Ohio Energy
Luncheon Sponsored by FirstEnergy Solutions
Luncheon Keynote:
Competitive Electricity Markets
Spur Economic Growth &
Create Jobs
Jonathan A. Lesser, Ph.D., President, Continental Economics, Inc., Sandia Park, NM
Wednesday, February 20, 2013
12:45 p.m. to 1:30 p.m.
JONATHANA.LESSER,PHD
Dr.JonathanLesseristhePresidentofContinentalEconomics,Inc.,andhas
almost30yearsofexperienceworkingforregulatedutilities,government,
and as an economic consultant. He has analyzed economic, policy, and
regulatoryissuesaffectingtheenergyindustry,includingmarketstructure
anddesign,environmentalandregulatorypolicy,renewableenergypolicy,
economicimpactsofenergyinvestment,andutilityfinancingandthecost
ofcapital.
Dr. Lesser has extensive experience in valuation and damages analysis,
from estimating the damages associated with breaking commercial leases to valuing nuclear
powerplants.Dr.Lesserhasperformedduediligencestudiesforinvestmentbanks,testifiedon
generating plant stranded costs, assessed damages in commercial litigation cases, and
performed statistical analysis for class certification. He has also served as an arbiter in
commercialdamagesproceedings.
Dr.Lesserhasprovidedexperttestimonyonmanyelectricandnaturalgasregulatoryissueson
thecostofcapital,costofservice,costallocation,ratedesign,anddepreciationincasesbefore
state utility commissions; before the Federal Energy Regulatory Commission (FERC); before
internationalregulatorsinLatinAmericaandtheCaribbean;aswellastestifiedincommercial
damages cases in U.S. federal and state courts. He has also testified on energy policy and
marketdesignbeforestatelegislativecommittees.Dr.Lesserhasalsoservedasanindependent
arbiter in disputes involving regulatory treatment of utilities and valuation of energy
generationassets.
Dr.Lesseristheauthorofnumerousacademicandtradepressarticles.Heisalsothecoauthor
of Environmental Economics and Policy, published in 1997 by Addison Wesley Longman,
Fundamentals of Energy Regulation, published in 2007 by Public Utilities Reports, Inc., and
Principles of Utility Corporate Finance, published in 2011 by Public Utilities Reports, Inc. Dr.
Lesser is also a contributing columnist and Editorial Board member for Natural Gas &
Electricity.
6 Real Place • Sandia Park, NM 87047 • main: 505.286.8833 • DC Office: 202.446.2062 www.continentalecon.com www.competecoalition.com
STATE SUBSIDIZATION OF ELECTRIC GENERATING PLANTS AND THE THREAT TO WHOLESALE ELECTRIC COMPETITION Report prepared for: COMPETE Coalition Prepared by: Continental Economics, Inc. December 2012 Copyright © 2012, Continental Economics, Inc.
The information contained in this document is the exclusive, confidential and proprietary property of Continental Economics,
Inc. and is protected under the trade secret and copyright laws of the U.S. and other international laws, treaties and conventions.
No part of this work may be disclosed to any third party or used, reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying and recording, or by any information storage or retrieval system, without prior
express written permission of Continental Economics, Inc.
TABLEOFCONTENTS
ExecutiveSummary
I.
Introduction....................................................................................................................................1
II. WhyCapacityMarkets?...............................................................................................................3
A. CapacityMarketOpposition..................................................................................................................5
III. HowGovernmentSubsidizationofGenerationHarmsMarketsandConsumers..6
A. BuyerMarketPower—NoSuchThingasaFreeLunch............................................................7
B. TheBriggs‐KleitModel:EffectsofGovernmentSubsidizedEntry
intoCapacityMarkets...............................................................................................................................8
IV. ConclusionsandPolicyImplications....................................................................................10
Appendix1:TheBriggs‐KleitMultipleMarketModel
Appendix2:ReviewofElectricRestructuringandCreationofCompetitiveWholesale
Markets
Appendix3:TheNewJerseyandMarylandPrograms
EXECUTIVESUMMARY
“Theexpectationofgovernmentinterventioncreatesaself‐fulfillingconditionthatincreases
costsanddrivessuppliersfromthemarket”
Competitive wholesale electricity markets in the U.S. have shown themselves to be an
economic success. Today, more than 40% of total generating capacity is owned and
operatedbyindependent,competitivegenerators.1And,unlikemonopolyelectricutilities,
whose ratepayers must bear the financial risks of investment decisions, competitive
generatorsbearthoserisksthemselves.
Because electricity cannot be stored cost‐effectively, ensuring reliable electric service
requires that reserve capacity be available to meet unforeseen events, such as sudden
increasesindemandorunexpectedgeneratoroutages.Toprovidethesecapacityreserves,
a number of competitive market solutions for capacity have been developed. One such
solution is a forward capacity market that uses an auction to secure commitments for
sufficientcapacityresourcesintothefuture.ThePJMInterconnection(thegridoperatorin
themid‐Atlanticstates)administersthelargestsuchmarket,whichiscalledtheReliability
Pricing Model (RPM). The goal of capacity markets like RPM is to provide economic
incentivesthatensurethereisenoughcapacityneededtomeetreliabilitystandardsatthe
lowestpossiblecost.2
Some opponents have argued that capacity markets are not working properly because
higher capacity prices in their states—typically, the result of sub‐regional transmission
constraints—havenotincentedtheconstructionofnewcapacityresources.Consequently,
capacity market opponents conclude that direct government intervention in capacity
markets is required to develop local generating resources, despite the regional nature of
theinterstatewholesalepowermarkets.3
This paper examines the impact of such intervention in the form of mandated state
government subsidies for new generation resources. Using the results of recently
published work by the Pennsylvania State University Electricity Markets Initiative,4 we
concludethatgovernmentsubsidiesfornewgenerationresourcesbothraisecapacitycosts
for the very customers whom the subsidies are supposed to benefit and jeopardize
resourceadequacyandreliabilityinthelongrunforallconsumers.
1
Source:U.S.EnergyInformationAdministration,ElectricPowerAnnual2010,November2011,Table1.3.
2
ForanintroductiontoRPM,seeCOMPETECoalition,“KeepingtheLightsOntoPowerOurFuture:“RPM”,
PJM’sReliabilityProgram,”March2012.
http://www.competecoalition.com/files/Keeping%20the%20Lights%20On%20to%20Power%20Our%
20Future%20‐%20FINAL%2031212.pdf.
3
See,forexample,2011NewJerseyEnergyMasterPlan,pp.21‐22.
4
R.BriggsandA.Kleit,“ResourceAdequacyandtheImpactsofCapacitySubsidiesinCompetitive
ElectricityMarkets,”WorkingPaper,Dept.ofEnergyandMineralEngineering,PennsylvaniaState
University,October22,2012(BriggsandKleit).
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2165412
EX‐1
The new work by Professors R.J. Briggs and Andrew Kleit finds that consumers in states
wheresubsidiesaremandatedreceivelessoftheinitial“benefits”fromstategovernment
intervention in capacity markets than customers in surrounding states. However, these
“benefits” quickly disappear, as government intervention drives out otherwise economic
existinggenerationandhindersthedevelopmentofnewresourcesinallstateswithinthe
market.Thus,whengovernmentintervenesonbehalfofonegeneratoritdrivesoutother
generators, taking with it not only competitive generation capacity, but also the jobs and
taxbaseassociatedwithgenerationthatexitsthemarket.Mostimportantly,theyfindthat
the adverse long‐run impacts in all states far outweigh any short‐term “benefits” of
temporary price reductions. Their work demonstrates that mandates by state
governmentstosubsidizeinvestmentingenerationresourcesareneveroptimalandresult
inlossestoallconsumers.
MostU.S.electricmarketstypicallyextendbeyondindividualstates,operatingregionallyto
achieve economies of scale and scope. As a result, individual state government
intervention, like the recent government actions in Maryland and New Jersey,5 imposes
long‐run costs on neighboring states. Such “beggar‐thy‐neighbor” policies are not only
counterproductive, they also invite policy “retribution” that will further damage
competitivemarkets.Forexample,notingthatsignificantnewgenerationinvestmenthad
taken place in Pennsylvania, that state’s public utility commissioners expressed concern
that
[t]he ability to bid in new capacity at potentially artificially low prices can
skew the capacity market leading to less investment in new and existing
capacity,includinginPennsylvania.Withoutsuchinvestment,theendresult
from the consumer's perspective, ultimately, could be higher rates in
Pennsylvaniathanwithoutthisstate‐mandatedsubsidy.6
It is a basic economic fallacy that price distortions caused by subsidies from government
interventioninafreemarketare“benefits.”Therealityissuch“freelunch”policiesnever
work,becausetheyincorrectlyassumethatsupplierswhoseelowerpriceswillnotchange
theirbehavior.
Perversely, when a state government intervenes in a regional capacity market so as to
benefitconsumers,itactuallyforcesthosesamein‐stateconsumerstosubsidizeshort‐run
benefitsforneighboringstates’consumers.Oneconsequenceofsuchinterventionisthat
merchantgenerationwillnotinvest,eitherinnewresourcesorexistingresources,forfear
that governments will intervene and eliminate the ability to realize a needed economic
return on the investment that compensates developers for the risks they take. In other
words, even the expectation of government intervention createsa self‐fulfilling condition
thatincreasescostsanddrivessuppliersfromthemarket.Thisharmsallconsumersand,
in turn, will harm long‐run grid resource adequacy. Eventually, the only suppliers in a
market will be subsidized ones, and the market will cease to exist, eliminating the real
benefitsthatcompetitivemarketsprovide.
5
Appendix3providesadescriptionofthegenerationsubsidyprogramsinMarylandandNewJersey.
6
LettertothePennsylvaniaCongressionaldelegationfromthePennsylvaniaPublicUtilityCommissioners,
July13,2011.
EX‐2
StateSubsidizationofElectricityGeneration
I.
December2012
Introduction
CompetitivewholesaleelectricitymarketsintheU.S.havebeenaneconomicsuccess.With
passageoftheEnergyPolicyActof1992,Congressunleashedcompetitiveelectricmarkets,
breaking the virtual monopoly on utility‐owned and operated generation.1 Today, more
than40%oftotalgeneratingcapacityisownedandoperatedbyindependent,competitive
generators.2 Competitive generators have also improved their reliability and operating
efficiency more than their fully‐regulated, utility counterparts.3 And, unlike monopoly
electric utilities, whose ratepayers bear the financial risks of investment decisions,
competitivegeneratorsbearthoserisksthemselves.
Because electricity cannot be stored cost‐effectively, the assurance of reliable electric
service requires that reserve capacity be available to meet unforeseen events, such as
sudden increases in demand or unexpected generator outages. To provide such capacity
reserves, a number of competitive markets for capacity have been developed. Unlike
competitive wholesale electric energy markets, whose prices fluctuate each hour,
competitivecapacitymarketsarelonger‐terminnature.PJMInterconnection4operatesthe
largest such market called the Reliability Pricing Model (RPM). The goal of capacity
marketslikeRPMistoprovideeconomicincentivesthatensurethereisenoughcapacityto
meetreliabilitystandardsatthelowestpossiblecost.5IndependentreviewsofRPMhave
found that it has worked well, adding over 13,000 Megawatts (MW) of new generation
since2007.6
Someopponentsofforwardcapacitymarketshavearguedthattheyprovideawindfallfor
generatorsattheexpenseofcustomers,especiallyforbaseloadgeneratorslikenuclearand
coal‐fired power plants, without providing a more reliable electric system. In essence,
1
Appendix2providesabriefhistoryofelectricrestructuring.
2
Source:U.S.EnergyInformationAdministration,ElectricPowerAnnual2010,November2011,Table1.3.
3
Seee.g.,P.Joskow,“MarketsforPowerintheUnitedStates:AnInterimAssessment,”TheEnergyJournal
27(January2006),pp.1‐36.http://econ‐www.mit.edu/faculty/download_pdf.php?id=1219;N.Rose,K.
Markiewicz,andC.Wolfram,“DoesCompetitionReduceCosts?AssessingtheImpactofRegulatory
RestructuringonU.S.ElectricGenerationEfficiency,”MassachusettsInstituteofTechnology,Centerfor
EnergyandEnvironmentalPolicyResearch,04‐418WP,November2004.
http://web.mit.edu/ceepr/www/2004‐018.pdf.
4
PJMInterconnectionisthelargestregionaltransmissionorganizationintheUS,covering13statesand
theDistrictofColumbia.
5
ForanintroductiontoRPM,seeCOMPETECoalition,“KeepingtheLightsOntoPowerOurFuture:“RPM”,
PJM’sReliabilityProgram,”March2012.
http://www.competecoalition.com/files/Keeping%20the%20Lights%20On%20to%20Power%20Our%
20Future%20‐%20FINAL%2031212.pdf.
6
TheBrattleGroup,“SecondPerformanceAssessmentofPJM’sReliabilityPricingModel,”August26,2011.
http://www.pjm.com/~/media/committees‐groups/committees/mrc/20110818/20110826‐brattle‐
report‐second‐performance‐assessment‐of‐pjm‐reliability‐pricing‐model.ashx.
‐1‐
StateSubsidizationofElectricityGeneration
December2012
these opponents believe (wrongly) that they had previously been receiving capacity “for
free,”andthatpayingforcapacitythroughaformalmarketmeansthattheyjustpaymore.
Of course, that was never true. Before competitive electric markets existed, utilities still
maintainedcapacityreserves,whichratepayerspaidforintheirmonthlybills.Thecostof
capacitywasneithertransparentnorsubjecttocompetitivemarketforces,butitcertainly
existed.
Other opponents have argued that formal capacity markets are not working properly.
Theseopponentsarguethathighercapacitypricesintheirstates—typically,theresultof
transmission constraints caused by higher costs within the various regions—have not
incented the construction of new capacity resources. Consequently, they conclude that
directgovernmentinterventionincapacitymarketsisrequired.7
Insomestates,suchasMarylandandNewJersey,thisinterventionhastakentheformof
requiring local distribution (i.e., “poles and wires”) utilities to either build generating
capacity themselves, or auction off the right to build new capacity to independent
developers.8 Whatever the specific form of intervention, the costs of the newcapacity in
these state‐subsidized efforts are underwritten by the local “poles and wires” utility’s
customers,thuseliminatingthenormalfinancialriskthatcompetitivegenerationsuppliers
bear. (In fact, allocating financial risk back to generation suppliers was one of the key
purposesofelectricindustryrestructuring.)Thenewgeneratingcapacityisthenoffered
into the capacity andenergy markets so as to purposefully “suppress” market prices and
supposedly“benefit”consumers.
However, it is a basic economic fallacy that price distortions caused by government
subsidies in a free market are “benefits.” The lower prices made possible by subsidized
entry are short‐lived because they drive competitive suppliers from the market. For
example,inanOctober2011hearingbeforetheNewJerseyBoardofPublicUtilities,Zamir
Rauf, the Chief Financial Officer of Calpine Corporation—the largest developer of
independentlyownedandoperatedgenerationinthecountry—testifiedthat:“[O]neofthe
biggest risks we currently face is the regulatory uncertainties created by various states'
interest in trying to jump‐start the process of developing new capacity via long‐term
contracts. … [i]t will be exceedingly difficult for merchant projects to compete with
ratepayersubsidizedprojects.”9
Inthisreport,wefocusontheeffectswhengovernmentsmandateconsumerpaidsubsidies
ofgeneratingcapacitythatcanbeusedtosuppresspricesincapacitymarkets.Thereality
is that such policies never work: they are a form of ‘free lunch’ economics that fails to
account for market dynamics. This report provides a non‐technical introduction to the
7
See,forexample,2011NewJerseyEnergyMasterPlan,pp.21‐22.
8
Appendix3providesadescriptionofthegenerationsubsidyprogramsinMarylandandNewJersey.
9
IntheMatteroftheBoard’sInvestigationofCapacityProcurementandTransmissionPlanning,DocketNo.
E011050309,NewJerseyBoardofPublicUtilities,Transcript,October14,2011,pp.135‐136.
‐2‐
StateSubsidizationofElectricityGeneration
December2012
important work of Briggs and Kleit (2012),10 who demonstrate how such subsidized
capacity actually imposes the greatest economic harm on intervening states’ own
electricityconsumersandleadstohigherlong‐runelectricitypricesforconsumersinthese
statesandneighboringones.
Briggs and Kleit demonstrate that the adverse long‐run impacts in both markets far
outweigh any short‐term “benefits” of temporary price reductions. They show that even
theexpectationofgovernmentinterventioncreatesaself‐fulfillingconditionthatincreases
costsanddrivessuppliersfromthemarket.Eventually,theonlysuppliersinamarketwill
besubsidizedones,andthemarketwillceasetoexist,harmingallconsumersandlong‐run
grid resource adequacy. This violates one of the founding tenets of electricity
restructuring: less government intervention on supply and price, and more reliance on
market signals to attract investment. As Briggs and Kleit conclude, “[i]n the context of
restructured electricity markets, subsidized additions of base capacity by state
governments are, at best, a costly and undue burden for taxpayers. At worst, these
subsidieshavetheperverseeffectofreducingtheincentivesforresourceadequacyinthe
longrun.”11
II.
WhyCapacityMarkets?
As discussed previously in the introduction, because electricity cannot be stored cost‐
effectively, ensuring the reliability of the power system requires there to be reserve
marginstoaddresscontingencies.Inregionswhereelectricmarketswererestructuredto
allowmarketincentivestoimproveoperatingefficiency,betterallocatefinancialrisk,and
reduce costs, the old regulatory model of having vertically integrated utilities meet their
reserve requirements by building their own generating resources no longer applies.
Instead,restructureddistributionutilitiesbecame“polesandwires”companiesthatobtain
generationfromcompetitivewholesaleelectricmarkets.
