NETWORKS AS LEVERS · 2018-07-23 · MEDITERRANEAN ENERGY REGULATORS ... (MedReg) and the Energy Regional Regulatory Association (ERRA) of Central and Eastern Europe. 8 Introduction:

NETWORKSASLEVERS:

EMERGENCE,FUNCTIONINGANDEXPORTOFTRANSNATIONALNETWORKSOFENERGY

REGULATORS.

Author:FrancescaPiaVantaggiato

Supervisors:ProfessorChrisHanretty,ProfessorHusseinKassim

ThesissubmittedforthedegreeofDoctorofPhilosophyatthe

UniversityofEastAnglia

SchoolofPolitics,Philosophy,LanguageandCommunicationStudies(PPL)

UniversityofEastAnglia

Norwich,UnitedKingdom

"Thiscopyofthethesishasbeensuppliedonconditionthatanyonewhoconsultsitisunderstoodtorecognisethatitscopyrightrestswiththeauthorandthatuseofany

informationderivedtherefrommustbeinaccordancewithcurrentUKCopyrightLaw.Inaddition,anyquotationorextractmustincludefullattribution.”

2

TableofContents

ABSTRACT........................................................................................................................7

INTRODUCTION:RATIONALEOFTHERESEARCHANDKEYFINDINGS................................8

1. PREPARINGTHEGROUNDFORANALYSIS:FROMTHELITERATURETOTHESTRUCTUREOFTHETHESIS.............................................................................................17

TRANSNATIONALNETWORKSOFREGULATORS............................................................................18THEGAPSINTHELITERATURE:THEEMERGENCEANDTHEEVOLUTIONOFTRANSNATIONALREGULATORY

NETWORKS...........................................................................................................................25WHYTHEFOCUSONENERGYREGULATORS?...............................................................................29ANALYSINGTRANSNATIONALNETWORKSOFENERGYREGULATORS:THEQUESTIONS,THECONCEPTSANDTHEHYPOTHESESGUIDINGTHISRESEARCH.................................................................................34

Definitionsofkeyconcepts...........................................................................................41THEMETHODSOFENQUIRYANDCASESELECTION........................................................................44

PARTONE-EMERGENCEANDCONSOLIDATIONOFNETWORKSOFENERGYREGULATORS.................................................................................................................49

PARTI.INTRODUCTION:THESPONTANEOUSEMERGENCEANDCONSOLIDATIONOFTRANSNATIONALREGULATORYNETWORKS...................................................................50

2. THEHISTORYOFTHENARUC:DEFENDINGTHETURFOFTHEPUBLICUTILITYCOMMISSIONS...............................................................................................................55

THEEMERGENCEOFTHENARUC:THEPUBLICDEMANDFORREGULATIONOFTHERAILROADS............55TheShreveportcaseandWWI:questioningthelegitimacyofstateregulation...........59

THECONSOLIDATIONOFTHENARUC:POST-WWIARRANGEMENTS..............................................61CriticaljuncturesinthehistoryoftheNARUC:battlingattemptstogetridofPUCs...64Electricity:theutilities’demandforregulationofthemselves.....................................66

3. THEHISTORYOFCEER:THEQUESTFORLEGITIMACY..............................................72BEFORETHEREGULATORYSTATE:THEROLEOFENGINEERSINPROMOTINGINTERCONNECTIONSAND

CROSSBORDERENERGYTRADE.................................................................................................73THEEMERGENCEOFTHECEER:NETWORKINGFORLEGITIMACY.....................................................76THECONSOLIDATIONOFTHECEER:THETENSIONSOFMULTI-LEVELGOVERNANCE...........................82

ERGEGandtheofficialisedCEER:cherishingtheNRAs’self-rule..................................85Thetrainingofregulators:theestablishmentoftheFlorenceSchoolofRegulation....88TheCEERasapolicylever:policyinfluenceandinstitutionalgoals.............................89

RE-INVENTINGTHECEER:LAYERINGORCONVERSION?................................................................91CriticaljuncturesinthehistoryoftheCEER..................................................................93

PARTI.CONCLUSION:NETWORKINGFORCONTROL.......................................................96

PARTTWO–EXPLAININGNETWORKSTRUCTURE,NETWORKACTIVISMANDNETWORKINFLUENCE...................................................................................................................100

PARTII.INTRODUCTION:MEASURINGNETWORKCOOPERATION.................................101

4. THEDRIVERSOFTRANSNATIONALREGULATORYNETWORKING:VARIETIESOFCAPITALISMBETWEENHOMOPHILYANDCONVERGENCE.............................................104

INTRODUCTION...................................................................................................................104THEDETERMINANTSOFNETWORKTIES:LITERATUREREVIEWANDHYPOTHESES..............................106DATAANDMETHOD:ANEXPONENTIALRANDOMGRAPHSMODELOFSTRONGTIESAMONGEUROPEANREGULATORS......................................................................................................................117

Resultsofthemodel...................................................................................................122

3

Modelfit.....................................................................................................................131DISCUSSIONOFTHERESULTS.................................................................................................133CONCLUSIONS.....................................................................................................................136

5. NETWORKINGFORRESOURCES:HOWREGULATORSUSENETWORKSTOCOMPENSATEFORLACKINGRESOURCES......................................................................139

INTRODUCTION...................................................................................................................139NETWORKINGFORRESOURCES:LITERATUREREVIEWANDHYPOTHESIS..........................................142DATAANDMETHODS...........................................................................................................147

Analysis.......................................................................................................................151Discussionoftheresults.............................................................................................157

CONCLUSIONS:NETWORKINGFORRESOURCES..........................................................................160

PARTII.CONCLUSION:THERATIONALEFORREGULATORYNETWORKING....................162

PARTTHREE–EXPORTINGNETWORKS:THEROLEOFAMERICANANDEUROPEANREGULATORSINFOSTERINGREGULATORYNETWORKSABROAD.................................164

PARTIII.INTRODUCTION:EXPORTINGNETWORKS.......................................................165

6. THEESTABLISHMENTOFTHENETWORKOFENERGYREGULATORSOFCENTRALANDEASTERNEUROPEORFOREIGNPOLICYVIANETWORKS...............................................170

BILATERALAIDANDTRANSNATIONALREGULATORYCOOPERATION:STRANGERLITERATURES.............172“CONSULTANTSWERENOTENOUGH”:HOWTHENARUCBECAMEINVOLVEDINTHEUSAIDMISSIONIN

CENTRALANDEASTERNEUROPE............................................................................................176Theutilityandregulatorypartnerships......................................................................178NARUC:theobviousintermediary..............................................................................183TheestablishmentoftheEnergyRegionalRegulatoryAssociation(ERRA)................186TheendofERRAasaregulatorynetworkandthebeginningofERRAastrainingprovider......................................................................................................................194

NETWORKEXPANSIONANDFOREIGNPOLICYGOALS...................................................................195

7. NETWORKSAS“FIRSTBEST”?THEESTABLISHMENTOFTHENETWORKOFEURO-MEDITERRANEANENERGYREGULATORS......................................................................198

THEEU’SEXTERNALENERGYREGULATORYGOVERNANCE:TOP-DOWNANDLEGALISTIC....................200THEESTABLISHMENTOFTHEMEDREG:NETWORKSAS“FIRST-BEST”............................................205

ReplicatingtheCEERexperience.................................................................................210THELIMITSANDTHEPOTENTIALOFMEDREGFOREURO-MEDITERRANEANENERGYCOOPERATION....214

PARTIII.CONCLUSION:NETWORKSFROMNETWORKS.................................................219

8. CONCLUSIONS......................................................................................................222THEFINDINGSANDSTRENGTHSOFTHISRESEARCH....................................................................222THELIMITSOFTHISRESEARCH................................................................................................231GENERALIMPLICATIONSOFTHISRESEARCHANDFUTURERESEARCHAGENDA.................................233

REFERENCES.................................................................................................................237

APPENDICES................................................................................................................268APPENDIX1–INTERVIEWSINFORMINGCHAPTER2AND3.........................................................269APPENDIX2–DATAUSEDINCHAPTER4-ELECTRICITYSECTORSTRUCTUREINEUCOUNTRIES.........271APPENDIX3–ADDITIONALMODELSFORCHAPTER5.................................................................290APPENDIX4–INTERVIEWSINFORMINGCHAPTER6AND7.........................................................292

4

Figures(chapternumberthenfigurenumber)

Figure4-1-Visualizationofthenetwork................................................................120Figure4-2-GoodnessofFitofModel2:indegree,outdegree,edge-wiseshared

partners,minimumgeodesicdistance............................................................132Figure5-1–Relationshipbetweenstafflevelsandnetworkactivism....................152Figure5-2–Relationshipbetweenbudgetlevelsandnetworkactivism................153

Tables(chapternumberthentablenumber)

Table4-1–Hypotheses,VariablesandMechanisms..............................................123Table4-2-ERGModelsofthenetworkofEuropeanenergyregulators................125Table5-1-Associationbetweenregulators’networkactivism(dependentvariable),

stafflevelsandcovariates...............................................................................154Table5-2-Relationshipbetweennetworkactivism(dependentvariable),budget

levelsandcovariates.......................................................................................156

5

AcknowledgementsWritinghundredsofpagesonthefeatures,thehistoryandthepolicyrelevanceofnetworksofcollaborationbetweenregulatorshassomehowproveneasierthanputtinginwordsthegratitudeIfeeltowardsthepeoplewhohaveaccompaniedmeinthisjourneyandsupportedmethroughit.

My primary supervisor, Chris Hanretty, has trustedmewith his time and dedicationwayearlierthanIwasofficiallyhisPhDstudent.Heconsistentlypushedmetoimprovement,tobettermentandtosharperanalysis.Ilearntfromhim,hispracticeandhissupervisionwhatethicalandrobustresearchis.Iwillforeverbegratefulforthebetterthinkerhemademein.Mysecondarysupervisor,HusseinKassim,immenselyhelpedmenotonlywiththethesis,butalsotodeveloparesearchagendabeyondthedoctoralwork. Iwillstrivetoreplicatethatdepthoftheanalyticaldiscussionswehadtogetherinallmyfutureendeavours.

I should spend many words thanking the Director and the colleagues of the Centre forCompetitionPolicyattheNorwichBusinessSchool,fortheirconstantsupportsinceIsetfootinNorwich.Iwillbeforeverthankfulfortheopportunitiestopresent,toresearch,andtolearnthey providedmewith. The School of Politics, Philosophy, Language and CommunicationStudies(PPL)hasprovenagreatsettingformyresearch,offeringcountlessopportunitiestowidenmyresearchhorizonsthroughseminars,coursesandinteractionwithotheracademicsatallstagesofthecareer.Finally,thetimespentatGeorgetownUniversityinWashington(DC) helped me deepening my research scope; I am thankful to Abraham Newman forwelcomingmyrequest,totheBMWCentreforGermanandEuropeanStudiesforhostingme,andtotheFulbrightFoundationandtheEuropeanCommissionforfinancingmytimethere.

Importantly, Iwholeheartedly thankmy interviewees: regulators, consultants, academics,policy-makersandallofthosewhohavedevotedtime,attentionandefforttocontributingtothisresearchproject.Withoutthem,thisworkwouldhaveneverbeenpossible.Ihopetheywillfindthisanalysisusefultoreflectontheirexperienceandinformtheirfuturepracticeofcooperation.

IthinkIhavecometothepartintheacknowledgementswhereIshouldmentionthepeoplewhosupportedmethroughthesethreeyearsina–sotospeak–“informalcapacity”.Iamnotafanofbigpublicdisplaysofaffectionandwillnotspendtoomanywordsdescribingfeelingswords cannotexpress. It feels easyandnatural, however, to recall the strenuoussupportofpeoplelikeAnnaRita,Lisa,Veronica,Lorella,Alberto,Natalya,Viviane,Liz,andmany,manymorefriends,impossibletonameall,thathavemademytimeinthefinecityofNorwich truly unforgettable, including themany (non-academic) housemates that I havestressedaboutmywritinggoalsandmy fellowPhDsatUEA.My friendselsewhere in theworldandmyfamilyhaveallbeenincrediblypatient,supportive,andrelievingevenintimeswhen Iwastired,moaning,and insufferable.But if Iwereto justwriteasimple,heartfeltdedicationofthisthesistoexpressthedeepestessenceofmygratitudefortheseyearsandthisachievement,itwouldread

6

“Tomymother,andtoAnna”.

7

Abstract

This thesis investigates the emergence, functioning and evolution of voluntary,

informal networks of regulators. Via a combination of inductive and deductive

reasoning,qualitativeandquantitativemethods,thisresearchshedslightonthusfar

unexplored mechanisms of networked regulatory collaboration. These are: the

conditions leading to spontaneous network emergence and consolidation into an

institutionalstructure;thefactorsdeterminingnetworkmembers’tiestoeachother;

the strategies that network members deploy to ensure network survival; the

conditionsfacilitatingnetworkentrepreneurship;andtheroleofinformalnetworks

intheimplementationofforeignpolicyagendas.

Through six empirical chapters, divided in three parts, this thesis explains why

regulatorsnetwork.Thecoreargumentisthatregulatorsusenetworksaslevers:they

leverage their collective collaboration in order to obtain goals that are both

individually and collectively desirable. The first part shows that they network for

control:regulatorsformnetworkswhenevertheyfaceconcretethreatstothescope

of their authority and the extent of their autonomy. The secondpart shows that

regulatorsnetworkforresources:similarity inthepoliticaleconomyandexpertise

explain the structure of regulators’ relationships together with resources, as

regulatorsrelyontheirpeerstocompensatefortheirscarcestaffnumbers.Thethird

partshowsthatregulatorsnetworkforsurvival,andthattheinstitutionalintegration

ofnetworksfacilitatesregulators’networkentrepreneurship.Further,itshowsthat

international organisations and regulators deem informal networks capable of

fostering policy change; hence, they export regulatory networks to target

jurisdictionsintheexplicitattemptofreplicatingtheirsuccessformula.

Thethesisaccomplishesthisambitiousresearchagendabyfocusingonfourempirical

casesoftransnational/trans-jurisdictionalnetworksofenergyregulators:theCouncil

ofEuropeanEnergyRegulators(CEER),theNationalAssociationofRegulatoryUtility

Commissioners (NARUC) of the USA, and their respective progeny, i.e. the

AssociationofMediterraneanEnergyRegulators(MedReg)andtheEnergyRegional

RegulatoryAssociation(ERRA)ofCentralandEasternEurope.

8

Introduction:rationaleoftheresearchandkeyfindings.

- Whydoyouthinkregulatorstakepartinthesenetworks?- Theword“network”,Ithinkthat’stheanswer.

(interview17)

The motivation for this thesis is to bring new theoretical, empirical and

methodologicalperspectivestothestudyofregulatorynetworks.Itisbasedonsix

empiricalchaptersdivided inthreepartsandaddressingfourdifferent literatures.

The essentialmotivation for such an ambitious undertaking is that thewealth of

contributions on regulatory networks examine them through a limited set of

methodological and theoretical lenses. This research features new findings on

regulatorynetworksofsubstantiveacademicandpolicyinterest.Byinvestigatingthe

emergence,operationandtheevolutionofregulatorynetworks,itoffersadifferent

approachtothestudyofnetworkstothosetakenintheexistingliteratureandopens

newpathsofinquiry.

The key argument of this thesis is that regulators use their networks as levers;

networksafford regulators leverage,atbothdomesticand supranational level, to

pursuegoalsthatarebothindividuallyandcollectivelydesirable.Thefocusofthis

analysisisontheinnerworkingsoftrans-nationalortrans-jurisdictionalregulatory

collaboration:howregulatorsrelatetoandrelyononeanotherintheirrespective

contexts.Specifically,Iinvestigatetheconditionsleadingtospontaneous,voluntary

networkemergence,waybeforeendorsementbypoliticalprincipals (which is the

point of departure of most existing literature). I also investigate the drivers of

network collaboration, i.e. the factors explaining regulators’ specific ties to their

most regular, frequentcollaborators.Finally, Iexplore the leverage thatnetworks

have even outside their jurisdiction (which I define as the sum of the territorial

jurisdictions of network members), whenever they play a role in foreign policy

agendas,andinvestigatewhytheychoosetoplaysucharole.

9

Theresearchisinnovativeinseveralways.Firstandforemost,myapproachtothe

studyofregulatorynetworksstartsfromwithinnetworksthemselvesandfromthe

perceptions,evaluations,andactionsofnetworkmembers.Thisapproachallowsme

toavoidtakingtheexternalenvironmentofnetworksasagiven,andtoarticulate

their relationship with it in terms of resource dependency (Pfeffer and Salancik,

1978), aswell as to explore how regulators navigate this dependency over time.

Secondly, and relatedly, I show that there is a relationship between regulators’

institutional self-interest and their contribution to regulatory policy, which is

mediated by their networks. Regulators leverage their network to manage their

relationshipwiththeirpoliticalinterlocutorssoastostrengthenandpreservetheir

institutionalstatus,whilealsoproactivelyfeedingintoregulatorypolicy.

Thirdly,Irelinquishtheassumption,implicitinexistingliterature,thatregulatorsare

allthesame:theydiffergreatlyintheextentoftheirresources,intheirexpertise,in

termsofthemarkettheyoverseeandinthegeographicallocationofthosemarkets.

Ishowthesefactorstoaffectregulators’choicesregardingwhomtonetworkwith,

i.e. which relationships they maintain on a bilateral basis, beyond the policy

requirements of their institutional setting. In turn, these choices determine the

structureofthenetwork.Thisstructureissusceptibletochange;inthisthesis,Ionly

manage to provide a snapshot of network structure in a given moment. If the

structureofthenetworkmightchange,thedeterminantsofthatstructurearenot

likelyto:expertise,resources,similarityinthepoliticaleconomyemergefromthis

thesisasthekeydriversofregulatoryinformalnetworking.

Fourthly,Ialsodismissanotherassumptionthatisimplicitinmostexistingliterature;

namely, that regulatory networks in different parts of the world have emerged

independently of one another and can or should be studied as independent

observations.Inthisthesis,Ishowthatregulatorynetworksareinterdependent.The

historiesofthenetworksexaminedinthisthesisareentwined,andpertaintoforeign

policy agendas: regulatory networks are “exported” from economically advanced

countriesorregionstoeconomicallylessadvancedjurisdictionswithinthecontextof

foreignpolicyprogrammes,intheattemptofentrenchingorfosteringpolicychange.

10

Iexaminetherolethatregulatorynetworksplayinforeignpolicyprogrammes,once

again,fromtheperspectiveoftheregulatorsinvolved.Inthisregard,Ifindevidence

ofmulti-causality:namely,theinvolvementofregulatorsintotheexportoftheirown

“networkformula”maycorrespondtodifferentcausalmechanisms.Iidentifytwo:

networksurvival,andinstitutionalentrepreneurship.

This thesis does not view networks as simply a metaphorical recourse to

conceptualize informal relationships amongst groups of regulatory professionals.

Thoserelationshipscreatenetworkstructuresthatcanbemeasuredandmodelled.

Intakingthisapproach,thisthesisrediscoversandbuildsonsomeofthekeytenets

ofthepoliticalscienceliteratureonpolicynetworks(e.g.powerdependence,policy

influence, advocacy) but infuses them in with a dynamism that has largely been

absent from existing scholarship, where networks have been largely static. By

contrast, this thesis views networks as the results of generative processes, and

structuresthatcanchangeovertimeasregulatorscreatenewandseveroldlinks,

redefinethefunctionsofthenetworkandexpandthemissionbeyondtherealmof

theirterritorialjurisdiction.

The thesis draws on, and contributes to, several literature strands; from public

administration and public policy to political economy, international relations and

internationalpoliticaleconomy.Bytalkingtothesedifferentliteratures,theresearch

highlightsthebreadthofscopethatthestudyofregulatorynetworkscanandshould

have, and thereby reaches beyond well-trodden, low-risk paths of research. The

thesisbringsintothestudyofregulatorynetworksanemphasisonemergence,which

isrevelatoryofactorsandinterests,incentivesandideasdrivingtheestablishment

ofnetworks.Itadoptsanevolutionaryperspective:networkschangeovertimeasthe

externalenvironment inwhichtheyareembeddedchanges; regulators learn,and

modifytheirnetworktiesaccordingly.Ittakesserioustheagencyofregulatorsinsider

networks,whichisinvariablyoverlookedinexistingliterature.Italsoaimstobreak

newgroundbyaddressingahithertounexploredconnectionbetweennetworksand

foreignpolicy.

11

The analysis is pursued through an examination of a single sector and four case

studies.Thefirst twocasesarethenetworkofenergyregulatorsoftheEuropean

Union,theCouncilofEuropeanEnergyRegulators,(CEER)whichiscomparedwith

theNationalAssociationofRegulatoryUtilityCommissioners(NARUC)intheUnited

StatesofAmerica(USA).Thesecondtwocasestudiesinvestigatenetworks,which

aretheoffspringofthefirsttwo.TheyaretheAssociationofMediterraneanEnergy

Regulators (MEDREG), which brings together regulators from Southern European

MemberStates,Balkancountries,MiddleEastandNorthAfricancountries,andthe

EnergyRegionalRegulatoryAssociation(ERRA)ofenergyregulatorsofCentraland

EasternEurope.Thecomparativeapproachmakesitpossibletoanalyseandexplain

varianceinnetworksofthesamesectorbutacrossdifferenttimeperiods,macro-

regional,andinstitutionalcontexts.

Thereasonto focusonnetworksof regulatorsofelectricityandgas– forbrevity,

‘energyregulators’–isthatthesenetworkshavebeenlessofteninvestigatedthan

networks of other infrastructure sectors, notably telecommunications. Yet, the

energysectorisoneofthemostcontroversialandpoliticizedofallsectorssubjectto

economic regulation. In addition, the characteristics of energy infrastructure are

interesting to juxtapose to the study of networks of institutions charged with

regulatingit.Energygridscancrossnationalborders,butfallshortofglobalreach.

They link the local, the national and the transnational dimensions of regulatory

governance. Moreover, the energy grid is nothing but a carefully structured,

constantly balanced network. Hence, in this thesis the geographical and

technologicalaspectsofenergyinfrastructurefeatureprominently.

The thesis is divided into three parts. The first part speaks to the public

administrationliteraturethatfocusesonregulatorynetworksandtheirrationale.It

stems from the acknowledgement that regulatory networks often emerge

spontaneously (Kenis and Provan 2008) and are endorsed by political institutions

(e.g. the European Commission, see Eberlein and Newman 2008 or Coen and

Thatcher 2008) only later. It, therefore, asks: what are the conditions for

spontaneousregulatorynetworkemergence?Inrespondingtothisquestion,thetwo

12

chapters in this part rely on inductive reasoning and on a comparative research

design.IcarryoutacomparativehistoricalanalysisoftheemergenceoftheCouncil

ofEuropeanEnergyRegulators (CEER)and theNationalAssociationofRegulatory

UtilityCommissioners(NARUC)onthebasisofinterviewmaterial(42interviewswith

regulators from both networks), documentary and archival analysis (only in the

NARUCcase). I find thatnetworkemergencecanbeusefullydistinguished froma

slightly later phase, which I name “network consolidation”. Network emergence

resultsfromtheemergenceofregulatoryauthoritieswhich,inbothcasesstudied,

resultedfromtheemergenceoftheRegulatoryState.Newtotheirprofessionand

taskedwithreformingthemodeofgovernanceofsectorsofcrucialeconomicand

political relevance, such as infrastructure sectors, regulators are alone in their

territorialcontexts,withnoprecedenttorefertoandnopeers.

Hence, theybegin interactingwithpeers in their jurisdictionalcontext inorder to

compareandcontrastexperiencesandlearningfromeachother.Thislearningeffort

is not just motivated by professional aims: regulators need to establish their

legitimacy tomakedecisionsof great distributional consequences andbuild their

reputation.However, this remains a loose set of informal relationships driven by

informational asymmetries and expertise. The spark for network consolidation

occurswhenever regulators face the real threat of either losing their powers, or

seeingtherealmoftheirauthorityseverelydiminished,orboth.BoththeUSAand

theEUaresystemsofgovernancearticulatedacrosslevels.AstheRegulatoryState

tookshape,ineithercontext,regulatorycompetencewasdistributedacrosslevels

andhorizontally across different institutions. Tensions for control andprimacy of

regulatory power ensued. As it became clear that these tensions would become

permanenttraitsoftheregulatorygovernancesystem,networksconsolidatedinto

semi-formalorganizations,withaname,headquartersclosetothecentreofpower,

setbudgetarycontributions,workplans,etc.

Thechaptersidentifyafurtherkeychallengeforregulatorynetworks:inordertobe

viable,theyhavetobepolicyrelevant.Regulatorsareconstantlyestablishingandre-

establishingtheir legitimacyatbothdomesticandsupranational (federal) level: in

13

order to credibly claim their individual legitimacy, regulators have to prove their

collectiveonebycontributingtothepolicyprocess.Inturn,feedingintoregulatory

policyallowsregulatorstoaccessandinfluencetheearlystagesofpolicyformulation,

thusaffordingthemanenviablepositionforfurtheringpolicygoals.InboththeUS

and the EU, the interdependence across levels of governance creates a resource

dependency between regulators and their supranational peers (the European

CommissionintheEuropeancase;federalregulatoryagenciesintheUScase):the

latterrelyontheformerforinformationonnational(state)marketsandcompanies;

theformerrelyonthelatterforpoliticalbacking.Althoughtheserelationshipshave

changedovertime(particularlyradicallyintheUScase),interdependencybetween

levelsofregulatorygovernancehasprovencrucialfornetworkconsolidation.

Overtime,however,therationaleunderpinningthisinterdependencymaydeclinein

importance.Tworeasonsforthisdeclineemergefromthechapters:intheUScase,

thegraduallyincreasinglegislativeempowermentoffederalagenciesbeganeroding

thestatelevelregulators’sphereofcontroloverutilitiesintheirstate,renderingthe

US regulatory federalism increasingly confrontational; in the EU case, the

formalization of previously informal relationships, e.g. the establishment of a

European Agency for the Coordination of the Energy Regulators, rendered the

European Commission’s reliance on the CEER for national market information

redundant.Thefinalfindingofthisfirstpartofthethesisisthat,inordertosurvive,

networksresorttotwostrategies:layeringand/orconversion.Chapter3showsthat

theCEERisconvertingitsfocustotopicsitpreviouslydidnotconcernitselfwith,e.g.

consumerpolicy,andacquiringnewfunctions,e.g.trainingprovision.Chapter2hints

atthelayeringstrategydeployedbyNARUC,bygainingtechnicalassistancefunctions

inaidprogrammes,whichisfurtherdetailedinchapter6.

Thesecondpartofthethesisreliesondeductivereasoningandfocusesonexplaining

network structure by uncovering the drivers of regulators’ informal ties to each

other. It comprises chapters 4 and 5, and bridges between network theory and

analysisandthepoliticaleconomy,aswellasthepublicadministration, literature.

ThispartreliesonoriginalnetworkdatathatIgatheredfromthefullpopulationof

14

EuropeanNationalEnergyRegulatoryAuthorities,comprising29elements.Iasked

respondentstonametheregulatoryauthoritieswithwhichtheirauthorityismost

regularlyincontactwithasconcernsexchangesofinformation,opinionsandadvice.

I obtained replies from 28 regulatory authorities. For the missing one, I simply

considered the ties they received as reciprocated. Respondents are all informed

individuals at their national regulatory authority, because they are in charge of

managingand/orsupervisingtheexternalaffairs.Irelyonthisdatainbothchapters

inthispart.

Inchapter4, I formulateseveralhypothesesasconcernsthedriversof regulatory

networking.Thecorehypothesisofthechapteristhatregulatorsprefertonetwork

with peers facing the same sort of sector structure and political economy. I

operationalizenationalpoliticaleconomiesbyrelyingontheVarietiesofCapitalism

typology. The results of the Exponential Random Graph Model I develop in the

chapter lendsupport to thehypothesis.Moreover, theanalysisdemonstrates the

importanceofexpertisebyshowingthatcertainregulators(thoseoverseeingmore

advancedenergymarkets)arerelativelymoresoughtafter,aswellasmoreactive,

than others. Finally, the results suggest that regulators with medium levels of

resourcesaremoreactivenetworkersthanregulatorswithhighlevelsofresources.

The pattern of electricity interconnections across Member States also bears

explanatorypowerfortheobservednetworkstructure,confirmingtherelevanceof

sectoralcharacteristicsinmotivatingregulators’ties.However,itsimportancewas

lesscompellingthanthepoliticaleconomyfactor.

Theanalysis reveals theexistenceofadividebetweenregulators fromEU-15and

those fromEU-13, particularly the newestMember States. This findingmay be a

functionoftime:newerMemberStateshavebeenmembersoftheCEERnetwork

(andoftheEuropeanUnion)forlesstimeandmay,therefore,havehadlesstimeto

buildstablerelationshipswithpeersfromtheEU-15.Atthesametime,however,this

finding suggests that spontaneous regulatory collaboration may lead to the

formation of cliques (i.e. small groups of close-knit communities) and that the

establishmentofaEuropeanAgencymayhavebeennecessarytoensurethatthe

15

flowofinformationacrossallEuropeanregulators,andcoordinationbetweenthem,

actuallyoccurred.

Inchapter5,Iexplorethelinkbetweenregulatorynetworkingandresourcesinmore

detailanddevelopthehypothesisthatregulatorsusenetworkstocompensatefor

lackingresources.Individualregulatoryauthoritiesareverydifferentintermsoftheir

resources.Eventhoughtheirhumanandfinancialresourcespositivelyandstrongly

correlatewiththeextentof theirpopulation (aproxy, if imprecise, for thesizeof

their market), the institutional economics literature argues that each regulatory

authorityneedstofulfilarangeofexperttasks,whateverthesizeoftheirmarket

(Glachant, Khalfallah et al. 2013). Hence, resource constrained regulators may

strugglemorethanwell-resourcedpeerstofulfilthoseexperttasks.Ipositthat,faced

withthoseconstraints,regulatorsresorttotheirnetworkofpeersinordertoaccess

theexpertinformationtheycannotafford.Theresultoftheanalysislendsupportto

thehypothesis. This suggests that, at least in theEuropean context, even though

regulatorycoordinationwithinnetworkshadscarceimpactonconvergence,itmay

havehadsubstantial impact intermsof improvingnationalregulatorygovernance

beyondwhatnationalresourceswouldhaveallowedfor.

Thethirdandfinalpartoftheliteraturegoesbacktocomparativehistoricalanalysis

and inductive reasoning and asks: what explains US and European regulators

involvementintoforeignpolicy/externalgovernanceprogrammessuchasthosethat

ledtotheestablishmentoftheERRAandtheMedReg?Inordertorespondtothis

broadquestion,inchapter6IretracethehistoryoftheestablishmentofERRA;in

chapter7,IretracethehistoryoftheestablishmentofMedReg.Thispartofthethesis

relieson interviewmaterialaswellasdocumentaryanalysis.This finalpartofthe

thesis is very dense empirically and contributes to several strands of research.

Besidescontributingtotheliteratureontransnationalregulatorynetworksastackled

in the international relations and international political economy fields, it also

addresses the literature on the politics of aid and technical assistance through

network collaboration. Indeed, the two chapters in this part demonstrate that

16

informalnetworksarethedeliberatelychoseninstrumentsofforeignpolicyandthat

fosteringinformalnetworkscanbepartofasecuritystrategy.

Theempirical chapters are followedby a short concluding chapter,where I draw

together the findings of the research, its contributions and its limitations. I also

outline two issues that represent promising avenues for future investigation of

regulatory cooperation and coordination in the energy sector: electricity

decentralizationandtheintegrationofdigitalandenergytechnologies.

17

1. Preparingthegroundforanalysis:fromtheliteraturetothestructureofthethesis.

Theliteratureontransnational/trans-governmentalregulatorynetworkshasliterally

boomedsincetheearly2000s.Thisliterature,steepedintheinternationalrelations

tradition, primarily focuses on regulatory networks of economic sectors of global

extent, suchas finance, securities,andbanking (Slaughter1997,Slaughter2000a,

Raustiala2002,Slaughter2004,Verdier2009,Zaring2009,Ahdieh2010,Bach2010,

BachandNewman2010a,BachandNewman2010b,Cao2012,BachandNewman

2014,FarrellandNewman2014,NewmanandPosner2016,CaoandWard2017,

Henriksen and Ponte 2018). This literature understands and analyses networks

throughthelensofmilestonecontributionsonneoliberalinstitutionalism(Keohane

and Nye 1974, Keohane 1982, Keohane 1988, Keohane 1998) and international

regimes(Haas1975,Krasner1981,Krasner1982)asalternativestorealismtobring

orderintotheanarchyoftransnationalrelations.

In the same period, a burgeoning literature focused specifically on networks of

Europeanregulatorsalsoemerged,andhaskeptdevelopingtothisday(Eberleinand

Grande2005, Tarrant andKelemen2007,CoenandThatcher2008, Thatcher and

Coen2008,LavrijssenandHancher2009,Levi-Faur2011,MaggettiandGilardi2011,

Yesilkagit 2011, Van Boetzelaer and Princen 2012,Maggetti 2013, Danielsen and

Yesilkagit2014,Egeberg,Trondaletal.2014,MaggettiandGilardi2014,Blauberger

and Rittberger 2015, Vestlund 2015, Blauberger and Rittberger 2015b, Mathieu

2016). For the most part, the contributions in this literature stem from an

acknowledgementoftheuniquelypeculiareconomic-politicalconstructiontheEUis,

andthereforetendtounderstandandanalysenetworksthroughthatlens,variously

relying on the insights provided by the public policy and public administration

literature,ascondensedinthemorerecentmulti-levelgovernanceliterature(Marks,

Hoogheetal.1996,Scharpf1997,BenzandEberlein1999,Falkner2000,Hoogheand

Marks2001,HoogheandMarks2003,Papadopoulos2005,Piattoni2010),which,in

itself,hasalwayshadanalmostexclusivefocusontheEuropeanUnion(Stephenson

18

2013).TheEuropeanfocustranscendedtherelevanceofsectors:thisliteraturehas

examined networks of European regulators from a variety of sectors, from the

environmenttocompetitionpolicytobanking.Since,however,infrastructuresectors

have always been at the core of the integrationist aspirations of the EU, many

contributionsinthisliteraturefocusonsuchsectors,primarilytelecommunications

andenergy.

Thesetwostrandsofliteraturedominatethefieldofinquiryonregulatorynetworks.

Indeed,theirtheoreticalbasis,methodsandfindingspermeatealsothemuchsmaller

set of contributions that devoted attention to regulatory networks in emerging

markets(BergandHorrall2008,Horrall2008,Jordana2011,LaBelle2012,Dowdle,

Gillespieetal.2013,Bianculli, Jordanaetal.2015,Fernández-i-MarínandJordana

2015),alsooftenfocusedoninfrastructuresectors.

In this chapter, I outline the strands of literature that this thesis addresses and

contributesto.Ibeginbynotingthat,somehowsurprisingly,thisrichliteratureon

regulatory networks rarely relies on a body of knowledge, which is extremely

relevant tothestudyofcollaborationbetweenpublicadministrators,wherever in

theworld:thepoliticalscienceliteratureonpolicynetworks.Ithensurveythekey

tenets and findings of existing approaches to the study of regulatory networks.

Subsequently, I identify themain gaps in the literature as concerns approaches,

topicsandmethods. I then justify the focusonenergy regulatorsandexplain the

differentchallengesthattheyencountercomparedtoregulatorsofothereconomic

sectors. Finally, I lay out the research questions and the hypotheses guiding this

research,themethodologicalframework,andtherationaleforcaseselection.

Transnationalnetworksofregulators.

The literature on regulatory networks has an important, if rarely acknowledged,

predecessor in the literatureongrouppolitics (John2012). In turn, the theoryof

groups spawned the literature on policy networks and the many associated

19

definitionalandsubstantivedebates.PrecedingtheliteratureonnetworksistheUS

notionof“sub-governments”,i.e.sub-systemsofpolicymakingcomposedofgroups

withprivilegedaccesstopolicy-makers(RipleyandFranklin1980).Thefamous“iron

triangles” label derives from this literature and described US policy making as

happeningwithin a triangle featuringCongressional committees, bureaucrats and

interest groups (Freeman and Stevens 1987). This literature entered the studyof

BritishpoliticsvialandmarkcontributionssuchasHecloandWildavsky(1974),who

studied the UK budgeting process and Richardson and Jordan (1979), who

distinguished specific “policy styles” determining the structure of interpersonal

relationsinthenetworksofpublicpolicymaking.

To the best ofmy knowledge, the only contribution understanding a network of

regulators(moreprecisely,ofnationalcompetitionauthorities)throughthelensof

thepolicynetworkliteratureisCengiz(2013).Inherbook,Cengizoutlinesthemain

featuresofthethreerecognizableEuropean“schools”ofstudiesonpolicynetworks:

theBritish, theGermanand theDutch school. These three schools differ in their

understandingofthefunctionofpolicynetworks,althoughtheyarerelativelysimilar

intheirevaluationofthedesirableoutcomesofnetworkcollaboration:betterpolicy.

The British school understands policy networks as instruments of interest

intermediation.Keycontributions(MarshandRhodes1992,MarshandSmith2000)

focused on central-local government relations in the UK. The German school

understandsnetworksasinstrumentsofgovernanceandemphasisestheimportance

oftrustasthemainengineofnetworkcollaborationforsuperiorpolicyoutcomes

(KenisandSchneider1991,Schneider1992,Schneider2001).TheDutchschoolof

policynetworkstudiesfocusesprimarilyonnetworkmanagementtechniques,which

it considers essential to ensure productive network collaboration (Kenis and

Schneider 1991, Raab and Kenis 2006, Provan and Kenis 2008, Kenis and Provan

2009)

Themain criticismaddressed topolicynetwork studies concerns their inability to

demonstrate that networks are a determinant of policy decision and their overly

descriptive nature (John 2012). Policy network studies pioneered the usage of

20

networkanalysis in thestudyofpolicy relationshipsandtheircharacter (Brandes,

Kenisetal.1999),although,initially,mostlytodescriptivepurposes.Fiercecriticsof

thepolicynetworkapproach,suchasDowding(1995),(2001),contendedthatthe

policynetworkapproachwasoflimitedtheoreticalrelevanceandbesetbyinefficient

analytical methods. At any rate, this literature was first in problematizing the

complex patterns of informal networking among stakeholders, and seeking to

evaluatetheirrelevancetopolicy-makingandinfluenceonpolicyoutcomes(O'Toole

Jr1997).Theperceivedfailureofthepolicynetworkapproachtoexplainhowpolicy

outcomes are due to networks may explain why contributions on transnational

regulatory networks make scarce, if any reference to this important precedent

literature.Afurtherreasonmaybethat,differentlyfrompolicynetworks,regulatory

networksarehomogenous:theycomprisesimilarentitieswithsimilartasks(Borzel

1997).Therefore,regulatorynetworksaremoreamenabletothoroughanalysisas

theydonotsufferfromtheboundaryproblemthatKassim(1994)identifiedinthe

study of European policy networks: networks comprising all sorts of actors are

virtually boundless, rendering the network concept unable to clearly specify its

object. Sectoral networks of regulators, instead, are not subject to definitional

ambiguities.

Becausethereisusuallyoneregulatoryauthority(coveringoneormultiplesectors)

per country, regulatory networks tend to be transnational – which attracted the

interestofinternationalrelationsscholars.Indeed,veryoften,literaturereviewson

transnational regulatory networks begin by mentioning Anne-Marie Slaughter’s

(2004c) influential contribution, which acknowledged the emergence of “a new

world order”. This contribution and the literature it spawned emerged at least a

decade after the debate on policy networks in political science, and portrayed

regulatorynetworksasembodyinganimportantconceptualshiftintheinternational

system: from a system of unitary states negotiating within the framework of

supranational institutions to a system of interactions among various (legislative,

regulatory, judicial) components of the state, which interact across borders in

disaggregated form. Trans-governmental networks, Slaughter (2004) contends,

expandregulatoryreach,buildtrustamongtheirparticipants,exchangeinformation

21

anddevelopbestpractices, besidesoffering technical assistanceandprofessional

socialization to theirmembers.These findingsarecommontopreviousstudiesof

policynetworks.Inherview,networksaretheresponsetotheglobalizationparadox

of needing global government but fearing the centralization of authority in

supranationalstructures:throughthepowerofsuasionandtheleverofreputation,

regulatorynetworkscancreateglobalrulesinabsenceofcentralizedstructuresof

power,whileremainingaccountabletotheirnationalprincipals1.

AfterSlaughter’s(2004c)contribution,scholarlyinterestintransnationalnetworksof

governmentalofficialsblossomed.Inparticular,thefocushasbeenontheextentto

which they are able, and should therefore be relied upon, to improve global

governance(Raustiala1997,Slaughter2002,Slaughter2004b,SlaughterandZaring

2006).Hence,theinternationalrelationsliteraturealsoinheritedthepolicynetwork

literatureconcernwithestablishingclearlinksbetweennetworksandpolicychange,

orpolicyoutcomes.Thisquest,accordingtosome,hasprovenjustaselusive(Verdier

2009).Thedebateontransnationalregulatorynetworksisnowlongstandingandhas

come to include a vast range of issues, such as the accountability of networks

(Slaughter2000a,Slaughter2004a,Maggetti2010)andthehomogenizingpowerof

regulatorytransferofbestpracticesthroughouttheworld(Raustiala2002).Atany

rate, this research agenda set the foundations of continued academic and policy

interestinunderstandingtransnationalnetworksofregulatoryofficials.

TheEUpioneered“regulationbynetworks”asamodeofgovernancewhenfaced

with the conundrum of trying to achieve the extent of regulatory harmonization

necessary to bring about the Single Market while avoiding the delegation of

administrative and regulatory powers to European institutions (Sutherland 1992,

Hancher1996,Dehousse1997).TheEuropeanCommunity’sdecisiontoencourage

the coordination of regulatory practice between the representatives of national

1Inhermostrecentbook(Slaughter,A.M.,2017,“Thechessboardandtheweb:Strategies

ofConnections inaNetworkedWorld”,YaleUniversityPress) theauthorgoes furtherby

advocatingforanetworkapproachinAmericanforeignpolicy,muchinlinewiththenotion

oftransformationaldevelopmentdiscussedlaterinchapter6.

22

administrationswasmeanttoachievethenecessarydegreeofregulatoryuniformity.

Several contributions, notably Nicolaides (2004), Mintrom and Vergari (1998),

Eberlein and Grande (2005), Eberlein and Newman (2008), Van Boetzelaer and

Princen(2012),convergeonidentifyingasthemainrationaleofEuropeanRegulatory

Networksthetaskoffillingthegovernancegapbetweenthenationalandthesupra-

national/European level, thus engendering (the conditions for) harmonization.

RecentrefinementsofthisfunctionalistargumentseetheEuropeanCommissionas

anorchestratorof regulatorycompliancevianetworks (BlaubergerandRittberger

2015).

Inasimilarfashiontoscholarsofpolicynetworks,scholarsofEuropeanregulatory

networkshaveassessedtheirabilitytoaffectpolicyand,specifically,tofulfil their

stated aim of engendering regulatory convergence across the Member States

(Maggetti 2009,Maggetti and Gilardi 2011,Maggetti 2014,Maggetti and Gilardi

2014).Thefindingsofthesecontributionsshowamixedrecordofeffectiveness:for

instance,whereasthenetworkofEuropeanregulatorsofsecuritiesappearstohave

been able to design and implement soft rules across European Member States

(MaggettiandGilardi2011), thenetworkofEuropeanenergyregulatorsdoesnot

(Maggetti2014).Ishouldnotethat,inthistypeofcontributions,verylittleattention

isdevotedtotheexplanatorypowerofthesectoritselfinmakingsenseoftheease

withwhichregulatorsare,orarenotabletofosterregulatoryconvergence.

Otherauthorsarguethatexpectingpolicyimpactfrominformalregulatorynetworks

isnaïve:networksarecreatedinsteadofformalcooperationframeworkswhenever

the distributional implications of regulatory harmonization would be politically

damaging(KelemenandTarrant2011,TarrantandKelemen2017).Accordingtothis

view,networksarecreatedwiththedeclaredintentnottoachieveconvergence.It

has alsobeen suggested that the interest of regulators innetworkingmay reside

moreinbureaucraticpoliticsthanincontributingtogovernance(Bach,DeFrancesco

etal.2016).Indeed,recentliteratureonregulatorynetworksinaEuropeansetting

hasrelinquishedregulatoryconvergenceas itssolefocusofanalysisandgradually

come to encompass the point of view of network members, i.e. regulators

23

themselves.Recentfindingsarethatcooperationwithinnetworkshastheeffectof

increasing regulators’ powers (Maggetti 2013), even though not their budgets,

expanding theirautonomy (Yesilkagit2011,DanielsenandYesilkagit2014),create

newcommonresources(Vestlund2015).Moreover,recentcontributionshavetaken

an historical perspective on European regulatory networks, given that, although

apparentlyunabletobringaboutconvergence,theyhavebeenoperatingforover

two decades. These contributions have illustrated that, through their networks,

national regulatorswereable toestablish thedialogic relationshipwithEuropean

institutions and that they played an active role in consolidating the European

regulatoryframework,inthecasesexamined,intelecommunications(Mathieu2016,

BoegerandCorkin2017).

TheliteratureontransnationalnetworksoftendiscussestheleadingroleoftheUSA

andtheEUontheglobalgovernancesceneforstandardsandregulationsinavariety

ofsectors.However,theconspicuous literatureonpolicydiffusionacrosstheUSA

rarely discusses its networked aspects (even though Mintrom and Vergari 1998

representanimportantexception).Thesamecouldbesaidoftherichliteratureon

theUSAregulatoryfederalism(GormleyJr1983,PierceJr1984,Hedge,Scicchitano

etal.1991,LawtonandBurns1996,GerberandTeske2000,ChoandWright2004,

McGuire2006,DavisandHoffer2012,RadaelliandMeuwese2012),whichfocuses

overwhelmingly on the character of the relationships between and across

bureaucraticaswellasgovernmentallevels.However,theserelationshipsareusually

phrasedintermsoffederalismandintergovernmentalism.Childs(2001)andBeecher

(2012) are important exceptions. Their insights on the network relationships

betweenUSstatelevelenergyregulatorswillinformchapter2.

TheUSAandtheEUsystemsofgovernancesharetwokeyattributes,whichrender

themsuitableforcomparison(asIexplaininmoredetaillater):botharepremised

on the choice of independent regulation as a means to achieve superior policy

outcomes;andbotharearticulatedacross levelsofgovernance.The literatureon

multi-levelgovernance,however,hasexclusively focusedon theEuropeanUnion.

Theconceptofmulti-levelgovernance,formulatedinaseriesofworksbyMarksand

24

Hooghe(Marks,Hoogheetal.1996,Marks,Scharpfetal.1996,HoogheandMarks

2001,HoogheandMarks2003),representsthescholarlyefforttointegratedifferent

territoriallevelsinaunifiedframeworkofanalysisinordertoconceptualize,explain

and understand the workings of EU governance (Eising 2008). The initial

conceptualizationoftheEUasaMLGsystemderivedfromanalysesofthepoliticsof

thestructuralfundsandtheroleofregionsintheEU(Marks,Hoogheetal.1996).

Overtime,theterritorialfocusoftheMLGliteraturehasbroadenedtoincludeamore

metaphoricalconnotation,portrayingthecomplexityof interactionsamongactors

placedatdifferentpoliticalandsectorallevels(Stephenson2013)aswellasoccurring

atdifferenttimes(Goetz2010).

ThekinshipbetweenMLGandpolicynetworksisamplyrecognizedintheliterature

(EisingandKohler-Koch2003,Kohler-KochandRittberger2006)anddrawnuponto

conveytheinterdependenceofactorsacrosslevelsandtheircollaborativeattempts

to influenceEuropeanpolicy-making.Thepopularnotionof theEUasa“network

governance” polity defines the formation of coalitions of interest around certain

“issueareas”(Heclo1978),encompassingdifferenttypesofactorsplacedatdifferent

governance levels (Piattoni 2010). Networks of national regulators, however,

exemplifyatypologyofnetworkgovernancethatisnotadequatelycapturedbythe

network governance metaphor (Piattoni 2010). First and foremost, networks of

nationalregulatorsgatheronetypeofactor(regulators)placedatoneterritoriallevel

(national); they are, therefore, homogenous networks. Secondly, networks of

nationalregulatorsdefinethemselvesalongsectoral,not“issuearea”,lines.Thirdly,

thesenetworksfeaturerepresentativesfromallMemberStates,ratherthanvariable

constellationsofactors.

These differences might explain why the MLG literature and the literature on

regulatorynetworkshaverarelymet.Yet,themulti-levelgovernanceliteratureand

the inter-governmental and cooperative federalism literature in the USA share

importantcommonalitiesintermsofthegoalstheypursueandthequestionsthey

ask:whataretheconditionsforcollaborationacrosslevelsofgovernancetodeliver

desirable outcomes? This question has often been asked in the literature on

25

regulatorynetworks. I contendthat,besides focusingonoutcomes, this literature

shouldre-discovertheimportanceofthecollaborationprocessandhowthatfitsinto

thebroaderpolicyprocess,whatbenefitsitentailsfornetworkmembers,andhow

theyconcretelyusetheirrelationshipstoeachother.

Regulatory networks are structures of interaction generated by the agency of

individualregulators,giventhestructuralconstraintstheyfaceatbothdomesticand

supranational(federal)level.Theyareheldtogetherbyinterdependenciesandthe

realization,vianetworking,ofbothindividualandcollectivebenefits(deBruijnand

ten Heuvelhof 1995). Studying what these are leads enquiry from description to

causalexplanationandinference.Importantly,manyoftheassumptionsimplicitin

existingliteratureneedtobeabandonedinordertobreaknewgroundinthestudy

ofnetworks.Theassumptionsdebunkedinthisthesisarethoseof:stationarity(in

the literature, networks never change in either shape, mission or rationale);

homogeneity(althoughregulatoryauthoritiesarethesametypeofinstitutions,they

differgreatlyfromjurisdictiontojurisdiction);passivity(networkmembersarenot

merely implementing the directives of their political principals); and isolation

(networks are often discussed as if they were isolated from their external

institutionalenvironment).Thereneedstobespaceintheliteraturefordiscussing

the conditions leading to spontaneousnetwork collaboration; howandwhen the

rationale of network collaborationmay change; and how the differences among

networkmembers in terms of resources, expertise, independence etcmay affect

patternsofcollaboration.

The gaps in the literature: the emergence and the evolution of

transnationalregulatorynetworks.

Firstandforemost,Icontendthatenlargingthetimespanofanalysisofregulatory

networksisparamountinordertoseriouslyinvestigatetheirimpactonpolicyandon

the structure of policy-making. Existing literature is overwhelming ahistorical and

focusedonthetimehorizonthatPaulPiersonrecognizedasincreasinglyprevalentin

26

thesocialsciences:thetimehorizonofatornado,whichdevelopsrapidlyandonly

lastsforashortperiod;“Inchoosingwhatweseektoexplainandinsearchingfor

explanations,wefocusontheimmediate;welookforcausesandoutcomesthatare

both temporally contiguous and rapidly unfolding. In the process,wemiss a lot.”

(Pierson 2003, p. 178). Many contributions focusing on European Regulatory

Networks(ERNs)areclearexamplesoftheseattitude:ontheonehand,contributors

disregardthevoluntarynetworksofregulatorsthatprecededtheestablishmentof

ERNs and emerged as a result of the bottom up initiative of national regulatory

authorities(seeforinstanceMaggetti2013aand2013b);ontheotherhand,virtually

no contributors have seriously attempted to explain the persistence of these

voluntarynetworksoncethecorrespondingEuropeanRegulatoryNetworkoreven

the corresponding European Regulatory Agency had been created (although see

Thatcher2011).

ThecontributionbyKaiser (2009) representsanotableexception instudiesofEU

governanceinthatitexplicitlycallsformorehistoricalresearchintotheemergence,

theevolutionandtheimpactofnetworksofpoliticians,bureaucracies,industryand

otherstakeholdersinEuropeangovernance.Inhiscomprehensivereview,however,

Kaiser(2009)doesnotmentionnetworksofregulators.Althoughrecent literature

has engaged in investigation of regulatory networks in a historical perspective

(Boeger and Corkin 2012, Mathieu 2016), it has focused on networks of

telecommunicationregulators,ontheemergencephaseonlyandontherelationship

withtheEuropeanCommissionratherthanofregulatorswitheachotherwithinthe

network.

Second, I consider that the predominant functional approach to explaining the

existence of regulatory networks is overly constraining in analytical as well as

conceptual terms. Moreover, it attributes all the agency of establishing and

maintaining networks to the single goal of achieving regulatory convergence.

Recently,theliteraturehasbegunshiftingitsattentiontothestudyofnetworksfrom

the perspective of regulators, emphasising their rational motives (see Bach and

Ruffing(2013),Maggetti(2013),Bach,Ruffingetal.(2014),DanielsenandYesilkagit

27

(2014)).However,thesestudieseitherstudynetworksfromtheoutside,orrelyona

very limited number of cases in order to gauge regulators’ perceptions of their

networks. Although studies emphasising the socialization aspect of regulatory

networks exist (Bianculli 2013, Danielsen and Yesilkagit 2014) and represent an

additionalanalyticalnuance,theyrarelyventurebeyonddescription.Inthisthesis,

socializationisunderstoodasamechanismpromotingnetworklongevity.

Third,theliteraturehasscarcelyreliedonthepowerofcomparativeanalysisinits

studyofnetworks.Manycontributionsconsidermorethanonenetwork;however,

they do not carry out comparative analysis, but rather consider similarities and

differencesbetweennetworksinhowagivenindependentvariableaffectsthemor

inhowthesenetworksaffectagivendependentvariable.Othercontributionseither

discussnetworksinverygeneralterms,orfocusonasinglecase.Inthelattercase,

withincaseanalysisisrare.Contrarilytoexistingresearch,thisanalysisexploitsthe

leverageofferedbysystematiccasecomparisoninordertoperformareasonedquest

for discovering the rationale of transnational regulatory cooperation. Specifically,

this research looks within networks to find these responses, rather than outside

networks,andinvestigatestheirthrustthroughtheperceptions,thememoriesand

theassessmentsofregulators.

This point leads to the main contribution that this thesis makes to existing

scholarship:investigatingthereasons,theperceptions,andtheexpectationsofthe

agentsthatareembeddedintransnationalregulatorycooperation–theregulators.

Toooftenthemotivesofregulatorsforcollaborationwithinnetworksareassumed

away,orsimplyneglected.Thisattitudeiscommontobothenthusiastsandsceptics

of regulatory networks and is deeply problematic because, by overlooking the

motivations of the agents of cooperation, it neglects its very essence. Relational

patternswithinnetworksaswellasbetweennetworksandpoliticalreferents(inthe

case of networks of regulators in a macro-regional or federal setting, like the

European Union or the United States, these exist at both domestic and

supranational/federal level) represent a wealth of opportunity structures for

regulatorstofurthertheirpreferences,whichcurrentresearchignores.

28

The fifth limitation of existing scholarship addressed in this thesis concerns the

methodsofinvestigation.Atfirstsight,itissurprisingtonoticethatthevastmajority

oftheliteratureonregulatorynetworksdoesnotrelyonnetworkanalysisinorder

to investigatenetworks. Intheir insightful reviewof thetreatmentofnetworks in

publicadministrationscholarship,Isett,Mergeletal.(2011)highlightthreemainuses

of the term “network” in that body of research: metaphorical, utilitarian, and

methodological.Earlierliteratureusestheword“network”inautilitarianway,asa

tooltounderstandcoordinatedpublicserviceprovision;morerecentliteratureuses

thetermmetaphorically,asanorganizingconceptofthefeaturesofacertainsocial

context2.

The utilitarian and metaphorical approaches predominate over methodological

contributions on network structures. A close reading of the literature, however,

reveals the reasons for this shortcoming:on theonehand,gatheringdataon the

actualrelationsbetweentheregulatorsthataremembersofanetworkisadaunting

task,asregulatorsarereluctanttoprovidesuchinformation;ontheotherhand,most

analysisofnetworksareactuallynotfocusedonnetworksbutoneitherthepolitical

decisionsdrivingtheirestablishmentbypoliticalprincipalsortheireffectsonpolicy

convergence or harmonization. Yet, quantitative network analysis offers a rich

reservoirofpossibilitiesfortheresearcherofnetworks,allowingforthediscoveryof

patterns and drivers of transnational cooperation, a better knowledge of which

provesinestimableforthisfieldofresearchaswellasfortransnationalgovernance.

2Inthisregard,theseauthorscorrectlyhighlight“theimportantandcriticalissueofwhether

the actors in an attributed network (meaning a group where the network paradigm is

applied)mustacknowledgeandacceptthattheyoperateinanetworkforittoactuallybea

network.”(Isettetal2011,p.i160).Thisnecessityhasrepresentedafurtherstrongreason

to rely on interviews within the framework of this research: transnational networks of

regulators usually do not call themselves networks, but “associations”. Therefore, it is

importanttomentionthatnoneofmyintervieweeshasobjectedtousingthetermorhas

evenenquiredaboutwhether itwastherighttermtouseorhasuseddifferentwordsto

indicatethesubjectsofthisresearch.

