transformation bp 180112 2

8/3/2019 Transformation BP 180112 2

1/126

The great transformation:decarbonising Europesenergy and transport

systemsBY GEORG ZACHMANN, MICHAEL HOLTERMANN,JRG RADEKE, MIMI TAM, MARK HUBERTY, DMYTRO NAAND ANTA NDOYE FAYE

BRUEGEL BLUEPRINT 1


2/126

The great transformation:decarbonising Europesenergy and transportsystems

BY GEORG ZACHMANN, MICHAEL HOLTERMANN, JRG RADEMIMI TAM, MARK HUBERTY, DMYTRO NAUMENKO AND

ANTA NDOYE FAYE

BRUEGEL BLUEPRINT SERIES


3/126

BRUEGEL BLUEPRINT SERIESVolumeXVIThe great transformation: decarbonising Europesenergy and transport systems

Georg Zachmann, Michael Holtermann, Jrg Radeke,Mimi Tam,Mark Huberty,Dmytro Naumenkoand Anta Ndoye Faye

Bruegel 2012. All rights reserved. Short sections of text,not to exceed twoparagraphs,may bequotedin theoriginallanguagewithoutexplicitpermissionprovided thatthesource isacknowledged.TheBruegelBlueprint Seriesis published under the editorial responsibility of Jean Pisani-Ferry, Director of Bruegel.Opinions expressed in thispublication are those of the author(s) alone.

Editor: StephenGardnerProduction:MichaelT. HarringtonCover graphic: Jean-Yves Verdu

BRUEGEL33, rue de laCharit, Box 41210 Brussels, Belgiumwww.bruegel.org

ISBN: 978-9-078910-25-1


4/126

This report has been produced by Bruegel and the European School for Managementand Technology (ESMT). The research leading to these results has received fundingfrom the Fuel Cell and Hydrogen Joint Undertaking (FCH). The views expressed inthis publication are the sole responsibility of the authors and do not necessarilyreflect the views of FCH.

The authors are grateful to Nicolas Brahy, Klaus Bonhoff, Bert De Colvenaer, RayEaton, Pierre Etienne Franc, Patrick Francoisse, Peter Frschle, Ian Hodgson, KarelKapoun, Guillaume Leduc, Gunnar Muent, Armin Riess, Jurriaan Ruys, SandroSantamato, Koen Schoots, Franz Sldner, Reinhilde Veugelers, Gijs Vriesman, JensWeinmann, and Guntram Wolff for having shared their time and thoughts.


5/126


6/126

Contents

About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .viiForeword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i x

Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

1 Rationale for supporting the transition to a new energy andtransport system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61.1 Benets of a new energy and transport system . . . . . . . . . . . . . . . . . . . . . . . . .61.2 Market failures that impede an optimal transition . . . . . . . . . . . . . . . . . . . . . . .81 .2 .1 Cl imateex te rna l i ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 .2 .2 Innovat ionexternal i ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111.2.3 Pathdependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .131.2.4 Coordination externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171.2.5 Infrastructure externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221.2.6 Businessexplorationexternality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251.2.7 Insurance externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261.2.8 Industr ialpolicyexternali ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .291.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36

2 Analysis of commercial and policy gaps: the case of fuel cell

e l e c t r i c v e h i c l e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 82.1 Identifying the relevant factors for success of fuel cell electricv e h i c l e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 9

2.1.1 A model-based approach to detecting the technological andcommercial gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41

2.1.2 Technical pre-conditions forcommercial deployment of hyd rogenveh i c l e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

2.1.3 Costdevelopmentofkeycomponents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432.1.4 Production infrastructure hydrogen production cost . . . . . . . . . . . . . . . . .50

2.1.5 Retaildistributioninfrastructurenetworkdensity ....................512.2 Consumer acceptanceand the technological and commercialgap . . . . .57

v


7/126

2.2.1 Consumer acceptance of FCEVs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .572.2.2 FCEVmarketpenetrationunderselectedscenarios ....................592.3 Current Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .622.3.1 Climateexternality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .622.3 .2 Innovat ionexternal i ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .642.3.3 Infrastructure externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .692.3.4 Othermarketfailuresaddressedbycurrentpolicies ...................74

3 Policy response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .783.1 Resolving the climate externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .783.1.1 Inclusion of road transport in the ETS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .793.1.2 Financial instruments to lock in a long term carbon price . . . . . . . . . . . . . . .803.1.3 Schemestodrivesupplysideinvestments .... . . . . . . . . . . . . . . . . . . . . . . . .813.2 Resolving the infrastructure externality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .823.2.1 Option 1:public funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .823.2.2 Option 2:establishment ofan infrastructure consortium . . . . . . . . . . . . . . .833.3 Financial support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .853.4 Shifting risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .893.5 Public procurement mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .953.6 A consistent policy response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .963.6.1 Limitsoftechnologyneutral i ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .963.6.2 Challengeoftechnologychoice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .973.6.3 Status quo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .983.6.4 A consistent and predictable support mechanism . . . . . . . . . . . . . . . . . . . . .983.6.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107A.1 Modelling consumer acceptance and policy impacts . . . . . . . . . . . . . . . . .107A.2 Appendix: Selected existing support instruments . . . . . . . . . . . . . . . . . . . . .109

THE GREAT TRANSFORMATION

vi


8/126

About the authors

Michael Holtermannis Project Leader of the European School of Management andTechnologysMarket Model Electric Mobilityproject. Hisexpertise includes regulatedindustries, energy systems and networks and public-private interdependencies.Previously,hewasa Programme Director atESMT CustomizedSolutionsGmbH (2004-2009),anda consultant atAccenture(19982004). Healso workedasSeniorProjectManager at the Treuhandanstalt, advising companies from eastern Germany (19961997), and was a project manager and editor at the Klett Group (19941995). Hereceived his MA from theFreie Universitt Berlin.

Mark Hubertyis a research associateat theBerkeley Roundtableon theInternationalEconomy, and a doctoral candidate in political science at UC Berkeley. His researchconcerns the political determinants of comparative advantage, climate change andindustrial policy,andrmbehaviourin industrialiseddemocracies.Outside academia,Mark has consulted for Mandag Morgen, Accenture and A.T. Kearney. His work hasreceived support from theFulbright Foundation and theUnited States EnvironmentalProtection Agency

Dmytro Naumenkoisa SeniorResearch Fellow at the Institutefor EconomicResearchandPolicyConsulting(IER),a Kyiv-based think-tank, advisingpolicymakers onenergypolicy issues and nancial markets development. He was previously a Research

Assistant atBruegel, and workedasa nancialanalyst for KINTO,anasset managementcompany in Kyiv. He received his MA degree in nance from Kyiv National EconomicUniversity.

Anta Ndoye Fayeis anAdvisor to theExecutive Director for France at theInternationalMonetaryFund inWashingtonDC,andwaspreviously a ResearchAssistant atBruegel.Shehas a PhDinMicroeconometricsfromBETA (Strasbourg)and theKonstanz DoctoralSchool in Quantitative Economics and Finance. She also holds a Masters degree inEconometrics and Financial Markets and aMagistre dEconomiste Statisticien from

the University of Toulouse.

vii


9/126

Jrg Radeke is a member of the European School of Management and TechnologysMarket Model ElectricMobilityproject team.Previously hewasa ConsultantEconomistat the Centre for Economics and Business Research, London, commentating onEuropean macro-economic trends. His expertise is in economic impact assessment,economic modelling and forecasting.

Mimi Tamhas been a Research Assistant at Bruegel since September 2011. Shecompleted her MSc in Economics at the Barcelona Graduate School of Economics,specialising in theoretical microeconomics. Previously she worked for large con-sultancy rms including KPMG and Deloitte & Touche. Mimi is also a graduate inComputer Science from New YorkUniversity.

GeorgZachmannhas been a Resident Fellow atBruegel sinceSeptember 2009.He isalso a member of the German Advisory Group in Ukraine and the German EconomicTeam in Belarus, advising policymakers in these countries on energy-sector issues.HepreviouslyworkedatLARSENinParis,attheGermanMinistryofFinanceandattheGerman Institute for Economic Research (DIW Berlin). He holds a doctoral degree ineconomics from Dresden University of Technology.


viii


10/126

Foreword

The euro-area crisis dominates the economic news. Yet, the world and Europe mayface even more important challenges that will shape our lives and the lives of ourchildren. World population isprojected to increase to 9 billion or more by 2050. At thesametime, currenttrends indicateanincreasein livingstandards and a growing middleclassaroundtheworld.Thesetwomega-trendswillhaveprofoundimplications,andtheway theyare managed willbeone of the key determinantsofprosperityand peace inthe decades or even centuries to come. A number of factors are important in thisrespect.