Inoverseeingwholesaleelectricmarkets,FERChassoughttoensurethatgeneratingunits
needed primarily for maintaining reserve margins could remain economically viable.
Beforerestructuring,suchunitswereownedbyutilitiesandpaidforbyratepayers.Thus,
post‐restructuring, a market mechanism was needed to ensure these same types of
generatingresourcescouldremaineconomicallyviableeveniftheywererarelyusedand
thusdidnotroutinelyearnsufficientrevenuesfromwholesaleelectricenergymarkets.
However, FERC also imposed price caps in wholesale electric energy markets to address
potential market‐power concerns, especially when electricity demand peaked. But these
10
R.BriggsandA.Kleit,“ResourceAdequacyandtheImpactsofCapacitySubsidiesinCompetitive
ElectricityMarkets,”WorkingPaper,Dept.ofEnergyandMineralEngineering,PennsylvaniaState
University,October22,2012(BriggsandKleit).
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2165412Appendix1providesamoredetailed
reviewoftheBriggsandKleitmodel.
11
BriggsandKleit,p.27.
‐3‐
StateSubsidizationofElectricityGeneration
December2012
two goals—sufficient electric supplies to ensure reliability12 and price caps to curb
potentialmarketpowerabuses—resultingeneratorsnotrecoveringtherevenuesthatthey
needtocontinueoperating.Thisissue,oftencalledthe“missingmoney”problem,hasbeen
well‐documented.13
Thesecondproblemthataroseisthatelectricsystemreliabilityhascharacteristicsofwhat
economists call a “public good.”14 Because reliability for one is really reliability for all,
thereistoolittleincentiveforsupplierstoinvestfullyinresourceadequacy,andthereisan
incentiveforend‐useelectricityconsumersto“free‐ride”bynotpurchasingthefullamount
ofreliability.
Atleasttwobasictypesofpoliciescanaddressthereliabilityproblem.First,regulatorscan
raisethepricecaponwholesaleenergypricestothefulleconomicvalueofgenerationat
the peak.15 This is an “energy only” solution. The Electricity Reliability Council of Texas
(ERCOT)16 has adopted this approach. The Public Utility Commission of Texas, which
overseesERCOT,recentlyvotedtoraisethecaponenergypricesto$9,000permegawatt‐
hour(MWh)byJune1,2015.17Incomparison,FERChascappedwholesaleelectricmarket
offersat$1,000perMWhinPJM.18
12
Therearereallytwodifferenttypesofreliability.Thefirstis“systemsecurity,”whichdealswith
ensuringtheelectricsystemcanmeetminute‐to‐minutechangesinelectricdemandandcompensatefor
unforeseenevents,suchassuddenlossesoftransmissionlinesandgenerators.Thesecondis“resource
adequacy,”whichaddressesthelong‐termneedforresourcestomeetfutureelectricdemand.Thefocus
ofcapacitymarkets—andthisreport—isthelatter.
13
See,e.g.,R.Shanker,“CommentsonStandardMarketDesign:ResourceAdequacyRequirement,”Federal
EnergyRegulatoryCommission,DocketRM01‐12‐000,January10;W.Hogan,“OnanEnergyOnly
ElectricityMarketDesignforResourceAdequacy”FederalEnergyRegulatoryCommission.
http://www.ferc.gov/EventCalendar/files/20060207132019‐hogan_energy_only_092305.pdf
14
SeeJ.LesserandG.Israilevich,“TheCapacityMarketEnigma,”PublicUtilitiesFortnightly,December2005,
pp.38‐42.Seealso,P.CramtonandS.Stoft,“ACapacityMarketthatMakesSense,”TheElectricityJournal
18(August2005),pp.43‐54.
15
Inthelimit,thehighestvalueofelectricityisthevalueoflostload(VOLL),orwhatelectricityconsumers
wouldbewillingtopaytoavoidaforcedoutage.
16
ForadiscussionofresourceadequacyinERCOT,seeS.Newell,K.Spees,J.Pfeifenberger,R.Mudge,M.
DeLucia,andR.Carlton,“ERCOTInvestmentIncentivesandResourceAdequacy,”TheBrattleGroup,June
2012.
http://www.ercot.com/content/news/presentations/2012/Brattle%20ERCOT%20Resource%20Adequ
acy%20Review%20‐%202012‐06‐01.pdf.
17PUCRulemakingtoAmendPUCSubst.R.25.505,RelatingtoResourceAdequacyintheElectricReliability
CouncilofTexasPowerRegion,OrderAdoptingAmendmentsto§25.505asApprovedattheOctober25,
2012OpenMeeting,October25,2012.ThePUCTincreasedthepricecapto$5,000perMWh,effective
June1,2013;$7,000perMWh,effectiveJune1,2014;and$9,000perMWh,effectiveJune1,2015.
http://www.puc.texas.gov/industry/projects/rules/40268/40268adt.pdf.
18
NewlyadoptedruleswithinPJMforscarcitypricingwilleventuallyallowthelevelofcappedpricestorise
to$2,700perMWh.
‐4‐
StateSubsidizationofElectricityGeneration
December2012
Second,regulatorscanrequireload‐servingentities(LSEs),whetherlocalelectricutilities
orcompetitiveelectricityproviders,tocontractintothefuture(orforward)foraccessto
specified amounts of generating capacity, such that each LSE will be individually reliable
andthesystemwillbereliableasawhole.19Thesecontractsmaybenegotiatedbilaterally
ortheirpricemaybedeterminedusingacentralizedmarket‐clearingmechanism,hence,a
“capacitymarket.”20
Althoughtheyhaveevolvedovertime,theforwardcapacitymarketsthatareoverseenby
ISONewEngland(ISO‐NE),theNewYorkISO(NYISO),andPJMallshareacommondesign.
Specifically,theyallusemarketmechanismsthataredesignedtosendthemarketsignals
needed to incent developers to construct additional generating capacity (or supply
alternativeformsofcapacity,suchasdemandresponseresources,whichareagreementsto
curtail electric usage when called on by the RTO), and thus ensure the “desired” level of
system reliability over time.21 Retail electric utilities and competitive firms who supply
electricitytoretailcustomersarerequiredtoobtainaminimumquantityofcapacity,soas
toensuretherearesufficientreservestomeetpeakdemand.
A.
CapacityMarketOpposition
To be sure, the capacity markets operated by PJM, NYISO, and ISO‐NE have vocal critics.
Thesecriticismsfallintotwomainareas:
1. Capacity markets provide “windfall” profits to existing generators, especially
baseloadgeneratorsforwhichratepayershavealreadypaid;and
2. Capacity markets are not incenting new generating resource development in
constrained regions, despite capacity prices that are higher than the overall RTO
prices.
Althoughthefirstargumenthasbeenraisedinthepast,especiallywhennaturalgasprices
weremuchhigherthantheyaretoday,ithasbeeneffectivelyrefutedandisrarelyraised
today.22 Instead, it is the second argument that capacity market opponents now raise.
19
ForadiscussionofhowthePJMcapacitymarketisimplemented,seeJ.Chandley,“PJM’sReliability
PricingMechanism:(WhyIt’sNeededandHowItWorks),”2008.
http://www.pjm.com/documents/~/media/documents/reports/pjms‐rpm‐j‐chandley.ashx
20Foracomparisonofbilateralcapacitycontractsandcapacitymarkets,seeP.CramtonandS.Stoft,“Why
WeNeedtoStickwithUniform‐PriceAuctionsinElectricityMarkets.”TheElectricityJournal20(January
(2007),pp.26‐37(CramtonandStoft2007).
21
PJMandtheNYISOuseadministrativelydetermined,downwardslopingcapacitydemandcurvesandan
auctioninwhichsuppliersofferincapacityresources.ISO‐NEusesaverticaldemandcurvesettothe
installedcapacityrequirementandwhatiscalleda“descendingclockauction,”inwhichsuppliers
respondtodescendingpricesuntilthecapacitysuppliedequalstheinstalledcapacityrequirement.The
NYISOIndependentMarketMonitorhascalledforNYISOtoadoptasimilardownward‐slopingdesign.
SeePotomacEconomics,2011AssessmentoftheISONewEnglandElectricityMarkets,June29,2012,pp.
117‐121.http://www.iso‐ne.com/markets/mktmonmit/rpts/ind_mkt_advsr/emm_mrkt_rprt.pdf.
22
Forarefutationofthe“windfallprofits”argument,seeCramtonandStoft2007,supranote20.
‐5‐
StateSubsidizationofElectricityGeneration
December2012
Maryland and New Jersey officials have argued that the lack of new generating capacity
being built in their states, despite higher capacity prices reflecting transmission
constraints, requires government intervention to subsidize new generating capacity and
reduce market prices. For example, the 2011 New Jersey Energy Master Plan stated,
“Despite high capacity prices in New Jersey as a result of the RPM capacity market
construct,newgenerationhasnotbeenbuiltrecentlyinNewJersey.”23
Theproblemwithsuchanargumentisthatrisingcapacitypricesarenotapersesignalof
theneedfornewgeneratingcapacityinvestment.Theentirepurposeofcapacitydemand
curvesistoprovideanefficientmarketsignal:onlywhenpricesrisetolevelsthatsupport
newgeneratingcapacity,calledthe“costofnewentry”(CONE),24shouldnewcapacitybe
developedtoensureresourceadequacy.Instead,therehasbeenanexplosionoflower‐cost
supply alternatives, including capacity “uprates” of existing generating units, and a rapid
increaseindemand‐response(DR)resourcesthathaveenteredthemarketatacostlower
thanCONE.Forexample,sincePJMimplementeditscapacitymarketin2007,almost6,000
MWofcapacityuprateshaveclearedintheRPMauction.Similarly,inthemostrecentRPM
auctionthatwasheldinMay2012,clearedDRresourcesincreasedtoalmost15,000MW.25
III.
HowGovernmentSubsidizationofGenerationHarmsMarketsand
Consumers
FERChasexpressedconcernthatcapacitymarkets,arevulnerabletobuyermarketpower.
Theissuewasfirstraisedin2006aspartofthesettlementthatestablishedthePJMRPM.26
More recently, FERC became concerned about state‐mandated intervention designed to
suppresscapacitymarketprices.Thisissuewasraisedextensivelyinhearingsaboutboth
theNewJerseyLongTermCapacityAgreementPilotProgram(LCAPP)andtheMaryland
Request for Proposal (RFP) capacity procurements, in which opponents of these states’
23
See2011NewJerseyEnergyMasterPlan,December6,2011(2011NJEMP),p.81.
24
CONErepresentsthe20‐yearlevelizedpriceoftheleast‐costnewpeakinggeneratingresourcethat
wouldbebuilt,netoftheexpectedrevenuessuchaplantwouldearnfromenergyandancillaryservice
sales.CONEistheestimatedaveragemarketpriceforapeakingunitthatwouldberequiredtomakean
investmentinsuchaplanteconomicallyworthwhile.SeePJM,Updated2015/2016RPMBaseResidual
AuctionPlanningPeriodParameters,April6,2012,p.8.http://www.pjm.com/markets‐and‐
operations/rpm/~/media/markets‐ops/rpm/rpm‐auction‐info/2015‐2016‐planning‐period‐
parameters‐report.ashx.
25
HistoricclearedDRresourcedatacanbefoundinthe2011StateoftheMarketReportforPJM,Volume2,
p.89,Table4‐9.http://www.monitoringanalytics.com/reports/PJM_State_of_the_Market/2011/2011‐
som‐pjm‐volume2‐sec4.pdf.DataforthemostrecentRPMauctionheldinMay2012isavailablefrom
PJM.See2015/2016RPMBaseResidualAuctionResults,p.9.http://www.pjm.com/markets‐and‐
operations/rpm/~/media/markets‐ops/rpm/rpm‐auction‐info/20120518‐2015‐16‐base‐residual‐
auction‐report.ashx.
26
SeePJMInterconnection,L.L.C.,117FERC¶61,331(2006).TheoriginalMOPRrulesexemptedresources
undertakenpursuanttostatemandate,whichledtofurtherFERCordersacceptingmodificationstothe
MOPRrules.SeePJMPowerProvidersGroupv.PJMInterconnection,L.L.C.etal.,135FERC¶61,022
(2011)(“PJMInterconnection2011”).
‐6‐
StateSubsidizationofElectricityGeneration
December2012
plannedinterventionwarnedthatthestateswouldunderminethePJMRPM,aswellasin
complaints brought before FERC.27 For example, the PJM Independent Market Monitor
(IMM) argued that the Maryland RFP would have significant long‐term adverse
consequencesonthecapacitymarketandalsodisagreedwiththecontentionthattheRPM
failedtoprovideadequateincentivesforbuildingnewgeneration.28
A.
BuyerMarketPower—NoSuchThingasaFreeLunch
Althoughthephrase“marketpower”typicallyconnotesactionstakenbysupplierstoraise
pricesartificiallyabovecompetitivelevels,withregardtostate‐mandatedinterventionin
capacitymarkets,suchastheNewJerseyandMarylandprograms,capacitysupplierswere
particularlyconcernedthatcapacitypurchaserswouldmanipulatethecapacitymarketto
lower prices artificially below competitive levels.29 State policymaker mandates for local
distributionutilitiestoacquirecapacityviatheuseofratepayer‐fundedsubsidiesarethe
principalmanifestationofthisconcern.
In formal capacity markets, which hold auctions to determine clearing prices, capacity
suppliers raised concerns that LSEs, which are required to obtain sufficient capacity
resources to meet their forecast demand plus a reserve for contingencies, could exercise
buyermarketpower.Specifically,capacitysuppliersexpressedconcernthatLSEscouldbe
ordered to build their own or contract for new generating resources—paid for by
customers—and offer the resulting capacity into the market at a zero price so as to
guarantee its clearing in the auction.30 By adding to the overall capacity supply, the
subsidizedgeneratingcapacitywouldthuslowerthemarket‐clearingprice.
Althoughsuchanapproachmayseemlikeareasonablestrategyforbenefitingconsumers,
it is actually a textbook example of buyer market power. The reason is that, in a well‐
functioning competitive market, neither an individual supplier’s nor an individual
consumer’s decisions will affect the market price. When buyers or sellers have market
power,however,theirindividualactionsdoaffectthemarketprice.Theextremecaseisa
monopolist (single supplier) or a monopsonist (single buyer). A sole buyer or seller, left
unchecked,willclearlyaffectthemarketprice,becausetheyarethatsideofthemarket.
27
AsFERCstatedinPJMInterconnection2011,“Themountingevidenceofriskfromwhatwaspreviously
onlyatheoreticalweaknessintheMOPRrulesthatcouldallowuneconomicentryhascausedusto
reexamineouracceptanceoftheexistingstateexemption,whichweapprovedaspartofthe2006RPM
SettlementOrder.Forthesereasons,weacceptasjustandreasonablePJM’sproposaltoeliminatethe
currentstateexemption.”(Par.139,footnotesomitted).Inthatsameorder,FERCagreedwiththe
PennsylvaniaPublicUtilitiesCommissionthat,“thereisnovalidstateinterestinensuringthat
uneconomicofferscansubmitbelow‐costoffersintotheRPMauction.(Id.,Par.142).
28
IntheMatterofWhetherNewGeneratingFacilitiesareNeededtoMeetLong‐TermDemandforStandard
OfferService,CaseNo.9214,CommentsoftheIndependentMarketMonitor,January28,2011.
29
AlloftheestablishedcapacitymarketshaveIndependentMarketMonitors(IMMs),whoalsoensurethat
capacitysuppliersdonotexercisemarketpowertoraiseprices,suchasbywithholdingsuppliesfromthe
market.
30
SeePJMInterconnection2011,Par.20.
‐7‐
StateSubsidizationofElectricityGeneration
December2012
Because they must meet specific capacity requirements for reliability purposes, LSEs
cannot manipulate capacity market prices by arbitrarily reducing the amount of capacity
they purchase. However, left unchecked, LSEs can manipulate capacity market prices by
artificially increasing capacity supplies. But in doing so, not only will LSEs harm the
market, but their own customers will end up worse off. Therefore, states that promote
subsidized capacity development to suppress market prices are really pursuing a “free
lunch” economic theory; they assume (contrary to basic economics) that lower market
priceswillnotchangethebehaviorofexistingandpotentialcapacitysuppliers.
Artificial price reductions caused by subsidized entry will cause existing power plants to
shut down prematurely or their owners to abandon plans to expand. Potential market
entrants, fearing further government intervention, will not build new power plants. And
investors, facing greater risks, will demand higher returns to compensate them for those
risks, thus raising thecost of capital for all suppliers. In the end, the marketwill lose as
many, or more, megawatts of supply as it gains through subsidies, which will result in
risingcosts.Consumerswillnotpaylessforelectricity;theywillpaymore.Thisiswhythe
“promise”ofbenefitsarisingfromgovernment‐subsidizedinterventionincapacitymarkets
willalwaysremainunfulfilled.
In their paper, Briggs and Kleit formally explore this “promise” of benefits from
government‐subsidized intervention in capacity markets. Specifically, they pose the
following question: if a utility’s retail customers are better off because that utility has
subsidizedinvestmenttoloweritsoverallcapacitybill,thenwhydon’tallstatesintervene
in the market by subsidizing capacity? As Briggs and Kleit show, the answer is market
dynamics.Inotherwords,whenstatesintervenetolowercapacitymarketpricesthrough
subsidized investment, that intervention harms existing and potential suppliers, who
respond accordingly. Thus, Briggs and Kleit prove that, contrary to state promises, but
consistentwithbasiceconomics,thereisno“freelunch.”
B.
