June 2013...Doreen Griffiths, NEPI Administrative Director Charles Wohlforth, author and journalist,...

Tony Knowles • Mary Haddican • Charles Wohlforth

www.nepinstitute.org

June 2013

Building America’s Energy Future - A Portfolio of Promising Policies

NEPI

The National Energy Policy Institute (NEPI) is based in Tulsa, Oklahoma, at The University of Tulsa. NEPI was createdin 2008 to provide policymakers with better research with which to design energy policy. NEPI provides expertenergy economic analysis for national policies that reduce oil use and oil imports as well as carbon dioxide andother air emission pollutants.

NEPI Project Acknowledgements

Tony Knowles, NEPI PresidentMary Haddican, NEPI Project Manager, Research DirectorDoreen Griffiths, NEPI Administrative DirectorCharles Wohlforth, author and journalist, and text writer of the reportHillard Huntington, Senior Adviser on the project, and Executive Director of the Energy Modeling Forum,

Stanford University, Stanford, CA OnLocation, Inc., consultant for NEMS-NEPI modeling, especially Lessly Goudarzi, Frances Wood

and Sharon Showalter, Vienna, VACreative Matters Studio, especially Mark Leavitt, Senior Art Designer of the report, Tulsa, OK

NEPI would like to extend special thanks to the following authors of papers which were commissioned by theNational Energy Policy Institute and provided technical and background information for the project:

Stephen P. A. Brown, Ph.D., Professor of Economics and Director of the Center for Business and Economic Research,University of Nevada, Las Vegas, and Ryan T. Kennelly, Economic Analyst at the Center for Business and EconomicResearch at the University of Nevada, Las Vegas.

Anna Lee Deal, Coordinator of Green Initiatives for Lynden Incorporated.

Mark A. Foster, M.A.F.A., Utility Consultant and former State Public Utilities Commissioner from Alaska.

Kenneth Gillingham, Ph.D., Assistant Professor of Economics at Yale University.

Geoffrey Rothwell, Ph.D., Senior Lecturer, Department of Economics, Stanford University.

Roberton C. Williams, Ph.D., Associate Professor, Department of Agricultural and Resource Economics, Universityof Maryland.

NEPI acknowledges the generous funding for this project from the George Kaiser Family Foundation.

Recommended bibliographic listing: National Energy Policy Institute. Building America’s Energy Future - A Portfolioof Promising Policies 2013

Copyright ©2013 by the National Energy Policy Institutewww.nepinstitute.org

PROLOGUE

America’s economic prosperity is inextricably linked to its use of energy. While past energy practices havefueled this country’s unparalleled economic growth, it is time to acknowledge the challenges we face in theabsence of national energy policy. Excessive dependence on foreign oil and continued massive emissions ofcarbon dioxide and other airborne pollutants threaten our national and economic security, our health, and ourenvironment. To meet these threats, we need domestically produced energy that is affordable, efficient, lesscarbon-based, and more sustainable. Such goals demand change. That change is development of acomprehensive national energy policy.

Why do we need a policy to achieve these goals? In our economic system, the marketplace provides themost efficient allocation of goods and services. But some costs of the energy we use are not reflected in themarket price of that energy, including the costs imposed by our vulnerable foreign oil supply on our economyand security, the illnesses and premature deaths caused by air pollution, and the unabated direction of climatechange. It is the responsibility of our elected leaders to craft a national energy policy that addresses thesesocial costs, primarily through comprehensive, market-driven policies, while preserving the vital link betweenenergy and our economic prosperity.

To help our leaders address America’s energy future, NEPI has engaged some of the country’s leadingenergy economists to construct a comprehensive methodology, parameters, and components for a nationalenergy policy.

Our study begins with quantifiable targets and timetables. Our goal is to reduce oil use and imported oil by16% and carbon dioxide emissions by 15% from 2010 levels by 2035. This improvement would includeachieving 80% of electric generation from clean energy.

Next, building upon a previous study, we used a common model and metrics to analyze the five mostpromising policies, all of which use existing technology. Finally, the combination of these potentialcomplementary policies was analyzed for synergistic results.

This work told us how far we would advance toward the targeted goals with each policy, as well as with thecombined portfolio, and helped determine the individual and portfolio costs. The five policies consisted of aclean energy standard, an oil security dividend, LNG for transportation, automotive fuel economy, and energyefficiencies. Each policy was tested for cost effectiveness and ability to address the scale of the problem inaccomplishing the two targets as well as the societal welfare cost, effect on the national economy and federalbudget, and out of pocket consumer costs.

The results are both surprising and encouraging. We learned we can achieve solutions that are viable, low-cost, and realistic without calling upon undiscovered technologies or political miracles. The work is not offeredas a prescriptive presentation. Rather, it is an exemplary exercise showing the possibilities for a comprehensivechange and offering a blueprint for its structure. Our purpose is to make a positive contribution in the urgentlyneeded debate about building America’s energy future.

Tony KnowlesNEPI President



Table of Contents

Part 1: Need, context and challenges for America's energy policy................................................................11.1 Purpose of the energy policy.............................................................................................................2

1.1.1 The goal of reducing carbon dioxide emissions and air pollution..................................................41.1.1.1 New concern about extremes..........................................................................................41.1.1.2 Climate-related threats........................................................................................................51.1.1.3 Toward a climate solution....................................................................................................61.1.1.4 Reducing air pollution........................................................................................................6

1.1.2 The goal of reducing oil vulnerability...........................................................................................61.2 Current status of achieving an energy policy.........................................................................................8

1.2.1 Past energy policy initiatives......................................................................................................91.2.1.1 Sulfur dioxide and nitrous oxide cap-and-trade.....................................................................91.2.1.2 CAFE Standards.............................................................................................................101.2.1.3 Energy efficiency standards for appliances and buildings....................................................101.2.1.4 Federal Air Quality Act and EPA requirements...................................................................111.2.1.5 State renewable energy portfolio standards.......................................................................111.2.1.6 Regional carbon reduction schemes..................................................................................111.2.1.7 Federal ethanol and renewable energy policy....................................................................12

1.2.2 Results of current policy........................................................................................................131.3 The purpose of this report............................................................................................................14

1.3.1 Pricing policy: benefits and lack of political viability....................................................................141.3.2 Finding alternatives to pricing policy..........................................................................................15

1.4 Previous work and key updates......................................................................................................151.5 Targets.........................................................................................................................................16

1.5.1 Oil use reduction target............................................................................................................161.5.2 Carbon dioxide emission target...............................................................................................171.5.3 Air pollution reduction targets..................................................................................................18

1.6 Metrics..........................................................................................................................................181.6.1 Welfare costs..........................................................................................................................19

1.6.1.1 Welfare costs and market efficiency....................................................................................191.6.2 Additional measures of policy impacts.......................................................................................201.6.3 Allocating costs to combined goals...........................................................................................20

1.7 Modeling.......................................................................................................................................201.8 Refining the five most promising policies............................................................................................21

Part 2: Five Promising Policies: Explanation and impact of the components..................................................222.1 The NEPI Clean Energy Standard (NCES)..........................................................................................22

2.1.1 How the NCES works.............................................................................................................232.1.2 Alternative designs of a CES....................................................................................................242.1.3 How the NCES differs from an RPS...........................................................................................252.1.4 NCES benefits and modeling results..........................................................................................26

2.1.4.1 Electricity generation by fuel source...................................................................................262.1.4.2 Carbon dioxide emissions................................................................................................272.1.4.3 Electricity sales...............................................................................................................282.1.4.4 Electricity prices..............................................................................................................282.1.4.5 Natural gas prices.......................................................................................................292.1.4.6 Emissions of other pollutants.............................................................................................29

2.1.5 Policy costs..............................................................................................................................302.1.6 Regional and state price impacts of the NCES..........................................................................30

2.1.6.1 Regional and state-by-state results......................................................................................312.1.6.2 Solutions to price differences..............................................................................................31

2.1.7 Technology limits and renewable energy....................................................................................322.1.8 Nuclear power challenges and opportunities.............................................................................33

2.1.8.1 Financial issues for nuclear power......................................................................................332.1.8.2 Safety issues for nuclear power...........................................................................................342.1.8.3 Environmental issues for nuclear power..............................................................................35

2.2 Oil Security Dividend: A fully rebated oil tax.......................................................................................352.2.1 How the Oil Security Dividend works.........................................................................................362.2.2 Alternative designs of the Oil Security Dividend...........................................................................372.2.3 Oil Security Dividend benefits and modeling results.....................................................................372.2.4 Policy costs..............................................................................................................................38

2.3 Automotive fuel economy..............................................................................................................392.3.1 How the CAFE Standard works.................................................................................................402.3.2 An alternative policy: feebates..................................................................................................412.3.3 CAFE Standard benefits and modeling results.............................................................................412.3.4 Policy costs.............................................................................................................................42

2.4 LNG Trucks.................................................................................................................................432.4.1 A policy supporting LNG trucks..............................................................................................432.4.2 Alternative policies not selected................................................................................................442.4.3 LNG truck benefits and modeling results ....................................................................................442.4.4 Policy costs.............................................................................................................................442.4.5 Business case and challenges for LNG trucks..............................................................................45

2.4.5.1 Limited fueling infrastructure..............................................................................................462.4.5.2 Fuel temperature issues.....................................................................................................462.4.5.3 Cost to upgrade maintenance shops..................................................................................462.4.5.4 Limited engine options......................................................................................................46

2.5 Energy efficiency............................................................................................................................462.5.1 A policy for energy efficiency....................................................................................................472.5.2 Energy efficiency benefits and modeling results...........................................................................482.5.3 Policy costs............................................................................................................................482.5.4 Challenges for building energy codes........................................................................................49

Part 3: Conclusion: Combining policies into a complementary, integrated whole...............................................503.1 Review of policies and goals.........................................................................................................503.2 Policy benefits and targets.............................................................................................................51

3.2.1 Policy combination modeling results..........................................................................................523.2.1.1 Effectiveness of combined policies.....................................................................................533.2.1.2 Policy cost of combined policies.........................................................................................55

3.3 Conclusions....................................................................................................................................56

Appendix................................................................................................................................................58



List of Figures and Tables

Figures Part 1

Figure Title

Fig. 1.1 Petroleum and Other Liquid Fuels Use by Sector - 2010.............................................................1Fig. 1.2 Energy-related CO₂ Emissions by Sector - 2010.......................................................................1Fig. 1.3 U.S. Primary Energy Use and Real Gross Domestic Product 1949 - 2011......................................3Fig. 1.4 Imports of Petroleum and Other Liquids as a Percentage of US Consumption 1949 - 2011.............3Fig. 1.5 Global CO₂ Emissions and Temperature Anomalies 1880 - 2011...............................................3Fig. 1.6 U.S. CO₂ Emissions and Temperature Anomalies 1895 - 2012...................................................4Fig. 1.7a Arctic Sea Ice Extent 09/18/2007........................................................................................4Fig. 1.7b Arctic Sea Ice Extent 08/26/2012.......................................................................................4Fig. 1.8 U.S. Consumption, Production and Imports of Petroleum and Other Liquids 1949 - 2011...............6Fig. 1.9 World Oil Supply by Country - 2011.....................................................................................7Fig. 1.10 U.S. Primary Energy Use and Energy Intensity 1949-2011.......................................................8Fig. 1.11 Power Sector CO₂ Emissions by Fuel - 2010.............................................................................8Fig. 1.12 U.S. Electricity Generation by Source - 2011..............................................................................9Fig. 1.13 Comparison of CAFE Standards and Compliance 1975 - 2011.................................................10Fig. 1.14 Renewable Portfolio Standard Policies - February 2013...........................................................12Fig. 1.15 Atmospheric CO₂ as Measured at Mauna Loa Observatory 1960 - 2012..................................13Fig. 1.16 World Primary Energy Consumption by Country - 2011...........................................................14Fig. 1.17 World Per Capita Primary Energy Consumption - Selected Countries 2011................................14Fig. 1.18 Oil Use Reference Case and Target 2010 - 2035..................................................................17Fig. 1.19 Cumulative Total Energy-related CO₂-e Emissions Reference Case and Target 2010 - 2035.........17Fig. 1.20a Sources of SO₂ Emissions 2008............................................................................................17Fig. 1.20b Sources of NOx Emissions 2008...........................................................................................18Fig. 1.20c Sources of Mercury Emissions 2008......................................................................................18Fig. 1.21 Schematic of National Energy Modeling System.....................................................................21

Figures Part 2

Fig./Table Title

Fig. 2.1 CO₂-e Emissions Reference Case Total and Power Sector, and Target Total Emissions 2010 - 2035...22Fig. 2.2 CO₂ Emissions and Credits Earned as % of New Conventional Coal Plant Emissions................23Fig. 2.3 NCES Technology Credits vs Requirement............................................................................24Fig. 2.4 Percent of Electricity Generation by Fuel, NCES, 2010 - 2035...............................................26Fig. 2.5 Power Sector CO₂-e Emissions Reference Case, RPS and NCES, 2010 - 2035.........................26Fig. 2.6 Percent of Electricity Generation by Source, Reference Case, RPS and NCES 2010 and 2035......27Fig. 2.7a RPS Generation (change from Reference Case) 2010 - 2035.................................................27Fig. 2.7b NCES Generation (change from Reference Case) 2010 - 2035..............................................27Fig. 2.8 Annual Total CO₂-e Energy-related Emissions Reference Case, RPS and NCES 2010 - 2035......28Fig. 2.9 Electricity Sales by Sector Reference Case, RPS and NCES 2010 - 2035.................................28Fig. 2.10 Average Electricity Prices Reference Case, RPS and NCES 2010 - 2035..................................28Fig. 2.11 Natural Gas Spot Prices Reference Case, RPS and NCES 2010 - 2035...................................29Fig. 2.12a SO₂ Emissions Reference Case, RPS and NCES 2010 - 2035................................................29Fig. 2.12b NOx Emissions Reference Case, RPS and NCES 2010 - 2035...............................................29Fig. 2.12c Mercury Emissions Reference Case, RPS and NCES 2010 - 2035...........................................29Fig. 2.13 Percent of Non-hydro Renewable Electricity Generation by Source, Reference Case, RPS

and NCES 2010 and 2035...............................................................................................32

Fig. 2.14 Non-hydro Renewable Electricity Generation by Source, Reference Case, RPS and NCES 2010 and 2035.....................................................................................................33

Fig. 2.15 Components of Net Burden by Income Quintile, Oil Tax with Distribution-Neutralizing Tax Change 2035.............................................................................................................36

Fig. 2.16 Oil Security Dividend, Net Cost per Household Under Three Recycling Options 2013 - 2035........37

Figures Part 3

Fig./Table Title

Fig. 3.1 Oil Use, Reference, Oil Security Dividend, CAFE Standards, LNG Truck Policy, Combination of Policies, Target 2010 - 2035.......................................................................52

Fig. 3.2 Cumulative Total Energy-related CO₂-e Emissions, Reference, NCES, Energy Efficiency, Combination of Policies, Compared to Target 2010 - 2035...................................................52

Fig. 3.3 Annual Total Energy-related CO₂-e Emissions, Reference, NCES, CAFE Standards, Oil Security Dividend, Energy Efficiency, Combination of Policies 2010 - 2035........................52

Fig. 3.4 Natural Gas Heavy Trucks as % of Heavy Truck Fleet, Reference, LNG Truck Policy, Combination of Policies 2010 - 2035..................................................................................53

Fig. 3.5 Motor Gasoline Use, Reference, CAFE Standards, Oil Security Dividend, Combination of Policies 2010 - 2035.....................................................................................................53

Fig. 3.6 Electricity Sales, Reference, NCES, Energy Efficiency, Combination of Policies 2010 - 2035......53Fig. 3.7 Average Electricity Price, Reference, NCES, Energy Efficiency, Combination of Policies 2010 - 2035.....54Fig. 3.8 Electricity Generation by Fuel, Reference, NCES, Energy Efficiency, Combination of

Policies 2010 - 2035........................................................................................................54Fig. 3.9 Total Non-hydro Renewables Generation by Source, Reference, NCES,

Energy Efficiency, Combination of Policies 2010 - 2035........................................................54Fig. 3.10 Natural Gas Spot Prices (at Henry Hub), Reference, LNG Truck Policy, NCES,

Energy Efficiency, Combination of Policies 2010 - 2035........................................................55Fig. 3.11 Annual Total Energy-related CO₂-e Emissions, Actual, Reference Case, NEPI Combination

of Policies 1949 - 2035.....................................................................................................56Fig. 3.12 U.S. Consumption, Production and Imports of Petroleum and Other Liquids 1949 - 2011

Actual; 2012 - 2035 NEPI Combination of Policies................................................................57Fig. 3.13 Electricity Generation by Fuel Reference Case, NEPI Combination of Policies 2010 and 2035........57

Tables Part 1

Table Title

Table 1.1 Emissions, Heat Input, and Emission Rates from Acid Rain Program Sources 1990 - 2010............9

Tables Part 2

Table Title

Table 2.1 NEPI Clean Energy Standard Policy Key Metrics.....................................................................30Table 2.2 Oil Security Dividend Policy Key Metrics................................................................................38Table 2.3 CAFE Standards Policy Key Metrics......................................................................................42 Table 2.4 LNG Trucks Policy Key Metrics.............................................................................................45Table 2.5 Energy Efficiencies Policy Key Metrics....................................................................................48

Tables Part 3

Table Title

Table 3.1 Policy Key Metrics...............................................................................................................55


Part 1: Need, context and challenges for America's energy policy

A coordinated energy policy that can meetmajor American objectives for reducing fossilfuel use and emissions is well within the financialmeans of society. That is the conclusion of thisproject. We studied a portfolio of fivecomplementary policies that would substantiallyease U.S. vulnerability to oil imports and cutcarbon dioxide output. By focusing on off-the-shelf technology and existing policy tools, weheld costs exceptionally low. By 2035,consumer out-of-pocket costs were flat ornegative; federal outlays were insignificant, andimpact on Gross Domestic Product was too smallto model. Extending tax incentives reducedfederal revenue $120 billion over 20 years. Thecombination policy isn’t free—costs areexplained throughout the report, andsummarized in the conclusion—but even ourmost pessimistic estimates compare favorablywith the cost society already incurs for livingwithout an energy policy. A smart, aggressivepolicy to reduce carbon emissions and oilimports is something we can easily afford.

We have examined the many interlockingparts of our energy system through a single lenscapable of bringing a comprehensive policy intofocus. The clarity of our lens relies uponquantifiable, achievable targets to state ourgoals; consistent metrics to look at costs; acommon modeling system to forecast policyoutcomes; and analysis of how policies andpolicy combinations can scale up to themagnitude needed to meet the targets. The resultis a package of five policy initiatives, usingreliable, familiar technology, that cansignificantly mitigate U.S. climate changeimpacts and reduce oil use and imports over thenext 20 years at a reasonable cost.

The starting point for our analysis is the year2010. The Energy Information Administration(EIA) reports a wealth of energy data to paint apicture of the energy economy at that time.Figures 1.1 and 1.2 show liquid fuels use andenergy-related carbon dioxide emissions bysector for the year 2010.

The challenge of developing a credible energy

1

Fig. 1.1 Petroleum and Other Liquid FuelsUse by Sector - 19.4 million barrels per day

2010

Data source: EIA

Fig. 1.2 Energy-related CO2 Emissions by Sector - 5,634 million metric tons

2010

Data source: EIA

Residential &Commercial

6%

Electric Power1%

Industrial22% Transportation

71%

Residential6%

Commercial4%

Transportation33%

Electric Power41%

Industrial16%

policy is to exercise imagination within the realm ofthe possible. Choices must be evaluated on an evenfooting, taking into account economic and socialimplications, not only technological promise. Anappealing alternative without political viabilitycannot be included. A meaningful energy policymust be practical to adopt and implement in thenear term, and able to achieve benefit in the long-term for the nation and its people.

We have addressed this complex planningchallenge through an approach that is conceptuallysimple. Beginning with our 2010 report, written inpartnership with Resources for the Future,1 weconsidered the broad array of energy options. Weselected specific oil consumption and carbonemission goals that were achievable and justifiablybeneficial. We compared 35 policies andcombinations of policies to work toward our goalsusing consistent metrics and a common modelingsystem. Finally, here, we have combined the mostpromising of those options in a clear, integratedpackage derived from the analytical process—thepackage that could take us toward our goal at areasonable social and economic cost.

The five elements in our package of policies are:

• NEPI Clean Energy Standard: Establishes a 2035goal for clean electrical generation and requirestechnology-neutral, market-driven decisions byutilities to use low-carbon and renewable fuels toachieve this target.

• Oil security dividend: Imposes a modest,graduated oil products tax that is fully rebatedthrough the tax system to the public.

• Automotive fuel economy: Follows the CorporateAverage Fuel Economy (CAFE) Standards asadopted in 2012, and considers an alternative‘feebate’ mechanism with similar impact andpotential advantages.

• LNG trucks: Provides a temporary fuel taxreduction to encourage the nascent shift in theheavy truck market from diesel to liquefied naturalgas (LNG).

• Energy efficiency: Extends existing incentives forenergy-saving appliances and building codes andhome-based renewable energy.

Each strategy works by itself, but thecomprehensive policy is more than a shopping cartof unrelated initiatives. Like a well-cooked dishwhose flavors combine into a successful whole,these five individual policies are designed tocomplement one another for greater collectivebenefits. The whole is greater than the sum of itsparts, with significantly increased impact, becauseof the way the integrated policies interact to boosttheir collective effectiveness. Through modeling andanalysis, only policies with complementary impactsare included. The combination of policies isdiscussed in detail in Part 3.

Our intent is not to advocate for these particularchoices. An adopted national energy policy willsurely have other features, and some options wehighlight may not be favored through the politicalprocess. We instead present this work as an aid forpolicy makers to move forward, asking them to useits analysis as a light showing the way tomeaningful reductions in our use of fossil fuels,dependence on foreign oil, and emissions ofgreenhouse gases.

1.1 Purpose of the energy policy

Our country’s well-being depends onfulfilling national needsthat are not currentlyvalued in the energymarket. Americans payreal costs for the effects ofclimate change, the healthimpacts of air pollution,and the economic andnational security burdenof dependence uponinsecure oil supplies. Thepremise of this project isto reduce the total socialand economic price ofenergy by embeddingthese costs in the energymarket rather thancontinuing to allowenergy users to imposethem on others without

2

1Resources for the Future and National Energy Policy Institute.Toward a New National Energy Policy: Assessing the Options.Washington, DC: Resources for the Future, 2010.

paying. With initiatives in place to account for thereal costs of energy use, the market itself caneffectuate our policy goals. We need not pickwinners among technologies or industries.

Implicit in the use of the market to achievenational goals is the recognition that affordable,readily available energy is essential to theAmerican way of life. A subsidiary goal of anypractical national energy policy must be to avoidinterference in energy markets that wouldsignificantly diminish standards of living. Figure1.3 shows the historical relationship betweenenergy use and the growth of the economy. Amarket-based policy approach supports thissubsidiary goal by blending national prioritieswith private needs. The alternate approach, ofmandating technologies and programs, often costsmore and works less effectively because it does nottake advantage of the market’s ability toharmonize these conflicting interests.

Society’s cost for climate change cannot becalculated precisely, but ample evidence points toa developing threat to national security and theeconomy, as well as peril to the natural world.These tangible and intangible costs, already real,

will increase over decades, if allowed, andeventually become catastrophic. Reducing emissionsfrom burning fossil fuels is the most effective way tomitigate climate change. The human health gainsfrom reduced emissions of air pollution are anadditional and immediate benefit.

America’s excessive dependence on oil andimported oil (Figure 1.4) also threatens our nationalsecurity, at the cost of lives and treasure, and

exposes our economy to painful burdens andshocks. International agreements prevent us fromlimiting oil use by reducing imports through quotasor tariffs. Instead, we can adopt other policies thatreduce our use, while allowing domestic productionto grow. Reducing our use of oil also assists withthe goal of lowering carbon dioxide emissions.

The clarity of these problems allows us to stateour policy goals simply. We must reduce ouremissions of carbon and other pollutants from fossil

fuels. We must reduce our use of oil as long aspolitically unstable regions are an important sourceof global supplies. These two goals, meanwhile,must be achieved in the context of America’scontinued prosperity, without diminishing thefreedom and economic vitality empowered byaffordable access to energy.

3

0

2

4

6

8

10

12

14

16

0

20

40

60

80

100

120

Tri

llio

ns

of

20

05

$

Qu

ad

rillio

n B

tus

Energy Use Real GDP

Data source: EIA

Fig. 1.3 U.S. Primary Energy Use and Real Gross Domestic Product 1949 - 2011

(quadrillion Btus; trillion 2005 $)

0%

10%

20%

30%

40%

50%

60%

70%

Imports of Petroleum and Other Liquids as a Percentage of U.S. Consumption

1949-2011

Imports as % of Consumption

Data source: EIA

Figure 1.4

Fig. 1.4 Imports of Petroleum and Other Liquidsas a Percentage of U.S. Consumption

1949 - 2011

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

0

5

10

15

20

25

30

35

Deg

rees

Fah

renhei

t D

evia

tion f

rom

Bas

elin

e

Gig

atons

of CO

2 E

mis

sions

Global CO2 Emissions and Temperature Anomalies 1880-2011

CO2 Emissions Temperature Anomalies

Data sources: CDIAC, EIA and NASA

Figure 1.5

Fig. 1.5 Global CO2 Emissions and Temperature Anomalies

1880 - 2011

0

1.1.1 The goal of reducing carbon dioxide emissions and air pollution

The destruction caused by Superstorm Sandy in2012 renewed the public’s sense of urgency aboutclimate change by making manifest theconsequences of ever-increasing carbon dioxideconcentrations in the atmosphere. Although thestorm’s existence cannot be attributed directly toclimate change, some portion of its severityprobably does relate to elevated sea surfacetemperatures caused in part, over a span of time,by atmospheric warming due to greenhouse gases.Likewise a historic drought across the United Statesin 2012 focused attention on the issue, regardlessof its proximate cause.

These events illustrated the sudden, patchy andunpredictable nature of costly climate changeimpacts. More than a decade ago the scientificcommunity disposed of doubt that human carbondioxide emissions would drive atmospheric warming,but the focus of discussions centered on a seeminglygradual and distant increase in global meantemperature. Through 36 consecutive years of above-average global temperatures (Figure 1.5), theperception of an incremental threat led to a cautiousaccounting of incremental mitigation strategies, up to2012, the warmest year ever in the U.S. (Figure 1.6)Now a different calculation is required by therecognition that the climate’s response could beviolent and regionally disastrous even beforeatmospheric concentrations of CO2 reach themilestone levels cited in international goals. With theprospect of unacceptable costs, mitigation effortsbecome insurance premiums against catastrophe.