Morepeopleand moreincomewill increasetheglobaldemandfor energy.Choosing therightsourcesofthisenergywillbeoneofthedeterminingfactorsofglobaltemperature.The continued relianceonfossil-fuel energysources isone of the mainfactors behindtherisk of significantglobal temperature increases. The internationallyagreedgoal of limiting the temperature rise to less than two degrees Celsius above pre-industriallevels appears increasingly illusory.Currently, fossil energysources dominate manyeconomic areas. For instance, our transport infrastructure is largely based on fossilfuels, and is thereby one of themain contributor of thecarbon dioxide emissions thatarelinkedtoglobaltemperature. Thinkingabouta decarbonisationstrategy is thereforea key challenge with a global dimension.

Economic growth in Europe will be affected by the costs of this transition from thecurrentenergy and transportsystem.A smoothtransition towardsa low-carbon energyand transport system could come at comparatively modest cost. Furthermore,identifying themost economicallybeneficial solutionsearlyonandbecominga globaltechnology leader and standard setter offers vast opportunities for exports andeconomicgrowth.Hence,ourdecarbonisationstrategymayeventuallyhavea greaterimpact on long-term European growth than thecurrent economic crisis.

Bruegel iscontributing to thisdebatewith this report, which isbased on research that

receivedfundingfromtheFuelCellandHydrogenJointUndertaking.Theauthorsarguecarefully that to make decarbonisation growth friendly, a consistent policy approach

ix


11/126

is needed. Policy intervention appears indispensable as the energy and transportsystem is so based around and locked-in into an incumbent technology. Overcomingthis lock-in is crucial. The report makes three main proposals. First, the scope,geographical coverage and duration of carbon pricing should be extended. By settinga higher carbon price, incentives for developing and investing in new low-carbontechnologiesarecreated.Second, temporaryconsortia fornewinfrastructureto solveearly-phasemarketfailurescouldbeput inplace. This isdiscussed using theexampleof hydrogen vehicles. Lastly and importantly, an open and public transition model isneededso that second-best transport solutions donotgeta head start that afterwardscannotbe reversed.

The technological, economic and political challenge ahead is vast. But choosing theright decarbonisaton strategy offers huge economic, environmental and societalbenets. We should not overlook this debate because of the euro crisis.

Guntram Wolff, Deputy Director, BruegelBrussels, January 2012


x


12/126

Executive summary

Transition is necessary

The major challenge facing the energy and transport system is the reduction of itsfossil fuel consumption and carbon footprint. This requires a shift in the way weproduce and consume energy. Due to the limited carbon-reduction potential of incumbent technologies, new low-carbon technologies will have to enter themainstream market. Some of those new technologies offer signicant side-benetssuch as reducing localpollutantandnoiseemissions. Furthermore,decarbonising theeconomy based on new technologies could induce growth.

Transition is a complex endeavour

The current energysystem, in its complexity,has developed overcenturies.The rapiddiffusion of new technologies requires either that they have serious advantages overincumbent technologies, or that downstream changes are minimal. Presently, mostlow-carbon energyand transport technologiesmeet neither of these criteria: theyaremore expensive than the technologies they replace but offer little, if any, advantage,and they require substantial downstream changes to the incumbent energy ortransport system in order to accommodate different primary inputs and differentoperatingcharacteristics.Consequently, transitionrequires that stakeholders roll-out

all parts of thenew system synchronously.

Market failures impede transition

Markets alone will not encourage the development and deployment of un-competitive technologies, even if they are necessary for a low-carbon future. Inorder to encourage the development of these technologies, it is important tomonetise the societal benefits they provide by puttingpriceson greenhouse gases,pollution, noise, and import dependency. In the absence of first-best solutions

(for example a global long-term carbon price) policymakers should build onexisting instruments in order to provide sufficient incentives for early investment

1


13/126

into research and development, demonstration, and deployment of low-carbontechnologies.

Innovation is essential to develop the required new technologies. Without effectivepolicies to protect intellectual property or alleviate the private costs of innovation,therewill beunderinvestmentinR&D. Buteven ifcompetitivelow-carbon technologiesareavailable,pathdependenciesdue to institutions,risk-aversionandnetworkeffectspreventaquickrollout.Thisposesahugechallengetopolicymakersiftheyaretohelpnew technologies supplant the incumbent system without favouring one of thealternatives.

A key to success is domestic and international, and public and private cooperation.Leaving coordination entirely to the market might result in late deployment andfragmented networks and markets. Some technologies require a completely newunderlying infrastructure. This infrastructure has a high cost that may not be fullyrecoverable by the initial providers, when the business is regulatedex post or lateentrants face lower costs. To recoup their initial investment, providers might have anincentive tocapturecustomersby implementing articial barriers topreventswitching.This can lead to fragmented markets and slow adaptation of new technologies.

A similar problem is also faced by companies in other parts of the value chain. Thecostsofexploring,andbuilding,newmarketsishighandmaynotbefullyrecoverablegiven that later entrants may reduce prot margins. Thus, early movers might not bewilling to take risks.This isunfortunate because exploringnewlow-carbon technologybusiness models has a high social value. It provides important information to con-sumers, competitors and politics about the viability of technologies.

Some low-carbon technologies might never be commercialised because better

alternativesexist. However, continuing to fundthese technologies might be essentialin case the rst-best alternative fails to deliver. In this case, having a back-stoptechnology on the shelf for quick deployment might save valuable time in the ghtagainst climate change.

Finally, low-carbon technologies do not only offer environmental benets. Deployingand exporting them might offer business opportunities. Under certain conditions it iseven conceivable that economiesasawhole mightbenetfrom low-carbontechnologyindustriesthatwerebuiltonearlylocaldeployment.Thisearlydeploymentofstillnon-

competitive low-carbon technologies will, however, often not happen without publicsupport.


2


14/126

Thus, private investment intonewtechnologies provides manypositive spillovers forsociety. Asmarkets donot compensatefor thesespillovers companieswillbereluctantto make the necessary investments. Consequently, without public intervention, thetransition will only happenslowly ormay not takeplace at all.

Fuel cell electric vehicles will not be provided by the market alone

We use the example of fuel cell electric vehicles to demonstrate that certain low-carbon technologies only enter the market if at least some of the market failurespreviously described are resolved. This example was chosen because fuel cellelectric vehicles promise to be a carbon-free transport alternative with significantrange and no local pollutant and noise emissions,but their deployment isheldbackby thevery high initial cost andtheabsenceof therequireddedicated infrastructure.Under the existing framework conditions, fuel cell electric vehicles will be virtuallyabsent from the vehicle market in 2050 while incumbent technologies (gasoline,diesel) will still play a major role. We show that this changes when policies areimplemented to account for the emission cost of conventional vehicles, support toR&D is provided and the infrastructure externality is overcome. With such a con-certed approach, fuel cell electric vehicles might become an important transporttechnology by 2050, accounting for more than 10 percent of the market. Early costreductions (such as through R&D) are essential to overcome the gap that preventsdeployment. In the most optimistic scenario based on industry forecasts, fuel cellelectric vehicles might capture more than 25 percentof the market according to ourmodelling.

Existing tools are insufficient

There isanextensivemenuofcurrentpolicies at regional,national andEuropeanlevels

that are intended to address the market failures. Fuel taxes, vehicle emissionstandardsandR&Dfunding,forexample,canbeeffectivetoolsfortacklingsomeofthebarriers. However, the totality of current policies is insufficient to resolve the marketfailuresthathamperthetransition.ThereareinsufficientfundsforR&D,nogloballong-term carbon price, and deployment efforts are not coordinated. Most importantly, nosolution for the infrastructure externality is being implemented, and support fortechnologies is not predictable.

Smart policy tools for transition

To enable theprivate sector tomake the necessary investments for development and


3


15/126

deployment of the technologies needed for the energy and transport systemtransition, a set of smart policies needs to be implemented.

First of all, the cost of carbon in different sectors needs to be aligned in order tostimulate efficient emission-mitigation behaviour. Thus, all forms of transport shouldbe included in the European Union emissions trading system (ETS). A correspondingadditional carbon component in the fuel price would ensure that consumers dailymodal choicedecisions take the carbon cost intoaccountand thus prevent lower fuelconsumption incentivising increased vehicle use. Second, policymakers need toconvince companies that carbon will continue to be sensibly priced beyond 2020.Thus, policy should nancially commit their future budgets vis--vis companies thatinvest in low-carbon technologies to preserve the operability of the EU ETS beyond2020.Thiscould,forexample,beachievedbylettingpublicbanksissueoptionsonthecarbon price. Signicant exposure of public banks to the carbon price could serve asa tool to commit future policymakers to ensuring the reliability of the system overdecades. Third, tightening average emission standards for certain appliances is aneffective second-bestsolution for incentivising theprovision of low-carbonappliancesin the absence ofa global and long-term carbon price.

