuel

To meet the required carbon emissions reductions, the
mix of energy fuels consumed would change dramatically from that projected in the reference case
(Figure 8). Relative price changes cause a reduction in
coal and petroleum use, coupled with greater reliance
on natural gas, renewable energy, and nuclear power
(see Figures 9 through 13). Coal, with its high carbon
content and relatively low end-use efficiency, is severely
curtailed in the more stringent cases, replaced by more
use of natural gas, renewable fuels, and nuclear power
in electricity generation. Coal's share of generation is
reduced from 52 percent in 1996 to 42 percent, 26
percent, and 15 percent in 2010 in the 1990+24%,
1990+9%, and 1990-3% cases. By 2020, coal is nearly

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



eliminated from electricity generation in the 1990-3%
case (Figure 9). Some reduction in coal use, compared
with the reference case, occurs before the start of the carbon
permit program in 2005. These changes occur as the
result of anticipatory behavior in the electricity industry,
where capacity planning decisions in advance of 2005
are affected by the prospects of carbon prices in the
future.
Natural gas consumption is higher than in the reference
case, as greater use of natural gas in the generation
sector outweighs the reductions in the residential,
commercial, and industrial sectors (Figure 10). In those
cases with less stringent carbon reduction targets, and
correspondingly lower carbon prices, generators find it
more economical to substitute natural gas for coal than
to invest in renewable technologies. In the more
stringent cases, with high carbon prices, increasing use
of renewable fuels eventually leads to reductions in the
demand for natural gas by generators. This pattern is
reflected in Figure 10, as natural gas consumption in the
more stringent cases falls below that in the less stringent
cases toward the end of the forecast period. In the earlier
portion of the forecast, the rapid growth of natural gas
use exerts pressure on suppliers and distributors to
increase production and pipeline capacity. The ability of
the gas industry to respond to higher demand growth is
discussed in Chapter 5.
Petroleum, used primarily for transportation, is lower
in all the carbon reduction cases (Figure 11). Motor
gasoline demand, accounting for 43 percent of total
24 Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity
Reference
(111.2 Quadrillion Btu)
1990+24%
(106.5 Quadrillion Btu)
1990+9%
(99.6 Quadrillion Btu)
1990-3%
(93.9 Quadrillion Btu)
39.4%
26.1%
5.6%
6.6%
0.7%
21.7%
40.2%
27.8%
6.3%
6.9%
0.3%
41.4%
18.5%
34.6%
7.1%
7.9%
8.7%
0.2%
41.3%
32.0%
11.8%
7.0%
7.7%
0.2%
Other
Renewable
Nuclear
Coal
Natural Gas
Oil
Figure 8. Projections of Fuel Shares of Total U.S. Energy Consumption, 2010
Note: ÒOtherÓ includes net electricity imports, methanol, and liquid hydrogen.
Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.D080398B, and
FD03BLW.D080398B.
1970 1980 1990 2000 2010 2020
0
5
10
15
20
25
30
Quadrillion Btu
1990-3%
1990
1990+24%
Reference
1990+14%
1990+9%
1990-7%
History Projections
Figure 9. Projections of U.S. Coal Consumption,
1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV
.D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
1970 1980 1990 2000 2010 2020
0
5
10
15
20
25
30
35
40
Quadrillion Btu
Reference
1990+24%
1990+14%
1990+9%
1990
1990-3%
1990-7%
Series 8
History Projections
Figure 10. Projections of U.S. Natural Gas
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.


1970 1980199020002010202005101520253035404550QuadrillionBtuReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
HistoryProjectionsFigure 11. Projections of U.S. Petroleum
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
197019801990200020102020012345678QuadrillionBtu1990-3%
19901990+24%
Reference1990+14%
1990+9%
1990-7%
HistoryProjectionsFigure 12. Projections of U.S. Nuclear Energy
Consumption, 1970-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
petroleum consumption in 1996, is lower by 15 percent
in 2010 in the 1990-3% case, by 8 percent in the 1990+9%
case, and by 3 percent in the 1990+24% case than in the
reference case. Consumers respond to higher gasoline
prices by reducing miles driven and purchasing more
efficient vehicles.

Nuclear power, which produces no carbon emissions,
becomes more attractive under carbon reduction targets.
While no new nuclear plants are allowed to be built in
the carbon reduction cases, extending the lifetimes of
existing plants is projected to become more economical
with higher carbon prices. In the reference case,
approximately half of the nuclear capacity now in
operation is expected to be retired by 2020, reducing U.S.
nuclear capacity by 53 gigawatts between 1996 and 2020.
Much of that capacity would be life-extended in the
carbon reduction cases (15 gigawatts, 26 gigawatts, and
38 gigawatts in the 1990+24%, 1990+9%, and 1990-3%
cases, respectively). As a result, the use of nuclear power
for electricity generation is projected to be higher in all
three cases than in the reference case (Figure 12).

Consumption of renewable energy, which results in no
net carbon emissions, is projected to be higher with
carbon reduction targets (Figure 13). Most of the
increase is in electricity generation, primarily with
additions to wind energy systems and an increase in the
use of biomass (wood, switchgrass, and refuse). The
share of generation supplied by renewables increases
from 9 percent in 2020 in the reference case to 11 percent,
15 percent, and 20 percent in the 1990+24%, 1990+9%,
and 1990-3% cases, respectively. Most of the increase in
renewable generation occurs after the 2008-2012
compliance period, reflecting a relatively prolonged

199019952000200520102015202002468101214QuadrillionBtu1990-3%
19901990+24%
Reference1990+14%
1990+9%
1990-7%
HistoryProjectionsFigure 13. Projections of U.S. Renewable Energy
Consumption, 1990-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV
.D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
period of market penetration as renewable technology
costs and performance improve over time.

Electricity generation, which accounted for 35 percent of
energy-related carbon emissions in 1996, is also significantly lower across all the cases (Figure 14). In the
1990-3% case, electricity sales in 2010 are 15 percent
below the reference case projection, with percentage
reductions of about 13 percent occurring in the residential and industrial sectors and about 19 percent in the
commercial sector. The relative changes in electricity

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



199019952000200520102015202008001,6002,4003,2004,0004,800BillionKilowatthoursReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
HistoryProjectionsFigure 14. Projections of U.S. Electricity
Generation, 1990-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, and FD07BLW.
D080398B.
199019952000200520102015202002468101214161820KilogramsCarbonperMillionBtuReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
HistoryProjectionsFigure 15. Projections of U.S. Carbon Emissions
per Unit of Primary Energy
Consumption, 1990-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
sales by sector are similar in the 1990+9% and 1990+24%
cases, but the overall percentage reductions are smaller
(9 percent and 4 percent). One factor mitigating the
response of electricity demand to higher electricity
prices in these sectors is the relative change in energy
prices. For example, the percentage changes in
electricity prices, relative to the reference case, are
smaller than the changes in natural gas prices. With a
smaller percentage price increase, electricity becomes
relatively attractive in those end uses where it competes
with natural gas, such as home heating.

As the results have indicated, reductions in carbon emissions are also met through substitution away from
carbon-intensive fuels, not just through energy efficiency improvements and reductions in energy services.
The degree to which this occurs is indicated by the
change in aggregate carbon intensity of energy use, or
carbon emissions per unit of energy consumption. For
example, natural gas has a carbon intensity at full combustion of 14.5 kilograms per million Btu, whereas coal
averages about 25.7; thus, switching from coal to natural
gas tends to reduce carbon intensity. Aggregate carbon
intensity declined from 16 kilograms per million Btu in
1990 to 15.6 in 1996, but it is projected to increase in the
reference case after 2000, reaching a level of 16.1 kilograms per million Btu by 2010 (Figure 15), even though
energy intensity continues to decline. In the carbon
reduction cases, carbon intensity begins to decline with
the phase-in of the carbon targets. By 2010, carbon intensity declines to 15.7 kilograms per million Btu in the
1990+24% case, 14.7 in the 1990+9% case, and 13.8 in the
1990-3% case.

Sectoral Impacts

Energy demand across each of the end-use sectorsÑresidential, commercial, industrial, and transportationÑwill respond to different degrees to the incentives
imposed by a carbon permit price. In all sectors, however, consumers will have greater incentive to conserve
energy, switch to lower-carbon energy sources, and
invest in more energy-efficient technologies.

Figure 16 illustrates the contribution of each sector
toward meeting the carbon reduction goals in 2010
under three of the cases. The residential and industrial
sectors (including electricity losses) account for the
greatest carbon reduction, and transportation accounts
for the least. As shown in Figure 16, most of the carbon
reductions for the four end-use sectors occur in electricity, stemming from both reduced electricity demand and
the use of more efficient, less carbon-intensive sources of
generation. Reductions in carbon emissions from electricity generation account for about 75 percent of the
total carbon reductions in both the 1990+24% and
1990+9% cases in 2010, and for about 70 percent in the
1990-3% case. A variety of factors contribute to the central role played by the electricity sector in meeting the
carbon reduction targets: the industry's current dependence on coal; the availability and economics of technologies to switch from coal to less carbon-intensive energy
sources; and the comparative economics of fossil-fuel
switching in other sectors, particularly at lower carbon
prices. As discussed in more detail in Chapter 3, the
extent to which end-use energy consumers respond to
prices is often limited by institutional factors.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



ResidentialCommercialIndustrialTransportationResidentialCommercialIndustrialTransportationResidentialCommercialIndustrialTransprtation04080120160MillionMetricTonsElectricNon-Electric1990+24%1990+9%1990-3%
Figure 16. Projected Reductions in Carbon
Emissions by End-Use Sector Relative
to the Reference Case, 2010
Note: Electricity emissions are from the fuels used to generate elec-
tricity and are attributed to the sectors relative to their shares of total
electricity consumption.
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
In the industrial sector, some of the carbon reductions
can be attributed to reductions in manufacturing output
that result from the impact of higher energy prices on the
economy. In addition, industrial firms respond by
replacing productive capacity faster, investing in more
efficient technology, and switching to less carbon-
intensive fuels. Improvements in efficiency are
indicated by reductions in energy intensity, as measured
by the energy use per dollar of gross domestic product
(GDP). In 2010, industrial energy intensity is reduced
from 4.2 million Btu per dollar of GDP in the reference
case to 4.1 million Btu per dollar in the 1990+24% case,

4.0 million Btu per dollar in the 1990+9% case, and 3.9
million Btu per dollar in the 1990-3% case (Figure 17).
Taking into account fuel switching and efficiency
improvements, carbon emissions per unit of GDP in
2010 for the industrial sector are reduced from 60
kilograms per thousand dollars of GDP in the reference
case to 55, 50, and 46 kilograms per thousand dollars of
GDP in the 1990+24%, 1990+9%, and 1990-3% cases,
respectively.
Carbon reductions in the transportation sector occur
primarily as the result of reduced travel and the
purchase of more efficient vehicles in response to higher
energy prices. Compared with the reference case, light-
duty vehicle travel (cars, vans, pickup trucks, and sport-
utility vehicles) in 2010 is lower by 1 percent in the
1990+24% case, by 5 percent in the 1990+9% case, and by
11 percent in the 1990-3% case (Figure 18). At the same
time, more efficient cars and light trucks are purchased,
raising overall fleet efficiency (Figure 19). In 2010, the
average fuel efficiency for the light-duty vehicle fleet is
20.7, 21.2, and 21.5 miles per gallon in the 1990+24%,

1995200020052010201520200.00.51.01.52.02.53.03.54.04.55.0ThousandBtuper1992DollarofGDPReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
Figure 17. Projections of U.S. Industrial Energy
Intensity, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
19952000200520102015202005001,0001,5002,0002,5003,0003,500BillionVehicle-MilesTraveledReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
Figure 18. Projections of U.S. Light-Duty Vehicle
Travel, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B..
1990+9%, and 1990-3% cases, respectively, compared
with 20.5 miles per gallon in the reference case. The
results of those increases are reductions of 3 percent, 8
percent, and 15 percent, respectively, from the reference
case level of motor gasoline demand in 2010 (Figure 20).
Travel reductions and efficiency improvements also
occur in the air and freight sectors, further reducing
carbon emissions. Overall, transportation energy
consumption in 2010 is lower by 2 percent in the
1990+24% case, by 6 percent in the 1990+9% case, and by
12 percent in the 1990-3% case, than in the reference case.

In the residential and commercial sectors, higher energy
prices encourage investments in more efficient equipment and building shells and also reduce the demand

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1995200020052010201520200510152025MilesperGallonReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
Figure 19. Projections of Average Fuel Efficiency
for the Light-Duty Vehicle Fleet,
1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
199520002005201020152020020406080100120140160BillionGallonsReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
Figure 20. Projections of U.S. Motor Gasoline
Consumption, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
for energy services. In the residential sector, delivered
energy use per household in 2010 drops by 4 percent in
the 1990+24% case, 10 percent in the 1990+9% case, and
15 percent in the 1990-3% case compared with the reference case. Energy consumption for space conditioning
accounts for 59 to 62 percent of the change in the three
cases. Those energy services for which appliance efficiency standards are already in place, such as for refrigerators and freezers, are not expected to change greatly
in the carbon reduction cases, because the standards
reflect very efficient technology that already reduces
fuel consumption substantially in the reference case. The
fastest-growing segment of residential electricity consumption, categorized as miscellaneous and including a
variety of appliances such as computers and VCRs,
accounted for approximately 22 percent of residential
electricity consumption in 1996. Relative to the reference
case, miscellaneous electricity consumption per household is lower by 5 percent in 2010 in the 1990+24% case,
by 10 percent in the 1990+9% case, and by 14 percent in
the 1990-3% case.