TheBriggs‐KleitModel:EffectsofGovernmentSubsidizedEntryintoCapacity
Markets
To evaluate the economic impacts of state intervention in capacity markets, Briggs and
Kleitdevelopedamathematicalmodelofacapacitymarket.Themodelassumesthereisa
transmission‐constrained, “downstream” market and an unconstrained “upstream
market.31 They then examine how existing and potential capacity suppliers in these two
marketsrespondtosubsidizedgovernmententryinthedownstreammarketthathasthe
goal of artificially suppressing the market prices in that downstream market. Their
analysisshowsthefollowing:

31
Subsidized baseload capacity investments have significant potential for adverse
effectsinboththedownstreamandtheupstreamelectricitymarkets.
Appendix1providesamoredetailedreviewofthemodelandgraphicalanalysisoftheeconomicimpacts
ofmarketintervention.
‐8‐
StateSubsidizationofElectricityGeneration
December2012

Subsidizedinvestmentinbaseloadcapacityisneveroptimal.Inotherwords,such
subsidiesalwaysreduceconsumerwell‐being.

Government intervention reduces the incentive for private investment. If
governmentsrespondtothisreducedincentivebyinterveningagain,aviciouscycle
can arise and government’s perceived need to intervene becomes a self‐fulfilling
prophecy.Ultimately,theyconcludethattheentirecapacitymarketcancollapse.

Well‐designed Minimum Offer Price Rules (MOPR) can reduce the ability of the
subsidized generating capacity to distort the market‐clearing price by preventing
subsidizedgeneratorsfromsubmittingzero‐priceoffers,butcannotrestoremarket
optimality.
ThestartingpointfortheBriggsandKleitanalysisisthecostofbuildingnewgenerating
capacity, which is called CONE.32 If CONE is higher than the expected capacity market
price, no competitive supplier will build capacity. Thus, a utility would only build new
capacity if, by doing so, it could exert market power to drive down the market price.
Although,bydoingso,autilitymightreapsomeshort‐termbenefitsforitscustomers,these
are not economic benefits. Rather, they are short‐term wealth transfers from existing
competitivesuppliers.
Althoughproponentsofsuchsubsidiesmaydismisscomplaintsabouttransferringwealth
from existing suppliers, perhaps based on the “windfall” profits argument discussed
previously,thepromisedbenefitstoutilityconsumersassumethatexisting(andpotential)
suppliers do not respond to the changed circumstances in the market. In other words,
proponentsofstatesubsidizedinterventionassumeanunchangingor“static”market.
However, markets are never static. Instead, markets are dynamic, which is one of their
great benefits. As circumstances change, market participants adjust their behavior. For
example,asgasolinepriceshaveincreased,consumersdemandedmorefuel‐efficientcars
and trucks. Automakers responded with technological innovations to improve fuel
efficiency,whilemaintainingperformance.Asaresult,theimpactofhighergasolineprices
hasbeenreduced.
Notsurprisingly,marketdynamicsarealiveandwellincapacitymarkets.Therefore,when
oneaccountsforthedynamicmarketresponsesofexistingandpotentialcapacitymarket
suppliers in the long‐run, the wealth transfer vanishes and prices increase. Because
governmentsubsidiesdistortmarketsignals,theyforcesupplierstochangetheirbehavior.
Theartificiallyloweredmarketpriceforcesothercompetitivegenerationsuppliersoutof
themarket.Forexample,onOctober22,2012,DominionResourcesannounceditwould
shut down its 556 MW Kewaunee Power Station, which is located in Wisconsin, in early
32
Ineconomicparlance,CONEisthemarginalcostofnewentry.Ifthemarketpriceislessthanthe
marginalcostofadditionalsupplies,thenitdoesnotmakeeconomicsensetoenterthemarket.
‐9‐
StateSubsidizationofElectricityGeneration
December2012
2013 because of low market prices.33 As existing suppliers exit the market, capacity
supplies decrease. Not only do capacity market prices increase, but the reduction of
capacitycancreatereliabilityissues‐thesamereliabilityissuesthatsomeproponentshave
citedasjustifyingmarketintervention.
Governmentsubsidiesalsoincreasethecostofnewentryintothemarket,byincreasingthe
costofcapital.Investorswillnotwanttorisktheircapitaltocompetewithplantsthathave
been artificially subsidized by consumers. In other words, investors will demand
additionalreturnstobeartheriskoffuturegovernmentinterventionthatwouldreducethe
economic value of their investments. Alternatively, investors may simply wait for their
own subsidies. Ultimately, such a system would return us to the pre‐restructuring,
monopolyutilityworld.Itwouldresultinlessefficientplantsandconsumersonceagain
havingtobeartherisksofutilityinvestments.
Ultimately, market dynamics mean that artificially subsidizing new generation cannot
“trick”themarket.Subsidiesdonotloweractualcosts,butinsteadraisethem.Subsidies
donotencourageinnovation,butreducetheincentivetoinnovateandlowercoststhrough
greater operating efficiency. Subsidies do not improve long‐term reliability, but instead
harm it by driving out existing and potential competitors. To believe otherwise is to
believein“freelunch”economics.
Finally,BriggsandKleitshowthatawell‐designedMinimumOfferPriceRule(MOPR)34can
limittheabilityofsubsidizedgeneratingcapacitytodistortthemarket‐clearingprice,but
cannoteliminatealloftheharmcausedbysubsidizedgeneration.Inessence,aMOPRrule
meansgovernment‐subsidizedgenerationcannotbeofferedintothecapacitymarketata
zeroprice.Thislimitsthemagnitudeofthepotentiallossestoexistingmarketparticipants
arisingfromsubsidizedintervention,butcannotrestorethemarketoptimum.Insummary
they find, “Policies like PJM’s MOPR may mitigate this situation if they correctly and
crediblyscreenoutnon‐economiccapacityadditions.”35
IV.
ConclusionsandPolicyImplications
The Briggs‐Kleit model shows that government intervention imposes costs on both
consumers who are supposed to benefit and on consumers in interconnected markets.
Because government intervention, or even the expectation of intervention, imposes costs
on capacity suppliers, they will be less likely to invest. As a consequence, government
intervention to “correct” for a perceived absence of market‐based investments in
generatingcapacitysimplydrivessuppliersaway.Thus,governmentinterventionisaself‐
fulfilling prophecy that, if left unchecked, can eventually drive all suppliers from the
capacitymarket.
33
Inpart,lowershort‐runmarketpricesarebeingdrivenbyrapidincreasesinsubsidizedwindgeneration.
About20%ofallUSinstalledwindcapacity,approximately10,000MW,hasbeenbuiltwithinMISO.
34
SeePJMInterconnection,L.L.C.,139FERC¶61,011(2012).
35
BriggsandKleit,p.27.
‐10‐
StateSubsidizationofElectricityGeneration
December2012
Because electric markets are interconnected and extend beyond individual states,
governmentinterventioninindividualstates,suchasundertheMarylandandNewJersey
programs,imposeslong‐runcostsonneighboringstates.Itiseconomicfallacythatprice
distortionscausedbygovernmentsubsidiesinafreemarketare“benefits.”Such“beggar‐
thy‐neighbor”policiesarenotonlycounterproductive,theyinvitepolicy“retribution”that
willfurtherdamagemarkets.
Finally,justifyinginterventionasameans,notonlytoreducemarketprices,buttocreate
jobs, is a last refuge of the interventionist scoundrel. The short‐run “jobs” benefits from
buildingsubsidizedgenerationareunlikelytoaccruetolocalconsumers,whomustpayfor
the subsidies. By reducing incentives for new private investment in the long run and
increasingelectricprices,interventionwillcauseanoverallreductionineconomicgrowth.
Thus, such policies, pursued for economic development reasons, are “penny wise and
pound foolish.” At best, they are an undue burden on taxpayers. At worst, as shown by
Briggs and Kleit, “these subsidies have the perverse effect of reducing the incentives for
resourceadequacyinthelongrun.”BriggsandKleitconcludethat“ouranalysissuggests
thatto“keepthelightson,”stateseitherneedtoletmarketsworkorfacetheprospectof
continued, costly interventions over the long run.”36 Based on these findings and the
consequences discussed in this paper, future state subsidization of electricity generation
shouldbeavoided.Itharmstheveryconsumersitisintendedtobenefit,harmsconsumers
inneighboringstates,andwreckscompetitivecapacitymarkets.
36
BriggsandKleit,p.27.
‐11‐
StateSubsidizationofElectricityGeneration
December2012
Appendix1:TheBriggs‐KleitMultipleMarketModel
TheBriggs‐Kleitmodelconsiderstheimpactsofsubsidizedgenerationwhenthereareboth
upstreamanddownstreamcapacitymarkets.43Thedownstreammarketisassumedtobe
transmissionconstrainedatleastsomeofthetime.Asaresult,thedownstreamcapacity
market price, PD, is assumed to be higher than the price in the upstream market, PU, as
showninFigureA1‐1.44
FigureA1‐1:Briggs‐KleitTwo‐MarketModel
This price differential is assumed to underlie downstream policymakers’ desire to
subsidizenewgenerationentry.BriggsandKleitexaminetheimpactsofbothsubsidized
baseload capacity and subsidized peaking capacity to determine whether the type of
generationsubsidizedaffectstheoverallchangeineconomicwell‐beingtoconsumersand
existingsuppliers.Theirmodeladdressesfourpolicyquestions:

How does the presence of transmission constraints affect the optimal design and
functionofcapacitymarkets?

Doesitevermakesenseforgovernmentstosubsidizecapacity,andifsohow?
43
TheBriggsandKleitmodelextendsthesingle‐marketanalysispreviouslydevelopedbyJoskowand
Tirole,whichwasbasedona“Ramseyequilibrium”model.InplainEnglish,aRamseyequilibriumisone
inwhichthegovernmentknowshowtheprivatesectorwillreacttogovernmentintervention.The
governmentintervenesinawaydesignedtomaximizeoverallwell‐beingandtheresultingmarket
conditions,aftertheprivatesectorresponse,istheRamseyequilibrium.SeeP.JoskowandJ.Tirole,
“ReliabilityandCompetitiveElectricityMarkets,”TheRANDJournalofEconomics38(Spring2007),pp.
60‐84(JoskowandTirole2007).
44
Iftherewerenotransmissionconstraints,themarket‐clearingpriceswouldbethesameandthetwo
marketswouldeffectivelybetreatedasasingleone.ThereasonPJMandotherRTOswithcapacity
marketshaveseparatezonesistoreflecttransmissionconstraintsandprovidemoreaccurateprice
signalsinconstrainedmarkets.
A1‐1
StateSubsidizationofElectricityGeneration
December2012

Are price floors like the MOPR sufficient for ensuring the optimal functioning of
capacitymarkets?

How will markets react to subsidized capacity investments in the short and long
runs?
The Briggs and Kleit framework also allows for price caps in the energy market. As
discussed previously, such price caps create a “missing money” problem that capacity
marketsaddress.ExtendingtheJoskowandTirole(2007)model,theyshowthatcapacity
markets restore the appropriate market incentives for investment in peaking capacity as
longas(1)allgeneratingunitsarepaidthecapacityclearingprice;and(2)thatthemarket
price consumers pay incorporates the capacity market price, and not just the short‐run
energymarketprice.45
Wenowaddressthequestionofwhethergovernmentinterventioneverimprovesoverall
well‐being.Aswediscussedpreviously,reliabilityisapublicgoodthatmarketswilltendto
under‐supply. Briggs and Kleit show that, in the short‐run and in an ideal world, the
governmentcouldprovidethecorrectincentivesfornewpeakingcapacityinthemarket,
andthusensuresufficientcapacitytomeetreliabilitystandards.Theyalsoshowthisresult
doesnotholdforsubsidizedbaseloadcapacity.However,eveninthisidealized,short‐run
setting, Briggs and Kleit show that economic well‐being is increased if the market
determinesthecapacityprice.Astheystate:
Reducing the capacity price affects the incentive for the marginal producer
andreducescapacityinvestment.Thisintuitionessentiallyrestatesthelogic
at the heart of a capacity market policy: the policy supplies the market
information on the demand for reliability, and allows the market to
determine the price for that capacity. When governments act to affect this
price, they distort the information that the market provides, reduce
incentivesat the margin for capacity suppliers, and reduce the efficiency of
theoutcomeforconsumers.46
Briggs and Kleit consider both the short‐run and long‐run impacts of government
intervention in the transmission‐constrained downstream market. In the short‐run, they
showthat,becausedownstreaminterventionwithsubsidizedcapacityreducesexportsinto
the downstream market, upstream consumers benefit from downstream government
intervention. In other words, because downstream consumers must pay for the
government‐subsidized capacity, those consumers may not see any benefit, even in the
short‐run.
45
BriggsandKleitalsoshowthatthecombinationofanenergymarketpricecapandacapacitypayment
meansthatthegovernmentcannotcreateaRamseyequilibrium,whenevertherearemorethantwo
possiblestatesoftheworld.
46
BriggsandKleit,p.17(emphasisadded).
A1‐2
StateSubsidizationofElectricityGeneration
December2012
Short‐andLong‐RunImpactsofGovernmentSubsidizedEntryonCapacityMarkets
Below, we provide a general overview of the short‐run and long‐run impacts of
government‐subsidized entry based on the work by Briggs and Kleit. To understand the
short‐runimpacts,supposetherearefourcapacitysuppliers—A,B,C,andD—whosubmit
offers to provide capacity in the amount of QA, QB, etc., of increasing cost, as shown in
Figure A1‐2 (supply curve S0). Based on the administratively determined demand curve,
the market‐clearing capacity price is P*. Q* is the equilibrium quantity of capacity
purchased,withallofthecapacityofferedbysuppliersAandBtaken,andonlysomeofthe
capacityofferedbysupplierC.47
Next, suppose subsidized capacity supplier S is allowed to bid in a quantity of capacity,
QSUB,intothemarketatazeroprice.Inthatcase,theoffersoftheothersuppliersareall
pushedout,andthenewcapacitysupplycurveisS1.Thenewsupplycurveintersectsthe
demand curve at Q and the resulting market price falls to PSUB. At this lower price,
supplierCisknockedoutofthemarketentirely,andtheonlysuppliersofcapacityareA,B,
andthenewsubsidizedsupplier.
FigureA1‐2:Short‐runImpactofGovernmentSubsidizedCapacity
47
InFigureA1‐1,theamountofcapacitysuppliedbyCis:QC–(Q*–QB).
A1‐3
StateSubsidizationofElectricityGeneration
December2012
Intheshort‐run,consumersappeartogainfromthedecreaseincapacityprice.Thisgain,
called “consumers’ surplus,” accrues primarily from a transfer of wealth from capacity
suppliers48 (the large diagonally‐shaded rectangle) and a small increase because of the
increase in total capacity supplied (the small gray triangle). The net benefit, if any, to
consumersisthedifferencebetweenthemonetarygainfromthelowercapacitypriceand
the cost of the subsidized capacity. A key finding of Briggs and Kleit’s work is that the
consumerscalledontosubsidizecapacitythroughLCAPP‐typeprogramsareleast‐likelyto
enjoy even short‐run benefits, whereas consumers upstream will capture the majority of
theshort‐runbenefits.
Although subsidizing capacity can reduce market prices in the short‐run, focusing
exclusively on the short run ignores critical market dynamics, that is, effects over time.
Whenevaluatinggovernmentinterventionintendedtosuppressmarketprices,accounting
for these dynamiceffects is crucial, for several reasons. First, marketsare not static;the
entire point of a capacity market, for example, is to send appropriate price signals over
timeandincententry(andexit)asneeded.Second,marketparticipantsrespondtoprice
signals. Thus, subsidized entry in the capacity market will have long‐run impacts on,
unsubsidized suppliers. Understanding these impacts is crucial to evaluating the overall
impactsofgovernmentintervention.
Next, we consider the long‐run impacts. As Briggs and Kleit show, in the long‐run all
consumers lose. To see this, consider Figure A1‐3. Because the subsidy reduces the
market‐clearing price in the short‐run, Briggs and Kleit’s analysis finds that existing
suppliers A and B respond by either exiting the market or forgoing investment. Thus, in
our example, the capacity supplied by A and B decreases over time to QA’ and QB’,
respectively. Moreover, because investors perceive greater risk to supplying capital to
generators, the cost of capital increases, which further raises costs and the market price
(Psub)abovewhat itotherwisewouldbewithoutthesubsidizedgeneratingcapacity.This
resultsinthelongrunmarketprice(Psub)risingabovetheoriginalmarketprice(P*)asa
result of government‐subsidized entry into the market without mitigation to protect
consumers.
48
Thisisreflectedasareductionin“producers’surplus,”whichrepresentsthereturnstoindividual
suppliers.
A1‐4
StateSubsidizationofElectricityGeneration
December2012
FigureA1‐3:Long‐RunImpactofGovernmentSubsidizedCapacity
Long‐term market price increases
Because government intervention downstream reduces exports from upstream, the net
upstream capacity supply increases, causingthe marketprice todecrease.Consequently,
upstream market consumers benefit from the downstream market intervention. For
example, this suggests that the primary beneficiaries of the New Jersey and Maryland
capacity subsidy programs are consumers in Pennsylvania and Ohio, not in the states
themselves. Thus, advocates of intervention in the New Jersey and Maryland capacity
zonesareeffectivelyforcingtheirconsumerstosubsidizebenefitsforconsumersupstream
inPennsylvania.