1.1.1.1 New concern about extremes

Recent developments in climate science suggestmodeling predictions have been too conservative.The decades-long shrinkage of the Arctic Ocean icesheet, which came to a new record low ice extent

in 2012 (Figures 1.7a and 1.7b), is but one indicatorthat the mid-range estimates of climate sensitivity mayhave been too low—no model predicted ice retreatingat this rate. While the quantity of carbon dioxideadded to the atmosphere is known, the climate’s rateof response, or sensitivity, remains the mostimportant unknown in predicting climate change.

4

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

4.0

0

1

2

3

4

5

6

Deg

rees

Fah

renhei

t D

evia

tion f

rom

Bas

elin

e

Gig

atons

of C

O2 E

mis

sions

US CO2 Emissions and Temperature Anomalies 1895-2012

CO2 Emissions Temperature Anomalies

Data sources: CDIAC, EIA and NOAA

Figure 1.6

Fig. 1.6 U.S. CO2 Emissions and Temperature Anomalies

1895 - 2012

Fig. 1.7a Arctic Sea Ice Extent 09/18/2007

Data source: National Snow and Ice Data Center

Median 1979-2000

Fig. 1.7b Arctic Sea Ice Extent 08/26/2012

Data source: National Snow and Ice Data Center

Median 1979-2000

0

The lower boundary of sensitivity is well established,but the upper boundary has been relatively difficultto pin down and may be theoretically unknowablewith greater precision than already exists.2

The worst-case-scenario of climate sensitivity

becomes especially relevant if we think of carbonreductions as insurance against catastrophe. Forexample, with continued reliance on fossil fuels, theIntergovernmental Panel on Climate Change (IPCC)projects a 3 to 6 degree C increase in global meantemperature by 2100. Scientists predict that a 5degree C increase, which has a one-thirdprobability, would cause loss of major crop speciesand a decline in the human carrying capacity ofequatorial regions. At a global mean increase of 7degrees C, portions of the planet become physicallyuninhabitable due to heat stress, and at 11 to 12degrees C the uninhabitable regions cover most ofthe current population of the Earth.3 The likelihood

of such scenarios can be very low yet still justifysignificant current expenditures to make sure theydo not occur.

1.1.1.2 Climate-related threats

The costs of climate change impacts will reachunacceptable levels far short of global catastrophe.Economic losses are already accumulating fromextreme weather events, changes in waterdistribution and precipitation affecting industriesranging from agriculture to winter tourism, andforest destruction by fire and insect infestation inchanging biomes, to name but a few impacts. Incomplex ecological systems gradual change canreach unknown thresholds that trigger rapid,unanticipated change. Likewise, complex humansystems sometimes exhibit unexpected fragility, aswhen a minor fault in the electrical grid cascades intoa major blackout for a broad swath of the country.

The national security threats due to climatechange fall in this category of unpredictablechange with the potential to cascade intocatastrophe. The National Academy of Science,working with the U.S. intelligence community, hasidentified a long list of social stresses caused byclimate trends or extreme weather events that couldlead to violent internal or international conflict,including water, food and health security,epidemics, humanitarian crises and disruptivemigration. Environmentally vulnerable andpolitically unstable countries would flare first inthese scenarios. For example, drought and resultinghydropower outages in Pakistan in 2010 and 2011led to demonstrations and riots and increased tensionwith India over the Indus River. Both nations arearmed with nuclear weapons.4

Climate change not only imposes costs that aredifficult to know, it also threatens resources forwhich establishing a market price is impossiblebecause their value is intangible or is of a kind thattranscends materialism. IPCC estimates 20 to 30percent of plant and animal species are at risk ofextinction by the end of the century, a rate 10,000times that found in the fossil record.5 Scientistsstudying the acidification of the oceans due to

5

2Roe, et al, Science 318, 629 (2007).3Sherwood and Huber PNAS 2010

4Steinbruner, et al., Climate and Social Stress: Implicationsof Security Analysis; National Research Council, 2012.5http://www.epa.gov/climatechange/impacts-adaptation/ecosystems.html#Extinction

human carbon dioxide emissions say it is probablyalready too late for coral reefs to be expected tosurvive, and many other shell-bearing organismsare at risk if the increase in acidity does not cease.

1.1.1.3 Toward a climate solution

Embedding the price of insurance againstclimate catastrophe in the cost of fossil fuels wouldtend to drive energy use toward fuels that emit lesscarbon and thereby reduce the probability of anunlivable future. The combined uncertainties ofclimate science, economics and geopolitics havefoiled previous attempts to set a precise price onemissions. The error bar on IPCC’s 2007 carbonprice estimate spans two orders of magnitude, from$10 to $350 per ton of carbon dioxide. But lack ofprecision is no reason to ignore a danger that is, inits general outline, very certain. Our plan’s targetfor emission reduction is covered in Section 1.5 ofthis part, along with a discussion of the currenttrajectory of emissions levels.

When it unanimously rejected the Kyoto Protocolin 1997, the U.S. Senate resolved that the U.S.should not commit to reduce greenhouse gasemissions without a similar commitment fromdeveloping countries. But energy users’responsibility to pay the cost of their emissions doesnot depend on whether others also do so.Moreover, the benefit of reductions redounds to theUnited States as well as any other nation. As amajor energy user and emitter of carbon dioxide,the United States may need to take a leadership rolein stimulating action by other countries. As a

practical matter, international negotiations oncarbon reduction offer no real alternative. Theyhave proved costly, ineffective and even potentiallycounter-productive, as the expectation of aninternational agreement diminishes political willfor unilateral reductions. The vast majority ofsuccess to date has come through local, state andnational efforts. This is likely to continue.

1.1.1.4 Reducing air pollution

To cut greenhouse gas emissions, Americans will,among other changes, burn less of the dirtiest fuelsupon which we currently rely for power, the mostimportant of which is coal. Burning coal and oil,using current technology, produces air pollutionwhich injures human health (although oil is cleanerthan coal). Pollutants such as fine particulates,ozone, nitrogen oxides, and sulfur dioxide candamage the respiratory system, and methylmercuryfrom coal can impair children’s neurologicaldevelopment. Americans have already receivedenormous benefits relative to their costs by reducingthese pollutants through the Clean Air Act.Additional reductions in use of dirty fuels will furtherimprove human health and increase lifespans. Seethe sidebar on “Air pollution,” on page 7 for moreon the health aspects of these pollutants.

1.1.2 The goal of reducing oil vulnerability

Oil vulnerability has played an outsized role inthe economic and foreign policy history of theUnited States since the 1970s, when domesticsupplies began to fall seriously short of demand. Inthe past few years U.S. imports have been lowerdue to technological advances boosting domesticproduction, and economic weakness and energyconservation restraining demand (Figure 1.8).America’s oil use declined 2.5 million barrels a day(mmbd) from 2007 to 2011, and our domesticproduction from land and waters yielded anincrease of half a million barrels of crude oil a dayduring that period. Overall, we reduced importsfrom 61% to 50% of our consumption, which nowtotals 17.4 million barrels a day.

The relevant oil issue, however, does not concernpetroleum imports but our vulnerability to overseas

6

0

5

10

15

20

25

Millio

n b

arr

els

/d

ay

U.S. Consumption, Production and Imports of Petroleum and Other Liquids

1949-2011 (million barrels per day)

Consumption Production Imports

Data source: EIA

Figure 1.8

Fig. 1.8 U.S. Consumption, Production and Imports of Petroleum and Other Liquids

1949 - 2011(million barrels per day)

oil price shocks. The global economy continues to relydangerously upon supplies from unstable regions.Figure 1.9 shows oil supply by country. A majordisruption in those regions could send domestic oilprices substantially higher and cause an economicrecession even in countries like the United States thathave reduced their imports. By relying upon thesesources, we may also make ourselves vulnerable toour suppliers’ economic and political agendas.America cannot become totally energy independent—importing no oil at all—but we can achieve greaterenergy security than what we have now.

Sudden oil price increases caused by supplydisruptions, known as price shocks, have repeatedlydamaged the U.S. economy, as measured in lossesin Gross Domestic Product. Oil price spikes preceded10 of the 11 recessions that hit the United States sinceWorld War II.

The United States may reduce its exposure tofuture oil price shocks by reducing its oil use. Aneconomy that is less dependent upon oil would bemore insulated from the damages caused by oilprice shocks. Moreover, policies that reduce oil useoften decrease the amount of oil produced inregions where disruptions are more likely. If eventsshould cause disruptions in these regions, less oilwould be removed from the global market. Oilprices would not rise as much because the oilshortage would be smaller relative to total worlddemand. And finally, the United States economy is

7

Air PollutionThe air we breathe may be our most fundamental common resource. When it

is polluted with smokestack and tailpipe emissions, we have no escape or recourse,yet the illnesses caused thereby harm and kill thousands of Americans. The valueof clean air and the responsibility of industry and society to control emissions arebeyond serious debate.

Combustion for energy production and transportation emit a dominant share ofthe pollutants harmful to human health, including sulfur oxides, nitrous oxides,methylmercury, and particulates. Here are some health effects of these pollutants.

• Sulfur oxides can aggravate respiratory illnesses, especially affecting people withasthma. In the atmosphere, sulfur oxides form small particles that cause or worsendiseases such as emphysema, bronchitis and heart disease. In the environment, sulfuroxides cause acid rain.• Nitrous oxides have impacts similar to sulfur oxides, including respiratory effectsand formation of harmful particles. In addition, nitrous oxides react with sunlight andheat to produce ground-level ozone, which produces smog and harms the lungs.• Methylmercury is the organic form of the element mercury most often affectinghumans, because emissions into the atmosphere accumulate in fish. Among the manyways mercury harms human health, the most severe are to children and fetuses, inwhom it can damage brain development, causing life-long mental deficits. • Particulates are simply particles of chemicals or soot small enough to lodge deepin the lungs. Besides respiratory impacts, they can cause heart attacks, strokes, and avariety of other serious illnesses leading to death.

The health cost of these pollutants for Americans, overwhelmingly from coal andoil combustion in power plants and motor vehicles, totals $120 billion annually,according to a 2009 study by the National Academy of Sciences, mostly from the20,000 premature deaths they cause. (Figures are in 2000 dollars.) The Academycalculated the uncompensated health cost of electricity produced by coal-fired powerplants at 3.6 cents per kilowatt hour, almost a third of the retail price of power for allcustomers in the United States. The similar price for a gallon of gasoline or diesel cameto 23 to 38 cents per gallon. The Academy study found total air pollution health costswere attributable equally to coal and oil.1

Huge improvements can be made to these costs through regulations that requirestate-of-the-art pollution controls. A modern coal power plant can reduce the healthcost of its emissions by more than a factor of 10 over one not using the latesttechnology. EPA regulations that will phase in over the next four years will dramaticallyreduce this pollution. But improvements could continue to increase longer by switchingto fuels that produce less pollution in the first place—first to natural gas, and then toclean energy such as nuclear and renewables. A program to make that kind ofprogressive improvement requires a market-based strategy that includes the cost ofpollution in the price of the power and fuel we use.

This report proposes such market-based policies to address carbon dioxide emissionsand oil consumption, adding to and extending the benefits of the new EPA pollutionregulations on power plants (see Section 2.1.4.6). Carbon dioxide in fact does notdirectly harm human health. Its constituent elements, carbon and oxygen, are thefundamental material of life. But efforts to reduce carbon dioxide emissions to mitigatethe global warming will simultaneously curtail air pollution that damages our health.Likewise, reducing the use of oil, while addressing American dependency on foreignsources, will save lives at home by cleaning the air.

1Matthew L. Wald, “Fossil Fuels’ Hidden Cost Is in Billions, Study Says,”New York Times, October 20, 2009.

Fig. 1.9 World Oil Supply by Country -87.2 million barrels per day

2011

Data source: EIA

Saudi Arabia13%

Russia12%

United States12%

All others15%

China5%Iran

5%

Indonesia1%

Qatar2%

United Kingdom1%

Kazakhstan2%

Angola2%Algeria

2%Norway

2%

Nigeria3%

Venezuela3%

Kuwait3%

Iraq3%

Brazil3%

Mexico3% United

Arab Emirates4%

Canada4%

a large oil consumer. Coordinated policies thatreduce oil consumption within large economiesoften reduce the world oil price, particularly if thegovernments use domestic energy policies ratherthan international trade policies like tariffs or quotas.

The simplest method to reduce oil imports wouldbe an import tax or quota, options lacking politicalviability either domestically or in keeping with freetrade agreements. Instead, we propose to reduceoil imports by reducing oil use while allowingdomestic production to increase. This approachalso supports our goal of reducing greenhouse gasemissions from burning fossil fuels. Our targets forthis goal are covered in Section 1.5.

1.2 Current status of achieving an energy policy

Something like the weather, national energypolicy is a topic everyone talks about but no onedoes anything about. Eight presidents havedeclared the necessity of reducing oil use, yet nocomprehensive national energy policy has everpassed Congress. Instead, states and the federalgovernment have enacted piecemeal legislation forparticular issues in different parts of the energyeconomy. While these policies have accomplishedimportant individual goals, the overall need toreduce oil vulnerability, greenhouse gas emissionsand air pollution remain to be addressed in a broadand coordinated way. Never has the federalgovernment considered all the interlocking parts ofthese issues, the available technology, and the socialand economic costs and benefits. That is our goal.

In part, events have intervened where policy

makers have been unable to act. The last severalyears have seen encouraging trends of increasedenergy efficiency (Figure 1.10), reduced oil use,increases in domestic oil production, and a flatteningof carbon dioxide emission levels from developednations, including the U.S. But some of these changeswere unpredictable and did not all come fromwelcome causes. The world economic crisis andyears of slow recovery depressed use of petroleumand other fossil fuels worldwide. The lack of economicactivity was the primary reason why carbon dioxideemissions did not rise at their previous rate.

Domestic oil exploration and developmentrejuvenated, thanks to new fields and technological

advances for recovering oil from tight shale depositsand with horizontal drilling. The new availability ofabundant natural gas from shale also benefited thenational carbon footprint, as a sharp decline in gasprices put coal at a competitive disadvantage forpower production. Gas-fired plants emit half asmuch carbon, less than a third the nitrous oxides,less than 1 percent of the sulfur oxides andparticulates, and no methylmercury.6

Natural gas mainly improves carbon emissionsby displacing coal for electrical generation. Coalproduces 80% of carbon emissions caused bypower production (Figure 1.11), with power plants

8

0

2

4

6

8

10

12

14

16

18

20

0

20

40

60

80

100

120

Th

ou

san

d B

tus/

20

05

$ o

f G

DP

Qu

ad

rillio

n B

tus

U.S. Primary Energy Use and Energy Intensity 1949-2011

(quadrillion Btus; thousand Btus/2005 $ of GDP)

Energy Use Energy Intensity

Data Source: EIA

Figure 1.10

Fig. 1.10 U.S. Primary Energy Useand Energy Intensity

1949 - 2011(quadrillion Btus; thousand Btus/2005 $ GDP)

Fig. 1.11 Power Sector CO2 Emissions by Fuel - 2,271 million metric tons

2010

Data source: EIA

6http://www.epa.gov/cleanenergy/energy-and-you/affect/natural-gas.html

Petroleum1%

Coal80%

Natural Gas18%

Other<1%

accounting for 40% of U.S. energy-relatedemissions. In the longer run the lower natural gasprice may work against renewable sources otherthan traditional hydropower, such as wind andsolar. Currently, these renewable sources provideless than 5% of electrical generation (Figure1.12)7, even after years of rapid growth.Americans allocate more than 90% of our oil totransportation and industry (Figure 1.1, page 1),sectors which together are responsible for halfour carbon emissions.

1.2.1 Past energy policy initiatives

Despite the lack of a comprehensive nationalenergy policy, Americans have taken action onvarious individual issues affecting our energyeconomy. Some efforts have been small butworthwhile, such as decisions by individual familiesto drive hybrid vehicles or install efficient lighting,while others have been large but inefficient, suchas the federal government’s investment in ethanolas an alternative fuel. America is an experimentand our willingness to try new ideas is a strength.While only a few broad-scale initiatives havemeasurably affected outcomes on the national level,we can draw useful lessons from these individual

pieces of energy policy attempted by the states andthe federal government.

1.2.1.1 Sulfur dioxide and nitrous oxide cap-and-trade

Congress established the federal government’sfirst cap-and-trade emission reduction program in1990, under the leadership of President GeorgeH.W. Bush. The Acid Rain Program (ARP) aimed toreduce emissions of sulfur dioxide that damagelakes, forests and soils through acidic precipitation.Instead of mandating new technology on powerplants that emit sulfur dioxide, the program createda cap on emissions, but allocated permits toindividual emitters that they could buy and sell tomeet the cap. The market in permits allowed thosewho could reduce their emissions at the lowest cost

9

Fig. 1.12 U.S. Electricity Generation by Source4,101 terawatthours

2011

Data source: EIATable 1.1 Emissions, Heat Input, and Emission

Rates from Acid Rain Program Sources1990 - 2010

Data source: EPA

7http://www.eia.gov/electricity/monthly/xls/table_1_01.xlsx

Coal42%

Other0%

Nuclear19%

Natural Gas25%

Other RenewableSources

5%Hydroelectric

8%

PetroleumLiquids

1%Petroleum Coke

0%

SO2 Emissions From ARP Sources 1990 - 2012

NOX Emissions From ARP Sources 1990 - 2012

1990 2000 2005 2010 2012

1990 2000 2005 2010 2012

15.7

11.2 10.2

5.23.3

6.7

5.1

3.6

2.1 1.7

76543210

20

15

10

5

0

Mill

ions

Ton

sM

illio

ns T

ons

1990 2000 2005 2010 2012*SO2 (million tons) 15.73 11.20 10.22 5.17 3.32NOX (million tons) 6.66 5.10 3.63 2.10 1.71Heat Input (billion mmBtus) 19.68 25.62 27.14 27.00 25.28SO2 Rate (lbs/mmBtus) 1.60 0.87 0.75 0.38 0.26NOX Rate (lbs/mmBtus) 0.68 0.40 0.27 0.16 0.14*2012 are preliminary as of February 19, 2013

to profit by their improvements while imposingrelatively manageable costs on those who did nothave cost-effective options to reduce emissions.Later, the Environmental Protection Agency (EPA),which administers the Acid Rain Program, added asimilar cap and trade program for nitrous oxideemissions that create smog-producing ozonepollution and also contribute to acid rain,broadening the concept to reduce unhealthypollution that drifts from state to state.

The EPA’s cap-and-trade programs have gonethrough various changes and court challenges, buttheir record of success is accepted by academia,business and regulators (Table 1.1). EPA’s analysisclaims the programs have reduced tens ofthousands of premature mortalities and saved up to$100 billion in health costs. Sulfur dioxide andnitrous oxide emissions from U.S. power plants bothwent down by about 60% from 2000 to 2011.8

The General Accounting Office estimated the costof these reductions may have been halved by usingcap and trade rather than a command-and-controlregulatory approach.9

1.2.1.2 CAFE Standards

Corporate Average Fuel Economy Standards,first enacted by Congress after the Arab oilembargo of the 1970s, require vehiclemanufacturers to improve the fuel efficiency of theirproducts. Vehicle fuel efficiencies improved for carsbut (until 2005) not trucks as a result of theseprograms (Figure 1.13). By 2001, fuel use was one-third less than it would have been without CAFEStandards, according to the National ResearchCouncil, although higher gas prices also played arole. The federal government further tightened CAFEStandards during the George W. Bush and BarackObama administrations.

CAFE Standards represent the nation’s singularsuccess story and most powerful policy tool forreducing oil use for transportation, but the policy’sunintended consequences were significant as well.By exempting larger passenger vehicles,euphemistically categorized as light trucks,Congress created a perverse incentive for the

explosion in market share of gas-hungry sport utilityvehicles and minivans. By 2005, those categoriesconstituted more than half of vehicles sold.10

Changes in recent years have addressed some ofthe program’s flaws and introduced a market-basedelement, allowing manufacturers to trade credits formeeting the standard. However, the resulting CAFEStandard rules are complex. While incorporating theexisting program in our policies, we have alsostudied a simpler, more flexible market mechanismthat could produce the same result.

1.2.1.3 Energy efficiency standards for appliances and buildings

To reduce the waste of energy is among the mostobvious and familiar ways of reducing fossil fueluse, and one that is within the capability of everyhomeowner. Since 1987, the federal government

10

8http://www.epa.gov/airmarkets/progress/ARPCAIR11_downloads/emissions.xls9http://www.gpo.gov/fdsys/pkg/GAOREPORTS-T-RCED-97-183/pdf/GAOREPORTS-T-RCED-97-183.pdf

10Annual Energy Outlook 2012, DOE/EIA-0383(2012)June (p 29.)

Fig. 1.13 Comparison of CAFE Standardsand Compliance

1975 - 2011

Data source: EIA

Data source: EIA

has set efficiency standards for 55 products, andenergy-saving efforts have accelerated in recentyears. Building codes can also save energy byinfluencing design and materials, but they are notas easily managed at the national level, since codesare locally enacted by state and local governments.The International Code Council promulgates modelcodes that are widely adopted with local revisions,and the U.S. Department of Energy seeks to guidethis process by publishing analyses andrecommendations.

1.2.1.4 Federal Air Quality Act and EPA requirements

The Air Quality Act, passed in 1967, began thefederal government’s regulation of air pollution,followed by the Clean Air Act, passed in 1970, andcreation of the EPA the same year. Includingamendments that significantly strengthened it in1990, the Clean Air Act gives the EPA broad powerto create programs managing sources of airpollution. In 2007, the U.S. Supreme Court ruledthat the act also obliged the EPA to regulategreenhouse gas emissions unless the EPAdetermined the gases clearly do not contribute toclimate change.11 Political and legal challenges tothat additional authority have failed and EPA hasindeed issued rules to limit carbon dioxideemissions from vehicles and stationary sources.

The Clean Air Act stands as one of the nation’sgreatest environmental achievements. The EPAestimates that by 2010 more than 160,000 deathshad been averted by the 1990 amendments to theact, as well as 130,000 heart attacks, 1.7 millionasthma exacerbations, and 13 million lost workdays. It calculated that by 2020 benefits of $2trillion would come at a cost of just $65 billion.12

But the linkage of these efforts to national energypolicy remains haphazard, like the EPA’s court-mandated authority for climate change measures.Although the Supreme Court and EPA classifiedcarbon dioxide emissions as harmful, thosedecisions came only after a lawsuit brought by 12

states, not through a rational policy-making process.Consequently, the legal framework is a misfit for theproblem. A rational policy to address climatechange would look at the sources, uses, andconservation of energy, not only the contents of thetailpipe or smokestack.

1.2.1.5 State renewable energy portfolio standards

States have responded to the vacuum in nationalenergy policy with their own initiatives (Figure1.14). The mix of standards, incentives, rebates andother energy policies pursued at all levels ofgovernment and by utilities themselves cannot besummarized; they fill a database.13 These policiesrepresent progress on energy conservation andcarbon mitigation, and for that we can thank theU.S. system’s exceptional ability to respond toproblems regionally and at various governmentallevels. Indeed, some components of an energypolicy can only be accomplished with state andlocal involvement, such as improving local buildingcodes for energy efficiency. But the most importanttools for accomplishing our major goals—ofreducing oil use and greenhouse gas emissions andair pollution—lie beyond the reach of the states.

1.2.1.6 Regional carbon reduction schemes

In the absence of federal action on climatechange, a group of 10 northeast and mid-Atlanticstates created the Regional Greenhouse GasInitiative, which seeks to reduce carbon dioxideemissions from power generation by 10% by 2018.The cap-and-trade program began in 2009 and hasheld a number of auctions for allowances, withstates spending the proceeds on projects benefitingthe public and carbon reduction goals.14 Theinitiative raised less money than expected becausea surge in utilities’ use of natural gas caused carbondioxide emissions to abate on their own (the statesare considering a lower carbon cap to respond).15

Calling the program a failure, New Jersey

11

11NYT April 3, 2007, Linda Greenhouse.http://www.nytimes.com/2007/04/03/washington/03scotus.html?ex=1333339200&en=e0d0a1497263d879&ei=5124&partner=permalink&exprod=permalink12EPA CCA Second Prospective Study 1990-2020.http://www.epa.gov/air/sect812/prospective2.html

35Such a database can be found at http://dsireusa.org/14http://www.rggi.org/15http://www.ctmirror.org/story/18431/overhaul-ner-regional-greenhouse-gas-initiative

Governor Chris Christie withdrew his state in2011.16 The remaining nine states continue toparticipate, but New Jersey’s action highlights thevulnerability of state coalitions when trying to dothe work of the federal government.

California faces similar and additionalchallenges as it begins a more ambitious carboncap-and-trade program in 2013, which is intendedto cover many sources in addition to electricgeneration. Other western states that planned toparticipate did not follow through, and Californiawill go it alone. By itself, the state loses theadvantages of a larger market for carbonallowances, and is vulnerable to industry leavingthe state to avoid the new costs. Utilities using thepower transmission grid may simply send lower-carbon power to California and keep their overallgeneration profile unchanged. Finally, California’sprogram has been challenged on the grounds thatit interferes with interstate commerce, in violation ofthe U.S. Constitution.17

Only the federal government can overcome thedifficulties facing states addressing these issues ontheir own.

1.2.1.7 Federal ethanol and renewable energy policy

U.S. policy on biofuels and other renewableenergy has suffered from decision makinginfluenced by parochial political factors and with ashort-term outlook. America’s main biofuel is corn-based ethanol, which does reduce the need toimport oil, but does so at a high cost and withoutreduction of carbon emissions, since the carbonfootprint of corn production offsets the benefits oflowered fossil fuel use. In 2005, Pimentel andPatzek found that corn ethanol costs 29% more infossil fuel energy than is gained from the resultingethanol.18 In addition, the diversion of agriculturalresources from producing human food to makingfuel is unsustainable and can exacerbate povertyby increasing world food prices. As Oliveira finds,“The direct and indirect environmental impacts ofgrowing, harvesting, and converting biomass toethanol far exceed any value in developing thisalternative energy resource on a large scale.”19

While Congress has tailored biofuel policy to fitthe needs of corn producers, it has encouragedwind and solar energy in only a short-term, piece-meal fashion, with mixed results. Rapid growth ofthe renewable energy industry encouraged by

12

29 states,+Washington DCand 2 territories, have Renewable Portfolio Standards (8 states and 2territories have renewableportfolio goals).