To provide the refuelling stations for new fuels that existing markets will not deliver,we suggest the establishment of temporary infrastructure consortia for the differentlow-carbonfuels.Eachconsortiumwouldplanandorganisethedeploymentofitsres-pective fuelling station infrastructure. For this purpose, each consortium would begiven the exclusive right to sell local concessions for new fuel stations to interestedretailers. Consequently, competitionbetween different low-carbon fuels anddifferentretailers would be ensured. Finally, each consortium might organise internal cross-subsidisationbetween differentparts of thevaluechain (forexample, fuel andvehicleproducers might support infrastructure) and between different fuel stations (for

example,fueloutletsinremoteareasmightobtainsupportfromfueloutletsindenselypopulatedareas), if itnds that this encourages quicker roll-outof their technology.Toavoid abuse, all relevant stakeholders should participate in the consortia and theirconstitution shouldbe clearedex ante by competition authorities.

Furthermore, the public and private sectors should explore new ways of sharing risk.Governments might participate in the up-side of successful technologies by makinggrants reimbursable in successful cases. Meanwhile public nancing or guaranteesdedicated to business units with a high concentration of regulatory risks might

incentivise investment for two reasons. First, the corresponding company would beless exposed to regulatory uncertainty and might nd it easier to acquire private


4


16/126

nance for its low-carbon projects.Second,public exposure to regulatory risk signalscommitment to existing policies and reduces regulatory uncertainty in the privatesector asa whole.

One major improvement to current deployment policies would be to use publicprocurement strategically forexperimenting withalternative technologies.Wesuggestthat publicly nanced trials (for example, for municipal vehicle eets) should beallowed tofailcommercially inorder toavoid thefocus onlow-risktechnologies.For thispurpose, compensation for failed trials should be offered at a federal level, providedthat the individual trial is part of a coordinated experimental scheme.

Finally, themost important step for supporting new technologies isa transparent andpredictable support policy forall competing technologies. A consistentpolicy shouldprimarily comprise a set of horizontal policies to resolve existing market failures (egcarbon pricing). But in the absenceof horizontal rst-best solutions for some marketfailures, the public sector should return to technology-specic support instrumentsfor R&Dand deployment. In thiscontext, technologychoiceiscritical. In thepresenceofmultiplenewtechnologiesthatcompetenotonlyforamarketbutalsoforproductioninputs (such as capital, labour and raw materials), excessive support to onetechnologymight slow downthedevelopment ofothers.Consequently,a well-thought-throughand structuredapproach adapted tothecomplexityofthechallengeisneeded.For this purpose, government should adopt a choice mechanism that is dynamic andadaptable, able to digest new information and optimise support in a quick, reliable,and effective manner. Predictability and technology-neutrality can only be ensuredwhen technology choice is based on metrics and priorities dened by politics.Stakeholdersneed tobe incentivisedtoprovide unbiased forecasts of thecapabilitiesof their technology.These forecasts shouldbeprocessedinanopen multi-technologymodel toprovide guidancefor the targeting ofsupport.A corresponding model should

be built, maintained, extended and published by an independent public institution.This transparent mechanism would ensure that stakeholders can predict publictechnology decisions, and would thus nd it easier to commit to the long-term andrisky investments that are needed to make the low-carbon energy and transportsystem transition a reality.


5


17/126


18/126

system is the reduction of its fossil fuel consumption and carbon footprint.

With the current fuel mix5, even the most ambitious improvements to incumbenttechnologies are likely to be insufficient for reaching the reduction targets set by theEU. For example, improvements to motor vehicle internal combustion engines andconventional power plants are limited by physical factors. Fuel consumption wouldconverge to a technical minimum that is signicantly above zero. Consequently, thedeployment of new clean energy and transport technologies would be necessary tomaintain thecurrent service level at near zero emissions.

Anadditionalmotivationforcarbon-free technologies is thattheyoften offer signicantside-benets. For example, internal combustion engines are responsible for asignicant portion of local pollutant6 and noise emissions. For this reason, internalcombustion engines are, incontrast tosome oftheproposedalternative technologies,also detrimental to public health. Thus, signicant societal benets, in terms of greenhouse gas mitigation,decreasedfuel dependency, andreduced localemissionsof pollutants and noise, can be expected as results of a transition towards a cleanenergy and transport sector.

Furthermore, various authors have argued that decarbonising the economy couldinducegrowth(eg Hubertyetal ,2011).Policyargumentsforgreengrowthspanawiderangeofeconomic, environmental andsocial concerns. A samplingofsuch argumentsdemonstrates theirdiversity:

1. Keynesian demand stimulus for short-term job creation via decit-nancedinvestment in energyefficiency andenergy infrastructure: for example, Houseretal (2009, 2-5) nds that green stimulus in the US performed as well as or betterthan traditional stimulus, creating 20 percent more jobs than traditional infra-

structure spending.

2. Improvedtrade competitiveness viareducedexposure to terms-of-tradepressuresfrom fossilfuel imports,particularly petroleumandnaturalgas. Decreasing demandforfossilfuelimportsreducestheworldmarketpriceoffossilfuels.Thus,theterms-of-trade of energy importers improve, ie EU countries will have to export less inorder to pay for foreign fuels. Thus, domestic consumption and consumer welfarecan increase.


7

5. Many observers dismiss biofuels based on expected cost (Runge and Senauer, 2007; Ryanet al , 2006; Delgado

andSantos,2008)6. EgOxides of nitrogen (NOx), volatile organic compounds(VOC),ozone,particles andSulphur Oxides (SOx).


19/126

3. Increasedinnovation in response togreater administrative constraints (alsoknownas the Porter Hypothesis): Porter and van der Linde (1995) argue that stringentregulation pays for itself by inducing private sector innovation. Additionally, WEF-BCG (2011) argues that companies that comply with stricter standards do bettereconomically7.

4. Publicly supported deployment creates markets for new technologies that mighthaveahigher thanproportionate localvaluecontent .Thus,newjobsmightemerge(Weiet al , 2010).

5. Revenues from a polluter pays scheme such as emission allowance auctionreceipts or green taxes might be used to reduce distorting taxes on labour andcapital.Undercertainconditions adouble dividendintermsofhighergrowthmayarise (see Goulder, 1995).

6. Redirecting innovation and investmentsat anearly stage to thegrowing sector of clean technologies might help some countries retain or even strengthen theirinternational competitiveness, therebyboostingtheir economiesandcreating jobs.For example, Huberty and Zachmann (2011) argue that state-supported deploy-mentcanpartlyexplainthesuccessofthewindindustryinDenmarkandGermany.

Thus, thetransition toa new energyand transportsystem promisessignicant societalbenets.As thenext section demonstrates, a number ofmarket failures impedesucha transition.Thus,without public intervention the transition will only happenslowlyormay completely fail to take place.

1.2 Market failures that impede an optimal transition

The current energysystem, in its complexity,has developed overcenturies. Though itsuffers from an extreme degree of inertia, the energy system has undergone a seriesof transformations over time: from wood to coal, coal to oil, and to electrication. Ineachof these cases, the new energysourceprovedcheaper ormoreversatile than theone it supplanted or complemented. However, inertia in the energy system, due topathdependenciesandmarketfailures,suchasnetworkexternalities,ledtoveryslowtransitions. Despite the notable advantages of each successive fuel, transitions tooktime: perhaps 200 years for coal and 75 for both oil and electricity. Inertia inprevioustransitions has been due to extant market failures.


8

7. However, someeconomistshave argued thatthese ndingsaredueto thesocalled environmentalKuznetscurve,which postulates that, after a certain threshold, pollutionintensitydecreaseswith increasing economic activity.


20/126

Therefore, rapid diffusion of new technologies requires them to have either seriousadvantages over incumbent technologies or minimal downstream changes to work.Presently, green energy and transport technologies meet neither of these criteria:they generate more expensive but, at best, indistinguishable services compared tothe technologies they replace, and they require substantial downstream changes totheincumbentenergyor transport systeminorder toaccommodatedifferentprimaryinputs and different operating characteristics. Thus, the inertia witnessed in priortransformations may provide only a conservative estimate for the scale of the greenenergy and transport transition challenge. The deployment of new energy andtransport technologies will be hampered by their higher cost and technicalshortcomings (eg range for battery cars, temperature sensitivity for fuel cell cars,volatility of the electricity produced by wind turbines). Markets alone will notencourage the development and deployment of uncompetitive technologies, even if they are necessary for a low-carbon future. In order to encourage thedevelopment of these technologies, it is important to monetise the societal benets they provide byputting prices on greenhouse-gases, pollution, noise, and import dependency. Butevenaftermonetising thesocietal benets,thereare extantmarket failures whichmayhinder the development of new energy technologies or the transition to low-carbonfuels and technologies.