Iftheseupstreambenefitspersistedinthelong‐run,wewouldexpecttoseepolicymakers
inupstreamstatescheeringontheinterventionisteffortsindownstreamefforts.Because
we do not, it means there are adverse long‐term impacts stemming from the dynamic
response of capacity market suppliers in both upstream and downstream markets. To
evaluate long‐term, dynamic responses to government intervention, Briggs and Kleit
assumethatmarketparticipantsbasetheirinvestmentdecisionsonexpectationsoffuture
government intervention. Thus, if capacity suppliers believe there is a 50% probability
that the government will intervene in the market at some later time, and that the
probabilityofinterventionisaffectedbythemarketprice,thensupplierswilladjustchange
their supply decisions. In essence, the expectation of market intervention—even if that
intervention does not actually occur—imposes an expected cost on suppliers. The
A1‐5
StateSubsidizationofElectricityGeneration
December2012
expectedcostisaformofregulatorytaking.49Theresultofthisexpectationofintervention
is that suppliers will reduce future investment, reducing reliability in both the upstream
and downstream markets. According to Briggs and Kleit, “[A]s suppliers’ expectations of
government action become reinforced, the expected return to capacity investments
decreases.Thisstateofaffairscreatesaproblemforlongrungridresourceadequacy….”50
Theexpectationofintervention,whichistiedtothecapacitymarketprice,thusincreases
themarketpriceneededtoinducesupplierinvestment.However,ifsuppliersseehigher
marketprices,theywillalsoexpecttheprobabilityofgovernmentinterventiontoincrease,
which means they require an even higher market price to induce them to enter, and so
forth.Inotherwords,theexpectationofgovernmentinterventioncreatesaself‐fulfilling
conditionthatincreasescostsanddrivessuppliersfromthemarket.Eventually,theonly
suppliersinamarketwillbesubsidizedones,andthemarketwillceasetoexist.
Thus,whengovernmentpolicymakersindownstreammarketscomplainthatmarketprices
are not inducing new capacity investment and threaten to intervene with subsidized
capacity, the policymakers effectively signal to suppliers not to invest. And, because
downstream intervention also leads to lower short‐run prices in upstream markets,
threatened(oractual)interventionrestrictsinvestmentinupstreammarkets.TheBriggs
and Kleit model demonstrates that the long‐run impact of government intervention
reduces well‐being in both downstream and upstream markets, compared with the non‐
interventioncase.
Briggs and Kleit conclude their paper by analyzing how MOPR‐type instruments, which
limit the extent to which subsidized capacity can be offered into the capacity market,
affects the overall welfare impacts. They find that MOPR‐like instruments limit the
magnitudeofthepotentialregulatorytakingsarisingfromsubsidizedintervention(similar
towhatwasshowninFigureA1‐2),butcannotrestorethemarketoptimum.
Finally,FigureA1‐4providesanexampleoftheshort‐runimpactsofaMOPR.Specifically,
supposetheMOPRforthesubsidizedgeneratingresourceisbetweentheoffersofcapacity
resourcesBandC.Inthiscase,thesubsidizedgenerationcapacityisnotallowedtobida
pricebelowMOPRSUB.Thesubsidizedmarketpriceisstillbelowtheinitialmarket‐clearing
price,P*,buttheshort‐runimpact,intermsofthewealth‐transferfromexistingcapacity
supplierstoconsumersandoverallbenefittoconsumers,decreases.However,theMOPR
instrument cannot restore optimality to the market, because the threat of government
intervention, and the impact of that threat on investor expectations, will still distort the
market.
49
Foracomprehensivediscussionofregulatorytakings,seeW.Fishel,RegulatoryTakings:Law,Economics
andPolitics,(Cambridge,MA:HarvardUniversityPress1995).
50
BriggsandKleit,p.27.
A1‐6
StateSubsidizationofElectricityGeneration
December2012
FigureA1‐4:Short‐runImpactofSubsidizedCapacity–WithMOPR
A1‐7
StateSubsidizationofElectricityGeneration
December2012
Appendix2: ReviewofElectricRestructuringandCreationofCompetitive
WholesaleMarkets
By the late 1970s, the electric industry was in turmoil. After two OPEC oil embargoes,
rapidly escalating costs for nuclear power plants that threatened to bankrupt electric
utilities, new environmental regulations that were needed to reduce air and water
pollution, and looming shortages of natural gas (ironically caused by preventing market
pricesfromincentingnewexplorationanddrilling),theelectricindustrywasonitsknees.
A new direction clearly was needed, one which would restore the industry’s financial
health, ensure enough new generating capacity was built to meet increased demand, and
keepelectricitypricesfromskyrocketingoutofcontrol.
Thisnewdirectionentailedrestructuringtheelectricindustry,supplementingand,insome
cases, replacing the old vertically integrated structure with competitive wholesale and
retail markets. Electric utility restructuring was inspired by the success of introducing
market competition into other regulated industries, including airlines, trucking,
telecommunicationsand,perhapsmostimportantly,naturalgas.Inalloftheseindustries,
competitionhasledtoinnovation,improvedefficiency,andlowerprices.
In1992,CongresspassedtheEnergyPolicyAct(EPAct),whichcreatedthefoundationfor
competitivewholesaleandretailelectricmarkets.Specifically,EPActcreatedanewclassof
competitive generators, called Exempt Wholesale Generators (EWGs). EWGs were
designed to compete directly with generation built and operated by electric utilities
themselves,unlikethe“qualifiedfacilities”(QFs)createdaspartofthe1978PublicUtilities
ResourcePolicyAct.Thelatterwereprimarilyrenewablegeneratorswhoseoutputelectric
utilitieswererequiredtopurchaseatpricessetbystateutilityregulators—oftenathighly
inflated prices stemming from wildly inaccurate forecasts of future oil and natural gas
prices. Several years later, in 1996, the Federal Energy Regulatory Commission (FERC) issued
Order888,a major policy order designed torestructure the electric transmission system
and promote “open” competitive access for generators, thus enabling them to sell
electricity in wholesale markets.51 Although so‐called “power‐pools” – multiple utilities
combiningtheirgeneratingresourcestoreducecostsandimprovereliability–hadexisted
since the late 1920s,52 Order 888 led to creation of independent transmission
organizations, called “independent system operators” (ISOs) whose role was to better
coordinatetransmissionsystemoperationsandensurethatopenaccessrequirementsdid
not jeopardize the overall reliability of the bulk power system (i.e., the system of central
stationgenerationandhighvoltagetransmission).
51
PromotingWholesaleCompetitionThroughOpenAccessNon‐discriminatoryTransmissionServicesby
PublicUtilities;RecoveryofStrandedCostsbyPublicUtilitiesandTransmittingUtilities,OrderNo.888,75
FERC61,080(1996),
52
ThefirstsuchpowerpoolwasthePennsylvania–NewJersey–Marylandpool,theprecursortoPJM,
whichbeganin1928.
A2‐1
StateSubsidizationofElectricityGeneration
December2012
Three years later, FERC issued Order No. 2000, which was designed to further enhance
openaccesstransmissionandcompetitivewholesalemarketsthroughcreationofRegional
Transmission Organizations (RTOs), which can be thought of as enhanced ISOs.53 In
addition to operating wholesale spot markets for electric energy, several of these RTOs
developed separate markets for installed generating capacity—essentially payments to
generating firms to recover the fixed construction costs that were previously included in
rate base and in return for those to provide sufficient revenues for firms to constructing
additionalgeneratingcapacityforuseduringtimesofpeakdemand.Thoughthatcapacity
wouldbeuneconomicalinawholesaleenergymarket,itwasnecessarytoensuretherewas
sufficientgeneratingcapacitytomeetreliabilitystandards.54
53
RegionalTransmissionOrganizations,89FERC¶61,285(1999).Althoughtheyarenottechnicallythe
sametypesofentities,theresponsibilitiesandoperationsofISOsandRTOsarequitesimilar.Further
addingtotheconfusion,RTOscoveringNewEnglandandtheMidwestareISO‐NEandMidwestISO,
respectively.
54
Therearetwoflavorsofreliability:long‐termresourceadequacyandshort‐termsystemsecurity.
Capacitymarketsweredevelopedtoaddresstheformer.Thelatter,incontrast,focusesonspecific
minute‐to‐minuteoperationofthebulkpowersystem.Foramoredetaileddiscussion,seeJ.Lesserand
G.Israelivich,“TheCapacityMarketEnigma,”PublicUtilitiesFortnightly,December2005,pp.38‐42.
A2‐2
StateSubsidizationofElectricityGeneration
December2012
Appendix3:TheNewJerseyandMarylandPrograms
The New Jersey and Maryland subsidized generation programs were established for a
varietyofreasons,includingeconomicdevelopmentandjobcreation.Theprincipaldriver
oftheprograms,however,wastoreducethecostofgeneratingcapacitythestate’selectric
utilities are required to purchase to meet PJM’s reliability standards. For example, New
Jersey Board of Public Utilities (NJBPU) President Lee Solomon criticized both PJM and
FERC,arguingthatthePJMcapacitymarkethad“overcharged”thestatebymorethan$1
billion, without incenting new generating capacity within the state. This issue was also
raisedinthe2011NewJerseyEnergyMasterPlan,whichstates:
[m]any market participants argue that RPM has not brought enough
generation into the markets where and when needed. A key finding of the
June 2010 BPU Technical Conference was that generators proposing new
projectsarenotabletoobtainfinancingatreasonableratestodevelopnew
assetsduetouncertaincapacityrevenues.55
To address the perceived shortcomings of the PJM RPM, NJ Senate Bill 2381, which was
signed into law in January 2011, created the “Long‐Term Capacity Agreement Pilot
Program” (LCAPP). Under LCAPP, generation developers responded to an RFP to build a
total of 2,000 MW of new, in‐state generating capacity. The winning generators were
awarded 15‐year contracts with the state’s electric utilities and are required to bid and
clear all of this new capacity into the PJM RPM auction. The winning bidders are
guaranteed a fixed price for their capacity. If the market price is less than the bid price,
utilityratepayersarerequiredtomakeupthedifference.Thisrevenueguaranteeisaform
ofdirectsubsidy.
Three generation developers—Competitive Power Ventures (CPV), Hess, and NRG
Energy—were selected to build new capacity under the LCAPP solicitation. As a result,
New Jersey ratepayers have guaranteed CPV and Hess over $2.1 billion in revenues over
thenext15years.56TheCPVplantwillbeon‐linebeforetheJune1,2015startofthePJM
2015‐2016 energy year, and will be paid a price of $286.03/MW‐day for the first year.57
TheHessfacility,whichclearedintothe2015‐2016auction,willreceiveacontractpriceof
$200/MW‐daybeginninginthe2016‐17energyyear.TheNRGfacilityfailedtoclearinthe
PJMRPMauctionbecauseofPJM’sMinimumOfferPriceRule(MOPR).58
55
2011NJEMP,pp.21‐22.
56
SeeG.Thomas,“TimetoFold'em‐‐NewJersey'sBetonPowerPlantsGoesTerriblyWrong,”NJSpotlight,
June13,2012.http://www.njspotlight.com/stories/12/0613/1432/.
57
Incomparison,themarket‐clearingpricewas$167.46/MW‐day.SeeJ.Kaltwasser,“StateReleasesNew
LCAPPNumbers,”NJBIZ,October3,2012.
http://www.njbiz.com/article/20121003/NJBIZ01/121009938/State‐releases‐new‐LCAPP‐numbers.
58
ISO‐NEusesasimilarconcept,knownas“out‐of‐market”(OOM)resources.SeeISONewEngland,Inc.,et
al.,135FERC¶61,029(2011).
A3‐1
StateSubsidizationofElectricityGeneration
December2012
The Maryland program is smaller in scope, but is based on similar principles. The
Maryland program began in 2009, when the Maryland Public Service Commission
(MarylandPSC)issuedanorder“toinvestigatewhether[theCommission]shouldexercise
its authority to order electric utilities to enter into long‐term contracts to anchor new
generationortoconstruct,acquire,orlease,andoperate,newelectricgeneratingfacilities
inMaryland.”59
Based on similar concerns that the PJM RPM was not providing adequate incentives to
buildnewgeneratingcapacityinMaryland,theMarylandPSCrequiredtheutilitiestoissue
a request for proposals (RFP) for up to 1,500 MW of gas‐fired generating capacity.
UltimatelyCPVMarylandwontheRFP,proposingtobuilda660MWgas‐firedgenerating
unitinCharlesCounty,Maryland,withacommercialoperationdateofJune1,2015.Like
the New Jersey LCAPP, the Maryland program provides a revenue guarantee to CPV, in
which, if the RPM market price falls below CPV’s guaranteed revenue stream, Maryland
ratepayerswillmakeupthedifference.60ThetotalcostoftheMarylandrevenueguarantee
isover$800millionoveraten‐yearperiod.61
59
IntheMatterofWhetherNewGeneratingFacilitiesareNeededtoMeetLong‐TermDemandforStandard
OfferService,CaseNo.9214,OrderNo.82936,September29,2009,pp.2‐3.
60
Id.,OrderNo.84815,April12,2012.
61
CaseNo.9214,MemoofBostonConsultingGrouptoMarylandPublicServiceCommission,April3,2012,
p.5.
http://webapp.psc.state.md.us/Intranet/Casenum/NewIndex3_VOpenFile.cfm?ServerFilePath=C:\Casen
um\9200‐9299\9214\\138.pdf.
A3‐2
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Wind Generation Patterns and
the Economics of Wind
Subsidies
Dr. Jonathan Lesser is the founder
and President of Continental
Economics, Inc., an economic and
litigation consulting firm providing
services to utilities, industry, and
regulators on a broad array of market,
regulatory, investment, and
environmental issues affecting all
segments of the energy industry in
the U.S., Canada, and Latin America.
Dr. Lesser has coauthored three
textbooks, most recently Principles
of Utility Corporate Finance,
which was published in 2011 by
Public Utilities Reports, Inc. He
holds a B.S. in Mathematics and
Economics from the University of
New Mexico, and a M.A. and Ph.D.
in Economics from the University of
Washington. He can be reached at
[email protected].
Funding for this research was
provided, in part, by Exelon
Corporation. However, the views
expressed are solely those of the
author and do not necessarily
represent the views of Exelon
Corporation or its subsidiaries.
8
An analysis supports the conclusion that there is no
economic rationale for continued subsidization of wind
generation. At the federal level, direct subsidies, such as
the federal production tax credit, should not be continued.
State-level subsidies, whether feed-in tariffs established by
state regulators or statutory RPS mandates, further
exacerbate market distortions and raise electricity prices,
again to the detriment of consumers.
Jonathan A. Lesser
I. Introduction
The United States has
subsidized the wind industry for
35 years. At the federal level,
subsidies began with the Public
Utility Regulatory Policy Act
(PURPA) of 1978. Under PURPA
provided indirect subsidies for
renewable generation through
mandates that electric utilities
purchase the output of
qualifying facilities (QFs) based
on forecasts of avoided costs,
essentially ‘‘but for’’ cost
projections made by the utilities
and approved by state
regulators, or made by those
regulators themselves. With
passage of the Energy Policy Act
of 1992 (EPAct), wind subsidies
were increased through a variety
of programs. The most
prominent was the federal
production tax credit (PTC).1
Although not specifically limited
to wind generation,
approximately 75 percent of the
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
The Electricity Journal
Author's personal copy
total PTC credits claimed since
its inception have been for wind
generation.2 Wind generation
benefits from other subsidies as
well. Since 2009, for example, the
wind industry has received
payments under the $831 billion
American Recovery and
Reinvestment Act of 2009
(ARRA). Perhaps the largest
subsidy has been through statelevel renewable portfolio
standards (RPS), which mandate
minimum levels of renewable
generation that electric utilities
or competitive generation
suppliers must obtain as part of
their overall resource mix used
to serve customers. Currently, 30
states, plus the District of
Columbia, have such RPS
mandates.
nlike market prices for
other commodities, the
market price of electricity varies
by season, day, and hour. In part
because electricity cannot be
stored cost-effectively, the price is
highly dependent on daily
fluctuations in demand—higher
demand during the day and
lower demand at night—and
seasonal changes. In most of the
U.S., for example, electricity
demand now peaks in the
summer, driven by increased use
of air conditioning in commercial
and residential buildings. As a
result of this price variation, the
value of subsidized wind
generation also varies by season,
day, and hour. In some hours, the
value of electricity can be
thousands of dollars per MWh. In
other hours, the value actually can
be less than zero.
U
Jan./Feb. 2013, Vol. 26, Issue 1
The purpose of this article is to
examine the economic value of
subsidized wind generation.
Specifically, are taxpayers and
consumers who are forced to pay
for subsidized wind power
receiving high- or low-value
electricity? Answering this
question has important policy
implications. First, Congress is
currently considering whether or
not to extend the PTC for an
additional year, at an estimated
Are taxpayers and
consumers who are
forced to pay for
subsidized wind power
receiving high-value or
low-value electricity?
cost of over $12 billion. Second,
because the percentages of
renewable generation required
under state RPS requirements
continue to increase, electricity
consumers will be forced to
subsidize greater amounts of
wind power, which will have
larger impacts on electricity costs.
Third, continued subsidization of
wind generation will lead to
higher long-run retail prices for
electricity,3 which will have
adverse impacts on economic
growth. Given these reasons,
determining the value consumers
obtain for their subsidy dollar is
highly relevant to policy decisions
regarding continued subsidies.
II. Economic Costs of
Wind Power Subsidies
Renewable energy subsidies
have been advocated for a
variety of reasons, ranging from
common arguments about
protecting emerging or ‘‘infant’’
industries so they may become
established, 4 to ‘‘two wrongs
make a right’’ justifications, i.e.,
that because fossil fuel
generating resources have been
subsidized, it is only ‘‘fair’’ that
renewable generation be
subsidized, to arguments that
renewable subsidies offset
external environmental costs of
fossil fuel generation.5
Regardless of how they are
justified, subsidies distort
competitive markets, drive out
unsubsidized competitors, and
reduce the incentives to innovate
and improve operating
efficiency.6 In addition to these
economic costs, wind power
subsidies create four other types
of adverse economic spillovers
because of the nature of electric
markets and integrated power
grids.
irst, because baseload
generators, e.g., nuclear and
coal-fired power plants, cannot be
cycled easily, these generators
operate even when the market
price of electricity is less than their
variable operating costs.7 As a
consequence, when the demand
for electricity is sufficiently low,
market prices can fall below zero.