29

Whythefocusonenergyregulators?

The1980sand1990shaveseentherapidspreadofmarketreformsininfrastructure

sectors,inparticulartelecommunicationsandelectricity,acrosstheworld.Thecore

ofthesereformshasconsistedintheintroductionofprivatecapitalininfrastructure

sectors, followed by liberalization reforms aimed at promoting competition and

accompanied by the establishment of regulatory institutions. Socio-economic

interdependence (Lazer 2001, Raustiala 2002), membership in international

organizations(RodineHardy2008,Cao2009,Rodine-Hardy2015),electoralcalculus

(MurilloandMartínez-Gallardo2007),coercionbyInternationalFinancialInstitutions

(Henisz,Zelneretal.2005,DubashandMorgan2012),emulationofpolicychoices

perceivedaslegitimate(Meseguer2004),competitionforforeigndirectinvestment

(Elkins,Guzmanet al. 2006) are only someof the causes adducedby scholars to

explainwhyandhow,over thecourseof twodecades,nearlyall countriesof the

world adopted very similar kinds of economic reforms, including in their

infrastructureindustries.

The infrastructure sector privatization reforms undertaken by the Thatcher

governmentinthelate1970shadenormousimpactworldwide.Eventhoughutility

regulationhadbeenpracticedalreadyforoveracenturyintheUSA,ithadalways

had the purpose of replacing, not promoting, competition; the British reforms

introduced the notion of regulation promoting competition and markets. The

accompanying establishment of independent regulatory authorities was also

followedwithgreatinterestbypractitionersandobserversalike.Thecombinationof

privatization, re-regulation and introduction of competition has entered the

economicsbutalsothepublicpolicyliteratureunderthenotionof“Britishmodel”

(Stern2014).

This influence was primarily felt across the European Union. Before European

legislationon the subjectwas evendrafted, severalMember States governments

lookedattheUKasasourceofinspirationfortheliberalizationoftheirmarketsand

30

thecreationofregulatoryauthorities(apointmadeintheliterature–e.g.Kassim

andMenon 1996 – and confirmed by my interviews). Moreover, British officials

enjoyedsubstantialinfluencewithintheEUasconcernedthelegalframework,which

essentiallymandatedthesteadyerosionofstatecontrolovertheutilities(Hancher

1996).Thesubsequentuncertaintyregardingthemodesofsectormanagementafter

reformeasedtheemergenceofregulatorynetworkingintheEU,asshallbeseenin

thethesis.

GilbertandKahn(1996)scrutinizethespecificitiesoftheutilitiesanddescribethe

regulatoryproblemaslinkedtothefollowingcharacteristicsofnaturalmonopolies:

capital intensity and minimum economic scale; non-storability with fluctuating

demand; locational specificity generating location rents; essentiality for the

community; involving direct connection to customers (1996, p. 2). The last two

characteristics, which imply large exploitative power by the producer, render

regulation politically inevitable. The consumer demand for “fair” regulation of

electricitypricesmakesinvestorswarythat,oncetheyhavesunkcapitalinelectricity

infrastructure,theywillbe“unfairly”limitedinthepricestheycancharge.Authors

convergeontheacknowledgementthat,ifprivateinvestmentistobesuccessfully

attracted to and remain into the sector, “what is critical is that there be some

protectionagainstpoliticalintervention”(GilbertandKahn,p.5,seealsoLevyand

Spiller1994).Thisistherationaleforregulationofutilitiesbyindependentregulatory

authorities.

Energyinfrastructurehasimportantscaleandscopenetworkproperties.Simplifying

considerably, increasing the size of the interconnectednetworkof electricity grid

lines and gas pipelines leads to substantial increases in efficiency. The strategic

importanceofenergyforsecurityandeconomicpurposes,however,hasdetermined

thefactthatmostelectricityandgasinfrastructuresystemshavebeenbuiltinorder

to ensure national self-sufficiency (Lagendijk 2008). Nevertheless, engineering

reasonsofnetworkstabilityandreliabilitymotivatedtheconstructionofcrossborder

interconnectionlines.Incasesofshortagesorofovercapacityoneithersideofan

interconnection,nationalsystemoperators(usually,incumbentverticallyintegrated

31

state-ownedcompanies)wouldagreefortheone-offimportorexportofelectricity

inordertokeepbothsystemsstable.

Withtheadventofenergymarkets,eachofthoselinescametorepresentamarket

opportunity.Theallocationofthecostsofenergytradeinitsphysicalrealization(i.e.

thecoststhatthenatureofenergyinfrastructurenetworksimposesonthesystems

ofpartiesthatarenotinvolvedinthemarkettransaction)aswellastheallocationof

thecostsandthebenefitsofnewinterconnectionsaredeeplycontroversialpolitical

andeconomic issues.These issuesrepresentthepolicyrationale for transnational

regulatory cooperation. Integrating energy markets should bring considerable

benefits(Pérez-Arriaga,2014): increaseefficiency, improvesecurityofsupply,and

facilitatetheintegrationofrenewableenergysources.Thechallenges,however,are

great.Theycanbesummedupinthenecessitythatcountriessharereservesanddo

notdiscriminateamongmarketplayersandamongconsumers,givingprioritytotheir

localdemand.Thispresupposeswide-rangingregulatoryharmonisation,and,most

importantly,anacceptanceofthefactthataregionalmarketisaimedatimproving

globalsocialwelfare,whichimpliesitmay,eventhoughonlytemporarily,bringmore

benefittosomethantoothers(Pérez-Arriaga2014).

Theliteratureontransnationalregulatorynetworksandtheliteratureonelectricity

andgasreformshave,however,onlyrarelymet.Researchershavedevotedmostof

theirattentiontonetworkshavingglobalscope,suchasthoseinvolvingregulators

fromthebankingandfinancialservicessectors(Verdier2009,Bach2010,Bachand

Newman2010a,BachandNewman2014).Themostvisible,andamongthemost

thoroughly investigated, such network is the Basel Committee on Banking

Supervision (Zaring1998,Verdier2009,Zaring2009,Zaring2012,Goldbach2015,

Reisenbichler2015).

The diffusion of the regulatory authority as institutional form is, indeed, often

ascribedtothevirtuesofthefigureofthecentralbanker:expertise,clearobjectives,

anddepoliticizedoperation(Cukierman,Webetal.1992,Kapstein1992,McNamara

2002,SternandTrillas2003,SimonePolilloandMauroF.Guillén2005,Gilardi2007).

Overtime,thegrowthintheimportanceandvisibilityofnetworksofcentralbankers

32

or financial and insurance sector regulators eased the taskof investigating them,

giventheappearanceofreliablelongitudinaldataontheirmembership,decisions,

and meetings. Transnational regulatory networks have been understood as the

coordinated regulatory response to the mobility of capital, which would enable

regulatoryforumshoppingandracestothebottomofregulatorystandards.

Academic investigation of networks of regulators of infrastructure sectors (i.e.

electricity,gas,telecommunications,water,railways)islesswidespreadandmostly

focusedonthetelecommunicationssector(Barendse2006,MaitlandandvanGorp

2009, Jordana2011,BoegerandCorkin2012,Mathieu2016), followedbyenergy

(Vasconcelos2009,Beecher2012,Bianculli2013,Maggetti2014),asthesearethe

twosectorswherethemostextensivereformshavetakenplaceintheEUandaround

theworld. A considerable amount of research on EuropeanRegulatoryNetworks

displays the same sectoral focus. However, the specificities of these sectors are

rarely, ifever,acknowledged.Contrary to financial services, infrastructuresectors

areterritoriallyboundnaturalmonopolies.Thismeansthatitisanti-economicalto

buildtwoormoreinfrastructuresystemsinanygiventerritorialunit,henceaffording

a monopoly position to the owner and/or manager of the infrastructure. The

conditionsatwhichthismonopolyisheldaretheobjectofregulation.

Nowadays, the telecommunications sector appears to havenearly lost its natural

monopoly characteristics. Technological innovationhas rendered Information and

CommunicationTechnologies(ICT)marketsmoreandmorecontestable.Moreover,

telecommunicationsrepresentthe infrastructuresectorthat ismostembedded in

international institutional arrangements: the World Trade Organizations (WTO)

agreements cover telecommunications services, and stipulate the obligation for

countriestoestablishaseparateregulatoryagencyforthesector;theInternational

TelecommunicationsUnion(ITU)isthelong-standingUnitedNationsagencyforthe

ICT sector, involving a multiplicity of stakeholders and providing for technical

standards. None of this applies to the electricity and gas sectors. The relevant

infrastructures retain very strong natural monopoly characteristics. Liberalization

and privatization processes have not matched the levels reached in

33

telecommunications.TheWTOagreementsdonotcontemplateenergyservices.No

singleworldagencyororganization, comparable to the ITU,exists for theenergy

sector3.

Most importantly, the main difference between scholarship on financial sector

regulatorynetworksandextantresearchonnetworksofinfrastructureregulatorsis

thatthelatterisveryinformative,ifdescriptive,thuslackingfullappreciationofthe

politicsofthisnetworkingactivity.Inasectoraspoliticallysalientasenergy,thisis

animportantomission.Differentlyfromtheglobalscopeoftransnationalregulatory

networks in banking or securities regulation, transnational networks of energy

regulatorstendtohaveregionalscope.Thisismainlyduetothecharacteristicsof

the infrastructure:electricitygridsandgaspipelineshave limitstotheirextension

beforetheybecomeanti-economical.

Moreover, because the provision of energy service has a very visible, direct and

measurable(throughthebillsandthroughtheoutages)impactonthequalityoflife

of the citizenry, it has the characteristic of being at onceextremely complex and

extremely,immediatelypolitical.Ifthecoreoftheregulatoryprobleminglobalized

financialservicesistamingswiftcapitalmobility,regulatorsofinfrastructuresectors

have to tackle the opposite issue: sunk costs. In otherwords, capital invested in

infrastructure is unrecoverable. Theassets’ long lives and the time span required

beforerevenuescoverthecostoftheinvestmentmakeinvestingin infrastructure

riskyinabsenceofregulatorycredibilityandpolicystability(Spiller1996,Spillerand

Tommasi2005).

The regulatory problem in infrastructure sectors is providing private capital with

sufficientlycrediblecommitmentsandincentivestoinvestintheinfrastructure.The

3TheInternationalEnergyAgency(IEA)findsitsoriginsandisstillmostlyconcernedwiththe

politicsofenergysources(inparticularoilandgasandtheiralternatives),particularlyinsofar

as they involve relations between producer and consumer countries, but not regulatory

aspects.

34

rationale for network coordination for regulators of infrastructure sectors is,

therefore, qualitatively different from that of regulators of financial services. The

latter have to tame capital; the former have to attract it. Hence, the underlying

determinants of regulatory networking are sector specific. Moreover, the

inescapable territorial dimension of infrastructure renders regulatory authorities

constrainednotonly intheir jurisdiction,butalsointheirautonomyfrompolitical

principals. For this reason, researching networks of energy regulators presents

familiaraswellasnewchallengesandcorroboratesknownresultswhileyieldingnew

ones.

Analysing transnational networks of energy regulators: the

questions,theconceptsandthehypothesesguidingthisresearch.

Theliteratureontransnationalregulatorynetworkshasmaintainedarathernarrow

focusinsofarastheoreticalframeworks,approachesandmethodsareconcerned.As

mentioned,analysesofregulatorynetworkstreatthemasiftheywere“frozen”in

time.Theassumptionthatnetworkestablishmentwasdrivenbysomeexternalactor,

i.e.asupranationalinstitutionsuchastheEuropeanCommission,isomnipresentin

theliterature.Inthefewcontributionsrecognizetoregulatorsthepaternityoftheir

networks,theprocessofemergenceisusuallyneglected.Theneedforregulatorsto

network and exchange information is treated as almost self-evident, plausibly

becausethepolicy literaturehasestablishedthatexchangeof informationamong

stakeholders isconducivetopolicy input (KönigandBräuninger1998,Coen2005)

and/orbecausethenewinstitutionaleconomicsapproachrecognizedinasymmetric

information between regulators and companies one of the main obstacles to

effective regulatory performance (Ogus 2002, Guerriero 2010). Yet, network

formation(whichisanissueofemergence)mayrevealimportantinformationonthe

goalsofexchangeandontheincentivesoftheactorsinvolved.

Intheircontributionontheevolutionaryphasesofpolicynetworks,ProvanandKenis

(2008) distinguish between spontaneous, or member-driven regulatory networks

35

(thattheycallnon-brokered)versusnetworkswhoseestablishmentismandatedby

an external agent (that they call brokered). Within the brokered/non-brokered

dichotomy, these authors recognize three network governance modes: self-

governed; lead-organization;andexternally-governed. Inthefirstmode,members

areinthedrivingseat:theysetupthenetworkandexertsharedgovernanceoverit.

Inthesecondmode,oneofthenetworkmembers,usuallytheonewiththegreatest

interest in the network objectives, takes up the role of lead organization. This

membermayprovideadministration for thenetworkandunderwrite the relative

costs,orseekaccesstoexternalfunding.Inthethirdmode,networksaregoverned

“by a unique network administrative organization (NAO) which may be either

voluntarily established by networkmembers ormandated as part of the network

formation process” (p. 234). Provan and Kenis’ framework is parsimonious and

includes an element evolution of networks.However, it is very deterministic and

relinquishesconsiderationofthereasonswhybrokeredornon-brokerednetworks

areestablished.

Giventhisbackground,thequestionguidingthefirstpartofthethesisis:

1. Whatexplainstheemergenceoftransnational/trans-jurisdictionalnetworks

ofenergyregulators?

IrelyontheinsightsoftheliteraturebutalsolettheevidenceIgatheredthrough

interviews,documentaryanalysisandarchivalresearchspeaktothematter.Iexpect

this evidence to point to some of the motivations that have emerged from the

literature on regulatory networks but also on regulatory agencies: for instance, I

expectregulatorynetworkstoemergeduringpolicycrisesorpolicychange,inline

withassessmentsreachedbyBernstein(1955)andDowns(1967).Interdependence

is plausibly going to prove pivotal in the acknowledgement of the desirability of

transnational regulatory coordination (Haas 1975, Keohane 1982, Keohane 1998,

Lazer2001,Gilardi2002,BacciniandDür2012,VanBoetzelaerandPrincen2012,

FarrellandNewman2014,FarrellandNewman2015,Saz-Carranza,SalvadorIborra

etal.2016).Atthesametime,byfocusingontheregulators’motivestoestablish

networks, I extract from the research newmaterial that reveals how regulatory

36

institutions benefit from networking and how networks empower them towards

boththeirdomesticinterlocutorsandsupranationalones.

Justliketoindividualorganizationsdo,networksmay,overtime,grow(insizeand

importance)andlearntoperformbetterandroutinizetheirprocedures,therefore

undergoing a process of consolidation. Transnational regulatory networks usually

displaysomeelementsofformalization,suchaslegalregistration(usuallyasnot-for-

profit entity), a small coordinating secretariat, fixed budgetary contributions and

schedulesofactivities,meetings,roundtablesandthelike.Themomentofpassage

fromcompletelyinformalandspontaneousinteractionbetweenpeerstothekindof

“light”formalizationdescribedaboveisinvariablyoverlookedintheliterature.

The sociological literature on inter-organizational networks provides useful

instruments to understand the developments leading from a phase of network

existencetothenext.Inhisstudyofinter-organizationalrelationships,Powell(1995)

contends that trust, which is necessary for networks to be viable (Uzzi 1997), is

actually a product of interaction rather than a precondition for it (a point

underpinning the arguments made in Slaughter (2017)). Drawing from game

theoreticcontributions,Powellaffirmsthatorganizationsdecidetonetworkwhen

theyareawarethat theprobabilityof futureassociation ishigh. Inhisview, trust

cannotbecalculatedorenforced,asmuchasitisnotembedded:rather,itemerges

asaby-productofnetworking.However,cooperationimpliesvulnerability,froma

sociologicalpointofview,orcosts,fromapublicchoiceperspective;atanyrate,it

requires the establishment of some governance structures to allow for constant

monitoringandconsultation,leadingtoinstitutionalization.

Therefore,thesecondresearchquestionguidingthisresearchis:

2. What explains the consolidation/formalization of transnational/trans-

jurisdictionalnetworksofenergyregulators?

Theanswersthatarepresentintheliteratureincludegrowthinsizeandimportance,

awareness of the prospect of future cooperation, and desire to monitor

37

counterparts’contributionstothenetwork.Theseresponsesare,however,confined

to the dimension of the network. They do not appear to consider whether the

relationshipbetweennetworkmembersandtheirexternalinstitutionalenvironment

mayexplainthetransitionofanetworkfromaloosertoamoreconsolidatedstate.

Yet, this is a plausible hypothesis, given that regulators exist within complex

institutionalsystemsorregulatoryspaces(HancherandMoran1989),populatedby

amyriadotheractors.

One of the main goals of this research is investigating the concrete usage that

individual regulatory authorities make of their network ties. In examining the

purpose of networking, I deviate from the historical approach and analyse the

networktiesofthemembersofasinglenetworkatasinglepointintime.Thischoice

ismainlyduetolackofdataonregulators’tiesatdifferentmomentsintime.Despite

thisshortcoming,theanalysisfitswiththeoverallapproachofthethesisasitrelies

on a generative model that assumes that the currently observed network

configurationistheresultofthenetworkevolutionovertime.

3. Whatarethemaindeterminantsofnetworktiesacrossanetworkofenergy

regulators?

Iderivemymainhypothesisfromonewidelyacknowledgedtendencyofmembers

ofsocialnetworks:individualstendtoassociatewithsimilarothers.Thistendencyis

capturedbytheterm“homophily”.Inthecaseathand,Iexpectsimilaritiesinmarket

andsectorarrangementstoinfluenceregulators’choicesoftheirpreferredandmost

frequent network partners, therefore pushing them to establish direct linkswith

those. In order to assess this hypothesis, I rely on the ‘Varieties of Capitalism’

approach, which has recently been shown to affect the character of regulatory

institutions(GuardiancichandGuidi2016).Iusetheapproachasaheuristicinthe

caseathand;namely, tocategorize theregulatorsonthebasisof thespecificsof

electricitysectorarrangementsthatarefoundintheircountryoforigin.Moreover,

inspiredbytheworkofAlcañiz(2016),whofindsthatsuddenbudgetarycutsprompt

nuclearsectorregulatorsandexpertsinLatinAmericatorelyonnetworkinginorder

topooltheirscarceresources.

38

Iformulatethereforethefollowingcorehypotheses:

• Regulatorsaremorelikelytonetworkwithpeersfromcountriessharing

thesameVarietyofCapitalismasthemselves,allelseequal.

• Regulatorsaremorelikelytonetworkwithpeershavingmoreresources

thanthemselves,allelseequal.

Thefirsttworesearchquestionsconcernedspecifickeymomentsinthehistoryofa

transnationalregulatorynetwork:emergenceandconsolidation.Onceconsolidated,

howdonetworksevolve? Inparticular, Iaminterested inexaminingwhetherand

hownetworksareabletoreinventthemselvesoncetheiroriginal“regulatorypolicy

function”(toparaphraseDowns)hasdiminishedinimportanceasaresultofchanges

intheirinstitutionalenvironment.

Historical researchers havemadewide use of the notion of path dependence to

explainthelongevityofinstitutions.Accordingtothisapproach,institutionsemerge

whenapath-breakingevent (referred to in this literature as a “critical juncture”)

occurs,“movingthepressureonthestatusquotoanew,muchhigherlevel–very

close to the threshold level for major political change” (Pierson 2003, p. 213).

Outcomesatcriticaljuncturesinducepathdependentprocesses,creating“dynamics

of self-reinforcing or positive feedback processes even in the absence of the

recurrence of the original event or process” (2003, p. 214). By implication,

institutions,oncecreated,persistunaltered(inastateof“lock-in”,Thelen2003)until

anotherexogenousshockoccurs.Akeypointinthelogicofpath-dependencyisthat

thefactorsresponsibleforthegenesisofaninstitutionmaybedifferentfromthose

thatsustainitovertime.Moreover,theconfigurationsemergingfrompath-breaking

eventssetconstraintsonsubsequentdevelopments, leaving lessspaceforagency

andcontingencyandmoreforadaptationtoinstitutionalincentivesandconstraints.

Inherdiscussionoftheanalyticaldevicesthatcanbeusedtounderstandinstitutional

evolutionandchange,Thelen(2003)arguesthatpath-dependencyisill-suitedtothis

purpose.Thepath-dependentlogic,sheargues,ismoreaptatexplaininginstitutional

reproduction,i.e.howinstitutionspersist,thanatexplainingevolution.Hence,she

39

introducestheconceptsoflayeringandconversionasanalyticalconstructsthathelp

makingsenseofevolutionandchange.Shealsocallsforintroducingmorestructure

at the“frontend”of institutionalemergenceandmoreagencyat the“backend”

(2003, p. 225). This means, on the one side, considering the role that existing

structures played when path-breaking events led to institutional creation, by

constraining the rangeofoptions available to actors; on theother side, itmeans

investigatingtheagencyofinstitutionalactorswithintheframeworkoftheirexisting

institutionalstructureswhenfacedwithchangesintheirenvironment.Thefeedback

loopsreinforcingpath-dependentprocessesareusuallyconceivedasconstraintson

actors going forward; Thelen (2003) argues that they may also represents

opportunities.Thisapproach

“…helpsunderstandingwhy,overtime,institutionalarrangementsmaycome

to serve functions that are quire remote from those originally intendedby

theirdesigners,howtheycanaffect(ratherthanjustreflect)theprevailing

balanceofpoweramongsocietalgroupsandhowtheycanbecomeresources

for(ratherthanjustconstraintson)actorsengagedincontestsoverthetypes

ofpracticesthatarecodedasappropriateordesirable.”

(Thelen2003,p.220)

Over time, networks may evolve in response to a changed environment that

confrontsthemwithnewproblemsandchallengesthattheyaddressbyusingthe

existingstructuretonewpurposes.Theydosoviaaprocessoflayering,i.e.byeither

partially renegotiating some elements of their mandate, while leaving others in

place; or via conversion, i.e. by redirecting the network to new purposes. These

themes emerge in early American literature on “life cycle”models of both state

regulatory commissions (Bernstein 1955) and federal regulatory bureaus (Downs

1967).

AssummarizedinMitnick(1980),Bernstein(1955)arguedthattherearefourphases

inthelifecycleofaregulatorycommission:gestation,youth,maturityanddecline.

Gestation is a phase during which the demand for regulation emerges: rising

40

dissatisfaction,forinstanceduetobusinesspracticesperceivedasunfair,activates

affected groups who begin pressurizing government to protect their interests.

Conflictbetweenthesegroupsandthegroupstoberegulatedensues.Protracted

struggle ends when a statute with vague wording is passed and regulatory

commissionsarecreated.Duringitsyouth,theregulatorycommissionoperatesina

conflictual environment and aggressively fulfils its mandate, while accumulating

experience. Over time, the policy crisis giving rise to regulation dissipates as the

groupsthatcalledfor itretireandasregulatedindustry issuccessful inappeasing

regulators.Thematurityphaseseestheregulatoryagencyrelyingonprecedentand

routineandbecomingincreasinglyparochial;regulatorsareeventuallycapturedby

industry,leadingtoscandalsthatsetthelifecyclebackinmotion.

Downs’ (1967) lifecycleofbureausalsoseeks toexplain theultimate rigidityand

capture of regulatory agencies by the regulated industry. Initially, the bureau

emergesfromthedemandforregulationbutmustseekexternalsupportinorderto

survive, as the groups supporting it conflicts with other groups. The bureau is

vulnerable to termination initially, but overcomes its “survival threshold” and

stabilizesonceitisabletoofferusefulservicesandtoroutinizerelationshipswithits

major clients. The bureau grows because over time it learns to perform better,

develops rulesandproceduresandrecords itsownexperience.Competition from

other bureaus, or simply the decline of its social function, however, cause the

bureau’sdecline.Bureausreacttodeclinebyacquiringadditionalfunctionsinorder

tosurvive.“Astimepasses,bureaus,likefirms,tendtodiversitytoprotectthemselves

fromfluctuations indemand” (Downs1967,p.20).Hence,establishedbureausare

unlikelytodie,becausetheywilladapttheirfunctionstochangesintheenvironment

whilealsoexpandingtheirclientbase.

ThatcherandCoen(2008)haveindeedconceptualizedtheevolutionoftheEuropean

regulatoryspaceintermsoflayeringandconversion:onceseparatenationalpolicies

graduallyevolvedtoaEuropeanone.Thatcher (2011)arguedthatthecreationof

Europeanagencieshasinvolvedlayeringandconversionofthepre-existinginformal

41

regulatorynetworks.Atanyrate,thosenetworksdidnotdisbandwiththecreation

oftheagency,leavingopenthequestionoftheircurrentrationale.

4. Whatexplainsregulatorynetworkslongevity?

Network expansion may entail an expansion of the functions performed by the

network,of the issueareas itconcerns itselfwith,orof itsmembership.Network

expansion may also mean establishing connections with new clients for new

purposes. Expansion takes place as the networks continues to fulfil some of its

originaltasks,whoseusefulnesstonetworkmembersisstillhigheventhoughithas

diminished foractorsoutsideof thenetwork, fromwhich thenetworkderives its

relevance.

Definitionsofkeyconcepts

This research revolves around a few important concepts, that it is important to

define.Inmanyplacesinthechaptersthatfollow,Iarguethatregulatorsnetwork

for legitimacy. I alsomake references to regulators’ credibility and reputation as

beingenhancedbynetworking. Further, Ishowthattheagencyofregulators isa

crucial determinant of network structure; and that regulators can be network

entrepreneurs.Thetopicofthelegitimacyofregulatorydecision-makingpermeates

the literature on regulation (Majone 1999, Lodge 2002, Black 2008, Black 2009,

Maggetti2010,Prosser2010,Keegan,Craufurd-Smithetal.2013).Nominally, the

legitimacyofregulatorsdescendsfromtheactofdelegationofregulatoryauthority

fromelectedrepresentatives(Maggetti2010)andfromtheirexpertise.However,the

contestednatureoftheregulatoryprocessofdecision-makingandtheconsequences

ofregulatorydecisionsonmarketandsocietalinterestshaveoftentriggereddebate

onwhethertheactofdelegationsufficientlylegitimizesregulatoryauthorities.Early

literaturecontendedthatregulatorydecision-makingwouldonlyaffecttheefficiency

ofregulatedindustries,butnothavere-distributionalconsequences(Majone1997).

However, these expectations were not met in reality: regulatory decisions have

42

profoundre-distributionalconsequencesandtheregulatoryactivity is riddenwith

societalimplications(Prosser2010).

Inthecontextofregulatorygovernance,“astatementthataregulatoris‘‘legitimate’’

meansthatitisperceivedashavingarighttogovernbothbythoseitseekstogovern

and thoseonbehalfofwhom itpurports togovern” (Black2008,p.144). Inother

words,aregulatorneedstobeperceivedtobelegitimatebythoseaffectedbytheir

decision.Thelegallegitimacyinherentinthestatutesoftheregulatoryauthorityis

not sufficient formarket actors to deem the regulator legitimate; if they do not

considertheregulatorlegitimate,affectedinterestshavelittleincentivetoconform

totheregulator’sdecisions.Inotherwords,regulatorsneedtopossesslegitimacyin

order to induce the hoped-for behavioural responses (Black 2008, p.148).

Importantly,regulatorsarenotjustpassiverecipientsoflegitimacybutcanactively

construct their own legitimacy (Black 2008, p.146), including by forming and

maintaining ties to organizations in their institutional environment, which are

perceived to be legitimate by those whose legitimacy claims they want tomeet

(Meyer and Rowan 1977). In the context of multi-level governance, individual

regulators need to establish their legitimacy both in the eyes of industry and of

domesticandsupranationalpoliticalinstitutions.Bynetworkingacrossjurisdictions,

regulationconqueran intermediary spacebetween levelsofgovernance (Jordana

2017),whichtheycanleveragetoincreasetheirlegitimacyinbothdirections.

Furthermore, Icontendthatregulatorsnetwork inordertoboosttheircredibility.

One of themain rationale for delegation of regulatory authority to independent

regulatory bodies at national level has been, indeed, enhancing the credibility of

regulatorypolicy(Majone1994,Gilardi2002).Theneedforcredibleregulatorypolicy

isheightenedinthecontextofmarketintegration,asuncertaintyanddifferencesin

legal, economic and institutional traditions impose further costs on industry. As I

show in this research, in the context of multi-level governance systems, the

regulatory network represents a repository of a great amount of information on

individualmarkets,whichitwouldbeverydifficultandtime-consumingtogather.

Supranationalregulatoryinstitutionsbenefitfromhavingasinglepointofaccessto

43

suchinformation.Hence,regulators,viatheirnetwork,areabletocontributetothe

supranationalpolicyformulationprocess.Collaborativerelationshipsacrosslevelsof

governance in the formulation of policy strengthens the credibility of regulatory

commitmentsatbothdomesticandsupranationallevel.

The notion of reputation is very closely tied to the mechanism of regulatory

networkingand the regulators’ legitimacyand role inpolicy formulation. Idonot

tackle the link between networking and reputation directly. However, interviews

suggested that regulators perceived that collaboration within networks had

reputation-enhancingeffectsonthem.Arguably,reputationisoneofthemainassets

ofregulatorybodies.Soimportantthatregulatorswithlowerlevelsofexpertisemay

evenbeoverlygenerouswith industry inordertoprevent industryfromexposing

theirlowercompetenceandruiningtheirreputation(Leaver2009).Areputationfor

impartial, informed,expertdecision-makinghasseveralpositiveconsequences for

regulators: it enhances their legitimacy (Schrefler 2012), it produces desirable

behaviouralresponsesinregulatedentities(CambiniandRondi2011),anditprotects

the regulators’ autonomy (Carpenter 2001) from political interference. Recent

literaturehasarguedthatregulators’willingnesstocooperatewithotherinstitutions

depends on their calculations of the reputational benefits of the cooperation

(Busuioc 2016). Moreover, the position that regulators hold within horizontal

networkstructuresofcollaborationisadeterminantoftheirreputation(Ingoldand

Leifeld2014,FischerandSciarini2015).Insum,reputationisasourceofpowerfor

regulators(Carpenter2010).Icontendthatnetworkcollaborationhasareputation-

enhancingeffectonregulators,asaresultoftheircontrolontheinformationthatis

collectedandgeneratedbythenetwork.

Thefirstpartofthisthesisarguesthatregulatorsnetworkfor“control”. Iusethis

wordasmeaning,essentially, turf, i.e.aregulatoryauthority’s jurisdiction(Wilson

1980).Ialsousethewords“sphereofauthority”toconveythesameconcept.The

reasonwhyIuse“control”inthetext,ratherthan“turf”,isthat“control”istheword

usedbymanyofmy interviewees.All throughtheanalysis, Ioftenemphasisethe

importanceof regulators’ agency in shaping andusing their network. Thedebate

44

concerningwhetheragencyorstructureismoreimportantforexplainingbehaviour

is a crucial one in the institutional theory of organizations (Heugens and Lander

2009).Theresultsoftheprocess-tracinganalysisinthisthesissuggestthatregulatory

networksareabletoexertagencyontheirenvironmentbyrespondingtochangeand

pressuresinautonomousways,andthatindividualswithinthemareabletoexploit

thenetworkinstitutionalizationtodisplaypolicyentrepreneurship(DiMaggio1988).

Thisbecomesparticularlyevident inthe lastpartofthethesis,where Idepictthe

emergenceofthenetworkofEuro-Mediterraneanenergyregulators(MedReg)asan

actofnetworkentrepreneurship.

Thehistorical reconstructions in this thesis show thatnetworkemergencealways

appears to be due to the initiative of specific individuals. However, the case of

MedRegisdifferentfromallothers.Theotherthreeregulatorynetworksstudiedin

this thesis emerged in very similar circumstances: radical reform, uncertainty,

governance tensions. Individual initiative, in those cases, sparked a process of

collaborationforwhichahugewindowofopportunityhadbeenopenedbychanges

occurringintheinstitutionalenvironment.Incontrast,MedRegemergedinabsence

ofcomparablecircumstances.Initiativewasthedeterminantofitsemergence.Policy

entrepreneursareactorscapableofactingascatalystsforaction(Carpenter2001),

who possess considerable expertise and enjoy strong reputation (Arnold 2014).

Expertise and delegated authority, however, are not sufficient for policy

entrepreneurstofurthertheirgoals;networktiesarealsoessential(Newman2008).

Themethodsofenquiryandcaseselection.

Theinterestofcomparativehistoricalanalysiswithembeddingcausalpropositions

intohistoricalprocessesincomparativeperspectiveacrossarangeofcasesaffords

theresearcherconsiderablefreedomintermsofmethodsandconsiderableleverage

in terms of results (Mahoney and Rueschemeyer 2003). This approach combines

consideration of objectively changing circumstances and contexts with close-up

analysisoftheactorsinvolved:thewaysinwhichtheyunderstandthesituationsthey

45

facearenotassumedawaybutintegralpartofthedatacollectionandanalysis(2003,

p.16).Forthisreason,themethodofprocesstracingisoftenusedincomparative

historicalresearch.

Theinterestofprocesstracingisreconstructingthemechanismslinkingacauseto

an outcome within specific cases. As such, it finds its roots and most common

application in analyses concernedwith the sequencing of events and the role of

temporalityinleadingfromthecausetotheoutcome(BlatterandHaverland,2014).

Suitablestartingpointsforprocesstracingareoftencriticaljunctures(Bennettand

Checkel,2013),astheyoffertheconditionsofuncertaintyandcontingencyidealfor

aninvestigatortoreasononalternativeexplanatorypaths.Earlynotionsofprocess

tracing confined themselves to outlining themain thrust of themethod, that is,

exploring“thechainofeventsorthedecision-makingprocessbywhichinitialcausal

conditionsaretranslatedintocaseoutcomes”(VanEvera,1997,p.64).

Laterscholarlydebatesoughttostrengthentheinternalvalidityofprocesstracingas

instrument to open “the black box of causality” (Beach, 2013). It underlined the

invariantnatureofthemechanismstheresearcheroughttotestortofindwithinthe

casestudy,whicharenottobeconfusedwiththeinterveningvariablesbetweentwo

events(Beach,2013):theyarenecessarypartsofamechanismofcausation,whose

forcerestsinitsuniquenessandcertainty.Those,inturn,canbeassessedthrough

tests confirming or disconfirming the hypothesisedmechanism (Mahoney, 2012).

Processtracingcanbeusednotonlytotesttheory,butalsotogenerateit(VanEvera,

1997; Beach and Pedersen, 2012), raising issues of external validity, to be tested

throughapplicationofthetheorygeneratedtoothercases.

Fromthestandpointofhistorical institutionalism,institutionsarecharacterizedby

stability or constrained, adaptive change. The concept of path-dependence, as

previously mentioned, underlies this view. Any explanation of the regulatory

networksconsideredheremustadoptthelogicofpath-dependence,sincetheyhave

outlivedtheirinitialraisonsd'être,sometimesquiteconsiderably.Theperspectiveof

this school of thought on institutions’ formation and change emphasizes the

identificationofthemomentwhentheseoccur.Thechoiceofthe“when”isessential

46

tothenmakethecasethatthepatternscreatedinthatinitialmomentpersistedand

exertedlong-terminfluenceontheorganization(Peters,2011,p.76).Contributors

haveunderstoodmajorchangetohappenat“criticaljunctures”,beingthestarting

points for new path-dependent processes (Collier and Collier, 1991; Mahoney,

2001a,2001b).“Critical junctures”happen inmomentsofpolitical indeterminism,

whenmultiple courses of action are possible the usual structural constraints on

action are relaxed, allowing “wilful actors” to shape circumstances “in a more

voluntaristicfashionthannormalcircumstancespermit”(Mahoney,2001a,p.7).

Historicalinstitutionalismalsoemphasizestheroleofideasinshapingtheformation

of institutions. However, ideas require individuals to be translated into an

organization.Hence,agencyiscrucialinthisperspective.ThisiswhyIchosetorely

on elite interviews with individuals who were directly involved or well-informed

abouttheeventsretracedinthisresearch.Eliteinterviewsrepresentthenecessary

complementtodocumentaryandarchivalresearchinthiscontext,asmanyofthe

events, the actions, the choices made by actors at the inception of network

cooperationhaveneverbeenrecorded,ifnotinsomememonotesburiedinthese

actors’desks.Intervieweeswereselectedaccordingtopurposivesampling:myaim

was to talk to individuals who were involved in the establishment and/or the

managementofeachofthenetworksconsideredaswellastoindividualswhoare

(orhavebeen)“simple”networkmembers.Iwasacquaintedwithseveralkeyactors

(including regulators, academics, and consultants) of each of the networks

consideredthankstopreviousworkexperience.Thenamesthatsomeofthemmade

duringtheinterviewshelpedselectingfurtherinterviewees.

IuseSocialNetworkAnalysis(SNA)toinvestigatethedriversofnetworkconnections

among regulators. SNA is a powerful tool of analysis because it allows for the

measurement of physically immaterial relationship structures that constitute the

fabricofnetworkedcooperation.ThroughSNA,onecananalyseregulatorynetworks

in relational terms rather than in organizational ones. The interplay between

structure and agency that is encountered when analysing networks is also

acknowledged by researchers who have made use of the method in historical

47

research(PadgettandAnsell1993,Gould2003).TheSNAonlyconcernsthenetwork

ofEuropeanenergyregulators.

I gathered complete network data from regulators. I asked the Communication

Officersofeachmemberregulatoryauthoritythefollowingquestion:“Thinkofthe

individuals you (or somebody at your regulatory authority) exchange information

withmoreregularlyandfrequently.Whichregulatoryauthoritydotheybelongto?”.

Thequestionnaireaskedregulatorstolisttheregulatoryauthoritiestheyaremost

oftenintouchwithforexchangeofopinionsandsuggestions.Italsoaskedthemto

namethemostactivenetworkmembersintheiropinion,andthemostcompetent.

Finally,afewwordsonthecasesselectedforinvestigation.Inthisthesis,Iexamine

fourtransnationalortrans-jurisdictionalnetworksofenergyregulators:

1. The National Association of Regulatory Utility Commissioners (NARUC)

comprising50statelevelutilityregulatorsfromtheUnitedStatesofAmerica

(USA).

2. The Council of European Energy Regulators (CEER) comprising 28 full

membersfromthe28MemberStatesoftheEuropeanUnion(EU).

3. The Energy Regional Regulatory Association (ERRA) comprising initially 24

members from countries in Central and Eastern Europe and the New

IndependentStates(CEE/NIS).

4. TheAssociationofMediterraneanEnergyRegulators(MedReg)comprising24

members from21countries in theEU-definedEuro-Mediterraneanregion,

comprisingSouthernEuropeanMemberStates,AccessioncountriestotheEU

andmostofthecountriescomprisedintheEuropeanNeighbourhoodPolicy.

The first twonetworks are non-brokerednetworks, born out of the spontaneous

initiatives of theirmembers (Provan and Kenis 2008). NARUC and CEER are long

standing actors in their respective regulatory space that have had significant

influence on the policy process, as the chapters will outline. Chapter 2 and 3

investigate the conditions allowing them this opportunity. The second couple of

networksinvestigatedinthisthesisaremandated,orbrokerednetworks(Provanand

48

Kenis2008).Theirinceptionisinter-relatedwiththedevelopmentsoftheprevious

twonetworks.Specifically,theestablishmentofERRAwascoachedbyNARUC,which,

since1998,hasbeencooperatingwiththeUSAgencyforInternationalDevelopment

(USAID) inthedeliveryofregulatorytechnicalassistancethroughouttheworld. In

contrast, MedReg is a transversal case: although CEER was not involved in its

establishmentdirectly,MedRegoriginatedfromtheinitiativeofasmallgroupofEU

energyregulators,whoset ituponthebasisoftheirexperiencewithinCEER.The

analysisof these twocases,which is carriedout in the thirdand finalpartof the

thesis,showsthatregulatorynetworksareconsideredcapableofentrenchingpolicy

change.

Comparisonofall fourcaseswitheachotherrevealsthatNARUC,CEERandERRA

underwentsimilarevolutionaryphases,althoughatdifferentpacesandovertime

spansofdifferentlengths.Allthreeemergedinthecontextofradicalchangeinthe

policyparadigmoftheenergysector(Hall1993).Moreover,themembersofNARUC,

CEERandERRAbelongtoawell-definedpolity:theUSA,theEU,andinitiallywithin

ERRA, the countries of Central and Eastern Europe intending to enter the EU.

MedReg, in contrast, didnot emergeoutof apolicy crisis or aparadigmchange,

entailing uncertainty and opportunity structures. Moreover, the so-called “Euro-

Mediterranean” region is, at present, more of a denomination concocted by

Europeanpolicy-makersthananactualregionalpolitywithsharedinstitutions,rules,

oratleastgoals.ThesefactsdistinguishMedRegfromtheothercasesexaminedin

thisworkandmayexplainitslukewarmachievementsintermsofpolicycontribution.

At the same time,MedRegmaywell represent anempirical caseof collaborative

initiativehavingmutualtrustasitsoutcomeratherthanitspremise(Powell1995).

49

PART ONE - EMERGENCE AND CONSOLIDATIONOFNETWORKSOFENERGYREGULATORS

50

Part I. Introduction: the spontaneous emergence andconsolidationoftransnationalregulatorynetworks.

Theliteratureonregulatorynetworkshasoftendiscussedthecircumstancesoftheir

official establishment,usually tracing itback to thepreferencesofeitherpolitical

principals or supranational institutions. The spontaneous, bottom-up origins of

regulatory networks are sometimes acknowledged in the relevant literature, but

rarely investigated. Yet, understanding the rationale of spontaneous regulatory

networkingacrossjurisdictionsiskeytounderstandingwhyregulatorsnetwork,or,

moreprecisely, how theybenefit fromnetworking. This is particularly relevant in

systems of governance, articulated across multiple levels. In these systems,

regulatory authority is distributed across levels to regulatory institutions, whose

mission andmandatespartially overlap. Regulators at lower levels of governance

cometooccupyanintermediaryposition(Jordana2017)betweenlevelsastheyfulfil

theirtasksincompliancewiththepolicymandatesderivingfromeitherlevel.

Therefore,Inthisfirstpartofthethesis,Ipose,andanswer,thefollowingquestions:

- what are the conditions leading to the spontaneous emergence of trans-

jurisdictionalregulatorynetworks?

- whataretheconditionspromptingnetworkconsolidationfromaloosesetof

informaltiestoanentitywithsomeofthepropertiesoforganizations?

Thefindingsoftheanalysisarethatregulatorsbeginnetworkingwhenevertheyare

facinggreatuncertaintyinthefulfilmentoftheirtasks.Beinguniqueinstitutionsin

theirjurisdictionalcontext,regulatorsreachouttopeersinothercountrieswhoare

facing similar uncertainty and are therefore likely to be facing similar challenges.

Hence, the answer to the first question is that similar levels of uncertainty drive

networkformation.Regulatorsbeginnetworkingforexpertise,usingeachotheras

sources of information and of epistemic legitimacy in order to strengthen their

51

decision-makingdomestically.Initially,theircollaborationasanexclusivelypeer-to-

peerdimension,confinedtotheirprofessionalcommunity.

Consolidationoftheirloosesetofprofessionaltieshappensasaresultofthetensions

acrosslevelsofgovernancebecomepermanent.Networkconsolidation,thus,occurs

asaresultoftheinteractionbetweenregulatorsandtheinstitutionalenvironment

surrounding them. Regulatory decisions have distributional as well as economic

consequences.Inotherwords,theirdecisionscreatewinnersandlosers.Asaresult,

regulators face domestic opposition from interests that are damaged by their

decision-making.Hence,asgovernmentseekstoappeaseconstituencies,regulators

facetheconcretethreatoflosingtheirpowersand/orseetheirautonomycurtailed.

Hence, regulatory networks emerge as organizational entities as a result of

governance tensions that pose the continuous threat of partially reducing their

powersortheirauthority.

Whenever this threat is common, regulators strengthen their collaboration by

consolidatingtheirnetworkintoarepresentativebodyofalloftheregulatorsintheir

jurisdictionalsetting(whereaspreviouslysmallersubsetsofregulatorsinthemulti-

level governance system would gather more frequently) displaying some of the

properties of organizations: headquarters, legal personality, set budgetary

contributions,etc.Moreover, regulatorsusetheirnetworktoseekalliesathigher

levelsofgovernance,e.g.thesupranationallevel,byexploitingthelatter’squestfor

informationontheirlocalmarketsinordertoformulatepolicy.Partnership,inthe

formofasymbioticinterdependence(PfefferandSalancik1978),withsupranational

(federal) institutions wielding supranational regulatory authority represented an

enablingfactorfornetworkconsolidationandparticipationinthepolicyprocess.This

opportunity has the effect of increasing the reputation of the regulators and of

furtherstrengtheningthenetwork.

IfocusmyanalysisontheempiricalcasesoftheNationalAssociationofRegulatory

Utility Commissioners (NARUC) and the Council of European Energy Regulators

(CEER).Thesegather,respectively,statePublicUtilityCommissioners(PUCs)ofthe

USAandNationalenergyRegulatoryAuthorities(NRAs)oftheEU.NARUCemerged

52

in 1884 and was officially established in 1889; CEER emerged in 1997 and was

officiallyestablishedin2000.Bothorganizationshaverisenfrombeingconstituted

aslittlemorethaninformalprofessionalassociationstobecomingkeyactorsinthe

regulatory policy formulation process in their respective jurisdictions. In the two

chapterstofollow,Iretracetheiremergenceandconsolidation.Indoingso,Iidentify

thecriticaljuncturesthatshapedthenetworkidentityandmission.Intheconclusions

tothispartofthethesis,Idiscusstheirstoriesincomparativeperspective.

TheUSAandtheEUaretwo“regulatorypowerhouses”(Drezner2007).Bothhave,

over time, increasingly relied on regulatory policy, substituting “rowing” the

economy with “steering” market actors’ behaviour towards socially desirable

outcomes(OsborneandGaebler1993).Therefore,theUSAandtheEUrepresentthe

embodimentsofthenotionofthe“regulatorystate”(Sunstein1989,Majone1997,

Baldwin,Scottetal.1998,Moran2002,Lodge2010).WhereasintheUSAthenotion

of the regulatory statehasbeenassociatedwith the riseof federalagenciesand,

moregenerally,ofthefederallayerofgovernment,intheEuropeancontexttherise

oftheregulatorystatehasbeenqualifiedbothastheriseoftheEuropeanlayerof

governance at supranational level, and as the move from positive to regulatory

governance(Majone1997)atnationallevel,entailingtheretreatofthestatefrom

theprovisionofcertainservices,e.g.energyservices.Inbothpolities,thedeepest

rootsoftheregulatorystatearefoundintheregulationofeconomicsectors,and,in

particular,infrastructure(KanazawaandNoll1994,Lodge2008).

ForbothNARUCandCEER,themomentofofficialfoundationrepresentedlittlemore

than a formality, their real identities having formed at different stages in their

evolution in response to certainevents. In theirorigins andevolution“time itself

becomesanelementofthecausalexplanation,afactorinthemodel”(Büthe2002,

p.486).Investigatingcasesovertimeallowstheidentificationofcausalmechanisms

andexplanatoryfactorsrootedinpreviousphenomenaorevents,whosesignificance

informs later events (Farrell and Newman 2010, Farrell and Newman 2014).

However,definingthemomentortheeventtriggeringthepath-dependentsequence

of events that leads to the outcome is a delicate task, which researchers in this

53

traditionare sometimesaccusedofperformingarbitrarily (Capoccia andKelemen

2007). “Critical junctures” are moments presenting institutional actors with

opportunity structures that they may exploit in order to pursue their own

preferences.

CapocciaandKelemen(2007)addressamajorshortcomingintheexistingliterature

oncritical junctures: lackofmethodologicalguidanceonhowtorecognize,define

andanalysethem.Theliteratureisrichinmacro-historicalanalysesidentifyingthe

emergence of critical junctures as a result of broad, structural causes of change,

ratherthanfromactionsanddecisionsmadeduringthecriticaljunctureitself.Their

model,therefore,aimstoprovideasolidframeworktorecognizeandanalysecritical

junctures.Thecriteriatheyadvanceareasfollows:

• The critical juncture must be a moment characterized by political

indeterminism and uncertainty, when multiple courses of action are

possiblebutonlyoneischosenthankstocircumstancescombinedwith

theagencyofwilfulactors.

• Thelengthofthecriticaljuncturehastobemuchshorterthanthelength

of the process it generates: the briefer the juncture relative to the

outcome,themorecriticalitis(theycallthismeasuretemporalleverage).

• In order to identify a critical juncture, one should enquire on the

counterfactuals (alternative courses of action that might have been

chosen,butwerenot,asmuchasavailableinformationallows).

• Thefarthesttheprobabilityoftheoutcomebeingachievedinabsenceof

thejunctureisfrom1(theydefinethismeasureasprobabilityjump),the

higherthelikelihooditisacriticaljuncture.

• Thehighertheprobabilityoftheoutcomeproducingitselfpost-juncture,

thehigherthelikelihoodthatthejunctureisacriticalone.

• Contingencyisthekeyelementofcriticaljunctures.

Criticaljuncturesneednotbeperceivedaspath-breakingmomentsastheyoccur;to

thecontrary,theycouldbesmall,apparentlynotverymeaningfuleventsordecisions

whoseposteriorimpactisalmostsurprisinglystrong.Inthetwochaptersthatfollow,

54

Iinductivelybuildthecausalnarrativelinkingcriticaljuncturestonetworkemergence

and consolidation through process tracing, by relying on documentary analysis,

archivalresearchandeliteinterviewmaterial.Ararelyhighlightedfeatureofprocess

tracingconcernsitsrelianceonthetestimonyofthosewhoweredirectlyinvolvedin

theprocessunderstudy(VanEvera,1997)tomakesenseoftheirdecision-making

processandoftheeffectthatinstitutionalarrangementsandotherstimulihadonit

(GeorgeandMcKeown,1985).BlatterandHaverland(2014) include“confessions”

among theelementsneeded forprocess tracing, as actors’ perceptions, although

biased, can shed light on decisions as well as non-decisions, given anticipated

consequences. Such emphasis on contingent individual behaviour matches the

concernsofanalysesofcriticaljunctures,thusbringingthetwoframesofreference

toconverge.

Thefollowingtwochaptersareinformedbyatotalof41face-to-facesemi-structured

interviewsIcarriedoutwithcurrentandformerregulatorsandstaffexecutivesfrom

NARUCandCEER.Intervieweeshavebeenselectedonthebasisoftheirrole(past

andpresent)within thenetwork: chairs, vice-chairs, executive directors and staff

executives.Atthesametime,Iinterviewedasmanynetworkmembersnotholding

official leadership or coordinating roles in the network as time and resources

allowed,inordertograsptheirperceptionsofthenetwork.Finally,Ialsointerviewed

individuals who gravitated around both networks in different capacities, such as

privateconsultantsandrepresentativesofacademiaandinternationalorganizations.

Allintervieweeshavebeenguaranteedfullanonymity.Acompletelistofsummary

intervieweeinformationisprovidedinAppendix1.

NARUCandCEERaretheby-productoftheinnerworkingsoftheenvironmentthat

generated them. Their importance grew, perhaps unexpectedly, for the response

theymanagedtoarticulatetocertainevents.InthecaseofNARUC,regulatorswere

facingthethreatofprogressivedisempowermentbecauseoftheriseofthefederal

level of government; in the case of CEER, regulators enjoyed progressive

empowerment, and the network grew in importance, until regulators were

mandatedtocoordinate,becauseafederallevelofgovernmentwouldnotarise.

55

2. Thehistoryof theNARUC:defendingtheturfof thePublicUtilityCommissions.

ThelonghistoryofutilityregulationintheUSAisalsothelonghistoryoftheNARUC.

TheNARUCmembershipconsistsofutility regulators fromthe50states (plus the

District of Columbia, Puerto Rico and the Virgin Islands). Hence, it has a

homogeneous and horizontal membership structure. Despite its longevity, the

NARUChasrarelybeenobjectofin-depthstudy.PaulRodgers,NARUC’sExecutive

Directorfrom1965to1996(Beecher,2012),retracesthehistoryofNARUCfromits

originsto1978.Thebook,publishedbytheNARUCitself,onlyexistsinafewcopies4.

Childs (2001) provides the only thorough historical overview of the NARUC,

highlightingthe instrumentalrole ithadindefendingthePUCs’turf inthefederal

system.InvestigationofNARUCasnetworkismuchmorerecent(Beecher,2012).

Inhisseminalpaperonissuenetworks,Heclo(1978)remarkedhowintheUSAthe

politicsofpublicadministrationusuallyarousedlessinterest,andthusscrutiny,than

thePresidencyandCongress.Childs(2001)echoesthisview,notingthatstudyofthe

politicsof regulation inAmericahasoverwhelmingly focusedon federalagencies.

StudyingtheNARUCisanimportantsteptowardsaddressingbothoftheseconcerns.

TheemergenceoftheNARUC:thepublicdemandforregulationof

therailroads.

ThebirthofregulationintheUSAisentwinedwiththeappearanceoftherailroads

(Kolko 1965, Kerr 1968, Nice 1987, Himmelberg 1994, Kanazawa and Noll 1994,

Dobbin 1995,DowdandDobbin 2001). The sequenceof events leading from the

appearanceofrailroadstostateregulationcomprisestechnologicalprogress,abuse

of market dominance, popular discontent, litigation, and eventually political

4TheNARUC,verykindly,madethebookavailabletome.

56

interventionestablishinganew institution for theresolutionofcontroversies: the

PublicUtilityCommission(PUC).

In the late 1820s, the steam-powered locomotive technology revolutionized

transportation across the American states. The railroad increased the speed of

transportationacrossthemaincentresofcommerce,quicklygainingbusinessatthe

expense of the incumbent water and turnpike transportation companies. Private

railroad companiesquicklymultiplied, eachbuilding theirown infrastructure. The

technological disruption caused a series of abuses: wherever the railroads faced

competition from other technologies -- typically, over long distances, where

competition from canal and riverboats kept prices low (Law and Long, 2011) --

customersdemandedbelow-costservice;intheshorthaul,however,railroadshad

virtuallynocompetitorsandexploitedtheirmonopolypower.

The ensuing volume of litigation, between companies, shippers, and customers

quicklymadeevidentthelimitsofstateslegislaturesinregulatingrailroadoperations

andrespondingtobusinessandconsumercomplaintsovertheirpractices.Thecourts

lackedthenecessaryexpertisetodealwitheconomicregulationoftechnicalsectors

andhadatypicallypassiverole,astheywouldtakeactionuponfilingofsuitsonly

(Rodgers, 1979). In the early 1870s, the protests of the farmers prompted the

legislaturesoffourMidweststates(Iowa,Illinois,WisconsinandMinnesota)topass

legislation,imposingregulationofrailroadcompaniesbyregulatorycommission.

AsFainsodandGordon(1948)note,Americanregulationwasnottheproductofa

farsightedplan,butrather“aseriesofempiricaladjustmentstofeltabuses”.Railroad

companiesfeltthatregulatorylegislationdamagedtheirinterests,andchallengedits

constitutionality inaseriesof legalcases. Inonecase,MunnvIllinois(1876)5,the

SupremeCourtaffirmedthatwhenprivatepropertybecomesaffectedwithapublic

interest itcanbesubjecttogovernmentalregulation(Rodgers,1979).Thisverdict

representedthecornerstoneofthe legitimacyofstateregulationoftherailroads.

“Thefirstofthese“commissions”appearedinRhodeIslandin1839,NewHampshire

5Munnv.Illinois,94U.S.113(1876)

57

in1844,Connecticutin1853,NewYorkandVermontin1855,Mainein1858,Ohioin

1867,andMassachusettsin1869”(Rodgers,p.3).

TheseearlyPUCsdidnothaveratemakingpowers;theirresponsibilitieswerelimited

to undertaking an appraisal of theworth of a company’s property and enforcing

railroad safety standards. They often consisted of only one commissioner. Their

recommendations about rates were often plainly ignored by the railroads. Their

establishment had been motivated by the willingness to calm public unrest

(KanazawaandNoll1994).

Theseearly-establishedPUCsmetsixtimesbeforethecreationofNARUCitself.The

earliestof thesemeetings,held in1874, involvedCommissioners fromWisconsin,

IllinoisandMinnesota(Rodgers,p.5).Rodgersdoesnotcommentextensivelyonthe

content of those informal meetings. He mentions a common willingness of the

regulatorstoshareexperiences,compareandcontrasttheirpowersandobligations,

and thechallenges theywere facing in regulating railroads.Asmorecommissions

werecreatedinthevariousstates,theyalsojoinedtheseinformalmeetings.Thelast

recordedmeetingwasheldin1881.