In this section,wewill discuss marketimperfections responsible forunderinvestmentinnewenergyandtransporttechnologies,andhowtheyhavebeendealtwithinothercases.

1.2.1 Climate externality

Cumulative greenhouse-gas emissions cause global warming, implying potentiallyhugeeconomiccoststosociety8. Thus,each sourceofgreenhouse-gas emissionshas

a societal cost (a so called negative externality). To introduce the correct incentivesfor greenhouse-gas mitigation, various schemes have been proposed. The spectrumranges from administrative measures, such as the prohibition of certain pollutingtechnologies or emission restrictions, to the implementation of a polluter paysprinciple viacarbon taxes or tradable emission allowances.

Ideally, the introduction of a long-term carbon price reecting the true cost of emissions, via taxes or tradable allowances, would be the rst-best solution forreducing emissions at the lowest cost. It would ensure that emissions are reduced in


9

8. Stern (2006).


21/126

those sectors whereabatement ismost easilyachieved. However, therst-best isnotpossible because (1) The optimal level of emission abatement ie the level at whichthe cost of an additional abatement effort exceeds thebenet of the induced climatechange mitigation is unknown; (2) in the absence of an international agreement, alocal carbon price has only a limited effect on overall abatement. The reason is thatgreenhouse-gas-emitting companies or sectors might move to countries without acarbon price (carbon leakage) or that, due to reduced demand for fossil fuels in thecountries with a strict carbon price, fuel prices will decrease and result in higher fueldemand elsewhere (indirect leakage); (3) nally, there are numerous politicalconstraints. Transport and energy costs are important factors for regionalcompetitiveness, sopolicymakersareverycautious in implementinglegislation whichdirectly implies raising costs.

Unstableand inadequatecarbonpriceshavedeveloped in theEU. Othercountries (egAustralia, China) are also considering implementing incomplete carbon tradingschemes. In the presenceof only a local and short-term carbon price, there would beunder-investment innewenergyandtransport technologies.Companiesface smallerthan optimal current and future markets for clean technology, and, as a result, do notinvest in technologies thatmight incur a high cost per ton ofcarbonabated ina smallmarketin theshort term.However, learningandeconomies-of-scalesavings mayresultina much lower long-term carbon abatement cost.The mainbenetof limited deploy-mentofthenewtechnologiesisnotsomuchthedirectreductioningreenhousegases,pollution,noise,and fuel imports; rather it is the inducedcost savings due to learning-by-doing, learning-about-costs, and learning-through-R&D. This learning makes laterand larger deploymentscheaper and thus reduces thecost of achieving benets at alarge scale. Furthermore, cost reductions resulting from learning might make thetechnologies competitiveeven inenvironmentswith less ambitiouscarbon mitigationpolicies (eg developing and emerging countries). Therefore, in theabsence of a long-

term global carbon price, it is sensible to provide incentives for R&D and deploymentof these technologies, so as to approach thesocially optimal investment level. In theEU, for example, signicant support for renewable energy technologies and energyefficiency ispartly justiedascompensating the imperfections of thecarbon market.

The ipside of having multiple instruments to incentivise emission reductions is thatit inevitably leads to different prices for carbon in different sectors. According toFankhauseretal (2010), combiningtaxes, subsidies, or standardswithcap-and-tradeinstruments can undermine the carbon price and increase mitigation costs. That is,

the absence of a single carbon price signal to coordinate abatement decisions in allsectors is causing economic inefficiencies. Consequently, there is over-abatement in


10


22/126

some sectors (replacing classic lightbulbs in some applicationsimplies huge cost pertonofcarbon abated) andunder-abatement inmost othersectors (mostnew-builds of coalredpowerplantswouldnothappenatasufficientlyhighandstablecarbonprice).

Conclusion:Climate change is a pressing issue which poses huge negative exter-nalities. There is currently no effective policy tool for internalising the long-terminternational costs of climate change.

Recommendation:Policy shouldcontinue to strivefora globallong-termcarbon price.Moreover, policy should complement existing instruments with incentives for earlyinvestment into R&D, demonstration and deployment of clean technologies. Cleantechnologiesareessential to achieving timely carbon mitigation.

1.2.2 Innovation externality

Innovation, especially as it pertains tospecialised technologies, comesat a cost thecost of R&D. Although acquired knowledge may offset the cost of innovation for theinvesting rm, this knowledge may be non-rival and non-excludable. This means thatother rms may acquire the ability to imitate these innovations and lower their ownproductioncosts,withouthavingincurredthecostofR&D.Additionally,eveniftheydonot have theability to imitate the specic innovation, they may gain some benecialknowledge spillovers fromthe innovator. Therefore,R&Dinvestmentsconfer a positiveexternality to outside rms. This results in a situation where individual rms under-invest in R&D because they do not fully internalise the social benets of R&Dinvestments or because they anticipate costless benets to be gained from theinvestments of others.

This effect is present in all sectors. In order to facilitate the internalisation of this

externality, several policy instruments are available: protecting intellectual propertyrights(eg throughpatentsor tradesecrets), government fundingfor R&D and subsidiesfor private R&D.

Patents are a tool for removing the non-excludable aspect of innovations. Makinginnovations excludable would prevent rms that did not participate in R&D fromreaping thebenetsof theresultingtechnology at zero cost. Inaddition,excludabilityhastheaddedbenetof reducing incentives tosecrecy over technologicalknowledgewhich may benet society. However, patents are an imperfect tool. A strong patent

system increases incentives to innovate but decreases competition. As perSchumpeters theory of creative destruction, market power is a driving force of


11


23/126

innovation as innovation is a mechanism for destroying the market share of competitors. Therefore, in practice, patents are characterised by two dimensions:lifespan,andbreadth. These twodimensions inuence thedegreeofeffectivenessof patents in encouraging innovation. In addition, enforcement effectiveness andenforcement speed are issues which affect the impact of patents on innovation.Therefore, patent effectiveness also relies on effective institutions, and trust in theinstitutions of individual countries. The energy and transport system transition is aninternational effort and will rely on institutional strength in multiple countries. Thus,strong international patent protectionmight increase thenumberofgreen innovations.However, strong international patent protection also allows innovators to demandhigher prices for their more exclusive rights. This might decrease the rate of marketuptake. Another weaknessofpatents in internalising theinnovation externality is thatthey often cannot be applied to process knowledge (eg Fords assembly line), whichin many cases can only be protected via secrecy. Thus,patents alone cannot ensurean optimal level of innovation activity.

Anotheravenue of internalisationhas todowith relieving theimbalancebetween coststo innovators and the social benets of the innovations. This can be done viagovernment-fundedR&Dorvia government subsidies for private R&D.As the positivespillovers from energy and transportation technology innovations are essentially apublicgood, itmay makesense for governmentstocontribute to the costofproducingthem. Implementation is, however, key, as public R&D money risks simply replacingprivate R&Dmoneywithout increasing theoverall innovation level9.GovernmentsmaysponsorR&Dwhollyorviapublic-privatepartnerships.Insuchpartnerships,itisoftenthecase that intellectualproperty resultingfrom collaboration issharedviapatents orcontractual stipulations.

Consortiums of members from government, industry, and academia may provide a

way to direct R&D toward industry-applicable solutions. Consortiums, although theyproduce more general intellectual property, may be an important avenue forcoordinating efforts and may partially internalise the non-excludable nature of innovation. Furthermore, academic research which is wholly publically funded runstheriskofnotbeingadoptedoradoptablebyindustry.Consortiumsareawaytoshareintellectual property rights and the costs of producing intellectual property (see Box1). However, the R&D collaboration between competitors in the product market risksentailing anti-competitive effects10.


12

9. Davidet al (2000) argue that the literature on whether public R&D is a complement or substitute for private R&D

has been inconclusive.10. Seeforexample GoereeandHelland (2009).


24/126

Conclusion:Innovation carrieswith it a positive externality.Without properpolicies toprotect intellectualproperty oralleviate theprivate costsof innovative activities, theremay be underinvestment in R&D.

Recommendation:Government policy should augment investment in R&D for areaswhere intellectual property protection is not enough. Consortiums may be useful inencouraging industry-oriented innovation and may alleviate some of the issuescreated by non-excludability.

1.2.3 Path dependencies

Thetransition from oneenergysystemtoanother maybesubject topath-dependenceon,or lock-in effectsfrom,existingsystems.Path-dependence or lock-in, in themarketfailure sense, is the inability of the market to switch technologies despite theknowledge that the incumbent technology11 is inferior or undesirable relative to analternative (Liebowitz and Margolis, 1995). This is often due to the switching costsbeing higher than the benet for somepivotal actors in the system.


13

11. Path-dependence based on insufficient knowledge at the beginning is notex-ante inefficient but can beex-postinefficient.