Fig. 1.14 Renewable Portfolio Standard PoliciesFebruary 2013

Data source: DSIRETM

16http://www.nj.com/politics/index.ssf/2011/05/gov_christie_to_announce_nj_pu.html17http://www.washingtonpost.com/opinions/californias-climate-change-experiment/2012/12/30/942d4c40-4fb5-11e2-839d-d54cc6e49b63_story.html

18Pimental and Patzek, Natural Resources Research, March2005, vol. 14, issue 1, p 6519Oliveira, et al, Bioscience, July 2005, vol. 55, Issue 7, p 593

federal policy over the last decade led toretrenchment when incentives ran out andcompetition stiffened from subsidized overseasproducers and low natural gas prices. Also, solarand wind installations produce power only whenthe sun shines or the wind blows, not necessarilywhen customers want electricity. Withouttechnology to economically store this intermittentpower production, the usefulness of solar and windremain fundamentally limited. Like biofuels, theseenergy sources have the potential to contribute toour national energy goals, but achieving thatpotential requires a consistent, comprehensivepolicy to make them part of a predictable, rationalmarket for clean energy.

1.2.2 Results of current policy

America’s energy policies have beenuncoordinated, inefficient and sometimes irrational,but, as we have seen, they have not been entirelyineffective. The Clean Air Act saved thousands of livesand CAFE Standards slowed the growth of oil use, toname but two examples. But these improvements didnot fully internalize all of the excluded social costs.

From 1990 to 2011, the world’s carbon dioxideemissions increased more than 50%. (See Figure1.5 on p. 4) During that period, the atmosphericconcentration of carbon dioxide rose from 354parts per million to 392 (Figure 1.15).20 The United

States’ increase in carbon dioxide emissions wasmore modest, and some European countriesmanaged significant reductions in emissions aftersigning the Kyoto Protocol in 1997. But much of thisimprovement came about when industrialproduction moved to east Asia, especially China,which in 2007 surpassed the U.S. as the world’slargest greenhouse gas emitter. For example, the U.K.reduced its carbon emissions by 15% from 1990 to2005, but its carbon footprint—the carbon emitted tomake the products its citizens’ use—rose by 19%when imported goods are taken into account.21

The bottom line for climate change is not good.Carbon intensity of energy production worldwidehas increased due to the low price of coal. Scientistsnow agree the international goal of limitingwarming to 2 degrees C is out of reach. Leaders invarious nations have resigned to inevitable climatewarming and are emphasizing adaptation ratherthan avoidance of change.22 Two decades havebeen lost without significant progress.

America has made progress on reducing oilimports in the last few years, but, as discussedearlier, credit for this improvement belongs toimproved petroleum industry technology, highenergy prices that reduce energy consumption, anda prolonged economic slump rather than anyintentional policy moves by our government.Projections based on a business-as-usual approachto the energy market show that world oil prices willtrend higher over the next few decades. Perhapsmore worrisome, much of the world’s oil supply willremain in politically volatile regions where supplyinterruptions can disrupt the flow of petroleum.

At least two lessons may be drawn from thesemixed results. On one hand, we have the lesson thatpolicy does work to positively influence our energyuse and health when we take action as a nation.On the other hand, we have the lesson that to workwell, policy must take into account the economicdrivers of human behavior and be integrated intoour daily decision-making, allowing technology andthe marketplace to operate within the context of ourgoals, rather than be controlled by them.Command-and-control policies, often beset byunintended consequences, produce results farinferior to flexible, market-based solutions like thesulfur dioxide market in the United States.

13

Fig. 1.15 Atmospheric CO2 as Measured at Mauna Loa Observatory

1960-2012(parts per million)

20NOAA Earth Systems Research Laboratoryhttp://www.esrl.noaa.gov/gmd/ccgg/trends/

21This and other numbers in this subsection: Nature specialsection “After Kyoto,” Vol 491 p 653-667 (11/29/12)22ibid

Finally, the evidence of these decades shows thatour actions really do matter for the health andprosperity of our future, and for the integrity ofnatural systems on Earth. In 2011, the United Statesconsumed the energy equivalent of 2.3 billionmetric tonnes of oil, far more than double any othernation except China, which consumed just 15%more than the U.S. (much of it to produce productsfor our use). (Figure 1.16) Yet we have only 5% ofthe world population. U.S. per capita energy useremains four times that of China (Figure 1.17).23

Our actions, even in the absence of cooperationfrom other nations, will make an importantdifference, for bad or for good.

1.3 The purpose of this report

This report presents practical solutions to meetthe goals of reducing use of imported oil and ofreducing carbon dioxide emissions and airpollution. Simply stated, we can achieve these goalsby burning less fossil fuel for transportation andelectrical generation and by using a cleaner mix ofthe fossil fuels we do burn. Our analysis in thisreport and the related 2010 report points to thebest policies to achieve these ends, taking intoaccount technology, economics, legal mechanismsand, to the extent needed for real-world impact, thepolitical viability of the solutions we have studied.

1.3.1 Pricing policy: benefits and lack of political viability

The most efficient, simplest and least expensivetool to achieve our energy policy goals would be atax on carbon emissions. Such a tax, designed toincrease the price of high-carbon fuels, wouldreconnect the social cost of energy paid byeveryone to the users who benefit from its use.Higher prices would encourage conservation andsubstitution of cleaner fuels, lowering carbonemissions, reducing air pollution, and cutting theneed to import oil. The allocation of funds raised bythe tax would be less important than the incentive itwould create for energy users. When the price of

14

23World Bankhttp://data.worldbank.org/indicator/EG.USE.PCAP.KG.OE

0

2

4

6

8

10

12

14

16

18

World Per Capita Primary Energy Consumption Selected Countries 2011 (tons of oil equivalent)

Data sources: BP Statistical Review of Energy

Figure 1.17

Fig. 1.17 World Per Capita Primary Energy Consumption - Selected Countries

2011(tons of oil equivalent)

Fig. 1.16 World Primary Energy Consumption by Country - 12.3 million tons of oil equivalent2011

Data source: BP StatisticalReview of Energy 2012

Saudi Arabia2%

Russian Federation6%

United States18%

All others21%

China21%

Iran2%

Australia1%

Italy1%

South Africa1%

Ukraine1%

Spain1%

Japan4%

Germany2%

Indonesia1%

South Korea2%

France2%

India5%

Brazil2%

Mexico1%

United Kingdom2%

Canada3%

fuel reflects its true cost to society—including thecost for national security, economic security, climatechange and health—those externalities areaddressed by the change in user behavior,regardless of how the tax money is spent.

A cap-and-trade system for carbon reduction is,in effect, a type of carbon tax. The need for fossilfuel users to obtain carbon emission credits orallocations imposes increased cost on high-carbonfuels, discouraging their use, and thereby mitigatingthe externalized price previously borne by societyas a whole. Making allocations tradable allowsthose who can reduce their emissions relativelyeasily to profit by selling unneeded allocations tomore intensive emitters. Trading should addefficiency to the system by allowing the market tochoose the lowest-cost solutions.

Our research in the 2010 report confirmed that apricing policy such as a carbon tax or cap-and-tradeprogram would be the most cost-effective way to meetour goals. Unfortunately, the political climate in theUnited States does not support such a tax. TheWaxman-Markey American Clean Energy andSecurity Act of 2009, which would have establisheda cap-and-trade system, died in the U.S. Senate in2010 without coming up for a vote. With gasolineprices already a potent political issue, Congressappears unlikely to pass legislation that wouldincrease the cost of fossil fuels for consumers.

1.3.2 Finding alternatives to pricing policy

The task of designing a national energy policywithout a carbon tax begins with analyzingtechnology. Do we have technology available tomeet our goals of reducing use of imported oil andreducing greenhouse gas emissions and airpollution? To be included for consideration, anytechnology we screen must meet several criteria. Itmust be capable of conserving energy or deliveringcleaner energy. It must be affordable and ready touse. And it must be scalable, and thus able to makea significant, national contribution to the problem.

Our previous report established that thetechnological tools are at hand to meet nationalenergy goals. However, since the nation’s energyproblems have not solved themselves, the existenceof technology alone is not enough. We also needpolicies to mobilize technology for the benefit of

society. While a carbon tax might be the mostefficient way to do so, policies targeted at workabletechnological solutions can have much the sameimpact. The key unknown would be whether theycan accomplish these goals at little additional cost.This objective is most likely to be reached when theusers of energy pay the full costs of their energy use.

To find an optimal, comprehensive policy, weanalyzed suites of selected policies by cost andscale to determine their ability to meet our goals. Inpractice, more than one potential mix of policyoptions could be combined for a successful recipe.We completed this analysis in the 2010 report,offering a set of policies that, in combination,are nearly as effective as a broad-based carbontax. In this report, we highlight our preferred mixof five options, refine and deepen the analysis,and update our prescription for the changingenergy landscape.

1.4 Previous work and key updates

NEPI and RFF gathered top energy experts tostudy 35 different policies and their ability to reduceoil consumption and carbon dioxide emissions,leading to our November 2010 report, “Toward aNew National Energy Policy: Assessing theOptions.”24 The work covered in that report laid thefoundation of this document, with targets, metrics,modeling, cost analysis and identification of the fivemost promising policies.

Key changes are in targets for the changes wehope to achieve through policy. Our target forreduction in oil use has been adjusted to addresschanges in the oil market, the extension of our studyoutlook from 2030 to 2035, and more refinedthinking in our work about the cost of oilvulnerability. The emission reduction target takesinto account recent changes and an extendedoutlook, and also recognizes that, with the failureof international efforts, working toward a particularstabilization number for atmospheric CO2 isprobably not meaningful. In addition, we haveadded new metrics in this report and we have madecertain modeling adjustments.

The general approach to the problem remainsthe same. The details of our methods are coveredin the following section.

15

24http://www.nepinstitute.org/wp-content/uploads/2011/01/RFF-NEPI-Full-Report-Final-Nov-2010.pdf

1.5 Targets

Targets are desired outcomes expressednumerically. They allow comparison of policyoptions. Targets must be low enough to berealistically attainable but high enough tosignificantly affect the energy problems we want tosolve. They should be aggressive but fact-based. Thetargets selected here meet these criteria. However,their purpose here is exemplary, not prescriptive.

1.5.1 Oil use reduction target

A surge in domestic oil production usingenhanced technology, combined with continuingweakness in demand due to the sluggish economy,have facilitated a reduction in oil imports. The

principal concern about oil markets, however, is thetrend in the nation’s vulnerability to surprisegeopolitical events overseas. The work of Brownand Kennelly, summarized in Section 1.1.1.2 andcovered in more detail in the sidebar “The Costs ofU.S. Dependence on Oil Imports” below, quantifiesthe cost of oil vulnerability on America’s economyand national security. Those costs include theexpected loss in economic output attributable to oilprice shocks to our economy. It is much more difficultto incorporate the national security expenses ofdefending our oil supplies because our military forcesare present in the Middle East for reasons other thanoil. We have excluded costs that are much harder toquantify, such as the restriction that oil dependenceplaces on the prerogatives of U.S. foreign policy. This

16

The Costs of U.S. Dependence on Oil ImportsLeaders have recognized the need to reduce U.S. dependency on imported

oil for decades, but just how much of a reduction is needed remains controversial.To inform our target for reducing oil consumption—which is our primary means ofreducing imports—we commissioned an analysis of the social cost of U.S.dependence on imported oil by Stephen P. A. Brown, director of the Center forBusiness and Economic Research at the University of Nevada, Las Vegas, and Ryan T.Kennelly, an economic analyst at the center. Their paper is posted at the NEPI website http://nepinstitute.org/wp-content/uploads/2013/04/Brown-Costs-of-Oil-Dependence-Apr-20131.pdf.

The market price of oil paid by its consumers does not reflect the full cost ofoil use that is borne by society, and its price also can be distorted by OPEC’sexercise of monopoly power. These market failures justify the use of public policyto correctly assign the costs of oil use. One type of market failure occurs when acost is not included in the price, such as oil’s environmental impact. This kind ofcost is known as an externality, because it is external to the monetary transactionbetween oil buyer and seller. Another type of market failure occurs when thestructure of the market affects prices, such as in the case of OPEC’s monopolizationof the oil market. To address market failures through policy requires economicanalysis because market prices themselves do not inform us about the cost ofexternalities or monopolies.

Brown and Kennelly begin by reviewing various theories proposed by a widerange of scholars about oil market failures, and determine which should be includedin our calculation of the cost of U.S. dependence on imported oil. For example,they do not consider the cost of environmental damage from oil, because it affectsdomestic as well as international production. They also exclude the cost of U.S.foreign policy being inhibited by oil supply concerns, because such concerns cannotbe precisely defined or quantified, and many of these concerns derive from fearsof the economics losses associated with supply disruptions. Brown and Kennellyprefer to measure the economic costs of the disruptions instead.

In generating its broadest measure of the costs of U.S. dependence onimported oil, the paper evaluates five costs to find a total oil dependence premium.Brown and Kennelly show the size of each cost in a chart. Here we summarizethe meaning of several key items on the chart.

• The monopsony premium. The large role of U.S. oil purchases in theinternational market confers power to hold down prices that is analogous to thepower of monopoly sellers to support prices. Because U.S. oil buyers do not actin concert, however, this purchasing power is not exercised. We pay more for oilas a consequence, an increment called the monopsony premium. One way toexercise U.S. purchasing power would be to set a quota that artificially limits oilimports; although that would be illegal under international agreements, domesticpolicies that reduce total oil consumption can have the same impact when theylessen imports. A reduction of U.S. imports would depress the world price of oil.Multiplying that price reduction over all the oil used in the U.S., domestic orimported, yields large dollar-value savings.

• Price shock. International supply disruptions can cause large, sudden increasesin oil prices that drain wealth from Americans’ pockets and send it to oil producers.

• Change in GDP due to price shocks. A rapid increase in oil prices usually causesa drop in economic activity. Scholars have suggested various possible mechanismsfor this out-sized effect, but do not question its importance.

• Defense spending. The question of how much U.S. defense spending relatesto protecting oil importation is difficult and debatable. Brown and Kennelly tackleit by looking at defense spending increases following oil shocks, then price theseexpenditures on a per-barrel basis.

Brown and Kennelly combine these five elements to reach a figure of$27.96 per barrel as the cost of dependence on imported oil. They also finda comparable figure of $2.65 per barrel as the cost of dependence on domesticoil. Combining these estimates yields a figure of $25.31 per barrel fordisplacing oil imports with domestic oil production. Brown and Kennellytranslate the latter estimate into a target for reduction of oil imports that wecould use to measure policies in our study. To do so, they calculate the effecton imports if the premiums they calculate were to be imposed as taxes onimported oil. The consequent reduction in imports caused by this hypotheticaltax becomes our target for reducing oil consumption.

perspective has allowed us to take a clear measureof the value of reducing oil imports.

Converting the dollar cost of the oil dependencepremium into an import quota, Brown and Kennellyfind that a reduction of 1.9 mmbd from its level inthe reference case by 2035 would remove the U.S.exposure to oil vulnerability cost effectively.Subtracting that figure from the EIA’s ReferenceCase prediction of oil use in 2035, we find thatU.S. oil use in that year should be no more than15.2 mmbd, compared to actual oil use of 18.2mmbd in our base year of 2010 (Figure 1.18).

1.5.2 Carbon dioxide emission target

U.S. carbon dioxide emissions from fossil fueluse have slightly declined during the slow economicrecovery and recent period of low natural gasprices. Our Reference Case takes this change intoaccount. We have targeted a cumulative carbondioxide reduction of 21.7 gigatons by 2035compared to our Reference Case (Figure 1.19). Thetarget is consistent with reductions called for by the

IPCC 2007 report, the Waxman-Markey CleanEnergy Bill, and the goals called for by PresidentObama--all of which would reduce emissions in2050 by 83% compared to 2005.

The passage of time has helped illustrate that eventhis aggressive carbon dioxide target will not resolveclimate change. The adoption of the Waxman-Markey bill, which did not occur, would have beenconsistent with stabilization of atmospheric carbondioxide at 450 parts per million, a level estimated byIPCC’s 2007 report as necessary to limit climatewarming to 2 to 2.4 degrees C. But that stabilizationwould have occurred only with immediateparticipation of all emitters, including China andIndia. In the intervening years, internationalnegotiations have foundered and Asian emissionshave accelerated (see Sections 1.1.1.1 and 1.2.2).In these uncertain conditions, we cannot make a firmprediction of the atmospheric CO2 level. We canexpect, if our target is achieved, a smaller increasein the amount of warming. Achieving thesereductions may also make it easier to convince othercountries that they should also take action.

We have chosen a target of a cumulativereduction of 21.7 gigatons by 2035. This bringsthe cumulative emissions to a level of 123.3 gigatonswhich is a 15% reduction from the reference caselevel of 145 gigatons. Serving primarily as alandmark in a shifting landscape, this target is anambitious objective for achieving importantimprovement at an affordable cost.

17

13.5

14.0

14.5

15.0

15.5

16.0

16.5

17.0

17.5

18.0

18.5

2010 2015 2020 2025 2030 2035

Millio

n b

arr

els

/d

ay

Oil Use Reference Case, Target


Reference Target

Figure 1.18

Fig. 1.18 Oil UseReference Case and Target


0

25

50

75

100

125

150

2010 2015 2020 2025 2030 2035

Gig

ato

ns

of

CO

2-e

Cumulative Total Energy-related CO2-e Emissions Reference Case, Target

2010-2035 (gigatons of CO2-equivalents)

Reference Target

Figure 1.19

Fig. 1.19 Cumulative Total Energy-related CO2-e Emissions

Reference Case and Target2010 - 2035

(gigatons of CO2-equivalents)

Fig. 1.20a Sources of SO2 Emissions2008

Data source: EPA

Coal - electricgeneration

73%

Coal - other7%

Fuel combustion- other

2%Fuel combustion

- oil5%

Industrialprocesses

9%

Other2%

Mobile fuel use2%

1.5.3 Air pollution reduction targets

We have not established a numeric target forreduction of air pollution, such as sulfur dioxide,nitrogen oxide or methylmercury. The EPA’s existingprograms have substantially reduced sulfur dioxide,with improvements expected to level off in 2015,and have made significant impact on nitrogenoxide. New rules will go into effect in 2015 toreduce methylmercury emissions. These pollutantscome from fossil fuels, primarily coal. (Figures1.20a, 1.20b and 1.20c) The policies werecommend for reducing coal and oil consumptionwill yield health benefits by further cutting theseother pollutants. Due to the complexity of the taskwe have not established separate targets for thesepollutants, but we do report the impact of ourpolicies on their emissions.

1.6 Metrics

Metrics are standards of measurement forevaluating actions and policies. They are essentialto this project. Besides choosing a yardstick, wemust also select which dimensions to consider whenmeasuring a problem, a solution or a benefit,including whose interests are important to considerand which impacts are worth weighing. Since weseek a national energy policy that includes socialand environmental goals as well as material goods,

we need multi-dimensional metrics.For the two targets cited earlier—for reduced oil

consumption and carbon dioxide emissions—success can be measured simply, in barrels of oilsaved and tons of carbon dioxide emissionavoided. But the primary effort of our work isfinding the least costly way to make these changes.We look for bang-for-the-buck in policy options,justifying the benefits of our goals holistically, in theterms stated earlier in the discussion of goals andtargets. In measuring the cost effectiveness of apolicy we divide the costs of the policy by thephysical measure of effectiveness—the cost perbarrel of oil or ton of CO2 reduced. This measure isincluded for each policy.

We have not performed strict cost-benefitanalysis, which would monetize the positive effectsof each policy. Putting dollars to the benefits ofreducing oil use and carbon dioxide emissions isdifficult and subjective, and we believe the benefitof our overall targets is clearly established.

Our focus for policy comparisons has been on acommon metric of costs. We have evaluated whichcosts to include and what kind of impacts to count.In the free market of classical economics, each actordetermines value on his or her own, based onprivate wants and needs, and fair prices are setthrough exchanges among willing buyers andsellers. As we have seen, however, fossil fuel use 18

Fig. 1.20b Sources of NOx Emissions2008

Data source: EPA

Fig. 1.20c Sources of Mercury Emissions2008

Data source: EPA


48%

Other9%

Industrialprocesses

31%

Gasoline LDVs19%

Fuel combustion -other12%

Vegetation & soil6%

Diesel HDVs18%

Other2%


16%

Other mobilefuel uses

21%

Industrialprocesses

6%

Fuel combustion

other12%

imposes costs on others who are not party to theirmonetary purchase. The good of the nation and theplanet depend on managing such externalities,returning that burden to the energy user responsible.

1.6.1 Welfare costs

How shall we measure the cost of theexternalities? In our 2010 report, we focusedprimarily on welfare cost. A simple definition ofwelfare cost is the sum of all costs to the economyof a chosen action. A costly government actiondiverts economic resources that would otherwise beused to produce and consume goods and servicesof value for citizens. Welfare cost calculations takeinto account these lost opportunities. Althoughwelfare costs are more complicated to calculatethan other impacts, they provide a valuable metricif one wishes to focus on the size of the pie ratherthan on how it is divided among the various groups.

For example, if a policy would produceelectricity with cleaner but more expensive fuels, thetotal private cost of that action would be part of thewelfare cost, regardless of who would pay orbenefit from that action. A policy that discouragesdriving would impose a private cost on a family thatthereby drives less than it would like; the sum of thatcost across everyone’s altered driving preferenceswould be part of the welfare cost. As explainedbelow, these private costs may be partially offset byreduced damages from pollution and other socialcosts that are not priced in the market system.

Welfare cost can be more effective than GrossDomestic Product at measuring some of the costsand benefits that concern us. GDP is the monetaryvalue of all the goods and services produced in thecountry annually. Negative events such as anincrease in chronic disease can boost GDP, andenvironmental damage caused by economic activityis not considered in its calculation unless it affectsother economic activity. Welfare cost, on the otherhand, takes into account a more meaningful spanof costs. For example, if air pollution inducedasthma in a child, welfare cost could account tosome degree for that affliction as a cost, while GDPmight count it positively for the increased economicactivity in the health care industry.

While welfare cost is a powerful metric in concept,it also has serious weaknesses in practice. Importantcosts that are taken into account, including the

examples above, can be difficult to quantify.Calculations become complex and often requireassumptions about costs that cannot be directlymeasured. Different economists use different methodsthat can lead to different results. These issues cancombine to produce substantial uncertainty in theoutcome of welfare cost calculations, or results thatvary greatly among researchers.

Welfare cost, while subject to wide variations ofassigned costs, is an important metric. Federalagencies, including Congressional offices, includewelfare costs in mandatory cost-benefit analyses ofplanned regulations and proposed legislation toconsider society’s point of view in aggregate.

1.6.1.1 Welfare costs and market efficiency

A major puzzle confounds cost estimates foreconomists studying energy policy. Economic theoryindicates that consumers and businesses, givenadequate information, should invest as readily inenergy efficiency as in any other investment thatyields the same financial return. But they do not.Research shows that businesses and consumersundervalue energy efficiency improvements,demanding quick pay-back for long-lasting benefits.Typical decision-makers require a return oninvestment much higher for energy improvementsthan for other investments—as high as 40 percent.

Estimating welfare costs is complex enough if wecan assume that everyone behaves as would beexpected in an efficient market, but an additionalcomplication enters the picture if actors aremisinformed, irrational, or make bad decisions forany other reason, a condition called market failure.For example, suppose a building owner would forgothe financial benefit of installing energy-savingequipment, but is forced to accept that benefit by agovernment policy. Such a policy could produce anegative welfare cost—more than a free lunch.

But before we can be sure of such a policybenefit, we need to know for certain that theseinvestment decisions really are market failures. Ifpurchasers know of hidden costs that offset thebenefits of energy improvements, then they mayhave good reason for avoiding these investments.In that case, with an efficient market, or no marketfailure, the welfare cost of a policy would not beoffset, and there would be no free lunch.

19

Some researchers use an adjusted discount rateto simulate the partial market failure they see inenergy improvement investment decisions. Discountrates relate present dollars to future dollars,reflecting the fact that money we have now is worthmore than money we will get later. In this study,adjusting the discount rate downward to reflectpartial market failure would yield dramaticallylower welfare costs, producing essentially a freelunch for our entire comprehensive policy, beforeeven considering the benefits of reduced oil importsand carbon dioxide emissions.

We have not chosen that approach. While thelikelihood is high of at least some market failure inthe undervaluing of energy investments, our studywas not designed to quantify that factor or to lookfor hidden costs. Without such a factual basis,choosing a discount rate to simulate market failurewould be arbitrary. Instead, our welfare costsrepresent an efficient market, with no market failure.As such, they represent a highly conservative anddefensible policy cost, which can be thought of asa maximum. Any consideration of market failurewould make the welfare cost lower.

1.6.2 Additional measures of policy impacts

The uncertainties in welfare cost calculations arecompounded by the concept’s significant limitationsfor policy makers in a political environment. Welfarecost looks only at the total, so it does not take intoaccount whether one group within society gains orloses from an action or how it might create transfersfrom one part of society to another. Incometransferred between groups by a tax or governmentsubsidy would not affect welfare cost, but may havegreat significance for the affected groups. Likewise,a policy that leads to job losses in a particularindustry might not represent large welfare costs forthe total economy if the same policy were to expandjobs in another sector. Yet real-world decision-making, especially for democratically electedleaders, often ignores what is best for society andinstead focuses upon what is best for particulargroups. Complexity also undercuts the measure’susefulness in non-technical discussions. Variables canbe difficult to quantify and results difficult to explain.

For these reasons, we have expanded beyondlooking only at welfare costs. In this report we have

updated welfare costs for the package of fivepreferred policies we feature, which were selectedfrom 35 policies we screened for the 2010 report.But we have also added additional impact metricsto understand the varied consequences of thepolicies for specific groups within society. Thesenew metrics simply reflect impacts that many peoplewould want to know before adopting a policy.

Our new metrics cover the impacts of policyoptions for consumers and government. Forconsumers, we look at out-of-pocket expenses,primarily for fuel and taxes. For government, weconsider the impact of policies on the federal budget.

We studied policy impacts on GDP, but ultimatelyfound they were too small to be meaningfully quantifiedby the model that was the core of our project.