In such situations, baseload
generation owners are then forced
to pay to generate power and
inject that power into the grid,
F
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
9
Author's personal copy
which exacerbates economic
losses.8 With a current after-tax
PTC of $22/MWh, it is
economically rational for wind
generators to sell power into the
market even when prices are as
low as $34/MWh.9 As a result,
negative pricing periods are
exacerbated, which increases the
costs for baseload generators who
are unable to cycle their units and
may hasten their retirement.
econd, the inherent
intermittency of wind
generation increases the costs of
maintaining power system
reliability. The intermittent nature
of wind generation requires
additional generating reserve
capacity so as to ‘‘firm’’ wind
supply. Moreover, rapid
variations in wind output can
require additional voltage
support through automatic
generation control (AGC) that
automatically adjusts the output
of flexible generating resources
(e.g., gas-fired turbines) so as to
maintain voltage and frequency
within acceptable levels. A study
published by the National
Renewable Energy Laboratory
(NREL) estimated these
integration costs to be about $5/
MWh.10 In Texas, which has over
10,000 MW of installed wind
capacity, in 2011 these integration
costs added an estimated
additional $140 million in power
system costs. Nationally,
integration costs were over $500
million in 2011.11
Third, wind generation
requires additional investment
in high-voltage transmission
lines, because wind resources are
S
10
geographically dispersed and
typically located far from load
centers. The costs of highvoltage transmission lines are
generally socialized across all
transmission system users. Texas
alone spent over $6.9 billion on
Competitive Renewable Energy
Zone (CREZ) high-voltage
transmission lines to
interconnect wind power.12
Fourth, the demonstrated
inaccuracy of short-term forecasts
of wind generation increases the
overall cost of meeting electric
demand as system planners must
reimburse other generators who
had been scheduled to operate, but
were not needed because actual
wind generation was greater than
forecast, or had not been
scheduled, but were required to
operate because actual wind
generation was less than forecast.
Although generators can be
penalized for erroneous forecasts,
most of the resulting system costs
are socialized across all users.
Despite claims by wind power
advocates that wind generation
can be predicted accurately several
days in advance, allowing system
operators to reduce, if not
eliminate, the impacts of wind’s
volatility, actual data does not bear
this out.13
III. The Economic Value
of Wind Generation
To examine the economic value
of subsidized wind generation,
we analyzed wind generation in
three regions where there has
been extensive—and rapid—
development of wind power: the
PJM Interconnection, which
covers the mid-Atlantic states
and the Ohio Valley; MISO, which
covers much of the remaining
Midwestern States; and ERCOT,
which oversees the electric
system in almost the entire state
of Texas. Together, these three
regions account for over
27,000 MW of wind generating
capacity, more than half of the
approximately 50,000 MW of
installed wind generating
capacity in the U.S.14
Because of weather patterns
that can change from year to year,
we examined hourly wind
generation and load data over a
44-month period, Jan. 1, 2009,
through Aug. 31, 2012, to assess
the relative economic value of
wind power. We then evaluated
the performance and availability
of wind power in each of the four
seasons, where each season was
defined as including the months
shown in Table 1.15
From both a system planning
and customer perspective, the
highest-value generating
resources are those that are
available when electricity
demand peaks: like taxicabs that
never show up when it is raining,
generating resources that fail to
produce when most needed have
little value.
Table 1: Month-Season Mapping.
Season
Winter
Months
December–February
Spring
March–May
Summer
Fall
June–August
September–November
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
The Electricity Journal
Author's personal copy
[(Figure_1)TD$IG]
Figure 1: PJM Hourly Load and Wind Generation, July 1–8, 2012
Consider, for example, the
pattern of hourly load and
wind generation in PJM for the
week of July 1–8, 2012, when
much of the eastern U.S. was in
the grip of a record heat wave
(Figure 1).
ver that week, a strong
negative correlation
between hourly demand and
wind generation is apparent. The
actual correlation coefficient is
0.40.16 As Figure 1 shows, over
this week, wind generation
usually peaked in the late night
and early morning hours,
whereas peak demand occurred
in the late afternoon. Electricity
demand peaked at 5 PM on July 6
of this week, when demand was
over 151,000 MW. During that
same hour, 201 MWh of wind
power was generated by the
approximately 4,700 MW of
installed wind capacity in PJM,
O
Jan./Feb. 2013, Vol. 26, Issue 1
less than 5 percent of the potential
generation. As little generation as
that was, it represented an
increase from earlier in the day, as
only 14 MWh was generated
during the hour between Noon
and 1 PM.
In the Northern Illinois zone,
which encompasses Chicago, the
demand for electricity averaged
22,000 MW over the entire day;
the average amount of wind
power generated was just 4 MW.
From a system planning
standpoint, the ‘‘gap’’ between
high hourly loads and low wind
output makes wind a far less
valuable and far less reliable
resource than conventional
generating resources. This ‘‘gap’’
between peak electric demand
and low wind generation is not
only observable on a daily basis,
but can also be observed on a
seasonal basis.
T
o evaluate the load-wind
gap, we first calculated
average daily wind availability,
Wd,y, during a standard 16-hour
on-peak portion of each day,
7 AM–11 PM, as total wind
generation relative to total
potential generation based on
installed wind capacity, WC,m,y.17
Thus,
16
1 X
W d;y ¼
w
=W C;m;y ; (1)
16 h¼1 h;d;y
where wh;d;y equals hourly
wind generation on day d of
year y. Next, we average these
daily wind availability values in
each season of each year to define
¯ S;y .
seasonal wind availability, W
Thus,
Ds
X
¯ S;y ¼ 1
W :
W
Ds d¼1 d;y
(2)
Similarly, we define the annual
¯ A;y , as the
wind availability, W
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
11
Author's personal copy
average daily wind availability
over year y, or
365
1 X
¯
W :
WA;y ¼
365 d¼1 d;y
(3)
The seasonal wind ratio is
just equal to the ratio of the
seasonal and annual wind
¯ S;y =W
¯ A;y .
availability levels, or W
Next, we define the seasonal load
ratio, L¯S;y , as the average load
during season S of year y relative
to the average annual load in year
y, L¯y . Thus,
Ds L
X
d;y
L¯S;y ¼
;
(4)
¯
Ly
value of subsidized wind
generation, the load-wind gap
should be as large as possible
when load and market prices are
at a maximum. That is, the
economic value of subsidized
wind generation will be
maximized if the relative wind
generation is greatest when loads
are greatest. Intuitively, during
peak demand hours, wind
d¼1
where
365
1 X
L¯y ¼
L
365 d¼1 d;y
(5)
Finally, the load–wind ‘‘gap,’’
GS,y, equals the difference
between the seasonal wind
availability ratio and the seasonal
load ratio:
GS;y ¼
¯ S;y
W
L¯S;y
¯ A;y
W
(6)
For example, suppose the
seasonal load in spring of year y
equals 90 percent of annual
average load, but seasonal wind
generation is 120 percent of
annual average wind generation.
Then the spring load–wind gap,
GSpring,y equals 120–90 percent, or
+30 percent. A positive load-wind
gap value means there is
relatively more wind generation
available to serve load; a negative
load-wind gap value means there
is relatively less wind generation
available to serve load.
rom the standpoint of
maximizing the economic
F
12
generation will displace high-cost
fossil generating units; the greater
the availability of wind power,
the greater will be the cost savings
from displacing fossil-fuel
peaking units.
n contrast, when load and
market prices are low, wind
generation will displace lower
variable-cost baseload resources.
Moreover, when load is
especially low and baseload
resources cannot be cycled, wind
generation will not displace any
generation. Instead, wind will
simply force baseload generation
owners to pay to continue
operating, driving prices below
zero. In such cases, the value of
wind displacement is zero;
subsidized wind generation
I
simply results in a wealth
transfer from existing generation
owners to wind generators and
consumers. Although consumers
may benefit from lower
wholesale prices in the short run
if load-serving entities are
relying on the market, in the
long run, consumers will be
worse off, as demonstrated by
Briggs and Kleit.8
Figures 2–4 illustrate the
seasonal load-wind gaps for
ERCOT, MISO, and PJM.
As Figures 2–4 demonstrate,
however, the economic value of
subsidized wind generation does
not follow this pattern. In each
region, there is a strong lack of
wind generation during the last
four summers, when electricity
demand was greatest. Instead, in
all three regions, the highest
relative amount of wind
generation occurred when loads
were lowest, and the smallest
amounts of wind were available
when loads were greatest in
summer. In PJM, this effect has
been particularly pronounced,
with a summer load – wind gap of
almost 70 percent in summer
2010 and 2011, and 59 percent in
summer 2012.
Although we did not evaluate
wind generation in the Southwest
Power Pool (SPP), which has
about 4,800 MW of installed wind
capacity, the SPP Independent
Market Monitor reports similar
wind output behavior during
peak load hours. In 2011, for
example, wind availability during
all peak hours averaged just over
15 percent, whereas in the hours
where loads were lowest, wind
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
The Electricity Journal
Author's personal copy
[(Figure_2)TD$IG]
Figure 2: ERCOT Load–Wind Gap, 2009–2012
availability averaged over 40
percent.18
ext, we evaluated
availability ratios each
N
year during the hour when
demand peaked on the 10 days
with the highest greatest
electricity demand in each RTO.
We compared the median of the
availability ratios in each year
with the overall median
availability over the entire
[(Figure_3)TD$IG]
Figure 3: MISO Load–Wind Gap, 2009–2012
Jan./Feb. 2013, Vol. 26, Issue 1
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
13
Author's personal copy
[(Figure_4)TD$IG]
Figure 4: PJM Load–Wind Gap, 2009–2012
Table 2: Median Wind Availability, Peak Demand Days and Overall.
Year
ERCOT
2009
14.2%
1.8%
14.6%
2010
6.0%
2.5%
8.2%
2011
2012
15.9%
14.0%
7.6%
7.2%
14.0%
13.8%
Median, All-hours, All years
30.9%
27.0%
25.9%
four-year period, based on the
individual daily availability
ratios.19 The results are shown in
Table 2.
s Table 2 shows, in MISO,
median wind availability
ranged between 1.8 percent and
7.6 percent of total installed wind
capacity at the peak hour on the 10
highest-demand days. In ERCOT,
median wind availability ranged
between 6.0 percent and 15.9
percent. In PJM, the range was
between 8.2 percent and 14.6
percent. As shown, these
availability values are, at best, half
A
14
MISO
PJM
the median availability for the
entire period and, in the case of
MISO, at best less than one-fourth
of the median availability. From a
system planning perspective,
therefore, planners must assume
that little wind generation will be
available on the highest-demand
days.
inally, we examined wind
generation based on its
relation to an average daily load
profile, both seasonally and over
the entire year. This is shown in
Figure 5, which compares average
wind availability by hour in
F
ERCOT to average hourly electric
demand over the entire four-year
period, both in the summer
season and on an average annual
basis.
As Figure 5 shows, average
hourly loads in summer are
higher than during the year
overall, whereas average wind
availability is lower in summer.
Thus, we see the same high-load/
low-wind generation
relationship: high-load hours are
associated with low wind
availability.20
IV. Policy Implications
Our analysis shows that
continued subsidies for wind
generation represent both bad
economics and bad energy policy,
for at least three reasons. First and
foremost, wind generation’s
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
The Electricity Journal
Author's personal copy
[(Figure_5)TD$IG]
Figure 5: Summer Season and Annual Daily Wind Generation and Load Patterns
production pattern not only is
volatile and unpredictable, but
even more significantly, has low
economic value. Rather than
displacing high-variable-cost
fossil generating resources used
to meet peak demand, wind
generation’s observed availability
peaks when electricity demand is
lowest. As a result, wind
generation tends to displace lowvariable-cost generation or simply
forces baseload generators to pay
greater amounts to inject power
onto the grid because the units
cannot be cycled cost-effectively.
The low economic value of wind
power is comparable to the
government paying farmers to
plow under high-value crops in
order to plant low-value ones, or
even weeds.
Second, as with all subsidies,
subsidized wind generation
distorts electric markets by
artificially lowering electric prices
Jan./Feb. 2013, Vol. 26, Issue 1
in the short run, but leads to
higher prices in the long run. This
imposes economic harm on
competitive generators and
consumers, thus reducing
economic growth.
hird, because geographic
dispersion of wind
resources does not address
inaccurate forecasts of wind
availability, additional fossil
generating resources are required
to maintain system reliability.
Moreover, geographic dispersion
requires billions of dollars to be
spent on additional transmission
lines. These costs, along with most
of the system integration costs, are
socialized across all grid
customers, that is, borne by all
generators and, ultimately,
consumers. In other words, wind
generation imposes external costs
on other market participants.
After 35 years of direct and
indirect subsidies, there is no
T
economic rationale for continued
subsidization of wind generation.
At the federal level, direct
subsidies, such as the federal PTC,
should not be continued. Statelevel subsidies, whether feed-in
tariffs established by state
regulators or statutory RPS
mandates, further exacerbate
market distortions and raise
electricity prices, again to the
detriment of consumers.
Ultimately, continued
subsidization of wind generation
simply rewards the few at the
expense of the many. Given a
massive federal debt and anemic
economic recovery, this type of
pernicious redistribution cannot
be justified.&
Endnotes:
1. More recently, payments to the
wind industry have increased still
further with billions of dollars in
additional monies paid-out as part of
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
15
Author's personal copy
the $831 billion American Recovery
and Reinvestment Act of 2009
(ARRA).
2. M. Sherlock, CRS. ‘‘Impact of Tax
Policies on the Commercial
Application of Renewable Energy
Technology,’’ Statement Before the
House Committee on Science, Space,
and Technology, Subcommittee on
Investigations and Oversight &
Subcommittee on Energy and
Environment, April 19, 2012, p. 3.
3. The reasons why are discussed in
the next section.
increase. Moreover, the threat of
intervention raises the expected costs
of market entry, leading to higher
long-run market prices than would
prevail in the absence of subsidies. For
a discussion of subsidies and price
suppression in organized capacity
markets, see Briggs, Robert, and
Andrew N. Kleit, Resource
Adequacy and the Impacts of Capacity
Subsidies in Competitive Electricity
Markets, Working Paper, Dept. of
Energy and Mineral Engineering,
Pennsylvania State University,
4. The ‘‘infant industry’’ argument
historically was used to justify
protection of domestic firms from
international trade. It was first
developed by Alexander Hamilton at
the beginning of the nineteenth
century. A classic article discussing
why infant industries should not be
protected is Robert Baldwin, ‘‘The
Case Against Infant Industry
Protection,’’ Journal of Political
Economy 75 (1969), pp. 295–305.
5. Arguments that subsidies account
for external costs incorrectly assume
that the effects of subsidies and taxes
are equivalent. They are not. See
William Baumol and Wallace Oates,
The Theory of Environmental Policy, 2d
ed., (Cambridge: Cambridge
University Press 1988). See also,
Daniel Dodds and Jonathan Lesser,
‘‘Can Utility Commissions Improve on
Environmental Regulations,’’ Land
Economics 70 (1994), pp. 63–76.
6. There is an extensive literature on
the effects of subsidies in agriculture,
energy, housing, environmental
quality, and so forth. General
discussions on the impacts of
subsidies on markets can be found in
any intermediate microeconomics
textbook.
7. Equivalently, the marginal cost of
cycling the plant is greater than the
variable operating cost. Hence, it is
economically rational to continue
operation.
8. Some argue that price suppression
‘‘benefits’’ consumers. While subsidies
can reduce market prices in the very
short-run, markets are dynamic. Thus,
as competitors are driven out, prices
16
Encourage Wind Energy Predicated
on a Misleading Statistic?’’ The
Electricity Journal 25 (April 2012),
pp. 42–54.
14. Source: SNL Financial. Data
through August 31, 2012.
15. We defined the ‘‘Winter’’ season
contiguously. Thus, for example,
Winter 2012 is defined as the three
months December 2011 through
February 2012.
16. In ERCOT, the correlation
coefficient between hourly wind
availability and annual hourly loads is
0.83. The correlation coefficient for
the Summer season is 0.74.
17. We used wind capacity data as
published by SNL Financial, which
provided wind capacity installed in
each month of the 44-month analysis
period. The capacity used to calculate
wind availability in month m was the
amount of reported capacity installed
at the end of month m 1.
Oct. 22. 2012. http://papers.ssrn.
com/sol3/papers.cfm?abstract_id=
2165412.
9. This value is based on a federal
corporate tax rate of 35%.
10. NREL, Eastern Wind Integration
and Transmission Study, NREL/SR550-47086, Revised February 2011.
http://www.nrel.gov/docs/
fy10osti/47086.pdf.
11. This value is based on total wind
generation of just over 28.2 million
MWh in ERCOT in 2011. According to
the US Energy Information
Administration, total wind generation
was about 120 million MWhs in 2011.
12. Public Utilities Commission of
Texas, Competitive Renewable Energy
Zone Program (CREZ) Oversight,
CREZ Progress Report No. 8, July
2012, p. 6. http://
www.texascrezprojects.com/
page2960039.aspx.
13. Forbes, Kevin, Marco Stampini,
and Ernest Zampelli, ‘‘Are Policies to
18. SPP, Independent Market
Monitor, 2011 State of the Market, July 9,
2012, pp. 59–60. The Independent
Market Monitor reports that similar
wind availability patterns—
decreasing availability as load
increased—were observed in the three
previous years.