BOX 1: R&D CONSORTIA

The VLSI (Very Large Scale Integrated circuit) project was designed to help Japancatchupwith semiconductor technology.Theproject wasconducted between 1975and 1985 with a budget of 1.25 billion, of which 22 percent was nanced by thegovernment. It developedstate-of-the-art semiconductormanufacturingtechnology.All the major Japanese semiconductor producers participated in this project, andJapanesesemiconductor companies gainedworld leadership in thisperiod.Beyondthis anecdotal evidence, it was found that consortia have the effect of stimulating

innovativeactivityintheselectedrms.However,thiscomesatacost.Amongothercomponents, these costs include theeffectsof reducedcompetition,administrativeburdens on the research personnel of participating rms, and cost of governmentsubsidies.

Take-home message: R&D cooperation between competing companies mightstimulate innovation but can have high long-term cost.

Source: Sakakibara(1997).


25/126

The market failure can occur for a number of reasons: lock-in due to uncertain payoff functions; lock-in due to learning-by-doing; institutional lock-in; and lock-in due tonetwork effects.

Lock-in due to uncertain payoff functions

Often, when a new technology is introduced, its future payoffs are uncertain, ie eventhe distribution of payoffs is unknown. Cowan (1991) develops a model of lock-inreferring to technologies of unknown merit as tortoises and hares. He demonstratesthat the reduction of uncertainty stemming from the adoption process may lead tolock-in. Oneillustrativeexample of lock-indueto learning-about-payoffs is theexampleof the two-armed bandit slot machine. Each arm of the slot machine has a differentdistributionofpayoffs.However,overtime,theplayermayconvergeononearmifitisused more. As the player learns more about the payoff distribution of one arm (theone which is used more), he refrains from investing money to obtain knowledge onthe payoff distribution of the other arm. Similarly, this analogy can be applied totechnologies of unknown merit. Costly learning abouta priori uncertain payoff functions can create a lock-in effect.

Lock-in due to learning-by-doing

Learning-by-doing can lead to technology lock-in. A more frequently-used technologytendstomovealongitslearningcurvefaster,andmaythuscauseacost-relatedsnow-ball effect where adopters continue along the path even with the knowledge that thetechnology is inferiororundesirable.Thus, an inferiorbutmore-developedtechnologymay become locked-in. This lock-in is exacerbated over time. Acemogluet al (2009)show that even research tends to build on the shoulders of giants, the giants beingincumbent technologies.

Institutional lock-in

A potentially less obvious form of lock-in is institutional and policy lock-in. Theautomotive industry is anexample ofan industry forwhich institutional lock-in exists.Both formal and informal private institutions have developed alongside the internalcombustion engine technological system. Knowledge-based institutions, such astechnical schools, developed to train labour for servicing a growing auto network.Higher-learningdisciplinarydepartments like highway or automobile engineering are

intrinsically linked with the automobile industry. Industry approaches may becomelocked-inas a curriculumfor long periods. In addition,unions, industry associations,


14


26/126

and media (egMotor World, Motor Age) have emerged. The existence of specialisedlabour for this technology creates a sort of lock-in.

Public institutional lock-in may also occur. Subsidies or government institutions canhave long-term impacts and persist for longperiods of time. Williamson (1998) foundthat formal institutions change over decades while informal institutions, such ascultureornorms,changeovercenturies.Inthecaseoftheautomobileindustry,alargenetwork of institutions, including theAmerican Automobile Association, the AmericanRoad Builders Association, and National Automobile Chamber of Commerce, formedalongside the technology; the highwaylobby is still seenasone of themostpowerfulinterest groups (Unruh, 2000). The existence of government institutions specic to atechnology might lead to policy biases as the obsolescence of a technology wouldmean the obsolescence of that institution. Public support of certain standards ortechnologiesmayexacerbate lock-in. This wasthecase forlight-waternuclearreactoradoption(seeBox2fordetails).ThepoliciesadoptedbytheUSwerebiasedtowardonetechnology (Walker, 2000), and this alsocontributed to the choiceof that technologyin Europe, due to US aid (Cowan, 1990). Another example is the case of German coalsubsidieswhichhavepersistedlongafterGermancoal becamemuch more expensivethan imported coal. Path dependencies exist due to the skill-set of the workforce inthe Ruhr Valley of Germany,coupled with existing subsidies.

In general, institutional lock-in has the potential to create non-market forces thatenhance technological lock-in. Institutional policy can override the neo-classicalmarket forces of competition by removing uncertainty about the direction of technological development. Firms might then favour a certain technology not inresponse to market forces but to institutional ones. Care should be taken to developpolicies and institutions which are not prone to lock-in and which are exible tochanging environments ie not technology-specic. The fact that technology lock-in

occurs naturally due to imperfect information and learning curves provides a case forgovernment intervention. In the initial stages of a technology, when its merit is asyetunknown, government support should not be biased toward support for a singletechnology. Switching support at a later time is less effective and more costly due tolearning costs (if no investment or learning was done in the interim), and potentialnetwork effects from theadopted technology. Thus, prior to the creation of subsidiesand other institutions, governments should carefully consider their necessity anddetermine the ease of shifting from the respective subsidies and institutions in thefuture. This premeditation on the part of government would help in avoiding

institutional lock-in.


15


27/126


28/126

fuel producers, fuel distribution,andconsumers)createa classical lock-ineffect.Thisproblem isexacerbated whenswitching costsare high. Switchingcosts are a functionof: the investmentcostof thecomponents, theminimum(efficient)networksize, andthelevelofuncertainty about thenewsystem. Thus,highlycapital-intensivesystems,with signicant scale economies and a large number of alternatives will be the mostdifficult to replace. Consequently, bringing a new technology to market in such asystem requires that it is either largely compatible with the incumbent system (egthrough hybridisation or the use of adaptors), to benet from incumbent networkeffects,or that it manages to effectively deploy its own system.

Conclusion:When governments, rms and consumers must choose between tech-nologies of unknown merit, technological and institutional lock-in may occur wherepath dependency on a suboptimal technology develops. This can become even morepronounced when network effects are also at play.

Recommendation:Technology-neutrality in public support at the early stages, whenthe winners are unclear, is important. Support may be needed to overcome pathdependency on established networks.

1.2.4 Coordination externality

Any transition from one system to another requires coordination among thestakeholders. As the energy and transport sector is very capital-intensive, lack of coordination during the transition is very costly. Standardisation is a primarycoordinationmechanism. AccordingtoSwann(2010), there arefour differentpurposesforstandardisation: compatibility/interfacestandards andvariety reduction standardsare utilised to reduce the fragmentation of a network, and to provide compatibilityrequirements in order to allow entry into the market. Minimum quality standards are

important inensuring a levelofsafety forconsumers. Informationstandardsareusedto homogenise information in order to lower information costs (eg for comparingproducts).

Thus, standardisation is critical for the development of complex markets. However,standardisation itself is costly, and the incentives to sponsor standardisation do notnaturally lead to a welfare-optimal set of standards.

Firms involved in standardisation (typically the rst-movers in a market) will have

heterogeneous preferences. Ontheone hand,eachrm wouldprefer a set ofstandardsthat increases thevalue of thecapabilities, patents and business model of said rm.


17


29/126

On the other hand, all companies want to avoid a situation in which the adoptedstandard locks them out of the market. Due to the high number of stakeholdersinvolved, and the complexity of the technical questions, complicated negotiationsbetween stakeholders may emerge. These might take years and consume valuableresourcesandtime.Consequently, rst-moverswhoparticipate in thecoordinationof standards impose a positive externality on late-comers by absorbing the costs of standardisation. Late-comers can free-rideon thecoordinationefforts withoutpayingthe price for it.

Due to this market failure, if left to their own devices, rst-movers may prefer to formfragmented networks/markets to avoid laborious and costly coordination (see Box 3for a case-study on a failed standardisation effort). Alternatively, in the absence of otherstakeholdersinthecoordinationprocess,aminorityofengagedcompaniesmaypush through astandard that isclearly not in thebest interestofsociety.Consequently,public intervention might prove valuable for resolving this market failure. Publicintervention in standardisation should take place when there is a weak, or no,coordination mechanism between competing companies in a market. Governmentsmay intervene through administering the standardisation process, and/or throughnancial subsidies to overcome coordination problems at the R&D stage, and/or tomitigate the deployment barriers imposed by market competition. Public-privatepartnerships areanother avenuefor facilitating coordination.

It is worth noting that government interests may not necessarily be in line with theshort-term interests of companies active in the standardisation process. The publicsector puts greater importance on customer interests and on the long-term health of standards infrastructure. Standards-setting can have different effects on innovationandR&Dinvestmentsdependingupontheimplementationandstrictnessofstandards(see Box 4 for thepharmaceutical industry case study).