1.6.3 Allocating costs to combined goals

Our project has two goals, to reduce oil use andimports and to reduce carbon dioxide emissions. Threeof our five promising policies contribute to both goals,and the combination of all policies described in Part 3is designed to mix these benefits in a complementaryway. This blending of policies and goals is a strengthof the comprehensive policy. In determining the costeffectiveness of polices on a per unit basis—per barrelof oil or per ton of carbon dioxide reduced—achallenge arises as to determining the allocation ofcosts between the two goals. Economists refer to thisas the joint allocation of cost problem.

In Part 2, we present individual policies. Forpolicies that contribute to both our goals, we assignthe total welfare cost in full to each, the barrels of oilreduced and the tons of carbon dioxide emissionsreduced. While this, in effect, double-counts the cost,showing oil barrels and carbon dioxide tons as moreexpensive to reduce, it avoids an arbitrary andunhelpful allocation of costs among the goals. For thecombination of policies reported in Part 3, we simplypresent the present discounted value of the welfarecost (rather than cost-effectiveness) with other policymeasurements, as there is no rational basis forallocating costs per policy.

1.7 Modeling

This project brings value to the energy policydebate in part through the rigor of our process ofevaluating many alternatives using the sameeconomy-wide computer model. An energy-

20

economic model is a computer program usingmathematical formulae to represent the relationshipsof the many parts of the energy economy. Modelerscan change inputs representing economic conditionsor change assumptions about economic behavior tosee how those changes cascade through the systemand ultimately impact oil use, carbon emissions, andother energy-related outputs.

Since our work has sought to analyze manypolicies covering a wide variety of fuels,technologies and sectors, we chose to use animportant national model with sufficient detail andthe ability to disaggregate these factors, theNational Energy Modeling System (NEMS),maintained by the Energy Information Agency(Figure 1.21). This is a detailed energy-system modelthat incorporates facets of numerous technologieswhile balancing supply and demand throughout theeconomy’s energy system.25 Assumptions inform themodel about how prices and availability of productsinfluence human behavior—for example,concerning the fuel prices at which consumers switchto more efficient vehicles or hybrids—as well asfactors such as future fossil fuel discoveries orimproved oil recovery technologies. Like otherresearchers in the field, we have started with theEIA’s data and its Reference Case for projectionsthat incorporate existing laws and regulations.

Model runs conducted for NEPI by OnLocation,Inc. used data and Reference Case assumptionsfound in the EIA’s Annual Energy Outlook 2012(“AEO 2012”). This is the basis for creating aReference Case for our analysis. (A single exceptionis explained in Part 2, Section 2.4.3.) We refer tothis version of the model as NEMS-NEPI. The modeldoes not estimate welfare costs: instead, modeloutput was used to inform calculations of welfarecosts, which were done off-line.

1.8 Refining the five most promising policies

After analyzing 35 policy alternatives, our 2010report gathered the most promising into groups ofcrosscutting policy combinations in search of themost powerful and least expensive politically viablepackage for a national energy policy. The NEMS-RFF model allowed us to test various combinationswith differently interlocking pieces to find a suitewith the strongest complementary features. In thisreport, we focus on the five most promising policies,studying them more deeply, refining their qualitiesfor greatest benefit, and updating the data used toanalyze them to take into account recent changesin the energy outlook, as laid out in the EIA AEO2012 report (see the previous paragraph).

The results of that work are presented in Parts 2and 3 of this report.21

Fig. 1.21 Schematic of National Energy Modeling System

Source: Energy Information Agency

25This model is neither “bottom-up” nor “top-down” but insteadmight be considered as a hybrid framework that combinesthese two approaches.

Part 2: Five Promising Policies: Explanation and impact of the components

Five policies have emerged from our analysis asthe most promising to address, at low cost, thenational energy goals of reducing oil use andimportation and cutting emissions of carbon dioxideand air pollution. In this chapter we explore thefunction and impact of each policy individually. Thepolicies are:• NEPI Clean Energy Standard: Establishes a 2035goal for clean electrical generation and requirestechnology-neutral, market-driven decisions byutilities to use low-carbon and renewable fuels toachieve this target. • Oil security dividend: Imposes a modest,graduated oil products tax that is fully rebatedthrough the tax system to the public.• Automotive fuel economy: Follows the CorporateAverage Fuel Economy (CAFE) Standards asadopted in 2012, and considers an alternative‘feebate’ mechanism with similar impact andpotential advantages.• LNG trucks: Provides a temporary fuel taxreduction to encourage the nascent shift in theheavy truck market from diesel to liquefied naturalgas (LNG).• Energy efficiency: Extends existing incentives forenergy-saving appliances and building codes andhome-based renewable energy.

2.1 The NEPI Clean Energy Standard (NCES)

Americans have a multitude of opportunities toreduce carbon emissions and air pollution fromelectrical generation, but making reductionsrequires decision-making by many actorsaddressing conditions in individual localities. Policymakers have tried setting state-by-state RenewablePortfolio Standards (RPSs), requiring orrecommending a percentage of electricity in eachstate to come from renewable sources by a certaindate, but the national impact of these efforts islimited by regional jurisdictions and the problem ofscale. Current renewable energy technology simplycannot produce enough reliable power to replacefossil fuels. The NEPI Clean Energy Standard

establishes a 2035 clean energy goal that will beaccomplished by a national, market-basedapproach to identify the lowest-cost options for atransition to cleaner fuels, which may or may not berenewable, thereby yielding greater and lessexpensive emission reductions.

A number of CES policies have been proposedto reduce the carbon emissions of the electric powerindustry. Electric generation produces over 40% ofenergy-related greenhouse gas emissions in the U.S.(as seen in Figure 1.2.) However, the particulardesign of a CES policy significantly affects its costand effectiveness. Here we have refined the conceptby modeling various program designs to find thebest option, which we present and compare to othersimilar concepts under discussion. The resultingpolicy achieves 64% of our target for carbondioxide reductions.

After presenting the policy, we describemodeling results that show impacts on fuel sources,carbon dioxide emissions, energy prices, airpollution, and other parameters. Next we considerpolicy costs. In addition, we analyze the electricityrate impact of the program on a sample of statesand regions of the country, which differ dependingon their current fuel mix and their energy resources.Finally, we look at the feasibility of converting torenewable and nuclear sources of energy at thescale the policy would require.

In Figure 2.1, we see our Reference Case totalenergy-related emissions compared to power sectoremissions, and our target total emissions.

22

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2010 2015 2020 2025 2030 2035

Millio

n m

etr

ic t

on

s C

O2-e

Annual CO2-e Emissions Reference Case Total and Power Sector,

and Target Total Emissions 2010-2035

(million metric tons of CO2-equivalents)

Reference Case Total Reference Case Power Sector Target Total

Figure 2.1

Fig. 2.1 CO2-e EmissionsReference Case Total and Power Sector,

and Target Total Emissions2010-2035


2.1.1 How the NCES works

The NEPI Clean Energy Standard sets a specificgoal for clean electrical generation through 2035,to be accomplished with market-driven, least-costdecision making. NEPI and RFF designed asomewhat similar standard as part of our 2010report, in pursuit of the overall IntergovernmentalPanel on Climate Change carbon reduction targetrecommended in 2007. We also aligned ourconcept with the share of carbon reduction thatwould have been accomplished by the WaxmanMarkey Clean Air legislation proposed in Congressin 2009. Finally, the standard would actuate thevision that President Obama articulated in his 2011State of the Union Address when he said, “Ichallenge you to join me in setting a new goal: By2035, 80% of America’s electricity will come fromclean energy sources. Some folks want wind andsolar. Others want nuclear, clean coal and naturalgas. To meet this goal, we will need them all.”

Our analysis confirms that meeting the 80% goalcannot be accomplished at an acceptable costwithout a blended strategy that uses more naturalgas, renewables, and nuclear energy and lessconventional combustion coal. (The meaning of thephrase “clean coal” is ambiguous; coal-burningtechnology cannot currently produce economicallypriced electricity without high carbon and otherpollution emissions, but it may have that capabilityin the future with carbon capture and sequestration.)The path from our present energy mix to thePresident’s clean energy goal must take us throughintermediate technology that is currently availableon the scale necessary before we can reach low-carbon solutions not yet ready in adequateabundance. A policy to navigate this path requiresa market mechanism to mobilize electric utilities tomake a myriad of individual decisions pointing inthe right direction.

The NCES would phase in clean energy,beginning with a 45% standard in 2015 and thenincreasing by increments to reach 80% by 2035.Clean energy in 2010 reached 44% of U.S.generation thanks to low natural gas prices and theincreasing use of wind power, and was estimatedat 47% in 2012. But the standard would driveadditional clean energy immediately because every

electricity retailer, regardless of size, location orownership, would be responsible to meet thestandard or acquire credits to compensate forexcessive carbon emissions.

The NCES sets the cost of moving to cleanenergy through the trading of credits. Thosegeneration utilities that surpass the standard wouldearn credits to sell or to bank for use in future years.Existing renewable and nuclear energy producerswould get full credit for the clean electricity theygenerate, as would owners of distributedgeneration, such as rooftop solar panels or on-sitewind generators. The credits they earn and selleffectively lower the cost of the power they produce;the necessity to buy those credits converselyincreases the cost of power produced with coal.

A conventional coal-fired power plantrepresents the baseline definition of dirty energy.The NCES allocates credits to generatorsdepending on their carbon dioxide emissions incomparison to coal. A carbon-free renewable ornuclear plant receives a full credit per kilowatthour of electricity produced, a conventional coalplant receives no credits, and technologies inbetween receive partial credits: a combined-cyclenatural gas plant, which emits about half thecarbon dioxide of a coal plant, receives half acredit for each kilowatt hour of energy produced.Figure 2.2 shows the relative emissions and creditsearned by various technologies.

As the emission standard phases in through theyears, high emitters would be increasinglydependent on buying credits from low emitters,

23

80%

100%

2.2eurgiF

80%

100%

%sa

CO2 ndansoissimEaniotnevnoCw Nef o%

ionssmisE

dneraEtsiderCndniossmiEtnPlal aoCl a

ions tsCredi

sn

0%

20%

40%

60%

0%

20%

40%

60%

0%

0%

CCGICCGNS:CC

asGderatgetnI:CCCoasGral uatN:CCreutCapnoCarbS:

einmbConioatificasle;cCyeinmbCo

iorattseuqSedanre

le;cCyde

nio

Fig. 2.2 CO2 Emissions and Credits Earnedas % of New Conventional

Coal Plant Emissions

creating high prices for emission credits andstronger incentives for low-emission generation.The NCES offers no alternative compliancepayment system, such as a fine for failing to meetthe standard, which would allow generators toavoid participation in the emission credit market;such a system of fines fails to assure that emissionsare ever reduced.

The system of credits encourages movementtoward cleaner fuel sources without dictating thetechnology selected, thereby allowing individualproducers to choose the least costly path tolowering carbon emissions in their own locale. Likea Renewable Portfolio Standard, which sets apercentage goal for how much electricity must begenerated by renewable sources, the NCESestablishes a goal for the percentage of powerproduction that must be carbon-free. Unlike an RPS,however, the NCES does not pick the technology tomeet the goal, instead giving credit in proportionto an electrical generator’s contribution to meetingthe national standard. Under the NCES, carbonemissions avoided by moving from coal to naturalgas are equal in value to equivalent emissionsavoided by moving from natural gas to wind.

For example, in 2015 the proposed standardcalls for 45% of electric generation to be clean. Themost carbon-intensive energy technology, theconventional pulverized coal plant, is defined as100% dirty. In 2015, it would need credits tocompensate for falling 45 percentage points shortof the standard. A wind installation, zero percentdirty, would receive credits for the 55 percentagepoints by which it surpassed the standard. The windplant’s owners could then sell those credits to thecoal plant’s owners to meet their carbon reductionobligation. A conventional combined-cycle naturalgas plant, which produces half the carbon dioxideof a conventional coal plant per kilowatt hour,would be rated as 50% clean. In 2015, it wouldexceed the 45% standard by 5 percentage points,receiving credits that its owners could sell, or couldbank for later use. In 2025, when the standard risesto 60%, and the gas plant would need credits (as itwould now fall short of the standard) it could usepreviously banked credits to make up the difference,or buy credits on the open market (figure 2.3).

2.1.2 Alternative designs of a CES

Many versions of a clean energy standard arepossible. A recent version was embodied in theClean Energy Standard Act, a bill sponsored inFebruary 2012 by Senator Jeff Bingaman (D-N.M.),then chair of the Senate Energy and NaturalResources Committee. In our study and modeling,we have chosen several features different from theBingaman bill. Some of our changes are notimportant to the overall analysis but make modelingeasier. Others reflect real differences.

Bingaman’s bill exempted small utilities.Depending on the definition of “small,” this kind ofprovision can leave out up to 25% of electricitygeneration. We include all utilities to avoid thatproblem, and to avoid the complexity of selectinga proper size definition and quantifying its impact.We realize that some very small utilities that lackpractical fuel alternatives should and probably willbe excluded, but those adjustments would notsignificantly affect our analysis.

The 2012 bill also excluded existing nuclear andhydropower plants from receiving tradable creditsfor carbon-free generation. Proponents of thisprovision argue that providing credits to existinggenerators does not advance the purpose of theprogram, which is to increase clean energy. Inaddition, the inclusion of existing facilities maydisproportionately benefit regions of the countrywhere nuclear or hydroelectric projects alreadyprovide much of the electricity used. We havestudied that issue and address it in Section 2.1.6.

24

2015 2020 2025 2030 2035

Cre

dit

s is

sued

per

kilo

watt

ho

ur

NCES Technology Credits vs Requirement

Nuclear/Renewables (1.00)

Advanced NGCC (0.60)

NCES Requirement

Coal (0.00)

(0.45)

(0.80)

Figure 2.3

NGCC: Natural Gas Combined Cycle

Fig. 2.3 NCES Technology Credits vs Requirement

We have chosen to include existing generationin the NCES. This decision produces the mostcomprehensive policy, against which otheralternatives can be measured, with the market ableto function in the most pure form. It also wouldcreate a constituency with a vested interest fromyear one: utilities that receive credits would be ableto profit from them only with continuation of thepolicy. Finally, the greater availability of creditsbroadens the market, reducing the probability ofsudden increases in credit prices.

Two other elements of our NCES that differ fromsome other proposals underline the decision tomake this policy a market-driven solution. We donot allow high-carbon generators to avoid themarket for credits by paying a government fine forexcessive emissions. Such a work-aroundopportunity can keep the market price of credits

from going too high, but italso threatens toundermine the entirepurpose of the program,which is to meet thestandard.

The NCES also allowsproducers to bank creditsearned in the early years,which they could use laterwhen the standardbecomes harder to meet.By allowing banking ofcredits, we allow utilitiesgreater discretion to holddown their costs, whichshould tend to reduce thecost of compliance.Moreover, the bankabilityof credits increases theintrinsic value of credits,adding to the incentive tocut emissions in the earlyyears. But the bankabilityfeature also tends tomoderate increases in thecost of credits, since itincreases the supply.

Some proposed CES

systems would provide credits for energy efficiencyimprovements. While it is true that efficiencymeasures can be a cheap way to reduce carbonemissions, usually on a small scale, quantifying thecontribution of any particular project is difficult andcomplicated. An appealing feature of the NCES isthe simplicity of awarding credits. Consequently, wedo not allow energy efficiency projects to earncredits, but we do encourage efficiency throughanother component of our comprehensive nationalenergy policy.

2.1.3 How the NCES differs from an RPS

The majority of U.S. states and territories haveadopted Renewable Portfolio Standards, which arepolicies setting objectives for increasing renewablesources in a state’s mix of power generation. (SeeFigure 1.14) Considering the popularity of thisapproach, we have devoted modeling efforts tocomparing the RPS to the NCES. Modeling showsthe NCES yields superior results at lower cost, aswe show in the next section. Here we focus on thefunctional differences between the two approachesthat produce these differences.

The meaning of the term Renewable PortfolioStandard varies from state to state, depending onhow “renewable” is defined, whether the standardis mandatory or aspirational, how projects arecounted, and whether a system of tradable creditsexists. This diversity makes precise predictionsimpossible as to the impact of these measures, butthe Energy Information Agency (EIA) notes that, tothe extent the policies could be modeled, and if theyall work as hoped, renewables would consequentlyaccount for 10% of power production by 2025,with their share thereafter rising more slowly thantotal energy demand.26

For the purposes of our study, we have modeleda national RPS with ideal features. This approachallows a fair comparison of the two approachesbased on their fundamental differences rather thanon real-world compromises that occur when anypolicy is implemented.

The RPS seeks to reach 25% renewable energyby 2028. Improvements come only from certainrenewable energy sources. Electricity generated

25

26Annual Energy Outlook 2012, DOE/EIA-0383(2012) June (p 11.)

from nuclear energy or natural gas doesn’t count,nor do existing hydroelectric generation ormunicipal solid waste generation facilities. The RPSdoes allow generators to earn credits by surpassingthe renewable standard, and requires them to buycredits if falling short, but the credits cannot bebanked for later use. A utility unable to meet themandate or obtain credits can continue operatingby making an Alternative Compliance Payment,similar to a fine, which we modeled as being set at10 cents per kilowatt hour. This provision limits thecost of the policy to that ceiling, but creates thepotential that the national standard will not be met.

As has been explained elsewhere in this report,a clean energy standard does not dictate thetechnology to be used, such as wind or solar, butinstead encourages progress toward a nationalgoal of 80% clean energy by whatever means andin whatever area of the country improvements canbe accomplished at the least expense to society andenergy users. If switching a Midwest power plantfrom coal to natural gas avoids a ton of atmosphericcarbon more cheaply than building a west coastwindmill, that would be the option we choose forthe benefit of the environment and our nationalwellbeing. The market mechanism in the NCESallows these trade-offs to be valued and exchangedor saved for future use.

2.1.4 NCES benefits and modeling results

The NCES guides U.S. electricity generationtoward lower carbon emissions on whatever pathaccomplishes the goal at the least cost. Modelingshows this path uses natural gas as a bridge tocarbon-free energy (Figure 2.4).

The exchange of clean energy credits creates aflow of funds from dirtier to cleaner producers ofelectricity. Initially, this flow increases the price ofelectricity generated from conventional coal andreduces the price of power from natural gas,renewables and nuclear energy. Natural gas hasthe greatest potential for rapidly increasedgeneration, so it would gain market share in theearlier years of the program. But as the standardincreases in the later years of the program,natural gas burning utilities would begin needingto buy credits, which would encourage them tophase out fossil fuel use in favor of nuclear and

renewable production.The Renewable Portfolio Standard does not

achieve the same level of carbon reductions (Figure2.5.) It puts more renewable generation into service,but also leaves much more coal generation operating.

These scenario results emerge from runs of the

National Energy Modeling System, as modified forour purposes by OnLocation, Inc., which we call theNEMS-NEPI model. The Reference Case representsa no-action alternative, continuing current policiesand taking into account anticipated fuel costs fromthe EIA’s Annual Energy Outlook 2012(AEO2012). Details from the modeling follow.

2.1.4.1 Electricity generation by fuel source

The Reference Case shows electricity beingproduced by much the same mix of fuels in 2035as we currently use, but with coal accounting for38% of generation and non-hydro renewables 9%.

26

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2010 2015 2020 2025 2030 2035

Perc

en

t o

f ele

ctri

city

gen

era

tio

n

Percent of Electricity Generation by Fuel NCES 2010-2035

Figure 2.4

2010: Coal = 45%

2035: Coal = 16%

2010: Natural Gas = 24%

2035: Natural Gas

= 34%

2010: Nuclear = 20% 2035:

Nuclear = 29%

2010: Renewables = 10%

2035: Renewables

= 20%

NOTE: Petroleum + Other = 1%

Fig. 2.4 Percent of Electricity Generation by FuelNCES - 2010 - 2035

0

500

1,000

1,500

2,000

2,500

2010 2015 2020 2025 2030 2035

Millio

n m

etr

ic t

on

s C

O2-e

Power Sector CO2-e Emissions Reference Case, RPS, and NCES

2010-2035 (million metric tons of CO2-equivalents)

Reference Case RPS NCES

Figure 2.5

Fig. 2.5 Power Sector CO2-e EmissionsReference Case, RPS and NCES, 2010 - 2035


(Figure 2.6) All scenarios continue roughly thecurrent share of hydropower through the studyperiod, of around 6%. The RPS reduces coal to 32%while increasing non-hydro renewables to 20%. TheNCES reduces coal to 16%, increasing non-hydrorenewables to 13%.

The NCES achieves this much greater reduction

in coal use by shifting more power generation tonatural gas and nuclear (Figures 2.7a and 2.7b).Including hydropower and nuclear, the NCESachieves more carbon-free electricity generationthan the RPS, despite using less renewable energy.

Since the RPS targets only renewables such aswind, solar, and biomass, those sources alonereceive credits and a competitive boost in themarket. Natural gas receives no credits, and thuslittle market advantage compared to coal;consequently, construction of new natural gas plantsis more limited than in the NCES and Reference

Cases. The market in renewable energy creditsmakes scant impact on the more carbon-intensivecoal because renewables are not an effectivesubstitute for coal in baseload power applications.The reduction of coal that does occur comes mostlythrough substituting biomass in co-fired plants.Figure 2.7a shows the impact of the RPS comparedto the Reference Case.

The NCES issues credits to all generation cleanerthan the standard requires. At the beginning of thepolicy, natural gas generation earns partial creditfor being cleaner than coal, creating a marketadvantage that leads to gas displacing coal fromthe first year of the program. As time passes andthe standard rises toward 80% clean, natural gasno longer earns credits and itself begins to bedisplaced by zero-carbon generation, includingbiomass, nuclear and wind. Renewables alsodisplace additional coal. Figure 2.7b shows theimpact of the NCES compared to the Reference Case.

2.1.4.2 Carbon dioxide emissions

The Reference Case shows modest changes incarbon dioxide emissions through the study periodto 2035 (Figure 2.8). The RPS achieves a reductionin emissions to 2030, but then emissions begin torise again to 2035. This increase in emissionsoccurs because the 25% renewable goal is reachedin 2028 and because the policy includes a pricecap on the cost of adding renewables, in the formof the Alternative Compliance Payment. The NCESachieves much larger improvements. The rapidtransition from coal to natural gas in the early years

27

45% 38% 32%

16%

24% 27%

24%

34%

20% 18%

16% 29%

6% 6%

6% 7%

4% 9% 20%

13%

0%

20%

40%

60%

80%

100%

Actual Reference Case

RPS NCES

Perc

en

t o

f ele

ctri

city

gen

era

tio

n

Percent of Electricity Generation by Source Reference Case, RPS, and NCES 2010 and 2035

Other

Non-hydro Renewables Hydropower

Nuclear

Natural Gas

Petroleum

Coal

2035 2010

Figure 2.6

NOTE: Petroleum and Other are less than 1% each

Fig. 2.6 Percent of Electricity Generation by Source,Reference Case, RPS and NCES

2010 and 2035

-1,250

-1,000

-750

-500

-250

0

250

500

750

1,000

1,250

2010 2015 2020 2025 2030 2035

Billio

n k

ilo

watt

ho

urs

RPS Generation (change from Reference Case) 2010-2035 (billions of kilowatt hours)

Conv Coal Coal CCS Natural Gas CC Natural Gas CCS Other Gas/Oil Nuclear Hydro Biomass Wind Other Renew

Figure 2.7a

CCS: Carbon Capture and Sequestration; CC: Combined Cycle -1,250

-1,000

-750

-500

-250

0

250

500

750

1,000

1,250

2010 2015 2020 2025 2030 2035

Billio

n k

ilo

watt

ho

urs

NCES Generation (change from Reference Case) 2010-2035 (billions of kilowatt hours)

Conv Coal Coal CCS Natural Gas CC Natural Gas CCS Other Gas/Oil Nuclear Hydro Biomass Wind Other Renew

Figure 2.7b

CCS: Carbon Capture and Sequestration; CC: Combined Cycle

Fig. 2.7b NCES Generation(change from Reference Case)

2010 - 2035(billions of kilowatt hours)

Fig. 2.7a RPS Generation(change from Reference Case)


of the policy gives the NCES a large immediateadvantage in reducing carbon dioxide emissions.The advantage continues to broaden through thelater years of the study as the standard continues torise to 80% in 2035.

The NCES produces the best results in cumulativecarbon dioxide reduction by the 2035 target year.Compared to the Reference Case, the NCESreduces energy-related carbon dioxide emissions by13.8 gigatons, which is 64% of the 21.7 gigatontarget that NEPI has set as our overall target for anational energy policy. The RPS produces only a5.1 gigaton reduction, or 24% of the 21.7 gigatontarget. The NCES, on the other hand, continues toincentivize changes in the fuel mix until reachingthe 80% clean energy standard in 2035.

2.1.4.3 Electricity sales

The NEMS-NEPI model shows electricity useincreasing throughout the study period for theReference Case, the RPS and the NCES (Figure 2.9).Differences among the three are not large, but theNCES shows the smallest increase in power sales.The RPS depresses power sales by 2025, but theyrise faster and exceed the Reference Case in 2035.The reason is the same as identified in the previoussection: once the RPS standard is met in 2028, thepolicy produces no further pressure on prices.

2.1.4.4 Electricity prices

None of the scenarios drive substantial increasesin the price of electricity (Figure 2.10). TheReference Case shows a rise from 9.8 cents perkilowatt hour to 10.2 cents in 2035, an increase ofonly 4 percent over 25 years (these numbers are in

real dollars and include all sectors of power use).Under the RPS the price of electricity increases untilpeaking at 10.7 cents in 2028, when the 25%renewable standard is met. Alternative CompliancePayment (ACP) of 10 cents per kilowatt hour helpskeep the price increase modest but limits the amountof carbon dioxide reduced. After achieving thestandard, power prices fall under the RPS to 10.2cents in 2035, the same as the Reference Case.

The price of electricity rises faster under theNCES and continues to rise through the studyperiod. The price rises in 2015, to 10.2 cents perkilowatt hour, when the policy kicks in, andcontinues to rise in response to the steady phase-inof the standard, reaching 11.8 cents in 2035. Thetotal increase over the 25-year period is 21%, or less

than 1% per year, which is hardly enough to benoticeable in an electricity bill, and could easily beoffset in individual households that conserve power.