The energy demand response is somewhat stronger in
the commercial than in the residential sector. Overall,
delivered energy use per square foot of commercial
floorspace in 2010 drops by 5 percent in the 1990+24%
case, 13 percent in the 1990+9% case, and 21 percent in
the 1990-3% case. As in the residential sector, significant
energy reductions are projected for heating, cooling, and
ventilation (29 to 31 percent of the change in the three
cases); however, more than half the energy reduction
comes from more efficient lighting and office equipment
and in the category of miscellaneous electricity uses,

including such appliances as vending machines and
telecommunications equipment.

The electricity generation sector is expected to respond
strongly to the incentives imposed by a carbon price.
Generation from coal, which currently accounts for
more than half of all electricity, drops significantly as the
cost of coal to generators increases by factors of 3 to 8
times the reference case level in 2010. To replace coal
plants, generators build natural-gas-fired combined-
cycle plants, extend the life of existing nuclear plants,
and dramatically increase the use of renewables, particularly biomass and wind energy systems, which
become economical once a carbon price is imposed.
These changes, coupled with the expected reduction in
electricity demand, result in carbon emissions of 567
million metric tons in the 1990+24% case, 409 million
metric tons in the 1990+9% case, and 312 million metric
tons in the 1990-3% case. In comparison, actual 1990
emissions in the electricity generation sector are estimated at 477 million metric tons. The issues related to
plant capacity changes in the electricity industry are discussed in detail in Chapter 4.

The mix of fuels used for electricity generation is projected to change rapidly as new plants come on line (Figures 21, 22, and 23). In the aggregate, cumulative
investments by generators to reduce carbon emissions
tend to bring down the carbon price over time. A slowdown in most new plant additions occurs at the end of
the initial compliance period in 2012, but the growth in
renewable capacity continues throughout the forecast
horizon.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1996200020052010201520200510152025303540QuadrillionBtuRenewable/
OtherHydropowerOilNuclearNaturalGasCoalFigure 21. Projected Fuel Use for Electricity
Generation by Fuel in the 1990+24%
Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD24ABV.D080398B.
1996200020052010201520200510152025303540QuadrillionBtuRenewable/
OtherHydropowerOilNuclearNaturalGasCoalFigure 23. Projected Fuel Use for Electricity
Generation by Fuel in the 1990-3% Case,
1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD03BLW.D080398B.
1996200020052010201520200510152025303540QuadrillionBtuRenewable/
OtherHydropowerOilNuclearNaturalGasCoalFigure 22. Projected Fuel Use for Electricity
Generation by Fuel in the 1990+9%
Case, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run FD09ABV.D080398B.
Sensitivity Cases

Among the sources of uncertainty in the effects of carbon mitigation polices over the next 20 years are the
assumed rate of economic growth, the speed of adoption
of advanced technologies, and the role of nuclear power.
A series of sensitivity cases illustrate how these factors
influence the results of the carbon reduction cases. The
sensitivity cases were analyzed against the 1990+9%
case. The nuclear power sensitivity case was analyzed
against the 1990-3% case, because new nuclear power
plants were found to be economical only with the higher
carbon prices in that case.

Because each of the sensitivity cases is constrained to the
same level of carbon emissions as the case to which it is
compared, the primary impact is not on the carbon emissions levels, or even aggregate energy consumption, but
rather on the carbon prices required to meet the emissions target. For example, in the high technology case,
with an emissions reduction target of 9 percent above
1990 levels, projected carbon emissions during the compliance period are the same as in the corresponding reference technology case (1990+9%) with emissions at the
same level. What differs is the cost of meeting the target,
as reflected in the required carbon price or in expenditures for energy services. As a result, the carbon price
and energy expenditures are the primary measures by
which the sensitivity cases are compared in this report,
in contrast to the presentation of similar sensitivities in
AEO98. Because the technology sensitivities in the AEO
typically are run with energy prices and macroeconomic
assumptions held constant and without any target for
carbon emissions, sensitivities are normally compared
on the basis of levels of energy consumption.

Macroeconomic Growth

The assumed rate of economic growth has a strong
impact on the projection of energy consumption and,
therefore, on the projected levels of carbon emissions. In
AEO98, the high economic growth case includes higher
growth in population, the labor force, and labor productivity, resulting in higher industrial output, lower inflation, and lower interest rates. As a result, GDP increases
at an average rate of 2.4 percent a year from 1996 to 2020,
compared with a growth rate of 1.9 percent a year in the
reference case. With higher macroeconomic growth,
energy demand grows more rapidly, as higher manufacturing output and higher income increase the demand
for energy services. In AEO98, total energy consumption

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



in the high economic growth case is 117 quadrillion Btu
in 2010, compared with 112 quadrillion Btu in the reference case. Carbon emissions are 80 million metric tons,
or 4 percent, higher than the reference case level of 1,803
million metric tons.

Assumptions of lower growth in population, the labor
force, and labor productivity result in an average annual
growth rate of 1.3 percent in the AEO98 low economic
growth case between 1996 and 2020. With lower economic growth, energy consumption in 2010 is reduced
from 112 quadrillion Btu to 107 quadrillion Btu, and carbon emissions are 90 million metric tons, or 5 percent,
lower than in the reference case. Thus, the effect of
higher or lower macroeconomic growth can have a significant impact on the ease or difficulty of meeting the
carbon targets.

To reflect the uncertainty of potential economic growth,
high and low economic growth sensitivity cases were
analyzed against the 1990+9% case, using the same
higher and lower economic growth assumptions as in
AEO98. With higher economic growth, the industrial
output and energy service demand are higher. As a
result, carbon prices must be correspondingly higher to
attain a given carbon emissions target. With low economic growth, the effects are reversed, leading to lower
carbon prices. In addition to industrial output, some of
the most important economic drivers in NEMS are disposable personal income, housing stock, housing size,
commercial floorspace, industrial output, light-duty
vehicle sales, and travel.

Figure 24 shows the effect of the high and low
macroeconomic growth assumptions on the projections
for 2010 in the 1990+9% case. The carbon price in 2010 is
$215 per metric ton in the high economic growth case, or
$52 per metric ton higher than the price of $163 per
metric ton in the 1990+9% case with reference economic
growth. In the low economic growth case, the carbon
permit price in 2010 is $128 per metric ton or $35 per
metric ton lower than in the 1990+9% case.

The higher carbon prices necessary to achieve the carbon
reductions with higher economic growth will tend to
moderate the growth rates of the economy as a whole
and the economic drivers in the energy system. Despite
this price effect, total energy consumption in 2010 is
higher with higher economic growth, by 2.2 quadrillion
Btu relative to the 1990+9% with reference economic
growth. Similarly, the lower economic growth assumption results in lower carbon prices, which offset a portion
of the projected reduction in energy consumption that
would otherwise be expected when economic growth
slows. Lower economic growth lowers total energy consumption by 2.2 quadrillion Btu.

To meet a carbon reduction target with higher economic
growth and energy consumption, there is a shift to less

128163215243163121LowGrowthReferenceGrowthHighGrowthLowTech-
nologyReferenceTech-
nologyHighTech-
nology0501001502002501996DollarsperMetricTonFigure 24. Projected Carbon Prices in the 1990+9%
High and Low Economic Growth and
High and Low Technology Sensitivity
Cases, 2010
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs FD09ABV.D080398B, LMAC09.D080698A, HMAC09.
D080598A, FREEZE09.D080798A, and HITECH09.D080698A.
carbon-intensive fuels and higher energy efficiency;
however, economic growth affects energy consumption
in the industrial and transportation sectors more significantly than in the other end-use sectors. With higher economic growth, renewable energy and natural gas
consumption is higher, primarily for generation but also
in the industrial sector. Coal use for generation is lower,
and more nuclear capacity is life-extended as a result of
the higher carbon prices. Petroleum consumption is also
higher with higher economic growth, in both the transportation and industrial sectors. As shares of total
energy consumption, natural gas and renewables are
higher with higher economic growth, coal is lower, and
nuclear and petroleum remain approximately the same.
Opposite trends for fuel consumption and fuel shares
are seen when lower economic growth is assumed.

Total energy intensity is lower in the high economic
growth case, partially offsetting the changes in energy
consumption caused by the different growth assumptions. There are three reasons for the improvement in
energy intensity. First, although demand for energy
services is higher with higher economic growth, there is
greater opportunity to turn over and improve the stock
of energy-using technologies. In the AEO98 cases, aggregate energy efficiency in the high economic growth case
decreases at a rate of 1.0 percent a year through 2020,
compared with 0.9 percent in the reference case and 0.8
percent in the low economic growth case. Second, with
higher carbon prices, additional efficiency improvements are induced by higher energy prices. Finally, the
higher energy prices lead to some reductions in energy
service demand, moderating the impacts of higher economic growth. In the 1990+9% carbon reduction case,
aggregate energy intensity declines at an average annual
rate of 1.6 percent through 2010. In the 1990+9% high

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



economic growth sensitivity case, the annual decline
increases to 1.9 percent. In the 1990+9% low economic
growth case, the decline in energy intensity slows to 1.3
percent per year.

Technological Progress

The assumed rate of development and penetration of
energy-using technology has a significant impact on
projected energy consumption and energy-related carbon emissions. Faster development of more energy-
efficient or lower carbon-emitting technologies than
assumed in the reference case could reduce both consumption and emissions; however, because the AEO98
reference case already assumes continued improvement
in both energy consumption and production technologies, slower technological development is also possible.

To examine the influence of technology improvement,
two sensitivity cases were analyzed relative to the
1990+9% case. The high technology case includes more
optimistic assumptions on the costs, efficiencies, market
potential, and year of availability for the more advanced
generating and end-use technologies, assuming
increased research and development activity. This sensitivity case also assumes a carbon sequestration technology for coal-and natural-gas-fired electricity generation,
which would capture the carbon dioxide emitted during
fuel combustion and store it in underground aquifers;
however, use of the technology is not projected to be economical relative to other technologies within the time
frame of this sensitivity case because of high operating
costs and storage difficulties. The low technology case
assumes that all future equipment choices are made
from the end-use and generation equipment available in
1998, with building shell and industrial plant efficiencies
frozen at 1998 levels.

Because faster technology development makes
advanced energy-efficient and low-carbon technologies
more economically attractive, the carbon prices required
to meet carbon reduction levels are reduced. Conversely, slower technology improvement requires
higher carbon prices (Figure 24). In the 1990+9% case
with high technology assumptions, the carbon price in
2010 is $121 per metric tonÑ$42 per metric ton lower
than the price of $163 per metric ton in the 1990+9% case
with reference technology assumptions. With the low
technology assumptions, the projected carbon price is
$243 per metric ton in 2010.

Total energy consumption in 2010 is lower by 2.1 quadrillion Btu in the high technology case, about 2 percent
below the projection in the 1990+9% case, and average
energy prices, including carbon prices, are 10 percent
lower. As a result, direct expenditures on energy are 13
percent lower in the high technology case. Demand in
both the industrial and transportation sectors is lower as
efficiency improvements in industrial processes and

most transportation modes outweigh the countervailing
effects of lower energy prices. In the residential and
commercial sectors, the effect of lower energy prices balances the effect of advanced technology, and consumption levels are at or near those in the 1990+9% case. With
the high technology assumptions in the generation sector, coupled with the lower carbon permit price, coal use
for generation is 3.8 quadrillion Btu higher than the 9.7
quadrillion Btu level associated with reference technology assumptions.

In the low technology case, the converse trends prevail.
In 2010, total consumption is higher by 1.5 quadrillion
Btu with the low technology assumptions, and energy
expenditures are 17 percent higher. Industrial and transportation demand is higher, and residential and commercial demand lower, suggesting that industry and
transportation are more sensitive to technology changes
than to price changes, and that the residential and commercial sectors are more sensitive to price changes. With
the higher carbon prices in the low technology case, coal
use is further reduced in the generation sector, with
more natural gas, nuclear power, and renewables used
to meet the carbon reduction targets.

Nuclear Power

In the AEO98 reference case, nuclear generation declines
significantly, because 52 percent of the total nuclear
capacity available in 1996 is expected to be retired by
2020. A number of units are retired before the end of
their 40-year operating licenses, based on industry
announcements and analysis of the age and operating
costs of the units. In the carbon reduction cases, life
extension of the plants can occur, if economical, and
there is an increasing incentive to invest in nuclear plant
refurbishment with higher carbon prices; however, no
construction of new nuclear power plants is assumed,
given continuing high capital investment costs and institutional constraints associated with nuclear power.

A nuclear power sensitivity case was developed to
examine the potential contribution of new nuclear plant
construction to carbon emissions reductions, assuming
that new nuclear capacity would be built when it was
economically competitive with other generating technologies. In the nuclear power sensitivity case, electricity generators were assumed to add nuclear power
plants when it became economical to do so. In addition,
the reference case assumptions about higher costs
incurred for the first few advanced nuclear plants were
relaxed by reducing the premium in costs for the first
phase of new nuclear plant additions.

In the 1990+9% case, even with the nuclear power sensitivity assumptions, nuclear plants are not competitive
with fossil and renewable plants. In the 1990-3% case,
however, when the new nuclear assumptions are used,
1 gigawatt of new nuclear capacity is added by 2010, and

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



41 gigawatts, representing about 68 new plants of 600
megawatts each, are added by 2020. (In a trial case in
which first-generation cost premiums were left unchanged, only 3 gigawatts of nuclear capacity was
added.) The availability of this no-carbon capacity offsets about 25 million metric tons of carbon emissions
from additional natural gas plants in 2020; on the other
hand, more coal is used, because the projected carbon
prices are lower. Most of the impact from the new
nuclear plants comes after the commitment period of
2008 through 2012. As a result, there is little impact on
carbon prices in 2010. By 2020, however, carbon prices

are $199 per metric ton with the assumption of new
nuclear plants, as compared with $240 per metric
ton in the 1990-3% case with the reference nuclear
assumptions.