In1886,theSupremeCourt,intheWabashvsIllinois6case,affirmedtheexclusive

competenceofCongressover interstatecommerce,confiningtheauthorityofthe

statecommissionstotheirstate.Onlyayearlater,in1887,Congresscreatedthefirst

federalregulatoryagencyoftheUnitedStates:theInterstateCommerceCommission

(ICC). The first president of the ICC, Judge Thomas M. Cooley of Michigan,

“recognized that his new agency then lacked the resource to carry out even the

modesttaskssetforitbyCongress”7(Rodgers,p.9).Therefore,Cooleydecidedto

6Wabash,StLouisandPacificRailwayCompanyv.Illinois,118U.S.557(1886) 7TheICCpowersforeseenintheActwerelimitedandvaguelyworded:ithadthepowerto

require comprehensive information from carriers, as well as testimony; to deal with

complaints, including from PUCs; to request the carrier tomodify its behaviour if found

contraveningtotherequirementsoftheact,butincaseofincomplianceithadtoreferthe

58

call upon the state regulators to provide information on the operations of the

railroad and their regulation in the different states. In order to establish such

dialogue, Judge Cooley convened the firstmeeting of the Convention of Railroad

Commissioners, which was to become the National Association of Railroad

Commissioners,atICCofficesonMarch5,1889(Rodgers,p.8).Representativesof

21 PUCs attended the Convention8. Judge Cooley became the association's first

president. Hence, the ICC and NARUC first chairmen were the same person,

representingtheircooperativerelationship.

Theperiodduringwhichtheseeventsunfoldedrangesfromtheearly1890stoWorld

War I (WWI) and is commonly referred to as the Progressive Era. Progressives

nurtured the conviction that regulation should and could be depoliticized by

entrustingexpert regulatory commissionswith regulatorypowers tobeexercised

independentlyofpoliticsand industry (Bernstein,1977).These ideasconcurred in

theestablishment,in1907,ofthefirst“modern”statecommissions,properlystaffed

and overseeing several utility markets: New York and Wisconsin, which quickly

becamemodelsforallotherPUCs(Anderson1980).

casetoacourtoflawinthestatewherethecarrierheldofficeorwheretheinfractionhad

takenplace.http://www.historycentral.com/documents/Interstatecommerce.html8Theintentoftheconventiontoencourageregulatoryconvergenceintherailroadsector

emerged since the firstmeeting: the first reports to the convention concerned “Uniform

Railway Statistics” and “Uniformity in Reports from Railroad Companies”. A third report

concerned“UniformClassificationofFreights”.Thesereportsresultedinthepromotionof

resolutions. One of them, initially recommending the adoption of a uniform freight

classification for the roads across the country, was modified into a lesser prescriptive

formulation upon a roll call vote. Significantly, Rodgers remarked (p. 10) “only State

representativesvoted[…]andeachrepresentativehadonevoteirrespectiveofthenumber

ofrepresentativespresentfromaState”.Clearly,thiskindofvotingsystem,coupledwith

direct access to federal regulators, representeda strong incentive for state regulators to

attendconventionmeetingsinordertomaketheirinterestsheard.

59

TheProgressive Era-inspireddecisions to grantmorepowers to the ICC,with the

HepburnActof1906andtheMann-ElkinsActof19109,markedthebeginningofthe

conflicted relationship between state and national regulators (Childs, 2001).

Eventually,a logicof“pragmaticfederalism”(Childs,2001)prevailed,accordingto

whichfederalagenciesandstateregulatorscooperatedintheinterestofpreserving

their respective regulatory powers. However, the railroads’ lobbying, public

discontentwiththerailroads,andsomePUCs’discriminatoryratemakingpractices

resultedinalegalbattle,whichhadlastingconsequencesforregulatoryfederalism

intheUnitedStates.

TheShreveportcaseandWWI:questioningthelegitimacyofstateregulation.

ThestoryofhowtheprocesswherebyNARUCbecamethesinglevoiceofAmerican

PUCs centers on the Shreveport case10. Both Rodgers (1979) and Childs (2001),

identifythiscaseascrucialinthehistoryofPUCs.TheShreveportcaserepresented

the beginning of their conflicted relationshipwith the federal level of regulation,

sincethenperceivedtobeconstantlyattemptingatdeprivingthemoftheirpowers.

In1912,thePUCofLouisianafiledacomplaintwiththeICCagainstaTexanrailroad

company. The Louisiana PUC complained that intrastate railroad rates in Texas

(establishedbytheTexanPUC)weremuchlowerthaninterstateratesbetweenTexas

andLouisiana(establishedbythe ICC).Thisresulted inLouisiananshipperspaying

9TheHepburnActgaveICCrulingstheforceoflawandempoweredittochangearailroad

ratetooneitconsidered"justandreasonable,"afterafullhearingofacomplaint.TheMann-

ElkinsActplacedtheburdenofproofontherailroads;forthefirsttime,theywouldhaveto

activelydemonstratethataratewasreasonable.Interstatetelegraph,telephone,andcable

companies,weredeclaredtobecommoncarrierswithinthepurposeoftheAct,andwere

placedundertheregulatingauthorityoftheCommission.Onimpulseoftherailroads,the

ActalsocreatedtheCommerceCourttoreviewICCdecisions(Kerr,1968).10Houston,EastandWestTexasRailwayCompanyvsUnitedStates,234US342,58L.Ed.

1341–theparagraphsonthecasearebasedonRodgers(1979)andChilds(2001)accounts.

60

higherpricestoreachcommercialcitiesineasternTexas,closetotheborder,than

Texanshippersdid,eventhoughthelatterhadtocovermuchlongerdistanceswithin

Texas. The case took the name from the city in Louisiana, which the shippers

competedtoserveandoriginatedthecase:Shreveport.

TheinitialdecisionoftheICCwastocaptheinterstateratesandtoordertherailroad

companiestoadjustintrastateratessoastonotbediscriminatorywhencompared

tointerstateones.Theycoulddosobyeitherraisingintrastateratestotheinterstate

level,orbyloweringtheinterstaterates.Thedecision,insteadoffillingit,seemedto

maketheinstitutionalanddecisionalvacuumoverthemattermoreevident,asthe

railroadsseemedtobeauthorizedtomakethesedecisionswithouttheapprovalof

eithertherelevantPUC,whoseauthoritywasentirelybypassed,ortheICC.

Thislongcaseculminatedina1914SupremeCourtdecisionstatingthatthepower

toalterintrastaterateswhenthosediscriminatedagainstinterstateonesrestedin

the ICC (Childs,2001).Thiswas instarkcontrastwiththeprovisionsof theAct to

RegulateCommerce,institutingtheICC,whichstated“theprovisionsofthisactshall

not apply to the transportation of persons or property wholly within one state”

(Rodgers,1979).Thecourtconsidered,however,thatdiscriminatoryintrastaterates

wouldconstituteanundueburdenoninterstatecommerce.Ineffect,theprimacyof

federalregulationhadbeenestablished.

ThislegalbattlerepresentedtheconsolidationofNARUCastherepresentativebody

ofUSPUCs.ThroughNARUC,PUCsstronglyopposedtheSupremeCourtdecision,as

itwasfactuallysubordinatingtheir jurisdictiontothefederalone.Adelegationof

NARUCrepresentativesevenwenttoseeoneofthejudgesdecidingthecaseattheir

vacationhome(Rodgers,p.28)inordertomaketheircase.TheShreveportDoctrine

defactodeprivedPUCsoftheirauthorityoverrailroads.Bythistime,PUCsexistedin

most states and oversaw other utilities as well (i.e. telephone, electricity, gas)11.

11 The history of how PUCs came to regulate other sectors, besides the railroads, in

interwovenwiththehistoryoftheelectricityutilitiesintheUSAandinparticularwiththe

61

NARUCdraftedresolutionsmeanttoregainthePUCs’lostleverageoratleastavoid

being further deprived of authority (Childs, 2001). The political response from

Congresswastocreatea“specialjointsubcommittee[…]toconductaninvestigation

intogovernmentregulationofinterstateandforeigntransportationandgovernment

ownershipofallpublicutilities”.Thesubcommitteenevercompleteditstaskbecause

theUSAenteredWWI.

TheconsolidationoftheNARUC:post-WWIarrangements.

WhentheUSAdecidedtoengageinWWI,itbecameevidentthattherailroadsystem

wasunabletomeetthecoordinationdemandsofthewarandriskedplunginginto

chaos.ThispromptedtheCouncilofNationalDefencetoaskrailroadcompaniesto

halt competitive activities and to coordinate their operations nationally. The

companiescreatedtheRailroadsWarBoardtothateffect.Initially,theICCtriedto

preserve the extant system of regulation and to cooperate with theWar Board,

althoughrejectingtheircallsforgeneralrateincreases.However,itlateragreedto

unifiedoperationunderthedirectionofthePresident(Splawn1939).

In1918,PresidentWilsoncreatedthetemporaryUSRailroadAdministration(RA),to

managetherailroadsforthedurationofthewar.MrMcAdoo,SecretaryofTreasury,

President’s son-in-law and main advisor on the matter, became the RA Director

General. He populated the RA with railroad executives, who, empowered with

federal executive powers by the Federal Control Act of 1918, set off to improve

railroadsmanagement. In sodoing,however, theydisrupted thepolitical balance

characterizingtheProgressiveEra.Swiftrateincreaseswereauthorized,damaging

shippers.TheICCwasonlycalleduponforadvice,itspowersvirtuallyabrogated.Only

oneNARUCrepresentativesatinoneoftheRAcommittees(Kerr,1968),sothatthe

figure of Samuel Insull. Insull was a powerful and influential utility executive who was

instrumentalintheverycreationofmodernStatePublicUtilityCommissions,asshallbeseen

laterinthechapter.Bythe1910s,PUCsregulatedalloftheutilities.

62

representativenessofPUCswasreducedtotheminimumpossible.Recognizingthat

tensionwasmountingagainsttheRAanditsmanagers,McAdoo’ssuccessor,Walker

Hines, who was also a railroad executive, tried to re-gain the shippers’ and the

NARUC’ssupportbyincreasingtheirrepresentationwithintheRA.Hetriedtoco-opt

PUCsbysettingupaPublicServiceDivision,headedbythethenpresidentofNARUC.

These decisions improved the relations between state regulators and the RA;

however,stateregulatorswerepainfullyawareoftheirconsiderablelossofcontrol

over the practice of industry in their state and concerned about post-war

arrangements.

Attheendofthewar,theentireregulatoryandownershipsystemoftherailroads

was put under discussion. Hence, a window of opportunity opened for all

stakeholderstoadvancetheirpreferences.Morethanthirtyplansforreformwere

proposed to Congress (Waterman, 1919). Hines proposed the regionalization of

regulation(andofPUCs)withintheStatesundertheaegisoftheICC(Kerr,p.134).

The railroad companies called for the elimination of state regulation and

centralizationofregulatorypowersatthefederallevel(Post,1918;vanMetre,1918).

Partofpublicopinionwasinfavourofgovernmentownership(Stonex,1919).Inits

1919 annual report, the ICC recommended continued private ownership and

operationundergovernmental regulation (Eastman,1919;Splawn,1939).The ICC

wanted its powers expanded and regulation tomove beyond rates to involve all

aspectsofrailroadservice,ensuringitsstability,andefficiency.Shipperswanteda

returntothestatusquoante(Kerr,1968).

TheNARUC“urgedtheeliminationoffederalcontroloftherailroads[…],anICC-State

cooperative mechanism for ratemaking, and the eradication of the Shreveport

Doctrine.”(Rodgers,p.20).“Butthecommissioners’mostimportantroleinthepost-

war legislative debate was to communicate, through the National Association of

RailwayandUtilitiesCommissioners,with the InterstateCommerceCommission in

developingargumentstocountertherailroadattackontheregulatorysystem”(Kerr,

p.202).ThefinaloutcomeofthestrugglewastheTransportationActof1920,which

strengthenedthepowersoftheICC,butstillenshrinedtheShreveportDoctrine.

63

TheattitudeofPUCstowardstheICChadbecomeextremelyguarded,asexemplified

by the following extract from a NARUC record dated December 10, 1920 and

consistingofalettertheNARUCsenttotheICCinrelationtoapendingproceeding

broughttotheICCbyrailroadcompanies.

“Inthisconnection,Icallyourattentiontotheprovisionsofparagraph17of

Section1of the InterstateCommerceAct as amended,whichprovide that

nothingintheActshallimpairoraffecttherightofastateintheexerciseof

its powers to require just and reasonable service for intrastate business

“exceptinsofarassuchrequirementisinconsistentwithanylawfulorderof

theCommissionmadeundertheprovisionsofthisAct”.Ifthecommissionhas

power tomake, and doesmake, an order against carriers prescribing the

terms of side-track agreements, the effect of it may be to prevent state

authoritiesfrommakingorderscontrollingthetermsuponwhichside-tracks

shallbe installedandserviceprovided.Theseare localmatters,concerning

which,ifthefederalcommissionhaspower,itisaseriousquestionwhetherit

willdesiretoactinsuchawayastointerferewiththejurisdictionofthestate

authorities.”

ASupremeCourtdecision in1922(WisconsinPassengerFareCase)mandatedthe

ICCtobereceptiveofPUCsinstancesinthosecaseswhenitintervenedonintrastate

rates. Inthatyear,most ICCmemberswereformerNARUCmembers:6ofthe11

positions intheICCboardwereheldbyformerstateregulators,andeventheICC

headwasaformerNARUCPresident(Childs,2001).Cooperationbetweenthetwo

organizationsresumedandgraduallydeepened,untilNARUCmanagedtoeradicate

the Shreveport Doctrine and to prevent its reappearance in the statutes of

subsequent federal agencies. According to Childs, state regulators, “battling

attemptstocentralizeregulatoryactioninWashington”(p.702),managedtowork

out a role for themselves “in the emerging modern regulatory state” (p. 703).

Although PUCs eventually lost their powers in railroad regulation following the

deregulation reforms of the 1980s (Goldin and Libecap 2008), the experience

consolidatedNARUC’sroleastheirrepresentativeorgantowardsfederalinstitutions.

64

Critical junctures inthehistoryoftheNARUC:battlingattemptstogetridof

PUCs.

Critical junctures are events determining change from the status quo ante in

unexpectedways.TherewereseveraldefinitionalmomentsinthehistoryofNARUC:

the foundational moment, in 1889; the Shreveport case in 1912 and ensuing

doctrine;theperiodfollowingWWI,whentheallocationofregulatorycompetence

over US railroads was put into question. However, not all moments bear the

propertiesofcriticaljuncturesasdefinedbyCapocciaandKelemen(2007).Thefirst

two events lack the necessary contingency and unpredictability: PUCs had been

informallymeetingsince1874anditwasJudgeCooleytocallfortheirconvention,

establishingNARUC, not the regulators. As for the Shreveport case, although the

InterstateCommerceActsafeguardedPUC’sauthoritywithintheirownstates,the

practiceofcuttingaStateshipper’scoststobolstertheircompetitivenesswasbound

totriggerlitigationandthenstumbleuponfederaloversight.

Thepost-WWIandpost-RAsettingofuncertaintyoverthefutureoftherailroads,

instead,qualifyascriticaljunctures:

1) Uncertaintywashighandusualconstraintsonactionwererelaxed,sothatall

parties could advance their own preferences, including calls for public

ownership;

2) Thecounterfactualswere,therefore,many;influentialgroups(railroads,ICC,

politicians)allhadadifferentvision;

3) Indeterminacy was high. A complex puzzle of preferences had emerged:

railroadswantedtokeepprivateownershipbuttodisposeofcompetitionin

favourofcentralizedregulation.TheICCwascallingforanexpansionof its

ownpowers aswell. Therefore, PUCswere seriously at riskof losing their

competencesoverrailroads.

4) Eventually, NARUC and PUCs remained and subsequent legislation

institutionalizedtheirinvolvementinICCdecision-makingwhenthistouched

65

uponintrastaterates.ThismeantcementingPUCsandNARUCintotheUSA

regulatory system, which is the real outcome of the struggle. Hence, the

probabilityofNARUCbecomingastableactorinUSutilityregulatorypolicy

dramaticallyincreasedpost-juncture.

5) Contingencies were key in this case. In particular, the attitude of RA

executivesduringthewarperiodbroughtregulatorsatbothlevelstoforma

coalitionandtobringshippersontheirside.

Most importantly, the events surrounding the post-WWI struggle for regulatory

authoritytriggeredregulators’awarenessthatconflictoverthedegreeofregulatory

control between the state and the federal level had become inherent to the US

regulatory federal system. In other words, it was the moment of NARUC

consolidation;theoccasionthattransformedNARUCfromaloosesetoftiestoan

institution.

Afterthewar,Congressessentiallyabodebythepre-warsystem.Facedwithmany

contrastingpressuresfromdifferentsides, itshiftedtheburdentotheICC,whose

powers were expanded. However, at that timemany ICCmembers were former

NARUC members. Cooperation between the two organizations resumed and

graduallydeepened,untiltheNARUCmanagedtoeradicatetheShreveportDoctrine,

by preventing its appearance in the statutes of subsequent federal agencies. The

weapon PUCswieldedwas the NARUC itself: “once the state commissioners had

persuadednationalregulators,Congress,andthecourtstoacceptthecooperative

approach, they employed the power of their association, NARUC, to solidify their

victorieswithaconcertedmovementto"professionalize"thebusinessofregulation”

(p.704).Thefactthattheoutcomeofthejuncturewascontinuityshouldnotputin

doubtitscriticality.AsCapocciaandKelemen(2007)confirm,acriticaljunctureneed

notnecessarilybringaboutchange.

66

Electricity:theutilities’demandforregulationofthemselves.

Since the early 1900s, the relationships between PUCs and electricity industry

executives evolved in parallel to their relationships with the railroad industry.

However, they were markedly less confrontational. Whereas railroad executives

struggled to eliminate state level regulation of their business, electric utility

executivesengagedintheoppositebattle:theyactivelysoughtstateregulationand

protected it from federal and local encroachment already since 1898, when the

Supreme Court case reviewed the method of rate regulation (Anderson 1980,

Anderson 1981, Hausman and Neufeld 2011). Anderson’s narrative retraces the

inceptionof state regulationofelectricity in thesimultaneousaction, in theearly

1900s, of three types of actors: the utilities, civil society and certain reformer

governors,sharingacommoninterestintomaintainingregulationatlocalleveland

locatingitintoindependentcommissions.

Theutilitiescalledforstateregulationasalternativetostatefranchises,whichwere

subject to corruption, andmunicipal ownership. AsHausman andNeufeld (2011)

note, “state rate regulation did not replace open competition among utilities; it

replacedregulationbymunicipalities”(2011,p.727).Statefranchiseswerecontracts

whereby municipalities allowed the utilities to provide services by placing their

equipmentonmunicipalland.Theyusuallylastedforseveraldecades.Thetermsof

thefranchisesvariedconsiderablyamongmunicipalitiesanddidnotforeseeconstant

oversight.Also,utilitieswerenotnormallyofferedmonopolyprotectionbeyondthe

firstfewyearsofoperation,sothat“entrybycompetingutilitiescouldanddidoccur”

(2011,p.727).However,municipalgovernmentsoftenappropriatedthequasi-rents

generatedbythedifferencebetweentherevenueneededtocovertheutilities’short

run operating costs and the revenue needed tomaintain economic viability and

threatenedtheutilitieswithentryandcompetitionunlessabribewaspaid(Hausman

andNeufeld,2011).Theintroductionofstateregulationgavethosequasi-rentsthe

statusofaconstitutionalrightprotectedbybothstateandfederalcourts.

In1907,NationalElectricLightAssociation (NELA) releasedareport,outliningthe

utilities’ position: some form of public interference in their business appeared

67

inevitable,judgingalsofromtheeventsregardingtherailroads;theywouldrather

have it exerted by a regulatory commission than bymunicipal governments. The

report was extremely timely, as in that same year the first “modern” PUCs (i.e.

possessingactualregulatorypowers)hadbeencreatedinNewYorkandWisconsin.

InWisconsin,aRepublicanpoliticianwhowasalsovice-Presidentofalocalelectric

utility actively lobbied for electricity to be regulatedby PUC. In California,where

supportforelectricityregulationwasfuelledbydiscontentovertherailroads,utilities

wereamongitsearnestsupporters.AsAndersonputit,“thewidespreadhistorical

impressionthatthemostimportantleadersofallpublicutilitiesweredraggedkicking

andscreamingintoasystemofstateregulationissimplywrong”(p.9).Atthetime,

electricitynetworksdidnotextendacrossstateboundaries.Thetypicalutilityserved

asmallareafromasinglegeneratingplantinasinglestate(HausmanandNeufeld,

2011).Hence,PUCshadfulljurisdictionovertheutilitiesandregulationeffectively

guaranteedtheutilitiesprotectionfromcompetition.

ThesecondactorinAnderson’sreconstructionistheNationalCivicFederation(NCF),

agroupof“civicreformersandcorporateliberals”(p.10).TheNCFpublishedareport

in1907stating“publicutilitiesaresoconstitutedthatitisimpossibleforthemtobe

regulatedbycompetition[…]andthequestioniswhethertoregulateortooperate”.

Theirreplytothisquestionwasstateregulation.Theyfearedmunicipalownership

wouldhaveperpetuatedthecorruptionmachineryaroundthegrantingoffranchises.

It was the Progressive Era, and dissatisfaction with politics was at its peak. In

Wisconsin, the governor asked an NCFmember to draft the law extending state

regulationtomunicipalpublicutilities.Eventually,therefore,theNCFreportserved

asoutlineforstatelegislation,inWisconsinandelsewhereintheUSA.

ThelastactorsinAnderson’saccountareseveralgovernors,whousedtheissueof

publiccontrolofprivateenterpriseasaspringboardtolaunchtheirpoliticalcareer

at national level. He focuses on the case ofGovernor C. E. Hughes ofNew York.

Hughes,knownasclose toprivate interests,had ledan investigation intoutilities

ratesandtheinsuranceindustry.Contrarilytoexpectations,heexposedtheabusive

practicesoftheprivatesector.Thismadehimanationalfigure.Assoonaselected

68

governorofNewYork,hecreatedtwoPUCswhosemembershecouldremoveatwill:

oneforNewYorkCity,onefortherestofthestate.Amonthlater,theWisconsin

legislature followed suit. In the following six years most States created PUCs to

regulateelectricandotherutilities.Importantly,Andersonretrievedmostofitsdata

on PUCs relationships from NARUC’s archives, and underlines NARUC’s role in

gatheringregulatorsandprovidingthemwithaplatformforexchangeofexperiences

andexpertise.

DuringtheNewDeal,manynewfederalagencieswerecreated,suchastheFederal

Power Commission (FPC – created in 1920 to regulate hydroelectric projects,

reinforced in 1935 and given jurisdiction over the sale and transportation of

electricityandgas)and theFederalCommunicationsCommission (FCC,created in

1934).BothwereenlistedinNARUC’smembership.NARUCpurposefullyshapedits

relationshipswithallnewfederalagenciesinacooperativefashion.Asamatterof

fact,previousNARUCmemberswereoftenBoardmembersinthesenewagencies.

NARUCrepresentativeshelpeddraftingtheseagencies’statutes,carefultoprevent

theShreveportdoctrinefrombeingenshrinedinthem(Childs,2001).

The Federal Motor Carrier Act of 1935 was particularly important as it

institutionalizedtheICC–NARUCcooperation.TheActnegatedtheapplicationof

theShreveportDoctrinetointrastateratesandprovidedfortheuseofStatejoint

boards,nominatedbyPUCsandappointedbytheICC,todecidemotorcarrierissues

involvingnotmorethanthreeStates(Rodgers,p.31).AccordingtoRodgers,these

were NARUC’s proposals, which ended up in the law. The Act also provided for

NARUCtohaveofficespaceinorclosetotheICCbuildinginWashington.

The tension between federal agencies and state PUCs had become a permanent

feature of American regulatory federalism. However, for a relatively long period

betweentheendoftheSecondWorldWasandthe1970s,theenergysectorenjoyed

a relative calm (Hausman and Neufeld, 2011). As Anderson (1980) outlines, the

industry was growing and reaping the benefits of economies of scale; increased

consumptionwasrewardedwithlowertariffsandrateswereperceivedas“justand

reasonable”.

69

In1965,ProfessorHarryTrebing,aformeracademicturnedregulator,established

theveryfirst“schoolofregulation”intheUSAandtheworld:theInstituteofPublic

Utilities(IPU)headquarteredattheUniversityofMichigan(MillerandSamuels2002).

Commissioners from all over the USA began receiving their training at the same

institution,whichenhancedtheirknowledgeof regulation inboth theoreticaland

practical terms, and further strengthened their bond to NARUC as their “natural

home”(interview39).Tothisday,theIPUisthemainproviderofregulatorytraining

toUSPUCsstaff.Asshallbedetailedinthenextchapter,thecreationofa“schoolof

regulation”wasofprimaryimportanceinthedevelopmentandstrengtheningofties

andcooperationamongEuropeanregulators.

Before the 1970s, Anderson writes, the job of being a state Public Utility

Commissionerwasplacidandattractedformerpoliticianslookingforanoccupation

after retirement, with close links to state governors. The whole approach to

regulationatstatelevelwaspremisedonapactofsortbetweentheutilitiesandthe

statelegislature:utilitieswouldbeaffordedprotectionfromcompetition,provided

theyweresubjecttopubliccontrol.Regulationwouldperformbothprotectionand

controlfunctions.Untilthe1970s,developmentsinthesectorfollowedasmooth,

predictable linear pattern of growth in consumption and in utility revenue. It is

perhapsnotbychancethatthefirstcritiquesofregulationwereformulatedatthis

time,andthattheyfocusedontheregulationoftheelectricitysectoratstatelevel

(StiglerandFriedland1962,Peltzman1971)beforeproducingtheirwidelyinfluential

contributionsonthetopicofregulatorycapture(Stigler1971,Peltzman1976).

This period of calm ended once economies of scale were exploited to the then

possiblemaximum and utilities began facing difficulties in raising capital. The oil

embargoes of the 1970s gave a blow to the industry, by making rates soar and

pushingseveralbigutilitiesonthevergeofbankruptcy(Anderson,1980).Asudden

andmassive increase in rates was deemed inevitable. A conspicuous amount of

litigation ensued, with consumers and environmentalists becoming increasingly

involvedintheregulatoryprocess.Decidingwhatwas“justandreasonable”became

acontestedprocess,requiringskilled,timelyandinformeddecision-making.Public

70

utility commissioners got under the spotlight of the stakeholders, who, in turn,

becameprofessionalizedinhowtheydefendedtheirinterests.Thewholeregulatory

processbecamemuchmorecontested,asdidPUCs’relationshipswiththefederal

agencies.

Theexpansionofthefederalagencies’jurisdiction.

Sinceitwasconferredauthorityoverthesaleandtransportationofelectricityand

gas in 1935, the federal energy agency FERC has seen its jurisdiction constantly

expanding,byallowingitincreasingpolicyrelevanceandcontrol.Theintroductionof

competitivegenerationthroughthePublicUtilityRegulatoryPoliciesAct(PURPA)of

1978 and, in particular, through the Energy Act of 1992 entailed increased

competenciesforFERCasitwastaskedwithdevelopingnewrulesthatwouldopen

access to utility-owned transmission systems allowing the private independent

generatorstocompeteinwholesalemarkets.Thepoliticalsupportforrestructuring

andderegulatingbothgenerationandretailgrewunabatedacrossthestatesduring

the1980sandthe1990s;untiltheCaliforniaelectricitycrisisof2000-2003reversed

the trend (Hausman andNeufeld, 2011). At any rate, over time state regulators’

controlovertheutilitiesintheirstatediminished.Mostrecently,theregulationsof

theEnvironmentalProtectionAgency(EPA) inmattersrelatedtotheenvironment

and climate change (most notably, President Obama’s signature environmental

initiativecalledCleanPowerPlan)alsohavetangibleconsequencesforPUC’spowers

andforelectricityrates.

Theroutetowardscentralizationofregulatorypowerisalsomarkedincaselaw:the

PUCsandNARUC’effortstoexpungetheShreveportdoctrineinthe1920shavenot

preventedtheemergenceofanewdoctrine,whichPUCsperceiveas“givingcontrol

71

tothefeds”(interviews35and40),i.e.theso-calledChevrondeferencedoctrine12

(interview40).Stateregulatorsdefendedtheirpowers in thetelecommunications

sectorwithequalresolve(Wallsten2006).Eventsinthatsectormirrorthoseinthe

electricitysectortoagreatextent.Deregulationandgreaternetworkintegrationin

both the telecommunications and electricity and gas sectors determined a

diminished role of state regulation,which not even theNARUC could be able to

reverse.

It is inthiscontextthattheevolutionofNARUCtookplace. IdiscusstheNARUC’s

strategyofsurvivalasoneof“layering” (Thelen2003): in theearly1990s,NARUC

seizedtheopportunitytoexpandthescopeofitsactivitiesbyincorporatinganew

rationale.Justlikeatitsfoundation,thatopportunitypresenteditselfintheformof

apropositionfromafederalagency.Aschapter6discussesindetail,thenextcritical

juncture inNARUC’s history occurred in 1998,when theUSAID and theUS State

DepartmentinvitedtheNARUCtocooperateintoatechnicalassistanceprogramme

aimed at establishing regulatory authorities in Central and Eastern Europe. The

successofthatprogrammemarkedthebeginningofthecontinuedinvolvementof

theNARUCinregulatorytechnicalassistanceprovisionaroundtheworld,whichhas

risentorepresentthemainsourceofNARUC’sincome.

12“Chevrondeference”principleinadministrativelawholdsthatareasonableinterpretation

of an ambiguous statute by an agencywith subjectmatter jurisdiction prevails. Chevron

U.S.A.,Inc.v.NaturalResourcesDefenseCouncil,Inc.,467U.S.837(1984).

72

3. ThehistoryofCEER:thequestforlegitimacy.

Theemergence,consolidation,growthandevolutionoftheCEERhasalwaysbeen

entirelydrivenbynationalEuropeanregulators.ThissinglefactdifferentiatesCEER

fromall theothercasesexamined in this thesis.Differently fromothernetworks,

whichbenefitedfromfundingfromexternalsources intheir transitionfromloose

communicationtoconsolidatednetworkedorganizations,theCEERhasalwaysbeen

financedbyregulatorsexclusively.ThisfacttestifiedtotheclearintentofEuropean

regulatorstomaintaintheirautonomy,aswellastotheiragencyinshapingtherole

oftheircollaborativenetworkintheEuropeanenergyregulatorypolicyprocess.

Inmostof theUnitedStates, thecreationofPUCspredated theestablishmentof

federalagencies.Thesameoccurred in theEU:after the first regulatory“offices”

were created in the UK (Stern 2014), the institutional format of the regulatory

authority spread to other European countries. Their establishment was only

subsequentlymandatedbyEuropeanlegislation.ThecreationofNationalRegulatory

AuthoritiespredatestheestablishmentofEuropeanAgencies,whichhasoccurred

onlyveryrecently.

The key difference between these two cases is in the distribution of regulatory

authority across levels of governance. In the USA, which is a federal system,

regulatoryauthority isverticallysplitbetweenthefederalandthestate level.The

competenceoffederalagenciesislimitedtointerstateinfrastructureandissues.The

competenceof state level regulators is confined to theboundariesof their state.

However,thisapparentlyneatseparationofpowersdoesnotpreventtheemergence

ofconflictsofauthority,becauseoftheinherentinterdependentnatureoftheobject

of regulation: energy infrastructure is a network industry. Hence, infrastructure

placed at state level feeds into interstate infrastructure. The interdependent,

networkedcharacteroftheinfrastructurehascausedfederalagencies’authorityto

cometoimpingeontheauthorityofstateregulators.

Incontrast,therearenosupranationalregulatoryauthoritiesintheEU.Eachnational

regulatory authority is competent for the infrastructurewithin the boundaries of

73

theircountry.TheEuropeanUnionisnotafederalstate,butapeculiarpoliticaland

economic construction. The allocation of power and authority in the realm of

infrastructuresectorsisnotasclear-cutasintheUScase.Rather,variousinstitutional

entitiesplacedatvariouslevelsofgovernanceholdandcompeteformoreregulatory

authorityonthesector.Regulatorshavebeenvoluntarilynetworkingforadecade

beforetheEuropeanCommissionmandatedthemtocoordinatewithintheEuropean

AgencyfortheCooperationofEnergyRegulators(ACER),which,however,doesnot

haveactualregulatorypowers.ThisresultsfromtheMemberStates’reluctanceto

delegateregulatorypowerstotheEuropeanlevel,particularlysinceinterconnection

issues, and the repartition of the relative costs, are one of the longest standing

mattersofcontroversyintheEuropeanenergypolicy(OlmosCamachoandPérez-

Arriaga2007).Hence,currentlyregulatoryauthorityresidesmostlyatnationallevel.

However, the ACER is managed by the European Commission. Therefore, the

preferencesoftheEuropeanCommissionweighinonACER’sprocesses;regulators

arenotaloneinthedriver’sseat.

TheliteraturehasanalysedCEERempirically:itspolicyrelevance(Vasconcelos2005,

ThatcherandCoen2008,Vasconcelos2009);itsempoweringeffectsonregulators

(Maggetti2013)anditsabilitytosuccessfullybargainitsgradualabsorption(Eberlein

and Newman 2008) into the European institutional framework (Lavrijssen and

Hancher2009)bysecuringacentralroleforBoardofRegulatorsintotheAgencyfor

theCooperationofEnergyRegulators(ACER)(Thatcher2011).However,thereare

nodetailedanalysesonitsemergenceandtherationaledrivingit.Investigationsof

the agency of regulators within it, their motivations to network, and the CEER’s

evolutionaremissingfromtheliterature.

Before the Regulatory State: the role of engineers in promoting

interconnectionsandcrossborderenergytrade.

ThetopicofcrossborderenergytradehasdeeprootsinEuropeanhistory,asshown

by Vincent Lagendijk (2008) in his book on the history of electricity networks in

74

Europe.AlthoughoftenlinkedtotheintegrationeffortsoftheEuropeanCommission

and its tendency to self-aggrandizement, energy markets integration across the

Europeancontinenthasbeenalongstandingconcernof(initially)privateutilitiesand,

in particular, of the technicians and engineers that worked in the power sector.

Langedijk’shistoricalreconstructionunderlinesthekeyroleengineersplayedinthe

electrificationofWesternEuropeandinshapingthedebateofthetimeconcerning

thedesirabilityofinterconnectionacrossEuropeancountries.

Engineers maintained a central role in the European power sector also in the

aftermathofWorldWarII.Thefollowingextendedquotationsportraytheencounter

betweenUSandEuropeanengineersinthecontextoftheMarshallplan.

“In April, 1949, a group of European engineers was welcomed by their

Americanhosts.ThevisitorsfromEurope,mostofthemsystemoperatorsin

their respective countries, flew across the Atlantic to see firsthand the

Americanstate-of-the-artintheelectricityindustry.ThisTechnicalAssistance

(TECAID)Missionwasanintegralelementoftheelectricityprogramssetup

withintheframeworkoftheEuropeanRecoveryProgram(ERP),alsoknown

astheMarshallPlan.TheoverallintentionoftheERPwithregardtoelectricity

was to expand generation capacity, by building national and international

powerplantson theonehand,andmakingbetteruseofnewandexisting

capacitybycreatingEuropeanpowerpoolsontheother.Thesepowerpools,

should be brought about by building both physical and institutional

interconnectionsbetweencountries.”(p.107)

“Asanoutcomeoftheseefforts,organizationsrepresentingregionalpower

pools eventually became the face of European cooperation. (…) European

engineers clearly had differing opinions from American ERP officials, who

arguedforinternational–andevensupranational–ownershipandoperation

of power plants and networks. The ideas of Western European engineers

showed remnants of interwar plan, in stressing the solidarity effects of a

Europeannetwork.Atthesametime,theirproposedwayofcreatingsucha

networkwasrathersimilartotheconsensusthatemergedinthecourseofthe

75

1930sthatnationalandinternationalinterconnectionsshoulddevelopside-

by-side. To them,aEuropean systemshould consistofnationallyoperated

networks,workinginclosecoordination.”(p.108-109)

Producer organizations were the main promoters of cross border electricity

infrastructure integration in Europe. The UNIPEDE (Union Internationale des

Producteurs et Distributeurs d’Énergie Électrique i.e. International Union of the

Producers and Distributors of Electric Energy), set up in 1925, was the platform

wheretheideasofengineersregardinginterconnectionofEuropeanelectricitygrids

were first voiced. TheUCPTE (Unionpour la Coordinationde la Production et du

Transport de l’Électricité - the Union for the Coordination of Production and

TransportationofElectricity)aninformal,regionalnetworkofutilityrepresentatives

from eightWestern European countries established in 1951 that set up the first

mechanismsof regular communication amongutilities13. Bothorganizationswere

very influential in framingthedebateon infrastructure integrationacrossEurope.

They evolved into what today are the Eurelectric (the association representing

EuropeanelectricutilitiesinBrussels)andtheENTSO-E(theassociationofEuropean

systemoperators).

Lagendijk’s historical narrative compellingly attests to the relevance of technical

expertstotheconstitutionofasupranationalEuropeanorder,linkinginnotonlywith

well-establishedresearchonthecontributionandimpactoftransnationalnetworks

ofexpertsonnationalandtransnationalpolicy-making(Haas1975,Haas,Williamset

al.1977,Haas1992,DjelicandSahlin-Andersson2006,Ambrusandetal.2014)but

alsowithcontributionsfromthesocialandgeographicaltheoryunderscoringhow

technological interconnection and integration shaped the conceptualization of

13Similargroupingswerecreatedovertheyears,e.g.onegatheringScandinavianutilities,

another gathering utilities’ representatives from Spain, Portugal and France, another for

utilitiesfromItaly,AustriaandYugoslavia.CentralandEasternEuropeancountriesremained

relatively isolated from developments in Western Europe. Gradually, however, they

establishedsimilarstructuresofcooperationamongtheirutilities’representatives.

76

Europeanintegration(Barry1996),andstressingtheroleofexperts,engineersand

scientistsasphysicalcarriersofpolicy-impactingknowledge(LarnerandLaurie2010).

Moreover, itoutlinesthecorefeaturesoftheEuropeanenergy infrastructureand

marketintegrationpolicy:thetensionbetweennationalandsupranationalauthority

andapreference forexpertcoordinationrather thanthecreationof fully-fledged

supranationalinstitutions.Onthisbackground,theemergenceandevolutionofCEER

findtheirplace.IhaveinvestigatedthehistoryofCEERbywayofface-to-faceelite

interviewswith key individualswhowere directly involved into its establishment

and/orwho previously or currently work in its ranks. In the following sections, I

retracetheeventsleadingtoitsemergence,itsconsolidation,itsreproductionover

timeasinfluentialactorontheEuropeanenergyregulatorypolicy,anditsongoing

evolutionarypath.

TheemergenceoftheCEER:networkingforlegitimacy.

“Itisavery,veryinterestingstory…I’llgiveyouanoutlineofthekey

stages… frommypointof view,of course. I’ll tell youmypersonalhistory,

althoughIbelieveitissimilartothatofmanyothers”.(interview1)

InFebruary1997,representativesofthefewthenexistingenergyNRAs(fromItaly,

Spain,Portugal,Sweden,Norway,andtheUK)metforthefirsttimeataconference

onelectricitymarketrestructuring14organizedbytheEuropeanUnionandtheWorld

Bank.ThelawsestablishingSouthernEuropeanNRAshadbeenpassedin1995(Italy

andPortugal)and1994(Spain).Theirrespectivegovernmentshadfollowedinthe

steps of the United Kingdom, which had undertaken the utility privatisation and

liberalization process in the 1980s and had established independent energy

regulatory authorities in 1989.As for the Scandinavian countries, at the time the

14SecondWorldConferenceonRestructuringandRegulationoftheElectricityMarket,3-5

February1997,Vasteras(Sweden).

77

liberalizationofelectricitymarketswaswellunderway,withNorwaypioneeringthe

processin1992.ThefirstcomprehensiveEuropeanlegislationonenergymarketswas

releasedin1996.Theso-called“firstElectricityDirective”mandatedtheendofthe

energymonopoliesandsetthegoalofachievingthecreationofaEuropeanInternal

EnergyMarket(IEM).However,itdidnotmentionregulationorregulatorsatall.The

establishment of independent regulatory authorities depended exclusively on

nationalgovernments’initiative.

The head of the British electricity regulatory authority (called Office of Energy

Regulation, orOFFER)waspresenting at the conference.At the time,OFFERwas

perceivedasbeingtheparadigmaticNRA.ThenewlyestablishedSouthernEuropean

regulatorswereeager to learn from its experience (interviews1, 2, 3, 4, 13, 14).

Scandinavian regulators, also in attendance, had a long-standing history of

collaborationalready.Inthosecountries,governmentalregulatorydepartmentsfor

infrastructure sectors were created in the early 1900s. Differently from their

SouthernEuropeanandBritishcounterparts,whichwereconstitutedasIndependent

RegulatoryAuthorities(IRAs),Scandinavianregulatoryauthoritieshaddevelopedout

of former ministerial departments. The newly established Southern European

regulators were more interested in learning from the British experience of

independentregulation.However,Britishregulatorsobservedwithmoreinterestthe

massiveprivatizationprogrammesthenongoinginSouthAmericaandAustraliathan

theembryonicopeningofsouthernEuropeanmarkets.

United in their quest for benchmarks, and given the strong similarities in their

national markets, the three Southern European regulators agreed to start

communicatingon a regular basis to exchange information about the issues they

faced.Theystartedmeetingquarterly,onceineachcountry.Theirfirstjointmeeting

tookplace amonthafter theVasteras conference, inMarch1997, in Lisbon. The

secondmeetingwasheldinSanSebastian(Spain)inJune1997.Theirthirdmeeting

tookplaceinMilaninDecember1997.Theyestablishedthreeworkinggroups,each

chairedbyoneofthem.Theychargedtheirstaffwithmaintainingregularcontact

withtheothertworegulatorsin-betweenmeetings.Thefirsttopicstheyaddressed

78

were the new regulatory framework, in particular the development of incentive

regulation,andregulatory independencefromgovernmentand incumbentenergy

companies.Inthisway,thethreeSouthernEuropeanregulatorslaythefoundation

ofwhatwouldbecometheCEER.

TheEuropeanCommissionco-optstheregulators.

Atthetime,theCommissionwasseverelyunder-resourced,especiallyinsectorsfor

which itdidnothavewell-definedcompetence, suchasenergypolicy.Therefore,

officialsintheDirectorateGeneralforEnergyhadonlylimitedknowledgeofnational

energymarketsandactors.Theironlyprovidersofinformationatthetimewerethe

nationalenergyincumbents,whichhadclearincentivestoprovideabiasedpicture.

For these reasons, EuropeanCommission officials began reaching out to national

regulators.

In the course of 1997, the then Director of the EC Directorate General (DG) on

TransportandEnergy(TREN,nowDGENER)learntabouttheongoingmeetingsof

thethreeSouthernEuropeanregulators.Theirdevelopingcollaborationinspiredhim

toconceiveofamulti-stakeholderForum,whosetaskwouldbetofosterdialogue

betweenstakeholdersonenergymarketreforms,andencourageprogresstowards

the creation of the Internal EnergyMarket (IEM). He envisioned it as a regularly

scheduled gathering of “those actors who, like itself, felt the need for market

integration” (interview 3), i.e. regulators, operators, and other energy market

stakeholders.Itwasalsoamechanismtobringtransparencyintoasectorthat,atthe

time,wasimpervioustotheoutsideobserver.

TheEuropeanCommissionhadrealizedthatmandatingtheabolitionofthestate-

ownedutilitymonopolies,asthe1996FirstElectricityDirectivehaddone,wasnot

enoughtostimulatecrossborderenergytrade.However,itwastooearlytobegin

negotiating a second piece of legislation. As a matter of fact, Member States

79

representativeswerestillnegotiatingthefirstGasDirectiveasthesediscussionswere

takingplace.

The Director General’s idea resonated with sector stakeholders and rapidly

materialized intowhatwascalled theEuropeanElectricityRegulatoryForum.The

firstmeetingoftheForumwasheldinFlorence(Italy)inFebruary1998.Henceforth

it was commonly referred to as “the Florence Forum”. Florence appeared as a

suitablyneutrallocationbecauseithostedaEuropeanapoliticalinstitution,i.e.the

European University Institute (EUI15). The following year, the European Gas

RegulatoryForum,or“MadridForum”,wassetup,focusedonthegassector.Nowa

consolidated, taken-for-granted event concerning EU electricity markets and

regulatorypolicymatters,atthetimetheFlorenceForumrepresentedaveritable

revolution in the governance of the energy sector and a bold initiative of the

Commission.Theideaofbringingtogetherregulators16,operators,utilitycompanies

andEuropeanCommissionofficialswasliterallyunheardof.Attendeesofthevery

firstmeetingoftheFlorenceforum

“…included senior representatives of national regulators or ministries

responsible for electricity regulation, the EU Director General for Energy,

Pablo Benavides, European Commission officials (DGs XVII and IV),

representatives of the electricity industry and of major consumers. All EU

MemberStateswererepresentedaswellasUnitesStates,NewZealandand

Norway.Themainareasaddressedbytheforumcoveredtransmissionpricing

methods and cost accounting, (…) non-discrimination and unbundling, and

treatment of public service obligations and environmental ‘costs’ in a pro-

competitiveenvironment.(…)MrBenavidesemphasisedthattheobjectiveis

notsimplytheliberalisationof15nationalsystems,butalsotheestablishment

15TheEUIistheonlyEuropeanuniversity,whoseexistenceisbasedonalegalagreement

between the EU and 22 Member States, initially signed in 1970s1972,

https://www.eui.eu/About/HistoryofEUI.16BecausemostMemberStateshadnotestablishedaregulatoryauthorityyet,theirenergy

Ministerswereinvitedinstead.

80

ofanInternalMarketinelectricity:i.e.notjustliberalisationbutalso"internal

marketisation".Thismeansthattheimperativeofsubsidiarityco-existswith

theneedforacertaindegreeofhomogeneousinterpretationonthepartof

theMemberStates.”(FirstFlorenceForumMeeting1998-minutes)

Besides presentations from the European Commission, key Transmission System

Operators (TSOs – from the UK, Sweden, Germany, Spain) and European and

Americanconsultants,aroundtableofregulatorsalsotookplaceduringthemeeting.

Participants included regulators from Italy, Spain, Portugal, Sweden, Norway,

Finland,Denmark,theUK,andtheUS.Allregulatorsrecognizedtheneedtoallowfor

national diversity and to increase transparency. However, clear differences in

approachemergedinthediscussion.Accordingtotheminutes,whilethePortuguese

regulatormade the clearest calls for harmonized regulatory approaches, German

industrialconsumersexpressedsatisfactionwiththehands-offapproachinplacein

theircountryandclearlyauspicatedthat“strongerregulationwillnotbecalledfor”.

ThehopesofGermanindustrialconsumersstoodtobedisappointed:atthesecond

meetingoftheForum,whichtookplacesixmonthslater,thetopicoftheregulation

of TSO took centre stage together with key aspect of market opening, such as

unbundling.

ThefirstmeetingofthecorrespondinggasForum,heldinMadridforthefirsttime

andhenceforthrenamed“theMadridForum”tookplaceinthesameyear–1998.

The minutes of that first meeting suggest a cautious approach concerning the

integrationofEuropeangasmarkets.

“ThefundamentalobjectiveoftheForumistodevelopconsensusamongstall

thepartiesinvolvedinthisprocess;governments,regulatorsandindustry,in

ordertomakerapidprogressonthesehighlytechnicalissues.Thismechanism

actsasacomplementtoharmonisationmeasuresthatmaybenecessaryto

achievethebasicgoalofaneffectivelyfunctioningsinglemarket.”

Similartoitselectricitycounterpart,themeetingwasattendedbyofficialsfromthe

EuropeanCommission,nationalregulatoryauthoritiesandEUEnergyMinistries,as

81

wellas representatives from internationalorganisations,suchas the International

EnergyAgency,andassociationsrepresentingthegasindustryandgasandelectricity

consumers.WhilethemeetingsoftheFlorenceForumtookplaceeverysixmonths,

the second meeting of the Madrid Forum took place two years after the first,

testifyingtotheconsiderablereluctanceofMemberStatestoopeningsofthegas

sector. Nevertheless, theminutes from the secondmeeting show a considerably

strongerresolve:

“TherepresentativesoftheEuropeanCommission,theMemberStates,and

theEuropeanCouncilofEnergyRegulators,invitetheEuropeangasindustry

to establish, as quickly as possible, a new body or grouping that brings

together representatives of all those responsible for the operation of the

transmissionnetworkforgas inEurope.(…)Theobjectiveofthisbody isto

work, inter-alia in the context of the Madrid Forum, together with the

Commission, Member States and national regulators, to resolve issues of

mutual concernwith respect to the development of a competitive internal

European gas market, and to provide technical data regarding the

transmission systems within Europe at the request of the Commission,

nationalRegulators,andtheMemberStates.”

Thesecondmeetingof theFlorenceForum in1999hadsimilarlyput forward the

necessityforTSOstoformanassociationtodialoguewiththeCommissionandthe

association of the regulators. The Florence and Madrid Fora became important

appointmentsandboostedtheworkingrelationshipbetweennationalregulatorsand

theEuropeanCommission.

It is important to note that, at this stage, the CEER did not officially exist yet.

Regulators would sign the Memorandum of Understanding officialising their

cooperationonlyin2000.Underthesecompletelyinformalandloosearrangements,

EuropeanregulatorsemergedfromtheFlorenceandMadridForaaskeyplayersin

theformulationoftheregulatoryframework.

82

Thenecessityofhomogenousinterpretation,andimplementationoftheregulation

ofenergymarketsacrosstheMemberStateshasalwaysbeentheleitmotivofthe

EuropeanCommission’spolicyinitiativesinthissector.Theliteraturehasunderstood

theCommission’sinsistenceonregulatoryconvergenceandharmonizationasaself-

aggrandizing strategy (Schmidt 1998, Eberlein 2008,Mayer 2008, Diathesopoulos

2010, Torriti 2010, Maltby 2013, Goldthau and Sitter 2014, Herweg 2015): by

transferringtheformulationofenergyregulatorypolicytotheEuropeanlevel,the

European Commission inherently transferredmore power to itself. As previously

mentioned, the vast majority of scholarly contributions on European regulatory

networks stem from consideration of the European Commission’s integrationist

pushesinenergypolicy.Earlyliteratureonnetworks,inparticular,hasassumedthat

thepreferencesoftheEuropeanCommissionandtheregulatorscoincided.Ithasnot

investigatedwhetherthatis,indeed,thecase,andwhy.Thissectionsuggests,and

following sections confirm, that the regulators, similarly to the Commission,

endorsed the policy goal of Europeanizing energy regulation because thatwould

enhancetheirinstitutionalrole.

The consolidation of the CEER: the tensions of multi-level

governance.

“Inthebeginningmostoftheworkwasnational.Therewassomuch

to do nationally (…) the CEER was a club, it was interesting to go there

becauseyoumetcolleagues,onanationallevelyoudidnothavecolleagues

(…)youhadnobodytotalktoandfindout“Ohthisisausualproblemormy

issues are totally different from everybody else’s? Am I doing something

wrong?(…)Inthemeantime,theCEERcontinuedtodevelopandwefound

thatasagroupwehavesomeinfluenceatEUlevel,soit’snotjustexchanging

bestpracticesandlearningfromeachother,whichisalsoacomponent,but

wealsohaveanimpactonwhathappensnext”(interview28).

83

“Let’sputitthisway,ifmygoalwouldbeto…implementortoachieve

verynationalistic,specificsolutions,probablyitwouldnotbesuccessfultodo

it at EU level. But if the goal is to develop solutions thatmake sense in a

generaleconomicwayinaEuropeanperspectivethenyou’remuchbetteroff

totrytopushthatatEUlevel”(interview25).

TheCEERhadbegunasinformalcontactsbetweenahandfulofregulators,meeting

occasionallyinEuropeancapitals.WiththeexceptionsofFrance(thatdidsoin2003)

andGermany(in2005),allEU-15MemberStatesestablishedenergyNRAsquickly

thereafter.Thegroupofregulatorsexpandedinsize:newlyestablishedauthorities

didnothesitatetoreachouttotheformingnetworkofEuropeanregulatorsandto

establishtieswithit.Inordertotakestockofthestateofenergyregulationacross

theMemberStates,in1999,theItalianregulatoryauthorityorganizedaseminaron

“Criteriaforelectricitytariffsandpricing”,heldinRome(Italy).AllotherEuropean

NRAs were invited to take stock of their experiences. Representatives from 18

institutions attended and gave a presentation, including NRAs from Italy, Spain,

Portugal, the Netherlands, Finland, Ireland, Belgium, the UK, Northern Ireland,

Sweden,Norway,Hungary,Poland,Romania;MinistersfromFrance,Germanyand

Switzerland,andanobserverfromtheEuropeanCommission(AEEGSI,AnnualReport

2000,p.291;CNEdatabase17).

Theexchangesprecedingandfollowingthismeetingshowedthat,intheirdifferent

nationalcontexts,regulatorswerefacingverysimilarchallenges:primarily,domestic

opposition from industry and government as prices increased as a result of cost-

reflective tariffs, and government subsidies were phased out. Amonth after the

meeting,inDecember1999,duringthefourthmeetingoftheFlorenceForum,the

three regulators publicly expressed the intention to establish an association of

17CNEOnlineDatabasehttp://www.cne.es/cgi-

bin/BRSCGI.EXE?CMD=VERDOC&BASE=TODO&DOCR=14&SORT=-

FECH&RNG=20&SEPARADOR=&&desc-c=+DISE%D1O+Y+ESTRUCTURA+DE+PRECIOS+Y+TARIFAS

84

European energy regulators. Just like European Energy Ministers met at the

EuropeanCouncilinBrussels,regulators“wouldhavetheirowncouncil”(interview

1), aswell. Hence, they decided to call their association the Council of European

EnergyRegulators(CEER).

Notonlyhadtheregulatorsfoundthattheyfacedsimilarchallenges;theyalsohad

understoodthattheprocessofreformingnationalenergymarketsandintegrating

themwouldtakealongtime,andthatoppositiontotheirdecision-makingactivity

would not dissipate: the activity of regulation consists of imposing costs on (and

creating benefits for) differentmarket actors as a result ofmarket design.More

broadly,justliketheirUScounterpartshadexperiencedacenturyearlier,European

regulators realized that the tensions and contrasts inherent in thedistributionof

regulatoryauthorityacrosslevelsofgovernanceentailedissuesofregulatorycontrol.

Theywould always face the threat of having their powers or authority curtailed,

whether“frombelow”(i.e.bynationalgovernments)or,astheysoonrealized,“from

above”(e.g.bytheEuropeanCommission).Therefore,theyneededamechanismof

collective representation to address the supra-national level and to refer to at

domesticlevel.

Intheearly2000s,theregulatorsunderstoodthatestablishinganassociationwould

haveallowedthe“regulatoryposition”tocomeacrossdirectlytotheCommission

(interview4).ThisopportunitychangedtheattitudeoftheUKregulatoryauthority

(OFGEM), which had initially been lukewarm towards the idea of regularly

cooperating with their European counterparts. Therefore, in January 2000, the

representatives of 10 regulatory authorities met at OFGEM offices in London to

discusstheformandthegoalsoftheirassociation.On7March2000,representatives

from Belgium, Finland, the UK, Ireland, Northern Ireland, Italy, Norway, the

Netherlands,Portugal,SpainandSwedenmetinBrusselstosigntheMemorandum

of Understanding (MoU) establishing the Council of European Energy Regulators

(CEER). Several interviewees recalled theendorsementof theCEERbyOFGEMas

decisive for the establishment of the association. The British model of energy

regulationwas so influential in the EU and beyond thatmany felt an association

85

withouttheBritishNRAwouldhavebeensomehowlesscredible(interview3).The

MoUwasnon-bindingandmerelydeclaredtheintentionofregulatorstocoordinate

in the interest of achieving the IEM. The formalizationof regulators’ cooperation

occurred three years later, in response to the first attempt of the European

Commission toappropriate thespaceof cooperation that regulatorshadbuilt for

themselves.

ERGEGandtheofficialisedCEER:cherishingtheNRAs’self-rule.

In the year 2000, the European Council launched a policy programme called the

Lisbon Agenda. Among other things, it encouraged the full liberalization of the

nationalenergymarketsandtheachievementoftheIEM.Tothatend,theCouncil

askedtheCommissiontopreparelegislativeproposals.Intheirturn,theEuropean

Commissionsoughttheregulators’views.Atthispoint,CEERmembersfeltthattheir

network could no longer remain an informal association of professionals: the

constant interaction with the Commission required resources and updated

knowledgeofthelatestpolicyproposals.Moreover,providingpolicyinputrequired

closeness to the centre of policy formulation. This was difficult to achieve from

nationalcapitals.

Until thatpoint, the financial viabilityof theCEERhadbeenassuredbyvoluntary

contributions:regulatorsdidnothavefixeddues;theycoveredtheirCEER-related

expenses, which mainly consisted in travelling to meetings around Europe,

individuallyandautonomously.Theopportunitytocontributetothedraftingofthe

EuropeanCommissions’ legislativeproposals,however,representedarathermore

consequential commitment. Regulators decide to provideCEERwith structure, as

wellasheadquarters inBrussels, secretariatstaff,aworkplan,andsetbudgetary

contributions.