19. The median was selected as a more
representative planning value for the
data. Consider a simple (albeit
extreme) example: suppose wind
availability was 0% on nine of the ten
days, but 100% on the 10th day. In that
case, the median availability would be
0% and the average availability would
be 10%. However, system planners
who assumed 10% wind availability
each day in order to schedule
generating resources would have to
rely on replacement generation on nine
of the days, and be forced to back down
generation on the 10th. In contrast,
using the median availability, system
planners would only have to back
down generation on the 10th day.
20. The correlation coefficient
between average annual hourly wind
availability and average annual hourly
load is 0.83. The correlation
coefficient for the Summer season
is 0.74.
1040-6190/$–see front matter # 2012 Elsevier Inc. All rights reserved., http://dx.doi.org/10.1016/j.tej.2012.11.015
The Electricity Journal
The Economic Impacts of U.S. Shale Gas Production on Ohio Consumers Prepared by: Continental Economics, Inc. January 2012 Copyright © 2012, Continental Economics, Inc.
The information contained in this document is the exclusive, confidential and proprietary property of Continental
Economics, Inc. and is protected under the trade secret and copyright laws of the U.S. and other international laws,
treaties and conventions. No part of this work may be disclosed to any third party or used, reproduced or transmitted
in any form or by any means, electronic or mechanical, including photocopying and recording, or by any
information storage or retrieval system, without prior express written permission of Continental Economics, Inc.
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
TABLE OF CONTENTS
Executive Summary
I.
Introduction ........................................................................................................................... 1
II. Historical Overview ............................................................................................................... 2
A. The Emergence of Shale Gas ............................................................................................... 5
B. Overview of Shale Gas Production Economics ................................................................... 8
C. Natural Gas Prices and Demand ........................................................................................ 10
D. Trends in U.S. and Ohio Natural Gas Demand .................................................................. 12
E. Natural Gas Prices and Ohio Consumers’ Energy Costs. .................................................. 13
III. Estimating the Impacts of Shale Gas Production on U.S. Wellhead Natural Gas Prices
and Ohio Consumers’ Energy bills .................................................................................... 14
A. The Impacts of Shale Gas on Wellhead Natural Gas Prices .............................................. 14
B. Impacts on Ohio Natural Gas Consumers .......................................................................... 17
Appendix 1: Econometric Model Specification
The Economic Impacts of Shale Gas on Ohio Consumers
EXECUTIVE SUMMARY
January 2012
†
Shale gas has fundamentally changed the U.S. energy picture, providing a boon in an otherwise
moribund economy. A decade ago, shale gas and liquids production were inconsequential. As the
gas supply “bubble” of the 1990s ended and crude oil prices accelerated, so did wellhead natural
gas prices, because of the historic linkage between the prices of the two fuels. In 2005, the
damage caused by Hurricanes Katrina and Rita to the U.S. Gulf natural gas supply infrastructure
caused a further spike in wellhead prices, and concerns grew that natural gas prices would
continue to escalate.
As shale gas production has accelerated, U.S. natural gas prices have plummeted. Although the
severe economic recession that began in late 2008 and the resulting decrease in the demand for
natural gas have contributed to lower wellhead natural gas prices, much of that price decrease
stems from the rapid increase in domestic shale gas supplies, which increased almost tenfold
between 2005 and 2010.
The rapid expansion of shale gas production in the United States has created hundreds of
thousands of new jobs directly and in supporting industries. The effect of this expansion on
people and communities within the geographic areas of the shale plays has received considerable
attention. However, domestic shale gas developments have also been the catalyst for far broader
economic benefits throughout the country. More specifically, the lower wellhead natural gas
prices that have resulted from this expanding shale gas production have lowered businesses’ and
consumers’ energy bills, not only for natural gas, but also for electricity, an increasing
percentage of which is generated from natural gas. Without seeking to divert attention away
from the important economic development and retention benefits that shale gas development has
had or will have on local populations and communities, this report provides information about
the broader beneficial dividends that shale development is paying to the public at large.
While conventional natural gas production in the U.S. has decreased over time, shale gas has
become a rapidly increasing source of U.S. gas supplies, accounting for about 20 percent of total
U.S. onshore domestic natural gas production in 2010. The U.S. Energy Information
Administration (“EIA”) forecasts that, by 2035, shale gas could account for over 50 percent of
onshore natural gas production.
Of greater interest for Ohioans is the Utica Shale, which lies beneath the better-known Marcellus
Shale, and extends into the eastern half of the state. Although reserve data is based on
preliminary drilling in the Utica Shale, geologists expect the Utica Shale to be relatively rich in
†
Funding for this report was secured through the Industrial Energy Users-Ohio (IEU-Ohio), an Ohiobased organization of customers that helps customers address issues affecting the price and
availability of energy. Information on IEU-Ohio is available at http://ieu-ohio.org
EX-1
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
oil and natural gas liquids that are currently worth significantly more than natural gas on an
energy-equivalent basis. Preliminary estimates by Ohio's Department of Natural Resources
(ODNR) suggest a recoverable reserve potential of between 1.3 and 5.5 billion barrels of oil as
well as 3.8 to 15.7 trillion cubic feet (“Tcf”) of natural gas. The overall economic value of the
Utica Shale region in Ohio may be especially large, because it lies relatively close to the surface,
which reduces exploration and development costs.1
Although overall natural gas consumption in Ohio has decreased since 1997 (in part because of
reductions in the energy intensity of Ohio’s economy), expenditures on natural gas remain
significant. In 2009, Ohio consumers and businesses, including electric generators, consumed
724 billion cubic feet (“Bcf”) of natural gas, at a cost of $7.46 billion. Thus, lower natural gas
prices owing to shale gas production can have real benefits for Ohio energy consumers as well as
the public at large.
To estimate how much shale gas has contributed to the decline in wellhead natural gas prices and
how those price decreases have flowed through to benefit Ohio’s natural gas consumers,
Continental Economics developed a model to isolate the impacts of shale gas on wellhead prices.
Then, using the results of that model, we determined the savings to different classes of Ohio
consumers.2
Our analysis showed that, for each Tcf of shale gas produced, the average annual wellhead price
is $0.46 per thousand cubic feet (“Mcf”) lower that it otherwise would be. Equivalently, the
average wellhead price would be $0.46 per Mcf higher for each Tcf of shale gas not otherwise
produced. The results of our analysis are shown in Figure EX-1 and Table EX-1 on the following
page.
As the table shows, the impact of shale gas production on wellhead gas prices has increased
steadily as shale gas supplies have increased relative to total natural gas supplies. For example,
in 2010 we estimate that shale gas production, which was over 4.7 Tcf, caused observed average
wellhead natural gas prices to be $2.43 per Mcf lower than what they would have otherwise
been.
1
The depth of shale gas deposits below the surface is not uniform. All other things equal, the closer to
the surface, the lower are exploration and development costs.
2
We will address the impacts of lower wellhead natural gas prices on wholesale and retail electricity
prices for Ohio consumers in a subsequent report.
EX-2
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Figure EX-1: Estimated Annual Wellhead Natural Gas Prices Without Shale Gas
(1990–2010)
Table EX-1: Estimated Annual Price Impact of Shale Gas Production (1990-2010)
Year
Price Reduction
($/Mcf)
Year
Price Reduction
($/Mcf)
1990
($0.01)
2001
($0.13)
1991
($0.02)
2002
($0.17)
1992
($0.02)
2003
($0.20)
1993
($0.03)
2004
($0.24)
1994
($0.04)
2005
($0.30)
1995
($0.05)
2006
($0.43)
1996
($0.07)
2007
($0.70)
1997
($0.09)
2008
($1.14)
1998
($0.09)
2009
($1.68)
1999
($0.09)
2010
($2.43)
2000
($0.11)
Based on the results of the analysis described above and average use per customer data for 2010,
Table EX-2 provides an estimate of the resulting natural gas energy bill reductions for Ohio
commercial, industrial, and residential customers.
EX-3
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Table EX-2: Estimated Annual Cost Savings for Ohio End-Use Customers
Customer Class
Commercial
Industrial
Residential
Total
Average Use Per Customer (Mcf)
Price Reduction ($/Mcf)
2010 Estimated Cost Savings
Number of Customers
Estimated Savings
(Millions of $)
562.1
($2.43)
$1,366
258,422
$353.0
35,266.8
($2.43)
$85,698
5,738
$491.7
88.2
($2.43)
$214
3,198,883
$685.4
$1,530.2
As this table shows, we estimate that Ohio businesses and consumers saved over $1.5 billion on
their natural gas bills in 2010 because of lower wellhead natural gas prices. The average
residential customer, for example, burned 88 Mcf of natural gas and saved $214 in 2010. The
average commercial customer used 562 Mcf and saved $1,366, while the average industrial
customer used over 35,000 Mcf and saved almost $87,000. In addition, electric generators
reduced their costs because of lower wellhead gas prices. This translated into lower fuel charges
levied by electric utilities with fuel cost recovery mechanisms,3 such as Columbus Southern
Power and Ohio Power Company, and also contributed to lower wholesale electric prices paid by
retail electric suppliers.
The results of our analysis demonstrate that shale gas production has significantly reduced U.S.
wellhead natural gas prices and reduced Ohio consumers’ natural gas bills. The estimated
savings of $1.5 billion on natural gas bills alone in 2010 affect all sectors of the Ohio economy.
As Ohio’s Utica Shale gas resource is developed, Ohio businesses and consumers are likely to
benefit even more in the future. Furthermore, because of the increasing importance of natural gas
used in generating electricity, Ohio consumers are reaping even more benefits from lower
electric bills. (In a subsequent report, we will present the estimated savings for Ohio consumers
on their electric bills.)
The decreases in natural gas and electricity prices will benefit the Ohio economy, not only by
creating jobs directly in the shale gas development and extraction industries as the Utica Shale is
developed, but by lowering home energy bills and improving the overall competitiveness of Ohio
businesses and industry.
3
The default generation supply prices of Ohio Power and Columbus Southern Power (sometimes
referred to as AEP-Ohio) continue to be administratively set by the Public Utilities Commission of
Ohio (“PUCO”) based on a rate structure that includes a fuel adjustment clause (FAC). Other Ohio
electric distribution utilities (“EDUs”) establish default generation supply prices through a
competitive bidding process (“CBP”) conducted under the PUCO’s supervision. The downward
pressure that shale gas development has placed on electric prices is observable from the inputs that go
into the FAC as well as the pricing results of the CBPs that have been approved by the PUCO.
EX-4
The Economic Impacts of U.S. Shale Gas Production
on Ohio Consumers†
I.
INTRODUCTION
Shale gas has fundamentally changed the U.S. energy picture, providing a boon in an otherwise
moribund economy. A decade ago, shale gas and liquids production were inconsequential. As the
gas supply “bubble” of the 1990s ended and crude oil prices accelerated, so did wellhead natural
gas prices, because of the historic linkage between the prices of the two fuels. In 2005, the
damage caused by Hurricanes Katrina and Rita to the U.S. Gulf natural gas supply infrastructure
caused a further spike in wellhead prices, and concerns grew that natural gas prices would
continue to escalate.
As shale gas production has accelerated, U.S. natural gas prices have plummeted. Although the
severe economic recession that began in late 2008 and the resulting decrease in the demand for
natural gas have contributed to lower wellhead natural gas prices, much of that price decrease
stems from the rapid increase in domestic shale gas supplies, which increased almost tenfold
between 2005 and 2010.
The rapid expansion of shale gas production in the United States has created hundreds of
thousands of new jobs directly and in supporting industries. The effect of this expansion on
people and communities within the geographic areas of the shale plays has received considerable
attention. However, domestic shale gas developments have also been the catalyst for far broader
economic benefits throughout the country. More specifically, the lower wellhead natural gas
prices that have resulted from this expanding shale gas production have lowered businesses’ and
consumers’ energy bills, not only for natural gas, but also for electricity, an increasing
percentage of which is generated from natural gas. Without seeking to divert attention away
from the important economic development and retention benefits that shale gas development has
had or will have on local populations and communities, this report provides information about
the broader beneficial dividends that shale development is paying to the public at large.
†
Funding for this report was secured through the Industrial Energy Users-Ohio (IEU-Ohio), an Ohiobased organization of customers that helps customers address issues affecting the price and
availability of energy. Information on IEU-Ohio is available at http://ieu-ohio.org.
-1- The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Much attention has been paid to the jobs created by the shale gas industry, both directly and
indirectly.1 Much less has been focused on these broader economic benefits provided by shale
gas stemming from lower natural gas and electricity prices.2 The purpose of this report,
therefore, is two-fold. First, we estimate the magnitude of the decrease in wellhead natural gas
prices that has been caused by increased shale gas production. Second, we estimate how this
decrease in wellhead natural gas prices has reduced natural gas expenditures by Ohio businesses
and consumers.3
II.
HISTORICAL OVERVIEW
In the late 1960s, the conventional wisdom was that natural gas supplies would soon be
exhausted. Wellhead natural gas prices were regulated and capped. Supplies began to diminish as
the incremental cost of production exceeded the revenue available from regulated prices and
production from existing wells declined. Growth in the natural gas industry came to a standstill
because there was little economic incentive to undertake new, more costly exploration. By 1967,
estimated domestic reserves had peaked and actual production began to fall steadily. Natural gas
supply shortages on peak usage days began to occur. As these gas supply shortages became more
common, in states like Ohio new customer hookups were unavailable, supplies for industrial and
commercial customers were interrupted and curtailed, and predictions that “the spigot would run
dry” within a decade became prevalent.
By 1978, proved natural gas reserves had dropped by 30%. Something had to be done and policy
makers turned to market-based strategies to balance natural gas supply and demand. First, as part
of comprehensive energy legislation that year, Congress passed the Natural Gas Policy Act
(“NGPA”), which began to dismantle the complex historic system of natural gas price regulation
that stemmed from a 1954 decision by the United States Supreme Court decision regarding the
meaning of the Natural Gas Act passed in 1938.4 For example, Ohio initiated the Natural Gas
1
See, e.g., Kleinhenz and Associates, “Ohio‘s Natural Gas and Crude Oil Exploration and Production
Industry and the Emerging Utica Gas Formation,” (Sept., 2011).
http://www.oogeep.org/downloads/file/Economic%20Impact%20Study/Ohio%20Natural%20Gas%20and
%20Crude%20Oil%20Industry%20Economic%20Impact%20Study%20September%202011.pdf. For a
different viewpoint on job creation impacts, and the costs and benefits of shale gas development in
Ohio, see A. Weinstein and M. Partridge, “The Economic Value of Shale Natural Gas in Ohio,”
Department of Agricultural, Environmental, and Development Economics, Ohio State University,
December 2011. http://go.osu.edu/shalejobs.
2
In this report, we do not analyze employment impacts, which are not benefits per se. However,
businesses and consumers clearly do benefit from lower energy prices.
3
We will address the impacts of lower wellhead natural gas prices on wholesale and retail electricity
prices in a subsequent report.
4
The Natural Gas Act (“NGA”) of 1938 allowed the Federal Power Commission (“FPC”), the
precursor to the Federal Energy Regulatory Commission (“FERC”), to regulate the prices charged by
interstate natural gas pipelines. From 1938 to 1954, the FPC did not regulate wellhead natural gas
prices. Instead, independent producers sold natural gas to interstate pipelines at unregulated prices,
-2-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Self-Help Program5 that allowed retail customers to develop and obtain their own natural gas
supply and then use the unbundled delivery capabilities of natural gas companies to transport the
gas supply from the wellhead to the point of utilization. The NGPA provided for the gradual
elimination of a labyrinth of price rules, and full decontrol of prices was achieved by 1993.6
Not surprisingly, eliminating price controls and creating a truly competitive market for natural
gas supplies created the economic incentives needed for renewed natural gas exploration and
development. Coupled with FERC’s severing of the traditional connection between production,
pipeline transportation, and distribution in 1992,7 by the early 1990 the natural gas market was
vibrant; the predicted shortages had turned into a gas supply “bubble” that led to much lower
prices. Those lower prices, in turn, spurred development of competitive wholesale electricity
markets that were envisioned under the Energy Policy Act of 1992, as well as calls for electric
industry restructuring, because of advances in gas-fired generating technologies, such as
combined-cycle units that were energy efficient and could be constructed more quickly than
traditional coal-fired or nuclear baseload generating plants. As a result of the increasing reliance
on natural gas-fired generation, the demand for natural gas has increased since the mid-1990s.
Between 1997 and 2010, for example, total natural gas consumption increased by about six
percent, whereas natural gas consumption for electric power generation increased by 80%.8
__________________________
(cont.) with any subsequent sales for resale being regulated by the FPC. Advocates who sought to keep
consumer prices low through price regulation went to the FPC to close what they alleged was a
regulatory “loophole,” because the FPC exempted wellhead sales from price regulation as
“production and gathering” activities. The FPC rejected this claim, but in 1954, the U.S. Supreme
Court, in Phillips Petroleum v. Wisconsin, 347 U.S. 672 (1954), reversed the FPC, ruling that the
NGA applied not only to pipelines, but also to natural gas producers. This led to the FPC regulating
natural gas wellhead prices and creating the natural gas “shortages” beginning in the late 1960s.
5
Ohio’s Self-Help Program was one of the first unbundled natural gas transportation programs in the
Nation and during the 70’s it sparked a surge of exploration and development activities in Ohio. It
also served as a model for other unbundled open access transportation programs that evolved to
provide the foundation for the natural gas industry structure that is in place today.
6
In 1989, Congress passed the Natural Gas Wellhead Decontrol Act, which fully decontrolled
wellhead prices as of January 1, 1993.
7
Pipeline Service Obligations and Revisions to Regulations Governing Self-Implementing
Transportation Under Part 284 of the Commission’s Regulations, Regulation of Natural Gas
Pipelines After Partial Wellhead Decontrol, FERC Order No. 636, 59 FERC ¶ 61,030 (1992).