In the 1990-3% case, total energy consumption is about
the same in 2010 with new nuclear plants allowed and
higher by about 1.8 quadrillion Btu in 2020. Somewhat
lower energy prices induce higher consumption in all
sectors, and the greater availability of carbon-free
nuclear generation allows the carbon reduction target to
be met with higher end-use consumption.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



3. End-Use Energy Demand
Background

This chapter provides in-depth analyses of the carbon
emissions reduction cases for the four end-use demand
sectorsÑresidential, commercial, industrial, and transportation. Additional analyses are included for a
number of alternative cases, including low and high
technology sensitivity cases, which have the most direct
impacts on energy end use.

Primary and Delivered Energy
Consumption

In each of the reduction cases, carbon emissions are
reduced through a combination of switching to carbon-
free or lower-carbon fuels, reductions in energy services,
and increased energy efficiency. The latter two options
lower total energy consumption (Figure 25).

Electricity generation typically consumes about three
times as much energy, on the basis of British thermal
units (Btu), as is contained in the electricity delivered to
final consumers. In AEO98, total delivered energy
consumption in 1996 is estimated at 70.4 quadrillion Btu,
compared with total primary energy consumption of

94.0 quadrillion Btu (Table 3). The difference comes from
electricity-related generation and transmission losses
and, consequently, is relatively small for the transportation sector, where little electricity is consumed.
Although the delivered price of electricity per Btu
generally is more than three times the delivered price of
other energy sources, the convenience and efficiency of
electricity use outweigh the price difference for many
applications.
Because consumers base their fuel and equipment
choices on performance at the point of use, the analysis
of end-use energy consumption presented in this
chapter focuses on energy delivered to final consumers.
When consumers choose to purchase a particular type of

1990199520002005201020152020020406080100120QuadrillionBtuReference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
Figure 25. Projections of Primary Energy
Consumption, 1990-2020
Sources: History: Energy Information Administration, Annual Energy Review
1997, DOE/EIA-0384(97) (Washington, DC, July 1998). Projections: Office of
Integrated Analysis and Forecasting, National Energy Modeling System runs
KYBASE.D080398A, FD24ABV.D080398B, FD1998.D080398B, FD09ABV.
D080398B, FD1990.D080398B, FD03BLW.D080398B, FD07BLW.D080398B.
energy-consuming equipment or to use a particular fuel,
their decisions are based on the cost and performance
characteristics of the technology, mandated efficiency
standards, and energy prices. End-use energy prices
include all the direct costs of providing energy to the
point of use.

The distinction between end-use and primary energy
consumption is an important one for the evaluation of
efficiency standards and other energy policies. Reducing electricity demand through the use of more efficient
technologies reduces primary energy consumption by a
factor of three. In addition, although electricity at its
point of use produces no carbon emissions, reductions in
electricity use produce savings in emissions from the
fuels used for its generation.

Table 3. Primary and End-Use Energy Consumption by Sector, 1996

Sector
End-Use Consumption Primary Consumption
Quadrillion Btu Percent of Total Quadrillion Btu Percent of Total

Residential........................... 11.1 16 19.4 21
Commercial.......................... 7.5 11 15.0 16
Industrial ............................ 27.1 38 34.8 37
Transportation........................ 24.7 35 24.9 26
Total .............................. 70.4 100 94.0 100


Source: Energy Information Administration, Annual Energy Outlook 1998, DOE/EIA-0383(98) (Washington, DC, December 1997).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Integrated Energy Market Analysis

The analysis in this report is a fully integrated analysis of

U.S. energy markets, representing the interactions of
energy supply, demand, and prices across all fuels and
sectors. For example, initiatives to lower energy consumption may lower the prices of the energy supplied,
causing some offsetting increase in energy consumption. An integrated market analysis can capture such
feedback effects, which may be missed in an analysis
that focuses on end-use demand for energy without
accounting for impacts on energy prices.
The Energy Information AdministrationÕs Annual
Energy Outlook 1998 (AEO98), includes results from a
number of alternative sensitivity cases in addition to its
reference case projections. Sensitivity cases generally are
designed by varying key assumptions in one of the
demand, conversion, or supply modules of the National
Energy Modeling System (NEMS), in order to isolate the
impacts of the revised assumptions. For example, the
high technology sensitivity cases for the end-use
demand sectors in AEO98 do not include any feedback
effects from energy prices, and energy consumption in
each sector is lower than in the reference case solely due
to the revised assumptions about technology costs and
efficiencies. The sensitivity cases described in this
report, in contrast, were combined into an integrated
analysis. As a result, lower energy consumption in the
high technology case leads to lower energy prices,
which in turn produce some offsetting increases in consumption.

Carbon emission reduction targets and carbon prices
further complicate the integrated market analysis. In the
high technology sensitivity cases presented in this chapter, the carbon reduction targets are the same as those in
the comparable cases that use the AEO98 reference case
technology assumptions. For example, the 9-percentabove-1990 (1990+9%) case and the 1990+9% high technology sensitivity case have the same carbon emissions
target. The effect of the high technology assumptions is
to lower the projected carbon price that would be
required to achieve the same level of carbon emissions,
which also reduces the delivered price of fuel. With
lower carbon prices, adverse impacts on the macroeconomy and on energy markets are moderated. Assuming
that the technological advances posited in the high technology cases for the various end-use sectors could in fact
be achieved, energy consumption levels would not
necessarily be lower in each sector. Rather, the carbon

price would be lower, and it would be less costly to
achieve a given emissions reduction target.

Residential Demand
Background

As the largest electricity-consuming sector in the United
States, households were responsible for 20 percent of all
carbon emissions produced in 1996, of which 63 percent
was directly attributable to the fuels used to generate
electricity for the sector. Electricity is a necessity for all
households, and with electricity use per household
growing at 1.5 percent per year since 1990, the projected
increase in residential sector electricity consumption has
become a central issue in the debate over carbon stabilization and meeting the goals of the Kyoto Protocol.

The number of occupied households is the most important factor in determining the amount of energy consumed in the residential sector. All else being equal,
more households mean more total use of energy-related
services. From 1980 to 1996, the number of U.S. households grew at a rate of 1.4 percent per year, and residential electricity consumption grew by 2.6 percent per year.
In the reference case, the number of households is projected to grow by 1.1 percent per year through 2010, and
residential electricity consumption is projected to grow
by 1.6 percent per year. Strong growth in the South,
which features all-electric homes more prominently
than do other areas of the country, and the advent of
many new electrical devices for the home have significantly contributed to high electricity growth since 1980.
Although these trends are projected to continue through
2010, efficiency improvementsÑdue in part to recent
Federal appliance standards, utility demand-side management programs, building codes, and nonregulatory
programs (e.g., Energy Star)Ñshould dampen electricity growth somewhat as residential appliances are
replaced with newer, more efficient models.

Within the residential sector, all of the major end-uses
(heating, cooling, lighting, etc.) are represented by a
variety of technologies that provide necessary services.
Technologies are characterized by their cost, efficiency,
dates of availability, minimum and maximum life
expectancies, and the relative weights of the choice criteriaÑinstalled cost and operating cost. The ratio of the
weight of installed cost to that of operation cost gives an
estimate of the Òhurdle rateÓ used to evaluate the energy

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



efficiency choice.24 When more emphasis is placed on
installed cost, the hurdle rate is higher. The hurdle rates
for residential equipment range from 15 percent for
space heating technologies to more than 100 percent for
some water heating applications. The range in part
reflects differences in the way consumers purchase the
two technologies. In the case of water heaters, for example, purchases tend to occur at the time of equipment
failure, which tends to restrict the choice to equipment
readily available from the plumber. Space conditioning
equipment, on the other hand, is not used all year round,
allowing some latitude in terms of timing the replacement of an older unit. It is assumed that residential consumers expect future energy prices to remain at the
current level at the time of purchase when calculating
the future operating cost of a particular technology.

Technological advances and availability play a large role
in determining future energy savings and carbon emission reductions. Even in todayÕs marketplace, there exist
many efficient technologies that could substantially
reduce energy consumption and carbon emissions, however the relatively high initial cost of these technologies
restricts their widespread penetration. Over time, the
costs of more advanced technologies are assumed to fall
as the technology matures, one example being natural
gas condensing water heaters. In addition, technologies
that are not available today but are nearing commercialization are assumed to become available in the future.
Three technology menus are used in the analysis below:
a reference technology menu, a high technology menu
(reflecting more aggressive research and development),
and a ÒfrozenÓ menu limited to equipment available
today. In all cases, the menu options and characteristics
are fixed. In the high technology sensitivity case, for
example, the cost of a condensing natural gas water
heater is assumed to fall by almost 75 percent by 2005,
relative to the reference case, and a natural gas heat
pump water heater becomes available for purchase, by
2005.

In response to energy price changes, residential elasticities, defined as the percent change in energy consumed
with a 1-percent change in price, range from -0.24 to

0.28 in the short run, depending on the fuel type, to -0.33
to -0.51 in the longer term. The elasticities reported here
are derived from NEMS by a series of simulations with
only one energy price varying at a time, beginning in
2000.25 These price elasticities reflect changes in both the
demand for energy services and the penetration rate of
more efficient technologies. In the absence of energy
price changes, energy intensity, as defined as delivered
energy consumption per household, declines at an average rate of 0.5 percent per year through 2010. This non-
price-induced intensity improvement reflects the efficiency gain brought about by ongoing stock turnover,
equipment standards, new housing stock, and the future
availability of new technologies.

Energy consumption, including the combustion of
various fossil fuels, is the major source of U.S. carbon
emissions. Energy use in the residential sector is greatly
affected by year-to-year variations in seasonal
temperatures, particularly in the winter, as illustrated by
the decline in delivered energy use in 1990 (Figure 26),
which was one of the warmest winters on record. The
projections in this analysis assume normal seasonal
temperatures over the 1996-2020 forecast period.

In the 3-percent-below-1990 (1990-3%) carbon reduction
case, which assumes an emissions target of 3 percent
below 1990 levels for the United States, a sharp drop in
residential energy use is projected between 2005, when

1970198019902000201020200.91.01.11.21.31.4Index,1990=1.0Reference1990+14%
1990+24%
19901990+9%
1990-3%
1990-7%
HistoryProjectionsFigure 26. Index of Residential Sector Delivered
Energy Consumption, 1970-2020
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
24The ÒhurdlerateÓforevaluating energyefficiencyinvestmentshasalso beenreferredtoasthe Òimplicit discountrateÓ(i.e.,the empirically based rate required to simulate actual purchasesÑthe one implicitly used). These rates are often much higher than would be expected
if financial considerations alone were their source. Among the reasons often cited for relatively high apparent hurdle rates are uncertainty
about future energy prices and future technologies, lack of information about technologies and energy savings, additional costs of adoption
not included in the calculations, relatively short tenure of residential home ownership, hesitancy to replace working equipment, attributes
other than energy efficiency that may be more important to consumers, limited availability of investment funds, renter/owner incentive differences, and builder incentives to minimize construction costs. For a good discussion of potential market barriers and the economics of energy efficiency decisions, see Jaffe and Stavins, ÒEnergy Efficiency Investments and Public Policy,Ó The Energy Journal, Vol. 15, No. 2 (1994),
pp. 43-65.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



the target is implemented, and 2010 (Figure 26).
However, the projected decline is nearly identical to that
seen historically from 1978 to 1983, in terms of both
consumption and intensity (Figure 27). Housing starts, a
major predictor of residential energy use, fell from 2.02
million units in 1978 to 1.062 million in 1982.26 The drop
in housing starts was tied directly to mortgage rates,
which increased from 9.6 percent in 1978 to over 16
percent in 1981-1982. In addition, real energy prices to
the residential sector increased by 87 percent from 1978
to 1982, similar to the 82-percent real price increase
projected in the 1990-3% case. In the carbon reduction
cases, delivered energy consumption in the residential
sector never reaches its 1990 level, which has been used
as a benchmark in setting carbon reduction targets.
Given the uncertainty regarding technology and
consumer behavior in a high-price energy world,
additional sensitivities are examined here to analyze the
effects of variations in the level of optimism associated
with assumptions about both technology advances and
consumer responsiveness.

1970198019902000201020200.70.80.91.01.11.21.31.41.5Index,1996=1.0Reference1990+24%
1990+9%
1990-3%
HistoryProjectionsFigure 27. Index of Residential Sector Delivered
Energy Intensity, 1970-2020
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997) and Data
Resources Incorporated. Projections: Office of Integrated Analysis and
Forecasting, National Energy Modeling System runs KYBASE.D080398A,
FD24ABV.D080398B, FD09ABV.D080398B, and FD03BLW.D080398B.
Carbon Reduction Cases

Carbon emissions associated with electricity generation
are the largest component of emissions from the
residential sector, in terms of both the levels and
projected growth in the reference case, and in terms
of the projected declines in the carbon reduction cases.
In the reference case, which does not include the
Kyoto Protocol, 98 percent of the projected increase
in residential sector carbon emissions by 2010 results
from increasing electricity use and the fuels used for

electricity generation. In the 1990+9% case, 87 percent of
the sectorÕs decline in carbon emissions is related to
reduced electricity demand and changes in electricity
generation (Figure 28). The following discussion focuses
on the results of three carbon reduction casesÑ1990-3%,
1990+9%, and 24-percent-above-1990 (1990+24%)Ñin
which carbon emissions, averaged across all energy
sectors, reach targeted levels relative to 1990 in the 20082012 period.