Atthesametime,theEuropeanCommissionneededtoformalizeitsrelationshipwith

theregulators.Asoneintervieweepointedout,theECcouldnotcontinuetoexplicitly

86

andextensivelyrelyonaprivateassociationofindividualprofessionalstocarryout

importantmarketreforms(interview1).“Formalizing”theCommissions’relationship

totheregulatorswasnecessarytoensuretransparencyandaccountability.Atthe

sametime,regulatorsfearedthattheCommissionwouldhaveimposeditsdecisions,

shouldithavetheopportunity–theydecidedtoretaintheirowninformalnetwork

toretaintheirindependencefrombothoftheirprincipals(Egeberg2006):the“old”

one(i.e.nationalgovernments)andthe“new”one,i.e.theEuropeanCommission.

Hence,theCommissiongaveleewaytotheregulatorstoproposeamodelfortheir

coordinateddialogue.Theregulators’proposalswereveryambitious.Besidesmarket

designrules,inthefirstdraftofwhatthenbecamethesecondEnergyPackage,there

wasthecreationofaEuropeanbodyofregulatorswithampleregulatorypowers,“on

themodel of the American FERC [Federal Energy Regulatory Commission]; a true

European regulator, with real powers, so that you could move from national to

Europeanregulatoryagency”(interview4).Therefore,someregulatorsenvisioneda

regulator-only European agency, similar to the USA FERC, and their careers as

ascending from the national to the supranational level. The EC Legal Service,

however, in application of the Meroni doctrine18, restrained the scope of the

foreseenregulators’bodytonomorethanconsultativepowers.

18 The Meroni doctrine arose from cases C-9/56 and C-10/56 (Meroni v High Authority

[1957/1958]ECR133)andrelatestotheextenttowhichEUinstitutionsmaydelegatetheir

taskstoregulatoryagencies.TheEuropeanCourtofJusticerulesthatit"cannotbeexcluded"

thatpowermightbedelegated tobodieswhoseexistencewasnot contemplatedby the

Treaties, if doing so appears compatible with the regulatory powers conferred on the

institutions (in thatcase, theHighAuthorityof theCoalandSteelCommunity).However,

suchdelegationispermissibleonlywhen"itinvolvesclearlydefinedexecutivepowersthe

exerciseofwhichcan,therefore,besubjecttostrictreviewinthelightofcriteriadetermined

bythedelegatingauthority",whereasdelegationof"adiscretionarypower,implyingawide

margin of discretion" is to be excluded in all cases.

http://www.jeanmonnetprogram.org/archive/papers/02/020201-03.htmlThisistorespect

the institutional balance of powers of the European institutions,which is a fundamental

87

InNovember2003,aCommissiondecisioncreatedtheEuropeanRegulatoryNetwork

(ERN)ofenergyregulators:theEuropeanRegulatorsGroupforElectricityandGas

(ERGEG).Afewdaysearlier,theCEERhadbeenregisteredasanoprofitassociation

underBelgianlaw.

“The Commission wanted to consolidate the CEER as advisory

organtoitself.Itwastheregulatorswhosaid“Well,wedon’twanttobe

part of something that can only be convened by the Commission…we

want to be able to convene meetings and talk about our things, for

instancetraining,exchangesofinformation,ofhelp….Wehaveaseriesof

thingsthatinterestussowedon’tdissolvetheCEER”.Andsotherewere

theCEERandtheERGEG.”(interview1)

Astheabovesuggests,theactualprocessleadingtotheestablishmentoftheERGEG

followedaratherinformal,under-the-radarpath.Iftheformalaspectsofthisprocess

maybedescribedasconsistingofa“doubleroundofdelegation”(CoenandThatcher

2008),itsinformalcoreconsistedofregulators’acceptancesofclosermonitoringby

andcollaborationwiththeEuropeanCommission,providedthenetwork’sadvisory

roleremainedessentiallythesame.CoenandThatcher(2008)donotfocusonthe

energyregulators’network,becauseofitsweakpowersandformalautonomy.This

statement reinforces the importance of investigating the informal dimension of

regulatorynetworks,since,likeinthiscase,itmaybethedominantone.

TheCEERandtheERGEGcoexistedunderthesameroofbetween2003and2011.

These arrangements suited the circumstances – the Commission could barely

devolveanyresourcestotheERGEG;theexistingorganisationalinfrastructureofthe

CEER supported it entirely. The creation of the ERGEG faced regulators with the

option of terminating the CEER. SeveralNRAswere in favour of this solution, on

account that it would have reduced the expenses on travelling to Brussels for

guarantee afforded by the Treaty to European citizens. http://eur-

lex.europa.eu/summary/glossary/institutional_balance.html

88

meetings(interviews4,15,19).Nevertheless,themajorityofCEERmembersvoted

duringtheirGeneralAssemblyinfavourofretainingtheCEERastheregulators’own

association, independentofanyotherinstitution.Someregulatorsfearedthatthe

EuropeanCommissionwas“tooideological”(interviews4,5)andtooeagertospeed

upthepaceofmarketintegration,regardlessofnationalcircumstances.

The training of regulators: the establishment of the Florence School of

Regulation.

Since1997,Europeanenergyregulatorshadbeenregularlyexchanginginformation.

In the early 2000s, they reckoned that the body of knowledge that they were

developingonthefieldwouldbeusefultocurrentandfutureEuropeanregulators.

OneoftheSpanishCommissionersvolunteeredtoorganizeacourseonelectricity

tariffs.Hehadgainedconsiderableexperienceinprovidingregulatorytrainingwhile

workingasconsultantforLatinAmericangovernmentsandNRAsinthemid-1990s.

Thefirsttrainingcourse,attendedonavoluntarybasisbyonlyafewCEERmembers,

was held in March 2002 in a hotel conference room in Palma deMallorca. The

locationwaschosentokeepparticipationcostsataminimumthankstolowseason

hotelprices,giventhatmostoftheparticipantsfinancedtheirownparticipationout

of their own pocket (interview 1). The course was not labelled a “CEER training

course”.Notallmemberswouldbereadytoendorsesuchastatementatthatpoint

(interview1).ThecoursewasattendedbyallEUregulatorsaswellasbytheheadof

the Hungarian regulatory authority (interview 3), which had been established in

1994.HungarywasnotanEUMemberStateatthattime,butinteractionsbetween

EU and Eastern European regulators were becoming more frequent in the early

2000s,aswillberecalledinchapter6.

ThesuccessofthisfirsttrainingprogrammepromptedEuropeanregulators,astheir

North American counterparts had done in the 1960s, to set up a “school of

regulation”.TheydecidedtotesttheideabylocatingthefutureschoolofEuropean

energyregulatorswithinanexistinginstitution.ThechoicefellupontheEuropean

89

UniversityInstitute(EUI)basedinFlorence(Italy)–ithadhostedthefirstFlorence

Forummeetings and still appearedas a suitablyneutral location. Supportedby a

communicationtotheEUIpresidentfromtheECvice-PresidentLoyoladePalacio,

regulatorsarrangedforasecondtrainingcoursetobeheldattheEUI.Thecourse

took place in October 2003 and represented the foundational moment for the

FlorenceSchoolofRegulation(FSR),whichwasofficiallycreatedin2004.Thehead

oftheItalianNRA,aprofessorwhohadcreatedtheCEERtogetherwithitsSpanish

and Portuguese counterparts, became its first Director upon finishing his NRA

mandate.

TheCEERasapolicylever:policyinfluenceandinstitutionalgoals.

“…A strong unbundling (…) would not have found a majority in

parliament (…),even if theministryofeconomyhadsupported that (…) so

earlyonIsaidit’snotworthtrytolobbynationallybecausewewillfailbutat

EU level it’sadifferentstory…sowedidsucceedtogetquiteanumberof

thoseprovisions into the3rdpackage (…) like it’sa separatebranding, you

know,itwouldbetotallyimpossibletogetthisinXXXXXbutinBrusselsitwas

notsuchabigissuebecauseanumberofcountrieshadalreadydoneitsothey

felt,yeah,it’sagoodidea,let’sdoit,andthosecountriesthatwerestrongly

opposedwerefew,(…),buttheydidnothavethemajorityinthecomitology

process therefore we got some rules we would never have gotten on a

nationallevel”(interview25).

“Agoodexampleisthethirdenergypackagewherewithluckatthat

momentintimewewereabletoinfluencequiteabittheoutcome,sointhe

thirdenergypackagetherearemanyprovisionsthatwehavedrafted…and

atthattimewehadextremelygoodcooperationwiththeParliamentsowe

wereabletobringinmanyamendmentstotheParliamentandstrengthenthe

roleofregulators,theindependenceofregulators,strengthentheunbundling

90

rules, things thatwewouldnothavebeenable todoonapurelynational

basis”(interview15).

Preparatorywork for the Third Energy Package represented a keyopportunity to

further market integration. Regulators used the opportunity to shape the

institutional design of their nationalmarkets by taking part in the formulationof

bindingEuropeanlegislation.ThesecondEnergyPackagehadrepresentedthefirst

occasion for regulators to influence European legislation on energy markets.

However,itresultedintomuchlessambitiouslegislationthanregulatorswouldhave

preferred.Eventually,itmandatedregulatoryindependencefromindustry,butnot

fromgovernment,andprovidedforminimalmonitoringandadvisorypowers.

Furthermore,intheearly2000stheEuropeanCommissionhadrealizedthatmarket

integrationwas not progressing. In 2004, the Europeanenlargement brought ten

newMemberStatesintotheEUandtennewNRAsintotheCEERandtheERGEG.The

scopeoftheIEMhadthusgreatlyexpandedandcametoencompassjurisdictionsat

differentstagesofinstitutionalandmarketdevelopment.Asthepolicyrelevanceof

thecoordinationtakingplacewithintheCEER/ERGEGkeptgrowing,andthepushfor

market integration and liberalization of retail markets grew stronger, these

differencescametobeseenashamperingtheeffectivenessofsuchcoordination.

Specifically, some regulators could notmake commitments to their counterparts,

becauseveryoftentheyeither lackedjurisdictionovercrossborder issuesorthey

shared it with the government. Until the second half of the early 2000s, most

regulatorsstillneededgovernmentapprovalofmostoftheirdecisions.

However,noentityexisted,whichcouldimposeitsdecisionsonanyoftheactorsof

the EU energy regulatory space. The regulators, together with the Commission,

craftedan interimsolutionbypromoting, in2006, thedivisionofnationalenergy

marketsintoregionalblocks,whichwouldintegratetheirmarketsfirst.Theseblocks,

calledRegionalInitiatives,becamethebuildingblocksoftheInternalEnergyMarket

(IEM).“Therewasnograndschemebehindit.Itwasmorelike…whatthehelldowe

do!”(interview25).Hence,regulatorscametoseetheestablishmentofaEuropean

91

Agencyasnecessarytoforcemarketintegrationandasanopportunitytoenhance

theirdecision-makingauthority(Thatcher2011).

“I said we need to keep our autonomy from everyone. I cannot

graduallyloseautonomyasIgetclosertoBrussels.SoIwantedtheAgencyto

bearegulators’agency.Autonomous,independent,thathadtobeveryclear.

Iwantedittobefinanced,likenationalonesare,throughlevies.(…)Atthat

time,theParliament listenedtous.They listenedtousandeventuallythat

hybridresulted,withatleastastrongboardfortheregulators”(interview4).

TheEuropeanCommission,however,sawitselfinthedrivingseatoftheAgency.This

competition led to a hybrid agency, where the regulators retain considerable

authoritybutundertheaegisoftheCommission.AlthoughfallingshortofEuro-wide

regulatorycompetences,theACERgainedexclusiveauthorityovertheformulation

oftherulesthatwouldensuretheviabilityofanintegratedEUenergymarket–the

networkcodes(Jevnaker2015).TheentryintooperationoftheACERin2011ended

the path-dependent progression of the CEER and triggered the third and current

phaseofitsexistence:re-invention.

Re-inventingtheCEER:layeringorconversion?

“TheCEERhasbeenincrediblysuccessfulbutitsinitialroleisfinished,it’sgone,

prettymuch,it’stransferredtoACER.SonowCEERisstillworkingbutIthinkit’s

got a philosophical purpose, which is to preserve the interest of independent

energyregulators.(…)Itcanalsolobby,ofcourse…lobbyisthewrongword…it

canalsotrytoinfluencethepoliticalestablishmentinEuropeinwayswhichACER

can’t,really,becausetechnicallyACERiskindofanarmoftheECsoitwouldbe

wrong for it to lobby fornewpolicy initiativesor lobbypublicly inwayswhich

disagreewithwheretheECiscomingfrombecauseitwouldbeliketheECarguing

92

withitself,butCEERcandothesethingsandthatIthinkistherealexampleofthe

independentvoiceofregulators”(interview14).

VirtuallyallintervieweesacknowledgedthatthecreationoftheAgencyhaspartially

deprivedtheCEERofitsinstitutionalandpolicyrelevance.BycreatingtheACER,the

European Commission institutionalized regulatory coordination, rendering the

informalspaceofinteractionitpreviouslyhadwiththeregulatorsredundant.

IntheirdiscussionoftheinstitutionalizationofthenetworkofEuropeancivilaviation

authorities into a European Agency, Pierre and Peters (2009) conceptualize

institutionalizationasanoutcomedeterminedbytheenvironmentsurroundingthe

network.

“Institutionalizationshouldbeseenasasetofcontingentchoicesmadeby

actors,ratherthanassometypeofnaturalevolutionoforganizational life.

Institutionalization could in some instances be seen as a process in which

actorswithaninterestincontrollingtheregulationofapolicysector,orapart

thereof,struggletocontrolthedomainor‘space’oftheemerginginstitution

(…)Totheextentthattheenvironmentispopulatedbypowerfulactors,and

therearetechnicaldemandsforuniformity,therewillbegreaterdemandsfor

institutionalization.Forexample,public-sector institutions facedemandsto

create legal instruments rather than informal instruments for controlling

policysectors.”(PierreandPeters2009,p.342)

The outcome of the struggle for regulatory control over the European energy

regulatory policy resulted in a hybrid agency, where the Commission and the

regulatorssharethedriver’sseat.

However, CEERdidnotdissolve.Rather, it re-invented itself.On theonehand, it

beganfocusingonpolicyissues,outsideoftheofficialremitofACER,suchasretail

markets and consumer issues; on the other hand, it added a new function to its

missionbyengagingintrainingprovision.Finally,inthepastfewyearsCEERhasbeen

welcomingregulatorsfromFYROM,Montenegro,Moldova,Bosnia-Herzegovinaand

93

Kosovoasaffiliatemembers,inanefforttorevampitsrepresentativeness(interview

32) and potentially representing the European energy policy in the EU

neighbourhood(interview21).

Moreover, regulators emphasise the CEER’s twenty-year networking experience,

conferringitanadvantageovertheEuropeanAgencyintermsofspeedandflexibility:

“Wehavemoreresources.ACERisalwayslimitedbytheCommission

resources(…)todayACERhas60peopleandtheregulatorshaveabout180

fulltimeequivalentsthatworkforACER,(…)theregulatorsasawholegroup

probablyhave3500peopleinEuropesowecanquicklycomeupwith5or10

peopleonsomething,ACERwouldhavetogototheEC,totheadministrative

service,getabudgetincrease,hastohire,tofollowaprocedure,soittakes

themalmosta year to findanother 10people.We cando this in aweek”

(interview25).

“They [CEER representatives] really negotiatedwith the Parliament

andtherearealotofthingsintheThirdPackagethatareactually,even

ourwording…whichtheAgencycouldnothavedone,Imeanitdidn’texist,

but we couldn’t have done that through ERGEG, we could only do it

through…andwe’vestillgotthat”(interview15).

CriticaljuncturesinthehistoryoftheCEER.

ThehistoryofCEERismarked,similarlytoNARUC’s,bykeydefinitionalmoments.

FourstandoutfortheirsignificancetoCEERdevelopments,asemergingfromthe

interviews: themeeting of the Italian, Spanish and Portuguese regulators at the

conferenceinSwedenin1997,whichsetinmotiontheprocess;theendorsementof

theBritishNRA; the launchof the Florence andMadrid Fora in 1998; the call on

regulatorstocollaboratewiththeEuropeanCommissiontospeedupenergymarket

integration following the launch of the Lisbon Agenda (2000), which led to the

94

creation of ERGEG (the European Regulatory Network) in 2003; and the

establishmentoftheACERin2009.

Only one, however, displays theproperties of a critical juncture according to the

diagnosticguidelinesinCapocciaandKelemen(2007).Thefortuitousencounterof

thethreesouthernEuropeanregulatorsinVasterasandtheirdecisiontomaintainan

informalcollaborationwasabsolutelykey,inthatitrepresentedtheemergenceof

CEER.However,theFlorenceForumrepresentedthefirstoccasionwhereregulators

confronted other stakeholder’s preferences and interests concerning the future

directionoftheEuropeanregulatoryframeworkandrealizedthatthetensionsthey

experienced in their national context existed across the EU. In otherwords, they

realized that those tensions and divergences would remain a feature of the

regulatorysystemandthat,therefore,regulatorswouldbenefitfromcollaborating

andconstitutingasingleentity.Moreover:

• Thelengthoftheeventsfollowingthefirstmeetingoftheforaintermsof

regulators’networkrelationsandinteractionwiththeEuropeanCommission

greatlyexceededtheperceivedimportanceoftheeventatthetime,andits

duration.

• Afterthecriticaljuncture,thelikelihoodofnetworkparticipationinthepolicy

processincreasedexponentially.

• Contingency was paramount, as the initiative had originated from the

institutionalenvironmentoftheregulators…

• …thankstotheproactiveagencyofasingleindividual,theDirectorGeneral

forEnergyatthattime.

OneimportantfeatureofacriticaljunctureasperCapocciaandKelemen(2007)that

appearsmissingfromtheCEERhistoryistheavailabilityofcounterfactuals.Inother

words,theCEERdoesnotseemtohaveemergedoutofaperiodofcrisisorpolitical

uncertainty, when it became the favoured option amongst several alternatives.

However,amorethoroughreflectionshowsthatthisimpressionisinaccurate.Inthe

early to the mid-1990s, the existing European energy NRAs faced considerable

indeterminism:thelegalframeworkatEUlevelwasbeingestablished;NRAshadno

95

guidanceastohowtosteertheliberalizationprocess;thefirstDirectiveshadjust

beenreleased,andonlyafewMemberStateshadestablishedNRAs.AllthatNRAs

had to guide their decision-makingwas theWhitePaper theEChadpublished in

1995, outlining the objectives of the EU energy policy, and the first liberalization

Directive for electricity markets (96/92/EC) where NRAs as such are not even

mentioned.

Moreover,astheinitialoutcomeoftheVasterasmeetingin1997shows,regulators

came from national backgrounds thatwere sowidely different that they initially

ruledout thepossibilityof cooperating. TheUK regulatorwasmore interested in

establishing linkswith regulatoryagencies inoverseascountries,wherevast scale

privatizationswereunderway.Scandinavian regulatorswereembedded inavery

specificgovernancesettingsbearingnocomparisonto thatofSouthernEuropean

regulators.TheinitiativeofthethreeSouthernEuropeanregulators,however,had

twoeffects.Firstly,byexchanginginformation,regulatorssetuptheirnetworkasa

rapidly accessible repository of information on different national markets; this

attractedmorememberssincethemereexistenceoftheseconnectionsreducedthe

newregulators’uncertainty.Thisconditionledtotheexpansionofthethree-node

network formed by the three Southern European regulators. Secondly, by facing

stakeholders from different Member States at the Florence forum, regulators

realizedthattheyfacedcommonchallenges,whichwereunlikelytovanish.Hence,

they would all would benefit from a mechanism of collective action and

representationanddecidedtoconsolidatetheCEERintoanetwork,displayingthe

propertiesofanorganizationandtherules,normsandlongevityofaninstitution.

96

PartI.Conclusion:networkingforcontrol.

Having emerged in different historical epochs and in response to diametrically

differentinstitutionalandmarketcircumstances,NARUCandCEERhaveapparently

littleincommon.However,Iselectedthesetwonetworksascasestudiesbecause

bothgather regulatory authorities embedded in amulti-level governance system,

that hasmade the political decision to organize the provision of certain services

through independent regulatorybodieswithdelegatedauthority. Furthermore, in

boththeUSAandtheEU,theestablishmentofregulatoryauthoritiesderivedfroma

radicalchangeinthepolicyparadigmgoverningthesector:fromcompetitionunder

municipal franchise intheformercase; fromstateownershipandmanagement in

thelattercase.Inbothcases,regulatorynetworksemergedspontaneouslyfromthe

coordinatedactionofsmallgroupsofnetworkmembers(ProvanandKenis2008).

The literature, however, has rarely investigated their reasons, and the interplay

betweenregulators’networkinitiativeandtheirsurroundingenvironment.

Thequestionguidingtheanalysisintheforegoingtwochaptersis,therefore:what

are the conditions leading to network emergence? To answer this question, I

reconstructedtheeventsandcircumstancesleadingtonetworkemergenceinboth

cases. The origins of NARUC date back to the 1880s; hence, I relied on archival

sourcesandexistingliteraturetoreconstructitsemergence.TheoriginsofCEERare

muchmore recent, dating back to the late 1990s; hence, Imostly relied on elite

interviews with individuals involved in its emergence. The analysis is based on

inductivereasoningbasedontheseinformationsources,onthebackgroundofthe

findingsandinsightsoftheinstitutionaltheoryoforganizationbutalsotheoriesof

bureaucraticpolitics.

Spontaneous regulatory cooperation emerges soon after the establishment of

regulatorybodies:regulatorsareunsureabouttheboundariesoftheirroleandface

domesticopposition.Thewillingnesstolearnfromtheexperienceofcounterpartsin

theirbroaderinstitutionalenvironment,constitutedbythenascentRegulatoryState,

undergirds the emergence of both networks. Regulators pursued their peers’

expertise with the aim of increasing their domestic legitimacy, from which they

97

would derive their authority. Although the American process was marked by

litigation,whiletheEuropeanprocessbyinformalelitenetworking(John2012),the

inception of the network was the regulators’ response to the uncertainty

surrounding them. Idistinguish thisearlyphaseof regulatorynetworking from its

subsequent one, where loose professional ties transform into consolidated,

politicallyconsequentialcollectiveaction.

Thefindingsof thecomparativehistoricalanalysisarethatregulatorsnetworkfor

control.Namely,theydecidetoconsolidatetheirnetworksasentities,bearingthe

propertiesoforganizations(e.g.legalpersonality,physicalheadquarters,secretarial

staff,budgetarycontributions,workplansetc),astheyrealizethatthegovernance

tensions inherent in the multi-level governance system will always represent a

potential threat to the extent of their autonomy and authority. These tensions

originateintheallocationofregulatoryauthorityacrosslevelsofgovernance.Even

wherethisallocationhasseeminglyveryclearboundaries,asinafederalstatesuch

astheUSA,regulatorsatdifferentlevelsofgovernancecollideovertheextentofthe

primacyoftheirrespectivedecisionswithingiventerritorialjurisdictions.Incontexts

where the allocation of regulatory power across levels of governance is far from

havingclearboundaries,giventheabsenceofsupranationalregulatoryauthorities,

suchasintheEuropeanUnion,theissueofregulatorycontrolisevenmorepolitically

complex, as regulators derive their legal authority from the previous holder of

regulatorypower:thenationalgovernment.

Comparative examination of the cases shows that, in the course of their

development,NARUCandCEERfaceddifferentkindsofopposition. Inthecaseof

NARUC,railroadindustryrepresentativeslongbattledtoeliminatestateregulation;

Public Utility Commissioners were only able to retain their regulatory power by

allyingwiththenascentfederalagencyandtheprimarycustomersoftherailroads:

shippersofgoods.Soonaftertheirinception,Europeannationalregulatorsalsofaced

harsh opposition from the energy industry, organized in powerful “national

champions”havingclosetiestogovernment.Consumers,whostoodtobenefitfrom

regulation,werenotvocalsupportersofregulationalsobecausetheonsetofcost-

98

reflectivetariffsandtheeliminationofsubsidiesinEuropeimprovedthequalityof

servicebutalso impliedmuchhigherprices.European regulators’only strongally

were supranational institutions and, in particular, the European Commission.

RegulatorsandtheEuropeanCommissionentered intoarelationshipofsymbiotic

interdependence(“theoutputofoneistheinputoftheother”,PfefferandSalancik

1978,p.41)becausethefulfilmentofthelatter’spolicyagendabecausenecessarily

implied the strengthening of the role of the former. As discussed at the end of

chapter3,however,theestablishmentofaEuropeanAgencyfortheCoordinationof

EnergyRegulatorschangedthenatureoftherelationshipbetweentheCommission

and the CEER, transforming it from one of symbiotic to one of competitive

interdependence(“theoutcomeachievedbyonecanonlybehigheriftheoutcome

achievedbytheotherislower”,p.41).Arguably,thesamemechanismundermined

theinitiallycooperativerelationshipbetweenstateregulatorsandfederalagencies

intheUSA:asthelatterweregraduallymoreempoweredbylegislation,whathad

hithertobeenahorizontalcollaborativerelationshipbecameaverticalcompetitive

relationship.

Ireliedontheconceptofcritical juncture(CollierandCollier1991)tocapturethe

shift intheessenceofregulatorynetworkingfrommereprofessional interesttoa

strategy of preservation. I adopted Capoccia and Kelemen’s (2007) diagnostic

framework to pin down the regulators’ realization of the permanency of the

governancetensionstospecificmoments.ThemomentsIidentifiedasrepresenting

critical juncturesderivefromtheevidenceIgatheredtotracethehistoryofthese

twonetworks.Theirvaliditymaybedisputed:althoughclearlyemergingascritical

frommydata,onemightarguethatadditionaldatamayhavepointedtoadditional

momentsor events, or thatdifferent intervieweesmayhavepointed todifferent

eventsascritical.Thispotentialcritiqueiswelltaken,butdoesnotchangethevalidity

of the finding: regulators consolidated their networks into entities with

organizationalandinstitutionalpropertiesasaresultoftherealizationthattheirturf

wouldbealwayspotentiallysubjecttoerosionfromcompetingsourcesofregulatory

authorityatdifferentlevelsofgovernance.

99

Thus far, I have discussed networks of regulators as unitary actors (and units of

analysis).Inreality,however,notallregulatorsmakethesameuseofthenetworks

theybelongto.Theyarenotequallyactiveorequallyinfluentialwithintheirnetwork.

Interviewsclarifiedthattherealwaysisasubsetofnetworkmemberswhoaremore

active than the rest: they attend all meetings, contribute material, bring new

informationintothenetwork,andeventuallyshapetheagenda.Theseregulatorsare

the engines of network activities.What explains the actual structure of network

relationshipswithinatransnationalnetworkofregulators?Whyaresomeregulators

moreactivenetworkersthanothers?Thesequestionsareposedandansweredinthe

nexttwochapters.

100

PART TWO – EXPLAINING NETWORK

STRUCTURE,NETWORKACTIVISMANDNETWORK

INFLUENCE.

101

PartII.Introduction:measuringnetworkcooperation.

Theliteratureontransnationalnetworkshasoftentreatednetworksasorganisations

withinwhose framework regulatory officials interact. Various contributors to the

literatureonregulatorynetworks,aswellaspolicypractitioners,oftenusetheword

“network” metaphorically, to describe a certain kind of informal association of

professionals (Isett,Mergel et al. 2011). In social networks, nodes choose which

relationshipstopursueandmaintain.Networksaremeasurablestructures,whose

analysiscanrevealinformationabouttherelationshipsbetweenthenodesandabout

thecharacterofthoserelationships.Recently,inferentialsocialnetworkanalysishas

madeinroadsintothesocialsciencesallowingforthemodellingofnetworks.These

models afford the researcher considerable explanatory power as concerns the

driversofnetworktiesandthedeterminantsofnetworkstructure.

In the context of this research, the social network analysed is the network of

Europeannationalenergyregulators.Usingoriginaldatacollectedbetweenlate2015

andearly2017,inthispartofthethesisIinvestigatetheexplanatoryfactorsoftheir

networkstructureand thedeterminantsof their ties. Iaskedregulators to report

theirmostfrequentandregulartiesasconcernsexchangeofinformation,opinions

andadvice.Igathereddatafromall29Europeannationalenergyregulatorsthatare

membersof theCEER (i.e. fromall28EUMemberStatesplusNorway).Onlyone

regulatordidnotrespondtomyquestion;forthatregulator,Iconsideredtheirties

as reciprocated. In chapter 4, I develop an Exponential RandomGraphModel of

networkstructure.Inchapter5,Iinvestigatethedeterminantsofregulators’bilateral

tiestoeachother.

Inthemodeldeveloped inchapter4, IhypothesisethatthetiesamongEuropean

regulators would display a pattern of homophily (McPherson, Smith-Lovin et al.

2001).Homophilyisthetendencyofindividualstoassociateandbondwithsimilar

others,asintheproverb"birdsofafeatherflocktogether".Itspresencehasbeen

discovered in a vast array of network studies. Specifically, I hypothesise that

102

European regulators tend to networkwith peers overseeingmarkets that have a

similarstructuretotheirs.SincetheEuropeanregulatorypolicyforenergymarkets

is premised on liberalization, I further hypothesise that regulators overseeing

marketsatmoreadvancedstagesofliberalizationwouldreceivesignificantlymore

incoming ties, as their peers seek to learn from them. Furthermore, previous

literature points to resources as a powerful motivation for cooperation among

expertsandregulators(Cengiz2007,Alcañiz2010,Alcañiz2016);therefore, inthe

chaptersIalsotestthehypothesisthatregulatorswithlower(financialandhuman)

arelikelytohavemoreoutgoingtiesthantheirpeers.

The results of themodel show that, indeed, regulators are homophilous in their

choicesofnetworkpartners;inotherwords,theytendtoestablishdirecttieswith

regulators,facingsimilarsectorstructureandpoliticaleconomy.Furthermore,the

UK energy regulatory authority emerges as overall more sought after than their

peers, inaccordancewiththesecondhypothesisconcerningmarket liberalization.

Finally,human,ratherthanfinancial,resourcesappeartobeassociatedwithahigher

number of outgoing ties. The model also reveals that geographic contiguity and

commonmembership inoneof theEuropeanElectricityRegionsalsomatter,and

thatregulatorstendtoclusteraroundparticularlyactivenetworkers.

Chapter5furtherdevelopsanalysisonregulators’resourcesandhowtheyrelateto

theirnetworktiechoices.Themodelsdevelopedinchapter4suggeststhatregulators

withintermediatelevelsofstaffresourcesaremoreactivecomparedtopeerswith

largeresources. Incontrast,regulatorswithverylowstaffresourcesarenotmore

likelythanresourcedpeerstobeproactivelyusingtheirnetworks:thissuggeststhat

the link between resources and activism holds as long as regulators are not so

resource-constrainedastobeingunabletodevoteresourcestonetworking,orso

well-resourced as to be able to acquire or generate information in-house. In the

chapter, network activism is operationalized as the count of regulators’ non-

reciprocalties.Non-reciprocaltiessuggestaweaker(i.e.lessfrequent)relationship

thanthatimplicitinmutualties(Granovetter1973).This,inturn,suggestsdifferent

assessmentsofthestrengthoftherelationshiponthepartoftheinvolvedregulators.

103

This chapter stems from the notion that regulators with more outgoing non-

reciprocatedtiesareinformationseekers,andunderstandsthemasproactivelyusing

networktiestocompensateforlackingresources.

Thehypothesisunderlyingchapter5 is that lower resourcescorrespondtohigher

counts of non-reciprocated ties. The results of the analysis lend support to the

hypothesis, but only for staff resources. Budget resources are measured by a

categoricalvariableresultingfromthecategorizationofbudgetarydatareferringto

2013);explorationofthedatarevealsalinearrelationship,which,however,failsto

achievesignificance.However,thisdataisolderthanstaffdataandthereforemay

sufferfrommeasurementerror.

Thecombinedfindingsofchapters4and5arethatregulatorsarehomophilousin

theirtiechoicesaccordingtomarketstructure;thatregulatorsfrommarketsatan

advanced stage of development are more sought after; that better resourced

regulators are more sought after while less resourced regulators are the most

proactive, exception made for regulatory authorities having extremely small

resources, forwhom,plausibly, regulatorynetworking represents toohigha cost.

Arguably,theinterdependenceengenderedbytheprocessofEuropeanintegration

underpinsthismechanismofresourcecompensationbycreatingtheopportunityfor

regulators’interaction.Thepatternsemergingfromanalysissuggestthatregulatory

networkingmayhaveimproved(national)governancemorethannationalresources

alone would have allowed. This conclusion is noteworthy for the literature on

transnationalregulatorynetworks,whichoverwhelminglyfocusedonthepotential

fornetworkstoimproveglobalgovernance.Arguably,thisanalysiscallsforreflection

and,givenfurtherhypothesisdevelopmentandtesting,acknowledgmentthat the

feedback effects of transnational networking on national governance have been

consistentlyimprovingnationalregulatorypractice.

104

4. The drivers of transnational regulatory networking:Varieties of Capitalism between homophily andconvergence.

Introduction

Theliteratureontransnationalregulatorynetworkshasliterallyboomedsincethe

early2000s(Raustiala2002,EberleinandGrande2005,SlaughterandZaring2006,

TarrantandKelemen2007,BergandHorrall2008,CoenandThatcher2008,Verdier

2009,Zaring2009,Ahdieh2010,MaggettiandGilardi2011,Yesilkagit2011,Jordana

andLevi-Faur2012,Maggetti2013,NewmanandZaring2013).Thecombinationof

theirpervasivenessandtheirrelativemysteriousnesshasfuelledacademicandpolicy

curiosity about their inner dynamics. What are the criteria according to which

regulatorschoosewhomtonetworkwith?Inotherwords,whatarethedriversof

regulatorynetworking?Thischapteranswersthisquestionbyanalysingnetworkties

amongNationalEnergyRegulatoryAuthoritiesfromEUMemberStates.

The governance literature has converged on the overarching understanding that

transnational (or trans-governmental) networks are meant to improve the

governanceofeconomicsectorsorphenomena,whosereachextendsbeyondany

single country. Within networks, regulators have the opportunity to exchange

informationandtocoordinatetheir regulatorypractice inorder to facilitatecross

border tradeand investment.Theseconclusions resonatewith thestancesof the

liberal school of international relations (Keohane and Nye 1974, Keohane 1998),

which has emphasised how increased interdependence motivates transnational

cooperationandcoordination,leadingtothecreationofinternationalregimes(Haas

1980,Keohane1982). In this view, regulatorsnetwork transnationally inorder to

tacklethechallengesofinterdependencewhilereapingtheopportunities.

105

Beyond interdependence, the drivers of regulatory networking have rarely been

investigated.Morerecentexplanationshavepointedtotheimportanceofautonomy

and resources in prompting regulators to network transnationally (Reisenbichler

2015,Vestlund2015,Bach,DeFrancescoetal.2016).Overall,inexistingliterature

theword“network”usuallyconstitutesausefuldescriptivemetaphor,ratherthana

relationalstructure(Isett,Mergeletal.2011).Yet,therealthrustofnetworksresides

intheconnectionsbetweenthenodesformingthem.Regulatorsmaintaininformal

networktieswithpeersfromothercountriesbecausetheyfinditworthwhile.Even

whensemi-formalizednetworkedorganizations(suchasEuropeannetworks)exist,

individual regulators are unlikely tomaintain regular informal tieswith each and

everyoneoftheirpeers;moreplausibly,theysustainbilateralfrequentandstable

tiestoasubsetofthem,reachingouttoothersmoresporadically.

In this article, I investigate the drivers of regulatory networking by analysing the

directed network of connections between the 28 National Energy Regulatory

AuthoritiesofEUMemberStates,plusNorway. IdevelopanExponentialRandom

Graph Model (ERGM) premised on hypotheses aimed at testing whether the

VarietiesofCapitalism(VoC)framework(HallandSoskice2001,Hancké,Rhodeset

al. 2007) holds relevance for explaining network structure. The results show that

regulators are homophilous in their tie choices, which means that they tend to

establishtieswithcounterpartsthataresimilartothem;namely,thisstudyfindsthat

similarity in the structureof theenergy sector across countries –operationalized

throughtheVoCframework–isapowerfuldeterminantofnetworktiesforsome

clustersofregulators.Moreover,theBritishregulatoremergesasconsiderablymore

activeandinfluentialthanitspeers,andadivideisobservablebetweenregulators

fromEU-15andothers.Contrarytoexpectations,theresultsofthemodelindicate

that higher, not lower resources are associated with more active regulators,

suggestingthatmoreresourcefulregulatorsarebothmoresoughtafterandmore

activenetworkers.

106

Overall, theseresults lendsupport tocontributionsunderliningthe importanceof

expertise-driven policy learning as the driving force of transnational regulatory

networking.ThepaperconfirmstherelevanceoftheVoCframeworktounderstand

network industries and regulatory institutions anddemonstrates itsusefulness to

also explain the transnational networking choices they make. Although co-

membership inEUpolicyframeworksandtransnationalflowsofelectricityand/or

gasalsopartiallyexplainnetworkstructure,VoC-relatedfactorsareassociatedwith

muchhigheroddsoftieexistence.

ThehomophilyeffectisstrongestforsomenewerMemberStates,suggestingthey

formaratherdetachedcliquefromthecoreofthenetwork.Ifplacedinthecontext

ofEuropeanenergymarketintegration,thisresultsuggeststhattheestablishment

ofaEuropeanAgencyfortheCooperationofEnergyRegulators(ACER)wasprobably

necessary to compel all EU energy regulators to coordinatewith and learn from

counterparts regulating very differently organized markets, as well as to spur

convergenceacrossthewholeEU.

Thedeterminantsofnetworkties:literaturereviewand

hypotheses.

The structures of interconnection or interaction among a set of nodes can be

measured,explained,andpredictedusingSocialNetworkAnalysis(SNA).Theusage

ofSNAfor transnationalornationalnetworksofcivil servantsorotherregulatory

officialhasrecentlymadeinroadsintothesocialsciences,asshownbythegrowing

number of contributions employing the technique (Alcañiz 2010, Cranmer and

Desmarais 2011, Ingold, Varone et al. 2013,Maggetti, Ingold et al. 2013, Alcañiz

2016,Boehmke,Chyzhetal.2016,Cranmer,Leifeldetal.2017,Lazega,Quintaneet

al.2017).Patternsofregulatoryinteractionshaveoftenbeenoperationalizedusing

co-affiliations, such as co-membership in certain policy fora or platforms, or

107

perceived influenceandresources (Cranmer,Leifeldetal.2017).Further, thefact

that nodes prefer to connect to nodes that they perceive as being similar to

themselves in some theoretically or empirically relevant respect (a pattern called

“homophily”) has often emerged as having considerable explanatory power

(McPherson,Smith-Lovinetal.2001,Lee,Leeetal.2012,Maoz2012,Barberá2015,

Alcañiz2016).

Scholars have conceptualized transnational regulatory networks as the functional

response to interdependence. Indeed, networks have been shown to affect rule

adoptionacrosscountries(MaggettiandGilardi2011,MaggettiandGilardi2014)via

aprocessofregulatorycoordinationatsupranationallevelleadingtothecreationof

softrulestobeimplementedatnationallevel.Inasociologicalperspective,networks

havebeendepictedasarenaswhereregulatorscanexchangeinformation,opinions

andexperiencesandthereforelearnfromeachother(Majone1997,Humphreysand

Simpson 2008, Sabel and Zeitlin 2008, Bianculli 2013). In a policy learning

perspective,networkedcooperationthatissustainedovertimeenablesregulators

toexperimentwiththeoutcomesoftheircollaboration,toconceivenewapproaches

tooldandnewproblems(SabelandZeitlin2012).

More recently, the literature has complemented these understandings with

perspectivesemphasisingdomesticcircumstancesandchallengesasmotivationsfor

transnationalnetworking. Inparticular,contributorshavehighlightedthestrategic

use that regulators make of their networks to achieve more independence and

autonomy (Danielsen and Yesilkagit 2014, Ruffing 2015) from government by

exploiting the informational advantages deriving from transnational networking

(EberleinandGrande2005,Jordana2017).Additionally,regulatorsappeartopool

resources throughnetworking, thereby compensating for those they lack (Alcañiz

2010,Vestlund2015).

However, both sides of the literature have thus far overlooked the rationales

explaining regulators’ choices regarding their network connections. Regulators

108

maintain informal, bilateral network ties beyond the official policy framework,

because they find itworthwhile. Theyareunlikely todevoteanequal amountof

efforttonetworkingwitheachoftheirnetworkcounterparts;moreplausibly,they

choosetheirstrongties.Analysesofthemotivationsforthesechoicesare,however,

lacking. Filling this gap requires close-upobservationandanalysisof the ties that

eachregulatorhaswithinanetwork,asthiswouldallowforaclearergraspofthe

aims of regulatory networking and the formulation of generalizable statements

abouttransnationalregulatorynetworks.

Inthisarticle,Iuseoriginalnetworkdatagatheredthroughemailandphoneinquiries

fromall28energyregulatoryauthoritiesofEUMemberStates,plusNorway19.Their

self-reportedbilateraltiesconstituteanetworkstructure.Themodeldevelopedin

this chapter is aimed at explaining that structure. The literature on European

RegulatoryNetworks(keycontributions includeCoenandThatcher2008,Eberlein

andNewman2008,ThatcherandCoen2008,KelemenandTarrant2011,Levi-Faur

2011,MaggettiandGilardi2011)andmoregenerallyonnetworksofregulatorsof

infrastructure sectors has, thus far, scarcely reliedon the richness of insight that

networktheoryandnetworkanalysiscouldaffordit.Inotherwords,theliterature

has often used the term “network”without using themethods pertaining to the

quantitativeanalysisofnetworks.

There are notable exceptions: several contributions have used measurements

derived from network analysis in order to quantify influence and reputation of

19 Ihave included theenergy regulatoryauthorityofNorway in thisanalysisbecause the

regulatoryauthorityisamemberoftheCouncilofEuropeanEnergyRegulators(CEER),the

voluntary network of European energy regulators, and the country has an obligation to

complywith theprovisionsof theEUenergy law. Ihavenot included theNRAof Iceland

becauseitisnotcompelledtoadopttheEUlegalframeworkinenergy.

109

different bureaucratic and political actors in the Swiss telecommunication sector

(Ingold, Varone et al. 2013) or the independence and accountability of different

regulatory authorities (Maggetti, Ingold et al. 2013). Few contributions, however,

haveinvestigatedtheexplanatoryfactorsoftherelationalstructureconnectingthe

members of a network. Alcañiz (2010, 2016) is a rare exception: her model of

networkstructurefindsthatcommongeographicaloriginsandlackofresourcesdrive

the frequency of regulatory cooperation amongnuclear experts. At any rate, the

mere existence of semi-formalized frameworks of cooperation, such as European

Regulatory Networks, offers no insight onto the drivers of informal bilateral ties

among regulators. Yet, the importance of trans-governmental networking for the

shapingofEUenergypolicyinparticularcanhardlybeoverlooked(Eberlein2008,

Kaiser2009).

Recent scholarly contributions have shown that the characteristics of network

industries and national regulatory authorities are associatedwith the “Variety of

Capitalism(VoC)”ofthecountry(Thatcher2007,GuardiancichandGuidi2016).The

VoC framework subdivides OECD countries according to “theway in which firms

resolve the coordinationproblems they face” (Hall andSoskice, 2001,p. 7) in the

country where they operate. The main watershed is between Coordinated and

Liberal Market Economies (CMEs and LMEs). In LMEs, firms predominantly

coordinate their activities via markets; transparency is, therefore, essential and

regulatoryauthoritiesandpolicyaretaskedwithensuringit.InCMEs,firmsrelymore

heavily on relational modes of coordination with other actors in the political

economythanonmarketsignals.InContinentalEurope,CMEsarethemajority.The

onlyLMEsarefoundtobetheUnitedKingdomandIreland(HallandGingerich2009).

CMEscanbefurtherdifferentiated:Scandinaviancountriesappeartohaveaspecific

type of social-democracy, different from the coordinated economies of Germany

(theonly pureCME),Austria, Belgium, and theNetherlands.Moreover, Southern

Europeancountries(France,Italy,SpainandPortugal)havebeendefinedas“mixed”

110

(or“Mediterranean”)economies(orMMEs),wherecoordination(orlackthereof)co-

existswithaninterventionistandcompensatingstate.

Thatcher (2007)studiesthe interplaybetweentheVoCframeworkandregulatory

institutions in network industries. He examines the cases of the UK, France and

Germanyandstatesthat,whilethe1980sUKprivatizationandliberalizationreforms

ofnetworkindustriesoccurredlargelyindependentlyfromEUdevelopments,theEU

regulatory framework for network industries dovetails the LME paradigm, being

premisedontheeffort tobringaboutcompetitionthroughtheunbundlingof the

sector, the introduction of private capital, cost-reflective pricing, market

transparency,andcross-borderenergytrade.TheEUchoice,accordingtoThatcher

(2007), forced all other Member States to converge towards the LME model of

regulationofnetworkindustries.Indeed,beforetheEUbeganlegislatingonnetwork

industries, France and Germany displayed very different approaches to those

sectors: in France, the state owned or controlledmost of the sectoral firms and

steeredtechnologicalandsectordevelopment;inGermany,industrialassociations

performed self-regulatory functions and the state accepted to keep itself at a

distance,whileretainingamonitoringrole.

Afterlongresistances,FranceandGermany,asallotherMemberStates,eventually

compliedwiththerequirementsofEuropeanlegislationinformalrespects,interalia

bysettingupregulatoryauthorities.Uponcloserexamination,however,itbecomes

evidentthatbothretainedthekeyfeaturesoftheirdistinctivemodeofcoordination

inthe informalnetworksundergirdingthesector(Thatcher,2007).Thus,inFrance

thestatepreservesakeysteeringrole,whileinGermanyindustrycontinuesleading

sectordevelopment,eventhoughbothatleastformallyconvergedtoanLME-type

sectororganization.

TheVoCofthecountryalsoaffectsthecharacteristicsofregulatoryinstitutions:in

bothFranceandGermany,lackofpoliticalsupportpreventedregulatoryauthorities

from enjoying the wide-ranging independence that is characteristic of an LME

111

approachand,therefore,ofUKregulators.HisstancesechoCoen(2005),whofound

regulators in Germany to be less independent than in the UK. Guardiancich and

Guidi’s (2016) analysis confirms these patterns: in the UK, wheremarket signals

shapemarket players’ interactions, the independenceof the regulatory authority

from both government and industry is crucial to ensure the credibility of the

regulatorysystem;inCMEs,strategiccoordinationamongfirmsprecludesregulatory

detachmentfrommarketactorsandgovernment;inMMEs,thehighlydiscretionary

roleof the state leaves little room for regulators’autonomy. Indeed, institutional

complementarities play a role in this regard, as market players and government

demand of regulators to perform the role best suited to the extant mode of

coordination.Thesecontributionsshowthat theoriginal classificationbyHalland

Soskice (2001), by and large, is able to explain variation in national regulatory

arrangements.Moreover, these studies show that VoC is an inherently relational

framework.

Howdoesthistranslateonthetransnationalarenaofregulatorynetworking?Does

theVoCoftheircountryaffectregulators’networkingchoices?Iarguethatitdoes,

possiblythroughmorethanonemechanism:ontheonehand,regulatorsmaygain

morevaluefromcommunicationwithpeersdealingwithasimilarsectorstructure,

asthesearemorelikelytofacesimilarchallengesasthemselves;ontheotherhand,

regulatorsfromcertainVoCmayenjoymoreinfluencethanothers,promptingtheir

peers to seek themoutas frequent interlocutors. Since theEUenergy legislation

conforms to the LMEvariety, one shouldexpect regulators fromLMEs to receive

moreincomingtiesasallotherVoCslowlyconverge(oradapttheirextantmodeof

coordinationinnetworkindustries)totheLMEmodeofcoordination.

In a recent contribution on the European network of broadcasting regulators,

Papadopoulos(2017)emphasisesthatregulatorslearnfrompeerstheyholdinhigh

esteemand/orwhoarefacingproblemsperceivedassimilar,butdoesnotexplore

thispatternanyfurther. IntheirstudyoftheEuropeannetworkofpatent judges,

112

Lazega,Quintaneetal.(2017)findthat“Itisclearthatjudgesdosorteachotherin

socialnetworksbasedontheirbelongingtoblocksofcountrieswithsimilartypesof

capitalism.”(p.19).Thevagueexplanationtheyprovideisthatjudgesfromsimilar

VoC probably refer to similar bodies of law. Both studies also note that, within

networks,certainregulatorsaremuchmoreinfluentialthanothersandperceivedas

modelsbytheirpeers,whoseekaccesstotheirexperienceandsuggestions.

Thereasonwhyregulatorsbelongingtocountrieswithsimilarmodesofcoordination

inthepoliticaleconomyaremorelikelytochooseeachotherasnetworkpartnersis

that they are likely to encounter the same bottlenecks in the formulation and

implementationofregulatorypolicy.Theenergysectoriswellsuitedtoillustratethis

argument: EU energy legislationmandates the unbundling (or separation) of the

network infrastructure, consisting of the separation of high voltage/capacity

(transmission) and low voltage/capacity (distribution) grids from the potentially

competitivesegmentsoftheelectricityandgassectors,i.e.generation/production

and supply. Previously, production and investmentdecisionsweremadewithin a

singleverticallyintegrated,usuallystate-ownedenergycompany.Theunbundlingof

generation of electricity (and gas production) from transmission, distribution and

supply ismeanttoensurethatcoordinationbetweenenergydemand,supplyand

construction of infrastructure happens through market signals (as per the LME

model)20.

20TheEUlegislationhasmandatedfullseparation(i.e.functional,accountingandlegal);the

so-calledOwnershipUnbundling(OU).UnderOU,theTransmissionnetworkisownedand/or

managedbyaseparatecompany(so-calledTransmissionSystemOperator-TSO),havingno

but a minority interest in generation or retail activities. Alternative models deemed

admissibleincludetheIndependentSystemOperator(ISO)model,whichforeseesthatthe

113

The economics literature has recognized that national regulators need to make

informal“regulatorycontracts”(SternandTrillas2003,Helm2009,Stern2012) in

their national markets, with government, infrastructure managers, equipment

manufacturersandsuppliers.Thesecontractsconcerntheregulatoryapproach,the

objectivesoftheregulatorypolicyandtheboundariessettofirmsoperatinginthe

sector.Thestabilityandcredibilityofregulatorycontractsarepivotaltoregulators’

legitimacy and, therefore, authority. Regulators have to conclude regulatory

contracts thatareacceptable toallparties,hencecorresponding to theprevalent

modeofcoordination,whilealsofulfillingtheobjectivesoftheEUlegislation,which

is clearly LME-inspired. Therefore, I argue that European regulators seek out

counterpartsfromsimilarnationalmodesofcoordination,asthesearebestsuited

torespondtotheirrequestswithreplies,groundedintheknowledgeofthestructure

of interactiontheregulatorfacesathome.Atthesametime,however,regulators

use networks to access the experience and expertise of other counterparts,

particularlyinfluentialones.

Myfirsthypothesisis,therefore,thatEuropeanenergyregulators’connectionsare

governedbyapatternofhomophily(CranmerandDesmarais2011)drivenbysimilar

VoCasexemplifiedintheenergysectorstructure.

H1:RegulatorsnetworkmorewithregulatorsfromtheirsameVarietyofCapitalism.

networkisownedbythevertically-integratedcompanyandleasedtoaseparatecompany.

Thethirdmodel,theIndependentTransmissionOperator(ITO),impliesthatthetransmission

network is owned by the vertically-integrated company and managed by a department

withinit,walledbystrictindependencerequirementsinitsoperations.Thethreemodelsco-

existacrossEUMemberStates,althoughtheOUisthemostcommon.

114

MysecondhypothesisisthattheUKregulatorshouldbehighlysoughtafterbyits

peers, given that all EUMember States have had to converge, at least partially,

towardsanLMEtypeofenergysectororganization.

H2:RegulatorsfromLiberalMarketEconomieshavesignificantlymoreincomingties

thantheirpeersfromotherVarietiesofCapitalism.

OneofthemaincriticismslevelledattheVoCframeworkisitsstationarity.Indeed,

the VoC of a country is virtually time-invariant. Studying connections between

regulators as dependingon theVoCof their country implies considering themas

rather stationary. The data underlying this analysis, indeed, concerns regulators’

frequent,regularconnections.Yet,theflexibilityofinformalsocialnetworksisone

oftheirkeyattributes:participantscanaddorseverlinksasopportunityandneed

require. However, the reasons for tie formation and elimination may be highly

idiosyncratic, and last for limited time periods. There are, however, time-variant

factors, which may consistently affect the likelihood of tie formation: resources

(Glachant, Khalfallah et al. 2013). Alcañiz (2010, 2016) finds that the sudden

budgetarycuts,especiallyindevelopingcountries,representastrongincentivefor

nuclearexpertstocollaboratewiththeirpeersintransnationaljointprojects.They

compensateforthesuddendecreaseintheirresourcesbypoolingtheircompetences

with those of their peers in neighbouring countries in order to accomplish their

ongoing technical activities. Less resourceful regulators, therefore, should use

informalnetworkingtocompensatefortheirlackingresources.Theseconsiderations

leadmetoformulatethethirdhypothesis.

H3:Regulatoryauthoritieswith lower (budgetaryandhuman) resourcesaremore

activenetworkers(i.e.havesignificantlymoreoutgoingtiesthantheirpeers).

ThisanalysiscomprisesthefullsampleofEuropeanenergyregulatoryauthorities,

includingEasternEuropeanMemberStates.However,theoriginalVoCclassification

by Hall and Soskice (2001) only includes OECD countries, which then excluded

115

EasternEuropeancountries.ScholarsofVoChaveinvestigatedtheusefulnessofthe

framework forunderstanding thepolitical economyofnewerEUMemberStates.

ThereissomedegreeofconsensusabouttheBalticcountriesbeingclosesttothe

LMEmodel,andSloveniatotheCMEmodel(Feldmann,2006).Whileacknowledging

thesedefinitionalefforts,Hancké,Rhodesetal.(2007)underlinethatnewMember

States should be considered as transitioning towards models of capitalism, and

clusterthemasEmergingMarketEconomies(EMEs).NölkeandVliegenthart(2009)

relyona specific trait (i.e.provisionof capital through foreigndirect investment)

foundintheCzechRepublic,Hungary,Poland,andtheSlovakRepublictolabelthem

DependentMarketEconomies(DMEs).

IcategorizetheVoCofEUMemberStateswithrespecttothemarketstructureof

their energy sector. The resulting categorization differs from the “traditional”

breakdown insomerespects;principally, innotconsidering IrelandasanLME,as

doneinHallandGingerich(2009),giventhatitsenergysectorisalmostentirelyunder

government control. I find a neat distinction between countries where the

transmission and distribution segments are owned and operated by different

companies,andareseparatefromgenerationandretail(onlytheUK)andcountries

where companies active in distribution are also active in retail and sometimes in

generation,too(asinmostCMEs).InMMEs,thedominanceofformerlystate-owned

incumbents has been restrained through regulation (as in France), mandatory

divestment(asinItalyandGreece)orprivatizations(asinPortugalandSpain);still,

these former incumbents have the largest market shares. Scandinavian Market

Economies have a good level of competition in both generation and retail, even

though state-controlled incumbents are also active in those segments, and are

characterisedbylocally-owneddistributionsystemsandstate-ownedtransmission

systems.Thepenetrationofforeigncapital(mostlyfromWesternEuropeannational

companies) in the generation, distribution and retail segments of the electricity

sector in severalEasternEuropeancountries resonateswith their classificationas

Dependent Market Economies (DMEs). Finally, a prevalence of direct state

116

ownership and control in all segments of themarket is evident in other Eastern

Europeancountries,Cyprus,MaltaandIreland(details inAppendix).Theresulting

classificationusedinthischapterisasfollows:

- LMEs:onlytheUK;

- CMEs:Germany(theonlypureCME),Austria,BelgiumandtheNetherlands;

- “Government-owned”: Ireland, Croatia, Cyprus, Estonia, Latvia, Lithuania,

MaltaandSlovenia;

- Dependent Market Economies (DMEs): Czech Republic, Hungary, Poland,

SlovakRepublic,BulgariaandRomania;

- MMEs:Italy,France,Greece,PortugalandSpain;

- ScandinavianMarketEconomies:Finland,Sweden,NorwayandDenmark.

Thetiechoicesofregulatorsofnetworkindustries,however,mayhardlybechoices

atall:geographyplayssuchadominantroleinenergyinfrastructureastopotentially

overcomeorobscureanyotherrationaleforinformalcooperation,asregulatorsare

boundtocommunicateoftenwithregulatorsfromneighbouring(or,moreprecisely,

interconnected) countries. Regular communicationwith neighbours, in turn,may

engenderstablepatternsofexchangeofinformation.Inordertoassessthiseffect,

ratherthangeographiccontiguity,oneshouldconsidertheactualdirectionof the

flowsofelectricityandgasthataretransmittedacrossbordersandseaswithinthe

EU.Consequently,Iincludeelectricityandgasflows,bothacrosslandandsea,inthe

analysis,astheirdirectionalitymaybestronglydeterminingregulatoryinteractions

andpotentiallyobscuringanyotherexplanatoryfactor.