28

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2010 2015 2020 2025 2030 2035

Millio

n m

etr

ic t

on

s C

O2-e

Annual Total Energy-related CO2-e Emissions Reference Case, RPS, and NCES

2010-2035 (million metric tons of CO2-equivalents)


Figure 2.8

Fig. 2.8 Annual Total CO2-e Energy-related EmissionsReference Case, RPS and NCES

2010 - 2035(million metric tons of CO2-equivalents)

1,451 1,726 1,701 1,658

1,329

1,694 1,673 1,615

962

976 874 905 128

301 477 375

0

1,000

2,000

3,000

4,000

5,000

Actual Reference Case RPS NCES

Billio

n k

ilo

watt

ho

urs

Electricity Sales by Sector Reference Case, RPS, and NCES 2010 and 2035 (billions of kilowatt hours)

Direct Use

Transportation

Industrial

Commercial

Residential

2035 2010

Figure 2.9

Fig. 2.9 Electricity Sales by SectorReference Case, RPS and NCES


0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

2010 2015 2020 2025 2030 2035

20

10

cen

ts/

kilo

watt

ho

ur

Average Electricity Prices Reference Case, RPS, and NCES

2010-2035 (2010 cents per kilowatt hour)

NCES RPS Reference Case

NOTE: RPS includes Alternative Compliance Payment of 10 cents/kWh

Figure 2.10

Fig. 2.10 Average Electricity PricesReference Case, RPS and NCES

2010 - 2035(2010 cents per kilowatt hour)

2.1.4.5 Natural gas prices

The NCES and RPS have opposite impacts on theprice of natural gas, although both scenarios andthe Reference Case end at close to the same pricein 2035 (Figure 2.11). Since the NCES givespartial credit to natural gas for its lower carbonemissions than coal, it encourages a shift from coalto natural gas for electricity generation in the earlyyears of the policy, creating an upward pressure onthe prices, with a spike when the policy first bites,in 2015. Later in the study period, as theincreasingly stringent clean energy standard phasesin, natural gas no longer can earn credits andutilities move to carbon-free energy instead. That

shift moderates gas price increases and brings theprice close to the Reference Case. The RPS gives nocredit to natural gas and drives its use and pricebelow the Reference Case.

2.1.4.6 Emissions of other pollutants

Reducing air pollution is an important goal of anational energy policy. Sulfur dioxide, nitrogenoxides and methylmercury all contribute to illness,premature death, lost work time and health carecosts; all are products of fossil fuel combustion,especially coal. The rapid transition from coal tonatural gas in the NCES attacks these pollutants,and the improvements increase in the later yearswhen natural gas generation is displaced bytechnologies with no air emissions. Part 1 furtherexplains the goal of reducing air pollution.

EPA policies based on the authority of the CleanAir Act have already saved thousands of lives, andnew limits on the books and still taking effect will

save many more. These improvements are includedin the Reference Case modeled by the EIA in AEO2012. Our modeling demonstrates an additionalimprovement for the NCES above what the EPA isexpected to achieve (Figures 2.12a, b, and c).

The EPA rules drive down emissions of sulfurdioxide and methylmercury dramatically in this

290.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

2010 2015 2020 2025 2030 2035

20

10

$/

millio

n B

tu

Natural Gas Spot Prices Reference Case, RPS, and NCES

2010-2035 (2010 $ per million Btu)

NCES RPS Reference Case

Figure 2.11

Fig. 2.11 Natural Gas Spot PricesReference Case, RPS and NCES

2010 - 2035(2010 $ per million Btu)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

2010 2015 2020 2025 2030 2035 M

illio

n t

on

s/year

SO2 Emissions Reference Case, RPS, and NCES

2010-2035 (million tons per year)


Figure 2.12a

Fig. 2.12a SO2 EmissionsReference Case, RPS and NCES

2010 - 2035(million tons per year)

0.0

0.5

1.0

1.5

2.0

2.5

2010 2015 2020 2025 2030 2035

Millio

n t

on

s/year

NOx Emissions Reference Case, RPS, and NCES

2010-2035 (million tons per year)


Figure 2.12b

Fig. 2.12b NOX EmissionsReference Case, RPS and NCES

2010 - 2035(million tons per year)

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

2010 2015 2020 2025 2030 2035

To

ns/

year

Mercury Emissions Reference Case, RPS, and NCES

2010-2035 (tons per year)


Figure 2.12c

Fig. 2.12c Mercury EmissionsReference Case, RPS and NCES

2010 - 2035(tons per year)

decade, with reductions of around 80%. Emissionsof nitrogen oxides are significantly improved aswell, down by about 20%. These changes are thesame for the Reference Case, the RPS or the NCESuntil 2017, when EPA rules are fully in effect andemissions of all three pollutants begin creeping upslowly under the Reference Case.

By reducing use of fossil fuels, the RPS and NCESboth continue to hold down emissions beyond2017; however, the NCES is significantly moreeffective. The RPS essentially holds emissions flat,but the NCES contributes further gradual reductionsin the pollutants over the entire study period,finishing in 2035 68% below the EPA requirementsfor sulfur dioxide, 58% lower for mercury, and 24%better for nitrogen oxides.

The superior performance of the NCES derivesfrom the policy’s tendency to replace coalgeneration with much cleaner natural gas early inthe study period, followed by continuedimprovement as plants with zero emissions of airpollution replace natural gas generation. The RPSeliminates less air pollution because its incentivesfavor renewable energy primarily at the expense ofnatural gas, leaving much more coal generation inthe mix to the end of the study period.

2.1.5 Policy costs

The cost for implementing the NCES can bemeasured several ways. The results of our costanalysis are shown in Table 2.1. The welfare costfor the policy in 2010 dollars is $307 billion. Thisis a cumulative cost through the year 2035 andrepresents the total policy costs across society. It

does not include the environmental or public healthbenefits of the policy, which are expected to faroutweigh the costs. The cost effectiveness of thepolicy is determined by calculating the price per tonof carbon dioxide emissions reduced. For theNCES this figure is $22 per ton. We provide nocomparable figure for barrels of oil. Because so littleoil is used in power production in the United States,this policy has insignificant impact on oil use, anda per-barrel figure would not be meaningful.

The increased cost of electricity under the NCESwould likely be too small for residential consumersto notice. Modeling shows the residential out-of-pocket electricity cost of the NCES to consumers (ona per household basis) is a $122 increase by theyear 2035 or an average increase of $4.88 peryear. The residential out-of-pocket natural gas costshows a slight decline by 2035 under the NCESpolicy. Modeling shows a $6 decrease in perhousehold natural gas cost by 2035, essentially azero cost effect.

The federal government would not havesignificant costs for the program. Other than thesmall increases in the cost of its purchases ofelectricity and natural gas, federal expenses wouldbe limited to administration, including settingstandards and monitoring compliance, which isalready performed within the EPA budget.

2.1.6 Regional and state price impacts of the NCES

Nationally, we have projected reasonable costsfor electric utilities to reduce carbon dioxideemissions through the NCES, but differences in the

30

Table 2.1 NEPI Clean Energy Standard Policy Key Metrics

a. oil reductions are insignificant - this policy is directed at CO2 emissions reductionsb. administrative cost only

Progress onOil Targetby 2035

Reductionfrom 2010(mmbd)

3.0

1.4

Increment vRef case

a.

PDVWelfare

Cost

(2010 $,billions)

n/a

307

CostEffectiveness

Oil

(2010 $/barrel)

n/a

n/a

CumulativeCO2

Reductions vReference

Case

2010 - 2035(mmt CO2)

21,678

n/a

13,842

Policy

Target

Reference Case

NCES

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

22

ElectricityCost (2010 $/household)

n/a

122

Natural GasCost (2010 $/household)

n/a

(6)

Impact onFederalBudget

Total2010-2035 TaxRevenue Impact(2010 $ billion)

n/a

b.

Impact on Consumer Householdsin 2035

Key Metrics Table

mix of energy resources in electricity marketsaround the country create significant variations inindividual regions and states. Future prices dependon an area’s current use of coal, availability ofnatural gas and renewable options, the stateregulatory system, and the ability to move gas orelectricity from other regions. Looking at six statesthat are significant users of coal, we found that thecost to reach the 80% clean energy standard inthose states ranged from 10 to 40 percent abovethe Reference Case (for all sectors, not onlyresidential users).

These numbers come from a paper by utilityconsultant Mark A. Foster, whom NEPI asked in2011 to analyze the state and regional cost impactsof a clean energy standard similar to the NCES (wehad not yet fully designed and modeled theNCES).27 The All Clean standard that Foster lookedat had been developed by the EIA during its workon Senator Bingaman’s clean energy standardlegislation. (For purposes of Foster’s analysis, theReference Case is the AEO2011) Although Foster’sresults are not precisely aligned with the NCES,they are close, and more than adequate to shedlight on the issue of regional differences.

2.1.6.1 Regional and state-by-state results

The EIA modeled power cost impacts of the AllClean standard on regions of the country, findingthat some areas could see significantly reducedelectricity costs by 2035, while others, includingTexas, would experience equally large percentageincreases. Foster found that the cost of meeting thestandard depended not only on the ability of aregion to inexpensively transition to a cleaner fuelsource, but also on the interconnection of electricalmarkets and degree to which an area is constrainedby transmission bottlenecks within or across itsboundaries. The state regulatory system played apart, too, as regulated utilities are generally morecapable of obtaining financing for nuclear powerinvestments (as we will explain in Section 2.1.8.1).

Foster further focused his analysis on six statesthat represented a wide range of circumstances and

probable cost outcomes, and grouped them intopairs: Texas and Florida, Pennsylvania and Ohio,and Colorado and New Mexico.

In Texas, where coal is currently an importantfuel, a shift to natural gas, nuclear and renewables,combined with a continued need to buy credits,drives a price increase in 2035 of more than 40%above the Reference Case. In Florida, coal isreplaced primarily with nuclear, and somewhat lesswith renewables, and credits are still needed in2035. Florida pays 30% more for electricity thanthe Reference Case in 2035, but its cost per kilowatthour remains higher than Texas because Texas ratesare lower to begin with. Both states see a declinein power generation due to the increased price andchanges in household income and population.

Power cost increases in Ohio and Pennsylvaniaare 32 and 37 percent above the Reference Case in2035, respectively; both are required to buy creditsto meet the standard in 2035. Ohio, which in theReference Case produces 80% of its energy fromcoal, reduces its power production by 30% under theclean energy standard, a change achieved throughefficiency and conservation and with imports throughthe regional grid. Power load is also assumed bynatural gas and existing nuclear generation, while athird of electricity still comes from coal-burning plants.In 2035, coal can remain in the mix of a cleanenergy standard through biomass co-firing andcarbon capture and sequestration, or throughpurchasing credits. Pennsylvania increases naturalgas use to replace coal.

Colorado and New Mexico are heavy coal usersin the 2035 Reference Case as well—81% and60% respectively—but modeling shows they canmake dramatic gains in use of renewables and inefficiency and conservation. New Mexico meets thestandard in 2035 at a cost only 9% over theReference Case, in part by decreasing power useand with in-state and imported clean energy.Colorado switches by 2035 from coal to nuclear,renewables, and natural gas, and reduces totalgeneration. Its power cost rises 32% over theReference Case in 2035.

2.1.6.2 Solutions to price differences

Some price differences in electricity cost betweenregions of the country are to be expected. The

31

27Mark A. Foster, “A Review of the Impacts of an ALL CLEAN CleanEnergy Standard for Selected Regions & States”, 2013.http://nepinstitute.org/wp-content/uploads/2013/04/Foster-Region-and-State-Impacts-of-Clean-Energy-Standard-Feb-20131.pdf

2009 cost of power in the six states we studiedranged from 6.9 cents per kilowatt hour in Texas to11.6 cents in Florida, a 68% difference. Federalpolicymakers need not intervene to reduce suchdisparities, relying on the market to act, andnormally they do not.

Our approach to preventing onerous electricityprice differences under the NCES also relies onmarket forces. While it is true that states are endowedunequally with clean energy resources, the studyshowed that importing energy, through electric or gastransmission lines, could reduce these inequalities.Transmission bottlenecks caused higher prices, butthese bottlenecks can theoretically be overcome.Decisions about building power lines will continue tobe made by regional governments balancing powercosts against other considerations. Likewise, stateswhose regulatory systems prevent financing ofnuclear power plants are free to make that choice inexchange for the potential impact on electricity prices.Finally, states with only higher-cost options to produceclean power can work toward meeting the standardby reducing power consumption, throughconservation and energy efficiency.

The market mechanism of tradable credits alsocan prevent price differences from becomingextreme. When clean power cannot beeconomically produced or imported to a state, theexchange of credits can assure an equivalentenvironmental benefit while holding down the priceof power at the same time. Tradable, bankablecredits effectively free the economic worth of cleanenergy electrons from the limitations of space andtime, allowing the value of these perishableresources to cross geographic barriers and flow intofuture periods of greater need.

A key to realizing this solution is to create acredit market that is broad and liquid, so that pricesfor credits rise only gradually, without disruptivespikes. Permitting existing low-carbon generation toreceive credits helps in this goal, as does theprovision allowing credits to be saved for future use.For these reasons, a policy which might seeminequitable, of granting credits to resource-richregions with existing generating capacity, in factalso benefits states where the clean energychallenge is greater.

2.1.7 Technology limits and renewable energy

The NCES can only achieve its aims if the costsit demands of society are within the range of thepublic’s willingness to pay. Credit trading canmobilize the lowest-cost energy alternatives, butsuccess ultimately depends on the optionsthemselves: will they be ready in time, can theyachieve the scale necessary, are they sociallyacceptable, and will they be affordable? In thissection we examine the ability of renewable energyto fulfill the role projected for it in the clean energystandard and in the next section the challengesfacing our projected increase in nuclear energy,along with ideas to meet these challenges.

The NCES does not drive a revolutionaryincrease in renewable energy. The increaseforeseen by the model appears well within the likelycapacity of technology development over the timespan of the policy. The NEMS-NEPI model showsthe NCES would, by 2035, increase renewableenergy in the American power generation mix to20%, including 7% from hydropower and 13%from all other renewables (Figure 2.13).

The Reference Case, which assumes minimalpolicy intervention, shows non-hydro renewablesgenerating 456 billion kilowatt hours (9% of thetotal generation market share) by 2035. The NCES,with 626 billion kilowatt hours of non-hydrorenewables generation in 2035, brings about areasonable 37% increase in non-hydro renewablesabove the Reference Case level (Figure 2.14).

32

9% 10% 5% 9%

11% 4% 2% 3%

22% 32%

24%

35%

3% 9%

24%

8%

55% 44% 45% 46%

0%

20%

40%

60%

80%

100%

Actual Reference Case RPS NCES

Perc

en

t o

f n

on

-hyd

ro r

en

ew

ab

le

gen

era

tio

n

Non-hydro Renewables Electricity Generation by Source (As a percent of total non-hydro renewables generation)

Reference Case, RPS, and NCES 2010 and 2035

Wind

Solar

Wood & Other Biomass Municipal Waste Geothermal

2035 2010

Figure 2.13

Fig. 2.13 Non-hydro Renewable Electricity Generation by Source

(As a percent of total non-hydro renewables generation)Reference Case, RPS and NCES - 2010 and 2035

Renewable energy made a dramatic surge inoutput in recent years, while federal incentives werein place, increasing by 68 billion kilowatt hours, or66%, in just three years from 2007 to 2010. Theincrease anticipated for the NCES over theReference case is 170 billion kilowatt hours, adifference to be made in 25 years—a pace lessthan a third as fast. Quicker mobilization ofrenewable power is probably possible, but we havechosen a market mechanism to govern selection ofclean energy sources in order to help society findthe least expensive path to reduce carbon dioxideemissions. Under this policy, without directincentives, the model shows renewable energy

growing at a healthy pace, but not booming as inthe last few years.

In light of this projected slower growth ofrenewables, it appears safe to assume that thetechnology will be able to meet the need when themarket is ready.

2.1.8 Nuclear power challenges and opportunities

Nuclear energy is an essential part of the mixthat allows the United States to reach the NEPIClean Energy Standard of 80% by 2035. Itcomprises 29% of electric generation at that time,compared to 18% in the Reference Case and 20%today. Nuclear-generated electricity is importantbecause of its stable price, predictable production,ability to scale up to large output, and lack ofcarbon dioxide, sulfur dioxide, nitrogen oxide andmethylmercury emissions. No other energy source

can match these qualities. Without nuclear power,the 80% standard probably is not achievable.

However, nuclear energy is beset by financial,safety and environmental concerns that must beaddressed before it can assume a much-increasedrole. NEPI asked Geoffrey Rothwell of the StanfordInstitute for Economic Policy Research to examinethese issues and suggest possible solutions thatwould make nuclear a financially viable andsocially acceptable option for rapid increases inU.S. power generation. His resulting paper outlinestechnical innovations and policies—well within thereach of current technology—that could accomplishthe task.28

2.1.8.1 Financial issues for nuclear power

Nuclear power plants can produce relativelyinexpensive energy predictably over a long periodof time, but the up-front cost of a plant is high andthe predictability is low of permitting and buildingone on budget and on time. Most countries thatemphasize nuclear power do so with governmentpolicies that spread the financial risk to society as awhole. In the U.S., where relatively small utilities mustgo to Wall Street to raise money to build plants, thepremium imposed for perceived risk can beunaffordable, making nuclear energy uneconomic.

Several factors drive the high cost of capital fornuclear power, but the sheer expense of a nuclearpower plant is fundamental to all of them. The EIAestimates the cost of a two-unit advanced nuclearplant at $12 billion, excluding the cost of borrowingfunds (in other words, if the plant could be builtovernight). For many utilities, Wall Street financingis not available because the plant, when built,would constitute too large a portion of the totalassets of the entire utility company. Even for largerutilities such an enormous capital investment, ifvulnerable to large regulatory or technologicaluncertainty, makes investors nervous andcommands a high cost of capital.

The ability of utilities to charge ratepayers forplant construction also affects the cost of capital. In26 states, representing 40% of the U.S. population,traditional rate-of-return utility regulation permits the

3316 46 56 53 19 17 17 17 38

147 241 217

4

42

245

51

95

203

465

287

0

200

400

600

800

1000

Actual Reference Case

RPS NCES

Billio

n k

ilo

watt

ho

urs

Non-hydro Renewables Electricity Generation by Source Reference Case, RPS, and NCES 2010 and 2035

(billions of kilowatt hours)

Wind

Solar

Wood & Other Biomass Municipal Waste Geothermal

2035 2010

Figure 2.14

Fig. 2.14 Non-hydro Renewable Electricity Generation by Source

Reference Case, RPS and NCES - 2010 and 2035(billions of kilowatt hours)

28Geoffrey Rothwell, “Small Modular Reactors: Costs, Waste andSafety Benefits”, 2012. http://nepinstitute.org/wp-content/uploads/2013/04/Rothwell-Small-Modular-Reactors-Dec-20121.pdf

operator to charge ratepayers for the cost offinancing during construction, reducing the risk andthe cost of borrowing. Another 20% of thepopulation lives in states with less generous rules,and 40% lives in states with unregulated utilities,where investors shoulder all of the risk. This latter60% of the population cannot avail itself oftraditional nuclear power plant constructionbecause the market considers the risk too high. Inits modeling projections, the EIA expects the use ofnuclear power will decline, in part because theNEMS model assumes all utilities operate inunregulated markets with costly capital.

One solution to these barriers is for the federalgovernment to provide loan guarantees to utilitiesthat build nuclear power plants. As exponents ofmarket solutions in the balance of our policyrecommendations, NEPI has not included loanguarantees for nuclear power in our package ofsolutions. We do believe utility regulation and thespreading of cost burdens to ratepayers is areasonable solution to enable communities toobtain costly but beneficial energy facilities, but inour federal system of government a national policyhas no role in deciding how individual statesregulate utilities.

A technical solution could save the day fornuclear power. Nuclear power plants do not haveto be so large and expensive. Huge plants enjoyeconomies of scale that potentially lower the cost ofelectricity, but much smaller plants could haveadvantages in safety, simplicity and massproduction, with total prices low enough to makecapital affordable for utilities that want to buildthem. Such a unit, called a Small Modular Reactor(SMR), would produce about a sixth as much poweras a unit at a traditional plant.

The cost of a civilian SMR has not yet beenquantified through a prototype engineering processfunded by the Department of Energy, but Rothwellestimates that, once in routine production, unitscould cost about $2 billion each. SMRs save moneytwo ways. By being more common and usingstandardized designs, components could be mass-produced in factories rather than being built one ata time on-site. Also, the smaller investment wouldentail less risk for utilities, allowing them to borrowat lower cost. The economies of scale lost by

building smaller plants might be offset by thesesavings; however, a mix of small and traditionallarge plants probably makes sense.

Small Modular Reactors are not a new idea—they have a long, illustrious history in the U.S.Navy’s 60-year record of operating them on shipsand submarines without a significant accident.Bringing this technology ashore, to civilian plants,promises broader use of nuclear power, andextension of the safety benefits already developedthrough their extensive use.

2.1.8.2 Safety issues for nuclear power

Three iconic accidents—Three Mile Island,Chernobyl and Fukushima—have instilledunderstandable skepticism in the American publicabout the safety of nuclear power. Rothwell examinesthe causes of these accidents in his paper. In eachcase, human errors either caused the accident orcontributed to making it worse. Error threatens toinfect every human endeavor. Nuclear energy cannotbe quarantined from this threat, but technology canstrengthen its immunities so that major accidents areextremely rare and, more importantly, have far lesserconsequences when they do occur. National safetyregulators need to be active.

On the other hand, passive safety systemsreduce the need for human intervention in anaccident, ideally allowing nuclear material to cooleven if systems fail or are not activated. All U.S.SMRs being considered for developmentincorporate passive safety measures. For example,the 180-megawatt Babcock & Wilcox mPowerreactor would be installed underground, allowingthe reactor’s heat to dissipate into the earth after ashut-down. In November 2012, The Department ofEnergy selected this reactor for engineering anddesign certification, with plans to install six unitsnear Oak Ridge, Tennessee.

New large reactors also will rely on the passivesafety concept, but SMRs have additional safetybenefits. Their designs can be based on U.S. navalreactors, with their exceptional safety record. Andthey would be built in larger numbers, allowing thereplication of reliability found in the aviationindustry when airliners are produced on anassembly line.

34

2.1.8.3 Environmental issues for nuclear power

Long-term disposal of used nuclear fuel remainsunresolved. With the cancellation of the YuccaMountain Nuclear Waste Repository by the ObamaAdministration, the radioactive waste of U.S. powerplants remains stored on-site, near where it wasused. By the time of its cancellation, the designcapacity of the Yucca Mountain facility already hadbeen exceeded by the volume of waste stored on

plant sites.Secretary of Energy

Steven Chu noted that on-site storage is consideredsafe for many decades,allowing time for a BlueRibbon Commission onAmerica’s Nuclear Futureto study a new solution.Among its findings, thecommission recommendedcreation of MonitoredRetrievable Storage facilitiesthat could consolidatematerial for severaldecades while waiting fora permanent disposalsolution.29 Rothwell findsthat these facilities couldhold waste from newplants for a couple ofcenturies at reasonablecost to utilities.

While decades ofnational procrastinationover permanent disposalof nuclear waste is hardlya virtue, it does not byitself constitute an

environmental threat. The waste is currently secureand no special intervention is required to keep it sofor many years. But action by Congress is required.Under current law, without a permanent disposalfacility coming on line, temporary consolidatedstorage cannot be licensed, and without consolidatedstorage, licensing and relicensing of plants is on hold,

as they must demonstrate they can ultimately bedecommissioned and their sites reclaimed—notpossible without a place to send waste.

Is permanent disposal of nuclear wasteimpossible? The answer appears to be no: severaltechnical solutions exist, including reprocessing torecover energy and reduce waste volume and useof underground salt domes that are stable for manymillions of years. The energy content of nuclear fuelis very high relative to its waste volume. Comparedto the billions of tons of fossil fuel carbon alreadydisposed of in the atmosphere, the mass of spentnuclear fuel is miniscule and the environmentalchallenge of containment relatively manageable.

2.2 Oil Security Dividend: A fully rebated oil tax

A broad-based tax is the most efficient means toreduce both oil use (and imports) and carbondioxide emissions, but recent history has shown thatproposals for a cap-and-trade program or straightcarbon tax are not politically viable (as discussedin Section 1.3.1). For elected officials whoseconstituents are constantly aware of the price ofgasoline, the very purpose of these policies—toincrease the cost of carbon-intensive energy to theend user—smells of political poison.

NEPI’s Oil Security Dividend policy overcomesthis difficulty by delivering the benefits of a broad-based oil tax without financial cost to the public. Atax would slowly increase the cost of fuel, withproceeds from the levy rebated in full to taxpayers.These refunds would completely offset the cost ofthe levy for the average taxpayer. Those who usemore than an average amount of fuel would paymore in total, and those who use less fuel thanaverage would receive a net financial benefit.

The dividend concept simplifies oil pricingstrategy by removing it from the arena of federalrevenue collection and wealth distribution. Thispolicy is revenue neutral and also neutral in itsimpact on income equality. Its purpose and impactare focused not on taxation but on the goal ofenergy policy, to shift the true costs of oil use to theusers of oil, including the cost of importing oil onnational security and on America’s economicstrength, and the environmental costs for climatechange and air pollution.

35

29Blue Ribbon Commission on America’s Nuclear Future, “Reportto the Secretary of Energy,” January 2012. Accessed March2012 at http://cybercemetery.unt.edu/archive/brc/20120620211605/http:/brc.gov//

NEPI engaged Roberton C. Williams of theUniversity of Maryland, to evaluate three oil taxrecycling concepts. We focus on one that imposesno net cost on the economy or the averagehousehold, and may slightly increase wealth—evenbefore considering the positive impacts of the oil taxitself. As we’ve seen elsewhere in this report,reducing oil imports can bring economic andenvironmental benefits to Americans well in excessof the cost of a carbon tax, even when the revenuesare not recycled. Under this policy, those benefitsaccrue without any cost to taxpayers. While nopolicy provides something for nothing, the OilSecurity Dividend is as close to a free lunch aspolicymakers commonly encounter.

2.2.1 How the Oil Security Dividend works

This policy imposes a tax on oil to account forthe costs that importation and combustion imposeon the environment and economy, but which are notincluded in the market price paid by users. The taxaffects all oil products. It is levied on a Btu basis atan amount that is equivalent to a certain level oftaxation on gasoline. The tax escalates over time toavoid price shock, beginning at 8 cents pergasoline gallon-equivalent in 2013 and increasingannually by 8 cents until 2035, when the taxreaches $1.84 cents per gallon. In consumerperception, this gradual rise would disappearamidst the routine volatility of gasoline prices, whichcommonly change by 8 cents or more in a singleweek (since 2008, the average retail price of agallon of gasoline has ranged from $1.61 to$4.12). The purpose of the tax is to influence oilusers to conserve fuel, a well-established resultbased on modeling and past experience.