19901996Reference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
050100150200250MillionMetricTonsElectricNon-ElectricHistoryProjectionsfor2010Figure 28. Residential Sector Carbon Emissions,
1990, 1996, and 2010
Note: Electricity emissions are from the fossil fuels used to generate
the electricity used in this sector.
Sources: History: Energy Information Administration, Emissions of
Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington,
DC, October 1997). Projections: Office of Integrated Analysis and Forecasting,
National Energy Modeling System runs KYBASE.D080398A, FD24ABV.
D080398B, FD1998.D080398B, FD09ABV.D080398B, FD1990.D080398B,
FD03BLW.D080398B, and FD07BLW.D080398B.
Although the use of electricity contributes most to the
projected growth in emissions in the residential sector,
natural gas consumption, which emits relatively low
levels of carbon per Btu burned compared with coal (the
major fuel used to generate electricity), is projected to
remain the most important fuel in the sector as
measured by delivered energy. Figure 29 shows delivered energy consumption by major fuel as well as the
losses associated with electricity generation. On a delivered basis, natural gas use is projected to decrease the
most in the three carbon reduction cases by 2010. Relative to the projected level of consumption in the reference case in 2010, delivered energy consumption is
projected to be 10 percent lower in the 1990+9% case and
electricity-related losses 22 percent lower. Of the 2.0
quadrillion Btu savings in electricity-related losses in
2010 in the 1990+9% case, 43 percent (0.9 quadrillion
Btu) can be attributed to reduced electricity demand in
the residential sector. The remaining 1.1 quadrillion Btu
(57 percent) of the savings in electricity-related losses
comes from efficiency gains and/or fuel switching for

25The long-run elasticities reflect the effects of altered prices after 20 years for the last year of the forecast, 2020.
26U.S. Bureau of the Census, Construction Reports, series C20.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



197019801996Reference1990+24%
1990+9%
1990-3%
024681012QuadrillionBtuNaturalGasElectricityPetroleumElectricity-
RelatedLossesHistoryProjectionsfor2010Figure 29. Delivered Energy Consumption in the
Residential Sector by Major Fuel, 1970,
1980, 1996, and 2010
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
electricity generation. Thus, changes in electricity
supply, absent any major technological or behavioral
changes in residential end use over the next 12 years, are
the key to controlling carbon emissions for the
residential sector.

Energy is used in the residential sector to provide a
number of different services, which vary in end-use
intensity (energy consumption per household) (Figure
30). Space conditioning (which includes heating,
cooling, and ventilation) is clearly the most energy-
intensive end use in the sector, and it accounts for most
of the direct use of fossil fuels. ÒWhite goodsÓ (which
include refrigerators, freezers, dishwashers, clothes
washers and dryers, and stoves), lighting, and other
uses are almost entirely powered by electricity and,
therefore, are responsible for most of the electricity-
related losses.

In the reference case, most of the projected growth in
residential energy consumption between 1996 and 2010
comes from increasing use of miscellaneous electric
devices, such as personal computers and home security
systems (Figure 31). The rate at which energy consumption changes over time depends on factors such as
equipment turnover rates, ability to control unit operation (thermostatic controls), energy prices, household
size (people per house), housing unit size (square feet),
and the efficiency of newly purchased appliances. Stock
turnover can provide drastic reductions in energy intensity, even without future gains in appliance efficiency.
On average, a new refrigerator purchased in 1995 used

SpaceConditioningWaterHeatingWhiteGoodsLightingOtherUses020406080100MillionBtuperHouseholdElectricity-RelatedLossesDirectUseFigure 30. Residential Sector Energy Use per
Household, 1996
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System run KYBASE.D080398A.
SpaceConditioningWaterHeatingWhiteGoodsLightingOtherUses012345-1-2-3PercentperYearReference1990+24%1990+9%1990-3%
Figure 31. Average Projected Annual Growth in
Residential Sector Energy Consumption
by End Use, 1996-2010
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B..
62 percent less electricity than one purchased 20 years
earlier.27 Conversely, slow stock turnover can limit the
role of energy efficiency gains in the future. Equipment
purchased in the 1990s that lasts 20 years or more will
not be eligible for replacement until after 2010.

With the exception of white goods, increases in total
energy consumption for all the major residential energy
services are projected from 1996 to 2010 in the reference
case. The negative growth in total energy consumption
for white goods results from a decline in energy use for

27Association of Home Appliance Manufacturers, Fact Book 1996.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



refrigeration, as aggressive Federal efficiency standards28 taking effect in 1993 and 2001 reduce the amount
of energy needed to provide the same level of service. In
the carbon reduction cases, increasing energy prices act
to reduce the growth in energy consumption for all
major services relative to their growth in the reference
case. In the absence of mandatory standards, residential
consumers traditionally have been reluctant to purchase
highly efficient appliances. However, faced with the
higher energy prices projected in the carbon reduction
cases, it is expected that consumers will respond by
purchasing more efficient appliances (Table 4). The extent of consumer response and its impact on average
equipment efficiencies would also depend on the purchase price of the new equipment (the initial investment
required).

Table 4. Change in Projected Average Efficiencies
of Newly Purchased Residential
Equipment in Carbon Reduction Cases
Relative to the Reference Case, 2010
(Percent)
Technology 1990+24% 1990+9% 1990-3%
Air-Source Heat Pump ..... 1.3 3.6 5.7
Electric Water Heater ...... 0.3 2.4 13.6
Natural Gas Water Heater . . 1.1 3.7 4.8
Building Shell ............ 1.0 3.3 5.5
Source: Office of Integrated Analysis and Forecasting, National Energy Mod-
eling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
In the reference case, the real (inflation-adjusted) prices
of electricity and natural gas to residential consumers
are projected to decline between 1996 and 2010 (Figure
32), by 8 and 10 percent, respectively. The outlook for
prices in the carbon reduction cases, however, is much
different. Without major changes in energy policy,
technology, or consumer response, prices to the
residential sector are expected to be as much as 94
percent higher in 2010 in the 1990-3% case. In response
to the higher prices, total residential energy consumption is projected to decline by more than 20 percent by
2010 in the 1990-3% case.

The factors that contribute to lower consumption
include behavioral responses, such as adjusting the
thermostat or turning off the lights when leaving the
room, and, to a lesser extent, the acquisition of more
efficient appliances. The rate of improvement in average
appliance efficiency is constrained by the rate of stock
turnover. For example, it is not uncommon for major
energy-using appliances, such as furnaces, to last for 30
years or more. More immediate responses to higher

energy prices can be achieved through retrofits to
improve the thermal efficiency of building shells.
During the energy price shocks of the 1970s, for
example, homeowners increased insulation levels
substantially,29 with the immediate effect of conserving
energy and lowering energy bills. The potential for
similar improvement between 1996 and 2010 is reduced,
given the improvements already made.

197019801996Reference1990+24%
1990+9%
1990-3%
0.00.20.40.60.81.01.21.41.61.82.0Index,1996=1.0NaturalGasElectricityPetroleumHistoryProjectionsfor2010Figure 32. Index of Residential Sector Energy
Prices, 1970, 1980, 1996, and 2010
Sources: History: Energy Information Administration, State Energy Price and
Expenditure Report 1994, DOE/EIA-0376(94) (Washington, DC, June 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
Sensitivity Cases

High and Low Technology. Technology improvements
over time can take the form of increased efficiency,
decreased cost, or both. To examine the effects of
assumptions about the rate at which technologies will
improve in the future, two sets of sensitivity cases were
analyzed. The low technology sensitivity cases assume
that none of the improvements assumed in the reference
case will occur. In other words, future technologies are
assumed to be ÒfrozenÓ at their 1998 cost and efficiency
levels. Technological improvement occurs in this case as
older units are retired and are replaced with 1998 technologies. Engineering technology experts were consulted to develop the high technology case, which
assumes more rapid advances than those in the reference case, due to research and development (Table 5).30
In the high technology case, for example, the efficiency
of the best available natural gas water heater is
assumed to improve by 63 percent over the 1998 level by
2015, and the cost is assumed to decline by 15 percent,

28These standards represent updates to previous standards authorized by the National Appliance Energy Conservation Act of 1987.
29U.S. Department of Energy, Progress in Residential Retrofit, Based on Owens-Corning Marketing Research.
30Energy Information Administration, Technology Forecast UpdatesÑResidential and Commercial Building Technologies, Draft Report (Ar


thur D. Little, Inc., June 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Renewables and Dispersed Electricity Generation

Dispersed renewable energy use in the residential sector solar thermal collectors to the residential and commerincludes wood, solar thermal, geothermal energy, pho-cial sectors nearly doubled to 10 million square feet from
tovoltaic cells, and fuel cells.a Wood is used as a main or 5.8 million square feet in 1976. The annual growth
secondary heating source in some households. Geother-inshipments averaged 8 percent per year until 1985,
mal energy is used to power ground-source heat pumps, when the tax credits were repealed. Subsequently, ship-
which exchange energy with below-ground earth or ments fell sharply from 19.1 million square feet in 1985
water, extracting heat in the winter and delivering heat to 9.1 million in 1986. The energy tax credit was reintroto the earth (and cooling the building) in the summer. duced for the commercial sector in 1986, followed by a
Solar thermal energy is used mainly to heat water for small increase in shipments, but since 1991 there has
swimming pools and household use. Photovoltaics pro-been little growth in the industry. Residential sales of
vide small-scale electricity generation, often in remote solar thermal systems are not expected to increase sub-
locations, using semiconductors to transform sunlight stantially in the reference case, given current tax policy
directly into electricity, which may be used for a variety and projected declines in real energy prices.
of functions, such as water pumps or remote lighting
systems. Fuel cells convert liquid fossil fuels into elec-Domestic shipments in the photovoltaic market (includ

tricity through electrochemical processes. ing both dispersed and grid-connected system) have

grown significantly since the 1980s, but they also were
The share and quantity of wood as a primary heating affected by the repeal of the tax credit. From 10,717 peak
fuel in the residential sector has been falling for nearly kilowatts shipped in 1983, shipments were down to
two decades. In 1982, 6.7 percent of all U.S. households 3,224 peak kilowatts in 1986 after the tax credit repeal, a
heated with wood, but its share fell to 3.2 percent in 1993. 32-percent average annual decline.c The market recov-
The aggregate quantity of wood consumed as primary ered somewhat in the next decade, with 1992 shipments
heating in households has fallen as well, from 28.7 mil-reaching 5,760 peak kilowatts. Since then, the industry
lion cords in 1982 to 12.6 million cords in 1993.b The has been developing steadily, particularly after 1992,
decline has resulted in part from local laws restricting with 23-percent average annual growth to 13,016 peak
wood burning. In addition, the convenience of natural kilowatts shipped in 1996.
gas heating and the decline in real oil and gas prices over

the past decade have led many households to choose gas Fuel cells have the potential for future integration into
or oil over wood. both grid-connected and off-grid applications in every

sector. When their cogenerative capabilities are used,
While wood has declined as a primary residential heat capturing excess heat from the chemical reaction for
source, its use as a backup or secondary heat source has space and water heating, fuel cell efficiencies can rise to
not. Wood use as a secondary heat source increased two or three times those of typical energy combustion
from 16 percent of households in 1980 to 20 percent in plants, emitting only half the amount of carbon dioxide
1993, suggesting that wood stoves are being kept as per unit of useful energy obtained.d
backup heating systems. If the prices of other fuels rise

significantly, however, the use of wood as a primary To date, fuel cells have not been used extensively. With

household heating fuel may well increase. In the refer-their relatively recent development and only one major
ence case for this analysis, wood energy use is projected manufacturer worldwide, there are only 160 medium-

sized (200-kilowatt) units in use.e Smaller units have

to be 0.61 quadrillion Btu in 2010. In the most stringent

carbon reduction case (7 percent below 1990 levels), been tested in the space program and in the automobile

higher energy prices lead to wood use of 0.63 quadrillion industry, but the first unit designed for the residential

market was not built until 1998.f Fuel cells are a promis-

Btu in 2010, increasing to 0.67 quadrillion Btu in 2020.

ing technology for the residential sector, but their
The market for solar energy systems has undergone sub-current high costs do not favor extensive market
stantial changes over the past three decades, largely as a penetration. Costs can be expected to fall as production
result of the introduction, removal, and subsequent rein-volumes increase, and depending on the timing and
troduction of Federal energy tax credits for photovoltaic extent of the cost reductions, fuel cells could become an
cells and solar thermal collection systems. With the important source of dispersed electricity generation.
introduction of a Federal tax credit in 1978, shipments of

aDispersed renewable energy is the direct use of power from a renewable energy system such as a photovoltaic array, disconnected
from the electric power grid. The production and sale of electricity from utilities using renewable energy fuels are not included.

bEnergy Information Administration, Housing Characteristics 1980, DOE/EIA-0312 (Washington, DC, June 1982), p. 101; Housing
Characteristics 1982, DOE/EIA-0314(82) (Washington, DC, August 1984), pp. 47-98; and Household Energy Consumption and Expenditures 1993, EIA/DOE-0321(93) (Washington, DC, October 1995), pp. 37-62.

cEnergy Information Administration, Renewable Energy Annual 1997, Vol. 1, DOE/EIA-0603(97/1) (Washington, DC, February
1998), p. 19.

dWhen byproduct heat is used, average total efficiency of the system increases to approximately 80 percent, significantly more than
a standard coal-fired utility plant, which operates at around 30 percent efficiency. Source: U.S. Department of Energy, Office of Fossil
Energy, Technology Center, Climate Change Fuel Cell Program, NG001.1197M.

eFred Kemp, Manager of Government Programs, International Fuel Cells (South Windsor, CT), personal communication, August
1998.

fNew York Times (June 17, 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Table 5. Cost and Efficiency Indexes of Best Available Technologies for Selected Residential Appliances,
2015

(1998 Values = 1.00)

Cost Efficiency
1990+9% 1990+9% 1990+9% 1990+9%
Low High Low High
Technology Technology 1990+9% Technology Technology 1990+9% Technology

Air-Source Heat Pump ........... 1.00 0.99 0.98 1.00 1.09 1.18
Ground-Source Heat Pump........ 1.00 0.86 0.56 1.00 1.05 1.08
Natural Gas Heat Pump .......... 1.00 0.81 0.75 1.00 1.00 1.00
Natural Gas Water Heater......... 1.00 0.76 0.85 1.00 1.00 1.63
Solar Water Heater .............. 1.00 1.00 0.73 1.00 1.00 1.67
Electric Water Heater ............ 1.00 1.00 0.73 1.00 1.04 1.17

Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs FREEZE09.D080798A, FD09ABV.D080398B, and
HITECH09.D080698A, computed from Technology Forecast UpdatesÑResidential and Commercial Building Technologies, Draft Report (Arthur D. Little, Inc., June
1998).

while ground-source heat pumps, which do not realize
much gain in efficiency, are assumed to decline in cost
by 44 percent in the high technology case by 2015.