18


30/126


19

BOX 3: JAPAN AND THE CELLULAR GALAPAGOS

Japanesecellularphonecompanieshave long been leaders indelivering advancedhandset technologies and services to their consumers. The rst cellular dataservices were availableinJapanin1997 atextraordinarily low rates. Cellularphonesare widely usedinJapan assubstitutes for nancial instruments likedebit and creditcards. Yet despite technical leadership that is often years ahead of foreigncompetitors, Japanese rmshave hadalmostnosuccessselling intoforeign cellulardeviceand servicemarkets.

As Kushida (2011) explains, the Japanese inability to capitalise on leadership innetwork technologies stems from both public and private decisions. Domestically,device and service companies were vertically integrated, allowing them to resolvethe tension between network and product introduction by internalising the designand deploymentofboth.This led tovery early, highly competitivemarkets incellulardata services, perhaps ve years before similar markets emerged in the US andEurope. But in practice, competition in an isolated market meant that Japaneseproducts andnetworkstandardsdivergedfrom internationalnorms inorder todeliverever-more-exotic products to customers.

As a result, when Japanese rms then attempted to take their advanced tech-nologies abroad, they found they were incompatible with the networks thosemarkets depended on.It mattered little that Japanese technology wasyearsaheadof the competition. Japans failure to keep international standards and domesticmarketsinsynclockeditoutofcapturingexportbenetsfromitsdomestictechnicalleadership. Like Darwins nches, Japanese cellular rms were highly adapted totheir isolated market,but bizarre and strange creatureselsewhere.

Take-home message: International coordination and the development of inter-national standards is key. Government policies and rms should pay attention tothe developments of other nations and coordinate/adapt to changing internationalstandardssoas tobecompetitive on theglobal market. Technologicalsophisticationisnot enough to ensure the success ofa technology.

Source: Kushida (2011).


31/126

Public intervention may have different impacts on standards development. Positiveeffects are generated by the facilitation of coordination activities and theestablishment of special bodies for this purpose. Negative effects can emerge whenpreferencesaregiven tocertain technologies byofficials,from a political orshort-termperspective. Illustrations of a positive and a negative result are provided in Box 5.


20

BOX 4: STANDARDISATION IN THE PHARMACEUTICAL INDUSTRY:A BARRIER OR INCENTIVE FOR R&D?

Excessivelystrictstandardsmaystie innovation as they mayaffect theprotabilityof certain products. In the case of the pharmaceutical industry, investments inproductdevelopmentarenotmadeonthebasisofpotentialbenetstosociety,buton the basis of maximal future returns. Therefore, if standards are overly strict,companies may decide not to develop or produce medicines effective in treatinglife-threatening diseases but which do not meet protability criteria. Excessive

regulatory burden can result in a decrease or cancellation of R&Dfor certain drugs.Inthepast20years,severalcountriesandregions(Australia,theEU,JapanandtheUS), have adopted orphan drug legislation (ODL). Incentives included fast-trackprocedures for standardisation and reduced registration fees. This legislation hasbeen successful in the promotion of the development of drugs for rare diseases ordiseases which areprevalent in poorer countries.

Take-home message: Governments need to ensure in developing standards thattheir strictnessdoesnot make investment inproducts that are benecial tosociety

unprotable.Source: Reichetal (2009).


32/126


21

BOX 5: EUROPEAN COMMISSION INTERVENTION IN STANDARDISATIONIN TV AND TELECOMMUNICATION INDUSTRIES

Failure case: HDTV standardIn1986, theCommission initiated thedevelopment ofa high denition standardforTVset manufacturers andbroadcasters (HD-MAC). This effort was in responseto Japansattempts to set up the worlds rst high-denition TV standard. To stimulate earlydeployment theCommission made thedecision to favour a certain technology duringthe early stages of the process. Due to its high complexity the technology promised

highermarginstoTVmanufacturers.Butitwasalsomoreexpensiveforbroadcastersandconsumers.Consequently, thetake-upofthenewtechnologywaslimited. In1992,theCommissiontriedtointerveneagainandofferednancialsubsidiesforthebroadcastingcompaniesinexchangeforagreeingwithTVsetproducerstohavestandardsforanewversion of HD-MAC. However, it was too late for HD-MAC due to the advance of digitalbroadcasting technologies. Public support for TV standardisation has now beenredirected towardswidescreenTVstandards corresponding todigitalbroadcasting.

Success case: GSM standardIn the 1970s, the Commission initiated the Global System of Mobile Telecom-munications (GSM)project inordertonarrowthegap between US/JapanesetelecomcompaniesandEuropeanones.IncontrasttotheHDTVcasetheCommissionrefrainedfromusingsecondary legislation andconcentrated itseffortson thenegotiationofamemorandum ofunderstanding (MoU)between allstakeholders.TheMoUcommittedoperators to open procurement to foreign manufacturers, and manufacturers toprovide royalty free licenses and to have GSM operational by 1991. This allowedmanufacturers to deliver the new generation of GSM infrastructure equipment in atimelywayandatreasonablecost.Asaresult,deploymentwasrapid,andgeneratedaffordable margins throughout the value chain. This eventually tipped the entire

Europeanandglobalmarket towards thestandard.Togetherwith thenext generationUMTS standard, the GSM family of standards eventually captured 89 percent of theinternationalmarket, according to the Global Mobile SuppliersAssociation.

Take-home message: All stakeholders opinions regarding the new technologystandards have to be taken into account. For this purpose, public support shouldconcentrate strictly on technical problems and avoid politically motivatedpreferences regarding the choice of standard.An MoU is considered a good tool forbringing all the conicting interests together and nding a mutually benecialframeworkofcooperation toreducethebarriers to investment insidethevalue chain.Source: Meyer(2010).


33/126

Public authorities need to nd the correct balance between socially importantstandards (eg safety) and promoting R&D inthe sector. Itmay bereasonablenot tosetstandards too strictly during pre-market development, in order to maintain enoughstimuli for private investment in R&D.

Often,withweakornocoordinationmechanismsinplace,iflefttoitsowndevices,themarket produces too little, too much, or standardisation of the wrong sort. However,public sector involvement is only helpful if properly implemented. Additionally,coordination amongst governments themselves is a key factor in the success of astandard.

Conclusion:Both domestic and international, and public and private cooperation arekeytothesuccessofanenergyandtransporttransition.Thiscoordinationiscostlyandmay create a positive externality discouraging rms from entering as rst-movers.Fragmented networks and markets may arise.

Recommendation:Public intervention into standardisation should take place whenthere are weak, or no, coordination mechanisms between competing companies in amarket.Public authorities should nd the right balance when determining standards.Domestic standardisation should take the international standards environment intoconsideration. Financial subsidies to overcome coordination problems at the R&Dstage, and to mitigate thedeployment barriers imposed by market competition, are apopular form of intervention.Public-privatepartnershipsareanother avenue forfacili-tatingcoordination.

1.2.5 Infrastructure externality

A newlow-carbon energyandtransport systemmayrequirea large-scale infrastructure

shift due to the use of alternative fuels. First-moving infrastructure providers face adisadvantage because they must pay a large xed cost in order to establishinfrastructure that is the precondition for the deployment of the appliances. Late-coming infrastructure providers may benet from the established network withouthaving paid the costs of building the network. As latecomers can install the latesttechnology and larger units they face lower service cost. This cost-advantage allowslatecomers to cherry-pick the best locations once the network and market have beenestablished. Consequently, competition from latecomers may decrease prices, andthereby decreasetheprotof therst-movers.First-movers may potentially beunable

to recover their initial investments. This externality could lead to underinvestment orno investment where competitive prices do not allow recovery of initial investments,


22


34/126

and protability, in the long-term. In the past, several mechanisms to internalise theinfrastructureexternality have proven successful:

The natural monopoly solution

Anaturalmonopolyemergeswhenunitcostsdecreasewiththenumberofclients12. Inthe case of a natural monopoly the presence of an early-mover locks out anylatecomers.Water, gas, telephone, broadbandandelectricitydistributionnetworksarenaturalmonopolies.Consequently,theearly-movercanrecoverany initial investmentby raising prices when the network has been established. This stimulates rapiddeployment but raises the issue of excessive prices. Consequently, mature naturalmonopoliesaretypically regulated13.Thus,themainissueforsettingupanewnetworkwith natural monopoly characteristics is the uncertainty about its future regulation.

The state monopoly approach

Some network industries are not characterised by natural monopoly characteristics,due to the facility of competitive local provision (eg postal services, fueling stations).Governmentsoften create public monopolies in order to ensure country-wide accessto services and prevent cherry-picking of the most protable locations. Typically,governments have linked the provision of a state-monopoly to a certain (sometimesprivate) provider with a universal serviceobligation.