The Oil Security Dividend policy differs from atraditional fuel tax because the revenues arerecycled to the taxpaying public. The total cost ofthe policy is very low—and may even benegative—because of the efficiency of an oil taxand the way the recycling mechanism encouragesmacroeconomic activity. The money paid at thepump, or paid in slightly higher costs of goods andservices, comes back in reductions to payroll orincome tax rates. Although these annual tax ratechanges are small, they accumulate over time as the

oil tax increases. By 2035, the average middle-income earner would pay $3,353 in oil taxes (in2010 dollars), but would receive a 3.4% tax cut thatwould fully offset that cost. By using less oil-basedfuel, the taxpayer could even profit from the system.

The offset to taxes could be applied to payroll orincome taxes with reductions or credits (in oureconomic analysis it made little difference which taxwas credited, so for simplicity we will referenceincome tax only). Tax rate adjustments would differby income level to make the policy neutral withrespect to income distribution. These adjustmentsare necessary because low-income Americansspend a larger percentage of their income on fuelthan wealthier Americans; consequently, fuels taxesare regressive, hitting the poor harder than the rich.Under our policy, by 2035, a taxpayer whoseincome was in the lowest fifth of all Americans’incomes would receive a tax break of 4.27%,compared to a break of 3.12% for a taxpayer inthe highest fifth. For each, this would represent anet benefit of .1% of income. (Figure 2.15)

Why would recycling the oil tax produce anegative cost—or net benefit—for the policy? Lowerincome taxes create an incentive to work and save,boosting the economy. Also, the fuel tax is aneconomically efficient way of raising revenue. As itoffsets taxes that are burdened with theinefficiencies of our complex income tax code,those inefficiencies are reduced. The overalleconomic efficiency of the nation’s tax systemincreases somewhat, yielding the net benefit shownin the modeling.

36

Lowest Quintile

2nd Quintile

Middle Quintile

4th Quintile

Highest Quintile

Average

Oil tax 5.04% 4.84% 4.32% 3.85% 3.39% 3.90% Inflation indexing -0.87% -1.08% -1.03% -0.86% -0.37% -0.70% Income tax rate changes -4.27% -3.86% -3.40% -3.09% -3.12% -3.30% Net burden -0.10% -0.10% -0.10% -0.10% -0.10% -0.10%

-4.50%

-2.50%

-0.50%

1.50%

3.50%

5.50%

Perc

en

t o

f In

com

e

Components of Net Burden by Income Quintile Oil Tax with Distribution-Neutralizing Tax Change, 2035

Figure 2.15

Fig. 2.15 Components of Net Burden by Income Quintile

Oil Tax with Distribution-Neutralizing Tax Change, 2035

The implementation of the policy would requirea more detailed review of the tax code than waspossible in this report. As noted, our modelingshowed the income tax or payroll tax wouldperform similarly for rebating the oil tax, but inactual practice, some lower-income Americansdon’t owe income taxes and don’t have to filereturns. A refundable tax credit could provide cashfuel tax rebates to these low-income workers.

2.2.2 Alternative designs of the Oil Security Dividend

Williams modeled three options for recycling oiltax revenue to the public (Figure 2.16).

The three recycling options are all revenue-neutral for the government, because all pay out100% of the proceeds from the oil tax. The scenariowe selected also maintained neutrality with respectto income distribution: it reduces income tax invariable amounts necessary to correct for theregressive nature of a flat cents-per-gallon fuel tax.The other two options included an across-the-boardincome tax rate reduction of the same percentagefor everyone, or a lump sum cash payment of thesame dollar amount for everyone.

Economic factors unrelated to energy dominatedthe analysis of the recycling portion of the program.Using his custom-built model, Williams found theacross-the-board income tax cut to be the lowest-cost option by a slight margin, but also found thatit would hit low-income earners hardest, making itregressive. These results came about because lower-

income earners spend a larger proportion of theirincome on fuel; if they receive a tax rebate at thesame proportion as the wealthy, their net cost willconsume more of their income. The modelingindicated the bottom 40% of earners would paymore in fuel tax than they would receive in incometax reductions, while the highest 60% would get anet benefit from the program.

This regressive aspect of the proportionalreduction option contributed to its total cost beingthe lowest of the three options we examined,although the savings compared to the option weselected were quite small. The proportional taxoption yielded a net benefit (or negative cost) of.15% compared to the .1% benefit for our preferredincome-distribution-neutral option. This result cameabout because tax cuts for higher earners creategreater economic efficiency: since their tax rates arehigher, reducing them does more to lessen tax-caused distortions to their economic behavior.

The third option Williams modeled provides afuel tax refund to everyone of the same dollaramount. The payment could be made as a taxcredit, or it could be paid out in mass distributionsof checks or direct deposits. This approach has acouple of appealing aspects. Sending out paymentswould render the benefit obvious to taxpayers,building support for the program, similar to the waythe popular Alaska Permanent Fund Dividendprogram has made that state’s savings fundpolitically inviolate. Also, the lump sum paymentswould be highly progressive, enhancing incomeequality, as the Alaska program is noted for doing.

But this option has disadvantages, as well. InWilliams’s modeling, its costs were significantlyhigher than the other two approaches. The reasonsare complex, and are treated in Williams’s paper. Inessence, the lump-sum dividend affects the economymore negatively because it creates a disincentive towork, and because it provokes tax increases tocompensate for the lost economic activity and theinflation caused by the increased oil price.

2.2.3 Oil Security Dividend benefits and modeling results

Analysis of the Oil Security Dividend involvedtwo models: the custom-made model Williams used

37

-200

0

200

400

600

800

1,000

2013 2018 2023 2028 2033

20

10

$ p

er

ho

use

ho

ld

Oil Security Dividend Net Cost per Household under Three Recycling Options

2013-2035 (2010 $ per household)

Net Cost after Proportional Tax Cut Net Cost after Distribution-neutral Tax Cut

Net Cost after Annual Dividend

Figure 2.16

Fig. 2.16 Oil Security DividendNet Cost per Household Under Three Recycling Options

2013 - 2035(2010 $ per household)

to look at recycling tax revenues, and the NEMS-NEPI model that informed our understanding of theenergy and emissions savings of all the policies inthe report. This section reports results from theenergy model.

The essential working part of the dividend policyis an oil tax introduced at a slowly graduated pace,adding the Btu equivalent of 8 cents a gallon ongasoline to all oil products every year for 23 years,until reaching $1.84 per gallon. The reason for theslow increase in the tax is to prevent the economic orpolitical shock of a sudden oil price rise. Despite itsgentle rise, however, the policy does bite, increasingthe price of oil and thereby reducing demand.

The model shows that the cost of the tax isessentially passed on in total to consumers, in theprice of gasoline and in refined products, as wewould expect. Similarly, the increased cost of higherfuel prices imposed on businesses flows through tohigher prices for goods and services. Sinceindividual consumers and not businesses willreceive recycled revenues from the tax, it makessense that consumers should pay these costs.

Higher fuel prices drive consumer decisions thatsave fuel, thereby cutting oil imports and reducingcarbon dioxide emissions. The model shows thatvehicle miles traveled decline by 4% compared tothe Reference Case in 2035, and total gasoline usegoes down by 9%. Oil use declines by 500,000barrels a day, with 300,000 barrels a day beingreduced from oil imports. The reduction incumulative carbon dioxide emissions is 1.7gigatons, or 1% compared to the Reference Case.

2.2.4 Policy costs

For the most part, policy costs for the Oil SecurityDividend are addressed in the sections above thatexplain the revenue recycling concept. They aresummarized in Table 2.2.

The welfare cost of a policy is the sum of its costsacross society. Since the Oil Security Dividendreturns tax proceeds to the public, the direct impacton consumers is a wash, or essentially zero.However, as discussed in Section 2.2.1 above,there is a net benefit to the economy from shiftingsome taxes from less efficient (income tax) collectionmethods to more efficient (oil tax) methods. Thatbenefit shows up in the table as a negative cost of$117.1 billion (in 2010 dollars).

The problem of how to allocate the welfare coststo our two goals of oil use and carbon emissions isunimportant in evaluating the oil security dividendpolicy, because its total costs are negative—inaggregate, it provides more benefit than cost, evenbefore looking at how it accomplishes the goals. Thefigures in the key metrics table show the total costapplied to each goal. In reality, these negative costs,or savings, can apply to only one goal or the other.For an explanation of this issue, see Section 1.6.3.

Since the policy refunds all fuel taxes paid, it hasno out-of-pocket cost for taxpayers in the aggregate.While those who use an above-average amount ofoil-based fuel would pay more, others who use abelow-average amount would pay less.

The federal government’s cost for the programwould be in administration only. Since the federalgovernment already imposes taxes on fuel and

38

Table 2.2 Oil Security Dividend Policy Key Metrics

1Motor gasoline cost is for all light duty vehicles; declines in cost are due to reduced VMT and/or reduced gasoline price (excluding the phased oil tax component)a. impact on residential electricity and natural gas costs are insignificant - these policies target oil consumptionb. administrative cost only



3.0

1.4

Increment vRef case

0.5

PDVWelfare

Cost

(2010 $,billions)

n/a

(117)

CostEffectiveness

Oil

(2010 $/barrel)

n/a

(46)

CumulativeCO2


Case

2010 - 2035(mmt CO2)

21,678

n/a

1,659

PDV of Cumulative

change in Total

Gasoline Cost1 v

Reference Case

2010-2035, netof Recycled Tax(2010 $ billion)

n/a

(186)

Policy

Target

Reference Case

Oil Security Dividend

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

(71)


n/a

a.


n/a

a.



n/a

b.

Impact onConsumer Households

in 2035

Key Metrics Table

income, the incremental cost of changing thesetaxes would presumably be insignificant.

2.3 Automotive fuel economy

After the 1970s energy crisis, Congressmandated manufacturers to produce more fuel-efficient cars through a seemingly simple mechanismcalled the Corporate Average Fuel Economy (CAFE)Standards. The system has proved effective, but farfrom simple. Average fuel economy has risen to 26miles per gallon, and under new rules adopted in2012, will double to 54.5 MPG by 2025, a level ofefficiency that will stretch the bounds of technologicalinnovation. In light of that success, and theestablished regulatory system surrounding CAFEStandards, we recommend no more than continuingthe already-existing regulations. However, we havealso studied a simpler, market-driven, way of meetingthe same objective with a more flexible policy, as wedescribe in this section.

The original CAFE Standards required eachautomaker to sell vehicles whose average mile-per-gallon efficiency met a government target, but theprogram created unintended consequences on amassive scale. By exempting light trucks, the rulesbrought whole new classes of gas-guzzling SUVs

and minivans into the market, which grew inpopularity until they comprised half of Americans’auto purchases.

New law and new EPA rules have addressedproblems in the CAFE Standards, introducing moresophisticated ways of addressing vehicle size, andallowing car companies to trade credits, so buildersof less efficient cars can pay for that right tomanufacturers of more efficient cars. With theseupdates, the CAFE Standards are a significant forcein saving fuel and reducing our dependence onforeign oil and our carbon dioxide emissions fromtransportation. We should retain those benefits, andno additional policy action is required to do so.

However, the CAFE Standards are far from ideal,so NEPI also studied a market-based policy calledfeebates. A market-based policy, unlike a mandate,can do more than it was originally intended to do,and can respond to the day-to-day changes thatinevitably affect real-world decisions. It is also lesssusceptible to the unintended consequences thatbeset the much more complex CAFE program.

A feebate is a fee connected to the sale of low-efficiency vehicles which is used to provide rebatesto buyers of high-efficiency vehicles. Research by ourconsultant, Ken Gillingham, of the Yale School of

39

Forestry and Environmental Studies, shows that asimple feebate policy can exactly mimic the impactof CAFE Standards as they are presently configured,but do so without complex rules or mandate rigidity.30

As with any mandate, the CAFE Standards cannotachieve results better than the target set out in therules. As a market-based approach, the feebate ismuch simpler, and therefore less expensive toadminister, and it continues to provide incentives tocar buyers regardless of whether companies haveattained a particular fuel efficiency average.Suggestive behavioral economics results indicate thatfeebates could even shape consumer choices thatwould be more beneficial to car buyers themselves,as well as to society.

2.3.1 How the CAFE Standards work

A full understanding of CAFE Standards requiresadvanced mathematics, so this review is confinedto general terms and leaves out many details.Complexity is a major flaw of the program. Thecomplexity of the standards increases each timepolicy makers seek to correct unintendedconsequences or meet unexpected circumstances.After more than 35 years, these patches andrewrites have evolved into a program that defiesany simple description.

The concept begins with a fuel efficiency numberin miles per gallon (MPG) that each automaker mustmeet, averaged over all the cars it sells. Underchanges made in 2011, the average MPGrequirement differs based on the size of the vehicle.Size is defined by a complicated formula based onthe footprint of the wheels on the road. Vehicles witha larger footprint have an easier MPG target tomeet, which critics point out is a benefit to USautomakers that build more large vehicles.

The program also includes add-on taxes andpenalties beyond the MPG average mandate.Inefficient vehicles other than minivans, SUVs andlight trucks pay a “gas guzzler tax,” which begins at$1,000 for cars with a rating of 22.5 to 21.5 MPG,and rises to $7,500 at an MPG rating worse than12.5 mpg. Another mechanism allows automakersunable to comply with the average MPG standard toget around it by paying a civil fine based on the

degree to which they fall short, multiplied by thenumber of vehicles sold; however, this provision hasbeen little used except by European automakers.

The CAFE Standards also provide variouscredits, which, like other parts of the plan, phase inand out in different years. Credits are provided tomakers of ethanol flex-fuel vehicles, dedicatedalternative fuels vehicles, vehicles with airconditioners that are efficient or that reduce loss ofrefrigerants, cars that can run on natural gas, andvarious other factors. New credits added in 2012complicate the system and make it subject tocriticism for unfairness. For example, Toyota hascharged a bias for American automakers becauselarge pickup trucks get credits for using hybridtechnology, applicable only to a single GM model,while small hybrids such as the popular Toyota Priusdo not get credits.31

One earlier criticism of the program was that theaveraging concept limited the improvement realizedwhen low-emission vehicles are sold. An automakermeeting the standard has no incentive to do better;thus, the sale of each additional hybrid or electricvehicle merely becomes a license to sell anotherhigh-emission vehicle, canceling the improvementfor energy savings and the environment. A systemof tradable credits added for the 2011 model yearis intended to address this problem, and to reducethe cost of the program.

Under the credit trading system, automakers thatsurpass their standard can sell excess credits tocompanies with less efficient fleets, or canexchange them within their own company or bankthem for use in a future year when the standardbecomes more difficult to attain. The credit systemcreates an economic incentive to improve fuelefficiency even after meeting the standard, as thecredits have monetary value. However, the criticismremains that a buyer of a hybrid car merely createspermission for an automaker somewhere to sell aninefficient vehicle, even if that vehicle is built by adifferent company that buys a credit.

Like the tax code, the complexity and changingrules of the CAFE Standards reward cynicism andspecial-interest political pressure. The business andautomotive press analyze each change for new

40

30Ken Gillingham, “The Economics of Fuel Economy Standards vsFeebates,” 2013.http://nepinstitute.org/wp-content/uploads/2013/04/Gillingham-CAFE-Standards-vs-Feebates-Apr-20131.pdf

31Dave Hurst, “New CAFE rules anger just about everyone,”Forbes, Sept. 4, 2012, accessed at:http://www.forbes.com/sites/pikeresearch/2012/09/04/new-cafe-rules-anger-just-about-everyone/

loopholes and evidence of winners and losers. Thepossibility of relaxing or reformulating the standardsin the future undercuts the incentive fortechnological improvement, as automakers mayfind it more cost effective to invest in lobbying forspecial relief provisions than building better cars.The 2011 changes negotiated by the Obamaadministration with the auto industry call for anincrease in average efficiency to 54.5 miles pergallon in 2025, but also include a 2021 review todetermine if that will be technically feasible—whichautomakers may see as an “out” for escaping therule when the time comes.

Part of adopting CAFE Standards as an elementof our comprehensive energy policy should be thecommitment to hold the line on backsliding.Automakers must invest in research anddevelopment to meet the standards in time. Firmnessin maintaining a consistent program will provide thecertainty necessary for success.

2.3.2 An alternative policy: feebates

A key advantage of the feebate concept is itssimplicity and transparency. Instead of mandatingmanufacturers to achieve an average fuel economystandard, the feebate imposes a fee on sales ofhigh-fuel-use vehicles and rebates those funds toreduce the price of fuel-efficient vehicles. Thefeebate concept provides a financial incentive toincrease fuel economy without a mandate to do so.It functions independently of technologicalconsiderations. Loopholes and special provisionsare less likely to be inserted into the policy becausecomplex rules are not needed in the first place. Thissimplicity should also make the policy lessexpensive to administer.

The effectiveness of the feebate is determined bytwo variables: the MPG pivot point below whichfees are charged and above which rebates aregiven; and the rate charged for variation from thepivot point. A higher MPG pivot point would createa stronger financial incentive for fuel efficiency. Inhis work, Gillingham shows that a pivot point andrate can be selected to replicate the results of theCAFE Standards. The feebate can also be revenueneutral, like CAFE Standards, when the pivotmatches the actual MPG average of the fleet. Theprice differential created by feebates pushes carbuyers toward higher-MPG vehicles, which

motivates manufacturers to produce more efficientvehicles for them to buy. If the government decidesto give a break to makers of larger vehicle classes,as in the CAFE Standards’ footprint formula, thatcan also be implemented through the feebate policyby setting a different pivot point for each size class.

As an example, a feebate designed to match thecurrent CAFE Standards would, in 2017, provide arebate of $313 for a vehicle that gets 40 MPG andcharge a fee of $1,361 for one that gets 20 MPG(figures here are in 2010 dollars). Although thepolicy remains nearly revenue neutral, the feebateramps up through the years to produce higher fuelefficiency. In 2025, the 40 MPG vehicle would earna rebate of $1,493 while the 20 MPG vehiclewould cost an extra $6,493.

The feebate policy could be designed to work atthe manufacturer or consumer level withoutsubstantively affecting the results. We prefer theapproach of charging the fees and issuing therebates at the point of sale, perhaps showing thefeebate as an item on the price sticker. Suchtransparency would take advantage of a tendencyof consumers, found in some behavioral economicsstudies, to respond more strongly to costs that areshown explicitly than costs that are built into a totalprice. More research is needed to verify this effect,so we have not included it in our modeling, but itcould increase the cost-effectiveness of the policy.

Behavioral economics research also suggeststhat consumers undervalue fuel economy in vehicle-purchase decisions. Researchers disagree on thestrength or reason for this prejudice, which couldbe caused by buyers using an extreme discount rate(placing too low a value on future costs), or couldreflect simple inattention on the part of buyers totheir total costs. Either way, a consumer-levelfeebate could benefit car buyers by bringing theissue to the fore and influencing them to choosemore efficient vehicles that will also save themmoney in the long run.

2.3.3 CAFE Standards benefits and modeling results

The modeled benefits of the new CAFEStandards are substantial and are expected tostrain the technological limits of attainable vehiclefuel efficiency. The changes made in 2012 are

41

included in our comprehensive policy as animportant part of meeting our goals. The alternativefeebate policy does not strive for efficiency greaterthan that, but seeks the same benefit in a way thatis simpler and more transparent.

The benefit of either program amounts to areduction of 1.0 million barrels per day (mmbd) ofoil in 2035 (of which 0.7 mmbd is imports),compared to the 2010 baseline. Carbon dioxideemission reductions are a cumulative 1.8 gigatonsover the period.

2.3.4 Policy costs

The federal government has already adopted theCAFE Standards policy that we recommend as partof our comprehensive policy, so its inclusion herehas no incremental cost to society. However,consistent with the rest of our analysis, we havecalculated costs for the CAFE Standards andincluded them in our bottom-line costs. The costs forthis policy are shown in Table 2.3.

Welfare costs for the CAFE Standards calculatedby Gillingham have three major components: theincreased cost of the more efficient vehicles, less thesavings in fuel used, plus the cost incurred by therebound effect—which is the additional costimposed on society when greater fuel economyinduces more driving. Additional driving createscosts for congestion, accidents, road construction,and so on. Our modeling shows the cost of thisrebound effect to exceed the cost of the policy itself;as we shall see, however, combining the CAFEStandards policy in our comprehensive policyreduces overall vehicle miles traveled, thuseliminating this issue (see Section 3.2.1.1).

Our decision to use conservative assumptions incalculating welfare costs is particularly significantwhen looking at CAFE Standards. As explained inSection 1.6.1.1, consumers appear to undervalueenergy savings when making investment decisions,a phenomenon especially prevalent in car-buyingbehavior. If this so-called market failure is real, thena policy which produces more fuel-efficient cars willsave money for consumers, reducing the cost forsociety. However, this proposition is debatable: wedo not know consumers’ motivations, and they maybe responding to important hidden costs related tosafety or other product attributes. We have choseninstead to assume that auto buyers are rational intheir choices and accurately trade off their fuelsavings with other vehicle attributes. In this case,CAFE Standards do not benefit consumers, and thecost of the policy includes more than the individualfuel savings. Our decision increases the cost of thispolicy by $125 billion to $180 billion, dependingon the assumption used to quantify the marketfailure. The reason for our decision is that we lacka satisfactory basis upon which to quantify themarket failure. We could have reduced these costsby assuming that consumers do not choose theirvehicles rationally, but this approach would haverequired an assumption about how large themarket failure is.

In allocating these welfare costs to our two goals,of reducing oil use and carbon dioxide emissions,we report the entire cost in the cost per barrel andcost per ton for each goal. This way of reportingthe cost tends to double its impact: if the total costis assigned to oil use reduction, then the carbondioxide reduction would be free, and vice versa.This issue is explained in Section 1.6.3.

42

Table 2.3 CAFE Standards Policy Key Metrics

1Motor gasoline cost is for all light duty vehicles; declines in cost are due to reduced VMT and/or reduced gasoline price (excluding the phased oil tax component)a. impact on residential electricity and natural gas costs are insignificant - these policies target oil consumptionb. administrative cost only



3.0

1.4

Increment vRef case

1.0

PDVWelfare

Cost

(2010 $,billions)

n/a

196

CostEffectiveness

Oil

(2010 $/barrel)

n/a

55

CumulativeCO2


Case

2010 - 2035(mmt CO2)

21,678

n/a

1,754

PDV of Cumulative

change in Total

Gasoline Cost1 v

Reference Case


n/a

(361)

Policy

Target

Reference Case

CAFE Standards

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

112


n/a

a.


n/a

a.



n/a

b.


in 2035

Key Metrics Table

Consumer out-of-pocket costs for the policy arerelatively simple: the fuel savings realized by driversoffset the increased cost of more fuel-efficient cars,leaving little increased expense for auto owners.Federal government costs are insignificant: theincreased CAFE Standards cost EPA $8 million andthe National Highway Traffic Safety Commission$3 million in FY 2013.

2.4 LNG Trucks

Technology will transform the commercial truckfleet that carries America’s freight in coming years.For many applications, trucking companies will findliquefied natural gas (LNG) is the best fuel choicefor heavy and medium trucks. In our 2010 report,NEPI recognized this technology as an inexpensiveway to reduce oil imports. Since then, lower LNGprices, technical advances, and fueling station

construction have combinedto make this fuel aneconomic winner regardlessof policy interventions. Thus,our policy to encourageLNG can be modest,enhancing a beneficialchange that is alreadybeginning to happen.

Traditional internalcombustion engines canbe built or converted toburn LNG or compressednatural gas (CNG) insteadof diesel or gasoline.Natural gas becomescompact enough to use asa vehicle fuel when storedunder pressure, in the caseof CNG, or cooled tonegative 260 degrees F,for LNG. The smallervolume and lower pressureof LNG make it a morepractical alternative thanCNG in highway trucks.

For operators, LNG’srelatively clean exhaust eliminates the need for theexpensive emissions control equipment required fordiesel trucks. However, LNG adds cost to the engine

and fuel system and increases the weight of the truck.Also, the fuel can vent from tanks when it gets toowarm, creating safety issues and increasinggreenhouse gas emissions. Maintenance facilitiesneed special safety equipment to deal with lighter-than-air natural gas. An additional limitation, a lackof LNG engines powerful enough for manyapplications, is being resolved with the introductionof new equipment by manufacturers.

On the eve of game-changing technical andmarket changes, LNG is already the lowest-cost fueloption for long-haul trucks and is beginning to beused. But the industry is understandablyconservative about investing in new equipment withunfamiliar characteristics for maintenance,operations, fueling and safety. Since we seek apractical policy, we have put these potentialbarriers at the center of our analysis, evaluating thereal life issues and business case for an actualtrucking company. That analysis, by Anna Lee Deal,forms the basis for this section.32 Deal is Coordinatorof Green Initiatives for Lynden Incorporated, a multi-modal transportation company based in the PacificNorthwest and Alaska.

2.4.1 A policy supporting LNG trucks

Despite the many factors facing truckingcompanies in adopting LNG as a fuel, a simple,inexpensive policy prescription is most helpful tobring about conversions. Trucking companyexecutives are focused on the cost savings of usingLNG instead of diesel. The overriding importanceof fuel cost is such that when operators gainconfidence in sufficient savings, market forces willbring down barriers to conversion. Indeed, thisprocess is already happening, as we will explorelater in this chapter (see Section 2.4.5).

Our policy reduces the excise tax on natural gasas a motor fuel to zero, then brings it back to levelscomparable with diesel over a 10-year period.Currently, LNG is taxed at the same rate per gallonas diesel, which puts it at a disadvantage becausediesel contains more energy per gallon. Our policywould put LNG and diesel on a level footing after

43

32Anna Lee Deal, “What Set of Conditions Would Make the Business Case to convert Heavy Trucks to Natural Gas? – a CaseStudy”, 2012.http://nepinstitute.org/wp-content/uploads/2012/12/Natural_Gas_for_Heavy_Trucks_201211051.pdf

the 10-year tax break, by setting the tax at a rateequivalent by Btu (a unit of energy) rather than bythe gallon. The 10-year tax break would also affectcompressed natural gas, or CNG, but we do notexpect a large increase in use of CNG and itsimpact is insignificant in our modeling.

By temporarily lowering the price of LNG as afuel, the policy helps push trucking companies andthe fuel industry beyond the inertia that wouldotherwise slow conversion. New fueling stationsand maintenance shops must be built, employeesmust be trained in new equipment and safetyprocedures, and new operating considerations mustbe accounted for. With sufficiently large savings onfuel, these changes, which are already impending,will happen more rapidly.