Ground-source heat pumps, which draw stored heat
from the ground beneath the frost line, provide an
efficient and comfortable (in terms of delivered heat)
alternative to the more common air-source heat pumps.
The cost of the unit and the placement of the ground
loop have been major barriers to wide market acceptance, however. Different levels of stocks of ground-
source heat pumps are projected in the reference case,
the 1990+9% carbon reduction case, and the 1990+9%
case low and high technology cases (Figure 33). Given
that significant market acceptance is seen only in the
high technology case, it can be concluded that the costs
associated with the technology restrict its acceptance.
Space heating technologies, in general, have the lowest
hurdle rates (15 percent) of all residential appliances,
primarily because of the large energy costs of home heating, relative to other energy-using services.

Figure 34 shows that improvements in technology can
indeed dampen the impact carbon restrictions have on
residential energy prices. Given the amount of time
needed for technology to penetrate the market, one
would expect that over a longer period of time, the
prices in the high technology sensitivity would fall
relative to the other cases. After 2008, prices in the high
technology sensitivity begin to fall, as reduced energy
demand caused by more efficient technology penetrating the market begin to make an impact. Relative to the
price in the 1990+9% case, the composite real residential
energy price in 2010 is 11 percent less in the high technology case. Conversely, if technology were frozen at the
level available in 1998, 2010 prices are expected to be 17
percent higher than the 1990+9% case, indicating that
energy efficiency plays a significant role in the cases
with reference technology assumptions.

Energy fuel expenditures are a good indication of the
success that technological advancement achieves in

1995200020052010201520200123456MillionUnitsReference1990+9%HighTechnology1990+9%LowTechnology1990+9%
Figure 33. Projected Stocks of Ground-Source
Heat Pumps, 1995-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FREEZE09.D080798A, FD09ABV.
D080398B, and HITECH09.D080698A.
19701980199020002010202005101520251996DollarsperMillionBtu1990+9%
HighTechnology1990+9%
LowTechnology1990+9%
ReferenceHistoryProjectionsFigure 34. Average Residential Sector Energy
Prices, 1995-2020
Sources: History: Energy Information Administration, State Energy Price and
Expenditure Report 1994, DOE/EIA-0376(94) (Washington, DC, June 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FREEZE09.D080798A, FD09ABV.
D080398B, and HITECH09.D080698A.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



lessening the impact on the consumer in a carbon-
restricted environment. Figure 35 details residential
sector energy expenditures for the 1990+9% case and
technology sensitivities. For the high technology
sensitivity, energy expenditures in 2020 are 23 percent
less than those realized in the 1990+9% case, saving
consumers over $440 billion from 2008 to 2020.

199520002005201020152020050100150200250300Billion1996Dollars1990+9%HighTechnology1990+9%LowTechnology1990+9%
ReferenceFigure 35. Projected Energy Expenditures in the
Residential Sector, 1995-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FREEZE09.D080798A, FD09ABV.
D080398B, and HITECH09.D080698A.
Increased Consumer Response. Residential energy
consumers have traditionally been reluctant to invest in
energy efficiency, even with ample financial benefits.
Many market barriers tend to create what are known as
high hurdle rates for consumer investments in energy
efficiency. As of 1993, 35 percent of all homes were
occupied by renters,31 most of whom were responsible
for paying energy bills but not for purchasing major
energy-consuming appliances. Such households tend to
buy the least expensive equipment on the market, which
also tends to be the least energy-efficient. The same
reasoning can be applied to many newly constructed
homes as well, because the builders, not the occupants,
are tasked with equipping them with most of the major
energy-using appliances. Other barriers include
equipment availability (e.g., whether plumbing
contractors have high-efficiency water heaters available
when they make service calls) and lack of information.

To examine the effects that lower hurdle rates could
have on both energy prices and expenditures in the carbon reduction cases, and at the same time differentiate
those effects from the effects of technological advances,

31Energy Information Administration, Housing Characteristics 1993.

an increased consumer response sensitivity case was
analyzed. This sensitivity case includes assumptions of
lower discount rates, higher short-run elasticities of
demand, greater inclination to change fuels when purchasing equipment, and lower growth in miscellaneous
electricity use.32

Impacts of Increased Consumer Response and
Advanced Technology. In order to gauge the impact of
assumptions regarding technological advancement and
consumer behavior with respect to delivered energy
consumption, sensitivity cases were analyzed relative to
the 1990+9% case where delivered energy prices were
the same across all cases. These cases serve to isolate the
impact of each of the key variables separately, and to
understand the impact of implementing the sensitivities
simultaneously. This section evaluates the relative
impact that each of these concepts could have on future
energy intensity at a price level realized in the 1990+9%
case.

Changes in technological development and the value
residential consumers place on energy related issues can
significantly affect the pattern of energy consumptionÑand carbon emissionsÑin the future. The availability of high-efficiency technologies in itself does not
guarantee increased energy efficiency. Without the willingness of consumers to purchase the more efficient
products, which usually cost significantly more, technology may not have much of an impact on future
energy consumption patterns. Conversely, in a world
where energy conservation was of paramount concern
to energy consumers, yet at the same time high-
efficiency products were unavailable, future energy consumption patterns would probably not be greatly
affected either.

Given the detailed nature regarding technological
development and consumer choice with regards to different technologies, it is important to analyze the results
at the technology level, as well as the overall level. With
nearly 40 million households (38 percent) using electric
water heaters in 1995, and given the relatively high
intensity associated with using electric water heaters,
the projected impact of increased energy efficiency can
have a large impact on future electricity use for this service. Electric resistance water heaters have traditionally
exhibited slow growth in energy efficiency. In fact, the
highest efficiency unit available today is not likely to see
any efficiency improvement due to thermal limits and
diminishing returns on controlling heat loss.33 This
implies that future gains in efficiency for electric water

32Assumptions include lowering hurdle rates to 15 percent real, increasing the price sensitivity parameters to switch fuels, increasing
short-run price elasticities from -0.25 to -0.40, and decreasing miscellaneous electricity penetration.

33Energy Information Administration, Technology Forecast UpdatesÑResidential and Commercial Building Technologies, Draft Report (Arthur D. Little, Inc., June 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



heating must be achieved through the increased penetration of electric air-source heat pump water heaters,
which achieve higher efficiency levels by extracting heat
from the air surrounding the unit. The current cost of
this technology, however, is several times that of a traditional resistance unit, and coupled with observed
implicit discount rates of over 100 percent, has led to
very limited market penetration.

Assumptions regarding technological advances through
improved performance and reduced cost, as well as
changes in consumer behavior, can significantly affect
the market penetration of emerging technologies. Figure
36 details the relative importance of varying assumptions regarding technological advances and consumer
behavior with respect to the intensity of the electric
water heating end use.34 Relative to the 1990+9% case,
intensity drops faster when assumptions regarding
consumer behavior are changed, as compared to
changes in technology characteristics. Over time,
however, the intensity decline in the technology case
outpaces that projected for the behavior case as more
and more equipment is purchased at higher efficiency
levels. Combining both sets of assumptions, that is,
changing both technology characteristics and consumer
behavior together, results in over a 25 percent decline in
energy intensity for electric water heating over time.
This indicates that a combination of both technology and
consumer behavior changes can bring about large
declines in energy intensity for this service, all else being
equal.

1995200020052010201520200-5-10-15-20-25-30PercentChangeFrom1990+9%CaseHighTechnologyHighConsumerResponseHighTechnologyandHighConsumerResponseFigure 36. Changes From Reference Case
Projections of Energy Intensity for
Residential Water Heating in Three
Sensitivity Cases, 1995-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs FD09ABV.D080398B, KYTECH.D081098C, KYBHAVE.
D081098D, and KYBOTH.D081098A.
Overall annual energy consumption per household, or
energy intensity, for these sensitivity cases follows the
general pattern described for electric water heating.
Again, technology advances exhibit a greater potential
for energy intensity decline in the long run (Figure 37),
but the combination of the two cases yields roughly half
of the intensity decline projected for electric water
heating. This is due to the fact that all other major
technologies exhibit much lower observed hurdle rates
and less range in terms of high-efficiency products. For
example, natural gas furnaces, the largest energy
consuming product class in terms of delivered energy in
the U.S., has already matured in terms of product
efficiency, and at the same time hurdle rates are at 15
percent.

1995200020052010201520200-2-4-6-8-10-12-14PercentChangeFrom1990+9%CaseHighTechnologyHighConsumerResponseHighTechnologyandHighConsumerResponseFigure 37. Changes From Reference Case
Projections of Residential Energy
Consumption in Three Sensitivity
Cases, 1995-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs FD09ABV.D080398B, KYTECH.D081098C, KYBHAVE.
D081098D, and KYBOTH.D081098A.
Commercial Demand
Background

The commercial sector consists of businesses and other
organizations that provide services. Stores, restaurants,
hospitals, and hotels are included, as well as a wide
range of facilities that would not be considered ÒcommercialÓ in a traditional economic sense, such as public
schools, correctional institutions, and fraternal organizations. In the commercial sector, energy is consumed
mainly in buildings, and relatively small amounts are
used for services, including street lights and water
supply.

34Intensity here is the average annual consumption of electricity for water heating in homes with electric water heaters.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



The commercial sector is currently the smallest of the
four demand sectors in terms of energy use, accounting
for 11 percent of delivered energy demand in 1996. The
commercial sector is also responsible for fewer carbon
emissions than the other sectors, emitting 230 million
metric tons, or 16 percent of total U.S. carbon emissions,
in 1996. The sector has a larger share of emissions than
its share of energy use because of the importance of commercial electricity use. The emissions associated with
electricity-related losses are included in the calculation
of emissions from electricity use.

Several factors determine energy use and, consequently,
carbon emissions in the commercial sector. One of the
most important is floorspace. Building location, age, and
type of activity also affect commercial energy use. Currently, total commercial floorspace in the United States
exceeds the area of the State of Delaware and amounts to
about 200 square feet for every U.S. resident. Mercantile
(retail and wholesale stores) and service businesses are
the most common type of commercial buildings, and
offices and warehouses are also common.35

Because of the relatively long lives of buildings, the characteristics of the stock of commercial floorspace change
slowly. Over half of the commercial buildings in the
United States were built before 1970, and the reference
case used for this analysis projects that total commercial
floorspace will grow at about the same rate as population, 0.8 percent annually, through 2020. This limits the
effects that new, more efficient building practices can
achieve in the near term, but as time passes and building
stock ÒturnoverÓ occurs, current and future building
practices will have a greater effect on commercial energy
use.

The composition of end-use services is another determinant of the amount of energy consumed and the type of
fuel used. The majority of energy use in the commercial
sector is for lighting, space heating, cooling, and water
heating. In addition, the proliferation of new electrical
devices, including telecommunications equipment, personal computers, and other office equipment, is spurring growth in electricity use. Electricity use currently
accounts for 45 percent of delivered energy consumption in the sector, and that share is projected to grow to
about 48 percent by 2010 in the reference case.

Consideration of end-use services leads to another
determining factor in commercial energy consumptionÑthe effects of turnover and change in end-use technologies. The stock of installed equipment changes with
normal turnover as old, worn-out equipment is replaced
and new buildings are outfitted with newer versions
of equipment that tend to be more energy-efficient.

Equipment with even greater energy efficiency is
expected to be available to commercial consumers in the
future. Energy prices have both short-term and long-
term effects on commercial energy use. Fuel prices influence energy demand in the short run by affecting the use
of installed equipment and in the long run by affecting
the stock of installed equipment.

Legislated efficiency standards also affect energy use, by
imposing a minimum level of efficiency for purchases of
several types of equipment used in the commercial sector. Two mandates currently affect commercial appliances: the National Energy Policy Act of 1992 (P.L. 102486, Title II, Subtitle C, Section 342), which specifically
targets larger-scale commercial equipment and fluorescent lighting, and the National Appliance Energy Conservation Amendments (NAECA), which affect
commercial buildings that install smaller residential-
style equipment. Examples include standards for heat
pumps, air conditioning units, boilers, furnaces, water
heating equipment, and fluorescent lighting.

Effects of Technology Availability
and Choice

The degree to which energy-efficient equipment can
affect energy consumption, and in turn carbon emissions, in the commercial sector is limited by the level of
efficiency available to commercial consumers and the
rate at which more efficient equipment is purchased.
Technologies for all the major end uses (lighting, heating, cooling, water heating, etc.) are defined by their
installed cost, operating cost, efficiency, average useful
life, and first and last dates of availability. These
parameters are considered, along with fuel prices at the
time of purchase, in the selection of technologies that
provide end-use services. Commercial consumers are
not assumed to anticipate any future changes in fuel
prices when choosing equipment. The commercial sector encompasses a wide variety of buildings, and not all
consumers will have the same requirements and priorities when purchasing equipment. Major assumptions
that take these differences in behavior into account and
affect commercial technology choices are described
below.