8
Source: U.S. Energy Information Administration, Natural Gas Annual, 2011.
-3-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
The Many Sources of Natural Gas
Natural gas is produced from a number of sources. The diagram below is a schematic of the
different types of gas and their relative locations underground.
Nearest the surface is coalbed methane, which is just natural gas found in coal seams.9
Associated gas is natural gas that is found on top of crude oil deposits. Often, crude oil wells
produce both crude oil, natural gas, and so-called “natural gas liquids” (“NGLs”), which are
valuable types of hydrocarbons, such as propane and butane. In other cases, natural gas is found
in separate deposits, called non-associated gas. Further below the surface, one finds “tight-gas.”
Tight gas is natural gas that has migrated upwards into sandstone formations and which, because
sandstone has low permeability,10 cannot migrate further. Further below still lies shale gas.
Although market forces had eliminated fears of natural gas shortages, wellhead prices remained
linked to world crude oil prices. Thus, when the events of September 11, 2001, and the
subsequent invasions of Afghanistan and Iraq, led to a rapid increase in crude oil prices, natural
gas prices followed; the gas supply “bubble” had burst. Natural gas prices increased, spiking in
2005 because of the damage caused by Hurricanes Katrina and Rita to the Gulf Coast gas supply
9
Coal seams are often saturated with water, and the pressure of that water forces methane (natural gas)
into the coal. When the water is removed, the pressure drops, and natural gas can be extracted.
10
“Low permeability” means that gas molecules do not flow easily. For shale gas, that is the reason
producers use hydraulic fracturing techniques to release the natural gas and let it easily flow to their
wells.
-4-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
infrastructure, and causing renewed fears that U.S. natural gas production once again faced
inexorable decline and that consumers would face increasing prices.11
The search was on for new sources of natural gas. One of the first was liquefied natural gas
(LNG) that could be imported from the Middle East, where gas was still considered a waste byproduct from crude oil production and simply flared (burned) off. Plans for huge new facilities
capable of receiving LNG were developed, but many such facilities faced intense siting
opposition because of the perceived risks, such as explosions.
Other “unconventional” domestic natural gas supplies also emerged. By the mid-1990s, for
example, production of coal-bed methane (“CBM”) had increased to about 1 trillion cubic feet
(“Tcf”) per year, or about four percent of total U.S. natural gas production. In 2008, coal-bed
methane production peaked at just under 2 Tcf. The other unconventional resource—and the one
that has already provided huge economic benefits—is shale gas, which now accounts for over 20
percent of all total domestic natural gas production.
A. The Emergence of Shale Gas
By the late 1970s, natural gas was already known to exist in deep shales, such as the Barnett in
Texas and Marcellus in Pennsylvania (Figure 1).
Figure 1: U.S. Shale Gas Plays
11
The same was true for natural gas supplies exported to the U.S. from western Canada, which had also
increased over the previous decade.
-5-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
However, the technology to retrieve gas from such “low-permeability” areas did not exist. It was
not until the 1980s that improvements in hydraulic fracturing, a drilling technique that had been
widely used since the 1940s to enhance production in existing oil and gas wells, began to change
the economics of shale gas.
The U.S Energy Information Administration (“EIA”) has tracked shale gas production for each
of the major shale gas plays since 1990 (Figure 2). As can be seen in Figure 2, shale gas
production began to accelerate rapidly after the year 2000, as production ramped up in the
Barnett shale of Texas. By the middle of the decade, production in the Fayette and Haynesville
regions began to increase. Most recently, production in the Marcellus region, which is estimated
to have far larger reserves, has begun to accelerate.
Figure 2: U.S. Annual Shale Gas Production, (1990 – 2010)
-6-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Figure 3: U.S. Annual Natural Gas Production, by Source (1990–2010)
While conventional natural gas production in the U.S. has decreased over time, shale gas has
become a rapidly increasing source of U.S. gas supplies (Figure 3), and now accounts for about
20 percent of total U.S. onshore domestic natural gas production. The EIA forecasts that, by
2035, shale gas could account for over 50 percent of onshore natural gas production.12
Of greater interest for Ohioans is the Utica Shale, which lies beneath the better-known Marcellus
Shale, and extends into the eastern half of the state. Although reserve data is based on
preliminary drilling in the Utica Shale, geologists expect the Utica Shale to be relatively rich in
oil and natural gas liquids that are currently worth significantly more than natural gas on an
energy-equivalent basis. Preliminary estimates by Ohio's Department of Natural Resources
(ODNR) suggest a recoverable reserve potential of between 1.3 and 5.5 billion barrels of oil as
well as 3.8 to 15.7 trillion cubic feet (“Tcf”) of natural gas. The overall economic value of the
Utica Shale region in Ohio may be especially large, because it lies relatively close to the surface,
which reduces exploration and development costs.13
12
http://www.eia.gov/forecasts/aeo/IF_all.cfm#prospectshale.
13
The depth of shale gas deposits below the surface is not uniform. All other things equal, the closer to
the surface, the lower are exploration and development costs.
-7-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Figure 4: Marcellus and Utica Shale Plays in Ohio
B. Overview of Shale Gas Production Economics
As we discuss in the next section, shale gas production has helped reduce wellhead natural gas
prices. But what factors affect shale gas production? It turns out, there are a number of factors,
including not only day-to-day production costs, but also the costs of leasing land, the
productivity of the wells drilled, and the mix of natural gas and NGLs produced.
The economic benefits of drilling an oil or gas well—and, often, the same well produces both—
depend on a number of factors. Broadly, these are the expected future revenues from what the
well produces, and the fixed and variable production costs of drilling and operating the well.
Expected future revenues depend on how much a typical well is likely to produce over its
lifetime and future prices. For example, wells that produce both crude oil and NGLs tend to be
more profitable than wells producing just natural gas, given current and expected prices. The
reason is that world crude oil prices are much higher (on a Btu basis, i.e., the price per million
Btus, based on the relative heat content of oil and natural gas) than the price of natural gas.
Similarly, some NGLs, such as propane and butane, tend to sell at higher market prices than
methane, which is the major component of what we term “natural gas.” Thus, all else equal, a
developer is more likely to drill where natural gas is likely to be found with crude oil and natural
gas liquids.
-8-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Production costs can be broken down into fixed and variable costs. Fixed costs are those that do
not change with the quantity of oil or gas produced, such as the cost of obtaining a lease. For
example, recent data shows that leases for land in Athens County are costing developers $2.500
per acre.14 The costs of these leases, together with the cost of leasing the actual drilling
equipment, are the largest fixed costs associated with well-drilling.
Variable costs of production are those that depend on how much oil and natural gas is produced.
For example, the state of Ohio levies a severance tax on crude oil and natural gas producers
of$0.10 per barrel of oil and $0.0025 per thousand cubic feet of natural gas produced, regardless
of the market prices.15 Landowners typically assess a royalty fee on producers that, unlike the
state severance tax, is based on the value of natural gas produced. Finally, there are the direct
variable production costs, such as the cost of operating the well equipment every day. Because a
significant portion of the overall production costs are fixed, drillers will often continue to
produce oil and natural gas from wells even when the average production costs are greater than
market prices, which can tend to further decrease market prices.
Estimates of the overall average production cost of shale gas wells vary widely, because shale
gas plays differ in their characteristics, such as depth. Typically, drilling costs are reported on a
per-foot basis. Thus, the equivalent cost per MMBtu of natural gas produced depends on how
deeply a well is drilled, and the well’s average daily production.
Publicly available data on the costs of shale gas wells, and production costs per MMBtu, are
difficult to obtain. Moreover, because of technological advances, production costs continue to
decrease. A 2010 report by the World Energy Council states that estimates of average shale gas
production costs in North America range between $4 per Mcf and $8 per Mcf.16 However, in
2010, Chesapeake Energy estimated average direct production expenses, including taxes, of just
over $1 per Mcf.17 It reports another $0.44 per Mcf in administrative and general costs, and
$1.56 per Mcf in depreciation and amortization costs, for an overall average cost of about $3 per
Mcf. Moreover, because shale gas resources tend to be located near demand centers,
transportation costs on natural gas pipelines can be less than for natural gas sourced from
traditional supply basins, such as the Rocky Mountains, Western Canada, and the Gulf Coast.
14
“County oil and gas leasing just goes on & on,” The Athens (OH) News, December 15, 2011.
15
http://codes.ohio.gov/orc/5749. One thousand cubic feet (“Mcf”) is approximately 1.04 million Btus
(MMBtu). One barrel of oil has an average heat content of 5.6 MMBtus. Thus, on a per-Btu basis,
Ohio levies a slightly higher severance tax on oil than natural gas.
16
World Energy Council, “Survey of Energy Resources: Focus on Shale Gas,” 2010, page 14.
http://www.worldenergy.org/documents/shalegasreport.pdf
17
Chesapeake Energy, 2010 Annual Report, page 4.
http://phx.corporateir.net/External.File?item=UGFyZW50SUQ9OTEzODB8Q2hpbGRJRD0tMXxUeXBlPTM=&t=1
-9-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
C. Natural Gas Prices and Demand
The gas supply “bubble” of the 1990s caused an extended period of low natural gas prices, with
prices generally less than $2 per Mcf (Figure 5). Starting in 2001, however, gas prices, which
were historically linked closely with crude oil prices, began to increase rapidly, in response to
the events of September 11, 2001, and the subsequent invasions of Afghanistan and Iraq, which
caused crude oil prices to increase rapidly (Figure 6). Wellhead natural gas prices peaked in 2005
at an average of over $7 per Mcf, in part because of the damage to the production and gathering
infrastructure along the U.S. Gulf Coast caused by Hurricanes Katrina and Rita, and continued
increases in natural gas demand, especially for generating electricity. Prices then decreased to
about $6 per Mcf in 2006 and 2007. However, in 2008, prices spiked to their highest annual level
ever, about $8 per Mcf, caused by increased demand and surging crude oil prices.18
Figure 5: Average Annual U.S. Wellhead Natural Gas Prices (1990–2010)
The rapid increase in U.S. shale gas production has more than compensated for decreases in
conventional natural gas production from oil and gas wells, because advances in drilling
technology have made the economics of shale gas production so favorable. In fact, according to
18
The June 2008 wellhead price was $10.79 per Mcf, the highest nominal value ever. Historically, U.S.
natural gas prices and crude oil prices were closely linked, owing to the substitutability of oil and
natural gas. As discussed below, because of shale gas, that historic link is much weaker.
-10-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
IHS CERA, the cost of producing shale gas is now less than the cost of producing
“conventional” natural gas.19
Another important benefit of the rapid increase in shale gas production has been to weaken the
historical link between wellhead natural gas prices and volatile world crude oil prices. Because
of the limited ability to export natural gas overseas, increased domestic production has reduced
wellhead natural gas prices, even though world crude oil prices remain in the $100 per barrel
(“Bbl”) range.20 For example, between January 2009 and August 2011, the price of Brent crude
(one of the world’s benchmark oil prices) tripled, rising from under $40 per Bbl to almost $120
per Bbl. During that same period, U.S. wellhead natural gas prices remained relatively constant
(Figure 6).21 Thus, while natural gas prices and crude oil prices in Europe and Asia continue to
move in tandem, the weakening of the historic link in the U.S. is providing consumers with the
economic benefits of lower and more stable natural gas prices.
19
“Fueling America’s Energy Future,” IHS Cambridge Energy Research Associates Report, 2010. A
copy of the Executive Summary may be found at:
http://groundwork.iogcc.org/sites/default/files/IHS%20CERA%20Executive%20Summary%20Fuelin
g%20North%20America%27s%20Energy%20Future%20The%20Unconventional%20Natural%20Ga
s%20Revolution%20and%20the%20Carbon%20Agenda.pdf.
20
LNG facilities were developed to import natural gas, not export it. Today, some of these facilities are
being modified so that natural gas can be delivered to them, and then exported, but the cost is high.
21
A portion of the increase in the Brent crude price can be explained by decreases in the value of the
U.S. dollar relative to other major world currencies. According to data published by the U.S. Federal
Reserve, between January 2009 and August 2011, the dollar declined in value by 15% relative to
these other currencies.
-11-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Figure 6: U.S. Monthly Wellhead Natural Gas Prices and Brent Crude Price (2009–2011)
D. Trends in U.S. and Ohio Natural Gas Demand
The increase in the production of shale gas has also outstripped the overall growth in the demand
for natural gas (Figure 7). This has also contributed to the decrease in wellhead natural gas prices
since 2008 and provided greater price certainty. As shown in Figure 7, between 1997 and 2010,
total natural gas delivered to domestic customers increased from 20.8 Tcf to 22.1 Tcf. Between
1997 and 2010, residential and commercial sector demand remained essentially constant, despite
increases in the number of customers, because of increases in the energy efficiency of space and
water heating equipment. Industrial demand during this same time period decreased by just over
22%, owing to a general decrease in manufacturing output, as well as improved energy
efficiency. However, the demand for natural gas by electric generators increased over 80%
between 1997 and 2010, reflecting the rapid increase in gas-fired generating capacity.
-12-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Figure 7: U.S. Annual Natural Gas Deliveries, by Customer Class
Unlike the U.S. as a whole, natural gas consumption in Ohio decreased during this period,
especially in the industrial sector, where demand fell by about one-third between 1997 and 2009,
reflecting the loss of heavy manufacturing industry in the state during this period.
E. Natural Gas Prices and Ohio Consumers’ Energy Costs
Although overall natural gas consumption in Ohio has decreased since 1997 (in part because of
reductions in the energy intensity of Ohio’s economy), expenditures on natural gas remain
significant. In 2009, Ohio consumers and businesses, including electric generators, consumed
724 billion cubic feet (“Bcf”) of natural gas, at a cost of $7.46 billion, as shown in Table 1.22
Thus, lower natural gas prices owing to shale gas production can have real benefits for Ohio
energy consumers as well as the public at large.
22
Source: EIA, State Energy Data System.
http://www.eia.gov/state/seds/seds-states.cfm?q_state_a=OH&q_state=Ohio.
-13-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Table 1: Summary of Ohio 2009 Natural Gas Consumption and Expenditures
Sector
Consumption
Cost Average Price
(Bcf)
(Millions of $)
($/Mcf) Residential
292
$3,708
$12.70
Commercial
161
$1,676
$10.41
Industrial
233
$1,908
$8.19
38
$166
$4.36
724
$7,458
$10.30
Electric Power
Totals
Of the total expenditure, residential customers spent $3.7 billion, about half of the total, and paid
an average delivered retail price of $12.70 per Mcf. That delivered price includes the wholesale
cost of gas, which reflects the wellhead price, plus the cost of transportation via pipeline and the
cost of retail distribution. The reason that the average price for electric generators is so much
lower than other customers is that most electric generators are directly interconnected to
interstate pipelines, thus avoiding all of the costs associated with retail distribution.
III. ESTIMATING THE IMPACTS OF SHALE GAS PRODUCTION ON U.S.
WELLHEAD NATURAL GAS PRICES AND OHIO CONSUMERS’ ENERGY
BILLS
Shale gas has contributed to the decline in wellhead natural gas prices, but by how much? And,
how does the wellhead price decrease caused by shale gas translate to savings for Ohio natural
gas consumers? To answer these questions, we developed a model to isolate the impacts of shale
gas on wellhead prices. Then, using the results of that model, determined the savings to different
classes of Ohio consumers.
A. The Impacts of Shale Gas on Wellhead Natural Gas Prices
Natural gas supplies reflect complex relationships between expectations of future demand,
market prices, and technology. Moreover, because significant quantities of natural gas are
produced from oil wells, supplies are also influenced by expectations about crude oil markets.
The U.S. EIA, for example, uses a complex set of interdependent models to prepare forecasts of
natural gas production and prices.23 The EIA models combine engineering relationships, such as
exploration costs per drilled foot, with econometric models and economic projections, to
determine the economic returns from exploration and development. These factors further interact
with demand projections, which are based on macroeconomic forecasts of the U.S. and world
economies.
23
The EIA crude oil and natural gas supply model (“OGSM”) is part of its larger National Energy
Modeling System (“NEMS”). Documentation for the OGSM can be downloaded at:
http://www.eia.gov/FTPROOT/modeldoc/m063(2011).pdf
-14-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
For the purposes of this report, however, it would be difficult to modify this type of modeling
approach to determine what historic natural gas prices would have been without production from
shale gas. Thus, we developed an econometric framework that models annual natural gas supply,
demand, and average wellhead prices, and which isolates the impacts of shale gas production on
wellhead prices. (A detailed description of the model can be found in the Appendix.) The
advantages of an econometric approach include its relative transparency: factors that influence
natural gas supply and demand, such as the price of crude oil and the delivered price of coal used
by electric generators, are easily modeled and evaluated. The disadvantages of the econometric
framework used here is that it cannot incorporate all of the variables that affect natural gas
supply and demand.24
Once the model was estimated, we evaluated how well it predicted wellhead natural gas prices
(Figure 8). As this figure shows, the model predicted natural gas prices that closely follow the
actual annual prices.
Figure 8: Actual v. Predicted Wellhead Natural Gas Price (1990 – 2010)
To estimate the impact of shale gas production on average wellhead prices, we used the
estimated relationship between wellhead prices and the quantity of natural gas produced.
Specifically, our analysis showed that, for each Tcf of shale gas produced, the average annual
24
In economic terms, we have developed a “partial equilibrium” model, rather than a “general
equilibrium” one.
-15-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
wellhead price would be $0.46 per Mcf lower. Equivalently, the average wellhead price would
be $0.46 per Mcf higher for each Tcf of shale gas not otherwise available. The results of the
analysis are shown in Figure 9 and Table 2.