2.4.2 Alternative policies not selected

NEPI considered several other policy ideas toencourage faster LNG conversion. We wereinfluenced by the recognition that conversion isbeginning to happen without any policy influence,and by the desire to adopt a low-cost option.

LNG trucks are more expensive than traditional,diesel-fueled trucks, a difference that could beaddressed through a tax credit. But such tax creditswould be a comparatively expensive means toencourage conversions. Trucking companies’concerns with the cost of trucks is directly related tothe long-term cost of fuel. Thus, addressing the cost offuel through a tax break is simpler and probably moreeffective in bringing about purchase of LNG trucks.

Trucking companies would be strongly motivatedby a policy that added confidence to fuel prices bysomehow guaranteeing a given price differentialbetween diesel and LNG. Our policy goes partwaytoward doing this by lowering taxes on LNG whileleaving them in place on diesel. But a policy tocontrol fuel prices would be too complex andpotentially too expensive.

One could also imagine a policy that would godirectly to the barriers to LNG truck mobilization,by somehow speeding construction of fuelingstations and maintenance shops. But these issuesare already being resolved by the market.

Finally, we could conceive of a policy to allowheavier LNG trucks on the road without diminishingtheir payload weight, a strong incentive for truckers.

But vehicle weights are controlled by the states,which also pay for road damage from overweighttrucks. A federal policy overriding this authoritywould be unfair to the states and thereby unlikelyto be adopted.

2.4.3 LNG truck benefits and modeling results

To arrive at our modeling results, we adapted theAEO 2012 Reference Case. We used newinformation about the cost of LNG trucks, weincluded the increased weight of LNG equipment,which reduces truck payload capacity, and weincorporated changes the EIA had made for its2013 outlook on vehicle miles traveled. When theEIA released a preliminary version of its AEO2013, we found its results similar to ours, showinga faster increase in LNG truck sales than previouslyshown in the AEO 2012 Reference Case.

The NEMS-NEPI model shows LNG truckscontributing a reduction of 400,000 barrels a dayof oil use by 2035 compared to the Reference Casein the EIA’s Annual Energy Outlook 2012. Total oiluse in 2035 would be 16.7 million barrels a day.Of the 400,000 barrel reduction, our policy ofexcise tax reduction would contribute 100,000barrels, and the balance comes about due to ourupdates to the modeling assumptions.

In 2010, only 0.1% of the 5 million trucks on theroad operated on LNG. That share increases to just0.2% by 2035 in the AEO 2012 Reference Caseprojection. Our analysis of the market insteadshows LNG trucks increasing to 11.9% in 2035without policy intervention; adding our policybrings LNG to 14.5% of the truck fleet by 2035.

The increase in LNG trucks contributes 0.1gigatons to our carbon dioxide reduction goal.

2.4.4 Policy costs

Costs are summarized in Table 2.4. As with ourother policies, the welfare costs here contain theassumption that the market accurately assesses thevalue of investments in energy technology.However, it is possible that fleet managers’ cautionin adopting new technology may not beeconomically optimal. If that is the case, then ourcost estimate would be too high, because the policywould tend to save money for companies that

44

otherwise would not do so. However, because wecould not quantify the quality of these businessdecisions, we chose the most conservativeassumption, which maximizes the cost. See Section1.6.1.1 for a more complete explanation.

The numbers reporting the cost effectiveness ofthe policy also require explanation. The LNG truckpolicy impacts both of our goals, for oil use andcarbon dioxide emissions. The figures for costeffectiveness, in dollars per barrel or per ton, showthe entire cost of the policy allocated against eachgoal. This double-counts the costs. If the policy costsare attributed in whole to one goal, the cost ofmeeting the other goal would be zero. This issue isexplained in Section 1.6.3.

Out-of-pocket costs for consumers relate to theimpact of the policy on the price of natural gas perhousehold and the electricity gas generates. Usingmore LNG for transportation does push gas prices upslightly, but the impact on consumers is small, $6 ayear for electricity and $3 a year for natural gas forthe year 2035.

The cost to the federal treasury comes in thereduction of the excise tax in the first 10 years ofthe policy, and the equalization of taxes for LNGwith diesel by energy content, which results in alower tax on LNG compared to the Reference Case.The initial tax cut reduces federal revenues by $800million. After 2025, when the tax cut ends butequalization with diesel continues, the tax reductiontotals $4.6 billion. For the entire 20-year period ofthe analysis, the loss of tax revenues totals $5.4billion compared to the Reference Case. NEPIanalyzed other options, including eliminating theexcise tax entirely, but found little increase in LNGtrucks for the increased cost.

2.4.5 Business case and challenges for LNG trucks

Our economic analysis of the shift from diesel toLNG trucks focuses primarily on the cost of fuel andof trucks. Trucks that burn LNG currently cost$30,000 to $70,000 more than comparable trucksthat burn diesel. Conversion kits for existing dieseltrucks cost $25,000 to $40,000. At the time Dealdid her business analysis, natural gas prices were65% to 70% lower than oil on an energy equivalentbasis. As noted elsewhere, our modeling alsoincluded the reduced payload capacity of trucks fittedwith LNG tanks, which weigh more than diesel tanks.

At the current diesel price of about $4 pergallon, Deal shows that trucking companies canmake a return on investment of over 20% usingLNG in line-haul trucks that drive at least 70,000miles a year. With the introduction of newequipment by manufacturers, these results can beexpected to improve.

Despite the attractiveness of this analysis,barriers stand in the way of greater LNG adoptionthat are not easy to model. These barriers includedifferent maintenance and fueling needs, newsafety procedures, and the understandable cautionof trucking companies to embark on use of a newtechnology with unfamiliar characteristics. In eachcase, an increased return on investment canaddress these concerns, which is why our policyreduces LNG taxes for 10 years, followed by paritywith diesel on an energy basis.

Since it isn’t practical to model the barriers toconversion to LNG, we address them individuallyin the following sections to show how they can beaddressed by the market.

45

Table 2.4 LNG Trucks Policy Key Metrics

1Motor gasoline cost is for all light duty vehicles



3.0

1.4

Increment vRef case

0.1

PDVWelfare

Cost

(2010 $,billions)

n/a

0

CostEffectiveness

Oil

(2010 $/barrel)

n/a

0

CumulativeCO2


Case

2010 - 2035(mmt CO2)

21,678

n/a

35

PDV of Cumulative

change in Total

Gasoline Cost1 v

Reference Case


n/a

7

Policy

Target

Reference Case

LNG Trucks

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

0


n/a

6


n/a

3



n/a

(5)


in 2035

Key Metrics Table

2.4.5.1 Limited fueling infrastructure

An LNG fueling station receives deliveries intankers and stores it with liquid nitrogen coolingequipment. Trucks’ tanks look similar to diesel tanks,but are larger and heavier, since LNG has lowerenergy density than diesel and must be isolatedfrom the environment by a thermos-like vacuumcontainer. Fueling lines attach differently than dieselpumps and operators must wear gloves, aprons andface shields for protection from the super-cooledmaterial. Gas that escapes evaporates rapidly anddisperses in open air.

These differences create no insuperable barrier tobuilding LNG fueling stations if a profitable marketexists. An LNG fueling station costs about $1.5million to build. Until recently, stations were locatedalmost exclusively in California, and most of thosewere not public. But in 2013, major companies inthe truck fueling and natural gas industries plan tocomplete a national network of 150 stations toconnect major Interstate Highway routes at intervalsof 200 to 300 miles. By 2015, the investors expectto have covered all regional routes with 300 to 400stations. Companies with large, busy fleets could alsobuild their own fueling stations.

A network that covers major trucking routes on theinterstates will create a positive business opportunityfor truckers. However, early adopters of LNG truckswill still sacrifice the flexibility of diesel trucks, whichgo almost anywhere and find fuel for sale.

2.4.5.2 Fuel temperature issues

Truck fuel tanks keep LNG cold with highlyeffective insulation, but after four or five days thefuel warms and expands to the point that it mustvent to the atmosphere. Venting wastes fuel andemits greenhouse gas 25 times as potent as carbondioxide. Venting also degrades the fuel left in thetank, as the lighter portions of natural gas that ventleave behind residual fuel that is of lower quality.

To avoid the venting problem, trucks must beused enough to require refueling every two days.For line-haul trucks driving the number of miles thatwould support the LNG investment, this frequencyof refueling would be expected. However, the issuedoes add an operational consideration that truckingcompanies must account for.

The tanks also require special maintenance toretain the super-insulating quality that preventsventing. After about five years, the vacuumdegrades and must be re-established with specialequipment; subsequent servicing is needed everyyear or two. Accidents can also damage thevacuum layer of the tank and cause venting.

2.4.5.3 Cost to upgrade maintenance shops

The potential escape of natural gas createsspecial safety hazards in repair shops that serviceLNG trucks. Fire codes address these issues withrequirements such as improved ventilation andexplosion-proof lighting. A trucking companyadopting LNG must either contract with a shopalready equipped to meet these codes, whichincreases the cost of maintenance, or must build itsown LNG maintenance shop, at a cost of up to$200,000 per repair bay. Other than the capitalcost of the shop, maintenance costs for LNG trucksare about the same as for diesel trucks that haverequired emission control equipment.

2.4.5.4 Limited engine options

The lack of infrastructure for LNG trucks hasretarded adoption of the technology, which has inturn discouraged manufacturers from producing a fullrange of LNG truck engines. Large-displacementengines suitable for heavy trucks have beenunavailable. That logjam appears to be breaking,however, as three manufacturers are bringing outmore powerful engines. As more trucking companiesstart using these money-saving trucks, construction offueling stations and maintenance shops will becomeprofitable, which, when built, will in turn improve theeconomics of adding more LNG trucks.

2.5 Energy efficiency

Avoiding energy waste in homes and businessesis a simple, inexpensive and obvious way ofreducing carbon dioxide emissions. When efficientappliances and buildings require no sacrifices fromusers in terms of their costs or comfort, they willdeliver the same benefits for less energy. Thesesavings can be slow to accumulate, since policychanges mainly affect investments that last a long

46

time, but the savings made are real and permanent. For the most part, the federal government has

adopted effective policies to improve the energyefficiency of appliances and buildings. For example,an efficiency standard for lightbulbs, one of 55Standards adopted for various products since 1987,effectively made most incandescent bulbsunavailable as of 2013. Rather than recommend anew strategy, we can accomplish enough bycontinuing efficiency policies that would otherwisesunset, and by gradually strengthening standardsand codes through existing administrative processes.

2.5.1 A policy for energy efficiency

The Energy Information Agency gathered abasket of energy efficiency policies to study for itsAnnual Energy Outlook 2012, called theExtended Policies Side Case. The purpose of thisside case was to look at the impact of extendingand incrementally expanding the impact ofexisting federal energy efficiency and renewableenergy policies. We have adopted much of theExtended Policies Case as our own energyefficiency policy.

Using the EIA’s case yields several benefitsbesides saving effort with an already-completedanalysis. The policy follows a clear path that hasmainstream acceptance, as demonstrated by thefact that large portions have already been madelaw in some form. Like the rest of ourcomprehensive policy, the energy efficiencyelement relies on existing or predictably feasibletechnology, offers a real and measurable impact,and faces no insuperable political barriers.

The three main components of the policy are:Energy standards for appliances and otherequipment; energy-saving building codes; and taxcredits for investments in renewable energy andcombined heat and power production in homesand buildings, known as end-user tax credits. Ineach case, the policies exist today, but arescheduled to end or will need to be updated tomaintain effectiveness. We would continue andupdate the policies through 2035.

We left out portions of the EIA’s ExtendedPolicies Case that are addressed in other wayselsewhere in our plan. We excluded CAFEstandards, because we addressed them as a

separate policy. We also excluded tax creditsfor renewable energy investment andproduction by utilities. Our modeling shows thatthe NEPI Clean Energy Standard will bringabout as much renewable energy as would bedriven by the tax credits, and providing taxrelief as well is unnecessary.

Appliance standards require manufacturers toattain minimum efficiency in the products they sell.For example, the lighting standard set in 2007required bulbs to be at least 25% efficient, meaningat least 25% of the electricity flowing to the bulbwould create light rather than heat or another formof waste. Since incandescent bulbs could not meetthis standard, they were no longer legal afterregulations took effect.

Federal standards exist for a broad range ofproducts, and some states have their own additionalstandards. Past energy savings from these rules aredifficult to quantify precisely. A standards advocacygroup reported that electricity use in 2010 wasreduced by 7% over what it would have beenwithout federal standards.33 Other estimates forparticular programs have ranged from 4% to 25%.

To continue working, a standard must beupdated and tightened as technology improves. Thedetailed technical process can take place within theexisting federal system. To model the progress ofimproving standards, the EIA assumed thestandards would be based on voluntary Energy Starstandards or the Federal Energy ManagementProgram, which provides purchasing guidelines forfederal agencies, and also assumed introduction ofnew standards on products not now covered.

For building codes, energy efficiency typically isimproved by specifying materials rather than settingefficiency goals. For example, a code could requireinsulation of a certain material and thickness, orwindows with certain specifications. In the UnitedStates, building codes are established and enforcedat the state and local level, where model nationalcodes can be adapted to the local climate,materials, and cultural patterns. Some areas havenot adopted energy efficiency building codes, butmandating codes under federal law is problematic,as we discuss in Section 2.5.4.

This analysis assumes a 30% increase in a

47

33Appliance Standards Awareness Project,http://www.appliance-standards.org/national

national energy building code by 2020 relative tomodel codes set for residences in 2006 (by theInternational Energy Conservation Code) and forcommercial buildings in 2004 (by the AmericanSociety of Heath, Refrigerating and Air-Conditioning Engineers Building Energy Code90.1-2004). Additional 5% increments ofimprovement come in 2023 and 2026.

We have not independently evaluated thetechnical feasibility of these higher standards.However, when the EIA put together the policies,the standards were kept within maximumtechnological feasibility levels already establishedby the Department of Energy.

The final part of the policy provides tax creditsto building owners for investments in renewableenergy equipment and combined heat and powerequipment. These tax credits are already on thebooks; our policy merely extends them to 2035.

2.5.2 Energy efficiency benefits and modeling results

The outlook of NEPI-NEMS modeling extends to2035, but much of the benefit for energy efficiencywork, especially in buildings, will take many yearsto show up, and will last well beyond the end of ourmodeling horizon. Appliances and equipment lasta long time, and building energy codes drivechange primarily for new construction. But the long-lasting benefits of these changes should berecognized in policy evaluations even if the resultsdon’t accrue to a single generation.

Energy efficiency improvements for buildings onlymodestly reduce oil use, because oil heats only 6percent of U.S. homes and generates only 1 percent

of the electricity powering appliances. Compared tothe Reference Case, these efficiency policies wouldreduce oil use by a very small amount.

The policy reduces cumulative carbon dioxideemissions 2.1 gigatons, down from 144.5 gigatonsin the Reference Case, or 1.4%. Total deliveredenergy in the residential sector is reduced by 1.0quads (quadrillion Btus) in 2035, from 11.9 quadsin the Reference Case, a cut of 9%. In the commercialsector, the reduction is 0.5 quads in 2035, from10.3 quads in the Reference Case, or 5%.

2.5.3 Policy costs

Energy efficiency provides the most cost-effectivecarbon dioxide reduction of all of our policies.However, the scale of the improvement is limitedbecause of the limited opportunities for increasingenergy efficiency. Cumulative carbon dioxidereduction through 2035 is 2.1 gigatons. Figures areshown in Table 2.5. As we will see in Part 3, energyefficiency policies have positive interactions with theother components of the comprehensive policy.

In addition, encouraging energy efficiencyprobably saves money for building owners.Investments that deliver fuel savings in excess of theup-front costs would seem to need noencouragement from government. However,homeowners and other building owners showreluctance to make these apparently attractiveinvestments, for reasons that may be justified byhidden costs, or may be related to behavioral quirksor market inefficiencies. In our welfare cost analysis,we take no credit for building owners’ energysavings, assuming that all their decisions areeconomically optimal. While this assumption is not

48

Table 2.5 Energy Efficiencies Policy Key Metrics

a. oil reductions are insignificant - this policy is directed at CO2 emissions reductions



3.0

1.4

Increment vRef case

a.

PDVWelfare

Cost

(2010 $,billions)

n/a

2

CostEffectiveness

Oil

(2010 $/barrel)

n/a

n/a

CumulativeCO2


Case

2010 - 2035(mmt CO2)

21,678

n/a

2,060

Policy

Target

Reference Case

Energy Efficiencies

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

1


n/a

(182)


n/a

(41)



n/a

(82)

Impact on Consumer Householdsin 2035

Key Metrics Table

supported by the evidence, we have chosen it as aconservative upper bound on our policy cost, sincewe cannot quantify the degree to which these costsrepresent market failures. A fuller explanation isfound in Section 1.6.1.1.

The energy efficiency policy is not effective inreducing oil use. For this reason, our cost-effectivenessnumber references only our goal of reducing carbondioxide emissions, which is given in the cost per tonreduction of carbon dioxide emissions.

Out-of-pocket costs for the policy are negative.The energy efficiency policy reduces residentialexpenditures for electricity in 2035 by $182 perhousehold and natural gas in 2035 by $41 perhousehold, both by cutting energy use and bycutting energy price. Reductions in price comeabout due to lower demand.

Federal outlays for the energy efficiency policyare inconsequential, but the tax incentives affectrevenues, which impact deficits equally as if theywere expenses. As explained in Section 2.5.1, ourpolicy differs from the EIA Extended Policies SideCase in AEO 2012 by allowing expiration ofcurrent tax incentives for utilities to invest in andproduce renewable energy. Expiration of theseInvestment Tax Credits and Production Tax Creditsreduces the cumulative cost to the treasury from $19billion in the Reference Case in 2035 to $14.3billion in our policy.

The $4.7 billion in federal savings on taxincentives for utilities partly offsets extended taxcredits for end users. The Energy Efficiency Policyretains the Investment Tax Credit for renewableenergy facilities installed by end users, whichprimarily go to owners of residential properties. Thecredits total $95 billion by 2035, an average of$3.7 billion per year. With the expected expirationof these tax breaks in the Reference Case, thecumulative credits in 2035 would be $22 billion,or an average of $900 million per year.

2.5.4 Challenges for building energy codes

Building codes assure quality and durability ofconstruction—as well as safety—to protect theinterests of future owners of structures that lastgenerations, and the interests of society as a wholein its investment of resources. The market alone

cannot protect these interests, because the complexityof construction and longevity of buildings createbarriers to knowledge of construction practices andquality by future buyers. These reasons hold equallyfor codes for energy efficiency, as many jurisdictionshave already recognized.

Despite the clear-cut case to be made for energybuilding codes, however, two challenges limit theeffectiveness of this strategy as part of a nationalenergy policy.

First, the time span of modeling does not reachfar enough into the future to take into account thefull benefit of the improvements, or even the fullimplementation of the policy. Energy improvementsmade in products such as insulation and windowscan easily last 50 years or more, but our modelingreaches out only about 20 years. Even if allbuildings were brought to code immediately, ournumbers would capture only a fraction of the totalbenefit of the investment. But, of course, manydecades will pass before all buildings meet energycodes. For the most part, codes are relevant only tonew construction, and new buildings replace oldbuildings slowly. The impact is lasting, but takes along time to show up.

The other major challenge to the strategy arisesfrom U.S. federalism. Building codes are adoptedand enforced at the state and local level. While thefederal government probably could impose a nationalenergy building code, it lacks an enforcementmechanism. Replicating the myriad of building codeauthorities that function at various levels ofgovernment would be an expensive, bureaucraticand likely unwelcome change to the way communitieshave done business for generations.

Instead, the federal government should usepersuasion and incentives to influence state andlocal governments to adopt and enforceincreasingly stringent model codes on their own.This is the process that has been used to date. It isthe simplest and most politically palatableapproach, and it has the benefit of harnessing thedesire of citizens and their representatives toaddress energy and climate needs. Most Americanprogress on climate change mitigation has comefrom individuals, businesses and state and localgovernments, and it is reasonable to count on thoseefforts continuing.

49

Part 3: Conclusion: Combining policies into a complementary, integrated whole

The ideal national energy policy would combineelements that, like our 50 unique states, build uponone another’s strengths to create a united wholemuch more powerful than its individual or collectiveparts. Our five most promising policies do just that.Standing alone, each contributes effectively and atlow cost to meeting our goals, reducing oil use andimports and reducing carbon dioxide emissions.Together, they also interact to magnify theireffectiveness and produce a larger positive result.In fact, we find that when we combine the policies,key costs go down even as effectiveness increases.

This result is not a surprise. We originally set out,beginning with the many studies that contributed toour 2010 report with Resources for the Future, toidentify practical, low-cost policies that wouldinteract positively with one another. A broadpricing methodology such as a carbon tax or a cap-and-trade system could perform better than anyother single policy, but we recognized the politicalunlikeliness of such a tax being enacted. Instead,this package of five policies delivers nearly thesame result at low cost, but without politicalimplausibility. Indeed, two of the five policies in thepackage are substantially in place and a third istrending strongly in the direction we call for.

Other policies could be found that wouldaccomplish progress toward the same goals. Thepurpose of this report is not to advance aprescription or legislative agenda. Adopting policyfor the nation as a whole requires balancing ofinterests and consideration of complex social andpolitical forces. This work proves, regardless of thedetails chosen, that our nation’s energy problemsare not intractable, but can be addressedaggressively and well, using policies that carryaffordable costs in return for importantaccomplishments for our nation, our economy andindividual citizens.

The purpose of this part of the report is to deliverthe results of modeling that looks at the outcomeand costs of the combined policies. Next, wereview factors in the five policies that create positivesynergies for particular costs and outcomes. Finally,the report concludes with our overall message.

3.1 Review of policies and goals

Our national energy policy goals are to reduceoil use and imports and to reduce carbon dioxideemissions. The reasons for adopting these goalsand the numeric targets for the reductions areexplained in Part 1. Each of the five policies werecommend contributes in a different degree toeach goal. The policies and their importance foreach goal are explained in Part 2. To help orientthe reader, here is a summary of the policies, withqualitative notes on the importance of each incontributing to each of the two goals.• NEPI Clean Energy Standard: Establishes a 2035goal for clean electrical generation and requirestechnology-neutral, market-driven decisions byutilities to use low-carbon and renewable fuels toachieve this target. Impact: Major reduction of carbon dioxideemissions and other airborne pollutants,insignificant reduction of oil consumption.

• Oil security dividend: Imposes a modest,graduated oil products tax that is fully rebatedthrough the tax system to the public.Impact: Significant reduction of oil consumption,moderate reduction of carbon dioxide emissions,increases economic growth.

• Automotive fuel economy: Follows the CorporateAverage Fuel Economy (CAFE) Standards asadopted in 2012, and considers an alternative‘feebate’ mechanism with similar impact andpotential advantages.Impact: Significant reduction of oil consumption andmoderate reduction carbon dioxide emissions, whichare already built into adopted CAFE Standards.

• LNG trucks: Provides a temporary fuel taxreduction to encourage the nascent shift in theheavy truck market from diesel to liquefied naturalgas (LNG).Impact: Increases an already-expected moderatereduction of oil consumption, with modest reductionin carbon dioxide emissions.

• Energy efficiency: Extends existing incentives forenergy-saving appliances and building codes andhome-based renewable energy.Impact: Significant reduction in carbon dioxideemissions, insignificant reductions in oilconsumption.

50

3.2 Policy benefits and targets

The integrity of NEPI’s process demanded that weadopt numeric targets for our goals of reducing oilimportation and carbon dioxide emissions beforedesigning the policies to meet those goals. In our2010 report with RFF, we screened three dozenpolicies. In this report, we refined our understandingand design of the five most promising policies andmodeled the contribution and cost of each policy.Only at the final stage, in this part of the report, dowe model the combination of policies and comparethe outcome to our targets.

Given this sequential process that holds endsindependent of means, it is not surprising that thecombination of policies does not precisely deliverthe results called for in our targets. As detailed inthe next section, our policies overshoot the targetfor reducing oil consumption but accomplish 85%of the target for reducing carbon dioxide emissions.To some extent, this outcome is a result of ourprocess: we did not adjust the targets to match theresults. But the relative difficulty of achieving thetargets is also a factor.

Since we began our work, the nation’s outlook forreducing dependence on oil, and especially foreignoil, has markedly improved, making that target easierto reach. Much of the reduction in imported oilconsumption came about because of the economicrecession, an influence which presumably will reversewith an economic recovery, but technology andpolicy have also played a part in lowering EIAprojections of future oil use, including adoption ofmuch-improved vehicle fuel efficiency standards, thenew viability of LNG technology for heavy trucks andnew technology for increasing domestic productionof natural gas and oil. (See sidebar “Reduction of OilUse: Imports or Domestic?” below.)

We welcome and incorporate those positivedevelopments as part of our policy, and add more.In light of the reasonable cost of the five mostpromising policies, there would be no reason to pullback because the target has been surpassed.Instead, we would take advantage of the additionalbenefit that is available.

In contrast, the target for carbon dioxideemissions has become more difficult to reach. TheUnited States has seen progress in recent years, anadvance coming both from the reduction of energyuse due to bad economic conditions and becausetechnology has lowered the price of natural gas,leading to displacement of dirty coal-poweredelectricity generation. However, much moreprogress would be needed to avoid falling fartherbehind the scientific and internationally recognizedtarget for carbon dioxide reductions.

Time is a key factor. In this report, we haveadjusted our targets by pushing them farther into thefuture, in part to reflect the time that has passed sinceour 2010 report was released. The details of theseadjustments are beyond the scope of thisexplanation, but it is important to understand theimpact of the passage of time and our 2035 horizonfor analysis, because these issues illuminate afundamental part of the problem we are addressing.

By the nature of our two goals, the time horizondifferently affects the difficulty of attaining the twotargets. The target for reducing oil consumption (andhence imports) reflects a need to cut the amount ofoil used per day. Extending the target into the futurecan make it easier to attain, as policy-drivenchanges have more time to reduce daily oil use. Thetarget for reducing carbon dioxide emissions, on theother hand, reflects a need to cut the cumulativeamount added to the atmosphere, which onlybecomes more difficult as the years pass.

51

Reduction of Oil Use: Imported or Domestic?Oil is a fungible product. All units are interchangeable and consumers neither know

nor care about the source of the oil when they drive up to the gas pump. However,dependence on foreign oil is an issue of national interest and the question of whethera reduction in the use of oil will affect domestic or imported sources is important.