In making the tradeoffs between equipment cost and
equipment efficiency, the purchase behavior of the commercial sector is represented by distributing floorspace
over a variety of hurdle rates. Rates of return on investments in energy efficiency (referred to in financial parlance as Òinternal rates of returnÓ) are required to meet
or exceed the hurdle rate. Floorspace is distributed over
hurdle rates that range from a low of about 18 percent to
rates high enough to cause choices to be made solely by

35General characteristics of the commercial sector provided in the above paragraphs are from Energy Information Administration, A
Look at Commercial Buildings in 1995: Characteristics, Energy Consumption, and Energy Expenditures, DOE/EIA-0318(95) (Washington, DC, September 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



minimizing the costs of installed equipment (i.e., future
potential energy cost savings are ignored at the highest
hurdle rate).36 The distribution of hurdle rates used in all
the cases for this analysis is not static: as fuel prices
increase, the nonfinancial portion of each hurdle rate in
the distribution decreases.37

For a proportion of commercial consumers, it is assumed
that newly purchased equipment will use the same fuel
as the equipment it replaces. This proportion varies by
building type and by type of purchaseÑwhether it is for
new construction, to replace worn-out equipment, or to
replace equipment that is economically obsolete. Purchases for new construction are assumed to show the
greatest flexibility of fuel choice, while purchases for replacement equipment have the least flexibility. For
example, when space heating equipment in large office
buildings is replaced, 8 percent of the purchasers are
assumed to consider all available equipment using any
fuel or technology, while 92 percent select only from
technologies that use the same fuel as the equipment
being replaced. The proportions used are consistent
with data from EIAÕs Commercial Buildings Energy
Consumption Survey and from published literature.38
Considerations such as owner versus developer financing, past experience, ease of installation, and fuel availability all play a role in fuel choice. This assumption also
accounts for some of the factors that influence technology choices but cannot be measured. For example, a hospital adding a new wing has an economic incentive to
use the same fuel that is used in the existing building.

The availability and costs of advanced technologies
affect the degree to which they can contribute to future
energy savings and carbon emission reductions. Many
efficient technologies currently available to commercial
consumers could significantly reduce energy consumption; however, their high purchase costs and the current
low level of fuel prices have limited their penetration to
date. As more advanced technologies mature over time,
their costs are expected to decline (compact fluorescent
lighting is an example). New technologies, beyond those
available today, may also enter the market in the future.
For example, the high technology sensitivity case,
described below, assumes that by 2005 a triple-effect
absorption natural-gas-fired commercial chiller will be
widely available, and that ÒtypicalÓ heat pump water
heaters will cost 18 percent less than assumed in the
reference case.

The combination of technology and behavior assumptions determines the commercial-sector price elasticity
for each of the major fuelsÑthat is, how commercial-
sector demand projections are affected by changes in
energy prices. Specifically, the commercial-sector price
elasticity for a particular fuel is the percent change in
demand for that fuel in response to a 1-percent change in
its delivered price. In the reference case, short-run price
elasticities for fuel use in the commercial sector are -0.34
for electricity, -0.39 for natural gas, and -0.39 for distillate fuel oil. Long-term price elasticities in the reference
case are higher, reflecting changes in both the use of
existing equipment and the adoption rates for more efficient equipment: -0.36 for electricity, -0.44 for natural
gas, and -0.45 for distillate fuel oil.39 The similarity of the
short-run and long-run elasticities for electricity has two
main causes. First, electric equipment becomes more
efficient even with the reference case assumptions, thus
reducing opportunities for further reductions when
prices are higher. For example, electric lighting efficiency in the reference case increases on average by 0.6
percent per year from 1996 through 2020. Electric space
cooling and ventilation improve on average by 1.1 and

0.7 percent per year, respectively, over the same period.
Second, miscellaneous electric end uses capture a growing share of commercial electricity consumption and
exhibit the same response in the long run as in the short
run. Building codes, equipment standards, and
improvements in technology costs and performance
contribute to reduced energy intensity in the commercial sector (i.e., annual energy consumption per square
foot of floorspace) even in the absence of price changes.
With constant real energy prices, energy intensity
declines on average by 0.1 percent per year through
2010.
Carbon Reduction Cases

In the 1990-3% case, commercial sector energy use in
2010 is projected to be below the 1996 level (Figure 38),
and carbon emissions attributable to the commercial sector are projected to be 29 percent below their 1990 levels
(Figure 39), despite 1-percent annual growth in commercial floorspace from 1996 to 2010. Projected fuel prices in
2010 in the 1990-3% case are more than twice as high as
the reference case projection, and they are higher in real
terms than they have been in any year since 1980 (Figure
40). As a result, energy consumption in 2010 is 22 percent lower in the 1990-3% case than in the reference

36The hurdle rates consist of both financial and nonfinancial components, as described for the residential sector.

37For the purposes of this study, the financial portion of the hurdle rates is considered to be 15 percent in real terms.

38Currentassumptions use ananalysisofdata fromEIAÕs 1992 commercial buildings survey.Sources fordataon consumer behaviorare
listed on page A-18 of Energy Information Administration, Model Documentation Report: Commercial Sector Demand Module of the National Energy Modeling System, DOE/EIA-M066(98) (Washington, DC, January 1998).

39As in the residential model, the long-run elasticities are for 2020 and represent the effects after 20 years of altered price regimes.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1970198019902000201020200.80.91.01.11.21.31.4Index,1990=1.01990+14%
1990+9%
1990+24%
Reference19901990-3%
1990-7%
HistoryProjectionsFigure 38. Index of Commercial Sector Delivered
Energy Consumption, 1970-2010
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
197019801996Reference1990+24%
1990+9%
1990-3%
0.00.51.01.52.0Index,1996=1.0NaturalGasElectricityPetroleumHistoryProjectionsfor2010Figure 40. Real Prices for Delivered Energy in the
Commercial Sector by Fuel, 1970, 1980,
1996, and 2010
Sources: History: Energy Information Administration, State Energy Price and
Expenditure Report 1994, DOE/EIA-0376(94) (Washington, DC, June 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
19901996Reference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
050100150200250300MillionMetricTonsElectricNon-ElectricProjectionsfor2010HistoryFigure 39. Commercial Sector Carbon Emissions,
1990, 1996, and 2010
Note: Electricity emissions are from the fossil fuels used to generate
the electricity used in this sector.
Sources: History: Energy Information Administration, Emissions of Green-
house Gases in the United States 1996, DOE/EIA-0573(96) (Washington, DC,
October 1997). Projections: Office of Integrated Analysis and Forecasting,
National Energy Modeling System runs KYBASE.D080398A, FD24ABV.
D080398B, FD1998.D080398B, FD09ABV.D080398B, FD1990.D080398B,
FD03BLW.D080398B, FD07BLW.D080398B.
case, and expenditures for energy purchases are 52 percent higher. Energy consumption starts to increase again
later in the 1990-3% case, as demand reductions lead to a
decline in fuel prices. Energy consumption in the
1990+24% and 1990+9% cases does not rebound as
much, because prices do not fall at the rate seen in the
1990-3% case.

Floorspace expansion in the commercial sector will lead
to growth in energy consumption if other factors remain
the same. Figure 41 removes the effects of floorspace
growth by presenting commercial energy intensity in
terms of delivered energy consumption per square foot
of commercial floorspace. Although total energy
consumption continued to increase when energy prices
were rising from 1970 through 1982, commercial energy
intensity declined by about 12 percent. Delivered energy
intensity in the reference case is projected to remain
essentially flat throughout the forecast. Projected
commercial sector growth is offset by the availability
and continued development of energy-efficient
technologies, existing equipment efficiency standards,
and voluntary programs such as those for the Climate
Change Action Plan. In the carbon reduction cases, with
higher energy prices, the energy intensities projected for
2010 are below the 1996 level. The projections for
commercial delivered energy intensity in 2010 in the
1990+24%, 1990+9%, and 1990-3% cases are 5 percent, 13
percent, and 21 percent below the reference case
projection, respectively.

When energy prices rise, consumers are expected to
reduce energy use by purchasing more efficient equipment and by altering the way they use energy-
consuming equipment. In addition to buying more efficient boilers and chillers, commercial customers in the
1990-3% case are expected to choose more heat pumps,
heat pump water heaters, and efficient lighting technologies than they would in the reference case (Table 6).
The same trends toward purchasing efficient technologies and monitoring energy use are projected in the
1990+9% case and in the 1990+24% case, but to a lesser
degree than projected for the 1990-3% case.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1970198019902000201020200.00.20.40.60.81.01.2Index,1990=1.0Reference1990+24%
1990+9%
1990-3%
CommercialDeliveredEnergyIntensityandPrice,History1970197519801985199019950.00.51.01.5Index,1990=1.0HistoryProjectionsIntensityPriceFigure 41. Index of Delivered Energy Intensity in
the Commercial Sector, 1970-2020
Sources: History: Energy Information Administration (EIA), State Energy
Data Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997); EIA,
State Energy Price and Expenditure Report 1994, DOE/EIA-0376(94)
(Washington, DC, June 1997); and EIA, Commercial Buildings Energy
Consumption Survey 1992 Public Use Data. Projections: Office of Integrated
Analysis and Forecasting, National Energy Modeling System runs KYBASE.
D080398A, FD24ABV.D080398B, FD09ABV.D080398B, and FD03BLW.
D080398B.
The adoption of more efficient technologies reflects the
reaction to rising fuel prices and a change in the way
commercial consumers are expected to look at purchase
decisions involving energy efficiency if carbon
emissions are severely limited. Most commercial
consumers give some consideration to fuel costs when
buying equipment. A significant increase in fuel prices is
expected to cause consumers to give energy costs greater
weight in the purchase decision, by seeking out more
information about energy efficiency options and by
accepting a longer time period to recoup the additional
initial investment typically required to obtain greater
energy efficiency. While taking client comfort and
employeesÕ working conditions into consideration,
commercial energy consumers would also be expected
to turn thermostats down (up) a few degrees during
cooler (warmer) weather and to be more conscientious
about turning off lights and office equipment not in use.

The vast majority of the projected commercial sector
reductions in carbon emissions in the carbon reduction
cases are related to electricity use (see Figure 39). Two
factors contribute to electricity-related carbon savings:
reductions in the level of carbon emitted during the
generation of a given amount of electricity (as discussed
in Chapter 4), and reductions in electricity consumption.
The projections for delivered electricity consumption in
the commercial sector in 2010 for the 1990-3% and
1990+9% cases are 19 percent and 12 percent lower,
respectively, than the reference case projection (Figure
42), and the 1990+24% case is 5 percent lower.

Historically, steady growth in electricity consumption
has been seen in the commercial sector during times of
both rising and falling prices. The growth has resulted in
part from expansion in the sector and, more importantly, from an increasing number of end uses for
electricity (i.e., increasing electricity intensity). The reference case projects further growth in electricity use
between 1996 and 2010. In the 1990-3% case, however,

197019801996Reference1990+24%
1990+9%
1990-3%
0246810QuadrillionBtuNaturalGasElectricityPetroleumElecLossesHistoryProjectionsfor2010Figure 42. Delivered Energy Use and Electricity-
Related Losses in the Commercial
Sector, 1970, 1980, 1996, and 2010
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(95) (Washington, DC, December 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.
D080398B, and FD03BLW.D080398B.
Table 6. Change in Projected Penetration Rates for Selected Technologies in the Commercial Sector
Relative to the Reference Case, 2010
(Percent)
Technology 1990+24% 1990+9% 1990-3%
High-Efficiency Boiler .............................................. 19 97 205
Air-Source Heat Pump ............................................. 2 9 10
Ground-Source Heat Pump ......................................... 0 27 150
High-Efficiency Chiller ............................................. 4 18 23
Heat Pump Water Heater ........................................... 29 102 167
Compact Fluorescent Lights......................................... 6 14 24
Electronic Ballast Fluorescent Lights With Reflectors or Controls ............ 14 26 32
Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD09ABV.D080398B, and
FD03BLW.D080398B.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



the electricity consumption projected for 2010 falls to
1996 levels. The growth in commercial sector electricity
intensity is projected to slow in the reference case for the
same reasons that apply to energy intensity, and further
reductions are expected in the carbon reduction cases.

The projected share of total end-use energy services that
each major fuel provides to the commercial sector in
2010 is fairly stable across the different carbon reduction
cases, and each fuelÕs share of energy consumption
within specific end uses (space heating, cooling, water
heating, etc.) shows little change. Electricity does
increase slightly in share, howeverÑup to 2 percentage
points in 2010 in the 1990-3% and 7-percent-below-1990
(1990-7%) cases relative to the reference case.

Because the carbon prices required to meet emissions
reduction targets cause a greater percentage increase in
natural gas prices than electricity prices relative to those
in the reference case, commercial consumers are
expected to curtail their use of equipment powered by
natural gas more than their use of electrical equipment.
In addition, because of its critical nature, the usage pattern of existing commercial refrigeration equipment is
not assumed to change in response to price changes, limiting projected reductions in electricity use for refrigeration to those caused by potential earlier retirements and
purchases of more efficient equipment when prices are
higher.

Finally, the fastest-growing commercial end uses, under
reference case assumptions, include office equipment
and miscellaneous devices powered by electricity (e.g.,
telecommunications equipment, medical imaging
equipment, ATM machines), which are continuing to
penetrate the commercial sector. Although electricity
consumption for these end uses would be responsive to
the price signals resulting from emissions reduction
efforts, their growth still is expected to be faster than
growth in the end uses that consume fossil fuels (primarily space heating and water heating).