The articial monopoly/oligopoly approach

For some network technologies, infrastructure is not a natural monopoly (eg mobiletelecommunication, Facebook), ie multiple infrastructure providers can coexist andcompete for network users. Here, some companies attempt to lock in customers by

creating articial barriers (locked SIM cards for mobile phones, non-exportablecontacts in social networks) in order to be able to pay their xed investments innetwork infrastructure14. Technologies that failed to lock in customers (eg emailproviders) quickly saw their prot margins drop to zero. Companies that can lock in


23

12. More precisely, when the long-run marginal cost is below the average cost of providing the service, no secondcompany hasan interest to enter themarket.

13. There is signicant literature on thequestion of if andhowto optimally regulate monopolies (egTroesken, 1996)14. Therstelectricitycompanies startedby leasing light-bulbs toconsumers(Hughes,1977). Thus, theentire value

chain was in the hands ofone company and it could set a price thatcovered xed and variable cost.For the initialelectrication ofManhattan,Edison essentiallydesignedeach of thecomponentsaspartof an integrated system

wheretheresistanceofthebulblamentwaspickedtohelpmodulategridloadbasedonthenumberofexpectedservice subscribers.


35/126

customers by making interfaces incompatible have higher incentives to quickly roll-out of their networks. So they invest more. In welfare terms this positive effect iscounteracted by the long-term negative effects from having a monopoly.

The consortium approach

Consortiums are another way to help ensure that the investments in networkinfrastructure can be recouped. If consortiums can lock out competitors theconsortiums might internalise future benets of creating a new infrastructure. Thisagain might stimulate early deployment but may also have undesirable effects oncompetition.

The vertical integration approach

The existence of infrastructure is essential for appliance and service providers.Consequently, they might engage in setting up the infrastructure themselves. This islike selling camera bodies cheaply in order to be able to develop a market for lenses.This, of course, requires that competing service/appliance providers are locked outfrom the network (see articial monopoly).

The cross-payment approach

A less-integratedapproachwould be forservice/applianceproviders tocross-subsidiseotherpartsofthevaluechaininordertocreatethenetwork.InGermany,forexample,somenatural gas distributioncompaniesprovidesubsidiesfor natural gas cars inorderto develop the market (seeBox 12).

The public provision approach

Finally, governments might providethe infrastructure using public funds or regulatedreturns. To date much infrastructure (such as roads, railways and electricitytransmission) is provided by government in most European countries. It has to benoted, however, that theinitial technology choice inmany of thesecases (eg railways)was left to the market with governments nationalising the infrastructure only afterit appeared.

Conclusion:The huge xed cost of infrastructure construction may not be fully

recoverable ina competitiveenvironment,duetonewcomersenteringandcompetingwithout having had to pay the initial xed cost.


24


36/126

Recommendation:If infrastructuredevelopment costs arenotrecoverable in thelongterm by the rst-movers, support instruments should be considered to encouragedevelopment of the necessary infrastructure.

1.2.6 Business exploration externality

In marketing, rst-mover advantage is the advantage gained by the initial signicantoccupant ofa market segment. It may be also referred to as technological leadership.This advantage may stem from the fact that the rst entrant can gain control of resources or build a brand that later market entrants may not be able to match15.I t isimportanttonotethattherst-moveradvantagereferstotherstsignicantcompanytomoveintoamarket,notnecessarilytherstcompany.Inorderforacompanytotryto become a rst-mover, that company needs to work out if the overall rewardsoutweigh the initial underlyingrisks.Sometimes, rst-moversarerewardedwith hugeprot margins and a monopoly-like status. Other times, the rst-mover is not able tocapitaliseon itsadvantage, leavingopportunitiesforother rmstocompeteeffectivelyandefficientlyversustheearlierentrant.Thesecompaniesthengainasecond-moveradvantage16.

First-movers face risk in both exploring technologies (see section 1.2.2 for furtherinformationregardingtheinnovation externality), andindevelopinga market. Therstcompanies investing in a new technology face signicant risk, as their businessmodels are based on uncertain assumptions. If successful, implementation mayquicklybe imitated bycompetitors. Fallingmarginsmightmakeit impossiblefor a rst-movertorecoveritsinitialinvestmentatareturnwhichiscommensuratewiththeriskstaken. Followers clearly have some cost advantages of their own. They can, forexample, learn from themistakesandsuccessesof theirpredecessors, reducing theirown investment requirements as well as risks. In addition, followers can frequently

adopt new and more efficient processes and technologies, whereas pioneers oftenremain entrenched in their original ways of doing things. Finally, followers will havelower marketing budgets for convincing thepublic that the (now familiar) technologyworks.

Accordingto Boulding andChristen (2001), for instance,pioneers inconsumer goodsmarkets andindustrialmarkets gained signicant salesadvantages,but incurred largercost disadvantages. Pioneers in consumer goods had an return on investment thatwas 3.78 percentage points lower than later entrants. And the ROI of rst-movers in


25

15. Grant (2003).16. Lieberman and Montgomery(1988).


37/126

the industrial goods sector was 4.24 percentage points lower than later entrants.Pioneers were less protable than followers over the long run, controlling for all otherfactors which could account forperformance differences17.

In someof the standards races thathave taken place, suchas inpersonalcomputers,audio recording media and video cassette recorder formats, the winner was notnecessarily the rst-mover (Box 6).

The market failure is that the commercialisation of a technology reveals information.This isa valuable inputfor thedecisionsofpolicymakers, of industryand ofconsumers(see lock-in due to uncertainty in section 1.2.3). In a pure market solution, the riskof failure for a new technology is privatised, while the benet is socialised to somedegree, essentially leading toprivateunderinvestment. Somestrategiesforcompaniesto internalise thesebenets (egcoordinationamongst competitors,monopolisingthenew energy technology) raise competition concerns and might not be welfare-maximising in the long term.

Conclusion:The costs ofexploring, and building, new markets ishighand may not befully recoverable given that new entrants may reduce prot margins. Due to thepositive externality of business exploration, companies may be reluctant to be rst-movers in certain sectors.

Recommendation:Policy should address the business exploration externality wherenecessary, providing incentives for the exploration and development of promisingmarkets and technologies.

1.2.7 Insurance externality

Energy transitions are inherently subject to a high level of uncertainty. As they arenationalorglobalshiftsinthewayenergyisproducedandconsumed,theyaresubjectto exogenous economic, political, and even geological events.Energy transitions areoften characterised by long time horizons from initial R&D investments to fulldeployment. Large investmentsare required throughout theprocess, and, due to theuncertainnatureofenergytransitions,theseinvestmentscarrywiththemahighlevelofrisk.Transitionsareverydifficulttomodel.Predictionsaboutthedurationandspeedof transitions have seldom been accurate. For example, in terms of 2008 primaryenergy share, coal was still at 20 percent versus a 5 percent share predicted by


26

17. Boulding and Christen (2001).


38/126


27

BOX 6: VIDEO FORMAT WAR: VHS VS. BETAMAX, OR THE CASE OF CORRECTUSE OF NETWORK EXTERNALITIES

The classical illustration of technological lock-in was the war of the video cassetterecorder (VCR)formats whichoccurred in1980s betweentworst-movercompanies Sony and JVC. Each company began by releasing two different formats: VHS andBetamax.

The cornerstone of VHSs success at the early-market stage was its recording timeof two hours as compared with the one hour provided by Betamax. Sony believedthat cassette size and transportability were paramount to the consumer, andsacriced playing time in order to make smaller cassettes. Simultaneously, JVCconcentrated its efforts on the availability of VCR machines by setting up rentalchainssuchasRadioRentalsorDER.Theserentalchainsofferedanattractivechoicefor consumers who did not want to spend a lot of money on a system which mightbecome obsolete. The ourishing video cassette rental business of the1980s wasreliant on the VHS formatas a more suitablemeans of storing movies.

When themarketmatured, thewide availability of recordersandpre-recordedtapes

inVHS formatbecamea key factor inJVCs victory,allowing it tobecomeanabsolutemarket leader. AlthoughBetamax initiallyowned100percentof themarket, in1975,theperceivedvalue of longer recording timeseventually tipped thebalanceinfavourofVHS.Sony,astherstproducertooffertheirtechnology,thoughtitwouldbeableto establish Betamax as the leading format. This kind of lock-in failed for Sony, butsucceeded for JVC.For thirty years, JVC dominated the homemarketwith theirVHS,Super VHS and VHS-Compact formats, and collected billions in royalty payments.

Take-home messages:

First-movers do not necessarily prevail.

Consumer preferences may tip the scale in a competition between similartechnologies. Attentionshould be paid to accurate evaluation ofwhat aspectsof the technology are most important to consumers.

Investment in downstream (retail) suppliers may help resolve the chicken-and-eggproblem.

Sources: Besen andFarrell (1994), LeibowitzandMargolis (1995).


39/126

Marchetti in1970, and the 23percent share ofnatural gas was far below the 60percentpredicted (Smil 2010). Predictions even over the course of 40 years are not reliableeven though time horizons for energy transitions are very long18.