The international oil market is complex but the market demand and supply signals todomestic producers and consumers are generally straightforward albeit, sometimes indifferent directions. The supply of U.S. domestic production is influenced by world oil prices.High prices encourage exploration and production in areas often requiring new technologyand sometimes in remote and difficult locations. The demand for oil by consumers ispredominately based upon the economy, fuel prices at the pump, and fuel efficiencies.

Given these economic forces, the result is that imported oil becomes the marginalsource for market changes. If oil demand (consumption) is reduced the principal effect

will be a reduction of imported oil. Conversely an increase in consumption will initiallybe predominately supplied by imported oil, although, increasing prices may over timeincrease domestic production.

The historical role of imported oil at the margin is seen in the activity of oil flowfrom 2008 to 2011. The demand for oil in the U.S. declined by 0.6 million barrelsper day (mmbd) during this period. The primary cause was the severe recessionbeginning in 2008. However, even with the reduced demand, the production ofdomestic oil actually increased 1.5 mmbd due to new technologies and a reboundingprice of world oil. The consequence of an overall reduced consumption of oil with anincrease in domestic production was imported oil being reduced by 2.7 mmbd. Thedifference between domestic production and total use, whether up or down, is themarginal barrel of imported oil.

We look at a cumulative target for carbondioxide because our emissions are essentiallypermanent contributions to the climate warmingproblem. Each ton of carbon dioxide we emitremains in the atmosphere and drives warming untilit is removed by geologic processes, which act soslowly as to be irrelevant on human time scales.Each year the target is extended is another year thatcarbon dioxide can accumulate and worsen theproblem, making the goal progressively moredifficult to attain. This issue is basic to the climateproblem and its urgency, and so it is not surprisingit should arise in our analysis. Mostly, it underlinesthe need to take action as soon as possible alongthe lines of the policies we have studied.

3.2.1 Policy combination modeling resultsOur overall target was to reduce oil consumption

to 15.2 million barrels per day (mmbd) by 2035.The combination of policies reduces consumption to14.4 mmbd in 2035. This represents a reduction of3.8 mmbd from the 2010 baseline of oil use, anda reduction 2.4 mmbd better than the ReferenceCase. See Figure 3.1.

Our target for reducing energy-related carbondioxide emissions was a cumulative total reductionof 21.7 gigatons of CO2, or 15% of the ReferenceCase cumulative total emissions of 144.6 gigatons.The combination of policies achieves totalcumulative emissions of 126.2 gigatons. This is areduction of 18.4 gigatons, or 85% of our target.See Figure 3.2.

The NEPI Clean Energy Standard (NCES) doesthe heavy lifting in our carbon dioxide emissionscenario, contributing three-quarters of the benefitachieved. Looking at reductions within the NCES,the positive impact closely tracks the retirement ofcoal-fired generation capacity, which is replaced bynatural gas generation. Later in the policy, naturalgas is replaced by nuclear and renewable energy.See Figure 3.3.

52

0

2

4

6

8

10

12

14

16

18

2010 2015 2020 2025 2030 2035

Millio

n b

arr

els

/d

ay

Oil Use Reference, Oil Security Dividend, CAFE Standards, LNG Truck Policy, Combination of Policies, Target


Reference LNG Truck Policy Oil Security Dividend CAFE Standards Combination Target

Figure 3.1

Fig. 3.1 Oil UseReference, Oil Security Dividend, CAFE Standards,LNG Truck Policy, Combination of Policies, Target


0

25

50

75

100

125

150

2010 2015 2020 2025 2030 2035

Gig

ato

ns

of

CO

2-e

Cumulative Total Energy-related CO2-e Emissions Reference, NCES, Energy Efficiency, Combination of Policies Compared to Target 2010-2035 (gigatons of CO2-equivalents)

Reference Energy Efficiency NCES Combination Target

Figure 3.2

Policies targeting oil reductions have cumulative emissions close

to Reference Case: CAFE Standards, Oil Security Dividend,

LNG Truck Policy

3,000

3,500

4,000

4,500

5,000

5,500

6,000

2010 2015 2020 2025 2030 2035

Millio

n m

etr

ic t

on

s o

f C

O2-e

Annual Total Energy-related CO2-e Emissions Reference, NCES, CAFE Standards, Oil Security Dividend,

Energy Efficiency, Combination of Policies 2010-2035 (million metric tons of CO2-equivalents)

Reference Energy Efficiency CAFE Standards Oil Security Dividend NCES Combination

Figure 3.3

Fig. 3.3 Annual Total Energy-related CO2-e EmissionsReference, NCES, CAFE Standards, Oil Security Dividend,

Energy Efficiency, Combination of Policies2010 - 2035


The National Energy Modeling System recognizes these market dynamics in itsenergy forecasting. When we used NEMS-NEPI to model certain policies to reduce oiluse, underlying increases in oil production over time impact imports as well . For instancethe effect of increased automobile fuel efficiency as modeled decreases oil use 2.4mmbd by 2035 relative to 2010. Domestic production increased 1 mmbd over thesame period. Correspondingly, the amount of imported oil projected for this time wasreduced by 3.4 mmbd.

Therefore, we understand that an overall reduction in use of oil can still be consistentwith increased use of domestic production. The balancing factor is the larger reductionof imported oil. The result of this situation is greater energy independence, additionaleconomic benefits from domestic production, and a reduction in carbon dioxide andother pollutant emissions.

Fig. 3.2 Cumulative Total Energy-related CO2-e Emissions

Reference, NCES, Energy Efficiency, Combination of Policies Compared to Target

2010 -2035(gigatons of CO2-equivalents)

3.2.1.1 Effectiveness of combined policies

A significant part of the study’s reduction in oilconsumption results from the positive interaction ofthe combined policies. When the impact of the fivepolicies is merely summed, the oil use reduction isnot as great as when we model the interaction ofthe combined policies.

Much of this improvement results from interactionbetween the Oil Security Dividend and the LNGtruck policy. As we have seen in Section 2.4.5, thedifferential in fuel costs between diesel and LNG isthe primary factor influencing decisions by truckingcompanies as to whether to convert to LNG. The OilSecurity Dividend policy increases this costdifferential, leading to more LNG trucks on the roadand a greater reduction in oil use. This difference issubstantial: our Reference Case projects LNG trucksto achieve 12% market share by 2035; our LNGpolicy by itself increases penetration to 14%; ourcombined policies increase LNG to 20% of the totalheavy truck fleet. See Figure 3.4.

A similar positive interaction occurs between theCAFE Standards policy and the Oil SecurityDividend. The Reference Case, which includes CAFEStandards before the 2012 increases (see Section2.3), contributes a reduction in gasoline use of 9%from 2010 to 2035; adding the Oil Security Dividendnearly doubles that to 17%. The new CAFE Standardsproduce a 25% reduction in gasoline use by the endof the period (the feebate policy is designed toperform identically; see Section 2.3.2). The newCAFE Standards interact strongly with the Oil SecurityDividend, delivering a 32% reduction in gasoline usebetween 2010 and 2035 in the combination of

policies case. These benefits eventually plateau, asthe CAFE Standards become fully implemented in2025. See Figure 3.5.

Several other interesting beneficial impacts also resultfrom the combination of policies, in electricity demand,price, and generation mix, and in natural gas prices.

Two of our policies contribute to reducingelectricity demand. The NCES reduces demand byincreasing price, primarily because it drives aswitch from some coal-powered generation towardsmore expensive nuclear and renewables. Theenergy efficiency policy reduces electricity sales bysaving electricity and by inducing on-site renewablegeneration by users. In the Reference Case,electricity demand increases 18% from 2010 to2035. Taken individually, the NCES cuts thatincrease to 12% and energy efficiency cuts it to10%. Combined, they interact to produce ademand increase of only 4% over the entire 25 yearperiod. See Figure 3.6.

53

0%

5%

10%

15%

20%

2010 2015 2020 2025 2030 2035

% o

f h

eavy t

ruck

fle

et

Natural Gas Heavy Trucks as % of Heavy Truck Fleet Reference, LNG Truck Policy, Combination of Policies

2010-2035

Combination LNG Truck Policy Reference

Figure 3.4

Fig. 3.4 Natural Gas Heavy Trucks as % of Heavy Truck Fleet

Reference, LNG Truck Policy, Combination of Policies2010 - 2035

0

1

2

3

4

5

6

7

8

9

10

2010 2015 2020 2025 2030 2035

Millio

n b

arr

els

/d

ay

Motor Gasoline Use Reference, CAFE Standards, Oil Security Dividend,

Combination of Policies 2010-2035 (million barrels per day)

Reference Oil Security Dividend CAFE Standards Combination

Figure 3.5

Fig. 3.5 Motor Gasoline UseReference, CAFE Standards, Oil Security Dividend,

Combination of Policies2010 - 2035

(million barrels per day)

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

2010 2015 2020 2025 2030 2035

Billio

n k

ilo

watt

ho

urs

Electricity Sales Reference, NCES, Energy Efficiency, Combination of Policies 2010-2035


Reference NCES Energy Efficiency Combination

Figure 3.6

Fig. 3.6 Electricity SalesReference, NCES, Energy Efficiency,

Combination of Policies2010 - 2035


Electricity demand in turn affects price bychanging generation mix. The NCES causes theprice increase noted above because of the highercost of generation, but that increase is moderatedwhen the energy efficiency policy is included,because it lowers demand. The increase in theaverage price of electricity in the NCES by itselfamounts to 21% over the 25-year period. Theenergy efficiency policy, considered separately,holds the price flat. However, when we model thecombination of policies the total increase is limitedto 16%. See Figure 3.7.

The combination of policies also drives moregeneration to renewables than the policies takenindividually. As explained in Section 2.1, theincreasing standard and trading system of theNCES first cause replacement of coal generationby natural gas, and then produce nuclear andrenewable energy at the expense of natural gas.By adding the energy efficiency policy, the powergeneration mix in 2035 has less nuclear energyand an increased percentage of wind, solar andbiomass than in the NCES alone. Two factorsproduce this altered mix. Reduced demanddecreases the use of the more expensivegeneration source, which in this case is nuclearrather than renewables. And the energyefficiency policy provides incentives for end-userenewable generation, enhancing the role ofsolar energy. See Figures 3.8 and 3.9 to see thegeneration mix in each case.

Three policies impact the price of natural gas

through the study period: the LNG truck policy,which tends to increase demand and price fornatural gas; the energy efficiency policy, whichtends to reduce demand and price; and theNCES, which first pushes price higher, and laterhelps bring it down. The NCES has this effectbecause natural gas generation receives creditsin the early years of the policy, but later mustobtain credits as the standard becomes moredifficult to meet. Fitting its role as a bridge fueltoward cleaner energy, gas is prioritized thendiscouraged. The Oil Security Dividend isindirectly involved, too, as its tendency toincrease diesel pump prices helps put more LNGtrucks on the road, further increasing natural gasdemand and price. The combination of policesreflects these interactions, as the price of natural

54

0.0

2.0

4.0

6.0

8.0

10.0

12.0

2010 2015 2020 2025 2030 2035

20

10

cen

ts p

er

kilo

watt

ho

ur

Average Electricity Price Reference, NCES, Energy Efficiency, Combination of Policies 2010-2035

(2010 cents per kilowatt hours)

NCES Combination Reference Energy Efficiency

Figure 3.7

Fig. 3.7 Average Electricity PriceReference, NCES, Energy Efficiency, Combination of Policies

2010 - 2035(2010 cents per kilowatt hour)

45% 38%

16%

37%

18%

24% 27%

34%

27%

34%

20%

18%

29% 18%

25%

10%

15% 20% 17%

23%

0

1,000

2,000

3,000

4,000

5,000

Billio

n k

ilo

watt

ho

urs

Electricity Generation by Fuel Reference, NCES, Energy Efficiency,

Combination of Policies 2010 and 2035 (billions of kilowatt hours)

Other

Renewables

Nuclear Power Natural Gas

Petroleum

Coal

Figure 3.8


Fig. 3.8 Electricity Generation by FuelReference, NCES, Energy Efficiency, Combination of Policies

2010 and 2035(billions of kilowatt hours)

9% 10% 9% 9% 7% 11% 4% 3% 4% 2% 22%

32% 35%

29% 30%

3%

9%

8% 23%

19%

55%

44%

46%

35%

42%

0

100

200

300

400

500

600

700

Billio

n k

ilo

watt

ho

urs

Total Non-hydro Renewables Generation by Source Reference, NCES, Energy Efficiency, Combination of Policies

2010 and 2035 (billions of kilowatt hours)

Wind

Solar

Biomass

Muni Waste

Geothermal

Figure 3.9

Fig. 3.9 Total Non-hydro RenewablesGeneration by Source

Reference, NCES, Energy Efficiency, Combination of Policies2010 and 2035


gas rises above the Reference Case in the earlyyears and falls below it by the end, in 2035.See Figure 3.10.

3.2.1.2 Policy cost of combined policies

In this section and the associated key metricstable (Table 3.1) we report bottom-line costs for thecombination of policies in three metrics: welfarecost, consumer out-of-pocket cost, and federalbudgetary impact. It is worthwhile to note that theeffect on the Gross Domestic Product, ameasurement of the whole economy, is too small toaccurately model in the combination of policies.

Combining the policies significantly reduces thewelfare cost and out-of-pocket costs.

We do not include cost effectiveness, as we dofor individual policies in Part 2—wherein we quotecosts for reductions per unit, using barrels of oil andtons of carbon dioxide emissions. As explained inSection 1.6.3, we have no rationale to allocatethese costs among the two goals. This joint costallocation problem is more manageable whenlooking at individual policies, because the readercan evaluate the relative contribution of each policyto each goal. But if we were to show costeffectiveness figures for the mix of five policies toachieve two goals, the numbers would be difficultto interpret and potentially misleading.Consequently, we report only total costs for theentire policy, not per-unit costs.

Welfare cost (in 2010 dollars) for thecomprehensive policy is $313 billion or less, presentvalue, for the entire 25 years of the study. Amortizedover the study horizon, the cost averages $25.6 billionper year over the 23-years from 2013 to 2035.

As explained in Section 1.6.1.1, we haveconservatively chosen to calculate the welfare costassuming that all decisions in the energy market areoptimal: no credit is taken for policies thatencourage cost-saving behavior that individuals andbusinesses might otherwise overlook. We know this“no market failure” assumption is not true to life,55

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

2010 2015 2020 2025 2030 2035

20

10

$/

millio

n B

tu

Natural Gas Spot Prices (at Henry Hub) Reference, LNG Truck Policy,

NCES, Energy Efficiency, Combination of Policies 2010-2035 (2010 $ per million Btu)

LNG Truck Policy Reference Combination NCES Energy Efficiency

Figure 3.10

Fig. 3.10 Natural Gas Spot Prices (at Henry Hub)Reference, LNG Truck Policy, NCES, Energy Efficiency,

Combination of Policies, 2010 - 2035(2010 $ per million Btu)

Table 3.1 Policy Key Metrics

1Motor gasoline cost is for all light duty vehicles; declines in cost are due to reduced VMT and/or reduced gasoline price (excluding the phased oil tax component)a. oil reductions are insignificant - these policies are directed at CO2 emissions reductionsb. impact on residential electricity and natural gas costs are insignificant - these policies target oil consumptionc. administrative cost only



3.0

1.4

Increment vRef case

a.

0.5

1.0

0.1

a.

2.4

PDVWelfare

Cost

(2010 $,billions)

n/a

307

(117)

196

0

2

313

CostEffectiveness

Oil

(2010 $/barrel)

n/a

n/a

(46)

55

0

n/a

n/a

CumulativeCO2


Case

2010 - 2035(mmt CO2)

21,678

n/a

13,842

1,659

1,754

35

2,060

18,430

PDV of Cumulative

change in Total

Gasoline Cost1 v

Reference Case


n/a

n/a

(186)

(361)

7

n/a

(513)

Policy

Target

Reference Case

NCES

Oil Security Dividend

CAFE Standards

LNG Trucks

Energy Efficiencies

Combination of Policies

CostEffectiveness

CO2

(2010 $/ton CO2)

n/a

22

(71)

112

0

1

n/a


n/a

122

b.

b.

6

(182)

(94)


n/a

(6)

b.

b.

3

(41)

(44)



n/a

c.

c.

c.

(5)

(82)

(120)


in 2035

Key Metrics Table

especially with respect to investments in moreenergy efficient autos and buildings, whichconsumers tend to under-value compared to theirtrue economic worth. Since we cannot quantify suchmarket failures, we have chosen to ignore thosesavings, a decision that results in reporting a higherwelfare costs for the policies. The decisiveadvantage of this approach is analytical clarity. Byremoving the variable of market failure, we narrowour view to the benefit of addressing broad societalexternalities through a mix of complementarypolicies that are measured on a single basis.However, as a result, our reported cost of $313billion—while reasonable in scale and well withinthe capacity of the economy to absorb—does notrepresent the cost we actually expect to be incurred,but instead represents an upper boundary of cost.

The synergistic combination of the five policiessignificantly reduces the welfare cost of the entirepackage. By simply adding the welfare cost of eachpolicy without looking at their interactions, weobtain a total of $387 billion. The combination cutsthat cost by $74 billion, or 19%.

This reduction is caused mainly by savings in theCAFE Standards policy. Gillingham’s analysis of thispolicy includes a costly rebound effect, whichaddresses the expectation that drivers with more fuel-efficient vehicles will drive more, creating trafficcongestion and other social costs. This rebound effectdisappears in the combination of policies, becausethe Oil Security Dividend taxes fuel sufficiently tooffset drivers’ savings from owning more efficientvehicles. Without a reduction in the cost of driving,there is no increase in the miles of travel, and nosocial costs for additional congestion, trafficaccidents, etc. Under the dividend concept, alltaxpayers receive the oil tax funds back through theirincome taxes, so that policy carries no welfare cost(indeed, as seen in Section 2.2.1, the policy’s welfarecost is negative, because it reduces inefficiencies inthe tax system). Out-of-pocket costs are partlyaddressed in the discussion of electricity, natural gas,gasoline and diesel in Section 3.2.1.1.

At the end of the study period, in 2035, annualresidential electricity expenditures have decreasedby $89 per household compared to the ReferenceCase. The upward pressure on power prices causedby the NCES is more than offset by savings from the

energy efficiency policy, which drives down usageand helps reduce prices by lowering demand.

The cost of natural gas also goes down, with areduction of $37 per household compared to theReference Case. Savings in both price and demandcaused by the energy efficiency policy offset theupward price pressure caused by LNG trucks usingmore natural gas.

Combining policies does not reduce federal costsfor the energy policy. However, the numbers are small.Over the 25 years of the policies, federal tax revenuesare reduced $120 billion (in 2010 dollars) due to taxsubsidies in the combination of policies case.

3.3 Conclusions

NEPI has presented a coordinated portfolio offive energy policies that, by 2035, reduce carbondioxide emissions by a cumulative 18.4 gigatonsand reduce oil use by 2.4 mmbd compared to theReference Case, and do so at a total welfare costof $313 billion or less, present value. As seen inthe previous paragraph, federal tax subsidies,primarily credits for renewable energy, are $120billion. Consumer out-of-pocket costs for electricityand natural gas decline, federal administrativeexpenses are insignificant, and impact on GDP istoo small to accurately model.

The rubric of cost hardly fits our Oil SecurityDividend policy, a gradually increasing oil tax thatis fully rebated to Americans through the income taxsystem. Its cost is less than zero, because it yieldssignificant economic benefits even withoutconsidering its impact on energy. By raising revenuesmore efficiently, the policy boosts the economy by$117 billion over the period of the study. Thisnegative “cost” buys a policy that interacts positivelywith the other policies in our portfolio, significantlycutting their cost and increasing their effectiveness.Consequently, the total cost of the comprehensivepolicy is considerably lower.

These costs are affordable, but the cost ofinaction is not. Natural processes will notappreciably scrub our carbon dioxide emissionsfrom the atmosphere within the lifetime of anyonenow living. Each ton emitted by a power plant orvehicle is essentially a permanent contribution toa fundamentally altered planet. The business-as-

56

usual Reference case (Figure 3.11) shows that thetemporary, recession-driven annual carbon-dioxide reduction which began in 2008 willresume its multigenerational increase startingaround 2014. Only with the Combination ofPolicies is this trend reversed sufficiently to put theU. S. on the right track to accomplish nationalgoals. We cannot know with precision what thisnew planetary environment will look like, but wecan get an idea from recent events, such asSuperstorm Sandy and the ongoing nationaldrought. Patchy, sudden and potentiallycatastrophic changes can be expected toaccumulate as carbon dioxide accumulates in theair. The policy costs we cite in this study can beconsidered a small insurance premium to reducethese unknown hazards.

America’s need to break our addiction to oil is agoal of a different kind, but also critically importantfor the economic well-being and national securityof the nation. The modest price we pay for thiscomprehensive policy can fairly be called a downpayment for peace and stability, buying a measureof safety for service personnel who defend currentU.S. supplies of oil, and a measure of confidencefor businesses wary of supply and price shocks.Action is necessary despite the current, temporarydecline in oil imports. The Combination of Policiescontinues to decrease overall consumption of oilwhile domestic production actually increases. As aresult, imported oil use is driven down to the lowestlevel in 50 years. (Figure 3.12). In addition, theoverall reduction of oil consumption is an important

contributor to the reduction of CO2 emissions.We willlikely regret it later if we miss the current opportunityto fundamentally change the transportation andindustrial use of oil. Failure to set in place themechanisms for gradually achieving such a changewill only make the day of reckoning more painful.

These twin primary goals are compellingenough, but our subsidiary goals are important aswell. This comprehensive policy would saveAmericans’ lives and health by cleaning the air ofhazardous pollution. It would also maintain thefreedoms and prerogatives of reasonably pricedenergy for us all. And these are changes that canreally happen. We do not have to believe inpolitical miracles to envision its adoption in part orin whole, nor do we have to pin our hopes on atechnological silver bullet not yet discovered.

The most efficient single policy to reach ourgoals would be a single pricing methodology suchas a broad-based carbon tax or cap-and-tradescheme, but political events of recent years havedemonstrated that this is not a realistic prospect.Our combined policies instead deliver similarbenefits at an affordable cost and with greaterpolitical palatability. Our second most expensivepolicy, CAFE Standards, is already on the booksand needs only to be implemented as adopted toproduce all the benefits we call for. The energyefficiency policy merely calls for extension ofexisting policies, and the LNG truck policyrequires no more than a temporary LNG tax breakfollowed by tax parity among fuels. The OilSecurity Dividend costs nothing in the aggregate.

57

0

5

10

15

20

25

30

Millio

n b

arr

els

/d

ay

U.S. Consumption, Production and Imports of Petroleum and Other Liquids

1949-2011 Actual; 2012-2035 NEPI Combination of Policies (million barrels per day)

Start of Policy Consumption Production Imports

Historical data source: EIA Projections: NEPI Combination of Policies

Figure 3.12

Fig. 3.12 U.S. Consumption, Production and Imports ofPetroleum and Other Liquids

1949-2011 Actual; 2012-2035 NEPI Combination of Policies (million barrels per day)

-

1,000

2,000

3,000

4,000

5,000

6,000

7,000

Millio

n m

etr

ic t

on

s o

f C

O2-e

Annual Total Energy-related CO2-e Emissions

Actual, Reference Case, NEPI Combination of Policies 1949 -2035 (million metric tons of CO2-equivalents)

Actual Reference Combination of Policies

Figure 3.11

Historical data source: EIA; Projections: NEPI

Fig. 3.11 Annual Total Energy-related CO2-e EmissionsActual, Reference Case, NEPI Combination of Policies

1949 - 2035(million metric tons of CO2-equivalents)

58

Indeed, by recycling oil tax revenues, it makes thenation’s system of taxation simpler, saving moneyover all. Moreover, while the policy increases theretail price of fuel, the change comes so slowly asto be unnoticeable. The NCES would produceAmerica’s biggest contribution to addressingclimate change and significantly decreasing otherhealth hazardous emissions. The fundamentalstructural change in our generation of electricity,from a coal-dominated fuel mix to substantiallymore low carbon and renewable fuels, is shownin Figure 3.13. This will be accomplished withinan economic framework of market-driven carbondioxide credit trading among utilities, providinga least-cost pricing approach rather than aregulatory mandate.

No doubt an actual national energy policy thatis adopted will contain different policy ingredientsin a mix that will come into focus through thepolitical process. Our goal has not been to provide

a recipe to Congress. Instead, we have shown that,for reasonable cost, America has within its graspsolutions to energy problems that threaten thenational security, our economy, and the environmentthat sustains us.

45% 38%

18%

24% 27%

34%

20%

18%

25%

10%

15%

23%

0

1,000

2,000

3,000

4,000

5,000

2010 Reference 2035

NEPI Combination of Policies 2035

Billio

n k

ilo

watt

ho

urs

Electricity Generation by Fuel Reference Case, NEPI Combination of Policies

2010 and 2035 (billions of kilowatt hours)

Other

Renewables

Nuclear Power Natural Gas

Petroleum

Coal

Figure 3.13


Fig. 3.13 Electricity Generation by FuelReference Case, NEPI Combination of Policies

2010 and 2035(billions of kilowatt hours)

59

Appendix

The findings in this study draw on various technical and background papers commissioned by the NationalEnergy Policy Institute as part of this project. These papers are or will be available on the NEPI website(http://nepinstitute.org/publications/).

• Consequences of U.S. Dependence on Foreign Oil. Stephen P. A. Brown and Ryan T. Kennelly (Center for Business and Economic Research, University of Nevada, Las Vegas)

• Small Modular Reactors: Costs, Waste and Safety Benefits. Geoffrey Rothwell (Department of Economics, Stanford University)

• The Economics of Fuel Economy Standards versus Feebates. Kenneth Gillingham (Yale University)

• A Review of the Impacts of an ALL CLEAN Clean Energy Standard for Selected Regions and States. Mark A. Foster, MAFA

• Oil Security Dividend: Research on the economic effect of a federal oil tax and the consequences of recycling of funds using various methods. Roberton C. Williams (University of Maryland)

• What Set of Conditions Would Make the Business Case to Convert Heavy Trucks to Natural Gas? – a Case Study. Anna Lee Deal (Lynden Incorporated)

www.nepinstitute.org

June 2013...Doreen Griffiths, NEPI Administrative Director Charles Wohlforth, author and journalist,...

Documents

Transcript of June 2013...Doreen Griffiths, NEPI Administrative Director Charles Wohlforth, author and journalist,...