The expected effects of carbon emission reduction
efforts on the average efficiencies of equipment stocks in
the commercial sector are exemplified by the projections
for natural-gas-fired space heating equipment. In the
reference case, the average efficiency of natural gas
space heating systems in the commercial sector is projected to increase by 0.6 percent per year through 2010,
and gas heating equipment purchased in 2005 is projected to be about 6.4 percent more efficient than the
average system in use at that time. The 1990+24% case
projects the same level of efficiency improvement and
purchased efficiency. With 2010 natural gas prices
expected to be near 1996 levels in this case (see Figure
40), there is little incentive for purchasers to invest

additional capital in more efficient gas heating systems.
In the 1990+9% case, however, the projected higher gas
prices yield a projected 0.7-percent annual increase in
average stock efficiency and an average efficiency for
new equipment purchases in 2005 that is 7.2 percent
higher than the stock average. Similarly, in the 1990-3%
case, the average stock efficiency for gas heaters in the
commercial sector increases by 0.8 percent per year, and
new gas heating systems are 7.5 percent more efficient,
on average, than the stock average in 2005. Heating
systems typically are purchased only for new construction, for major renovations, or when an existing system
needs to be replaced. Once in place, they typically last
over 20 years. Therefore, the energy savings realized
from purchases of more efficient equipment take time to
accumulate.

Sensitivity Cases

Sensitivity case assumptions were developed for the
1990+9% case, to examine uncertainties about technology development in the commercial sector. Similar
assumptions were developed for each of the demand
sectors, and results were derived from integrated model
runs requiring the entire U.S. energy system, not the
commercial demand sector individually, to meet the
specified emission reduction goals. Much different
results might be expected if only commercial sector
assumptions were modified and/or only the commercial sector was required to meet a specific emissions target, independent from other demand sectors and
utilities.

The low technology sensitivity case assumes that all
future equipment purchases will be made only from the
equipment available to commercial consumers in 1998,
and that commercial building shell efficiencies will
remain at 1998 levels. Alternatively, the high technology
sensitivity assumptions were developed by engineering
technology experts, considering the potential impact on
technology given increased research and development
into more advanced technologies.40 The high technology
sensitivity case includes technologies with higher efficiencies and/or lower costs than those assumed to be
available in the reference case.

The projected carbon prices and fuel prices in the
different sensitivity cases (Table 7) reflect the possible
impacts that changes in the level of technological
progress, across all sectors, may have on the fuel costs
required for the United States to meet a specific
emissions level. Different actions expected in the
residential, commercial, industrial, transportation, and
electricity generation sectors all contribute to meeting
the emissions target. The combination of these actions
results in the projected carbon prices, as each sector is

40Energy Information Administration, Technology Forecast UpdatesÑResidential and Commercial Building Technologies, Draft Report (Arthur D. Little, Inc., June 1998).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Photovoltaics and Fuel Cells
In every carbon reduction case considered in this report,
neither photovoltaics nor fuel cells are projected to gain
significant market penetration, because of their high
costs. With payback periods of more than 20 years, the
success of these technologies seems largely dependent on
reducing production costs and increasing efficiency
(which would result in further cost reductions for the con-
sumer). Federal financial assistance would also play a role
in their success.
Currently, electricity from photovoltaics and fuel cells is
approximately 1.4 to 5.8 times the price to consumers of
electricity from utility grids. Average prices in 1998 were
79 mills per kilowatt for utility power, 112 mills for phos-
phoric acid fuel cells (with no cogeneration), and 461 mills
for photovoltaic systems. To increase the market penetra-
tion rates of the alternative technologies, their costs
would have to be more competitive.
Photovoltaic and fuel cell technologies are examined here
on the basis of their potential for further market penetra-
tion in 2010 for the 1990-3% case and in sensitivity cases
assuming cost reductions (30 to 50 percent), performance
improvements (50 percent for fuel cells, 70 percent for
photovoltaics), and Federal subsidies and credits. Pay-
back periods are calculated for the regions where these
technologies are most likely to penetrate.
The effects of various private and government-assisted
financing plans, such as rolling the cost of the alternative
technology into a mortgage plan, tax credits, and depre-
ciation, are summarized in the chart below. The first pair
of bars shows the projected payback periods in 2010 for
the 1990-3% case with current technology performance
and costs. The other projections incorporate performance
improvements of 50 percent for fuel cells and 70 percent
for photovoltaics, as well as the cumulative effects of vari-
ous methods for reducing the payback periods. The sec-
ond set of bars shows the effects of the assumed
performance improvement. The third includes a 30-
percent production cost reduction, the fourth includes a
50-percent cost reduction, the fifth includes the incorpo-
ration of capital costs into a mortgage plan, and the sixth
includes a tax credit for photovoltaics and depreciation
adjustments for businesses. It is important to note that the
substantial cost reductions and improvements in effi-
ciency (50 percent for fuel cells, 70 percent for photovol-
taics) are merely arbitrary assumptions and are not
calculated projections for future costs and efficiencies.
These assumptions are not included in the carbon reduc-
tion cases or sensitivity cases presented in this report.
Under the most favorable assumptions shown in the
graph, payback periods could be reduced to less than 1
year for fuel cells and 2 years for photovoltaics. Although
penetration levels are hard to predict from payback peri-
ods, it can generally be assumed for the commercial and
residential sectors that paybacks within 3 to 4 years
would be needed for significant penetration. In the
National Appliance Energy Conservation Act, the Federal
efficiency payback standard for appliances is 3 years or
less for investments to be non-burdensome to the con-
sumer. Although some utilities may have payback peri-
ods on their plants of 20 years, building consumers are
more likely to spend their money for efficient technolo-
gies elsewhere if payback periods are over 4 years. To
achieve 3-to 4-year paybacks, both the current perform-
ance and the costs of these alternatives would have to be
improved by the levels shown here; however, the likeli-
hood of such substantial improvements in the next two
decades is small.
Production costs for photovoltaic modules have fallen
from $100 per watt to $4 per watt over the past three dec-
ades, an 11-percent annual decline, but since 1990 they
have declined by an annual average of only 3.9 percent.a
To meet the cost reduction assumptions in these scenar-
ios, the production costs for photovoltaic cells and mod-
ules would have to decline at an average annual rate of 5.6
percent through 2010.
The energy production efficiency of photovoltaic mod-
ules has also improved, to approximately 12 percent
today from 9 percent in 1980.b Reaching the goal of 70 per-
cent improvement in performance, as assumed for this
sensitivity analysis, would require an efficiency level of
20 percent in 2010. Since 1980, the rate of improvement in
performance for photovoltaics has been less than 2 per-
cent annually, whereas a 4.3-percent annual rate would
be needed to achieve a 70-percent improvement by 2010,
and that improvement would also have to be accompa-
nied by cost improvements to achieve a 3-to 4-year pay-
back period. Fuel cells have been on the market for only a
short time, and historical information is not available.
Neither technology appears to be on course to accomplish
such a goal during the period of this analysis, however,
and thus extensive market penetration is not probable for
either photovoltaics or fuel cells.
aEnergy Information Administration, Solar Collector Manu-
facturing Activity 1991, DOE/EIA-0174(91), p. 18; and P. May-
cock, ÒPhotovoltaic Energy Conversion: PV Technology, Cost,
Products, Markets, and SystemsÑForecast 2010,Ó ASES Con-
ference (Albuquerque, NM, June 1998).
bPaul Maycock, PV Energy, personal communication,
August 1998.
05101520YearsFuelCellsPV1990-3%
PerformanceImprovementPerformanceImprovementand30%CostReductionPerformanceImprovementand50%CostReductionPerformanceImprovement,
50%CostReduction,andMortgageIncorporationPerformanceImprovement,
50%CostReduction,
MortgageIncorporation,
andTaxCreditsProjected Payback Periods for Photovoltaic and
Fuel Cell Purchases Under Different Assumptions,
2010
Source: Energy Information Administration, Office of Integrated Analysis
and Forecasting.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



Table 7. Projected Carbon Prices and Average Fuel
Prices for the Commercial Sector in
Technology Sensitivity Cases, 2010
Analysis Case
Carbon Price
(1996 Dollars per
Metric Ton)
Average
Fuel Price
(1996 Dollars per
Million Btu)
Reference ............. Ñ 11.51
1990+9%.............. 163 17.99
1990+9%
Low Technology ........ 243 21.66
1990+9%
High Technology........ 121 15.75
Source: Office of Integrated Analysis and Forecasting, National Energy Mod-
eling System runs KYBASE.D080398A, FREEZE09.D080798A, FD09ABV.
D080398B, and HITECH09.D080698A.
expected to reduce demand in a way suitable to that
particular sector.

Among the technology cases the highest carbon prices,
and thus the highest fuel prices, in 2010 are projected in
the 1990+9% low technology sensitivity case. Due to the
lack of technological progress in all sectors, higher fuel
prices are required to achieve the demand reductions
needed to reach the emissions target. The projected price
of fuel to the commercial sector is 20 percent higher in
the low technology case than in the 1990+9% case,
resulting in 7 percent less commercial energy use.
Commercial expenditures for fuel are also expected to
be highest under these conditions (Figure 43). Fewer
options for increased efficiency limit the potential for
energy savings in the low technology case. The average
efficiency of the equipment stock in this case continues
to improve as normal turnover takes place and older
equipment is replaced, but the most energy-efficient
equipment available for purchase in 2010 or 2020 is what
is available today (Table 8).

199520002005201020152020050100150200Billion1996Dollars1990+9%
LowTechnology1990+9%
1990+9%
HighTechnologyReferenceFigure 43. Projected Fuel Expenditures in the
Commercial Sector in Low and High
Technology Cases, 1996-2020
Source: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FREEZE09.D080798A, FD09ABV.
D080398B, and HITECH09.D080698A.
In the 1990+9% high technology sensitivity case, advanced technologies are expected to penetrate the
market in all sectors over time as normal stock turnover
results in the replacement of older, less efficient
equipment. Projected technological advances throughout the energy market result in a carbon price in 2010
that is 25 percent lower than that projected in the
1990+9% case (see Table 7). In turn, the expected commercial fuel price in 2010 is 12 percent lower than in the
1990+9% case, resulting in 4 percent more energy consumption. Even though more advanced technologies are
available in the high technology case, with less price
incentive, commercial consumers are not as likely to
purchase more costly equipment. For technologies such
as commercial natural gas water heaters, where high

Table 8. Projected Highest Available and Average Efficiencies for Newly Purchased Equipment in the
Commercial Sector, 2015
Technology 1998
1990+9%
Low Technology 1990+9%
1990+9%
High Technology
Highest Available Efficiencya
Air-Source Heat Pump .................... 2.70 2.70 2.93 3.22
Natural Gas Chillers and Air Conditioners ..... 3.52 3.52 3.81 4.40
Heat Pump Water Heater.................. 2.00 2.00 2.50 2.80
Natural Gas Water Heater ................. 0.91 0.91 0.91 0.91
Average Purchased Efficiencya
Electric Space Heating.................... 1.10 1.13 1.13 1.11
Natural Gas Space Cooling ................ 1.32 1.73 1.62 1.59
Electric Water Heating .................... 0.95 1.03 1.00 0.98
Natural Gas Water Heating ................ 0.79 0.82 0.82 0.84
aThe efficiencies shown (Btu of output divided by Btu of input) generally are seasonal efficiencies or include some measure of losses incurred
during normal use.
Source: Office of Integrated Analysis and Forecasting, National Energy Modeling System runs FREEZE09.D080798A, FD09ABV.D080398B, and HITECH09.
D080698A.
Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



technology assumptions specify lower costs in 2015 for
the most efficient equipment, as compared with the reference case technology assumptions, more consumers
are expected to adopt the efficient technology (see Table
8). The projected reduction in energy demand in other
sectors causes commercial fuel prices to decline in the
later years of the forecast, lowering commercial expenditures for fuel (Figure 43).

Industrial Demand
Background

The industrial sector includes agriculture, mining, construction, and manufacturing activities. The sector consumes energy as an input to processes that produce the
goods that are familiar to consumers, such as cars and
computers. The industrial sector also produces a wide
range of basic materials, such as cement and steel, that
are used to produce goods for final consumption.
Energy is an especially important input to the production processes of industries that produce basic materials.
Typically, the industries that are energy-intensive are
also capital-intensive. Industries within the sector compete among themselves and with foreign producers for
sales to consumers. Consequently, variations in input
prices can have significant competitive impacts. The
most significant determinant of industrial energy consumption is demand for final output.

Although energy is an important factor of production, it
is not large in terms of annual manufacturing expenditures. In 1995, for example, purchased energy expenditures were 2.3 percent of annual manufacturing
outlays.41 Technology usually plays a minor role in the
pattern of energy consumption, because technology
tends to be used to produce new and improved final
products rather than to reduce energy consumption;
however, when new investments are undertaken to
introduce improved production technology, steps to
increase energy efficiency also are undertaken. Overall,
energy prices and technological breakthroughs tend to
have a rather small impact on industrial energy consumption.42

The influence of energy prices on industrial energy consumption is modeled in terms of the efficiency of use of
existing capital, the efficiency of new capital additions,
and the mix of fuels used. This analysis uses Òtechnology
bundlesÓ to characterize technological change in the
energy-intensive industries. This approach is dictated
by the number and complexity of processes used in the
industrial sector and the absence of systematic cost and
performance data for the components. These bundles
are defined for each production process step (e.g., coke
ovens) for five of the industries and for end use (e.g.,
refrigeration) in two of the industries. The process-step
industries in the NEMS model are pulp and paper, glass,
cement, steel, and aluminum.43 The industries for which
technology bundles are defined by end use are food and
bulk chemicals.