Figure 9: Estimated Annual Wellhead Natural Gas Prices Without Shale Gas
(1990 – 2010)
Table 2: Estimated Annual Price Impact of Shale Gas Production (1990-2010)
Year
Price Reduction
($/Mcf)
Year
Price Reduction
($/Mcf)
1990
($0.01)
2001
($0.13)
1991
($0.02)
2002
($0.17)
1992
($0.02)
2003
($0.20)
1993
($0.03)
2004
($0.24)
1994
($0.04)
2005
($0.30)
1995
($0.05)
2006
($0.43)
1996
($0.07)
2007
($0.70)
1997
($0.09)
2008
($1.14)
1998
($0.09)
2009
($1.68)
1999
($0.09)
2010
($2.43)
2000
($0.11)
-16-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
As Table 2 shows, the impact of shale gas production on wellhead gas prices has increased
steadily as shale gas supplies have increased relative to total natural gas supplies. In 2010, for
example, we estimate that shale gas production, which was over 4.7 Tcf, caused observed
average wellhead natural gas prices to be $2.43 per Mcf lower than what they would have
otherwise been.
To gauge the reasonableness of this price impact, we reviewed a 2004 analysis prepared by the
EIA at the request of Representative Barbara Cubin, Chairman of the Subcommittee on Energy
and Mineral Resources of the U.S. House Committee on Resources. The EIA analysis examined
the projected impacts on U.S. natural gas production and wellhead prices under three different
“low-supply” scenarios, and a combination of all three scenarios:25

No increased availability of Alaska natural gas;

No significant increase in production of tight sands natural gas (or other nonconventional
sources); and

Inability to permit more than three additional average-sized liquefied natural gas offloading facilities.
The EIA study estimated that, in 2010, the combination of these three restrictive supply
assumptions would increase the average wellhead price of natural gas by $0.47 per Mcf (2002$)
and reduce production in the lower 48 states by 0.96 Tcf. This implies an average increase of
$0.49 (2002$) per Tcf of reduced production. Adjusting for inflation to 2010 dollars, this
translates into an average price impact of $0.59 per Mcf for each Tcf reduction in natural gas
production in the lower 48 states. As discussed above, we estimated a somewhat smaller price
impact, $0.46 per Mcf for each Tcf reduction in gas supplies.
B. Impacts on Ohio Natural Gas Consumers
Natural gas retail distribution customers typically pay for natural gas on a pass-through basis.
Thus, if the wholesale price increases by, say, 10 cents per Mcf, the retail customer will see an
additional 10 cent charge on his bill. Thus, we believe it reasonable to assume that the full
impacts of the wellhead price reductions stemming from increased production of shale gas would
be fully reflected on customers’ bills.
Based on the analysis described above, Table 3 provides an estimate of the resulting natural gas
energy bill reductions for commercial, industrial, and residential customers, using average use
per customer data from 2010.
25
EIA, Office of Integrated Analysis and Forecasting, “Analysis of Restricted Supply Cases,” February
2004. http://www.eia.gov/oiaf/servicerpt/ngsupply/pdf/sroiaf%282004%2903.pdf . There do not
appear to be any more recent EIA analyses that have estimated wellhead prices and production related
specifically to unconventional gas.
-17-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Table 3: Estimated Annual Cost Savings for Ohio End-Use Customers
Customer Class
Commercial
Industrial
Residential
Total
Average Use Per Customer (Mcf)
Price Reduction ($/Mcf)
2010 Estimated Cost Savings
Number of Customers
Estimated Savings
(Millions of $)
562.1
($2.43)
$1,366
258,422
$353.0
35,266.8
($2.43)
$85,698
5,738
$491.7
88.2
($2.43)
$214
3,198,883
$685.4
$1,530.2
As this table shows, we estimate that Ohio businesses and consumers saved over $1.5 billion on
their natural gas bills in 2010 because of lower wellhead natural gas prices. The average
residential customer, for example, burned 88 Mcf of natural gas and saved $214 in 2010. The
average commercial customer used 562 Mcf and saved $1,366, while the average industrial
customer used over 35,000 Mcf and saved almost $87,000. In addition, electric generators
reduced their costs because of lower wellhead gas prices. This translated into lower fuel charges
levied by electric utilities with fuel cost recovery mechanisms,26 such as Columbus Southern
Power and Ohio Power Company, and also contributed to lower wholesale electric prices paid by
retail electric suppliers.27
The results of our analysis demonstrate that shale gas production has significantly reduced U.S.
wellhead natural gas prices and reduced Ohio consumers’ natural gas and electric bills. The
estimated savings of $1.5 billion in 2010 affect all sectors of the Ohio economy. As Ohio’s Utica
Shale gas resource is developed, Ohio businesses and consumers are likely to benefit even more
in the future. The decreases in natural gas and electricity prices will benefit the Ohio economy,
not only by creating jobs directly in the shale gas extraction industry as the Utica Shale is
developed, but by improving the overall competitiveness of Ohio businesses and industry.
26
The default generation supply prices of Ohio Power and Columbus Southern Power (sometimes
referred to as AEP-Ohio) continue to be administratively set by the Public Utilities Commission of
Ohio (“PUCO”) based on a rate structure that includes a fuel adjustment clause or FAC. Other Ohio
electric distribution utilities (“EDUs”) establish default generation supply prices through a
competitive bidding process (“CBP”) conducted under the PUCO’s supervision. The downward
pressure that shale gas development has placed on electric prices is observable from the inputs that go
into the FAC as well as the pricing results of the CBPs that have been approved by the PUCO.
27
A subsequent report will estimate how much Ohio consumers saved on their electric bills because of
the lower wellhead natural gas prices.
-18-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Appendix 1: Econometric Model Specification
This appendix provides supporting detail on the econometric model used to estimate the impact
of shale gas production on U.S. wellhead natural gas prices. The appendix first discusses general
issues in estimating econometric models of supply and demand, and how those issues affected
the specification of the model we developed. We then present the details of the model itself, the
data used to estimate it, and the results of the estimation.
The Identification Problem
One well-recognized problem in modeling supply and demand is called the “identification”
problem.28 What this means is that, when modeling supply and demand, observed data can
reflect changes in both, as shown in Figure A-1.
Figure A-1: Identifying Supply and Demand Curves
Regression line Price D1
observed D2
D3
D4
S4 S3 S2 S1 Quantity Figure A-1 illustrates four annual supply-demand equilibrium points, each corresponding to a
different supply-demand curve combination. In this example, the four observed points do not
trace out a single demand or supply curve. Thus, if we performed a simple linear regression of
price on quantity, the resulting regression line (shown as the bright red line in Figure A-1),
would not correspond to either a supply or demand curve.
28
For a discussion, see P. Kennedy, A Guide to Econometrics, 6th Ed., (Malden, MA: Blackwell
Publishing 2009), pp. 173-76.
-19-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
If we graph the annual average natural gas wellhead prices and total natural gas withdrawals, we
see a similar problem. This is shown in Figure A-2.
Figure A-2: Average Annual Wellhead Natural Gas Prices and Gross Withdrawals
(1990–2010)
In Figure A-2, we have graphed the supply-demand combinations for the years 1990 through
2010. Although the trendline shown looks like an upward sloping supply curve, it is far more
likely that it reflects changes in demand and supply curves over time, as in Figure A-1.
Model Specification
To address this simultaneity issue, we developed a 4-equation model reflecting the supply and
demand for natural gas, as well as the supply and demand of coal for electric generating
purposes, because natural gas is increasingly being substituted for coal to generate electricity.29
Thus, we write the general model structure as:
29
We assume the world crude oil price (measured as the published price of Brent crude) as independent
of the U.S. natural gas market and the market for coal used to generate electricity.
-20-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Pt G  f1 (QtG , Pt O , Pt C )
QtG  f 2 ( Pt G , Pt O , Pt C )
Pt C  f3 (QtC , Pt G , Pt O )
(A-1)
QtC  f 4 ( Pt C , Pt O , Pt G )
where:
Pt C
=
Average annual price of coal delivered to electric generating plants, year t.
Pt G
=
Average annual wellhead price of natural gas, year t.
Pt O
=
Average annual price of Brent crude, year t.
QtG
=
Gross withdrawals of U.S. natural gas from all sources, year t.
QtC
=
Receipts of coal at electric generating plants, year t.
Thus, the model consists of demand and supply equations for natural gas and coal delivered to
electric generating plants. The specification treats the wellhead price of natural gas, gross
withdrawals of natural gas, the quantity of coal delivered to electric generators, and the price of
coal delivered to electric generators as endogenous variables. The price of Brent crude is treated
as exogenous.
To address the data shown in Figure A-2, we evaluated a number of functional forms for the
general model structure in (A-1), ultimately settling on the following specification, (t-1
subscripts indicate one-year lagged values).
QtC   02  12 Pt C   22 Pt O1  32 Pt C1   43GDPt   t2
(A-2)
Pt C   03  13 Pt C1   23 Pt G   33 Pt O1   t3
(A-3)
Pt G   04  14QtG   24 Pt O   34 Pt C1   44 D2002  54 D2005   t4 (A-4)
QtG   05  15 Pt G   25 Pt G1   35 Pt C1   45 Pt O1   t5
(A-5)
where:
D2002
= Dummy variable for the year 2002.
D2005
= Dummy variable for the year 2005.
GDPt
= Real U.S. gross domestic product, year t.
 tj
= Random error term, equation j.
The inclusion of lagged price variables reflects the fact that gas and coal production and
consumption decisions are influenced by observed historic prices, as well as contemporaneous
-21-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
prices. The inclusion of a “dummy” variable for the year 2002 reflects the economic downturn
brought on by the events of September 11, 2001. The inclusion of the dummy variable for the
year 2005 reflects the damages caused by Hurricanes Katrina and Rita to the natural gas drilling
and gathering infrastructure off the U.S. Gulf coast.
To estimate the impact of shale gas production on wellhead natural gas prices, we note that
QtG  QtShale  QtNon  shale
(A-6)
where QtShale and QtNon  shale represent production shale gas and gas from all non-shale sources,
respectively. Thus, the estimated wellhead price of natural gas without shale gas, Pt G, NS is
Pt G, NS  Pt G  14QtShale ,
(A-7)
where the minus sign in equation (A-7) reflects the fact that total wellhead production is reduced
by the quantity of shale gas produced.
Data Sources
All of the data used to estimate the model is publicly available and published by the U.S. EIA.
The specific data sources are as follows:
Pt C
Electric Power Annual. Available at: http://www.eia.gov/electricity/data.cfm
Pt G
Available at: http://www.eia.gov/dnav/ng/ng_pri_sum_dcu_nus_m.htm
Pt O
Based on daily spot prices, as published by EIA from ThomsonReuters.
QtG
Natural Gas Annual. Available at:
http://www.eia.gov/dnav/ng/ng_prod_sum_dcu_nus_a.htm
Electric Power Annual
QtC
Analysis Results
Equations (A-2) – (A-5) were modeled using the three-state least-squares (3SLS) method in
STATA.30 The results are summarized in Table A-1.
30
See Kennedy (2009), pp. 180-81.
-22-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Table A-1: Three-stage Least-squares Regression Summary31
No. of Obs Parameters RMSE "R‐sq" Chi‐
square Significance (A‐2) Qcoal 20 3 50662.89 0.7064 50.17 0.0000 (A‐3) Pcoal 20 4 0.6913167 0.9783 909.27 0.0000 (A‐4) Pgas 20 5 0.3771036 0.8931 167.70 0.0000 (A‐5) Qgas 20 5 0.3141652 0.9670 584.57 0.0000 Equation The estimated coefficients for all of the equations are shown in Table A-2.
Table A-2: Three-Stage Least Squares Regression Results
Equation Dep. Variable
(A‐2)
(A‐3)
(A‐4)
(A‐5)
Coef.
Std. Err.
z
P>|z| [95% Confidence Interval]
Pcoal(t)
Qcoal(t)
‐5.01E‐06
3.95E‐06
‐1.27
0.205
‐0.0000128
2.73E‐06
Pcoal(t‐1)
0.9101757
0.0439476
20.71
0
0.82404
0.9963114
Pgas(t)
0.6102044
0.1782589
3.42
0.001
0.2608233
0.9595854
Poil(t‐1)
0.0627027
0.0122623
5.11
0
0.0386691
0.0867364
Constant
2.756187
3.41147
0.81
0.419
‐3.930172
9.442546
Pcoal(t)
‐2065.633
4251.958
‐0.49
0.627
‐10399.32
6268.052
Poil(t)
2004.676
1515.881
1.32
0.186
‐966.3957
4975.748
Real_GDP
24.01612
16.67534
1.44
0.15
‐8.666949
56.69919
Constant
625558.4
253000.7
2.47
0.013
129686.1
1121431
‐0.4617642
0.0809128
‐0.2108965
0.7724999
‐1.175227
18.91655
0.0873941
0.0047686
0.0164349
0.3563397
0.3269946
2.210884
‐5.28
16.97
‐12.83
2.17
‐3.59
8.56
0
0
0
0.03
0
0
‐0.6330535
0.0715666
‐0.2431083
0.0740869
‐1.816125
14.5833
‐0.290475
0.0902591
‐0.1786848
1.470913
‐0.5343296
23.2498
Pgas(t)
‐0.2000595
0.0905154
‐2.21
0.027
‐0.3774665
‐0.0226526
Pgas(t‐1)
‐0.6352557
0.1392652
‐4.56
0
‐0.9082104
‐0.362301
Pcoal(t‐1)
‐0.0856178
0.0413864
‐2.07
0.039
‐0.1667337
‐0.004502
Poil(t‐1)
0.0523611
0.0178196
2.94
0.003
0.0174353
0.0872869
Year
0.2041171
0.0379631
5.38
0
0.1297108
0.2785235
Constant
‐380.1301
76.49281
‐4.97
0
‐530.0532
‐230.2069
Qcoal(t)
Pgas(t)
Qgas(t)
Poil(t)
Pcoal(t‐1)
Dummy_2005
Dummy_2002
Constant
Qgas(t)
31
Note that the “R-sq” is not the same as the traditional “goodness-of-fit” measure in OLS regressions,
because with 3SLS “R-sq” can be negative. For comparison purposes, Appendix 1 provides the
results of OLS regressions of each of the four equations (A-2) – (A-5).
-23-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
The regression results for the coal equations are consistent with economic theory. For example,
equation (A-2) shows that the delivered price of coal to electric generators increases as the price
of natural gas and oil increase, which would be expected for substitute fuels. Equation (A-3)
shows that the quantity of coal deliveries increases as the price of oil increases, but decreases as
the price of coal increases. It also shows that coal deliveries increase as economic growth,
measured by real GDP, increases.
Equation (A-4) shows that the wellhead price of gas tends to increase as the price of oil
increases, as was generally the case until 2005. Equation (A-4) also includes the two dummy
variables for the years 2002 and 2005, respectively. The coefficients for both dummy variables
have the expected signs.
Equation (A-4) also shows that the wellhead price of gas is strongly related to the previous year’s
price of utility coal receipts. However, the coefficient is negative. That is, an increase in last
year’s utility price of coal tends to decrease this year’s wellhead price of natural gas. We
hypothesize that this initially counterintuitive result stems from the response by natural gas
producers to anticipated increases in the demand for natural gas. In other words, producers
respond to higher coal prices and, hence, expected increases in the demand for natural gas, by
increasing production. These production increases more than offset the expected increase in
demand. In fact, this phenomenon has clearly contributed to the reductions in natural gas prices
generally; technological improvements in drilling technology have enabled rapidly increasing
quantities of shale gas to be produced, more than compensating for the general increase in
natural gas demand.
Equation (A-5) yields the expected results for the coefficients on the prices of natural gas and
crude oil. Thus, we expect increases in the price of crude oil to increase production of natural
gas. Not only does this stem from increased demand for gas, but higher oil prices encourage
additional exploration of development of domestic oil resources, and significant quantities of
natural gas are produced from oil wells (so called “wet gas”). Equation (A-5) also shows the
same counterintuitive result with respect to coal prices.
Impacts of Shale Gas Production on Wellhead Prices
Equation (A-7) and the coefficient 14 from equation (A-4) are used to estimate the impacts of
shale gas production on wellhead natural gas prices. Thus, the impact of shale gas production in
year t on the average wellhead natural gas price in year t is
Pt G, NS  Pt G  (0.4617642) x QtShale
,
which implies that, for each Tcf of shale gas produced, the wellhead price decreases by just over
$0.46 per Mcf.
-24-
The Economic Impacts of Shale Gas on Ohio Consumers
January 2012
Limitations of the Model Specification
As discussed in Section II of the report, econometric modeling has both advantages and
disadvantages, with a specific disadvantage being the partial equilibrium framework that we
have used. Such a framework necessarily trades off accuracy for greater simplicity. The model
estimated presumes there is a linear relationship between shale gas production and wellhead
prices.
In fact, the relationship may be nonlinear, especially as shale gas production accounts for an
increasing proportion of total U.S. natural gas production. In 2010, for example, shale gas
accounted for almost 20% of total gross production in the U.S., and that percentage is expected
to increase over time, barring environmental regulations that restrict or curtail shale gas
production in the future. To the extent the relationship is nonlinear, the predicted impacts of
shale gas production on wellhead natural gas prices may be overstated. Moreover, additional
shale gas production, to the extent it reduces wellhead prices, may reduce production of
conventional natural gas. The impact will be a function of the market price and the production
cost of conventional gas: the higher the cost to produce conventional gas, the larger the likely
reduction in conventional production, and the smaller will be the net price impact. Finally, higher
natural gas prices could have other macroeconomic impacts that are not considered in the
econometric model, such as reductions in economic growth that would reduce overall natural gas
demand and temper such price increases. Thus, a recommended next step in researching the
impacts of shale gas on wellhead prices is to use more complex econometric specifications that
test for and account for these potential nonlinearities.
-25-