Furthermore, within the context of the EU, exchange of information is, beyond

beneficial,actuallymadecompulsorybythepresenceofasharedpolicyframework

thatnationalregulatorshavetoimplementatnationallevel.Thedeclaredaimofthe

EUenergypolicyisachievingafullyintegratedInternalEnergyMarket(IEM).Asan

interim step towards theachievementof the IEM, theEuropeanCommissionhas

launched the so-called Regional Energy Initiatives for Electricity and Gas. The

117

RegionalInitiativesgroupregulatorsintoeightregionsforelectricityandthreefor

gas with the intent of achieving integrated regional markets for both. Frequent

interaction in the framework of the Regional Initiatives may have engendered

socializationdynamicsleadingtotrustandthustothemaintenanceoftiesbeyond

theofficialpolicyframework.Ithereforeincludeco-membershipintotheRegional

Initiatives as a suitable control and proxy for the relevance of European policy

requirementstoexplainregulators’informalties.

Dataandmethod:anExponentialRandomGraphsModelofstrong

tiesamongEuropeanregulators.

Theliteratureontransnationalregulatorynetworks,bothEuropeanandglobal,has

scarcelyexploitedthepowerofquantitativenetworkanalysistoexplainpatternsof

transnationalnetworking.Thisisplausiblyduetolackofdataontheconnectionsthat

individual regulatorsmaintain. Those contributions that use the tools of network

analysis usually concern multi-level and multi-actor networks of experts around

specificissueareas,suchastheenvironment,andusuallyrelyondataconcerningco-

membershipincooperationinitiativesand/orco-attendanceofcertainevents.The

assumed link between co-membership and collaboration, however, is not self-

evident, as actors may be members of the same initiative but not collaborate

regularly.Veryrecentcontributionsinthepolicystudiesliteraturehaveresortedto

asking networkmembers about their regular and frequent ties to other network

members(Fischer,Ingoldetal.2017,HamiltonandLubell2017)inordertoattempt

tocapturetheessenceofcoordination.Thisanalysisadoptsasimilarapproach in

studyingtheempiricalcaseofahomogenousnetwork(i.e.comprisingonetypeof

actor)oftransnationalscope:thenetworkoftieslinkingEuropeanenergyregulators,

asreportedbyregulatorsthemselves.

118

Igatheredthedatausedinthischapterbetweenthesecondhalfof2015andlate

2016.Specifically,IwrotetotheHeadsofInternationalAffairsdepartmentsandto

Communication Officers. Not all European energy regulatory authorities have

dedicatedInternationalAffairsoffices,butallhavestaffdedicatedtointernational

affairs,suchasCommunicationOfficers. Iaskedtheserespondentstoreplytothe

following question21: “Think of the individuals you (or somebody at your NRA)

exchangeinformationwithmoreoften.WhichNRAsdotheybelongto?”.Network

analysisisverysensitivetomissingdata;it isimportanttopossessinformationon

thewholenetworkinordertomakeaccurateanalyses.Therefore,Ichosetorelyon

a single question in order to maximize my chances of receiving a reply from all

networkmembers.Indeed,IhaveobtainedrepliesfromallEuropeannationalenergy

regulatoryauthorities,barone.Forthatmissingrespondent,Ihavejustconsidered

thenominationsofotherregulatorsasreciprocated.

Being aware that the notion of “most frequent” may mean different things to

different people, and that regular exchanges of information may include mostly

routineexchangesduetosharedbordersandinterconnectedinfrastructurerather

thanbeoccasionsforlearning,Iaddedtomyrequestsexplanatorytextspecifyingto

respondentsthattheyshouldnamethepeerstheygetintouchwithwhentheyseek

advice or an exchange of opinions or suggestions, not just routine exchanges of

information.Moreover, Icomplementedthequestionwitharequesttonamethe

regulatoryauthoritieswithwhichtheyareincontactaboveandbeyondEuropean

policy requirements (including participation into the European Agency for the

CoordinationofEnergyRegulators).Iguaranteedanonymitytoallrespondents.

21 InhalfofthecasesIobtainedregulators’repliesoverthephone.Becausethequestion

askedduringphoneconversationswasidenticaltothoseintheemailmessages,thereisno

needtoaccountforwhetherregulatorsrespondedtotheemailorwerecontactedbyphone

inthemodels.

119

Theresultingnetworkisa“thinned”network(CranmerandDesmarais,2011),i.e.a

networkconsistingofonly thestrongrelationshipsbetweenthenodes. If theties

acrossEuropeanregulatorswereavaluednetwork(withtieshavingdifferentweights

dependingontheir importance),thenetworkstudiedinthischapter istheoneof

highly valued ties. I chose to focus on strong ties because energy regulatory

cooperation in the EU has a long history, dating back since the late 1990s

(Vasconcelos2005).Moreover,Europeanenergylegislationimposesanobligationon

EuropeanregulatorstocooperatewithintheEuropeanAgencyfortheCooperation

of Energy Regulators (ACER). Therefore, every European energy regulator is

connected to all others. I was specifically interested, however, in the informal

bilateraltiesthatregulatorsmaintainmoreregularlyandfrequently.

IsetupanExponentialRandomGraphModel (ERGM)ofthenetworkofrelations

amongEuropeanenergy regulators. ERGMaregenerativemodels: theunderlying

assumption is that the observed network structure has emerged from an

evolutionaryprocessoftieformationovertime,explainedbyaseriesofpredictive

factors(Robins,Lewisetal.2012).Theresearcherselectstherelevantfactors,which

couldbeattributesofindividualnodesorattributesofdyadicties,ontheoreticaland

knowledgebases;theanalysisrevealstheprobabilitythattheobservednetworkis

drawnfromthedistributionofthenetworkstructuresthatareplausible,giventhe

numberofnodesandthenetworkdensity,andthefactors.Thecoefficientsofthe

modelaretobeinterpretedaslogodds,asinalogitmodel.ERGMsalsoallowthe

modelling of the interdependencies existing within networks, and represent a

techniqueforinferentialnetworkanalysiswheretheoutcomeofinterestisasetof

relationships(i.e.thetiesamongthenodesofthenetwork)(Cranmer,Leifeldetal.

2017).

ThegraphinFigure1reportsthestructureoftherelationsamongEuropeanenergy

regulators. The network appears characterized by a small number of highly

connected nodes, a small number of peripheral nodes, and a majority of nodes

120

havinganintermediatenumberofconnections.Figure1showsthattwonodeshave

not been nominated by any of their colleagues as their most frequent contacts

(hence have an in-degree of zero). Figure 1 also shows thatmost of the ties are

reciprocal,whichvalidatesthedata,consideringthatIdidnotsetaminimumora

maximum number of nominations for regulators. The promise of anonymity

concerning respondents’ identities and their replies prevents me from assigning

labelstothenodesinthegraph.

Figure4-1-Visualizationofthenetwork.

EU National Energy Regulators Network

121

Asmentioned,Iexpectthetiesinthisnetworktodependonbothexogenousand

endogenousfactors.Asfortheexogenousfactors,thehypothesesIdevelopedinthe

previous section point to homophily, activism and influence according to VoC,

controllingforEUpolicyrequirements,interconnectionandflowsacrossborders.I

also include in themodel several endogenous dependencies to account for likely

patternsofsocialinteractionthatmayhavecontributedtodeterminethenetwork

structure:thedensityofthenetwork;thereciprocityofties;andthetransitivityof

ties,whereby ifnode i isconnectedto jand j isconnectedtok, there isahigher

probabilitythatiandkarealsoconnected.Ialsoincludedependenciestoaccount

for the centralization of the network, i.e. to verifywhether the network ismore

centralized around particularly active (i.e. many outgoing ties) or particularly

influential(i.e.manyincomingties)nodesthanwouldbeexpectedbychance.

Therefore,thepredictorsemployedinthemodelincludeelectricity22andgas23flows,

overlandandsea,acrossEUMemberStates(plusNorway);co-membershipinthe

EuropeanRegional InitiativesforElectricityandforGas24;dataoneachregulatory

22AmatrixreportingelectricityflowsinbothdirectionsacrossEUMemberStatesinGWhin

2015. Data from ENTSO-E website,

https://www.entsoe.eu/publications/statistics/electricity-in-europe/Pages/default.aspx

(lastaccessed3November2017).

23AmatrixreportinggasflowsinbothdirectionsacrossEUMemberStatesincubicmeters

of gas in 2015. Data from UK government website

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/579632/

Physical_gas_flows_across_Europe_in_2015.pdf(lastaccessed3November2017).

24 An affiliationmatrix of the Regional initiatives in Electricity and Gas. Data from ACER

website, http://www.acer.europa.eu/en/electricity/regional_initiatives/pages/default.aspx

122

authority’sbudgetandstaffnumbers25;andtheVoCcategorisation.Aspreviously

outlined, the VoC categories derive from my assessment regarding similarity of

electricityandgassectorstructureacrossEUMemberStates,asindicatedessentially

bythenumberandtheidentityoffirmsactiveindifferentmarketsegmentsandthe

extentofpublicownership(detailsintheAppendix).Thequantitativedatahasbeen

standardizedbeforeproceedingtotheanalysisbysubtractingthemeananddividing

bythestandarddeviation.

Resultsofthemodel.

Forthesakeofclarity,letmerecallthehypothesesIformulatedtogetherwiththe

variablesdescribingthemandthecorrespondingmechanismsintheERGM,aswell

as the factors I am using as controls for the effect of interconnection and co-

participationinEURegionalInitiatives.

and http://www.acer.europa.eu/en/gas/regional_%20intiatives/pages/gas-regional-

iniciatives.aspx(lastaccessed3November2017).

25Thebudgetandstaffnumbersofeachtheregulatoryauthorityin2012(forlackofmore

recent data). Data from the European Commission DG Energy website

https://ec.europa.eu/energy/en/topics/markets-and-consumers/single-market-progress-

report Country reports 2014, last accessed on 3 November 2017, complemented with

regulatoryauthorities’annualreportsinsomecases.

123

Table4-1–Hypotheses,VariablesandMechanisms

HYPOTHESES VARIABLE MECHANISM

1 Regulatorsnetworkmorewith

regulatorsfromcountries

belongingtothesameVarietyof

Capitalismasthemselves.

VoC Homophily

2 RegulatorsfromLiberalMarket

Economieshavesignificantly

moreincomingthantheirpeers

fromotherVarietiesof

Capitalism.

VoC Moreincomingties

3 Regulatoryauthoritieswith

lower(budgetaryandhuman)

resourcesaremoreactive

networkers(i.e.have

significantlymoreoutgoingties

thantheirpeers).

Budget

Staffunits

Moreoutgoingties

Control1 Thestructureofthenetworkof

relationshipsexistingamong

Europeanenergyregulators

mirrorsthepathsofelectricity

andgasflowsacrossEUMember

States.

Electricity

flowsand

gasflows

Matrixofnetworkties

mirrorsmatrixof

electricityandgasflows

124

Control2 Thestructureofthenetworkof

relationshipsexistingamong

Europeanenergyregulators

mirrorsthesubdivisionoperated

throughtheRegionalInitiatives

forElectricityandGas,

respectively.

Regional

Initiative

(electricity)

and

Regional

Initiative

(gas)

Matrixofnetworkties

mirrorsmatrixofco-

participationinEU

RegionalElectricity

and/orGasInitiative

TheresultsoftheERGmodelsarereportedinTable2.Iperformedtheanalysisusing

Rpackage“ergm”(Handcock,Hunteretal.2017).Thecoefficientarelogodds,that

is,afterexponentiation,they indicatetheprobabilitythatanedgeexistsbetween

twonodes,allelseequal,i.e.conditionalontherestofthegraphbeingfixed.Positive

and high coefficients indicate higher odds, while negative and high coefficients

indicateloweroddsofatieexistingbetweentworegulatorsonthebasisofthegiven

parameter.Eachexplanatoryfactorwasfedintothemodelaccordingtoitsexpected

effect26ontheoddsoftieexistence.

26ThesyntaxofERGModelscomprisesawealthofterms.Theonesusedinthismodelare

outlinedherebelow:

-“edgecov”:theinputisamatrixofcovariates;apositivecoefficientindicatestheprobability

that two nodes sharing the same characteristic are also ties (e.g. are part to the same

Regionalgrouping);

-“nodeicov”:ittestswhetheracertainattributeofthenodeaffectsitsin-degree(e.g.more

resourcesareassociatedwithsignificantlyhigheroddsofincomingties);

125

Table4-2-ERGModelsofthenetworkofEuropeanenergyregulators

(1) (2) (3) (4) (5)

Networkdensity -4.865*** -5.494*** -5.477*** -6.015*** -5.135***

(0.376) (0.496) (0.519) (0.576) (0.451)

HomophilyaccordingtoVarietyofCapitalismCoordinatedMarketEconomies

1.252*** 1.178** 1.167** 1.143* 1.177**

(0.484) (0.569) (0.593) (0.564) (0.553)

DependentMarketEconomies

0.744** 1.670*** 1.836*** 2.250*** 1.168***

(0.319) (0.514) (0.577) (0.605) (0.440)

Governmentownership 0.577** 0.800** 0.841* 0.928* 0.536

(0.257) (0.402) (0.444) (0.479) (0.360)

MixedMarketEconomies 0.586 0.416 0.397 0.452 0.911*

(0.394) (0.507) (0.512) (0.560) (0.470)

NordicMarketEconomies 1.494* 1.508 1.583* 1.625 1.448*

(0.801) (0.918) (0.917) (0.949) (0.843)

-“nodeocov”:sameasnodeicov,butforout-degree;

-“nodeifactor”:sameasnodeicov,butforcategoricalvariables;

-“nodeofactor”:sameasnodeocov,butforcategoricalvariables;

-“nodematch”:ittestsforhomophily,i.e.theprobabilitythattwonodesthatmatchonthe

givencharacteristic (e.g. tworegulatorswhoareboth fromWesternEuropeancountries)

shareatie;

-“absdiff”:similartonodematchbutforcontinuouscovariates.

126

VarietiesofCapitalismandincomingtiesCoordinatedMarketEconomies

0.980** 0.977** 1.106**

(0.462) (0.483) (0.502)


Referencecategory

Referencecategory

Referencecategory

Governmentownership 0.654 0.635 0.667

(0.475) (0.498) (0.518)

LiberalMarketEconomies 1.989*** 1.871*** 1.863***

(0.517) (0.610) (0.615)

MixedMarketEconomies 0.993** 0.958* 0.974**

(0.460) (0.492) (0.494)

NordicMarketEconomies 0.895* 0.883* 0.942*

(0.492) (0.510) (0.523)

EffectofresourcesonoutgoingtiesStaff(2012) -0.144

(0.188)

Budget(2012) 0.197 0.261*

(0.172) (0.143)

Staffsize(Fulltimeequivalents,2016):

Large(>170) Referencecategory

Medium(90-140) 0.978***

(0.353)

Medium-small(50-75) -0.173

(0.506)

127

Small(12-50) 0.750

(0.473)

Micro(>12) -0.173(0.690)

VarietiesofCapitalismandoutgoingtiesCoordinatedMarketEconomies

0.522

(0.415)


Referencecategory

Governmentownership 0.465

(0.434)

LiberalMarketEconomies 1.416***

(0.492)

MixedMarketEconomies 0.059

(0.458)

NordicMarketEconomies 0.464

(0.469)

ControlsCo-membershipinRegionalInitiativesforElectricity

0.361** 0.550*** 0.578*** 0.685*** 0.501***

(0.172) (0.196) (0.204) (0.219) (0.192)

Co-membershipinRegionalInitiativesforGas

0.154 0.130 0.101 -0.092 0.106

(0.238) (0.246) (0.251) (0.276) (0.247)

Crossborderelectricityflows

0.627*** 0.576*** 0.580*** 0.580*** 0.588***

(0.191) (0.188) (0.197) (0.181) (0.171)

Crossbordergasflows 0.052 0.016 0.013 0.085 0.035

128

(0.097) (0.096) (0.095) (0.102) (0.105)

Endogenousdependencies

Reciprocity 1.967*** 1.956*** 1.899*** 1.801*** 1.905***

(0.394) (0.383) (0.393) (0.396) (0.383)

Activity 3.022** 2.263 2.360 4.099* 2.800*

(1.487) (1.468) (1.451) (2.126) (1.623)

Popularity -0.187 -0.158 -0.209 -0.418 -0.336

(0.830) (0.914) (0.886) (0.928) (0.866)

Sharedpartners 0.051 0.061* 0.067* 0.077** 0.053*

(0.031) (0.036) (0.037) (0.036) (0.028)

Transitivity 1.073*** 0.932*** 0.908*** 0.880*** 1.008***

(0.218) (0.223) (0.232) (0.235) (0.234)

AkaikeInf.Crit. 494.559 489.868 492.298 483.653 496.646

BayesianInf.Crit. 565.052 583.858 595.687 601.840 590.636

Note: *p<0.1;**p<0.05;***p<0.01

Table2offerstwomaintakeaways:first,regulatorsdodisplayatendencytomaintain

closerelationshipswithpeersfromtheirsameVoC,allelseequal,andthispatternis

strongestandmostconsistentacrossspecificationsforregulatorsfromDependent

Market Economies and from Coordinated Market Economies; second, the UK

regulatorisnotonlymorelikelytobeatthereceivingendofatiethanitspeers,but

alsomuchmoreactivethanthem,allelseequal.Hypothesisoneandtwo,therefore,

areconfirmed.Thehomophilypatterncouldnotbetested inthecaseoftheLME

becauseitisuniqueinthedataset,hencecannotformhomophilousties.Overall,all

regulatorsaresignificantlymorelikelytoreceivetiesthanregulatorsfromDMEs(the

129

reference category in the models), bar regulators from countries where the

electricityandgassectorsaremostlyunderpublicownershipandcontrol.

In model 3, budgetary and staff resources are operationalized via a continuous

predictorandrefertotheyear2012,andappeartonotsignificantlyaffecttheodds

of tieexistence.Toshedmore lightonthematter, inmodel4 Iusemorerecent,

categoricaldataonstafffiguresreleasedbyEuropeanAgencyfortheCooperationof

EnergyRegulators(ACER)27.Allelseequal,regulatorswithmediumstaffnumbers(i.e.

90to140fulltimeequivalentemployees)aremorelikelytobeactivenetworkers

compared to regulators with large resources. Higher budgetary figures are also

associatedwithhigheroddsofoutgoing ties,but theeffect is ratherweak.These

resultsdonot fullyconfirmhypothesis three,whichexpectedregulatorswith low

staffnumberstosendsignificantlymoreoutgoingtiesthanregulatorswithhigher

staff numbers. In model 5, I examine whether VoC is associated with a higher

likelihoodofbeingactivenetworkers.Onceagain,theBritishregulatorappears,all

elseequal,morelikelytohavemoreoutgoingtiesthanitspeers.

Furthermore,theresultsforendogenousdependenciesshowthattiesareverylikely

to be reciprocated. The coefficient on the dependency called “shared partners”

shouldbereadinconjunctionwithtransitivity.Sharedpartnersindicatethetendency

for the nodes in the network to have connections in common,whether they are

relatedornot.Transitivityindicateswhethertworegulatorsthatshareastrongtie

27Categoricaldataonstafffiguresbyregulatoryauthoritiesreferringto2016ontheACER

website,

http://www.acer.europa.eu/official_documents/other%20documents/acer%20taking%20st

ock%20of%20the%20regulators’%20human%20resources%20summary%20of%20findings.

pdfIcombinedthe“Large”and“MediumLarge”categories,asthelatteronlycontainedtwo

regulatoryauthorities(lastaccessed3November2017).

130

are alsomore likely to have sharedpartners thanwould be expectedby chance.

Hence, inthisnetworkthere isaweaktendencytohaveconnections incommon,

which becomes significantly higherwhen two regulators are connected. In other

words,iftworegulatorshaveaconnectionincommon,theyhavehigheroddstobe

connected by a strong tie, as well. Finally, there are signs that the network is

centralizedonnodeshavinghighout-degree,i.e.havingmanyoutgoingties,while

the parameter for centralization of the network around highly influential nodes,

althoughpositive,failstoachievesignificance.Thissuggeststhatregulatorscluster

aroundactivenodes,butnotaroundinfluentialones.Plausibly,regulatorsuseactive

nodes to increase theiraccess to information, including theiraccess to influential

nodes.

Asfortheotherpredictors,themodelsshowthatregulatorsthataremembersto

thesameRegionalInitiativeforElectricityaremorelikelytobeconnected,andthat

thedirectionoftheelectricityflowsacrosstheEUmirrorsthedirectionalityofties

acrossregulators.GasRegionalInitiativesandgasflowsneverachievesignificance.

Nevertheless, theseeffectsdonotsufficetoexplainthenetworkstructure,which

appearsevenmorestronglydetermined,judgingbytheparametersinthemodels,

byVoC-relatedvariables.ThecoefficientsoftheERGMshouldbeinterpretedaslog

oddsoftheprobabilityofatieexisting,giventhefeatureinvestigated.Inthenetwork

describedbymodel2,theprobabilityofatieexisting(equivalenttotheinterceptin

aregression) is0.4%.Itthereisamutualtie,theprobabilitybecomes17%.Iftwo

nodesarebothfromCoordinatedMarketEconomies,theprobabilityjumpsto40%.

Inordertochecktherobustnessoftheresults,Irunothermodels(notshownhere

butavailableonrequest)testingforhomophilywithrespecttotheextentofmarket

liberalizationasoperationalizedbythemarketshareofthelargestgeneratorineach

131

country28.Moreover,Iusedthesamevariabletocheckforwhetherregulatorsfrom

more liberalizedmarketshavehigheroddsofreceiving incomingoroutgoingties.

Thevariabledidnotprovesignificant,however.Additionally, IusedOECDdataon

networkindustriesregulation(Koske,Wanneretal.2015)concerningtheextentof

governmentownershipinthelargestfirmactiveineachofthesegmentsofboththe

electricityandgassectors:generation(productionorimportforgas),transmission,

distribution,andsupply.Themodelsdetectedhomophilybetweenregulatorswith

lowergovernmentownershippercentagesinelectricitygenerationanddistribution

(nosignificanceforgas).Finally,whenre-runningthemodelsusingadifferentVoC

categorisation,whereIrelandappearsasLME(asperHallandGingerich2009)and

Member States that entered the EU in 200429, 200730 and 2013 (Croatia) are

clusteredasasingle“Other”category,theresultsarevirtuallyunchanged.

Modelfit.

TheERGMdefinesaprobabilitydistributionacrossallnetworksof thesizeof the

networkinthemodel.Ifthemodelisagoodfittotheobserveddata,thennetworks

drawnfromthisdistributionarelikelytoresembletheobserveddata.Asmentioned,

ERGMsaregenerativemodels.Theyrepresenttheprocessoftieformationfroma

local perspective: the perspective of the factors that are used to predict tie

probabilityintheERGMitself.Theselocallygeneratedprocesseseventuallyproduce

networkproperties,evenifthosearenotspecifiedinthemodel.Onewaytoassess

28 Eurostat energy statistics http://ec.europa.eu/eurostat/web/energy/data/main-tables

(lastaccessed3November2017)

29Cyprus,theCzechRepublic,Estonia,Hungary,Latvia,Lithuania,Malta,Poland,Slovakia,

andSlovenia.

30BulgariaandRomania.

132

thefitofamodel,then,istoexaminehowwellitreproducesnetworkpropertiesthat

arenot in it.Oneobtainsanassessmentofmodelfitbycomparingthevalueofa

network statistic of choice between the original network and the simulated

networks. I calculated goodness of fit over a series of network characteristics:

edgewise sharedpartnerdistribution,minimumgeodesicdistance, in-degree,and

out-degree. The four plots emerging from the simulation from Model 2 (the

preferredmodel)arereportedinfigure2.Model2appearsabletocapturenetwork

structureconsiderablywell,givenitsparsimonioussetupandclarity.Thedarklines,

correspondingtoobservedvalues,fallintheboxplots(simulatedvalues)fornearly

allconfigurations.

Figure4-2-GoodnessofFitofModel2:indegree,outdegree,edge-wisesharedpartners,minimumgeodesicdistance

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

in degree

prop

ortio

n of

nod

es

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

0.0

0.1

0.2

0.3

out degree

prop

ortio

n of

nod

es

0 1 2 3 4 5 6 7 8 9 10 11 12

0.0

0.1

0.2

0.3

0.4

edge−wise shared partners

prop

ortio

n of

edg

es

1 2 3 4 5 6 7 8 NR

0.0

0.1

0.2

0.3

0.4

0.5

0.6

minimum geodesic distance

prop

ortio

n of

dya

ds

Goodness−of−fit diagnostics

133

Discussionoftheresults.

Twomainfindingsemergefromthisanalysis.Firstly,regulatorsappearmorelikelyto

maintainconnectionswithregulatorsfromtheirsameVarietyofCapitalism;thisis

particularlythecaseforregulatorsfromCoordinated,NordicandDependentmarket

economies. Secondly and simultaneously, the regulator from theonly LME in the

dataset (the UK) has much higher odds of receiving ties, given its peripheral

geographic location, than its peers. Given that the whole EU energy regulatory

frameworkandrelevantlegislationarebasedonaLME-typemodeofcoordination,

this finding can cautiously be interpreted as a manifestation of regulatory

convergence.ThecombinationofthesetwopatternsappearstosupportThatcher’s

(2007)analysis,whichdiagnosedconvergenceofotherVoCtowardstheLMEmodel,

atleastinnetworkindustries,butatthesametimenotedthepersistenceofextant

modesofcoordination.RegulatorsfromMixedMarketEconomiesaretheleastlikely

tobesignificantlyhomophilousintheirtiechoices;thissuggeststhatthevarietyofa

regulators’connectionsisasimportantastheiractualnumber.

AlthoughcoefficientsarehigherforregulatorsfromCMEsandLMEs,allregulators

appear significantly more likely to receive incoming ties than regulators in the

reference category (DMEs), bar regulators from countries where government

ownership and control across the whole energy sector is prevalent. These two

categoriesofregulators,forthemostpart,belongtonewerMemberStates,which

entered the EU after 2004. In short, as far as transnational energy regulatory

networkingisconcerned,thereappearstobeadividebetweenregulatorsfromEU-

15andregulatorsfromnewerMemberStates.

Overall,theanalysisofthisnetworkstructureappearstotellastoryofpolicylearning

drivenbyexpertiseaswellasbycommonalityinsectorstructureandtherefore,as

permyhypotheses,commonchallenges.TheapproachIadoptedindevelopingthe

134

modelisveryconservative:Iincludedseveralendogenousdependenciestoaccount

for ties that exist because of structural properties of the network, rather than

homophily;IalsoincludedcommonmembershipinEuropeanRegionalInitiativesand

electricity and gas cross border flows in order to test the strength of regulators’

associations against a powerful constraint for network industries, as geography.

Furthermore,coordinationandcollaborationbetweenEuropeanenergyregulators

hasalonghistory,datingfromthelate1990s,andisembeddedinaverydeveloped

andwell-formedsupranationallegislativeandregulatoryframework.Thesefeatures

show in the network structure, which is overall dense and comprises a single

component.Indeed,thatanyeffectisvisiblebeyondthosecontrolsistellingofthe

strength of the national political economyof the sector as a driver of regulatory

networking.

Asregulatorsareplacedattheinterfacebetweentheirnationalmarketsandmarket

playersandtheEuropeandimension,theybuildtheirinformalbilateraltiesaccording

tobothhomophilyand,arguably,convergence. Interestingly, theBritish regulator

alsoemergesassignificantlymoreactivethanitspeersinmodel5,allelseequal.This

findingresonateswithThatcher’s (2007)remarkthatBritishpoliticianshaveoften

complainedoftheslowprogressofliberalizationreformsinotherMemberStates.In

truth, theBritish regulatory authority has repeatedly expressed concernover the

effectthatthelowerextentofliberalizationinotherEuropeanmarketsmayhaveon

British consumers, and underlined its leading role in the network of European

regulatorsinvirtuallyallofitsannualreportstotheEuropeanCommission,released

since200731.

Model2ispreferredforitsparsimony.Indeed,evidenceregardingtheimportance

ofresourcesforexplainingnetworktiesisnotconclusive,eventhoughsuggestiveof

31 Annual Reports of National Regulatory Authorities can be downloaded at

https://www.ceer.eu/eer_publications/national_reports(lastaccessed3November2017).

135

interestingpatterns.Usingcategorical insteadofcontinuousdataforstaff figures,

which splits regulators into groups according to the number of their full time

equivalentstaff,avoidsthecollinearitydrivenbytheveryhighnumbersofstaffof

the British regulatory authority. Results show that regulators with intermediate

numbers of staff are most likely to be active networkers. Regulators with small

numbersofstaffappearlessunabletocultivateanextendednetwork.Irunother

modelswhichindicatehomophilyamongregulatorswithlargenumbersofstaff(i.e.

>170),i.e.regulatorswithlargenumbersofstafftendtonetworkwitheachother.It

bears pointing out that, perhaps surprisingly, regulatory authorities with large

numbersofstaffarenotnecessarilythosefrombiggerMemberStates;thatsubgroup

comprises regulators from the UK and Germany, but also fromHungary and the

CzechRepublic.Theenergyregulatoryauthorityofa largecountry likeFrancehas

between40and90fulltimeequivalentstaffunits,endingupinthemediumgroup.

In short, the impact of resources on the likelihood of tie existence is difficult to

discernclearly.Ifurtherexplorethelinkbetweenresourcesandactivisminthenext

chapter.

Thecoefficientsforendogenousdependenciesindicatethattransitivityisaproperty

ofthisnetwork,asisrelativelycommonininformationexchangenetworks(Fischer,

Ingoldetal.2017).Moreover,themodelshowsthatsomeregulatorsaremoreactive

thanothers,hencehavemoreoutgoingties,causingthenetworktobecentralized

onout-degree,eventhoughthecoefficientisnotalwayssignificantacrossmodels.

Theseactivenodesareplausiblybridgingacrossthenetwork,whichwouldotherwise

comprisesomeisolatednodesordisconnectedcommunities.

The literaturehas found that veryoftenpolicynetworksdisplay a core-periphery

pattern(Knoke1990,Knoke1996,Carpenter,Esterlingetal.2003),wherebythereis

acohesivecoreofdenselyconnectednodesandaperipherywhosemembersare

poorlyconnectedbothtothecoreandamongthemselves.Theintuitiveconception

of core-periphery structures entails a dense, cohesive core and a sparse,

136

unconnected periphery. Core-periphery structures have been investigated in the

literatureonnetworks(BorgattiandEverett2000)aswellasintheliteratureonthe

EuropeanUnion (Magone,Laffanetal.2016). In thecontextof theEU, thesame

concept has been applied to frame relations between “old” and “new”Member

States(BohleandGreskovits2012).

Finally, the irrelevance of gas in explaining the patterns of European energy

regulators’networkingisapuzzlingresultsoftheanalysis.Neithergasflows,norgas

regionalinitiativesappeartohavestatisticalorsubstantivesignificancewithregard

to thisnetwork.Thismaybeduetothe lowercontrol regulatorshaveonthegas

sectorandthedevelopmentofgasmarketscomparedtoelectricity.

Conclusions

Themainquestionthischaptersoughttoanswerconcernedthedriversofinformal

regulatory networking at transnational level. Specifically, the empirical case

examinedinthischapteristhatofEuropeanNationalEnergyRegulatoryAuthorities,

taskedwithregulatingtheelectricityandgassectorswithintheirnationalborders

and simultaneously asked to coordinate in order to bring about regulatory

harmonization and foster market integration across the EU. Far from being

straightforward,thistaskisriddenwithdifficultiesandsetbacksas,astheliterature

onthesemattershasoftendiscussed,MemberStateshavedifferentadministrative,

legalandinstitutionaltraditions.Differencesintheirpoliticaleconomyarelikely,and

have been shown to be equally important not only in affecting the design of

regulatoryinstitutionsbutalsotheirnetworkingpractices,astheyseektofulfilthe

tasksbestoweduponthem.

Ihypothesisedthat regulatorswouldtendtochoosecounterpartsembedded ina

similar political economy as their most frequent and stable network partners. I

137

operationalizedthisconceptthroughtheVarietiesofCapitalismframework,slightly

modifiedtotakeaccountofthespecificcharacteristicsandstructureoftheelectricity

andgassectorsinthecountriesconsidered.Ialsohypothesisedthatregulatorsfrom

LMEswould receive significantlymore ties, since theEuropeanenergy regulatory

policyandlegislationareshapedaccordingtothatmodeofcoordination.Finally,I

expected resources to also matter for regulators’ networking choices, as less

resourcefulseektofilltheirinformationalgapsbylinkingtomoreresourcefulones.

TheresultsshowthattheVoCframework iswellsuitedto investigateandexplain

transnational regulatory networking, and able to make sense of the interplay

betweentransnationalinterdependenceandnationalcircumstances.Thehypothesis

concerninghomophilyisconfirmedformostcategoriesofregulators.Thehypothesis

concerningtheinfluenceofLMEisalsoconfirmed,andcanbeinterpretedasasign

of ongoing convergence, on the background, however, of persisting dynamics of

coordinationtypicalofnationalpoliticaleconomies,asfoundinThatcher(2007).The

hypothesised linkbetween lower resourcesandhighernetworkactivism failed to

emergefromanalysis:rather,mediumsizedregulatoryauthoritiesappearasmore

likelytobeactivecomparedtotheirmoreorlessendowedcounterparts.Moreover,

higher budgets are associated with more outgoing ties, suggesting that more

resourcedregulatorshavemoreresourcestodevotetonetworking.

Theendogenousnetworkdependenciesindicatethattiesinthisnetworktendtobe

reciprocal;theeffectforthisdependencyisverystrong,representinganimportant

validationof thedata as I didnot specify aminimumor amaximumnumbersof

nominations to the respondents. Moreover, regulators tend to close triangles,

particularlywhentwonodesarealreadyconnected;thistestifiestotheimportance

ofinformationexchangerelationshipsinfosteringtrustamongtheactorsinvolved.

Theobservednetworkstructurefeaturesahandfulofregulatorshavingconsiderably

more outgoing ties than their peers; to account for this, I included in themodel

138

dependenciesaccountingfornetworkcentralizationaroundactivenetworkers.The

effectisstrong,butnotconsistentlysignificant.

Overall,regulatorsfromnewerEUMemberStates(i.e.thosewhoenteredfrom2004

onwards) appear less integrated into the network structure, suggesting a core-

periphery pattern. Regulators from DependentMarket Economies appear as the

moststronglyhomophilousandtheleastlikelytoreceiveincomingties.Regulators

fromcountrieswheregovernmentownershipandcontrolacrossallsectorsegments

predominatealsoemergeaslesslikelytoreceivetiesthantheirpeers.Thepresence

ofscarcelyconnectednodesinthisdensenetworkofregulatorssuggestthatforms

of structured cooperation, such as the European Agency for the Cooperation of

Energy Regulators, are probably necessary in order to achieve energy market

integrationintheEU.Structuredcooperationimpedestheformationofcliquesor

disconnectedcommunitiesof regulatoryauthorities,andencourages learningand

exchangealsoacrosswidelydifferentinstitutionalcontexts.

Concluding,thischaptercarriesoutananalysisoftheglobalstructureofthenetwork

connectionsconsidered;exceptionsarenotableacrossallcategoriesofregulators.

Forinstance,theregulatorsofHungaryandLatviaaremuchmoreactivenetworkers

thansomeoftheirEU-15counterparts.Thepurposeoftheanalysiswasgraspingthe

invisible and undocumented drivers of transnational networking; this entails the

importantlimitationoftheimpossibilityoftriangulatingdatawithothersourcesof

information.Thisconcernisassuagedbythepractitionerknowledgeofthepersons

whokindlyagreed,underpromiseofanonymity,torespondtomyinquiry.Further

researchmayseektostudynetworkevolutionovertimebyrelyingonlongitudinal

data,asthismayhelpcapturingtheenginesofphenomena,onlycautiouslyalluded

tohere,suchasconvergence(orlackthereof).

139

5. Networking for resources: how regulators use

networkstocompensateforlackingresources.

Introduction

Thischapterunpackstherelationshipbetweenregulators’resourcesandtiesinmore

detail. Theanalysis in theprevious chapterwasnot conclusiveon thispoint. The

hypothesis concerning regulators’ resources expected regulators with lower

resourcestobemoreactivenetworkers,inthesenseofpossessingsignificantlymore

outgoingtiesthantheirpeers.Theanalysis includedcontinuousdataonstaffand

budgetary figures (dating back from 2013) and a categorical measure of staff

resources (dating from2016).The lattermeasure is thought tobemoreaccurate

sincethedatacollectiononregulators’tieswascarriedoutin2015/2016.Theresults

oftheExponentialRandomGraphModels(ERGMs)showednosignificanceforthe

continuous measurements. As for categorical data, the results showed that

regulators with medium resources are more likely to be active compared to

regulators with large resources. Regulators with medium small, small or micro-

resourceswerenotsignificantlymoreactivethantheirpeerswithlargeresources.

I investigatetherelationshipbetweenstaffandbudgetaryresourcesandnetwork

ties by testing the samehypothesis of linear relationship between resources and

activism,butbyoperationalizingnetworkactivismdifferently.Iintroducearefined

notion of network activism that draws on the sociological literature on social

networksandoutlinesnetworkactivismasthepossessionofhighernumbersofweak

ties(Granovetter1973).

Ipositthatregulatorsconceiveoftheirinformaltiesasresources,inaverypractical

sense: their peers are repositoriesof information andexperiencewhich they can

easilyaccess.Ithenconsiderthattheregulatorswhomostneedtorelyonnetwork

140

resourcesarethosepossessinglowerin-houseresourcestofulfiltheirinformation

requirements.Theseregulatorsarelikelytoreachouttotheirpeersmorefrequently

thantheirpeersarelikelytoreachouttothem,particularlyifthelatterhavehigher

resources.Thisimpliesthatactive,orproactivenetworkershavehighernumbersof

non-reciprocatedties.

Non-reciprocated ties indicate an overall weaker relationship than reciprocated

ones.Inthecontextofexchangeofinformationwithinnetworks,theysuggestthat

sendersare information-seekers,whohaveperhaps less information toexchange

thanwithinareciprocalrelationship.Stemmingfromthisreasoning,Iusethecount

ofEuropeanenergyregulators’weaktiestooperationalizenetworkactivism.Iuse

bothcontinuousandcategoricalmeasurementsofstaffandbudgetsaspredictors.I

carry out linear regression analyses to test the significance of the correlation

between resources and activism. I then include indicators of independence and

market liberalization in the regression. Thereafter, I run ERGmodelswith a new

specificationofresources,whilestillincludingtermsrelatedtohomophilybasedon

commonVarietyofCapitalism,asemergedinthepreviouschapter.

Theresultsoftheanalysislendsupporttothehypothesis:overall,lowerresources

are significantly associated with higher network activism. The significance of the

association is different, however, for different categories of budgetary and staff

resources:mediumandsmall levelsofstaffresourcesandsmall,butnotmedium,

budgets are associated with higher activism. Overall, however, the relationship

betweenstaffresourcesandactivismistheonlyoneofthetwothatisconsistently

significant.

Recentliteraturehasfeaturedcalls(MastenbroekandMartinsen2018)andattempts

(Maggetti2014,Ruffing2014,BoegerandCorkin2017)toshiftscholarlyattention

fromdebatingtherationaleofEuropeanregulators’collaborationaccordingtothe

preferencesofeithertheEuropeanCommissionortheMemberStatestostudying

thewaysinwhichregulatorsconcretelyusetheirinformalnetworks.Severalrecent

141

contributions have begun exploring these aspects and found that network

participation increases bureaucratic autonomy (Danielsen and Yesilkagit 2014),

thanks to the information exchange occurring within regulatory networks. Other

contributions showed that participation into networks correlates with increased

powersfortheregulatoryauthority(Maggetti2014),andfacilitatesthesharingof

existingresourcesandthecreationofnewonesasoutputsofnetworkcollaboration

(Vestlund2015).

Thetopicofregulators’resourceshasoftensurfacedintherelevantliterature,but

hasrarelybeentackledinitsownmerit.Yet,resourceconstraintssignificantlyaffect

regulatoryperformance,pushingregulatorstooptimizetheresourcestheyhavein

thefaceofincreasingsectorcomplexity(Glachant,Khalfallahetal.2013).However,

higher staff numbers are associated with higher quality regulation (Koop and

Hanretty 2017). Hence, whereas well-resourced regulators are likely to possess

sufficient expertise and means to acquire the information they need, less well-

resourcedregulatorsmaystruggletoaccomplishtheirtaskswhilealsostayingahead

oftheinformationcurve.

Although the resources available to a regulatory authority positively correlate to

country andmarket size (suggesting that regulators from smaller countries need

lowerresourcestobeginwith),allregulatoryauthoritieshaveasetrangeofexpert

taskstobefulfilled,whatevertheirsize(PollittandStern2009,Glachant,Khalfallah

etal.2013).Therefore,lessresourcedregulatoryauthoritiesmayrecurtotheextra

resourcestheyhaveavailable,suchastheirpeers,tocompensate.

Tounderstandwhethernetworkscanimprovegovernance,itisimportanttostudy

how regulators use them and the benefits they derive from them. This analysis

suggests that regulators use their bilateral network ties to compensate for their

lackingresources.Arguably,Europeanenergyregulators’commonembeddednessin

the EuropeanUnion and its single legal and regulatory framework are important

structuralpremisesfortheexchangeofinformationandexpertisetooccur.Absent

142

thatinterdependence,theonsetofamechanismwherebylessresourcedregulators

can attempt to compensate for their lacking resources via networkingwithmore

resourcedpeersappearslessplausible,sincetheassessmentoftheworthofanon-

reciprocaltieis,bydefinition,unequalbetweenthetwonodes.Finally,theresultsof

this analysis suggest that, if European regulatorynetworksmayhavehaduneven

influence on regulatory convergence, they probably have a consistent impact on

improvingregulatorypracticeatnationallevel.

Networkingforresources:literaturereviewandhypothesis.

Oneofthekeyquestionsintheliteratureonregulatorynetworksconcernstheextent

towhichtheyareabletoimprovegovernance(CoenandThatcher2008,Kelemen

andTarrant2011, Levi-Faur2011,MaggettiandGilardi2011,Thatcher2011,Van

Boetzelaer and Princen 2012,Maggetti andGilardi 2014, Bianculli, Jordana et al.

2015,BlaubergerandRittberger2015).Assessmentsofnetworkeffectivenesshave

often tried to capture their impact on regulatory convergence,whichhas proven

hardtodiscern(Bach,DeFrancescoetal.2016).Theliteraturehasoftenmaintained

thatregulatorynetworksmayimprovegovernancebyimprovingregulators’practice

andstrengtheningtheirprofessionalethosbyaffordingthemaccesstohighquality

informationandexpertiseheldbytheirpeers(Bianculli,Jordanaetal.2015,Jordana

2017,Papadopoulos2017).

InthecontextoftheEuropeanUnion,theliteraturehasconsideredthenecessityof

collaborationandexchangesofinformationbetweennationalregulatoryauthorities

asalmostobvious,giventheexigenciesofregulatoryconvergencetocompensatefor

thenecessaryvaguenessoftheEuropeanDirectives,whichboundMemberStatesto

goalsbutnottomeans(Dehousse1997,Nicolaides2004,EberleinandGrande2005).

Until recently, this focus has been predominant in the literature on European

RegulatoryNetworks.Contributionsontransnationalregulatorynetworkssteepedin

143

the international relations literature, however, have flagged the issue thatmere

participation intonetworks revealsnothingofhow individual regulatorsuse their

networks(BachandNewman2014,Ahdieh2015).Networkmembershipneednot

implyactivism.

In agreementwith that statement, this paper does not study networking among

European energy regulators in terms, functional to the achievement of European

integrationgoals;rather,thischapterstudiestheinformal,voluntary,bilateralties

thatregulatorsentertainwitheachotherbesidesthecontextortherequirementsof

the EU. I am interested in “pick-up-the-phone” relationships, i.e. those informal

collaboration ties that regulators maintain besides the meetings, schedules, and

requirementsofEuropeancoordination,withthepreciseintentofconsultingeach

other onmatters of their daily, national regulatory practice. The interest of this

research is exploring the innerworkingsof actual network relationships between

national regulators,which is aprominentgap in this literature (Mastenbroekand

Martinsen2018).

Recently, the literature has begun filling this gap by exploring the benefits that

regulators obtain by networking with their European peers. The relevant

contributionshavemostlyfocusedonindependenceandautonomy(Thatcher2011,

Yesilkagit2011,DanielsenandYesilkagit2014,MaggettiandVerhoest2014,Ruffing

2014).Severalcontributionshavementionedtheimportanceofadequateresources

forregulatoryauthoritiestobeabletoperformtheirtasks(CoenandThatcher2008,

MaggettiandGilardi2014)butdidnotconnectthemtonetworking.Maggetti(2014)

investigated the hypothesis that participation in European Regulatory Networks

correlatedwithanincreaseinthebudgetaryresourcesmadeavailabletonational

regulators.He foundnosupport for thishypothesis.Vestlund (2015)explores the

connectionbetweennetworksandresources,but inadifferenttakethantheone

adoptedinthispaper:inthatcontribution,networksareproducerofresources,i.e.

theoutputsofregulatorynetworking,suchasreports,meetingsandworkplans.In

144

contrast, this contribution, focuses on financial and human resources, i.e. those

resources that the regulatory authority needs in order to perform its duties, and

whethertheycanbethoughttoaffectthepatternofregulators’networking.

To assess the relationship betweennetworking and resources, as in the previous

chapter,IconsidertheempiricalcaseofthenetworkofenergyregulatorsfromEU

Member States (plus Norway). The data consists of regulators’ self-reported

information on their most frequent and regular bilateral relationships with their

EuropeanpeersinresponsetoaquestionnaireIsubmittedtothembetween2015

and 2016. The data gathering process aimed specifically at investigating the

regulators’ perceptions of theirmost frequent, regular collaborators among their

European colleagues, i.e. the colleagues they would call upon for suggestions,

exchangesofopinions,advice,andexperiences.

Thusfar,thefewcontributionsrelyingonnetworkdatahavelinkedthepossession

ofmanytiestonetworkinfluence(MaggettiandGilardi2011,Maggetti2013),inline

withthekeytenetsofnetworktheory.Hence,severalcontributionshaveconcluded

thatregulatorshavinghighdegree(i.e.ahighernumberofties)comparedtotheir

peers are influential. Examples include Ingold, Varone et al (2013),whomeasure

reputationacrossthevariousbureaucraticandpoliticalactorsinvolvedintheSwiss

telecommunication sector; Maggetti, Ingold et al (2013), who examine the

independenceandaccountabilityofdifferentregulatoryauthoritiesinSwitzerland;

and Alexander, Lewis et al. (2011), who explain the networking strategies of

politiciansandbureaucratsacrossmultiplemunicipalgovernments in thestateof

VictoriainAustralia.Thesecontributionshaveexaminedtiesasundirected.

Yet,thedirectionalityoftiesisasinformativeastheirnumber:beingatthereceiving

end of many ties may indicate influence, while sending many outward ties may

indicate activism (Desmarais and Cranmer 2012). The two analytically distinct

conceptsofinfluenceandactivismhaveoftenbeenconflatedundertheassumption

that influential regulators are also active networkers (Maggetti 2014). This

145

contributionassessesactivismnot intermsof theabsolutenumberof ties,but in

termsofthenumberofoutgoingtiesthatexceedregulators’strongest(i.e.mutual)

relationships, as this indicates additional effort to gather information from the

network.

In his seminal contribution on the strength of so-called “weak” ties, Granovetter

(1973)showedthatthevastmajorityofhisintervieweeshadfoundtheirjobthanks

toinformationreceivedfromthefriendsoftheirfriends,withwhomtheyhadonly

infrequent contact. In that article and in the conspicuous literature it spawned,

“weakties”bridgingacrosscliquesofstrongties(e.g.closefriends)havethusbeen

showntoplayacrucialroleininformationdiffusionacrossnetworks(Djelic2004).By

obtainingnovel,non-redundantinformation,networknodeswithweaktiesareable

tobridgeacrosscliques(BerardoandScholz2010)andtomakebetterdecisionsand

innovate more effectively (Aral 2016). In Granovetter’s research, respondents

reached out to their infrequent ties in the pursuit of (in that case, job-related)

information.Informationisthemainassetregulatorspursue;givenresourcescarcity,

theymaypursue itviaothermeans.Networkingwiththeirmoreresourcedpeers

appearsaparticularlysuitableandrelativelyinexpensiveone.

In social surveys, respondents are typically asked to nominate their contacts

accordingtothefrequencyoftheirinteractions.Theabsenceofreciprocitysuggests

aweakerrelationthanwouldbeinferredfromreciprocalnominations.Inthecontext

ofthisinquiry,theabsenceofreciprocitybetweentworegulatorsindicatesdifferent

assessmentsofthefrequencyoftheirrelationship.Activismisconceptualizedasthe

active pursuit of information from peers, given one’s resources.More resourced

regulatorsarelikely,inabsolutenumbers,tohavemoretiestotheirpeers,precisely

asa resultofhavingmoreresources tospendonnetworking.The interestof this

paper,however,isinvestigatingnetworkactivismrelativelytoresources.Todoso,I

operationalizenetworkactivismas thedifferencebetweena regulator’soutgoing

andincomingtiesor,inotherwords,theirnon-reciprocalties.

146

Acountofnon-reciprocaltiesequaltozerosuggeststhattheregulatorstickstotheir

strongestties.Anegativecountsuggestsinfluence(i.e.thefactofbeingsoughtby

others in the network more than one seeks them). In contrast, the fact that a

regulatoryauthorityhasapositivenumberofnon-reciprocalrelationshipssuggests

thattheyconsiderrelativelyweakertiesas,still,importantresourcesofinformation

andadvice.Hence,Iunderstandthepresenceofnon-reciprocaltiesasindicatingan

effort, on the part of the sender of the tie, to peruse its environment formore

information than the one available within the clique of their strong ties. I thus

contend that regulators reporting non-reciprocal outgoing ties have a stronger

incentive than others to seek information from peers. Given the importance of

resourcestoanyorganization,Iformulatethehypothesisthatthisincentivederives

fromlackofsufficientresourcestocarryouttheirregulatorytasks.Therefore,the

mainhypothesisunderlyingthisanalysisis:

H1: The lower the resources of the regulatory authority, the higher its network

activism.

Thetopicofregulatoryindependencehasbeenverywidelydiscussedandanalysed

in the public policy and public administration literatures, including in relation to

regulatorynetworks (DanielsenandYesilkagit2014,Monti2014,Ruffing2014). In

thatregard,severalcontributionshaveshownthatnetworksempowerregulatorsby

providingthemwithinformation,notavailabletotheirdomesticpoliticalprincipals.

By virtue of their intermediary position between levels of governance, the

information collected via networking expands regulators’ autonomy in the

transnationalpolicyspace(BachandRuffing2013,Ruffing2014).Arguably,theeffect

ofnetworkingonindependencemaypushlessindependentregulatorstobemore

activenetworkers.Therefore,Iincludeameasureofregulatoryindependenceinthe

analysis,toassesswhetheritmatterstoexplainactivismalongwith(orinsteadof)

resources. The indicator refers to regulators’ statutory, orde jure independence,

147

whichmaydifferfromactualindependence(Maggetti2007)butstillhasbeenfound

tobepositivelycorrelatedtoit(Hanretty2010).

Furthermore,theextentofmarketliberalizationinagivenregulator’scountrymay

alsoaffectitsnetworkingbehaviour.Indeed,theEuropeanenergymarketmodelis

premisedon liberalizationandthe introductionofprivatecapital in infrastructure

sectors(JamasbandPollitt2005,Thatcher2007).Hence,regulatorsoverseeingless

liberalized markets may seek interaction with peers overseeing more liberalized

marketsastheyattempttofoster liberalization intheirownnationalsettings.For

thisreason,Ialsoincludeintheanalysisindicatorsofmarketliberalization.However,

theseare likelytocorrelatewiththesizeofthemarket,which, inturn,correlates

withtheresourcesavailabletoanationalregulatoryauthority.

DataandMethods

Inordertogatherdataonregulators’networkties,Iemailedtheenergyregulatory

authoritiesfromallEUMemberStates,plusNorway.Allrespondentsareinformed

peopleattheirregulatoryauthority,whohaveexcellentknowledgeof itsexternal

cooperationpatternseitherbecausetheyareinchargeofsupervisingitorbecause

theyoccupyseniorpositions.Iaskedregulatorstonametheregulatoryauthorities

theyaremostoftenintouchwithasconcernsexchangeofinformationwithinthe

EU. I specified that they should mention their most frequent informal contacts,

beyond routine interaction and schedulednetworkmeetings or EU-relatedpolicy

events.

Underpromiseofanonymityoftherespondents’identityaswellasoftheirreplies,I

obtainedrepliesfromall29regulatoryauthorities,barone.Forthatone,Iconsidered

theirincomingties,asresultingfromotherregulators’nominations,asreciprocated.

Ididnotspecifyanupperorlowerlimitsonthenumberofnetworkpartnersthat

148

regulators could name as their most frequent contacts, in order to capture the

differentextentsofindividualregulators’networks.Moreover,Ialsolefttheprecise

frequencyofcontactunspecified.Regulatorswereaskedtoreportontheir“most

frequent” contacts. Clearly, the word “frequent” may mean different things to

differentrespondents.Also,differentregulatorsmayengagemoreorlessoftenwith

peersfromothercountriesbasedonavarietyoffactors.Then,reciprocityisafirst

good indicatorof the strengthof two regulators’ relationship. Lackof reciprocity,

instead,suggestsimbalanceinthetworegulators’assessmentofthefrequencyof

therelationshipand,thus,aweakerone.

Yet, the sender of a non-reciprocal tie, by the very fact ofmentioning the tie in

responsetomyquestion,indicatesthattheyassessthatrelationshipasimportant.

Possessionofnon-reciprocatedoutgoingties,then,isaproxyofamorepronounced

networkactivismcomparedtoregulatorswhoonlyengageinmutualrelationships

(anactivismofzeroaccordingtomyoperationalization)ordisplaynegativevalues,

suggestingthatinformationissoughtfromthemmorethantheyseekinformation

fromothers.A regulatoryauthoritycanbe,ofcourse,veryactiveandhavemany

mutualties.Thiswouldimplyahighinvestmentinnetworkcollaboration.However,

itwouldalsoimplypossessionofsufficientresourcestosupportabroadportfolioof

regularcontacts.The interestofthisanalysis iscapturingwhetherregulatorswith

lowerresourcesaremoreactivethantheirresourceswouldleadtoexpect,asthis

wouldsuggesttheyusenetworkstocompensatefortheirresources.

WhileIgatheredoriginaldataonEuropeannationalenergyregulators’ties,Irelyon

twomainsourcesofsecondarydataforstaffandbudgets:forstaff,Iusecategorical

data,asperthereportonNationalRegulatoryAuthoritiesresourcesreleasedbythe

European Agency for the Cooperation of Energy Regulators (ACER) in 201632; for

32 ACER Taking stock of the regulators’ human resources - Summary of findings (2016),

https://bit.ly/2GQnzFP(lastaccessed9April2018)

149

budgets,IrelyonfigurespublishedinthecountryreportsreleasedbytheEuropean

Commissionin201433,forlackofmorerecentdata.

TheACERdocument categorising regulators’ staff levels comprises six categories:

“large”(over220FullTimeEquivalents);“medium-large”(between170and175FTE);

“medium” (between90and140 FTE); “medium-small” (between50and75 FTE);

“small”(between12and50FTE)and“micro”(fewerthan12FTE).Asshowninthe

regressionmodel in Table 4 in the Appendix, category “large” is not significantly

different from “medium-large”. Since the whole dataset only comprises 29

observations,andsincethecategory“medium-large”onlycomprisestwonational

regulatoryauthorities (from ItalyandSpain), Imerge this category into“large” in

ordertosavedegreesoffreedom.Moreover,Isplitregulatorsinthe“medium-small”

category between group “medium” and group “small” on the basis of their staff

figuresin2013(derivedfromtheEuropeanCommissioncountryreports).Hence,the

categorizationofstafffiguresusedintheanalysissectioncomprisesfourcategories:

“large”(thereference,with9observations),“medium”(with9observations),“small”

(with 7 observations) and “micro” (with 4 observations). In further re-

categorizations,Imaintainedcategory“medium-small”separateormergedcategory

“micro” into “small”, thus reducing the number of categories to three: “large”,

“medium”and“small”;theresultsoftheanalysisdonotchange.

Althoughbudgetaryfiguresmayhavechangedsince2013,theyareunlikelytohave

changedveryconsiderably. Initially, Icategorizedbudgetfigures inavariablewith

five levels, ranging from “large” (over 20 million euros; only 3 observations) to

33 European Commission, (2014), “EU Energy Markets in 2014”,

https://ec.europa.eu/energy/sites/ener/files/documents/2014_energy_market_en_0.pdf

(lastaccessed9April2018)

150

“micro”(lessthan2millioneuros;5observations)34.Thiscategorization,however,

resultedinover-fitting;allregulatorsaremuchmorelikelytobeactivenetworkers

thanthethree“richest”authoritiesinbudgetaryterms.Inadditionalcategorizations,

Icreatedasingle“large”categorymerging“large”and“medium-large”budgets(thus

comprisingregulatorswithbudgetshigherthan10millioneuros);Ialsoreducedthe

numberofcategoriestothree(“large”,“medium”and“small”):theresultschange

very slightly and concur in showing that regulators with small (but notmedium)

budgetsaremorelikelytobeactive.