The rate at which the average industrial energy intensity
declines is determined primarily by the rate and timing
of additions to manufacturing capacity. The rate and
timing of additions are functions of retirement rates and
industry growth rates. Typical retirement rates range
from 1 percent to 3 percent annually. The current model
also allows retirement rates and the energy intensity of
new additions to vary as a function of price. Price elasticity of demand, which indicates the responsiveness of
energy consumption to changes in energy prices, is not
an explicit assumption in the model; however, the typical 20-year price elasticity ranges between -0.2 and -0.3,
which indicates that a 1-percent price increase would
reduce demand by 0.2 to 0.3 percent. Because the reference case approximates a constant price regime, the reference case results do not differ greatly from a situation
in which all prices are held constant.

In 1996, the industrial sectorÕs consumption of 34.6
quadrillion Btu accounted for more than one-third of all

U.S. energy consumption. The associated emissions of
476 million metric tons of carbon accounted for one-
third of all U.S. carbon emissions. In 1996, although
industrial energy prices were more than 50 percent
lower than in 1980 (Figure 44), delivered energy consumption was only 13 percent higher than in 1980.
Industrial output increased by more than 30 percent
over that period. As a result, energy intensity (thousand
Btu consumed per dollar of output) fell by 20 percent.
41Calculated form U.S. Department of Commerce, 1995 Annual Survey of Manufactures, pp. 1-7 and 1-36.

42For a variety of views, see Boyd et al., ÒSeparating the Changing Composition of U.S. Manufacturing Production from Energy Efficiency Improvements: A Divisia Index Approach,Ó The Energy Journal, Vol. 8, No. 2 (1987); Doblin, ÒDeclining Energy Intensity in the U.S.
Manufacturing Sector,Ó The Energy Journal, Vol. 9, No. 2 (1988); Howarth, ÒEnergy Use in U.S. Manufacturing: The Impacts of the Energy
Shocks on Sectoral Output, Industry Structure, and Energy Intensity,Ó The Journal of Energy and Development, Vol. 14, No. 2 (1991); Jacard,
Nyober, and Fogwill, ÒHow Big is the Electricity Conservation Potential in Industry?Ó The Energy Journal, Vol. 14, No. 2 (1993); Steinmeyer,
ÒEnergy Use in Manufacturing,Ó in Hollander, ed., The Energy-Environmental Connection (Island Press, 1992), Chapter 10; and U.S. Department of Energy, Comprehensive National Energy Strategy (Washington, DC, April 1998), pp. 13-14.

43The refining industry is modeled separately in the Petroleum Market Module of NEMS.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



200020052010201520201.01.52.02.53.0Index,2000=1.0197019751980198519901.01.52.02.53.03.5Index,1970=1.0IndustrialDeliveredEnergyPrices,
HistoryReference1990+9%
1990-3%
1990+24%
Figure 44. Index of Industrial Sector Energy Prices,
2000-2020
Sources: History: Energy Information Administration, State Energy Price and
Expenditure Report 1993, DOE/EIA-0376(93) (Washington, DC, December
1995). Projections: Office of Integrated Analysis and Forecasting, National
Energy Modeling System runs KYBASE.D080398A, FD24ABV.D080398B,
FD09ABV.D080398B, and FD03BLW.D080398B.
Most of the drop in energy intensity in the U.S.
industrial sector occurred between 1980 and 1985, when
prices for both energy and capital inputs were rising and
the ability of U.S. manufacturers to compete internationally was deteriorating. The recessions of 1980 and
1981-1982 forced many less efficient plants to close,
many permanently. Particularly hard hit were the
primary metals industries and motor vehicle manufacturing. Output of the U.S. steel industry has never
recovered to the levels of the late 1970s. Manufacturing
profits did not return to the levels attained in 1981 until
1988.44 Energy prices certainly played a role in shaping
these changes in the industrial sector, but general economic conditions, recession, record high interest rates,
and reduced ability of key industries to compete in international markets were more important determinants of
change.45

In the reference case, industrial energy prices are projected to increase very slightly or fall through 2010. For
example, the price of natural gas is projected to increase
by 0.5 percent, and the price of electricity is projected to
fall by 16 percent. From 1996 to 2010, industrial output is
projected to grow by 39 percent and energy consumption by only 16 percent. Industrial intensity falls by 17
percent during the same period, approximating the
intensity decline between 1980 and 1996. The factors that
are expected to produce the rapid decline in industrial
energy intensity despite moderate changes in energy

prices include a relative shift from energy-intensive to
less energy-intensive industries; replacement of existing
equipment with less energy-intensive equipment as
existing capacity is retired; adoption of improved and
less energy-intensive technologies; and the pressures of
international competition.

Carbon Reduction Cases

In the carbon reduction cases, the combined effect of
reduced demand for U.S. industrial output and higher
energy prices produces lower energy consumption than
in the reference case. Compared with the reference case
in 2010, industrial output is $69 billion (1 percent) lower
in the 1990+24% case, $157 billion (3 percent) lower in
the 1990+9% case, and $308 billion (6 percent) lower in
the 1990-3% case (see Table 29 in Chapter 6).

Compared with the reference case, average energy
prices in the industrial sector in 2010 are projected to be
22 percent higher in the 1990+24% case, 55 percent
higher in the 1990+9% case, and 95 percent higher in the
1990-3% case. In comparison, the industrial sectorÕs
average energy price increased by almost 189 percent
from 1970 to 1980. Prices of all fuels are projected to be
higher in the carbon reduction cases, with coal prices 135
percent higher than the reference case in 2010 in the
1990+24% case and natural gas prices 33 percent higher.
The projected price increase for coal is attributable solely
to the projected carbon price, whereas the carbon price
and higher demand contribute about equally to the
increase for natural gas. In the 1990+9% case, natural gas
and coal prices are projected to be 93 percent and 328
percent higher, respectively, than in the reference case,
and in the 1990-3% case they are 162 percent and 589 percent higher.

Lower projections of industrial output and higher
projected energy prices reduce the projections for delivered energy consumption in the industrial sector by

0.7 quadrillion Btu (2 percent) in the 1990+24% case, by
1.3 quadrillion Btu (4 percent) in the 1990+9% case, and
by 2.3 quadrillion Btu (7 percent) in the 1990-3% case in
2010 relative to the reference case (Figure 45). In the
1970-1980 period, industrial consumption was unchanged even though prices increased by 189 percent.
Year-to-year industrial energy consumption began to
fall in 1980, and the decline accelerated when general
economic conditions began to deteriorate during the
1980 and 1981-1982 recessions. Energy consumption
reached its minimum in 1983, even though prices had
begun to decline. These events reinforce the concept that
while energy prices do play a role in industrial energy
44Council of the Economic Advisers, Economic Report of the President (Washington, DC, February 1995), p. 381.

45For example, see Boyd and Karlson, ÒImpact of Energy Prices on Technology Choice in the U.S. Steel Industry,Ó The Energy Journal, Vol.
14, No. 2 (1993). More general discussion can be found in Berndt and Wood, ÒEnergy Price Shocks and Productivity Growth: A Survey,Ó in
Gordon et al., eds., Energy: Markets and Regulation (Cambridge, MA: MIT Press, 1987); and Berndt, ÒEnergy Use, Technical Progress and Productivity Growth: A Survey of Economic Issues,Ó Journal of Productivity Analysis, Vol. 2 (1990).

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



1970198019902000201020200.70.80.91.01.11.21.3Index,1996=1.0Reference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
ProjectionsHistoryFigure 45. Index of Delivered Energy Consumption
in the Industrial Sector, 1970-2020
Sources: History: Energy Information Administration, State Energy Data
Report 1995, DOE/EIA-0214(96) (Washington, DC, December 1997).
Projections: Office of Integrated Analysis and Forecasting, National Energy
Modeling System runs KYBASE.D080398A, FD24ABV.D080398B, FD1998.
D080398B, FD09ABV.D080398B, FD1990.D080398B, FD03BLW.D080398B,
and FD07BLW.D080398B.
consumption, general and industry-specific economic
conditions also play an important role.

Coal consumption is projected to drop sharply in the
carbon reduction cases, given its extreme price
disadvantage. In the 1990+24% case, coal consumption
in 2010 is lower by 422 trillion Btu (16 percent) than in
the reference case; in the 1990+9% case it is 737 trillion
Btu (28 percent) lower; and in the 1990-3% case it is
about 1 quadrillion Btu (36 percent) lower. The projected
reductions in coal consumption are predominantly due
to projected reductions in boiler fuel use.

The industrial sector consumes coal mainly as a boiler
fuel and for production of coke in the iron and steel
industry. For example, 75 percent of manufacturing consumption of steam coal was used in boilers in 1994.46
Coal-fired boilers have substantially higher capital costs
than do gas-fired boilers, because of their materials handling requirements. For large steam loads, however,
coalÕs price advantage over natural gas offsets its capital
cost disadvantage. But in the carbon reduction cases,
coal suffers from both a capital cost and a fuel cost disadvantage. As a result, a substantial amount of boiler fuel
use switches from coal to natural gas and petroleum
products.

The projected reduction in total steam coal consumption
in the industrial sector in 2010 (including for uses other
than boiler fuel) in the 1990-3% case relative to the reference case is more than 50 percent. Still, the reduction is
less severe than that projected for the electric utility

sector. Electricity generators, in addition to switching to
natural gas, also have the available options of nuclear
power and renewable energy sources.

Consumption of metallurgical coal, which is used to
produce coke for iron and steel production, also is
reduced sharply in the carbon reduction cases. The
reduction has several causes: substitution of natural gas
in production processes, replacement of domestic coke
production with coke imports, replacement of some
coke-based steelmaking capacity with electricity-based
capacity, and reduced production of domestic steel.

In the carbon reduction cases, natural gas consumption
is subject to two countervailing effects. The effect of generally higher energy prices, and consequent lower levels
of industrial activity, is to reduce natural gas consumption. On the other hand, natural gas prices do not
increase by as much as the prices of competing fuels. As
noted above, this results in relatively greater use of natural gas as a boiler fuel. The carbon reduction cases also
induce additional cogeneration using natural gas, which
increases natural gas consumption and reduces requirements for other boiler fuels.

In the 1990+24% and 1990+9% cases, natural gas consumption is projected to increase slightly, because the
impact of increased boiler fuel use outweighs the reduction caused by lower industrial output. In the 1990-3%
case, natural gas consumption is unchanged from the
reference case in 2010. Here, the drop in industrial output and the substitution for other boiler fuels have offsetting effects.

In the reference case, industrial carbon emissions are
projected to be 83 million metric tons higher in 2010 than
they were in 1996 (Figure 46). Emissions attributable to
increased electricity consumption account for more than
half the increase. In contrast, electricity-based emissions
account for more than 70 percent of the emissions
reductions in the carbon cases. For example, in the
1990+9% case, electricity-based carbon emissions in
2010 are 79 million metric tons lower than in the
reference case. A reduction of 19 million metric tons in
carbon emissions from the combustion of fossil fuels
brings industrial sector emissions to approximately
their 1990 level. Carbon emissions in the 1990-3% case
fall to 418 million metric tons, 58 million tons below the
1996 level and 35 million tons below the 1990 level.
Again, electricity-based emissions account for three-
fourths of the reduction from projected levels in the
reference case.

Part of the reduction in electricity-based carbon
emissions for the industrial sector is due to lower
electricity consumption in the carbon reduction cases

46Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94) (Washington, DC, December
1997), p. 168.

Energy Information Administration / Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity



19901996Reference1990+24%
1990+14%
1990+9%
19901990-3%
1990-7%
0100200300400500600700MillionMetricTonsElectricNon-ElectricProjectionsfor2010HistoryFigure 46. Industrial Sector Carbon Emissions,
1990, 1996, and 2010
Note: Electricity emissions are from the fossil fuels used to generate
the electricity used in this sector.
Sources: History: Energy Information Administration, Emissions of
Greenhouse Gases in the United States 1996, DOE/EIA-0573(96) (Washington,
DC, October 1997). Projections: Office of Integrated Analysis and Forecasting,
National Energy Modeling System runs KYBASE.D080398A, FD24ABV.
D080398B, FD1998.D080398B, FD09ABV.D080398B, FD1990.D080398B,
FD03BLW.D080398B, and FD07BLW.D080398B.
(Figure 47). A larger part of the reduction results from
sharply lower carbon intensity of electricity production.
In the reference case, approximately 16.5 million metric
tons of carbon are emitted in the production of 1
quadrillion Btu of delivered electrical energy, as
compared with only 12.6 million metric tons in the
1990+9% case and only 10.2 million metric tons in the
1990-3% case (38 percent less than in the reference case).

Industrial energy intensity fell by 17 percent between
1980 and 1996. In 1996, approximately 7,100 Btu of
energy was required to produce a dollarÕs worth of
industrial output. In the reference case energy intensity
continues to fall, and in 2010 it is projected that only
5,900 Btu will be required for each dollar of industrial
output. The impact of the carbon reduction cases on
industrial energy intensity results from opposing
effects. The effect of higher energy prices is to reduce
energy intensity, whereas reduced or falling output
growth limits the amount of new, less energy-intensive
capital equipment that will be added to the existing
stock, thereby retarding the rate of decline in energy
intensity. Additional structural shifts in the composition
of industrial output further reduce energy intensity.
(Fuel switching contributes to reduced carbon but does
not affect energy intensity.)

The projected rate of decline in industrial energy
intensity is smaller in the more stringent carbon
reduction cases (Figure 48). Some process steps in the
energy-intensive industries approach the minimum
level of energy intensity assumed to be practically
achievable. In addition, in the more stringent carbon
reduction cases, industrial output is more severely