Despite theuncertainty ofsuccess forindividual energytransitions, it isapparent thatwe are currently at the brink of a new era in energy. A drastic change in the way weproduce and consume energy must occur in order to avert a global environmentalcrisis. The SternReview(2006)estimates the costof inaction (with respect toclimatechange)tobeequivalenttolosingatleast5percentofglobalGDPperannumnowandforever, possibly rising to 20 percent if including other risks and impacts19.

It is difficult to know, at the current stage, the cost-effectiveness or feasibility of different green technologies.Earlyperceptionsofnuclear as too cheap tometer havebeen incorrect; nuclear is much more expensive than predicted (Cohn, 1997).Similarly, thecostsof new technologies may change due to materials availability (eglithium for batteries, platinum for fuel cell membranes) technological constraints(nuclear fusion)orchanging publicacceptability (carbonsequestration,nuclear, shalegas). Attempts to predict affordability have sometimes included the use of learningcurves savings in cost due to learning-by-doing and R&D. R&D investments createsteeper learning curves (higher cost savings) but their actual impact is difficult tomeasure due to the lackof availability of private data. Learning curves may be a wayto predict future costs but are an imperfect tool as much depends on external factorsunrelated to learning. In addition, technologies may have differently-sloped curvesandthussometechnologieswhichappearcurrentlyexpensivemaybecheaperinthelong run, whereas some immediately viable technologies may have learning curvesthat level out and will thus remain expensive.

Anenergytransition formitigatingGHGemissionswill require decisionsonboth public

and private investment. High levels of uncertainty, coupled with positive networkexternalities,maylead individual rms toconvergeona technologyorenergysystemthatproves suboptimalex post .Rationalagentsmaybehaveoptimallybycopyingthebehaviour ofothers inorder to reducerisk, asopposed toactingsolelyon the basis of their own information, due to an information cascade effect. This is called herding


28

18. Energytransitionsare dynamic innature, andoften subject tounforeseeable/unpredictableevents(egtheGermanreaction to the Fukushima accident).

19. Thishas beena widely debated report,for example,Nordhausresponse(Nordhaus, 2007) stronglyquestionsthediscounting method used in the review. However, there is growing concern about the environmental impacts of

GHG emissions, as evidenced by the Kyoto Protocol and the EU 2050 goals, and that there will be a huge costincurred by inaction.


40/126

behaviour and can occur when there is uncertainty and imperfect information(Bikhchandaniet al , 1992). Herding to a single technology might make it very costlyfor society if this technologyprovestobean inefficient or insufficient solutiontomeetemissions targets20. Thus, public intervention to discourage herding and to nurturealternative technologies, as an insurance against the risky nature of any energytransition, might increase expected welfare.

Energy security literature asserts that portfolio diversication is important in themaintenance of current energy security and fuel mix portfolios for individualcompanies(Roquesetal ,2008;Lesbirel,2004).Theheavyrelianceofsomecountrieson a single technology (eg France on nuclear, Ukraine on gas) poses energy securityrisks. Investments in newenergiesandfutureenergysystems shouldsimilarlyadopta portfolio approach. Investment in a backstop technology may prove to provide ahuge positive insurance externality in the case that the chosen primary technologyfails. The benets of this externality would not be automatically internalised. This isanopportunityfor thedevelopment ofpolicyinstruments toencourageinvestment inbackstop technologies. Although it is expensive to invest in the development of newtechnologies as an insurance policy, the potential cost savings, in the event that abackstop technology is needed, are huge and there is risk of substantial cost in notdoing so.

Conclusion: Diversied investment in alternative technologies may provide aninsurance against failure of primary technologies in meeting the energy and climatechallenge.

Recommendation:Governments shouldadopt a long-termperspective in technologyinvestments and diversify support over technology portfolios, possibly includinginvestments in backstop technologies.

1.2.8 Industrial policy externality

Fromcost tobenet: can newenergyand transport technology policy generate jobsand growth?

Inventing, buildinganddeployingtheinfrastructureandcapital requiredfora newlow-emission energy and transport system presentsa range of opportunities to generateemployment, renew rm competitiveness in existing sectors, and foster rm


29

20. Seesection 1.2.3 forfurther information regardingpossible path dependency and lock-in.


41/126

competitiveness in new sectors. These opportunities have not escaped the notice of national and regional governments. But most climate policy has emphasisedminimising thecostsof transition toa low-emissions transport system. Stateshopingto capitalise on that transition to generate employment and technological leadershipface a very different set of challenges.

Modern energy and transport systems provide an array of sophisticated services toindustrial economies. At present, it doesnt appear that a low-emissions transportsystem would make substantial improvements to these services. Rather, it seeks tocontinue to provide thesame services withoutgeneratingthedamaging by-products.Thus the introduction of a new transport system does not offer to radically transforman array of economic domains in the way that the introduction of the railways or theprivate car did. This reality restricts the potential economic benets that states maycapture during the transition process itself.

As Hubertyet al (2011) have shown, these benets come down to essentially threedomains:

1. Creation of domestic jobs to build and operate the new energy and transportinfrastructure;

2. Creationofdomesticjobs tomanufacture thecapitalequipmentrequired toreplacethe old fossil fuel-based capital stock;

3. Creation of globally-competitive rms in green export markets through domesticinvestments in research, development and deployment of new goods.

Thissection discusses eachof these inbrief. All threeof thesepotentialbenets from

theadoptionof low-emissions transport networkspose novelchallenges fortransportpolicy. Inparticular, translatingdomesticmarket growthinto commandofglobal exportmarkets traps policymakers between two dilemmas of network selection. Domestic-ally, the choice of the optimum network may require a lengthy process of experi-mentation to guard against the risks of network lock-in discussed above. But abroad,command of export markets requires that states position their industries to sell intoglobal network standards. Balancing the optimum choice of networks at home withinuence on global standards and access to export markets abroad becomes theprimary challenge for industrial policy. But, as we shall argue, this challenge differs

bystate size andexisting industrial capacity, furthercomplicating thedilemma posedby green industrial policy for the state and rms alike.


30


42/126

Job creation at home: limits to duration and size

Transport and energy systems in advanced industrial economies are largely alreadybuilt, and demand for new capacity grows slowly. In that context, domestic jobs fromthe creation of a low-emissions energy system can come in two forms. In the initialphase,thereplacement ofsignicantnetwork infrastructurecan createjobs insectorssuch as construction and services. Since these jobs are in non-traded sectors, theywill almost certainly appear. However, they are necessarily time-limited. With thecompletion of infrastructure replacement,weshouldexpectthelabourdemands fromthe energy and transport sector to return to the pre-replacement equilibrium.Furthermore, these jobs will arrive regardless of the particular kind of technologyeventuallychosenfor thereplacement.Because they arenotexposed to internationalcompetition, jobs in these sectors pose relatively few challenges to thestate beyondthegeneralproblem of inducingandmanaging theenergyand transportationsystemstransition.

States may also wish to keep some or all of the jobs associated with capital goodsmanufacture for infrastructure and appliance replacement at home. This poses fargreater challenges to industrialpolicy. Creating high-productivity manufacturing jobsin new energy and transport technologies is only justied by substantial demand fortheseproducts. Smallstates lacktheeconomies ofscale required to justify investmentinlargesegmentsofthevaluechaintheysimplylackthevolumeofdemandrequiredto pay those investments back. Likewise, states poorly positioned in the coretechnologies and industrial capabilities required to build either the components of alow-emissions energy and transport system, or act as systems integrators for thingslike low-emissionsautomobiles,facevery signicantstart-up costsat thesector level,apart from thecosts to invent, pilot and deploy low-emissions technologies.

Theopportunitiesforjobcreationbasedondomesticmarketsalonearethuslimitedinboth time and space. Most economies will capture jobs in non-traded sectors likeconstruction as they replace the infrastructure of todays fossil fuel energy andtransport system with low-emissions substitutes. But capturing high-productivity,high-technology manufacturing jobs to supply the capital goods required will requirethat states makecareful choices about thescope and positionof their investments inthe value chain that will supply those goods.


31


43/126

Capturing export markets abroad: learning-by-doing, international standards, andthe risks of network mismatch

Thelimitstodomesticmarketsinallbutthelargesteconomieshaveencouragedstatesto look toexportmarkets assources of green growth. Ifnationaleconomiescan trans-late domestic expertise in new low-emissions energy and transport technology intointernational competitiveness,export-ledmarketgrowthcan createa rangeofdomesticbenets through access to markets much larger than domesticdemand alone.

The history of export-led market growth provides some evidence that national policycan generate signicant returns. As Box 7 shows, the experience with green goodssuch as wind turbines has shown that aggressive domestic market expansion in low-emissions goods can generate signicant le

transformation bp 180112 2

Documents