As for thecovariates, Iderivedataconcerningregulators’ independencefromthe

OECDSectorRegulationindicators,releasedin2015butreferringtotheyear201335.

Theindependenceindicatoristheaverageofthreeindicators:onereferringtothe

regulator’saccountability,onetotheirregulatorypower,andonetotheextentto

whichtheyhavetotakeinstructionsfromtheexecutiveintheirregulatorypractice.

TheOECDdatafeaturestwoseparateindicatorsforindependence;oneforelectricity

regulation and one for gas regulation. Although all of the European regulatory

authoritiesthatregulategasalsoregulateelectricity,theirscoresforelectricityand

gas regulation may differ36. For this reason, I include each indicator in separate

models.Further,I includeameasureofmarketliberalization(fromthementioned

OECDSectorRegulation indicators) toaccountforthepossibilitythat itmaydrive

activism,aswell.

34 Othercategoriesare:“medium-large”(between20and10millioneuros;7observations);

“medium”(between10and5millioneuros;10observations);and“small”(between5and2

millioneuros;4observations).35 OECD, (2015), Indicators of Sectoral Regulation, http://www.oecd.org/gov/regulatory-

policy/indicators-sectoral-regulation.htm(lastaccessed9April2018). 36 Indeed, the Pearson’s correlation coefficient between the independence gas and

electricity indicators is 50% and their Spearman’s correlation coefficient is only slightly

higher.

151

To test the association between resources and network activism, I carry out an

AnalysisofVariance(ANOVA)wherebyIregress,firstly,stafflevelsandthenbudget

levelsonnetworkactivism.Secondly,IrunOrdinaryLeastSquaredmodelsincluding

the mentioned covariates (concerning independence, liberalization and

infrastructurelinks).Beforerunningregressions,however,itisusefultoprovideplots

describingtherelationshipbetweenthevariablesofinterest.

Analysis

Figure1andFigure2,depicting,respectively,therelationshipbetweenstafflevels

and activismandbudget levels and activism. Figure1 shows that regulatorswith

intermediatelevelsofresourcesaremorelikelytobeactivenetworkerscompared

totheircounterpartswithlargeorverysmallresources.Theupperpartoffigure2is

basedonacategorizationsettingasideregulatoryauthoritieswithveryhighbudgets

fromtheirpeers:allothercategoriesshowmuchhigherlevelsofactivism.Sincethe

group with much higher resources than the rest of European regulators only

comprisesthreecountries, I relyonadifferentcategorization inthe lowerpartof

figure2,whichrevealsalinearrelationshipwherebyasbudgetsdecrease,activism

increases.

152

Figure5-1–Relationshipbetweenstafflevelsandnetworkactivism

153

Figure5-2–Relationshipbetweenbudgetlevelsandnetworkactivism

154

Beforeassessingthesignificanceof thedifferencesbetween levelsofactivismfor

regulatorswithdifferentlevelsofbudgetsandstaff,Itestforthesignificanceofboth

factorvariables,asawhole,inexplainingthedependentvariable.Todoso,Icarry

out an analysis of variance (ANOVA). The results of the Levene’s tests on the

relationshipbetween,respectively,stafflevelsandbudgetlevelsonnetworkactivism

arenotsignificant.Thismeansthattheassumptionofhomogeneityofthemeans,

necessarytocarryoutanalysisofvariance(ANOVA),isnotviolated.Hence,Icarry

outANOVAsforbothindependentcategoricalvariables.Theresultsshowthatthe

variable related to staff levels, whatever the categorization used, is statistically

significant (i.e. p-values are lower than 0.05). In contrast, the variable related to

budgetaryfigureisnotstatisticallysignificant.

Table1andTable237 report theresultsofOrdinaryLeastSquaredmodelshaving

networkactivismasdependentvariableand,respectively,staffandbudgetlevelsas

themainpredictorinthefirstmodel.Subsequentmodelsinclude,onebyone,the

covariatesselectedforanalysis.

Table5-1-Associationbetweenregulators’networkactivism(dependentvariable),

stafflevelsandcovariates

Model1 Model2 Model3 Model4 Model5

(Intercept) -2.00* -1.82* -1.76* -1.35 -1.67

(0.79) (0.77) (0.81) (0.91) (0.81)

Staffcategory

large reference

37TheresultsoftheregressionanalyseswerecarriedoutinRandexportedusingpackage

“texreg”(Leifeld2013).

155

medium 2.89* 2.65* 2.63* 2.37 2.42*

(1.11) (1.09) (1.13) (1.16) (1.15)

small 3.29* 3.18* 2.91* 2.39 2.68*

(1.19) (1.15) (1.22) (1.34) (1.25)

micro 2.25 1.70 1.79 0.28 1.69

(1.42) (1.41) (1.46) (2.02) (1.85)

Independence(electricity) -0.73(0.45)

Independence(gas) -0.55

(0.47)

Liberalization(electricity) -0.87(0.64)

Liberalization(gas) -0.74

(0.49)

R2 0.28 0.35 0.32 0.33 0.36

Adj.R2 0.20 0.25 0.21 0.22 0.24

Num.obs. 29 29 29 29 27

RMSE 2.36 2.28 2.34 2.32 2.35

The indicators for independence, while not significant in table 1, are statistically

significant in conjunctionwithbudget levels and show that regulatorswith lower

independencearemoreactive. Lackof significanceon thestaff coefficientswhen

regressed together with the extent of electricity market liberalization is due to

collinearity:regulatorswithsmallerstaffresourcestendtobefromsmallercountries

andhavesmallermarkets,which,inturn,tendtobelessliberalized.Irunthesame

regressions using different categorizations of staff levels, and the results are

156

remarkablyconsistentlyshowingthatintermediatelevelsofstaffareassociatedwith

higheractivism.

Table 5-2 - Relationship between network activism (dependent variable), budget

levelsandcovariates

Model1

Model2

Model3

Model4

Model5

(Intercept) -1.10 3.61 4.65 0.54 2.39 (0.80) (2.21) (2.87) (2.90) (2.48)

Budgetlevels large reference

medium 0.90 0.44 0.22 0.80 0.04

(1.13) (1.07) (1.11) (1.16) (1.25)small 1.85 3.00 2.20 1.48 1.27

(1.49) (1.47) (1.41) (1.64) (1.51)micro 3.10* 2.77* 2.81* 2.28 2.82

(1.38) (1.29) (1.31) (1.98) (1.84)Independence(electricity) -1.57*

(0.69) Independence(gas) -1.89*

(0.91) Liberalization(electricity) -0.39

(0.66) Liberalization(gas) -0.82

(0.56)R2 0.18 0.32 0.30 0.19 0.29

Adj.R2 0.08 0.21 0.19 0.06 0.16Num.obs. 29 29 29 29 27

RMSE 2.52 2.34 2.37 2.56 2.47

***p<0.001,**p<0.01,*p<0.05

157

Itbearsnotingthat,sinceIamusingnetworkdata,linearmodelsmaybeunreliable

because they assume the independence of the errors. Typically, network data

featuresinterdependencies.MethodssuchasExponentialRandomGraphsModels

(ERGMs)havebeendevisedspecificallytodealwiththoseinterdependencies.The

ERGMs run in the previous chapter showed that regulators with medium staff

resources appeared to have significantlymore outgoing ties than regulatorswith

large resources. No other category reported significant differences. I run further

ERGMmodels,whichincludedthevariablesthatprovedsignificantintheprevious

chapter (i.e. homophily for Variety of Capitalism, an effect for incoming ties

dependingonVarietyofCapitalism,electricityflows,electricityRegionalInitiatives,

anddependenciesforhighoutdegree,reciprocity,sharedpartnersandtransitivity)

andstafflevelsasoriginallypresentedintheACERdocumenttheyaredrawnfrom.

Next, I run the same model including my re-categorization of staff levels that

comprisedfourinsteadoffivecategories.Inbothcases,resultsarenearlyidentical

and show that, all else equal, regulators withmedium and small staff levels are

significantlymore likely tohavemoreoutgoing ties than their peers, thus largely

confirmingtheresultsofthestatisticalanalysisaswellastheresultsintheprevious

chapter.By the same token, regulatorswith large staff resourcesare significantly

morelikelytoreceivemoreincomingtiesthantheirpeers.Incontrast,ERGmodels

including the categorical variable for budget levels do not result in significant

statisticalcoefficients.

Discussionoftheresults.

Theresearchinterestmotivatingthisanalysisisinvestigatingthedeterminantsofties

between regulators within regulatory networks. Regulatory networks are the

protagonistofarichliteraturewhich,however,hasrarelyinvestigatedtheincentives

driving ties across network members. Contributions on networks of European

regulators usually focus on their potential or actual influence on regulatory

158

convergence,or,alternatively,onestablishinglinksbetweennetworkmembership

andoutcomessuchasincreasedindependence.Thiscontribution,instead,attempts

toassessthequalityofnetworkmembershipandtheconcreteusagethatregulators

makeoftheirinformalcollaborativetiestotheirpeers.

The results of the analysis suggest that resource imbalances across European

nationalenergyregulatorsmatterfortheirnetworkbehaviour.Theclearresultisthat

regulatorswithlargeresources,bothinstaffandbudgetarylevels,arelessproactive

thantheircounterpartsandtendtobeatthereceivingendoftheiroutgoingties.

Well-resourcedregulatorshaveresourcesthatareseveraltimeslargerthanthoseof

someof their counterparts. The small sizeof thedatadoesnot allow for ahigh-

powered analysis. Yet, the relationship between numbers of staff dedicated to

energyregulation(i.e.theACERcategoricaldata)andnetworkactivismisconsistent

andsubstantive(withanetasquareabove0.50)and,inthecategorizationincluding

category “micro” (i.e. regulators with less than 12 FTE dedicated to energy

regulation)showsaquadratictrend.

This suggests that lower resources correspond to higher activism up to a point:

regulatorswithextremelysmallresourcesarenotsignificantlymoreproactivethan

theirpeerswithlargeresources.Thissuggeststhattheseregulatorsaresoresource-

constrainedthattheopportunitycostofnetworkingistoohighintermsofstafftime.

Regulatorswithintermediatestaffresources,instead,maynothavealltheresources

they need but still they have enough for networking to be a cost-effective

compensatorystrategy.Additionalmodelsconsideringtheadditiveeffectsofbudget

and staff resources yield no significance for budget levels. There are too few

observationstotestwhetherstafflevelsmatterdifferentlyatdifferentbudgetlevels

(and vice-versa) or, in other words, to test for interaction effects. At any rate,

budgetary and staff levels are very highly correlated (Spearman’s correlation

coefficient of 0.62). Yet, only staff levels are consistently significant; after all, a

regulatorynetworkisasocialnetworkofindividuals.Theirnumberseemstoaffect

159

the extent to which a regulatory authority can strive to compensate for lacking

resourcesbyrelyingontheirpeers.

Theseresultsleadtoinferthatoneofthemainvirtuesofregulatorynetworksmay

wellbetheirimpactongovernanceatnationallevel.Asamatteroffact,thescope

ofEuropeannationalregulators’authorityisconfinedtothebordersoftheircountry.

Steepedinverydifferentmarketcontexts,nationalregulatorsneedtosettheright

incentivesfortheregulatedindustrytoachieveefficienciesandre-investitsprofit,

monitortheirconduct,whilepreventingabuses,ensuringconsumerprotectionand

empowerment,whileremainingaccountabletogovernmentandthegeneralpublic.

Theseare tremendous tasks; in their fulfilment, regulators facegreat information

asymmetrieswith the regulated industry (Pérez-Arriaga 2014). At the same time,

regulators need to preserve their legitimacy and credibility as adequately

accomplishing their tasks; otherwise, they face government intervention and,

potentially, thecurtailmentof theirpowers. Facedwith low resources, regulators

tackledthedemandsoftheirprofessionbyrelyingonafurtherresource:theirpeers.

Therefore, the understanding of transnational regulatory networks’ potential for

improving governance should be extended beyond transnational or global

governance to encompass governance improvements at national level. In other

words,transnationalregulatorynetworkingmayhaveimportantfeedbackeffectsat

national level, which need not necessarily push in the direction of convergence.

Access to informed and resourced peers may improve the problem-solving and

analyticalcapacitiesofregulatoryauthorities(LodgeandWegrich2014)atnational

level,eventhoughbarriersremaintotransnationalregulatoryconvergence,which

maydependonfactorsbeyondthecontroloftheregulatoryauthorityitself.Inthe

European context, the results of this analysis suggest that the interdependence

createdbydesigningthesinglemarket(includingtheinternalenergymarket)may

haveimprovedregulatorypracticeatnationallevelbeyondwhatwouldhavebeen

achievablebasedonnationalresourcesalone.

160

Conclusions:networkingforresources.

The literature has claimed that membership is a regulatory network does not

automaticallyguarantee regular involvement (BachandNewman2014).Although

agreeingontheimportanceofnetworksforinformationexchangeamongregulators,

fosteringtheirlearningandevenincreasingtheirdefactoindependencefromtheir

principals,theliteraturefailedtospecifythemechanismwherebyregulatorsnetwork

forexpertise.Thispapermakesa step in thedirectionof tackling this issueusing

original data on European energy regulators’ ties to each other (gathered under

promiseofanonymity)andraredataontheirbudgetaryandstaffresources.

The analysis tests the relationships between the difference between regulators’

outgoing and incoming ties, or, in other words, the number of regulators’ non-

reciprocalties,andtheextentoftheirresources.Non-reciprocaltiesareunderstood

as indicating the active pursuit of information beyond one’s strongest ties (i.e.

reciprocatedties).Thehighernumbersofnon-reciprocatedties,themoreactivethe

regulatoryauthority.Possessionofnon-reciprocaltiessuggeststhattheregulatory

authoritybranchesoutbeyondtheirclosestcontactsinthepursuitofinformation,

much like job-seekers contacted the friends of their friends in the seminal

Granovetter’s article on the strength of weak ties for information diffusion

(Granovetter1973).

Information is the main resource regulators need in order to fulfil their tasks.

Resource-constrained regulators are less able to set the right incentives for the

regulated industries (Glachant, Khalfallah et al. 2013) and therefore less able to

recognizeandpunishtheirwrongdoings.Moreover,stafflevelshavebeenfoundto

correlatewithhigherqualityregulation(KoopandHanretty2017).Thehypothesis

guidingthisanalysisisthatsomeregulatorsaremoreactivenetworkersbecausethey

use their informalnetworks to compensate for their scarce resources in termsof

budget and staff. The results lend support to the hypothesis. Indeed, regulators

possessinghigh resources tend tohavenegativeactivism, i.e.negativedifference

161

between incoming and outgoing ties, pointing to their influence and popularity

amongtheirpeers.

AlthoughtheinterdependencecreatedbythecommonmembershipintheEuropean

Union is plausibly at the root of these informal collaboration patterns, the

mechanism whereby regulators use networks as substitutes for their lacking

resourceshasalreadybeenidentifiedelsewhereintheliterature(seeAlcañiz2016

on the case of nuclear experts in Latin America). This analysis suggests that the

benefits of transnational collaboration do not only improve European regulatory

governance but also spill over to national regulatory policy, providing for more

informedregulatorsthanitwouldhavebeenpossibletoachieveonthesolebasisof

domesticresources.

Althoughlackingempiricaldataconfirmingtheargument,thischaptercontributesto

theory development regarding how regulators use their networks. Hence, this

analysiscanspawnfurtheranalysesassessinghypothesesconcerningtheeffectof

networksonregulatorydecision-making.Forinstance,acasestudyanalysisfocusing

on a relatively less resourced regulatory authority and testing the causal chain

leading from network exchanges to actual decision-making against empirical

evidencewouldrepresentasuitablefollowuptothisanalysis.

A finalword concerning theavailabilityofdataonnational regulatory authority’s

resources:gatheringthisdataisexceedinglydifficult,asnocentralplatformappears

tocollectitonaregularbasisandmakeitpubliclyavailable.Nationalreportsoften

donotcontainthis informationandoftenarenotdraftedinotherlanguagesthan

thenationalone.Asa result, researchusing thisdatahas to relyon impreciseor

slightlyoutdateddata.Makingthistypeofinformationmorereadilyavailabletothe

academicandscientificcommunitywouldnotonlyimproveresearch,butprobably

alsocorroborateclaimsforadequateresourcestobeprovidedtonationalregulatory

authorities.

162

PartII.Conclusion:therationaleforregulatorynetworking.

Thissecondpartofthethesisshiftedthefocusofanalysisintwoways:itmovedfrom

the historical narrative to the present and from networks as organizations to

networks as structures resulting from regulators’ voluntary connections to one

another. This shift enabled analysis of the determinants of regulatory network

collaboration. The explanatory model of network structure reveal that a strong

patternofhomophilydrives regulators’ connections.Moreover, themodel shows

thatregulatorsoverseeingmoreliberalizedmarketsreceivemoreincomingties,but

alsohavesignificantlymoreoutgoingones.Finally,theanalysisinchapter5showed

that energy regulators with higher resources are more likely to receive more

incomingties,andthatregulatorswithlowerstaffresourcesaremorelikelytobe

proactivelyusingnetworkstocompensatefortheirlackingresources.

Thepolicyimplicationstobederivedfromthejointconsiderationsoftheseresults

areseveral.Ontheonehand,thestronghomophilypatterndeterminingnetwork

tiessuggeststhatsomeformofstructuredcooperation(suchasaEuropeanAgency

oramoreformalizednetwork)isprobablynecessarytofostercooperationamong

regulators, where the overall goal is market integration. Otherwise, network

interactionsmayresultindisconnectedorweaklyconnectedcommunities.Thisnot

onlywouldslowdowntheflowofinformationacrossthenetwork,butitwouldalso

conferremarkableleveragetobridgingnodes.Ontheotherhand,therelationship

betweenresourcesandactivismemergingfromchapter5showsthatregulatorsare

strategic in their usage of networks; moreover, it suggests that regulatory

cooperation in the European Union has probably had the effect of improving

governance at national level more than national resources alone would have

allowed.Discerning theneteffectofnetworkcollaborationonpolicyoutcomes is

exceedingly difficult, and I do not have data to empirically confirm the pattern

emergingfromthatanalysis.However,thechapterstrivestocontributetotheory

163

developmentconcerningthereasonswhyregulatorsnetworkandtotheformulation

ofhypothesesforfutureresearchconcerninghowinformationretrievedviainformal

collaborationhelpedregulatorydecision-making.

Thesignificanceoftimeasakeyfactorinexplainingtheoriginsandtheeffectsof

transnationalregulatorycooperationunderliesthiswholethesis.Thesetwochapters

utilizedata gatheredat aprecise timeand thereforeprovidinga snapshotof the

network of European energy regulators. Yet, Exponential Random GraphModels

suchastheonedeveloped inchapter4aregenerativemodels,whichmeansthat

they assume that the observed network is the result of an evolutionary process.

Some of the mechanisms visibly at play in the foregoing two chapters may be

idiosyncraticforindividualregulators;forinstance,agivenregulatoryauthoritymay

havebeenparticularlyproactiveinthetimeframeoftheresearch,andnotbethe

following year. Yet, themechanismsdriving tie choices, suchashomophily anda

compensatory strategy for lacking resources, are likely to be relatively stable

opportunitiesthatnetworksoffertoregulators.

Inthenextpartofthethesis,Ireverttohistoricalanalysisandtothecomparative

perspectivebyexaminingthephaseofnetworkexpansionforbothEuropeanand

American energy regulators. Although following entirely different evolutionary

paths,bothsetsofcooperationstructuresreplicatedthemselvesbygeneratingnew

networksinotherareasoftheworld.Iwillshowhow,facedwiththerelativedecline

of itsmain function, theNARUC resorted to a layering strategy by taking on the

additionalmissionofexportingUSutilityregulationtothefourcornersoftheworld.

This contrasts with the conversion strategy adopted by the CEER as the

establishmentoftheEuropeanAgencydepriveditofmuchofitspolicyworth;the

CEERfoundnewtopics (e.g.consumersandretailmarkets)andfunctions (e.g. in-

housetraining)tokeepitsviability.Theconditionsleadingtoeitherstrategyarealso

discussedinthenextpart.

164

PART THREE – EXPORTING NETWORKS: THEROLEOFAMERICANANDEUROPEANREGULATORSINFOSTERINGREGULATORYNETWORKSABROAD.

165

PartIII.Introduction:exportingnetworks.

This thirdand finalpartof the thesis consistsof twochapters focusingonwhat I

definethe“exportofnetworks”.Exportingnetworks, inthiscontext,referstothe

attempt to foster regulatory collaboration in a given governance context by

reproducing the informal network structure embedding successful regulatory

collaborationinanothergovernancecontext.Tobespecific,inthispartIinvestigate

the reasons why the US Agency for International Development (USAID) and the

European Commission through its European Neighbourhood Policy programme

decided to invest on fostering regulatory networked collaboration in foreign

jurisdictionsaspartof,intheformercase,aforeignpolicyagendaand,inthelatter

case,whattheliteraturedefinesas“externalgovernance”(Manners2002).

Tackling this question across the two chapters brought to the fore two other

important themes: the theme of regulatory network evolution and the theme of

networkentrepreneurship.Thenexttwochaptersfocusontheemergenceoftwo

networks,whicharetheprogenyof,respectively,NARUCandCEER;thesearethe

EnergyRegionalRegulatoryAssociation(ERRA)andtheAssociationofMediterranean

Energy Regulators (MedReg). The analysis shows that, although via different

processes,theNARUCandCEERwereessentiallyexportedtootherregions,inthe

explicitattempttoreplicatetheirsuccessformula.

Investigationsof the rationaleof suchnetworkexportyielded important research

findings as concerns the dynamics of network evolution, the interdependencies

betweennetworksemergedindifferentareasoftheworld,andthewaysinwhich

entrepreneurial regulators can leverage their embeddedness in the multi-level

governance system to further collaboration initiatives with strategically crucial

foreignjurisdictions.Thefollowingtwochaptersshowthatdonors(inthiscase,the

USgovernmentandtheEuropeanCommission)purposefullychosenetworksasan

166

instrument of “soft” foreign policy. This finding partially echoes some earlier

contributions,suchasRaustiala(2002),whoarguesthatnetworkscanbesynergistic

withmore formalpolicyprogrammes in implementing regulatory transferofbest

practices from more developed to less developed markets. Raustiala (2002),

however,wasconcernedwithregulatorycomplianceandmoreeffectiveregulatory

enforcement. In this setting, instead, the policy goals are primarily political. In

pointingexplicitly to foreignpolicymotivations, Iexcludealternativeexplanations

that,althoughplausible,areunlikelytobethemaindriverofnetworkestablishment

in the cases of ERRA andMedReg. Themain competing argument to the foreign

policyoneisthatthecreationofERRAandMedReg(theformerlocatedinCentral

andEasternEurope,thelatterinanareadefinedastheEuro-Mediterraneanregion)

happenedasaresultofaprocessofdiffusion.

Thekeyargumentofthe(voluminous)diffusionliteratureisthatwhencertainpolicy

or institutional solutionsbecomeacceptedwisdomtheydiffuseacross theworld.

This diffusion process can result from different mechanisms. The four main

mechanisms identified in the (rational-choice-inspired) literature are coercion,

competition, learning and emulation (ormimicry). In other words, countries and

governmentsreplicatepoliciesorestablishorganizations imitatingothercountries

becausetheycompetewiththemforcapitalandtrade(Elkins,Guzmanetal.2006),

becausesaidpolicieshaveproventodeliversuperioroutcomes(ShipanandVolden

2014),becausetheywanttobeseenas“modern”orcompetent(Meseguer2004),

or,finally,becauseexternalagents,suchasinternationalorganizations,forcethem

to do so (Henisz, Zelner et al. 2005). Amore constructivist explanation of policy

diffusion relies on the insights of sociology and on the concept of institutional

isomorphism(PowellandDiMaggio1983)andemphasisesthesymbolicproperties

of the adoptionof a policy that is generally thought to be effective and efficient

(McNamara2002).

167

The fact that transnational (or trans-jurisdictional) networks of energy regulators

havesprungupnearly ineveryregionoftheworld(BergandHorrall2008),often

withthesupportofinternationalfinancialinstitutions,appearstolendsupporttothe

diffusionhypothesis,particularlythecoercionandmimicryversions.However,the

concept of policy diffusion applies to policy solutions; informal networks of

regulators,perse,arenotapolicysolution.Networksareplatformsfordiscussion

andexchangeaboutpolicysolutions,notpolicysolutionsinthemselves.Regulators

maychoosetoleveragetheirnetworktoobtaincollectiveandindividualbenefits,as

outlinedinthefirstpartofthisthesis,butthemereexistenceofanetwork,e.g.of

an acronym and a meeting schedule, does not imply any actual policy

implementationorchange.Moreover,policiesdiffuse,orarediffused,becausethey

areperceivedtobegenerallysuperiorandefficient.

Bycontrast,theargumentsofforeignpolicyhaveadistinctsecurityflavour.Foreign

policyisaimedatachievingspecificpoliticalgoalsincountriesthatareperceivedas

a security threat to the donor. As the next chapter shows, thiswas the thinking

underlyingtheUSAIDeffortstoshiftthegovernanceparadigmoftheenergysector

inCentralandEasternEuropeintheaftermathofthecollapseoftheSovietUnion.

This included the establishment of regulatory authorities, according to a well-

established “recipe” of infrastructure sector governance that international

organizations worldwide espoused in those years (Henisz, Zelner et al. 2005).

Fosteringregulatorynetworking,however,wasnotpartofthatrecipe.TheUSAID

decided to foster the formation of a network of energy regulators in the area

autonomously, and for two reasons: the experience of NARUC had shown that

regulatory networks can foster learning across members, and, in the context of

technicalassistance, theywereacost-effectivealternativetoschoolingregulators

individuallyontheeconomicsofenergymarkets;also,adirectlinktoregulatorsin

different countries of the area allowed monitoring the reform progress and

accompaniedtheirtransitiontoamarketeconomyundertheaegisoftheEuropean

Union.Initsownturn,NARUCutilizedtheUSAIDpartnershiptorevampitselfandits

168

image.TheprogrammewassosuccessfulthatUSAIDdecidedtoreplicateitsformula

inotherregionsof theworldwhere itwasoperating, inpartnershipwithNARUC,

eversince.

The following chapter, chapter 7, reconstructs the formation of the network of

energyregulatorsfromcountriesfacingtheMediterraneanSea(i.e.MedReg).The

eventssurroundingMedRegwerecompletelydifferentfromthoseleadingtoERRA.

TheestablishmentofMedRegdidnottakeplacewithintheframeworkofabroader

foreignpolicyprogramme;asamatteroffact,itwasnoteventheinitiativeofthe

European Commission. MedReg was the result of a bold display of policy

entrepreneurialism by a small group of European regulators, in the wake of the

Europeaneastwardenlargementof2004and,mostimportantly,itssignatureofthe

EnergyCommunityTreatyin2005.TheTreaty,signedwithSouthEasternEuropean

governments, obliged those governments to reform their energy markets in

compliancewithEuropeanlegislationasaninterimsteptowardstheirfullaccession

asMemberStatesoftheEU.Contextually,itencasedregulatorycooperationinthe

ratherformalizedsettingofaBoardwithadvisoryfunctionstoaBoardofMinisters.

Atthetime,EuropeanregulatorsdisagreedwiththeCommissions’formalizationof

theregulatorycollaborativerelationships.Determinedtopreventthereplicationof

theEnergyCommunityapproachtotheSouthernneighbourhoodoftheEU,in2006

theItalianregulatorandafewSouthernEuropeancolleaguestooktheleadinthe

establishmentofaregulatorynetworkencompassingregulatorsfromallcountries

facingtheMediterranean,thusincludingregulatorsfromcountriesintheMiddleEast

andNorthAfricaregion:MedReg.

Theanalysisinthefollowingchaptersisbasedontheinsightsgainedfrom41semi-

structuredinterviews(someofwhichalsoinformedthewritingofchapters2and3;

thetableinAppendix4reportsthecorrespondinginterviewnumbers).Asamatter

offact,very littlewrittenpublic informationinavailableontheeventsrecounted.

Theinformationgainedthroughinterviewswastriangulatedacrossdifferenttypes

169

of interviewees (international organizations, regulators, consultants) and cross

checkedwithothersources(reports,newsletters,andelectronicarticles–quotedin

thetextandreferencedinthelistofreferencesattheendofthethesis).

Thesechaptersalsocontributetotheliteratureontransnationalregulatorytechnical

assistance programmes. In a recent contribution, Broome and Seabrooke (2015)

show that international financial institutions (IFIs - inparticular, the International

MonetaryFundandtheWorldBank)usetechnicalassistanceprogrammestofoster

theemergenceof “sympathetic interlocutors” in recipient countriesby socializing

national officials through transnational policy training. In thisway, these authors

argue,internationalorganizationsincreasethenumberofdomesticreformerswho

aresympathetictotheirprescriptionsforpolicychange.Indeed,astrongeremphasis

onsocializationfollowedtheIFIs’acknowledgementthatlendingconditionality,by

itself,wasnotachievingthedesiredresults(WorldBank2006).Thekeyimportance

of socialization with peers from more advanced economies as well as hands-on

cooperationoverconcretepolicyissueshasbeenproventodeterminethesuccess

oftechnicalassistanceprogrammes(Freyburg2015).

Someofthefindingsemergedelsewhereinthisthesisareconfirmedinthecasesof

ERRAandMedReg:primarily,thatuncertaintydrivesregulatorstoseekeachother

out toestablish informalcollaborative relationships tocorroborate theirdecision-

making and reduce the uncertainty they are exposed to; that symbiotic

interdependencedrivescollaborationbetweendifferentlevelsofgovernance,asin

the case of USAID and ERRA, when the fulfilment of one’s agenda entails the

strengtheningoftheother.Moreover,itshowsthattheevolutionaryphaseofone

networkcanrepresenttheemergenceofanother,asinthecaseofNARUCandERRA,

and that experience of network collaboration is a valuable asset in regulators’

relationships to their political principals. Finally, chapter 7 shows how well-

established regulatory networks may function as templates for policy

entrepreneurialismofanynetworkmember.

170

6. The establishment of the network of energyregulators of Central and Eastern Europeor Foreignpolicyvianetworks.

Inchapter2,Iidentifiedinthepost-WorldWarIcrisisofregulatorycredibilityand

legitimacythecriticaljuncturepromptingregulatorstoconsolidateNARUCastheir

collectivehome.Muchlaterintime,anothereventhadatransformativeimpacton

NARUC, representinga secondcritical juncture (CollierandCollier1991,Capoccia

andKelemen2007).In1998,theUSbilateralaidagency(USAID)proposedtoNARUC

acooperativeagreementtodeliverthegoalsoftheirmissioninEasternEurope.The

cooperative agreement was a manifestation of the new, post-Cold War guiding

paradigm of USAID’swork around theworld: from strategicallymotivated aid to

“transformativedevelopment”(Natsios2006).DuringtheColdWar,USAIDgaveaid

tocountriesirrespectiveoftheirhumanrightsorgovernancerecords,providedthat

theybecameorremainedtheiralliesagainsttheSovietUnion;aftertheColdWar,

the rationale for USAID aid provision changed radically and re-oriented its focus

towardsinstitutions.

Until USAID’s request, the international exposure of US state utility regulators

individuallyandofNARUCasawholewasvirtuallynon-existent,exceptforsporadic

interaction with Canadian counterparts. After the ERRA programme, USAID

partneredwithNARUCforregulatorycooperationprogrammesinotherareasofthe

world, including Asia and Africa. The federal grants that NARUC receives for its

nationalandinternationalworkrepresentnowoverhalfof itsannualbudget.The

consequencesofUSAID’srequestwerethereforefar-reachingandlong-lasting.Inits

absence, it ishighlyunlikelythatNARUCwouldhavemusteredtheresources, the

politicalsupportandtheinternalcohesiontodeveloptheseactivitiesindependently.

171

This chapter retraces the history of ERRA to unveil the under-investigated link

between technical assistanceandnetworks; it also represents the concrete route

takenbyNARUCtore-inventitselfandevolvebyacquiringnewclients(Downs1967).

TheestablishmentofERRA isacaseofbrokerednetwork fosteredbyanexternal

agent(seeProvanandKenis2008).Inthecontextofthecase,USAIDplaysasimilar

role to the European Commission towards the embryonic CEER: it was a policy

partnerplacedatahigherlevelofgovernance,pursuinganagendawhosefulfilment

was dependent on the regulators and entailed the strengthening of their policy

relevance.

TheemergenceofERRAisinscribedinthehistoryofNARUCandunfoldsinparallel

with that of CEER. Hence, the three cases are nested within each other. The

commonalitiesintheirdevelopmentarestriking:allthreeemergedinacontextof

radicalpolicyoverhaul;allthreestemmedfromloose,irregularprofessionalties;all

threefoundinthesymbioticinterdependencewithpoliticalactorsthekeytotheir

growthinsizeandimportance.However,theERRAdiffersfromCEERandNARUCin

onerespect:ERRAmembersareregulatorsfromseparatecountries,whichdonot

formasinglepolicyentity.AstheEuropeanenlargementenglobedsevenCentraland

EasternEuropeancountriesandtheirenergyregulatorsbecamemembersofCEER,

the taskERRAwas created forwasaccomplishedand themulti-level structureof

governance, comprising the USAID and national governments (as well as

International Financial Institutions and the European Union) underpinning it

vanished.TheERRAthatprovidedinputtopolicyformulationfortheregionvanished

as well. Still, the network did not disband. Regulators kept finding value in the

familiarity of the network and its focus on issues thatmattered for that region.

Hence,theERRAunderwentaprocessofconversionandtransformedintoadifferent

organization,bearingthesamename:primarilyatrainingproviderandplatformof

regulatoryinteractionswith,declaredly,nointentionofcontributingtopolicyorof

espousingaspecificpolicyagenda.ThisnewidentityenabledERRAtoexpandandto

172

welcomeanynewmember:indeed,ERRAmembersnowadayscompriseregulators

fromcountriesallovertheworld.

Bilateral aid and transnational regulatory cooperation: stranger

literatures.

TheinterventionofUSAIDinCentralandEasternEuropeinthe1990s,followingthe

demiseoftheSovietUnion,waspartofabilateralaidprogrammefinancedbythe

USgovernment. Thedeterminantsof recipient selectionbyaidproviders is a key

topicintheliteratureonforeignaidprovision(AlesinaandDollar2000,Berthélemy

and Tichit 2004,Neumayer 2005, Bermeo2008, Younas 2008, Bearce and Tirone

2010,Winters2010,Bermeo2011,Reinsberg2015,EichenauerandReinsberg2017,

Findley,Milneretal.2017).Thisliteraturehasconvergedonseveralfoundingclaims.

One of them concerns the higher politicization of bilateral aid compared to aid

provided by multilateral institutions such as the World Bank and the Regional

DevelopmentBanks.Bilateralaidagenciesarefinancedbyindividualgovernments,

which have biases and preferences concerning their aid recipients. The most

important bilateral donors include theUSA, theUK, Japan, France,Germany and

otherEuropeancountries(Neumayer2005).Ofthese,theUSAprovidesthelargest

sumsinabsoluteterms.

During the Cold War, strategic and foreign policy considerations dominated the

allocation of aid (Berthélemy and Tichit 2004, Natsios 2006), bringing USAID to

provideaidtocountrieswithdubiousorweakhumanrightsandgovernancerecords.

AftertheColdWar,however,allWesterndonorsbegantoexplicitlyconditionaidon

the quality of governance in recipient countries (Dollar and Levin 2006, Bermeo

2008).Eventhoughstrategicforeignpolicygoalsremainrelevanttoexplainbilateral

aid allocation (Milner and Tingley 2010), economic performance and institutional

qualityareatleastasimportant(AlesinaandDollar2000).Inhisassessmentofthe

173

rolethatgoodgovernanceindicatorsplayinaidrecipientselectioninthepost-Cold

Warperiod,Neumayer(2005)findsthattheonlyconsistentpatternacrossthemajor

worlddonors is the leveloftheregulatoryburden imposedontheprivatesector.

Respectforhumanrightsandtheruleoflawdonotemergeasequallyimportant.

Thequalityofthegovernancestructureofacountryisrecognizedasanimportant

determinant of the effectiveness of aid (Neumayer 2002, Dollar and Levin 2006,

Dietrich 2013). Donors have always used soft power techniques in order to elicit

compliancewiththeirdesiderataconcerningtheiraidprovision:Neumayer(2005)

lists persuasion, capacity building, policy conditionality and selectivity as the four

main strategies of donor soft power. Particularly after the ColdWar, the donor

community,andtheUSAinparticular,haveshiftedthetargetsoftheirsoftpower

efforts: not only governments, but also institutional actors. The USAID has been

foundtoprovidelessaidtopoorlygovernedcountries(Bermeo2008)orcountries

withweakgovernancestructures(EichenauerandReinsberg2017),unlessitcould

bypass government and target aid at non-state actors in order to increase the

likelihoodthatitwillachieveitsintendedoutcome(Dietrich2013).Thisistrue,more

generally, of bilateral donors from countries whose political economy places a

premiumonmarketefficiency,suchastheUSA,ratherthanonstate intervention

intotheeconomy(Dietrich2015).

Inotherwords,aftertheColdWarbilateraldonorsandtheUSAinparticularhave

startedrespondingtogoodgovernanceasasignaloftherecipient’scapacitytouse

aid effectively.Hence, recipient countries are rewardedwhen they show signsof

increased governance quality (Neumayer 2003). Since the quality of institutions

matters (Booth 2011), development aid has become increasingly targeted at

improving it, in particular through capacity building and technical assistance

programmes (Vandeveer and Dabelko 2001, International Competition Network

2003, Adams and Tirpak 2008, Urpelainen 2010, Marcoux and Urpelainen 2012,

BroomeandSeabrooke2015).

174

Bothbilateralandmultilateraldonorshave routinelyusedpolicy conditionality to

mandatecertaineconomicreforms inrecipientcountries.However,besidesbeing

blindtogovernanceindicators,duringtheColdWaraidwasalsoineffectivebecause

donor governments could not credibly enforce their conditionality (Bearce and

Tirone2010).Afterwards,thethreatofconditionalitybecamemuchmorecredible

andthereforetriggeredactualreforms.Overtime,however,donorsfullyrealizedthe

importance of a sympathetic institutional environment for correct reform

implementation(WorldBank2006).Therefore,theybeganinvestingresourcesinto

strengtheningthecapacityof institutionsinordertobuildlocalcoalitionsthatare

supportiveofdonor-promotedreforms(BroomeandSeabrooke2015). Asformer

USAIDdirectorNatsios(2006)explicitlyadvocates:“Asdevelopmentprofessionals,

weshouldbetryingtoinstitutionalizepressureswithinagivenpoliticalsystemthat

areconsistentwithoneanotherandthatworktofavourreformersandputthestatus

quoprotectionistsatadisadvantage.”(2006,p.137).

Thenotionof“transformationaldevelopment”,elaboratedbytheUSAIDintheearly

2000s (Essex 2013) is eminently focused on institutions and their potential for

entrenchingpolicychange inthedirectionofeconomicneoliberalism.The ideaof

transformativedevelopmentencapsulatesthestatedgoaloftransformingcountries

bytransformingtheirinstitutionalsetupthrough,inparticular,capacitybuildingand

partnershipwithdonors.ThelaunchofthisapproachtodevelopmentbytheUSAID

predates the State Department’s subsequent emphasis on “transformational

diplomacy”andajoint2007USAID/StateDepartmentstrategicplanincludingboth

concepts(Essex2013).InaspeechgivenatGeorgetownUniversity(DC)in2006,the

thenUSSecretaryofStateCondoleezaRiceremarked:

“Itwasalwaysassumedthateverystatecouldcontrolanddirectthethreats

emerging from its territory. It was also assumed that weak and poorly

governed states were merely a burden to their people, or at most, an

internationalhumanitarianconcernbutneveratruesecuritythreat.Today,

175

however,theseoldassumptionsnolongerhold.Technologyiscollapsingthe

distancethatonceclearlyseparatedrightherefromoverthere.(…)So,Iwould

definetheobjectiveoftransformationaldiplomacythisway:toworkwithour

many partners around the world, to build and sustain democratic, well-

governedstatesthatwill respondtotheneedsoftheirpeopleandconduct

themselvesresponsiblyintheinternationalsystem.”38

These words echo the stances of Slaughter (2017) but also Slaughter (2004),

Slaughter(2004b)andevenSlaughter(1997)withtheiremphasisontrans-national

cooperationandnetworksofinstitutionalactorsasapromisingavenueforfulfilling

the goals of American diplomacy. As shall be seen, the partnership programmes

implementedbytheUSAIDsincethemid-1990salreadybearthecharacteristicsofa

transformationaldevelopmentapproach,testifyingtotheentrepreneurialismofthe

organization(Essex2013).

Theliterature,however,lacksinvestigationofbilateralaidagencies’involvementin

thefieldofregulatoryreform.Mostimportantlyforthepurposesofthisresearch,

the literaturehasoverlookedthesignificanceofnetworks inthefieldof technical

development assistance and therefore the role that regulators may play in the

deliveryofcapacitybuildingandtechnicalassistanceprogrammes.Althoughmany

articlesonregulatorynetworksquoteSlaughter’s(2004c)contribution,theylackthe

foreignpolicy/diplomacyrationalethatinformsit.

Raustiala (2002) represents a notable exception. He suggests that trans-

governmentalcooperationofexpertsandregulatorsvianetworksissynergisticwith

government cooperation via international organizations. This author adds that

networkspromote"regulatoryexport":theexportofregulatoryrulesandpractices

from major powers to weaker states. This process, by facilitating regulatory

38 https://2001-2009.state.gov/secretary/rm/2006/59306.htm (last accessed 12 August

2017)

176

convergenceandbybuildingbureaucraticcapacity,canimprovedomesticregulation

and thereby enhance regulatory enforcement across the world. More recently,

Ahdieh(2015)hasremarkedtheopportunitiesthatcoordinationvianetworksoffers

forleveragingaregulators’ownpreferencesaswellastosignaltheextentoftheir

commitment.Thesecontributions,however,focusonthelegalandtheenforcement

side of regulatory cooperation, not on the eminently political aims that can be

pursued through networks. This chapter compensates for this lack of empirical

analysis.

“Consultantswerenotenough”:howtheNARUCbecameinvolved

intheUSAIDmissioninCentralandEasternEurope.

“Letmesetthecontext,that’stherightquestion.Thecontextwasback

to1990,1992…andCentralEuropeandtheformerSovietUnion.The

broader reforms, what we called the transition from a centralized

economy,centralizedcommunistgovernmenttodecentralizedmarket

economy and decentralized government required very fundamental

reforms. In terms of the economy itmeant breaking up centralized

monopolies,decentralization,intheelectricitysectorunbundlingwas

thetermthatwasused,politicallyitmeantbreakingupordestroying

a lot of institutions, procedures, processes… and… creating new

institutionsandradicallyrevisingexistinginstitutions”(interview29).

The USAID and the World Bank played an enormous role in accompanying the

transitionofCentralandEasternEuropeaneconomiesfromthecommunisttothe

capitalist economy. Privatization programmes were a very visible feature of the

transition (WorldBank1995,WorldBank2002),whoseessential thrust,however,

residedininstitutionalreform.Inallsectorsoftheeconomy,includinginfrastructure,

reformingorcreatinginstitutionsquicklybecameakeypolicygoal:ontheonehand,

177

newbodies–e.g.regulatoryauthorities–hadtobecreatedandstaffed;ontheother

hand, the reform effort consisted in introducing newmind-sets of economic and

socialorganization.

Indeed,energyutilitiesacrossCentralandEasternEuropeancountrieswerenotused

to conceive of utilities as businesses, whose commercial viability needed to be

ensured.Manyintervieweesconcurredinaffirmingthatengineeringandtechnical

competencesaboundedamongtheprofessionalsinthearea;themainobstaclewas

thatthey“hadnoideahowmuchakilowatthouractuallycosttoproduce,noidea

howmuch…installedcapacitycost….depreciationwasanalienconcept,whyshould

you ever allow an earned return, if you need more money you’ll get it from

somewhere.”(interview15).

It is important to note that an assessment of the success of electricity sector or

regulatoryreforminCEE/NIScountriesisbeyondthescopeofthischapter.Inactual

fact, theachievementsof the reformperiod in theelectricity sectorwere judged

unsatisfactory for most of the countries in the area insofar as competition is

concerned (vonHirschhausen andOpitz 2001, Krishnaswamy and Stuggins 2003),

eventhoughprivatizationshaveimprovedthequalityofservice(Vagliasindi2004).

TheaccessiontotheEuropeanUnionimprovedinvestmentlevelsinEasternEurope

(Vagliasindi and Izaguirre 2007). The countries of South East Europe have been

strugglingwithelectricitysectorreformforlonger;theestablishmentoftheEnergy

Community(Deitz,Stirtonetal.2009,HooperandMedvedev2009,Pollitt2009),a

precondition to these countries’ accession to the EU (Cambini and Franzi 2013),

improvedsectorperformancebutprogresshasbeenunevenanddonorsarestillvery

muchinvolvedinthearea(interview31).

178

Theutilityandregulatorypartnerships.

Inthecontextofpost-SovietCentralandEasternEurope,identifyingwhatBroome

and Seabrooke (2015) define sympathetic interlocutors was one of the primary

concerns.Theseareactorsinkeypositionsatvariouslevelofgovernancewhoare

willingandabletocollaboratewiththedonorsinordertocarryoutthereforms.

“Overtime,USAIDidentifiedprogressivechangeagentswithintheMinistries

that championed sector reform and led the Ministry’s transition from

operatingutilitytothepolicysettinganddatacollectionfunctions.Oncethey

were identified,USAIDprovidedthese individualswith legalandregulatory

technical assistance to draft new energy laws and shepherd their passage

throughtheparliamentaryprocess”(interview31).

The USAID organization comprises different regional Bureaus. Responsibility for

CentralandEasternEuropefellundertheremitoftheBureauforEuropeandEurasia.

The Energy and Infrastructure Division was in charge of carrying out the reform

programmeintheenergyinfrastructuresector(USCongress,1993,p.96).Inorder

to fulfil itsmission, theDivision adoptedan innovative approach, that laterwent

under thenameof“transformationaldevelopment”: investingnon-state (orpara-

state)actorswithpolitical support inorder to carryout thedesired reforms.The

scopeoftheUSAIDmissionintheareaisclearlyoutlinedina2007SpecialReport

focusedonitsworkinCentralandEasternEuropesincethecollapseoftheSoviet

Unionin1990.

“In this issue, we focus on a region where USAID has deployed a

transformationalapproachtodevelopmentforthepast15years:Europeand

Eurasia(E&E).ThefalloftheWallin1989andthecollapseoftheSovietUnion

in 1991 was followed by a US Government decision to provide assistance

throughUSAID to assistwith the transition to democracy and free-market

capitalism. (…) USAID focused most of its energy sector efforts on

179

transforming the economic and institutional foundations of the sector,

particularly the independent regulation and commercial operation of the

electricitysector.(…)TheE&EregionpresentedseveralfeaturesthatUSAID

hadnotencountered inother regions: virtually everyone in the regionhad

accesstomodernformsofenergy,andtheworkforcewashighlyliteratewith

impressive technical and scientific knowledge. (…) However, these services

were provided on a heavily-subsidized and non-commercial basis by

essentially self-regulating government departments. (…) With the end of

Sovietsubsidies,serviceprovisionbecameincreasinglyunreliableandlimited,

ineffectleavingcitizensinthedark”(USAID2007).

ThisReportmakesexplicitthelinkagebetweenenergyinfrastructureandsecurityas

well as between energy infrastructure and politics. The connection between

electricityandpoliticswasparticularlypoignant in formerlycommunist countries;

Lenin himself hadmade the electrification of the Soviet Union and the universal

provisionofelectricityoneofthepillarsofhispoliticalmessage.

“Theimportanceofpowersectorreformforsustainableeconomicgrowthis

widely accepted. What is not often acknowledged are the accompanying

political benefits. In the E&E region power sector reform is essential for

successfulpoliticalreformaswell.Powersystemswerecentraltoolsofformer

Communist governments. (…) Power sector reform and politics are

interwoven. Lenin’s aphorism about socialism plus electricity equaling

Communismreflectedelectricity’scentralrole”(USAID2007).

Giventhedirectlinkbetweenelectricityandcommunism,itwascrucialforUSAIDto

decoupletheprovisionofenergyservicesfromthestate.ThefirstmajorstepUSAID

undertook in this direction consisted in establishing the so-called “utility

partnerships”. Initially, these involved only four countries: Hungary, Poland, the

CzechRepublicandSlovakia.InitsreportontheprogressoftheUSinterventionin

theCEE/NIScountries,theUSCongressremarked:

180

“Underthisproject,AIDhasalsoenteredintoacooperativeagreementwith

USEA[UnitedStatesEnergyAssociation,aphilanthropicorganizationfunded

byelectricityandgasutilityexecutives] to form theU.S.-EasternEuropean

Utility Partnership Program (UPP). Begun in October 1991, UPP brings

togetherelectricutilitiesintheUnitedStatesandCentralandEasternEurope

for activities focused largely onmanagement issues but including also art

annual regulatory systems seminar and dissemination of information and

software resources. The Program has receivedwide support from theU.S.

utilityindustry,includingtheEdisonElectricInstitute,ElectricPowerResearch

Institute,andtheNorthAmericanReliabilityCouncil.U.S.utilitieswillshare

thecostsoftheprogram,contributing$1million,orone-fifthoftheprogram’s

budget. The first partnership was formed between New England Electric

SystemsandaHungarianpowercompanyinApril1992,andothersarebeing

developed between Commonwealth Edison and the Polish Power Grid,

HoustonLightingandPowerandtheCzechPowerCo.(CEZ),SouthernCo.and

theSlovakPowerUtility(SEP),andCentralMainePowerandtheBulgarian

NationalElectricCo. Furtherpartnershipsareplannedwith theBalticsand

Romania.AIDalsoreportsthatover20companieshaveexpressedinterestin

joiningthepartnershipprogram”(USCongress1993,p.100).

“ElectricallyknownasCENTREL”(interview31),theutilitiesofHungary,theCzech

Republic,SlovakiaandPolandwerethetargetoftheconcertedUSAID–USEAeffort

to improve their commercial viability, compromised by under-investment, scarce

revenuecollectionandtherefore inability to invest in infrastructuremaintenance.

Theprogrammealsohadastrongsymbolicimportance:theUSgovernmentwanted

to physically separate the electricity grids of these countries from the integrated

SovietgridandconnectthemtotheWesternEuropeanelectricalnetwork(interview

31).Bythemid-1990s,theCENTRELachievedthegoal,meetingamajorpost-Cold

Warpolicyobjective(interview31).

181

USAID’s strategy to foster cooperation and learningbetweenUS andCentral and

EasternEuropeanutilityexecutiveswasbasedonatwo-prongedapproach:onthe

onehand,USAIDprovided“traditional” technicalassistancebyhiringexperienced

contractorstoworkwithaidrecipientsoveranumberofyears,coupledwithshort

term technical assistance on selected topics; the other prong of the strategy

consistedoftheinnovativetwinningelementoftheutilitypartnerships.

“Whatthatdidwasgivethedecentralizedutilitiesthatweretryingtosortout

all kinds of issues direct contact and interaction for 3 or 4 yearswith aUS

commercialutilitysothattheycouldhaveadialogueintermsofspecificissues,

seehowtheyoperated,andthisinvolvedabout4exchangesayear;atthefirst

meetingtheutilitieswerepairedtogether,talkabouttheirissuesanddevelop

aplantomeettheneedsofthatunbundlingutilityandthentheywouldvisit

theUSforaweek,gotocertainspecificissues,thenextquartertheUSutility

wouldsendpeopleoverbaseduponwhattheywantedtotalkaboutsoitwas

decentralizedinthesenseofbeingverymuchdefinedbytheutility,whatthey

needed,butreallyfocused…”(interview29).

The decision to extend the partnership concept to the regulatory authorities

emergedrelativelysoonafterwards.“TheUSAIDledeffortsintheregiontoestablish

independentregulatoryagenciesasacountermeasuretopotentialbackslidingbythe

ministries” (interview 31) in order to remove political interference from the

operation of the electric power sector. As per the relevant literature, creating

regulatoryauthoritieswasdeemednecessarytorestrainthemonopolypowerofthe

newlyprivatizedutilitiesaswellastoenhancethepredictabilityandthecredibility

ofthecountries’commitmenttoreform(Majone,1997).Thefirstenergyregulatory

authoritiesintheregionwereestablishedin1994inHungaryandUkraine.Others

quicklyfollowedsuit.

“Now,inthedevelopmentofthoseorganizationswewerereallyfocusedon3

or 4 characteristics, we even put it down on a piece of paper, the AAAA:

182

Autonomy,Authority,Accountabilityandthenastimewenton,Ability,fourAs,

autonomy, separation from the utility, separation from the government;

authority tomakedecisions, set tariffs, approve licenses, andaccountability

throughopenpublicprocedures,removalforcauseonly,andthentheability,

thetraining,capacitybuildingskillsandpoliticalskills”(interview29).

USAID central managerial unit of the Europe and Eurasia Bureau convinced the

decentralizedUSAIDofficesineachcountrytocloselycooperateinordertoachieve

aconsistentapproachandforthenewinstitutionstodevelopatthesametimeand

learnfromeachother.ThesuccessoftheutilitypartnershipsencouragedtheBureau

toreplicatethesameschemewithregulatoryauthorities,hopingtoacceleratetheir

learningprocessgiventhespeedofthereformprocess.

“Sowe began again a two-pronged assistance approach. First the advisors

located in the country, workingwith the regulator, for passage of the law,

establishment of the organization, internal operations, development of

licenses,thoroughmethodologies,veryextensiveworkonthat,andconsultants

areveryusefulbutwefoundthat…regulators…werehearingdifferentthings

fromdifferentconsultantsandtheyhighlyvaluedtheconceptandapproachof

the energy regulatory partnership so we started doing that, partnering the

regulatorsincentralandeasternEuropewiththeUSregulators.Inordertodo

thatwedidacooperativeagreement,itwascalled,withNARUC,whothendid

manage, identifyandestablishthepartnerships.Whatdidapartnershipdo?

Againtherewas3or4visitsayeartostart,theregulatorswouldgettogether,

hear about each other, define their work plan and then have a series of

exchanges (…) sowe did that for 3, 4 or 5 years, that built a confidence, a

comfortanda trust thatalloweda lotof interactionbetween theEuropean

regulatorsandtheAmericanregulators”(interview29).

U.S.stateregulatorswerepartneredwiththeirnewlyestablishedcounterpartsbased

ontechnicalsimilaritiesintheirrespectiveenergysectors,theircommoninterestsin

183

energysectorrestructuringorsimplyonthebasisofasharedethnicity (interview

31). Initially, the regulatory partnerships focused on improving the institutional

capacity to manage regulatory institutions, focusing initially on the appointment

process for commissioners, ensuring independence from political interference,

developingorganizationalcharts,uniformsystemofaccountsandthepublichearing

processes.Asagenciesmatured, thepartnershipsbecame focusedon introducing

fundamentalconceptsofenergyregulation,e.g.costofserviceregulation,wholesale

electricitymarkets,renewableenergyintegration,crossborderelectricitytradeand

marketmonitoringfunctions(interview32).

Thepartnershipformatwasgreatlysuccessful:continuityofinteractionensurednot

only thegradual familiarizationofCentralandEasternEuropeanutilityexecutives

andregulatorswiththeneworganizationofthesector,butitalsohelpedUSAIDto

identifytheprogressiveagentsofchangethatcouldbetrustedtoendorsethereform

aimsintheirworkpractice:“therewasadegreeoftrustandcredibilitythatwehad

withpeople in thecountriesandthatwetrustedthemandknewwhoyoucould…

trustto…todothereforms”(interview29).

NARUC:theobviousintermediary.

Following the decision to implement regulatory partnerships, USAID needed to

decidewhether tomanageorganizational aspects itself orwhether to rely on an

externalbody.NARUCquicklyappearedastheobviouspartnerinthisendeavour,for

threemainreasons:thevalueofthepracticalexperienceoftheAmericanregulators

involved (their cognitive authority); the cost effectiveness of the solution (as

Americanregulatorswereonlyreimbursedfor their travels,butnotpaid for their

services);andthecoordinatingadvantagethatNARUChadassinglepointofcontact

forthewholeensembleofUSstatelevelutilityregulators.

184

“Iaskedthem[USAID]nottosendadvisorswhoarefocusingontheory“only”,

whoarenotfamiliarwithourregionandnotfamiliarwiththedailyregulatory

practice or the daily utility practice, just have elegant advice on it, on the

economic basic or… and they [USAID] followed our wish and they found

partnersontheutilitiessideandeachutilityintheregionhadapartnerwitha

US utility, andwe just copied thismodel whenwewanted to learn the US

regulatory practice it was evident that they should use the NARUC as a

facilitatorofknowledgetransfer/exchange”(interview25).

“Ibelievethat…thebasicconcepttherewasthat,sincetheutilityexecutives,or

utilitymanagers,werealreadyonsalary,andbecausethecommissionerswere

alreadyonsalary,basicallytheUSgovernmentreimbursedfor…outofpocket

expenses,travel,hotels,anyequipmentyouhadtobuy,aswellasamoderate

management fee for the sponsoring organizations, NARUC and USEA. In

retrospect I think itwas an innovative, creative, largely effective andpretty

cost-efficientmodelforthatregion”(interview15).

“TheobviousmanagersofthisprocesswasNARUC,theyaretheassociationof

the 50 state regulators, they had credibility, they knew the regulators, they

couldhelpregulatorsexplainwhythiswasintheUSinterest,intheinterestof

theregulatorsthemselvesaswellashelpingcountriesincentralandeastern

Europemakethetransitionsotheywereanobviouschoicetocarrythisout”

(interview29).

“In the Eurasia region, there were 6 countries where their public hearing

regulationsandprocedureswereadaptedfromUScommissioncounterparts.

Countriesdonotneedtostartfromablankpieceofpaper;itiseasiertolookat

a"goodpractice"andmodifyittofitthelocalcontextandneeds.”(interview

39).

185

USAID selected NARUC primarily for its access to reliable information across US

states.NARUCwasagaincalledupontoserveasaproviderof informationonthe

circumstances of each state, this time to export their learnt lessons to foreign

countries.NARUC’snetworkedstructureofcooperationwasthusputtoadifferent,

perhapsunexpectedusemanydecadesafteritsestablishment.Thisre-orientationof

purposetestifiestotheanalytical leveragetobegainedbystudyingandanalysing

networks along their evolutionary trajectory, rather than restricting the focus of

analysistotheirstatedaimsatthemomentoftheircreation.

TherewasonefurtherwayinwhichNARUCcouldservethepurposesofthemission:

byteachingregulatorshowtonetwork.Donorshadnoticedthatregulatorswere,

eachintheirowncountry,facingthesamekindsofissues.Moreover,justliketheir

EUcounterpartsinthoseverysameyears,theywerepoliticallyisolatedandstruggled

toestablishtheirlegitimacy.NARUCheldlong-standingexperienceofdefendingthe

regulators’institutionalturf.ItsfurthercontributiontoUSAIDmission,then,became

schoolingCEE/NISregulatorsabouthowtodealwiththeirdomesticinterlocutors.

“Therewearewithregulatorsin15,16countries.WhatIgettoseebyworking

inmanyofthemwassimilarissues,samequestions,samefrustrations,these

wereorganizationswithnopredecessor,nohistory,intermsofhowtodothings

beyondtheconsultant’sadviceandlookingattheUSprogram.(…)Forexample,

theUkrainian regulators,wanted to visit theGeorgian regulatorswhowere

doingtheexactsamethingastheywere,theyjumpedatthechancesowesent

3or4ofthemtoTbilisifor3or4daystotalktothem.Wedidthesamewith

theMoldovans,andthenIthinktheMoldovanstoRomania,andwebeganto

dotheseadhoconeoffvisitsandtheygreatlyvaluedit.ThenIcouldseewhere

thiswasheading,thevalueofthatexchange,andbeganthinkingabouthowto

bringthemtogethermoresowehadaworkshopinBudapestwherewegot11

or 12 countries together and had them each over their experiences and

presentations,discussions,andtheywesetuptwotechnicalworkinggroups

186

(…)ontariffsandlicenses(…).Sothatbeganonanadhocbasis,thetechnical

connection,butatthatworkshop4or5ofthechairmenstoodupandsaidwe

wouldliketoproposeandsupporttheestablishmentofaregionalorganization

(…)Sothatwas1997,Ithink,thatreallycrystalizedtheestablishmentofthe

regionalnetworkorwhatbecameERRA”(interview29).

TheestablishmentoftheEnergyRegionalRegulatoryAssociation(ERRA).

“Oneofthe,nowinretrospective,obviousobservationwasthattherewasno

peergroup,soyouhaveHungarysettinguptheHungarianEnergyOffice,you

haveBulgaria (…)withwhatwastheStateEnergyRegulatoryCommission,

youhavetheERGinPolandetcetera,andthesewere…wereobviouslyvery

committed, well-intended people, economists, lots of engineers, some

lawyers,butnopeergroup…andthe…theEUmodelthatwasemerging…of

associations of regulators aswell as the existingNARUCmodel in theUS,

whereyoucouldhavepeertopeerexchange,youcouldlookatbestpractices,

regionally, nationally, globally, you could facilitate dialogue. That model

helpedshapetheideathatthereshouldbearegulatoryassociationfornon-

EUmemberstatesthatweregoingthroughthistransition,foraspirationalEU

member states and to drive an ability to tackle common issues together”

(interview15).

By supporting theestablishmentof ERRA, theUSAID rendered the creationof an

informal cooperation network of energy regulators an essential part of its

operationalization of the policy goal of fosteringmarket institutions in the post-

Soviet space. Thekey reason for the involvementofUS regulators in themission

owedmore to their experience of network cooperation than to their regulatory

expertiseperse.Inotherwords,consultantscouldanddidprovideampleadviceon

187

variousregulatorymodels;buttheycouldnotcoachregulatorsonhowtocooperate

witheachother.NARUC’scomparativeadvantageresidedinitsnetworkexpertise.

“Asthenumberofregulatorybodiesintheregionincreased,USAIDsupported

networking among the national regulators to exchange experiences and

information. This led to a strong bottom up demand for an ongoing

institutional arrangement that allows regulatory bodies to continue their

exchanges. As a result, the EnergyRegulators RegionalAssociation (ERRA)

was established in Budapest, HungarywithUSAID support. The regulators

anticipated the evolution of the reformprocess toward regional electricity

marketarrangementsthatrequiredcross-borderregulatorycommunication

and cooperation. To support the establishment and development of ERRA,

USAIDformedaCooperativeAgreementwiththeU.S.NationalAssociationof

Regulatory Utility Commissioners (NARUC), an association of 50 state

regulators.NARUChasprovidedavaluablelinkfortheE&EregulatorstoU.S.

regulatory experience and practices. An indication of the relevance and

effectivenessof the regulatorywork is the fact that14E&EMissionshave

bought into the NARUC Agreement to complement Mission bilateral

regulatorydevelopmentefforts”(USAID2007).

Before the partnership with USAID in Central and Eastern Europe, NARUC had

virtually no international exposure. US regulators had occasional meetings with

regulators from neighbouring Canada and Mexico (interview 9). Besides these,

Americanstatelevelutilityregulatorswereconfinedtotheirstate,andNARUChad

a predominantly inward character and focus on the relationship with federal

agencies.Inparticularsincethede-regulationperiodofthe1970sand1980(Derthick

andQuirk 1985), federal agencies had seen their authority expanded: a series of

pronunciations of the Supreme Court greatly expanded the scope of federal

regulators’ discretion over that of state level regulators. Therefore, when USAID

188

proposed toNARUC toundersigna cooperativeagreement formarket reforms in

CentralandEasternEurope,itfoundareceptiveear:

“I think USAID realized they needed… they could get hands-on, very active

regulatorsthatareactuallyinofficeprovidingthoseservicessotheycameto

ustoseeifthatwouldbesomethingwecouldbeinterestedinandwewere,

andweare,that’swhywewentforward.(…)OurchargefromUSAIDwasreally

at least two-fold: firstwas toestablisha regionalassociation thatwouldbe

modelled after the NARUC model (…) the other task was to help the

commissionsofindividualcountries,establishindependentmissionsintheir…

totheir….legislatorsorgovernmentsandthelike,sowehadalotofbilateral,

wecallthempartnerships,betweenanindividualstateregulatorybodyherein

theUSwithapartner”(interview9).

The foundingmembers of ERRAwere the energy regulators of Albania, Armenia,

Bulgaria,Estonia,Georgia,Hungary,Kazakhstan,KyrgyzRepublic,Latvia,Lithuania,

Moldova, Poland, Romania, the Russian Federation and Ukraine. Several

intervieweesrecalledthat the ideaofanetworkofCentralandEasternEuropean

regulatorswasairedprimarilybytheregulatorsthemselves.Nevertheless,layingthe

foundation for cooperation was not an entirely smooth process. As emerged in

interviews concerning the early days of CEER, regulatorswere suspicious of each

other,unusedtoshareinformationandtocooperatewithpeers,andinsecureabout

theirownnationalrole.

“Itwasnoteasy,becausewhenyouwere sittingonameetingmanyof the

participatingcommissionsdidnotspeakEnglishatallandnotevenRussian,

theywerenotusedtostandupandspeakingtheiropinion,perhapstheywere

afraidoftheAmericans…thisissomethingelsethancooperation,Iguessthere

wassuspiciontowardsthiswholeexercise…andtheenergysectorwasalways

averysensitiveonealsointhatregionsoinformationsharingwassomething

very new and suspicious for those people… they were motivated to come

189

becausetheyreallywantedtolearnandgettheexperiencebutinthebeginning

itwasmuch…itwasadifficultstart…”(interview27).

Despitethesehesitations,sustainedcooperationquicklyensuedfromthefirstadhoc

meetings.Thenetworkregisteredasaregionalassociationofenergyregulatorswith

legalheadquartersinBudapestin2001.Inthatyear,theregulatorsofSlovakia,the

Czech Republic and Mongolia also officially joined. They were followed by the

regulatorsofCroatiaandTurkeyin2002andafewyearslaterbytheregulatorsof

SouthEastEurope.Thechoiceofcastingthenetworkinaregionalframeworkreflects

theenthusiasmofthetimewiththere-launchoftheEUSingleMarketProgramme

inthemid-80s,completedin1992withtheMaastrichtTreaty.TheUSaspirationwas

thatformercommunistcountrieswouldgraduallyandfinallybeenglobedintheEU

andthereforeintheEuro-Atlanticsphereofinfluence(interview29).

PracticalconsiderationsalsoplayedaroleinthedecisiontofostertheERRAnetwork;

not least, the budgetary constraints faced by regulatory authorities in small

countries.Byconveninginasingleplatform,regulatorswereabletoaccessstateof

theartknowledgeatcontainedcosts.Theadditionaladvantageofbeingabletotalk

topeersintheregioninaninformalcontextconsiderablystrengthenedthecasefor

thenetworktobeestablished.

“Itwasfranklyaneffectiveandefficientwayofbringingexpertisethatwould

beprobablyeconomicallyunfeasibleonaregulatorbyregulatorbasis, theX

regulatorIsuspectwouldhavegreatdifficultymarshallingenoughbudgetto

bringinleadingeconomists,licensingexperts,financingparties,totalktothem

andworkwiththeminworkshopsintheircapitalcityabouthowXshouldbe

lookingatthefuturetoevolvetoadiversified,competitive,sustainablepower

systemthatitwouldlike,butifXisoneof…24members,andyoucansendyour

tariffexperts totheannual tariffworkshop,andyour licensingexpert tothe

annual licensingworkshop, and your commissioners to the annual investors

190

conference, you’ve really gotten this ability and access to extraordinarily

knowledgeablepeopleat,quitefrankly,1/24thoftheprice”(interview15).

“Inthattime,themajordrivingforceforERRAmemberswastolearnwhatdoes

itmeanenergyrestructuringandhowtoinvolveprivatecapitalintotheenergy

industry.Thememberswantedtolearnfrompracticingregulatorsoutsideof

ERRAbutsomesubjectwasalreadypreparedanddevelopedinERRAmember

countries.ForexampletheHungarianregulatorhasalreadybeenestablished

and functioning and the country went through the privatization and

restructuringsosomeoftheknowledgetransferwasnotonlyfromUSorfrom

EuropebutamongtheERRAmembercountriesanditwasvery…therewasa

realappetitefromthecountriesintheregion,becausestepbystepeveryyear

anotherregulatorwasestablishedoranotherlicensingdepartmentwassetup;

theyhadtoissuelicenseconditions,tosettariffs,andtheydidnotknowhow

todoit.Thatiswhythemeetingswerefullofreallifediscussion,like;heyguys,

whattoputintothelicenseconditions?”(interview25).

ItisimportanttobearinmindthatCEERandERRAwereevolvinginparallel.Closer

analysisrevealsthattheywerealsomakinguseofsimilaropportunitystructures.In

thelate1990s–early2000s,EU-15energyregulatorsbegancooperatinginorderto

buildtheirownlegitimacyunderthepoliticalsupportoftheEuropeanCommission;

meanwhile,CentralandEasternEuropeanregulatorswereachievingthesamegoal,

undertheaegisandwiththesupportofUSAIDandNARUC.Althoughcircumstances

weredifferent,bothgroupsofregulatorsconsistedofnewlyestablishedinstitutions

having to assert themselves on their national scene and to overcome national

governments’andconstituencies’resistances.

“Especially when you are… yeah… being part of… those networks… really

provides you with a very strong professional background and community…

many times regulators deal with international companies operating in the

countries, having very strong global networks and verywell paid lawyers, I

191

mean,tremendousresources,now,youhavetoregulatethesecompanies,you

havetohurtthem,essentially,becausewithoutregulationtheywouldbehave

otherwiseandyouhavetosayno,youcandothisbutnotthat,andthenwhen

you do this you get into a conflict and you have to argue and defend your

decisions, and I think these associations what they mostly help for their

membersisthatthey…theyalsoprovidethatsortof…internationalnetwork…

thattheregulatedcompaniesalsohave…yougetcloserintoasimilarposition

that just knowinghow theybehave inother countries,what solutionsother

colleaguesfoundallaround,it’sveryimportant…”(interview27).

“Therealdrivingforcewasthediscussion,theknowledgetransfer,andgiving

astrongbackgroundforthememberswhocanuseitwhentheyarearguing

hometheycansaythat…heyguys,10ERRAmembersarefollowingthesame

way, then my position should be good - this strong basis was important,

because therewere no booklets, no theory for regulation in that region, so

that’swhyitwasarealtechnicalsupportfortheregulatorswhostartedfrom

scratchregulatingtheindustry”(interview25).

LikeCEERrepresentedthechannelofregulators’influenceontheconfigurationof

the nascent Internal Energy Market, ERRA helped the regulators of Central and

EasternEuropetoplayaroleininfluencingthecontentandthedirectionofreform

inthefaceoflukewarmgovernments’commitmenttochange.InboththeEU-15and

EasternEurope,therealizationofthenecessitythattheelectricityandgassectors

hadtobereformedclashedwiththepoliticalimplicationsthatreformswouldhave

hadintermsofhurtingconsolidatedincumbents’assetsandmarketsharesaswell

asaffectingtheincomeofenergyconsumers,bigandsmall.

“Atthattimewehadthefeelingthatthegovernments inSouthandEastern

Europe…werealizedthatthegovernments,theydidn’tfollowtheinternational

tendencies,theywerenotinterestedinrestructuring,sothattheERRAandthe

regulators in the region they were there facilitators of the changes, they

192

convincedthegovernmentswhytorestructure,whatisthereasonforit,whyis

ituseful,sothat’swhy…inmostcasestherewasnopoliticalwishto…tohave

a…stableregulatoryframework…

Butwhydoyoucreatearegulatorifyouarenotinterestedinreform?

Itwasfashionable,itwasawishfromdifferentinternationalorganizationsand

lendinginstitutions,therewereEUrequirements,andthegovernmentsrealized

thatalltheneighboursdoit,thereshouldbeareasonbehindit,andbecause

laterontheyrealizedtheywouldliketorestructure,theywouldliketoattract

private capital so they need this regulatory framework. The details were

prepared by the regulators; in most of the cases the regulators forced the

governmentstorestructure,toprepare,todraftbillsandtoamendthemarket

structure.Forusitwasevidentthatwhenwelearnfromeachotherandfrom

therestoftheworldregardingthebestregulatorypracticethenwecouldbe

onthegoodway.(…)ItwasnotjustbecauseofERRA,therewasaneedofthe

nationalmarkets to change and they had no guidance how to change, the

governmentscouldnotgivethemanyadvice,therewasnoliteraturetolearnit

anditwasevidentforthemthat…whenseveralotherERRAmembersaregoing

intoonedirection,whynottodothesame…itwasaverygoodmoment,during

the restructuring and establishing the regulation, learning from other ERRA

members”(interview25).

Funded by themagnitude of the foreign policy goals of the US government, the

programme enjoyed a rare opportunity for bilateral aid provision: continuity and

coordination.Veryoften,donorsarecriticizedfortheshortlengthsoftheirmandates

andforthelackofinter-donorcoordination(KleinandHarford2005).Thelengthof

theprogrammeandtheapproachadoptedmadeforanactualpartnershipbetween

donorsandlocalelitesandforanarticulatedapproachtoreform.

193

“Theapproachbecamethree-pronged: twoprongswas technicalassistance,

secondlythepartnershipsthatwerenationalinscopeandthenthethirdprong

ofassistancewastheregionalnetwork,sofromtheperspectiveofanenergy

regulatorsittinginMoldova,you’vehadatechnicaladvisorwhocanhelpyou

on the tariff methodology, you had access to the US pricing tariff process

throughthepartnershipandthenyoucouldgositdownwithyourneighbours

andhearatariffcommitteeandseewhattheyweredoing,whatproblemsthey

hadetc.sothreepointsoftestbeforetheymakedecisions, Ithinkthatgave

them enough confidence and capability to try to pursuewhat they could in

difficult political environments… We found that consultants only were not

enough.That’sthebottomline”(interview29).

Thenear-simultaneityofeventsintheEU-15andinEasternEuropemightexplainthe

conspicuousabsenceoftheEUintheprocessesoutlinedinthischapter:whileERRA

was being created, the CEER did not officially exist yet, and its members were

grappling with the same issues as their Eastern European counterparts. As one

intervieweerecalled:

“Wedidhavesomepersonalcontacts,youknow…buttheCEERwastooslow,

wewerenotready.Wegottherewheneverythinghadalreadybeendone,so

wedidnotreallyplayarolebackthen.(…)Theywereveryforward-looking,

theyunderstoodsomethingthattheEuropeanCommissionmadeitsownonly

lateron:wepreventfuturewarsifwecreatemanylevelsofcooperation;not

onlyatheadofstatelevelbutalsoamongthegovernancestructures.(…)They

understoodyouhadtodoinstitution-buildingincountriesthatyouwanttobe

democratic: you cannot just go there and tell them you have to hold free

electionsotherwiseyouarenotdemocratic,underconditionalitythreats;you

have to do something more, like encouraging learning and cooperation

amongtheintermediatelevelsoftheestablishment”(interview3).

194

TheendofERRAasaregulatorynetworkandthebeginningofERRAastraining

provider.

Afewyearsintotheircooperation,ERRAregulatorsrealizedthattheiraccumulated

experiencewould be helpful trainingmaterial for future staff of their regulatory

authorities.TheyenvisagedtheERRAtocontinueintothefutureandthesocialization

mechanism triggered by common regulatory training to sustain network

cooperation,asithadbeenforNARUCandCEER.InFebruary2003,ERRAandCEER

joinedeffortstoorganizethefirstERRAcourse–introductiontoenergyregulation–

addressedtoelectricityregulatorsfromEUaccessioncountries.Inthesummerofthe

2003, this introductory coursewas held again, organized in conjunctionwith the

Central European University (CEU), and offered to all ERRA members. After the

European enlargement, ERRA developed a comprehensive package of training

coursesthatcomprisebothfacetofaceandonlinetuitionandthat,today,represents

itsmainsourceofincome.

The expansion of ERRA began soon after the European eastward enlargement in

2004,whichinvolvedsevencountries,whoseenergyregulatorshadbeenfounding

members of ERRA: Estonia, Latvia and Lithuania and four former satellites of the

USSR(Poland,theCzechRepublic,HungaryandSlovakia).Asubsequentenlargement

took place in 2007,whenBulgaria andRomania becameEUMember States. The

correspondingenergyregulatorsjoinedtheCEERalmostbydefault.Theintegration

of Eastern European countrieswas finally underway, and their energy regulators

wouldbeobligedtocomplywithand implementEUenergy legislation;wouldthe

ERRAfallapart?

Withtheenlargement,ERRA’sinitialmandatewasfinished.Inordertoremainviable,

thenetworkedorganizationhad to re-invent itself. It decided to convert itsmain

missionfrompolitically loadedroleofaccompanyingthetransitionofCentraland

EasternEuropeanenergysectorstomarketprinciplestothelessconsequentialrole

of regulatory trainingprovider,witha specific focuson regulators fromcountries

195

where theenergysector is stillwhollyormostlyunder fullgovernmentalcontrol.

Initially,itexpandeditsmembershiptoincluderegulatorsfromSouthEastEuropean

countriesbetween2002and2006(i.e.exceptAlbania,whowasamongthefounders,

Croatia, Serbia, Bosnia-Herzegovina, Montenegro, UNM in Kosovo, FYR of

Macedonia).

Lateron,ERRAbeganpromotingitselfworldwideasapointofreferenceforenergy

regulatory training,with thesupportofUSAIDandNARUC.Sincetheearly2010s,

energyregulatoryauthoritiesfromanimpressiverangeofcountries,fromPakistan

toSaudiArabiatoNigeria,havejoinedtheERRAasaffiliateorassociatemembers,

therebyobtainingdiscountsontrainingandeventattendance.Internationaldonors

fuel this expansion by referring to ERRA as the regulatory training provider for

emerging markets. Nowadays, ERRA explicitly refrains from being perceived as

providing policy recommendations or from aspiring to do so. This shift can be

understood as aimed at gathering as ample amembership as possible, including

regulatorsthatarenotindependentandwhosemandateistooverseestate-owned

vertically integrated companies rather than attempting to achieve competitive

markets.

Networkexpansionandforeignpolicygoals.

In examining the case of financial services regulators, Macey (2003) argues that

nationalregulatorsbecomeinvolvedintocoordinationandcooperationwithpeers

from other countries (what he termed “regulatory globalization”) in response to

threephenomena:theirpotentialirrelevanceduetotheopportunityforcompanies

to engage in regulatory arbitrage; the impossibility of achieving their domestic

mandatebecauseoftheregulatorypractices inothercountries (whichtriggerthe

exportoftheirregulatorystandardstoothercountries-whatMaceycalls“regulatory

imperialism” – thereby gaining them considerable domestic prestige); or the

196

regulatoryagency’pursuitofitsownpolicypreferenceswhenfacedwithdomestic

oppositionfromaffectedconstituencies.

The first phenomenon is unlikely to find strong application in the world of

infrastructure investment: regulatory stability and predictability, rather than the

laxity of regulatory provisions, are among the main determinants of foreign

investment in infrastructure (Spiller and Tommasi 2005). Therefore, the

phenomenonofregulatoryarbitrageisunlikelytoapplytothiscontext.Thesecond

motivation that Macey (2003) adduces to explain transnational regulatory

cooperationpointstointerdependence.Wheneveraregulatoryagencyisprevented

fromfullyaccomplishingitsmandateasaresultofitsenforcementinterdependence

witharegulatorfromanothercountry,itwillinvestpoliticalcapitalintopersuading

the regulators fromthatcountry toadoptcommonrules.Yet, there ishardlyany

interdependencebetweentheenergysectorsorutilitybusinessesoftheUSAandthe

CentralandEasternEurope.

The third ofMacey’s hypotheses resonatesmore clearly with the context of the

establishment of ERRA (and of all comparable regulatory networks): periods of

radical policy change, inparticular in sectorswithdirect and visibledistributional

consequences, such as the utilities, createwinners and losers. The losers oppose

reform.EasternEuropeanregulators,exactlyliketheirEUcounterpartsinthesame

years, faced opposition from all sides who stood to lose from reform. The

endorsement of USAID and NARUC as well as the collective leverage of ERRA

representedanimportantsourceoflegitimacy.

“Sometimes, however, we observe the government expending significant political

capitaltoaffectinternationalpracticeinordertoimposeitswillontheinternational

community,evenwhensuchregulatoryglobalizationwillnotaffectdomesticpolicy

orpracticeatall”(Macey2003,p.1369).Thisstatementcapturestherationaleof

theUSgovernment’ssupporttothemission:thedesiretoeradicatetheremnantsof

communism from Eastern Europe represented a potent motivation to devote

197

resourcestotheprogrammeeventhoughithadnorealimpactondomesticpolicy.

Rather, it was part of a broader foreign policy strategy aimed at preventing the

backslidingofthedemocratizationofCentralandEasternEuropeancountriesand

theirtransformationintoliberalmarketeconomies.

USAIDchosenetworksoverotheroptionsasinstrumentsofprofessionalsocialization

abletofostersympatheticinterlocutors/changeagentsinrecipientcountries.Inturn,

NARUCexploited thisopportunity toadda layer to its functionsand increase the

breadth of the organization. By informing the recent history of NARUC and

juxtaposing itselftothatofCEER,thecaseofERRAprovidesausefulcomparative

casewithinacase.CEERandERRAdevelopedinparallelandnevercameincontact

astheywereemerging.Yet,regulatorsfrombothnetworksfacedsimilarchallenges,

e.g.uncertaintyanddomesticopposition,andpickedsimilarstrategies,e.g.usetheir

peers as a source of epistemic knowledge and legitimacy, and exploiting their

symbiotic interdependence with actors at higher levels of governance to obtain

accesstopolicyformulation.

Moreover,thecomparisonofthesecasessuggeststhattheprecisetimingofeach

phaseand/orthelengthoftimeeachofthephasesidentifiedinthisanalysislasts

varies across networks. Whereas in the case of NARUC and ERRA formalization

occurredrelativelysoon,CEERbecamelegallyregisteredsixyearsafteritsinception.

Thesedifferencesdonot,however,changethesubstanceandthesignificanceofthe

evolutionaryphaseseachofthesenetworkswentthrough.Thisisanacknowledged

factincomparativehistoricalstudies(MahoneyandRueschemeyer2003),whichare

agnosticastothepreciselengthofthespansbetweenphasesbutconvincedofthe

occurrenceandtherelevanceofthephasesthemselves.

198

7. Networks as “first best”? The establishment of thenetworkofEuro-Mediterraneanenergyregulators.

If thestudyofEuropeannetworkshasbeencontinuouslygrowinganddeepening

during thepast twodecades, the studyof theEuropeanexternalgovernancehas

remainedrathermoreconfinedtoitsformalmanifestations.Moreover,thepotential

oractualrelationshipsbetweenEuropeanregulatorsandregulatorsfromcountries

neighbouringtheEUhavebeenoverlooked inthe literature.Thischapter fills this

gap.ItprovidesanempiricalcaseofwhatIterm“networkentrepreneurship”:the

furtheranceofnetworkemergencebypolicyentrepreneurs.

Moreover,itrevealsanimportant,neglectedaspectofregulators’preferences.The

literatureconsidersinformalnetworksafall-backoptionwhenformalarrangements

cannot be achieved; a “second-best” solution (Sutherland 1992, Hancher 1996,

Dehousse 1997). This chapter shows that, from the point of view of regulators,

networks are a “first-best” option for regulatory collaboration. More precisely,

regulators think that formal arrangements should always follow, never precede,

informal collaboration. Collaboration between previously disconnected, stranger

regulatorsneeds tostart informally,or itwillbeunworkable.Regulatorsconsider

thatprolongedinformalcollaboration,notlegallybindingarrangements,canbreed

trust. This finding is important, because regulators are the subjects of regulatory

collaboration within networks. The policy-maker wishing to foster regulatory

collaborationshouldcareaboutwhichformatofcollaborationthatcanbeexpected

toachievethedesiredresult,andwhichdoesn’t.Thefindingsofthischaptershould

motivatefurtherstudy inthe informaldynamicsofEuropeanexternalgovernance

(Lavenex2004,2008,2009)andforeignpolicyinenergymatters(Youngs2014and

JudgeandMaltby2017).

199

Thenarrative focuseson the establishmentof theAssociationofMediterranean

EnergyRegulators(MedReg).MedRegdiffersfromtheotherthreecasesexamined

sofarinseveralrespects.Firstly,MedRegemergedduetotheinitiativeofanagent

(the Italianenergy regulatoryauthority)whoplayed the roleof leadorganization

(ProvanandKenis,2008)byprovidingfornetworkestablishment.Hence,MedReg

differsfromCEER,whichemergedduetothecollectiveinitiativeofasmallgroupof

regulators, and from NARUC and ERRA, which emerged from a combination of

regulators’initiativeandanexternalagent’ssupport(ICCandUSAID).Frustratedwith

thestandstillofenergycooperationbetweentheEUandthecountriesintheEuro-

Mediterraneanregion,asmallsubsetofEuropeanregulatorsfromSouthernMember

States invited their foreign counterparts to join efforts to create a new informal

networkfocusedontheregion.Cruciallyforthisresearch,theyledthisinitiativeon

thebasisoftheirownexperienceofinformalnetworkcooperationwithintheCEER

andfromobservationofthesuccessofERRAinsocializingandunitingregulatorsin

EasternEurope.

Secondly,MedRegdoesnotinvolveallEUenergyregulators,butonlythesubsetfrom

SouthernMember States. Therefore,MedReg gathers amix of energy regulators

from EU and non-EU countries, drawn from an EU-defined “Euro-Mediterranean

region”whoseboundariescompriseabroadrangeofverydiversecountries,from

EasternEuropetoNorthAfrica.

Thirdly,asthischaptershows,MedRegwasestablishedinthewakeofaboldpolicy

initiative(theestablishmentoftheEnergyCommunityofSouthEastEurope)butin

absenceofastrongmomentumforpolicychange(as itoccurred intheEUandin

CentralandEasternEuropebeginninginthelate1990s)orduringapolicycrisis(asit

happenedin19thcenturyUSAwiththediscontenttowardstherailroads).Fourthly,

theregulatorsinMedReghavenosymbioticinterdependencewithasupranational

policyactorsfurtheringaspecificagenda.Tenyearsafteritsestablishment,MedReg

is not perceived as an influential player in the energy policy of the countries

200

concernedoroftheregionasawhole(CambiniandFranzi2013).Recently,however,

theEuropeanCommissionhasendorsedMedRegandcommittedtokeepsupporting

thenetworkforthefourthcycleofitsoperations.

The EU’s external energy regulatory governance: top-down and

legalistic.

Relations between the EU and the countries to its southern border have a long

history.The firstattempts to forgeanexternalEUpolicy towards thesecountries

datebacktothe1970s,whentheGlobalMediterraneanPolicywasputforwardby

the then European Community, in the wake of the oil crisis (Cardwell, 2011).

However, it was only with the Maastricht Treaty and the stated objective of

increasing the EU presence abroad that a structured policy was planned for the

Mediterranean (Cavatorta and Rivetti 2014). It took the form of the so-called

BarcelonaProcess,initiatedwiththeBarcelonaconferenceof1995andbearingthe

officialnameofEuro-MediterraneanPartnership(EMPorEuroMed).TheEMPwasa

broadpolicy programme, covering items fromdemocracy to trade.However, the

centralityofenergyissueswasacknowledgedintheonlyECCommunicationentirely

dedicated to energy cooperation in the so-called “Euro-Mediterranean

region”(EuropeanCommission1996).TheEMPenergyvectorwascharacterizedbya

two-foldregionalapproach:politicaldialoguebetweenEnergyMinistersfromboth

shores in the Inter-Ministerial Conferences and expert dialogue between

representatives of the partner countries and the EC in the Euro-Mediterranean

201

EnergyForum.BoththeConferencesandtheForumbeganconveningsoonafterthe

BarcelonaConference,butfellshortofregularmeetingsthereafter39.

In2004,theEUaccomplishedthreemajorpolicyefforts:itcarriedouttheEuropean

eastwardenlargement;thelaunchoftheEuropeanNeighbourhoodPolicy(ENP);and

theestablishmentoftheEnergyCommunityofSouthEastEurope(officiallyinforce

since2006,butpreparedintheprecedingyears).Theenlargementbroughttennew

Member States into the EU. The European Neighbourhood Policy (ENP) was an

umbrellaprogrammegroupingthevariouscooperationprogrammeslinkingtheEU

anditsSouthernandEasternneighbourhood.TheENPmarkedtheboundariesofthe

EU, separating Members and Accession countries from countries without a

membershipperspective(Smith2005,Cardwell2011).Moreover,theENPcombined

regionalaspectswithapronouncedbilateraldimension:theCommissionproposes

bilateral Action Plans to each of its “neighbours”, framing cooperation around a

seriesofmutuallyagreedissues.TheActionPlansspecifytheextentofregulatory

convergencethateachcountryiswillingtoundertakeindifferenteconomicsectors

(Escribano 2010). Energy is an important topic in these bilateral relations, which

revolve also around the harmonization or adoption of the European energy

regulatoryframework.Inasmuchasnetworkgovernanceisacharacterisingfeature

of the EU (Boeger and Corkin 2017), the external dimension of its regulatory

39 The Inter-Ministerial conferences convened for the first time in Trieste in 1996.Other

meetings followed inRomeandAthens in2003.Theywere thendiscontinueduntil 2007

(meetinginLimassol,Cyprus)and2014(meetinginRome,Italy).TheEuro-Mediterranean

EnergyForumconvenedforthefirsttime inBrussels in1997,then inGranada in2000. It

proved extremely difficult to retrace the chronology of Forummeetings thereafter. The

Athensmeetingof2003appears tomix theForumand the Inter-MinisterialConferences

together. Following theAthensmeeting, the Energy Forumas conceivedwithin the EMP

seemstohavebeendiscontinued.

202

governance is, instead, highly formalized and premised on the extension of the

Europeanacquis(Bicchi2006).

At theverybeginningof the2000s, theEUand theUSAcrossed roads inEastern

Europe.ThisresultedinthecreationoftheEnergyCommunityofSouthEastEurope

(ECSEE)in2005.TheECSEEisaTreaty-basedorganization,originallycomprisingthe

countries that were deemed unprepared to join the EU as Member States

immediately, yetmaintained Accession status: namely, SEE countries and Turkey

(thatonly joinedasassociate,notasfullmember).Thepremiseofthisessentially

political initiativewas to extend the EU energy legislation to the countries in its

neighbourhood, as apreparatory step to full accession (Deitz, Stirtonet al. 2009,

Karova2009,Padgett2012,Schulze2015).TheEuropeanCommission,CEER,USAID

andERRA,aswellastheWorldBankandotherinternationaldonorswereallinvolved

inthesettingupoftheECSEE.

“WehadveryverycommoninterestwiththeEUintermsoftheintegrationof

the Balkans into the Euro-Atlantic community and integration intoWestern

Europe,theEU…itwasaveryverystaunchsupport,(…)withintheAIDbutalso

within the state department they understood the importance of that”

(interview29).

The political rationale for the establishment of the ECSEE was two-fold: the EU

wanted toexpand to reachof itsmarket framework to foster investments in the

infrastructure that would allow it to diversify its energy imports (Karova 2009,

Padgett2012);moreover,theintegrationofSouthEasternEuropeannetworkswould

endtheelectricalisolationofGreece.TheTreatybindsitssignatoriestoimplement

relevantpiecesoftheEUenergylegislation.Thesignificantcostsofadaptationfor

SEEcountrieswouldberewardedbyEUmembership(Karova,2009).TheECThas

alsobeenunderstoodasa signof theEU’s “enlargement fatigue” (Renner2009),

driving it to postpone full Accession while still increasing political stability in its

neighbourhood.BoththeenlargementandtheEnergyCommunitywereperceived

203

asimportantpoliticalachievementsfortheEU:theyportrayeditssuccessatexerting

and exporting its normative (Manners 2015) andmarket (Damro 2015) power to

third countries. Meanwhile, the Euro-Mediterranean Partnership was embedded

intotheENP.

TheEnergyCommunityTreaty foresawtheadoptionofEUenergy legislation tout

court in the countries of South East Europe via a formalizedprocedure typical of

accession processes (Schimmelfennig 2012). After 2004, the EU policy message

concerning regional energy cooperation became centred on the extension of the

“EnergyCommunityapproach”totheEuro-Mediterranean(Tholens2014).External

energygovernancebecamepremisedonthenotionofextending“theEU’sinternal

market,throughexpansionoftheEnergyCommunityTreatytoincluderelevantENP

countries.” (Joint Paper 2006, p. 6), whereas a decade earlier the European

Commission affirmed that regional energy cooperation with the Mediterranean

shouldaim“todevelopenergyplanningtoolsbasedonthehighlycomplementary

natureoftheNorthernandSouthernMediterraneanmarketsandsupplynetworks”

(EuropeanCommission1996,p.4)andleftroomforco-ownershipofthelegaland

regulatoryaspectofenergycooperation(“Anappropriatelegalframeworkshouldbe

devisedtoencourageandpromoteregionalandtransregionaltrade”,p.6).

In the followingyears,however,slowprogress in theachievementof the Internal

Energy Market and deterioration of energy relations with Russia increased the

perception of threat to the EU’s energy security. The gas disruption episodes

involving Russia and Ukraine in 2006 and 2009 prompted renewed European

Commission’sinterestintheMediterranean.Thistooktheformofbilateralstrategic

204

partnershipswithENPcountries(suchasAlgeria,Morocco,Egypt),aswellasthere-

launchofregionalcooperationinbothgasandelectricity40.

Soonafterwards,in2008,thethenPresidentofFrance,NicolasSarkozy,announced

theestablishmentoftheUnionfortheMediterranean,foreseeingclosercooperation

ineconomic issuesbetween the twobordersof theMediterranean.The initiative

should have concerned only EU members bordering the Mediterranean. The

European Commission, however, opposed the initiative, pointing out that an

economic cooperation programme that excluded most EU Member states was

inacceptable. The idea was therefore repackaged as another re-launch of the

EuroMed Partnership/Barcelona Process; its full name became the “Barcelona

Process:Union for theMediterranean”.The revampof thePartnership,however,

wasnotaccompaniedbysubstantialchangeintheEuropeanapproachtowardsits

Southernneighbours(Darbouche2011). Importantly,noneoftheEuropeanpolicy

initiativesincludedaregulatorycooperationorregulatorypartnershipcomponent,

noracapacitybuildingaspect.

EvenaftertherevampofEuro-Mediterraneanenergycooperationfollowingthegas

crisesinEasternEurope,theCommission’sstancedidnotchange:inMarch2011,a

JointCommunicationoftheCommissionandtheHighRepresentativeoftheEUin

reactiontothe“Arabspring”encouragedtheextensionoftheEnergyCommunity

approachtotheSouthernMediterranean(JointCommunication2011a,p.10).The

ideawas reiterated in theMay2011Reviewof the ENP: “a formof EU-Southern

MediterraneanEnergyCommunity.ExtendingtheEnergyCommunityTreatywiththe

Union's Eastern and South-Eastern neighbours, or building on its experience, this

40AtthefifthEuroMedEnergyMinistersconference,heldinLimassol(Cyprus)in2007.On

that occasion, the European Council jointly with the European Commission launched a

renewedEuro-MediterraneanEnergyPartnership.AnActionPlan for the regionwasalso

launched on that occasion, covering the 2008 – 2013 time span. On that occasion, the

MedRegwaspubliclypraisedforitseffortsatbringingaboutregulatoryharmonization.

205

communityshouldcoverrelevantpartsoftheEU'senergylegislationwithaviewto

promotingarealandreliableconvergenceofSouthMediterraneanpartners'energy

policieswithEUpolicy.”(JointCommunication2011b,p.15).

The European Commission’s approach stands in contrast to the assessments of

regulators, system operators and other technical experts involved in Euro-

Mediterranean energy cooperation: according to recent research, these are

convincedthattheideaofextendingtheEnergyCommunityisunworkableinthat

context and should not be used to frame the cooperation, lest jeopardizing its

progress(Tholens2014).

TheestablishmentoftheMedReg:networksas“first-best”.

“Therewastheenlargement…TheEUwasambitiousanditstartedto look

around itself. Inevitably they saw that both South East Europe and the

Mediterraneanwerecrucialforenergyissues,buttheyfocusedfirstonSouth

EastEurope.TheCommission,grabbingspaceasusual,imaginedthis“Energy

Community”.IimmediatelytriedtogetinvolvedbecauseIthoughtweneeded

to,astheclosestEuropeancountries…butthenwekindoflostthehandleof

thisandthecountriesendedupbeingrepresentedbytheEU.(…)Onthebasis

ofthatexperience,Ithought,well,I’lltakeashortcutandbepro-active,I’lltry

to anticipate the Commission… We will invent the community of the

Mediterranean…andsoIcreatedMedReg”(interview4).

TheonlyoccasionwhenCEERcollaboratedwiththeEuropeanCommissiontodeliver

foreign policy goals presented itself with the processes leading to the official

signatureoftheEnergyCommunityTreaty(interviews2,7,8and42).Inthoseyears,

theroleofCEERwas,incooperationwiththethenDirectorateGeneralforTransport

andEnergyoftheEuropeanCommission,aswellasUSAIDandotherinternational

206

donors,tofacilitatethesetupoftheEnergyCommunityRegulatoryBoard(ECRB)–a

boardofenergyregulatorswithadvisorypowersthatwouldrepresentintegralpart

oftheEnergyCommunityorganizationalstructure(interview7,42).CEERsetupan

adhocworkinggroupandhostedjointtrainingsandmeetingswithERRAaswellas

theregulatorsfromtheECSEE.Afterthatperiod,however,CEERwasneverinvolved

againinthedeliveryofEuropeanexternalenergypolicyobjectives41,whichareled

bytheEuropeanCommission.

Thequotationopeningthissectionnaturallyleadstoaskingaratherstraightforward

question: why would EU national regulators care what approach the European

Commission does or does not take as concerns energy cooperation with

neighbouringcountries?Limitationsonspacedonotallowanexhaustiveresponse

to this question, since a full account would need to discuss the institutional

boundariesofnationalregulatoryinstitutionsandthewaysinwhichtheyunderstand

theirroleindomesticandEuropeanpolitics.Forthepurposesofthisresearch,afew

details concerning the energy sector of the countries, whose regulators led the

initiative,mayshedlightontheirreasons.

The initiative to set up MedReg came from a small group of EU regulators, in

particular the Italian Regulatory authority, in 2006. Italy is the country with the

highestelectricitypricesintheEuropeanUnion42anddependsonfossilfuelimports

(particularlyfromRussiaandAlgeria)foritssupply.Moreover,theenergystrategy

developed by the Italian government foresees transforming the country into an

41 The CEER’s International Strategy working group coordinates the network’s informal

relations with similar networks, such as the NARUC, and the CEER participation in

internationaleventsonenergyregulation,butdoesnotremotelyresemblethestructure,

thescopeortheresourcesoftheNARUCInternationalAffairsdivision.42Eurostat,http://ec.europa.eu/eurostat/statistics-

explained/index.php/Energy_price_statistics(lastaccessed13August2017)

207

energyhubforthetransitofgasandelectricitygeneratedthroughrenewableenergy

sourcestotherestofEurope43.

SpainalsodependsonAlgeriangasandisinterconnectedwithMorocco,whichhas,

inturn,beenheavilyinvestinginrenewableenergygenerationinordertoreduceits

importdependencefromfossilfuelsandtradeelectricitywiththeEU44.Franceisa

formercolonialmasterinNorthAfricaandwasunlikelytoremainindifferenttothe

creationofanetworkofregulatorsincludingFrench-speakingcountriesintheMENA

(interview6).Bythesametoken,acountrylikeGreeceisnotablyinterestedinthe

progressofelectricitymarketintegrationinSouthEastEurope,whoseregulatorsare

alsomembersofMedRegaswellasoftheEnergyCommunity.Italy,SpainandGreece

areimportanttransitcountriesfortheelectricityandgasthatwouldbegeneratedin

theregionandexportedtoNorthernEurope(EscribanoandJordán1999,Escribano

2010,Escribano2011,EscribanoandSanMartínGonzález2014).Inturn,countries

likeEgyptandTunisia,butalsoTurkeyandAlgeriaareinterestedinleveragingtheir

resourcesand/ortheirgeographicalpositiontoimprovetheirenergyuseandengage

inenergytradewiththeEU.

Despite the clear rationale for energy trade and, in the longer term, market

integrationacrosstheEuro-Mediterraneanregion,themyriaddedicatedinitiatives

havelackednoticeableprogress.Intergovernmentalinitiativesinvariablycollapsed,

EU policy programmes covered many issues besides energy, and Mediterranean

countries are unwilling to undertake privatization and/or liberalization reforms.

43ItalianGovernmentEnergyStrategy

http://www.sviluppoeconomico.gov.it/images/stories/documenti/SEN_EN_marzo2013.pdf

(lastaccessed15August2017)44 Quartz Africa, https://qz.com/533187/tired-of-importing-almost-all-its-energy-morocco-

has-built-africas-biggest-solar-farm/(lastaccessed15August2017)

208

Moreover,thereishardlysuchathingasaMediterraneanregion:ancientconflicts

anddivisionspermeatetheMiddleEastandNorthAfrica(MENA)(interview3,11).

InthewakeoftheestablishmentoftheEnergyCommunity,theCommissionbegan

leveragingthatpolicysuccesstowardsthecountriesintheSouthernneighbourhood.

Fearingthiswouldfurtherhamperprogressinthedirectionofenergytradeand/or

market integration with the Southern neighbours (both options are technically

alreadypossible,butpoliticalandregulatoryobstaclespersist,seeKhalfallah(2015)),

regulators,inadisplayofextraordinarypolicyinitiative,madethefirstmove.

“Isentthefirst invites,wehadtwoorthreemeetings,andthenyouknow

howthesethingsgo…youhavethemeetingsthenyoubeginthinkingwellwe

shouldmaybejotdownsomerules,thenyoudraftastatute,thenyougoto

the notary and in the space of a year… therewewere. It was impressive

becausesometimeswhenyoubeginthesethingsyoucanfacesomelukewarm

reactionsbutno,everyonerespondedtotheinvite,alsotherepresentativeof

Ministries where regulators did not yet exist. Everyone was eager to

understandtheirownmandateandprofession, itwas,youknow,thesame

mechanismatthetime[oftheCEER],youseekatabletositatwhereyoucan

dialoguewithpeersinordertohelpeachothergrow,standonourfeet,and

everyonewasveryenthusiastic”(interview4).

The association of Mediterranean electricity and gas regulators (MedReg) was

established as theMediterraneanWorking Group on Electricity and Natural Gas

Regulation inMay 2006within the framework of the Euro-Med Partnership45. In

45MEDREG,todayaRegionalRegulatoryAssociation,bringstogetherEnergyRegulatorsfor

Electricity andGas of the following countries: Albania, Algeria, the PalestinianAuthority,

Bosnia-Herzegovina, Cyprus, Croatia, Egypt, France, Jordan, Greece, Israel, Italy, Lybia,

Malta,Morocco,Montenegro, Portugal, Slovenia, Spain, Tunisia and Turkey. Allmember

regulatorsofMedRegarealsoregulatorsofcountriesthatarepartoftheEnergyCommunity.

209

2007,MedRegwas legally registeredunder ItalianLaw.Hence, the timebetween

network emergence and consolidation was remarkably short. Soon after legal

registration,regulatorsaddressedtheEuropeanCommissionforsupportinensuring

theviabilityofthenetwork.“Theywerehappyaboutit,becausewehaddoneallthe

work.Weprovidedthemwithaready-madeframeworkforcooperation.”(interview

4).MedRegis,sincethen,supportedbyEuropeanNeighbourhoodPolicyfunds46.

MedReg’s“ConceptandStrategyPaper”47highlightstheparallelbetweentheEnergy

Community itself. It also stresses the possible involvement of Mediterranean

countriesintoanenergycooperationrelationshiponthebasisofmarketintegration,

butwithadifferentapproach:“Consideringthestrategicenergychallengesof the

Mediterraneanregion,MediterraneanEnergyRegulatorsconsiderastheirduty[my

emphasis] to constitute a strong institutional basis promoting a “bottom up”

approach (…) to guarantee greater harmonization of the energy markets and

legislationsandtoseekprogressivemarket integration in theEuro-Mediterranean

region”.

46Itisusefulinthisregardtore-useaquotationfromaninterviewthatIusedinchapter3

when discussing the growth of the CEER, for it aptly frames the context of MedReg

establishmentandtheregulators’successinsecuringfundsfromtheEuropeanCommission:

“Let’sput it thisway, ifmygoalwouldbeto…implementortoachieveverynationalistic,

specificsolutions,probablyitwouldnotbesuccessfultodoitatEUlevel.Butifthegoalisto

developsolutionsthatmakesenseinageneraleconomicwayinaEuropeanperspectivethen

you’remuchbetterofftotrytopushthatatEUlevel”.(interview25)47MeRegConceptandStrategyPaper

http://www.europarl.europa.eu/meetdocs/2004_2009/documents/dv/200/200903/20090

316Medreg_concept_EN.pdf(lastaccessed12August2017)

210

ReplicatingtheCEERexperience.

MedRegwouldprobablynotexist,haditnotbeenfortheinitiativeoftheregulators.

Theregulatorswouldprobablynothavetakentheinitiative,haditnotbeenfortheir

experienceofcooperationwithinCEER,whichtheyconsideredsuccessfulbecauseit

allowedthemtofeedintoEuropeanenergypolicy-making.

“The CEER was a bottom-up association born to preserve the regulators’

autonomyfromtheCommission,aswell…MedRegwascreatedonthesame

basis,wellactuallyallinstitutionsthathavebeencreatedintheareaarekind

ofreplicatingthingsthat“worked”intheEU.Ofcourse,thisdoesnotmean

they would immediately “work” in the Euro-Med. The CEER helped the

Europeanharmonizationprocessimmensely.Withtime,itshoweditsobvious

limitsandthereforeithasbeennecessarytocreatetheAgency.Butformany

manyyearsithasbeenacrucialforumfordiscussion(…)SoMedRegcouldbe

considered as mirroring the CEER but the thing is that they have

representativenessproblems;evenonverytechnicaltopicsofregulationyou

immediatelyencounterinstitutionalandgeopoliticalobstacles”(interview5).

“Yes,absolutely,theMedRegwassetuplikeaphotocopyoftheCEER.There

are two main differences. One is that regulators have less incentives to

cooperatewithintheMedReg,becausetheMedReghaslessimpactonwhat

happensatnationallevel.TheCEERhasalwaysbeenvoluntarybutwhatthey

saycounts,bothfortheEuropeanenergypolicyandatdomestic level.The

seconddifferenceistheSecretariat.TheMedRegSecretariatwouldneedmore

resourcesinordertocoordinatetheworkinggroupsmoreeffectively.Itbeing

financedbytheEuropeanCommission, there isnotmuchscopeconcerning

howtoallocateresourcesdifferently.TheCEERwasalwaysself-financedso

theregulatorsorganizeditastheysawfit”(interview14).

211

“Well, you know, the foundingmembers of theMedReg all had the CEER

experienceintheirbackground,aswellastheFlorenceSchool,sotheywere

used to starting and carrying forward this type of conversations, of

cooperation”(interview10).

Themechanism leading fromCEER toMedReg is similar to the one leading from

NARUC to ERRA: in both cases, a previous successful experience of regulatory

cooperation inspired its replication in a different context. However, there are

profounddifferencesbetweenthetwocases.TheestablishmentofMedRegdidnot

derivefromacoordinatedEuropeanforeignpolicystrategy,butfromtheinitiative

of national regulatory institutions. These institutions purposefully chose the

informal,networkedapproachnotonlybecauseitwastheonetheyknewbest,but

alsobecauseitallowedthemtobypasstherigiditiesofthesystemofgovernanceof

theEU.Asmentionedabove,Sarkozy’sinitiativetolaunchaFrench-ledUnionforthe

MediterraneanwasimmediatelyblockedbytheCommissionandtransformedintoa

Europeanpolicy.Accordingtotestimonyfrominterviewsconductedforthisthesis,

however, the energy policy preferences of the various Member States are too

different;someofthemaresimplynotinterestedintheEuro-Mediterraneanregion

(interview4,5,6).Moreover,Commission-ledpolicyprogrammesarepremisedon,

onceagain,compliancewithEUlegislation.Informalnetworks,instead,cangather

asmanyorasfewEuropeanregulatorsasneeded.

“I think it is important to cooperate voluntarily, because if you impose

legislationtop-downithastobeabletobringaboutsolutions.Otherwiseit

becomesanobstacletocooperation,ratherthananenablingfactor.Thisis

thefundamentaldifference;theoriginsoftheCEERweretheopposite[ofthe

EnergyCommunity].There,youhadnewinstitutionsthatwerefacedwitha

radicallynewcontextanddecidedtocooperateonconcretethings.MedReg

wasinspiredbythesamelogic.Imean,therehavebeeninitiativesintheEuro-

Medwhereyouhadagreementssignedbygovernmentsandyettheywere

212

neverimplemented.Sotheideawaslet’sdoitstartingfromtheinstitutions

thataredirectlyinvolvedintheseprocesses,let’sseeifwecanwork…andit

was immediately successful because everyone needs to understand how

things work in order to make progress. The Mediterranean is a natural

market,youarebasicallycondemned,andIsaythisinaverypositivesense,

condemnedtoworktogether”(interview5).

ThewordsinthepreviousquotationconfirmtherelevanceofCEERininspiringthe

establishmentofMedRegwhilealsohintingatthemaindifferencebetweenthetwo

networks. MedReg was an attempt to replicate the process of influential

transnationalregulatorycooperationthatunfoldedinCEERinabsence,however,of

sweepingpolicychange.Tothecontrary,therationaleforcreatingMedRegwasthe

frustration with the protracted standstill in energy cooperation in the Euro-

Mediterranean.Moreprecisely,regulatorsappeartohaveconceivedofMedRegas

an initiative that would be able to bypass the political deadlocks in the Euro-

Mediterranean cooperation by focusing on concrete, win-win, solutions. After

MedRegwasestablished,regulatorspushedforthecreationofelectricityandgas

systemoperators,onthemodeloftheEuropeanones(calledENTSO-EandENTSO-

G). From their point of view, no political institution was needed in order for

regulatorstoidentifybarrierstointegrationandagreeonsolutionstoliftthem.

“WedidMedTsoand[haveproposed]MedTsoforGas,theywereborninkind

ofthesameway,soyouhavethesymmetrywiththeCEER,theENTSO-Eand

theENTSO-Gandyoureallydon’tneedanythingelse”(interview4).

In other words, the establishment ofMedReg was based on the conviction that

cooperationbetweentechnicalexperts (regulatorsandoperators)couldsufficeto

deliver investment in energy infrastructure and eventually markets and market

integration.Thisconvictionseemstohavebeenbasedontheassurednessthatthis

was what had happened in the case of CEER: the obvious benefits of market

213

integrationhadovercomedomesticreluctancetoenergysectorreformasprecursor

toanintegratedEuropeanenergymarket.

Aschapter3shows,however,thedesirabilityofenergymarketintegrationandthe

capabilityofCEERtoproposeconcreteregulatorysolutionstoachieveitwereonly

part of the story. The other part concerns specific political circumstances: most

importantly, the opening of a window of opportunity caused by the enormous

uncertaintysurroundingenergysectorreformacrossEuropeanMemberStates,and

theparallel formationof a European energy policy framework that entwined the

EuropeanCommissionandtheCEERintoasymbioticdependenceononeanother.

The latter provided information and an impartial, technical input into the policy

processwhiletheformerbackedtheregulators’cooperationpolitically.Moreover,

the Commission and other European institutions (particularly the European

Parliament) supported regulators and couldmuster the authority to pressure all

different Member States into compliance with the legislation. In a completely

differentsetting,USAIDcouldexertitscloutoverthecountriesofCentralandEastern

Europethatreceiveditsaid.Thecreationofthesepartnershiprelationswithactors,

externaltothedomesticpoliticalsettingofindividualregulatoryauthorities,wasin

itselfanecessaryconditionfornetworkinfluence.

IfEuropeanMedRegmembersappearoverlyoptimisticonthepotentialofregulatory

networkcooperationtoimpactonpolicyoutcomes,themembersofMedRegfrom

the Southern neighbourhood appear more aware of the limits of the network’s

impact on domestic contexts in absence of political commitment to reform. The

quotationbelowpresentsadifferentperspectiveontherationaleandtheusefulness

ofMedReg, that conceives of network cooperation as laying suitable ground for

reform, so thatwhen reformoccurs,no timeshallbewastedonworkingout the

detailsofregulatoryframeworkimplementation.

“Strategicdecisionisnotbeingtakenbyregulators,it’stakenbygovernment,

wehaveachallengeforexampleintheXproject,whichisamarketforenergy,

214

it’squiteclearthatthisposesoflotofblocksbecausetherelationsbetweenY

andZcountriesarenotthatsmooth.MedRegplayedaroleinthisprojectfirst

in providing capacity building, secondly, we decided to resolve technical

challenges,sowheneverthereisapoliticalopeningintheprocess,itwillnot

behaltedbecausethere isa lackoftechnicalknowledge.(…)Waitingfora

politicaldecisionandthenstartworkingfromthattimewilltakemuchlonger,

soMedRegincapacitybuilding,inconvergingideas,inmakingpeopleinthe

samewayofthinking,allofthisisverymuchimportant”(interview21).

The limits and the potential of MedReg for Euro-Mediterranean

energycooperation.

TheestablishmentoftheEnergyCommunityofSouthEastEuropewasperceivedas

an outstanding policy success of the European Union in exporting its energy

governance framework (Lavenex and Schimmelfennig 2009). In the wake of that

experience,therewasmomentumfortheideaofapan-EuropeanEnergyCommunity

encompassing the whole of the so-called European neighbourhood. Seizing the

momentumfirst,asmallgroupofenergyregulatoryauthoritiestookaboldpolicy

initiativeandsetupaninformalnetworkgatheringregulatorsfromcountriesasclose

tothebordersoftheEUgeographicallyastheyaredistantfromitsmarketmodel

politically.

MedRegisanambitiousattempttohaveinformalregulatorycoordinationsucceed

where many other formal policy frameworks have failed: building infrastructure

around theEUborders (the so-called“energycorridors”, seeAhnerandGlachant

(2014) aswell as Escribano (2010), Escribano (2011),Abbasov (2014)) thatwould

achieve many important objectives at once: increase its security of supply by

diversifyingitssupply,representnewmarketsforitsenergycompanies,greeningits

energyproductionandconsumptionandeventuallyloweringitsenergyprices.

215

The experience of CEER had shown the remarkable influence that informal

transnationalregulatorycooperationcouldhaveonregulatorypolicyandevenonEU

legislationassuch;regulatorswishedtoreplicatethatexperienceinthecontextof

the Euro-Mediterranean energy cooperation. The experience of ERRAhad shown

that hands-on cooperation between regulators from advanced and emerging

markets couldnotonlyquickly familiarize the latterwith theprinciplesofenergy

marketsbutalsotransformtheminto“agentsofchange”intheirdomesticsetting.

Therefore,EUregulatorssawtheestablishmentofanetworkofregulatorsasa“first-

best”strategytofosterenergymarketintegration,ratherthanasecond-bestoption

toobviatetothelackofwillingnessofSouthernneighbourstocomplywiththeEU

regulatory framework. Interviewees explicitly argued that informal networked

cooperationwouldbemoreeffectivethantop-downimpositionofuniformruleson

wildlydifferenteconomies.ThreecitedERRAasanexampletheinitiativetosetup

MedRegwasinspiredto.Oneaddedthat,however,MedRegismuchmoreambitious

in that its aim is to contribute to the realizationofmarket integrationacross the

shores of the Mediterranean, not just transforming into a training providing

organization(interview3).

Yet, theMedReg thus far does not appear to be able to exert a real impact on

regulatory policy concerning the region (Cambini and Franzi 2013). Several

interviewees from within and without the MedReg have expressed scepticism

concerningitsabilitytoinfluencepolicy inthenearfuture.Theyhavearguedthat

differencesbetweenthetwoshoresoftheMediterraneanarestilltoosignificant.

“Fornow, theMedReg isdiffusingknowledge,bestpractices, youknow. It

does not haveany influenceon legislation, however, not evenanadvisory

function.It’snotlikeMedRegmakesproposalsthatarethendiscussed…Itis

succeeding in providing capacity building, in creating a shared body of

knowledge also by sharing the progressmade in the various countries on

216

differenttopics,forinstance,ItalyandSpaincansharetheirexperiencewith

renewablesetc.”(interview5).

Indeed, the evolutionary trajectory ofMedReg differs from that of the networks

previouslyexamined.Virtuallynotimehaselapsedfromthebeginningofdialogueto

semi-formalization. In the other three cases, regulators transformed the

accumulated output of their cooperation into training courses and repository

materialforfuturestaffattheirregulatoryauthorities;thisdoesnotseemtobethe

case for MedReg. Most importantly, the sort of radical policy overhaul that

characterized the cases previously considered seems missing. Therefore, most

national regulatory authorities in the Southern neighbourhood (as shown by the

missing links and nodes in chapter 5) appear mostly confined to their national

boundaries. There is no interdependence motivating continuous and growing

interaction.

Althoughfallingshortoftheambitiousgoalsitsfoundershadsetforit,MedReghas

managed,inthepasttenyears,topositivelyaffectitsmembersviathesocialization

mechanismsthatmanycontributors (IkenberryandKupchan1990,Sabatier1991,

Checkel2005,Beecher2012,Bianculli2013,Bianculli,Jordanaetal.2015,Freyburg

2015)andthisthesishaverecognizedtopolicynetworksgenerallyandtoregulatory

networksinparticular:

“MedReghasenabledaregionalvision(…).Cooperatingbetweenregulators

withoutanykindofcommitmentfromthegovernmentisveryimportantto

create mutual understanding. (…) Second, it created what we can call a

MedRegcommunity,weknoweachothernow,weareanassociation(…) I

justpickthetelephone,Icallsomebody,Iknowtheminperson,askwhatyou

do,Igetnormallyanappropriateresponsebecausetheyalsounderstandthey

respondtoapersonwhoknowsthemandexpectsthemtorespond,sothis

hascreatedakindofcooperationevenonabilateral…whichwasverymuch

important for exchanging experience. Third one is exchanging experience,

217

which is basically either from success storyor failed story, bothareof the

samevalue.Fourthone is capacitybuildingprogrammes (…).The fifthone

whichisverymuchimportantisworkingjointlyintheworkinggroups,actually

they representa forum,ora think tank for regulators to stick together, to

comeup,actuallywedon’thaveanyagendaimposedonusbysomebodyelse,

we created our agenda and work plan, this has helped in developing the

knowledgeofeachother,throughdevelopingthesereports,andIbelievethe

reports MedReg is producing are high quality and address many of the

problemsinawaythatmakesthemclearlyunderstood”(interview21).

However,certaincircumstancesmustbepresentforinformaltransnationalnetworks

ofenergyregulatorstoactuallyaffecttransnationalpolicyformulation:thepresence

of common goals, sufficient uncertainty to open opportunity structures, network

entrepreneurswillingandabletoshapeandimplementthenetwork’sagenda,and

partnership with a supra-national source of political authority. These conditions

were,andstillarelackinginthecaseofMedReg.TheParliamentaryAssemblyofthe

Mediterranean (PAM)48 could be a suitable interlocutor, but lacks the necessary

cohesionandlegitimacy(interview3,4).

Atanyrate,thecaseofMedReghasbroaderimplicationsforourunderstandingof

thedynamicsthatplayoutintheexternaldimensionofEuropeangovernanceand

theroleofindividualpolicyentrepreneursnavigatingthem.Theexistingliteratureon

thesetopicsis,onceagain,overwhelminglyfocusedontheEuropeanCommission’s

initiativesandstances.TheCommissionisdescribedasoverlyprescriptiveandrigid

in itsapproachtothecountriessurroundingthecurrentbordersoftheEU(Bicchi

2006) and overall unable to deal with countries whose will to comply with the

European legislativeand regulatory frameworkcannotbe shapedbymembership

conditionality (Schimmelfennig2012).Responses to these statements,whichmay

includeconsiderationofthekindofinstitutiontheEuropeanCommissionisandthe

48PAMwebsitehttp://www.pam.int/?m=news&id=575(lastaccessed15August2017)

218

well-definedlimitstoitsdiscretionindevisingtailoredpolicysolutionstoitsdiverse

neighbourhood,arebeyondthescopeofthischapter.

Thechaptershowedthatinformalnetworkscanbeconceivedas“first-best”solution

tothornypoliticalandeconomicissues.Howeverincompleteintermsofnecessary

conditions for success, the regulators’ perception that informal cooperation can

succeedininfluencingregulatorypolicyinorderformarketintegrationandtradeto

takeplacewasbasedontherealityoftheirlivedexperienceandinobservationof

developments in other areas of the world. This fact alone testifies to the

appropriateness of investigating transnational expert cooperation usingmethods

that stem fromacknowledgementof interdependenciesbetweenpolicydecisions

andoutcomes.

Finally,thecasesofCEERandMedRegembedacommoncallforthestudyofpolicy

entrepreneurshipintheEUasresultingfromtheabilityofindividualactors,bethey

placedatnationallevelorinanyofthelociofpolicypressure,tofurthertheirpolicy

goalsbydovetailingthemtothepolicyprioritiesoftheEU.

219

PartIII.Conclusion:networksfromnetworks.

Recentcontributions(e.g.Slaughter,2017)stronglyadvocatedformorerelianceon

social networks in the context of foreign policy. The bulk of the literature on

transnationalregulatorynetworksconcernsregulatorsfromeconomicallyadvanced

countries.Inthisliterature,theissueofnetworkestablishmentisusuallytackledin

either functional or intergovernmental terms, dependingonwhether emphasis is

placed on the advantages of coordination or on the incentives that powerful

countries have to set global standards close to their own. A smaller set of

contributions focusesonnetworksof regulators fromdevelopingcountries; these

contributionsareusuallydescriptiveandrarelyemphasisenetworkestablishment.

Somemention, in passing, the support of international organisations or financial

institutions. To my best knowledge, the literature does not, as yet, feature

contributionshighlighting and investigating the linkbetween regulatorynetworks

andforeignpolicy.

Thenarrativesdevelopedintheforegoingchaptersprovidesomeinsightconcerning

howconcretelynetworkfitforeignpolicyagendas.Thechaptersshowthat,inboth

cases,informalregulatorynetworkswerereliedupontofosterexperts’collaboration

increatingtheconditionsforpolicychangetooccur.Inthefirstchapter,Iretracethe

processofestablishmentofthenetworkofenergyregulatorsofCentralandEastern

Europe,whichemergedin1998andhas,sincethen,benefitedfromtheunwavering

supportoftheUSAgencyforInternationalDevelopment(USAID)andNARUC.Inthe

secondchapter,Iretracetheprocessesleadingtotheestablishmentofthenetwork

of Euro-Mediterranean energy regulators; a unique case of strong institutional

entrepreneurship of European national regulators to establish and consolidate

expertcollaborationwithenergyregulatorsintheEuropeanneighbourhood.

220

The reconstruction of their respective histories shows that the establishment of

regulatory networks, particularly in emergingmarkets,may not just followpolicy

change,butratheraccompanyorunderpinit.Moreover,thechaptersshowthatthe

emergence of ERRA and MedReg should not be understood exclusively as the

consequenceofinternationalinstitutions’coercivepressures,withlittleornoagency

onthepartofregulators.Tothecontrary,inbothsettings,regulatorstookanactive

stanceinadvocatingforsupporttotheircollaborativeactivitiesfrominternational

institutions.The interviewdata informingbothchapters juxtaposestheregulatory

andthedonorperspective,anddemonstrates theco-ownershipof theprocessof

networkestablishment.Hence, in these two cases, network formation involveda

dialogicprocessbetweendonorsandrecipients.

SeizingthewindowofopportunitycreatedbytheEuropeanenlargement,Southern

European pre-emptied the European Commission’s typically institutionalized

approachtoexternalgovernancebyespousingandpromotingtheconceptofbottom

upcooperationandthebenefitofpeer-to-peer interaction.Thereafter, leveraging

theaccessthat,arguably,adecadeofwithin-CEERcollaborationhadaffordedthem

totheEuropeanCommission,theywereabletobrandtheirinitiativeaspromoting

the goals of the European external governance and obtain funding for it. Twelve

yearslater,MedRegisatitsfourthfundingrenewalandexpandingbothitsstaffand

thescopeofitsinitiatives.

AsoneMedRegregulator(fromtheSouthernneighbourhood)put it,MedReg isa

success story in termsofpromoting thebenefitsof regulationper se inemerging

markets:

“Spreadingregulatorycultureisveryimportantbutmostimportantistoprove

thatregulatorscanbringaddedvaluetothesystembecausemanycountries

feel like, we have electricity, we are a running system, we don’t have a

shortage,whydoweneedaregulator?It’simportanttoprovefromakindof

successstorythatthisisverymuchimportant,thereisaconflictofinterestin

221

thesystemifthesupplierisaplanner,ifthesupplierisadecision-maker,ifthe

suppliersetsthetariffs,orthegovernmentsetsthetariffs.Thisisakindofa

culturewhichneedstobeprovedsoaswebuildasuccessstoryIbelievethis

wouldbeagoodexporttoothers”(interview21).

The initial aim of ERRA was socializing regulators with each other and with the

principles of the market economy, in preparation for their transition to liberal

democracyandtheirentranceintotheEuropeanUnion.OnceEuropeanaccession

occurred,ERRAlostitsinitialfunctionandhadtoreinventitselfinordertomaintain

its viability. ERRA operated a strategy of conversion, whereby it substituted its

previousmainrationalewithanewone.Incontrast,MedReghasnotyetreachedthe

point when its initial goal (promoting infrastructure investments and regulatory

alignment in order to foster energymarkets integration across the shores of the

MediterraneanSea)isexhausted.WhiletheexistingdistancebetweentheEUandits

neighbours’ institutionsandmarketmodels isstillverywide,the longroadahead

mayalsoimplythatMedRegwillbeneededfortheyearstocome.

222

8. Conclusions

Thefindingsandstrengthsofthisresearch.

Sincetheearly2000s,theliteratureontransnationalregulatorynetworkshasliterally

boomed.NumerouscontributionsonregulatorynetworksintheEuropeanUnionand

elsewhere have debated the rationale of their establishment, their composition,

theirstatedaims,andtheirfunctioning.Theliteraturehasrarely,ifever,empirically

investigatedtheevolutionof regulatorynetworksover time, therebyportrayinga

static image of regulatory cooperation. Importantly, existing literature has

overwhelmingly focusedontheformalattributesofnetworks:statutes,meetings,

workplans,membershiplistsetc.,overlookingtheinformaldimensionofregulatory

networking which, arguably, underlies and sustain their very existence. Further,

existingliteraturelooksatnetworksfromtheoutside.Itrarelyexaminestheinner

workings of regulatory networked collaboration, the drivers of the choices that

regulators make concerning how to use their informal ties to each other. This

research has addressed these important gaps in the literature on regulatory

networksbyexaminingfourempiricalcasesoftransnational(ortrans-jurisdictional)

networksofenergyregulators(intheUSA,intheEU,inCentralandEasternEurope

andintheEuro-Mediterraneanarea)incomparativeperspective.

Inotherwords,existingliteraturehasoverlookedagreatextentofthepolicyandthe

politics of regulatory networking. In addition,many empirical puzzles concerning

networkshave,tothebestofmyknowledge,neverbeenaddressed.Forinstance,if

the impact of networks on regulatory convergence and/or global governance is

scarcelydiscernible,asextantliteratureconcluded,whydotheystillexistandkeep

being established around the world? The literature agrees that networks are

instrumentsfortheexchangeofinformationandexperiencesandexperiences,but

rarelyendeavouredtostudyhowexactlythisexchangetakesplace,whichincentives

223

drive regulators to choose their closest network partners and how concretely do

regulators benefit from networks. Moreover, the literature seems to implicitly

assume that regulatory networks remain confined to the jurisdiction of their

members (like organisations do). This research shows this assumption to be

incorrect, thereby opening a promisinglywide avenue for future research on the

exportofregulatorynetworkstootherjurisdictionsandtheunderlyingrationale.In

pointoffact,thisresearchuncoverstheroleplayedbyinformalnetworksinlinking

verydistant(geographicallyandpolitically)areasoftheworldvirtuallyseamlessly.

Thisthesistacklestheseopenquestionsbymeansofaresearchdesignthatfocused

onfourinter-relatedstoriesofregulatorynetworkestablishmentandevolution,as

wellasquantitativenetworkanalysisofregulators’interactionswithinnetworksand

theirmotives.Isituatedthefournetworksanalysedinthisthesisintheirinstitutional

environmentandunderlinedtheopportunitiesandconstraintsitposedtothemat

differentmoments,aswellastheincentivesmotivatingtheirinteractionastheyco-

evolve(Trein2017)withtheirexternalenvironmentandwiththeprioritiesoftheir

maininstitutionalinterlocutors.

Theprecedingsixempiricalchaptersaredividedinthreeparts.Inpart1,Iinvestigate

the spontaneous emergence of regulatory networking, its consolidation and its

persistenceover time. Inpart2, I investigatehowregulators choose their closest

networkinterlocutors; Ifindthatsimilarity inthepoliticaleconomy,expertiseand

resources drive network ties. In part 3, I investigate how networks foster the

establishmentofothernetworks,aspartoftheirpoliticalprincipals’foreignpolicy

agendas, in order to re-invent themselves or as a result of institutional

entrepreneurship.Thenarrativesof the fournetworks Iexamine inter-relateover

time; they are all interconnected and, arguably, nonewould have existed as it is

withouttheothers.

Inpart1,Icarryoutacomparativehistoricalanalysisoftheemergence,consolidation

andevolutionofspontaneousregulatorynetworking in theempiricalcasesof the

224

NationalAssociationofRegulatoryUtilityCommissioners(NARUC)oftheUSAandof

theCouncilofEuropeanEnergyRegulators(CEER)oftheEU.Theformergathersthe

50 Public Utility Commissions (PUCs) of the USA; the latter, the 29 National

RegulatoryAuthorities(NRAs)oftheEU.Irelyoninductivereasoningtoidentifythe

conditionsforspontaneousregulatorycooperationtoemerge,toconsolidateintoa

semi-formal networked organization (this is the point of departure of existing

literature), and to evolve as their initial rationale declines in importance. The

historicalnarrativeshowsthatnetworksemergeasregulatoryauthoritiesdo.Inboth

theUSAand theEU, theestablishmentof regulatoryauthoritiesatnational/state

level represented a radical overhaul of the governance of the energy sector.

Appointedregulatorshavetocontendwithmanyconflictinginterestsandlearnthe

ropesofacompletelynewprofession.Beinguniquelyresponsiblefortheregulation

of the sector in their territorial jurisdiction, they have no counterparts and no

precedenttoreferto.Hence,theystartreachingouttooneanothertocompareand

contrastthechallengestheyfaceandthepossiblesolutions.Atthisstage,regulatory

networkinghasnostructure;regulatorskeepincontactoccasionally.

The nature of the ties linking regulators changes and becomes more politically

consequentialasregulatorsfacetheprospectoflosingtheirpowersorofseeingthe

scope of their authority greatly restricted. The fact of facing a common threat

cements their collaboration into an institutional structure with a name,

headquarters,abudget,workinggroupsandaworkplan;inotherwords,withthe

intentionof lasting.AsfederalagencieswereestablishedinthenascentAmerican

Regulatory State, they faced the same informational gap as their state level

counterpartsafewdecadesearlier:anewprofession,consequentialtasks,andvery

little informationor guidanceonhow to carry themout.Hence, they sought the

collaborationoftheirstatelevelcounterpartsandusedpartoftheirfederalfundsto

provideforastructureddialoguewiththemintheformofanassociation(NARUC)

headquarteredinWashingtonDC,closetofederalagenciesoffices.Hence,federal

agenciesprovidedtheseedfundingforNARUC.Astimewentby,however,legislation

225

increasinglyempoweredfederalagencies,effectivelyestablishingtheirprimacyover

state regulation. As a consequence, state level regulators witnessed the federal

agencies’ progressive empowerment at the expense of their own control over

utilitieswithintheboundariesoftheirstate.Asitbecameevidentthatthetension

between state and federal level regulatory agencies would become a trait of

American regulatory federalism, the foundations for NARUC to last as PUCs’

collectiverepresentationtowardsfederalregulatorsandgovernmentwerelaid.

At first sight, European regulators had no supra-national contenders to their

authority, asMember States always opposed the creation of supranational Euro-

regulators.ForEuropeanregulators,themainthreattotheirautonomyandauthority

wastheirnationalgovernments.AstheUKinfrastructuresectorreformsintroduced

privatecapitalandcompetitionintothe,thusfar,monopolisticenergysector,they

settheexampleforseveralotherEuropeanMemberStatesand,mostimportantly,

fortheEuropeanCommission.AsEuropeanlegislationmandatedtheliberalmarket

model in the energy sector across all of the EU, Member States had to create

independent regulatory institutions. A century later than their US counterparts,

Europeanregulatorsweretaskedwithimmenselyconsequentialmarketreforms,for

which theyhad todesign the rules.Theystarted reachingout toeachother,and

meetingoccasionallyandinformallyinvariousEuropeancapitals.

As the market reforms envisaged by European legislation triggered societal and

business backlash, governments across Member States sought to intervene in

regulatory decisions and to restrict the scope of the regulatory powers they had

delegated.Asitbecameevidentthatthetensionbetweengovernmentinterference

andregulatoryautonomousdecision-makingwouldremainatraitoftheEuropean

systemofmulti-levelgovernance,Europeanregulatorsdecidedtoconsolidatetheir

informalnetworkties intoan institutionalstructurewithaname,headquarters in

Brussels, i.e. close to the EU decision-making, budgetary contributions, etc.

Transforming their collaborative ties into the CEER provided them with a solid

226

platform of resistance to government’s intervention and of influence into the

Europeanpolicyprocess.DifferentlyfromtheirUScounterparts,Europeanregulators

self-financedtheirnetworkedcollaborationfromtheverystart.TheCEERhasalways

beenacompletelyindependentnetwork.

Hence,whereastheemergenceofahigherlayerofauthorityconstrainedUSPUCs’

powers,asthoseweresubjecttothefederalagencies’, itpartiallyfreedEuropean

regulators from government interference. In both cases, however, the glue of

regulatorynetworkingfortheirjointpursuitofcontrol,andauthority.Thequestfor

preserving, if not expanding, the scope of their authority was the benefit that

outweighed the costs of cooperation and the partial sharing of authority that is

impliedinregulatorycollaborationandcoordination(Macey2003).Thisisthefirst

key empirical finding of this thesis and, arguably, one of its main theoretical

contributions.

Ifcontrolwastheideationalrationaleunderlyingtheconsolidationofoccasionalties

into a semi-formal organisational arrangement with a name, a budget, and

headquartersclosetothecentreofpolicyandpowerintheirrespectivejurisdiction,

information and expertise are the assets that regulators aimed atmaximizing via

theircollaboration.Inchapters4and5IanalysethestructureoftheCEERnetwork

asatypicalcaseofregulatorynetworkwithalonghistoryofintensecooperationand,

therefore, one single component and no isolated members. Via the network

structureoftheircollaborativeties,regulatorscanreachanyoftheirpeersinasmall

numberofsteps.However,regulatorsareunlikelytoberegularlyincontactwithall

of their counterparts in their network. More plausibly, individual regulators will

maintainregular,frequentbilateraltieswithasubsetoftheirpeers.Themotivations

driving the choice of that subset are the dependent variables of the quantitative

networkanalysiscarriedout inchapters4and5.Amodelof thedeterminantsof

network structure showed that regulators choose partnerswho oversee similarly

structuredmarketsastheirs,asthosearethepeerswhoaremostlikelytohavecome

227

acrosssimilarchallengesastheirs.However,alltendtopursueregularrelationships

with regulators overseeing more advanced markets, as they are likely to have

accumulatedmoreexpertiseasconcernsdesigningfunctioningenergymarkets. In

chapter5, I investigate the linkbetweenresourcesandnetworkactivism(already

emergedinchapter4)indetail.Thefindingsarethatnetworkactivismisdetermined

byscarceresources,andindicatesthatregulatorsuseinformaltiestocompensate

fortheirlackingresources.Staffnumbersaremorerelevantthanbudgets.Moreover,

the relationship between resources and activism appears to follow a quadratic,

rather than linear, trend: active regulators are those with intermediate to small

resources,whileregulatorswith largeorverysmall resourcesarenotsignificantly

different.Thisshowsrationalityintheusageofnetworksandpointstothecategories

of regulators for which it may be most useful; for very resource-constrained

regulators,informalnetworkingbecomesprohibitivelycostly.

Ultimately,bothchaptersconcurinshowingthattheliterature’semphasisonglobal

governancewhendiscussingtheworthofregulatorynetworksisnotmisplaced,but

it is seriously incomplete. Regulators possess statutory powers over a specific

territorialjurisdiction;theyhavenoofficialpowersbeyondit.Hence,theyusetheir

networks to improve their regulatory practice in their own jurisdiction, alongside

attempting to coordinate to achieve more regulatory convergence across

jurisdictions.Thisisthesecondkeyempiricalfindingandtheoreticalcontributionof

this thesis. Regulators need information in order to produce quality regulation.

Qualityregulationistheonlyassettheyholdandcanwieldtoprotecttheirauthority.

Inordertoregulateeffectivelyandtobuildtheircredibilityandlegitimacy,regulators

needresources.Wherevertheylacksufficientresourcestocarryouttheirtasks,they

resorttotheirnetworktocompensate.

Thethirdandfinalpartofthethesistacklestheissueofnetworkexport:namely,the

consciousdecisiontoexportnetworks(their formaland informalstructures) from

economicallyadvancedjurisdictionstoemergingmarketsjurisdictions.Itacklethis

228

issueinchapters6and7,byreconstructingtheemergenceoftwonetworks,which

canbeconsideredtheprogenyof,respectively,theNARUCandtheCEER:theEnergy

RegionalRegulatoryAssociation(ERRA)andtheAssociationofMediterraneanEnergy

Regulators(MedReg).TheUSgovernment,throughitsdevelopmentagency,stood

behindtheestablishmentofERRAandsupporteditforitsfirsttenyearsofexistence.

Inturn,theEuropeanCommissionstandsbehindtheexistenceoftheMedReg,and

hasbeensupportingitsinceitsestablishment11yearsago.Theprimaryaimofthese

chapters is, therefore, to inductively identify thereasonswhytheUSgovernment

andtheEuropeanCommissionfinancedtheestablishmentof regulatorynetworks

modelledontheirowninotherpartsoftheglobe,andwhytheysupportedthemfor

a very long time. In other words, the aim of these chapters is discovering why

informal regulatory networks have been deemed capable of entrenching policy

change.

Astheanalysisunfolded,however,twoveryimportantadditionalthemesemerged

fromthedata.Thefirstone,emergedinchapter6(dedicatedtoERRA),isthetheme

ofnetwork survival and, specifically,ofnetworkevolutionover timeasa survival

strategy.As themain rationaleofnetworkexistencedeclines inpolitical salience,

network managers and members seek to re-invent their network to prevent its

decay.InthecaseofNARUC,theunexpectedopportunitytocontributeintheeffort

offacilitatingthetransitionofCentralandEasternEuropeancountriestothemarket

economyrepresentedalayeringstrategy,addingtothebaselinefunctionsofNARUC

aninternationalprong,whichquicklyrosetobeingitsmainsourceofincome.The

otherone,emergedinchapter7(dedicatedtoMedReg),isthethemeofinstitutional

– or, in this case, network – entrepreneurship. In this case, I define network

entrepreneurshipasthedecision,onthepartofanindividualregulatororasmallset

of individual regulators, to further regulatory coordination purposes by

autonomouslyinitiatinginformalnetworks,therebypre-emptingpoliticalprincipals

frompursuingthosesameobjectivesthroughtop-downinitiatives,suchassettingup

of formalorganizations. Inotherwords,networkentrepreneurs anticipate formal

229

institutionalactionbytheirpoliticalprincipalsanddecidetoshort-circuit themby

initiatinginformalnetworks.Inaddition,theymanagetopresenttheirinitiativeas

legitimateandproactiveandtosuccessfullyadvocateforfinancialsupportoftheir

efforts.

Chapter6narratesofhowNARUCbecamethefaceofenergyregulationworldwide.

In the aftermath of the collapse of the Soviet Union, the USAID undertook an

institutionbuildingprogrammeinCentralandEasternEurope,withtheultimateaim

toplanttheseedsoftheliberalmarketeconomyandfostertheaccessionofEastern

EuropeancountriestotheEU.Regulatoryinstitutionsfortheenergymarkets-to-be

were created in the first recipient countrieswith the aid of external consultants.

Soon, the mechanisms identified in chapters 2 and 3 started operating: newly

establishedregulatorsinCentralandEasternEuroperealizedthemagnitudeoftheir

taskandstartedinformallyreachingouttoeachothertocomparetheirexperiences

andchallenges.TheUSAIDstaffmonitoringtheprocessproactivelydecidedtofoster

thatspontaneouscollaborative incentive;theycontractedNARUCtoprovidetheir

expertsupporttotheemergingregulatorycooperation.NARUCacceptedthetask:

byworkingwithUSAID inEasternEurope, it gainedanew,politically crucial role;

moreover,itsmembersobtainedthepossibilitytotravelinternationallyanddiffusing

the principles of long-standing American regulators to the formerly communist

world. Collaboration within ERRA quickly gained pace and regulators grewmore

secure in their role and in their decision-making authority.As regulators onboth

sides of the European continent (the liberal and the communist) matured, the

Europeanbigeastwardenlargementof2004broughtsevenERRAmembersintothe

CEER.ThetaskunderpinningthecreationofERRAwasaccomplished.

Twoeventsintheearly2000screatedhugepoliticalmomentumforwhatwasnamed

the“externalgovernance”dimensionoftheEuropeanUnion.Ontheonehand,the

2004 eastward enlargement convinced many of the ability of the EU to export

prosperity;ontheotherhand,the2005signatureoftheEnergyCommunityTreaty

230

represented a first step towards the future accession of the countries of ex-

Yugoslavia,thenrenamedasSouthEastEurope,whichweredeemedunfittoenter

theEUin2004.TheTreatymandatedtheintegrationoftheenergymarketsofSouth

EastEuropewiththeEU’sthroughahighlyformalizedsetofrulesandinstitutional

structures. According to European regulators (primarily, the Italian, French and

Spanish),theestablishmentoftheEnergyCommunityasaninternationallegaltreaty

hadputa straightjacketon regulatorycooperationbystifling regulators’ initiative

andtyingtheircollaborationtoMinisterialdecision-making.

Hence,theydecidedtopreventtheextensionofthe“EnergyCommunityformula”

to the other side of the European neighbourhood: the Southern neighbourhood,

comprisingcountriesfromtheMiddleEastandNorthAfrica(MENA).Byexploiting

theirexistinginformalcontactswithregulatorsintheareaandtheirownexperience

ofcollaborativegovernancewithintheCEER,Europeanregulatorsjoinedforceswith

theirMENAcounterpartsandcreatedMedReginaveryshorttime.Contrarilytothe

literature on European integration, which saw networks as “second-best”

instruments of convergence (Dehousse 1997), European regulators proclaimed

networksasa“first-best”optionforframingtransnationalregulatorycollaboration,

offering regulators freedom from political interference, confidentiality and open

discussion.However,thischapteralsoconfirmsthefindingsofchapters2and3,as

well as chapter 6, as concerns the importance of uncertainty and windows of

opportunityforregulatorstoleveragetheirnetworktoinfluencepolicy,byshowing

theirabsenceinthecaseofMedReg.

This thesis shows that informal regulatory networks are levers. Regulators use

networksasleversthatraisetheirdomesticlegitimacyandhelpthemfendoffattacks

totheirauthority;theyshapetheirnetworksaccordingtotheinformationandthe

resourcestheyneedmost;theyleveragetheirnetworkstocontinuenetworkingand

retainingaprecious, cost-effective, resilient instrumentof regulatorygovernance.

Collectively,thefindingsofthisresearchshowthatnetworksareneitherrationally

231

designed by policy-makers, nor do they functionally descend from specific policy

targets,asmostoftheliteraturemaintains.Thereisnoexternal“mastermind”tothe

inception of cooperation: rather, regulatory networks are the outcome of a very

spontaneousaggregationprocessdrivenby the regulators’quest for information,

resources,andcontrol.

This research also makes a statement concerning the research approach. The

questions guiding the research (What explains network emergence? Why do

networks last? How can they influence policy?What is the agency of regulators

withinnetworks?)need tobeansweredby relyingonacombinationofmethods.

Comparative historical analysis helps identifying the critical junctures setting in

motionemergentprocessesofnetworkformationandappreciatingtheevolutionary

trajectory of networks, pointing to the lasting importance of sequencing and

feedbackeffectsfordeterminingwhichoptionsofnetworkevolutionareavailableat

any given time. Understanding regulatory networks as interdependent sets of

relationships, rather than as independent observations resulting from political or

administrative functional calculation, allows the identification of the

interdependencies linking regulators and networks across time and space;

quantitativenetworkanalysisallowsexplainingnetworkstructureparsimoniously.

Thelimitsofthisresearch.

Therearemanyissuesintheliteratureonnetworksthatlieoutsidethescopeofthis

thesis.Ididnotinvestigatepolicyoutcomesthatareconducibletonetworkinfluence

eitheratsupranational(federal)oratnational(state)level.Twomainmechanisms

ofinteractionacrosslevelsofgovernanceemergedfromtheanalysis:themechanism

whereby the coordination within networks and the resulting policy

recommendationsorpositionsinfluencesupra-national(federal)policy,whichisalso

the most studied in the literature; and the mechanism whereby information

232

gatheredfromthenetworkaffectstheregulators’abilitytoachievepolicyoutcomes

athome.Suchafocuswouldundoubtedlyenrichtheexistingconspicuousliterature

onpolicydiffusionacrossnetworks(MintromandVergari1998,Cao2009,Cao2010,

Cho2013,Desmarais,Hardenetal.2015,Boehmke,Ruryetal.2016).Therearetwo

reasons why I did not engage in analysis of these two immensely important

mechanismsofnetwork impactonpolicy:mymain interest lied in identifying the

reasonswhyregulatorsnetwork,thebenefitstheyobtainfromnetworksthatthey

couldnotachieveotherwise;settingouttoascertainingtheimpactofnetworkingon

policyrequiresdetaileddata,acompletelydifferentresearchdesign,and,inshort,

anotherPhDthesis.

Asecondlimitationofthisresearchisitsfocus:itisveryregulator-centric.Idonot

investigatehowandwhyregulatorsnetwork(individuallyorcollectively)withother

stakeholders.NordoIresearchinteractionsbetweenregulators’andstakeholders’

networks,whichisanotherimportant,butunder-investigatedtopic(Kalaitzake2017)

thatcroppedupinseveralofmyinterviews.Anumberofintervieweesreportedthat

regulators network in response to industry networks. As a matter of fact, the

establishmentofnetworksofregulatorsalwaysfollowstheestablishmentbusiness

orindustrynetworks:businessorganisesearlier,faster,andinamoreconcertedway

fromthestart,giventheircommongoalofensuringtheirownprofit.Thispolitical

economic angle on regulatory networking ties in with extant literature on the

importanceofbusiness’advocacyformarketintegrationintheEU(Mattli1999)and

integratesmyexplanationofregulatorynetworkingonthebasisofregulators’self-

interestswell.Onceagain,however,thisimportanttopicwouldneedtobeaddressed

on its own and would require a dedicated data gathering process and research

design,or,inotherwords,aseparatePhDthesis.Thetopicofstakeholders’networks

linkswiththerecentemphasisplacedonepistemiccapture(Sebenius1992,Dunlop

2009) and the growing interest in the frames throughwhich regulatory policy is

made. Epistemic capture means that, by interacting with regulated interests,

233

regulatorsincreasinglyseethesectorandthepolicyissuesthroughtheframesofthe

regulatees(Sunstein2014).

Finally, a third important topic that is missing from this research is the topic of

accountability. Slaughter (2000a) affirms that regulators’ accountability duties at

nationallevelsufficetolessentheproblemoftheopacityofnetworkcollaboration.

HelleinerandPorter(2010)andPapadopoulos(2007),however,woulddisagreeand

proposevariousmechanismsofcontrolbydemocraticinstitutionsoverthedesign,

the composition and the output of these networks. This research shows that

regulatorscherishtheconfidentialityandprivacyoftheirnetworkcommunication

aboveall.ThisisperhapsparticularlythecaseforEuropeanregulators;theonlyones,

ofthecasesexaminedinthisthesis,tohaveself-financedtheircooperativeactivities

from the start with the explicit aim of preserving their independence. Placing

networkunderpoliticalcontrolmayweakenthemechanismsofproductivenetwork

collaboration,whilenotnecessarily submittingnetworks topublic scrutiny; this is

what prompted Southern European regulators to establishMedReg: avoiding the

straightjacket of formalized cooperation overseen by politicians. These are

normativeissuesofextremeimportance,whichdeserveseparateconsiderationand

couldnotfindspaceinthecontextofthisresearch.

Generalimplicationsofthisresearchandfutureresearchagenda.

Theimplicationsofthefindingsofthisresearcharewide-ranging,andconcernboth

academic investigation and policy-making. From an academic point of view, the

findingsoftheresearchtestifytothenecessityoftakingabroaderperspectivewhen

investigating networks of regulators. Functional explanations of network

establishment and rationale only examine the surface of regulatory interactions.

Focusingattentionontheincentivesandinterestsofpoliticalprincipalswithregard

234

to networks only tells a part of the whole story, and not necessarily the bigger

portion.

Regulatorsnetworkforinformation,resources,andcontrol.Oncecreated,regulatory

institutionscanbeexpectedtostrivetoremainbyre-inventingthepurposesoftheir

cooperation as needed. Regulators are homophilous in their tie choices, which

implies that spontaneous interaction may lead to cliques and disconnected

communitiesorisolatednodes.Thisfactencouragesthecreationofsomeformof

structuredcooperationintheformofagencies,aswasthecaseinACER,particularly

ifregulatorycoordinationhastobefunctionaltotheachievementofcommongoals.

Moreover, the creation of agencies paralleling networks may assuage worries

concerning network accountability, yet avoid the temptation of setting external

constraints on informal networking, lest potentially hampering its advantages.

Networksarestructureswhereentrepreneurialregulatorscanthriveandpotentially

proposepolicysolutionstopressingpolicyproblems.Atthesametime,thefindings

of this researchpoint to the importanceof timeand repeated interaction for the

emergence of mutual trust. Particularly in the context of emerging markets,

withdrawingdonorsupporttooearlymayrisknetworkdisbandment.Forthepolicy-

makerintendingtofosternetworks,itisimportanttounderstandtheimportanceof

nurturingbutnotpilotingnetworks.

The ongoing technological overhaul in energy systemsmay provide the next big

opportunitywindowsfornetworksofenergyregulatorstopooltheirresourcesand

contributetopolicy-making,intheeconomicallyadvancedaswellasthedeveloping

world.Indeed,theworldofenergyproduction,transportationandconsumptionis

already experiencing the so-called “energy transition”. The shift away from fossil

fuels will causemajor reshuffles of winners and losers, and is already proving a

polarizing factor politically, with governments dividing between “green” and

“brown”andnurturingcorrespondingconstituencies(AklinandUrpelainen2013).

235

Theimportanceofregulationinthiscontextishardlyevermentionedintherelevant

policy debates. This is surprising, given that the energy transition is already

technicallypossible;oneofthemainreasonswhyitisnotyetaccomplishedisthe

absence of regulatory frameworks able to take simultaneously into account

technologicalnovelty,thestrandedcostsofextantinfrastructure,theframeworkfor

investments, the allocation of costs and benefits acrossmultiple actors, and the

myriadotherissuesthatcharacterizethesector.Indeed,modesofgovernanceare

animportantpartoftherelationshipbetweentechnologyandinstitutions(Künneke,

Groenewegenetal.2010).Thepursuitofcoherencebetweenthetwoshouldbeone

ofmainobjectivesofinfrastructurepolicy(Finger,Groenewegenetal.2005).

Thereasonwhyregulatorynetworksareplausiblygoingtobecrucialactorsinthe

implementation of the energy transition is the uncertainty it entails. Making

decisions under uncertainty is difficult and, as shown in this thesis, creates

opportunity structures for shaping the emerging new regime. In this perspective,

regulationisnotjustasetofrulesdesignedexogenouslyandregulatoryauthorities

are not just organizations tasked with enforcing them. Rather, regulation is an

emergent property of a complex system (Vazquez and Hallack 2017). Regulatory

networksalsoare.Asregulatorslearnfromtheoutcomesofthetechnologyandfrom

theoutcomesofpoliticaldecision-making,theyadaptandtheyshapetheregulatory

environment(andtherefore,theregulation)atthesametime(Künneke2008).The

regulatory frameworkof the futurewill need to keepdisplaying theattributesof

credibility, commitment, and sustainability in a radically changed technological

landscapeandanuncertainpoliticalenvironment(e.g.therecentwithdrawalofthe

USAfromtheParisAgreement).

Upcomingtransformationssuchaselectricitydecentralization(where,inshort,every

consumercanbea“prosumer”,oraproducerofelectricity)andtheintegrationof

InformationandCommunicationTechnologies(ICT)andelectricitygridsarelikelyto

multiplythescalesatwhichregulationisneededfromtheverylocaltotheEuropean

236

(or regional) to the international.Decentralization is likely toexpand the scaleof

regulatorymonitoringbydemandingmuchcloseroversightofDistributionSystem

Operators(gridoperatorsinchargeoflowvoltageelectricitygrids)thanperformed

atpresent.Inadecentralizedfuture,thelowvoltagegridwillbeunderamuchhigher

stress than it currently is, given that the number of consumers who are also

producers(i.e.peoplewithsufficientinstalledsolarpowertobeabletofeedintothe

grid) is set to increase thousand-fold.Digitalisationwill enable the full control of

one’s electrical consumption from remote – engendering significant privacy

concerns.Inotherwords,theworldofenergyregulationissettochangeinways,

unpredictableattheoutset.Theaccumulatednetworkexpertiseofenergyregulators

mayturnprovidentialinmanagingthenewsetofrelationshipsthatarelikelytoarise

across level of governance and dealing with uncertainty. The scope for

interdependenceisallbutenlarging,asallinfrastructuresectorshavetorespondto

thechallengesofclimatechangemitigationand,increasingly,startdevelopinglegal

andeconomicframeworksforfinancingadaptation.Regulatorsatalllevelsarelikely

to stand to exploit the structural properties of networks (reachability, flexibility,

interdependencies) as well as their relational properties (trust, routinized

communication, rapid access to information) to face upcoming challenges and

contributetonewpolicy.

237

References

Abbasov,F.G.(2014)."EU'sexternalenergygovernance:Amultidimensionalanalysisofthesoutherngascorridor."EnergyPolicy65:27-36.

Adams,E.andD.Tirpak(2008)."Bilateralandmultilateralfinancialassistancefortheenergysectorofdevelopingcountries."ClimatePolicy8(2):135-151.

Ahdieh,R.B.(2010)."ImperfectAlternatives:Networks,Salience,andInstitutionalDesigninFinancialCrises."U.Cin.L.Rev.79:527-551.

Ahdieh, R. B. (2015). "Coordination and Conflict: The Persistent Relevance ofNetworks in InternationalFinancialRegulation." LawandContemporaryProblems78(5):75-101.

Ahner, N. and J.-M. Glachant (2014). "In search of a'platform'for Mediterraneanrenewablesexchange:'EU-style'systemvs.a'corridor-by-corridor'approach."RobertSchumanCentreforAdvancedStudiesResearchPaperNo.RSCAS39.

Aklin,M.andJ.Urpelainen(2013)."PoliticalCompetition,PathDependence,andtheStrategy of Sustainable Energy Transitions." American Journal of Political Science57(3):643-658.

Alcañiz, I. (2010). "Bureaucratic Networks andGovernment Spending: A NetworkAnalysisofNuclearCooperationinLatinAmerica."LatinAmericanResearchReview45(1):148-172.

Alcañiz, I. (2016). Environmental andNuclearNetworks in theGlobal South:HowSkillsShapeInternationalCooperation,CambridgeUniversityPress.

Alcañiz, I. (2016). "Partnerselection in internationalenvironmentalnetworks:Theeffect of skills and money on cooperation in the Global South." EnvironmentalScience&Policy55,Part1:107-115.

Alesina,A.andD.Dollar(2000)."WhoGivesForeignAidtoWhomandWhy?"JournalofEconomicGrowth5(1):33-63.

Alexander,D.,J.M.LewisandM.Considine(2011)."HowPoliticiansandBureaucratsNetwork: a ComparisonAcrossGovernments." Public Administration89(4): 1274-1292.

Ambrus, M., ., eds. and et al. (2014). The Role of ‘Experts' in International andEuropeanDecision-MakingProcesses,CambridgeUniversityPress.

238

Anderson, D. D. (1980). State Regulation of Electric Utilities. The Politics ofRegulation.J.Q.Wilson.NewYork,BasicBooksInc.

Anderson, D. D. (1981). Regulatory Politics and Electric Utilities: A Case Study inPoliticalEconomy,AuburnHousePublishingCompany.

Aral, S. (2016). "The FutureofWeakTies."American Journal of Sociology121(6):1931-1939.

Arnold,G.(2014)."Policylearningandsciencepolicyinnovationadoptionbystreet-levelbureaucrats."JournalofPublicPolicy34(3):389-414.

Baccini,L.andA.Dür(2012)."TheNewRegionalismandPolicyInterdependence."BritishJournalofPoliticalScience42(01):57-79.

Bach,D.(2010)."Varietiesofcooperation:thedomesticinstitutionalrootsofglobalgovernance."ReviewofInternationalStudies36(03):561-589.

Bach,D.andA.Newman(2014)."Domesticdriversoftransgovernmentalregulatorycooperation."Regulation&Governance8(4):395-417.

Bach,D.andA. L.Newman (2010a). "TransgovernmentalNetworksandDomesticPolicy Convergence: Evidence from Insider Trading Regulation." InternationalOrganization64(03):505-528.

Bach, D. and A. L. Newman (2010b). "Governing lipitor and lipstick: Capacity,sequencing,andpowerininternationalpharmaceuticalandcosmeticsregulation."ReviewofInternationalPoliticalEconomy17(4):665-695.

Bach, T., F. De Francesco, M. Maggetti and E. Ruffing (2016). "TransnationalBureaucratic Politics: An Institutional Rivalry Perspective on EU NetworkGovernance."PublicAdministration94(1):9-24.

Bach, T. and E. Ruffing (2013). "Networking for Autonomy? National Agencies inEuropeanNetworks."PublicAdministration91(3):712-726.

Bach,T.,E.RuffingandK.Yesilkagit(2014)."TheDifferentialEmpoweringEffectsofEuropeanizationon theAutonomyofNationalAgencies."Governance28(3): 285-304.

Baldwin, R., C. Scott and C. Hood, Eds. (1998). A Reader on Regulation. OxfordreadingsinSocio-LegalStudies.Oxford.

Barberá,P.(2015)."BirdsoftheSameFeatherTweetTogether:BayesianIdealPointEstimationUsingTwitterData."PoliticalAnalysis23(1):76-91.

239

Barendse, A. J. (2006). Regional Regulation as a New Form of Telecom SectorGovernance: the Interactions with Technological Socio-Economic Systems andMarketPerformance,TechnicalUniversityofDelft.

Barry,A.(1996)."TheEuropeannetwork."NewFormations26:26-38.

Bearce,D.H.andD.C.Tirone(2010)."ForeignAidEffectivenessandtheStrategicGoalsofDonorGovernments."TheJournalofPolitics72(3):837-851.

Beecher,J.A.(2012)."NARUCasNetwork:aPerspectiveontheUSRegulatoryPolicyCommunity."Competition&Reg.NetworkIndus.13(4):362-388.

Benz,A.andB.Eberlein(1999)."TheEuropeanizationofregionalpolicies:patternsofmulti-levelgovernance."JournalofEuropeanPublicPolicy6(2):329-348.

Berardo,R.andJ.T.Scholz (2010)."Self-OrganizingPolicyNetworks:Risk,PartnerSelection,andCooperationinEstuaries."AmericanJournalofPoliticalScience54(3):632-649.

Berg,S.V.andJ.Horrall(2008)."Networksofregulatoryagenciesasregionalpublicgoods: Improving infrastructure performance." The Review of InternationalOrganizations3(2):179-200.

Bermeo,S.B.(2008)"DonorsandDevelopmentSectorAllocationinForeignAid."

Bermeo,S.B. (2011). "ForeignAidandRegimeChange:ARole forDonor Intent."WorldDevelopment39(11):2021-2031.

Bernstein,M.H.(1955).RegulatingBusinessbyIndependentCommission,PrincetonUniversityPress.

Berthélemy,J.-C.andA.Tichit(2004)."Bilateraldonors'aidallocationdecisions—athree-dimensional panel analysis." International Review of Economics & Finance13(3):253-274.

Bianculli,A.C.(2013)."TheBrazilianAssociationofRegulatoryAgencies:Integratinglevels, consolidating identities in the regulatory state in the south." Regulation&Governance7(4):547-559.

Bianculli, A. C., J. Jordana and A. G. Juanatey (2015). "International Networks asDriversofAgencyIndependence:TheCaseoftheSpanishNuclearSafetyCouncil."Administration&Society49(9):1246-1271.

Bicchi,F.(2006)."‘Oursizefitsall’:normativepowerEuropeandtheMediterranean."JournalofEuropeanPublicPolicy13(2):286-303.

240

Black, J. (2008). "Constructing and contesting legitimacy and accountability inpolycentricregulatoryregimes."Regulation&Governance2(2):137-164.

Black,J.(2009).LegitimacyandCompetitionforRegulatoryShare.LSELaw,SocietyandEconomyWorkingPapers. London, LondonSchoolof Economics andPoliticalScienceLawDepartment.

Blauberger, M. and B. Rittberger (2015). "Conceptualizing and theorizing EUregulatorynetworks."Regulation&Governance9(4):367-376.

Blauberger,M.andB.Rittberger(2015b).OrchestratingPolicyImplementation:EUGovernance Through Regulatory Networks. International Organizations asOrchestrators.K.Abbott,P.Genschel,D.SnidalandB.Zangl.Cambridge,CUP:39-64.

Boeger, N. and J. Corkin (2012). "How Regulatory Networks Shaped InstitutionalReformundertheEUTelecomsFramework."CambridgeYearbookofEuropeanLegalStudies14:49-73.

Boeger, N. and J. Corkin (2017). "Institutional Path�Dependencies in Europe'sNetworkedModesofGovernance."JCMS:JournalofCommonMarketStudies55(5):974-992.

Boehmke,F.J.,O.ChyzhandC.G.Thies(2016)."AddressingEndogeneityinActor-SpecificNetworkMeasures."PoliticalScienceResearchandMethods4(01):123-149.

Boehmke,F. J.,A.A.Rury,B.A.DesmaraisandJ. J.Harden(2016)."TheSeedsofPolicyChange:LeveragingDiffusion toDisseminatePolicy Innovations." JournalofHealthPolitics,PolicyandLaw42(2):285-307.

Bohle,D.andB.Greskovits(2012).CapitalistDiversityonEurope'sPeriphery,CornellUniversityPress.

Booth, D. (2011). "Aid, Institutions and Governance: What Have We Learned?"DevelopmentPolicyReview29:s5-s26.

Borgatti,S.P.andM.G.Everett(2000)."Modelsofcore/peripherystructures."SocialNetworks21(4):375-395.

Brandes,U.,P.Kenis,J.Raab,V.SchneiderandD.Wagner(1999)."ExplorationsintotheVisualizationofPolicyNetworks."JournalofTheoreticalPolitics11(1):75-106.

Broome,A.andL.Seabrooke(2015)."ShapingPolicyCurves:CognitiveAuthorityinTransnationalCapacityBuilding."PublicAdministration93(4):956-972.

241

Busuioc, E. M. (2016). "Friend or Foe? Inter-Agency Cooperation, OrganizationalReputation,andTurf"PublicAdministration94(1):40-56.

Büthe, T. (2002). "Taking Temporality Seriously:ModelingHistory and theUseofNarrativesasEvidence."TheAmericanPoliticalScienceReview96(3):481-493.

Cambini,C.andD.Franzi(2013)."AssessingtheEUPressureforRulesChange:ThePerceptionsofSouthernMediterraneanEnergyRegulators."MediterraneanPolitics19(1):59-81.

Cambini, C. and D. Franzi (2013). "Independent regulatory agencies and rulesharmonization for the electricity sector and renewables in the Mediterraneanregion."EnergyPolicy60:179-191.

Cambini,C.andL.Rondi(2011).RegulatoryIndependence,InvestmentandPoliticalInterference:EvidencefromtheEuropeanUnion.EUIWorkingPapers.Florence.

Cao,X.(2009)."NetworksofIntergovernmentalOrganizationsandConvergenceinDomesticEconomicPolicies."InternationalStudiesQuarterly53(4):1095-1130.

Cao, X. (2010). "Networks as channels of policy diffusion: Explaining worldwidechangesincapitaltaxation,1998–2006."InternationalStudiesQuarterly54(3):823-854.

Cao, X. (2012). "Global Networks and Domestic Policy Convergence: A NetworkExplanationofPolicyChanges."WorldPolitics64(03):375-425.

Cao, X. and H. Ward (2017). "Transnational Climate Governance Networks andDomesticRegulatoryAction."InternationalInteractions43(1):76-102.

Capoccia, G. and R. D. Kelemen (2007). "The Study of Critical Junctures: Theory,Narrative,andCounterfactuals inHistorical Institutionalism."WorldPolitics59(3):341-369.

Cardwell,P.J.(2011)."EuroMed,EuropeanNeighbourhoodPolicyandtheUnionforthe Mediterranean: Overlapping Policy Frames in the EU's Governance of theMediterranean*."JCMS:JournalofCommonMarketStudies49(2):219-241.

Carpenter, D. (2010). Reputation and power: organizational image andpharmaceuticalregulationattheFDA,PrincetonUniversityPress.

Carpenter,D.,K.EsterlingandD.Lazer(2003)."Informationandcontact-makinginpolicy networks: a model with evidence from the US health policy domain."RationalityandSociety15(4):411-440.

242

Carpenter, D. P. (2001). The forging of bureaucratic autonomy: Reputations,networks, and policy innovation in executive agencies, 1862-1928, PrincetonUniversityPress.

Cavatorta,F.andP.Rivetti(2014)."EU–MENARelationsfromtheBarcelonaProcessto theArabUprisings:ANewResearchAgenda." Journal of European Integration36(6):619-625.

Cengiz,F.(2007)."Managementofnetworksbetweenthecompetitionauthoritiesinthe EC and the US: different polities, different designs." European CompetitionJournal3(2):413-436.

Cengiz,F.(2013).AntitrustFederalismintheEUandtheUS,Taylor&Francis.

Checkel, J. T. (2005). "International Institutions and Socialization in Europe:IntroductionandFramework."InternationalOrganization59(4):801-826.

Childs,W.R.(2001)."StateRegulatorsandPragmaticFederalismintheUnitedStates,1889–1945."BusinessHistoryReview75(04):701-738.

Cho,C.-L.andD.S.Wright(2004)."TheDevolutionRevolutioninIntergovernmentalRelations in the 1990s: Changes in Cooperative and Coercive State–NationalRelations as Perceived by State Administrators." Journal of Public AdministrationResearchandTheory14(4):447-468.

Cho, Y. (2013).Why do countries implement Basel II? : an analysis of the globaldiffusionofBaselIIimplementationThesis(Ph.D.),LondonSchoolofEconomicsandPoliticalScience(UniversityofLondon).

Coen,D.(2005)."Business–RegulatoryRelations:LearningtoPlayRegulatoryGamesinEuropeanUtilityMarkets."Governance18(3):375-398.

Coen,D.andM.Thatcher(2008)."NetworkGovernanceandMulti-levelDelegation:EuropeanNetworksofRegulatoryAgencies."JournalofPublicPolicy28(01):49-71.

Collier,R.B.andD.Collier(1991).ShapingthePoliticalArena.CriticalJunctures,theLaborMovement and RegimeDynamics in Latin America. Princeton, New Jersey,PrincetonUniversityPress.

Cranmer, S. J. and B. A. Desmarais (2011). "Inferential Network Analysis withExponentialRandomGraphModels."PoliticalAnalysis19(1):66-86.

Cranmer,S.J.,P.Leifeld,S.D.McClurgandM.Rolfe(2017)."NavigatingtheRangeofStatistical Tools for Inferential Network Analysis." American Journal of PoliticalScience61(1):237-251.

243

Cukierman,A.,S.B.WebandB.Neyapti (1992)."Measuringthe independenceofcentralbanksanditseffectonpolicyoutcomes."TheWorldBankEconomicReview6(3):353-398.

Damro, C. (2015). "Market power Europe: exploring a dynamic conceptualframework."JournalofEuropeanPublicPolicy22(9):1336-1354.

Danielsen, O. A. and K. Yesilkagit (2014). "The Effects of European RegulatoryNetworksontheBureaucraticAutonomyofNationalRegulatoryAuthorities."PublicOrganizationReview14(3):353-371.

Darbouche,H.(2011)."ThirdTimeLucky?Euro-MediterraneanEnergyCo-operationundertheUnionfortheMediterranean."MediterraneanPolitics16(1):193-211.

Davis,C.andK.Hoffer(2012)."Federalizingenergy?Agendachangeandthepoliticsoffracking."PolicySciences45(3):221-241.

deBruijn, J. A. and E. F. tenHeuvelhof (1995). PolicyNetworks andGovernance.InstitutionalDesign.Dordrecht,SpringerNetherlands:161-179.

Dehousse,R.(1997)."RegulationbynetworksintheEuropeanCommunity:theroleofEuropeanagencies."JournalofEuropeanPublicPolicy4(2):246-261.

Deitz, L., L. Stirton and K.Wright (2009). "South East Europe's electricity sector:Attractions,obstaclesandchallengesofEuropeanisation."UtilitiesPolicy17(1):4-12.

Derthick,M.andP. J.Quirk (1985).ThePoliticsofDeregulation.Washington,TheBrookingsInstitution.

Desmarais,B.A.andS.J.Cranmer(2012)."Micro-LevelInterpretationofExponentialRandomGraphModelswithApplicationtoEstuaryNetworks."PolicyStudiesJournal40(3):402-434.

Desmarais,B.A.,J.J.HardenandF.J.Boehmke(2015)."PersistentPolicyPathways:Inferring Diffusion Networks in the American States." American Political ScienceReview109(02):392-406.

Diathesopoulos, M. D. (2010). From Energy Sector Inquiry to Recent AntitrustDecisionsinEuropeanEnergyMarkets:CompetitionLawasaMeanstoImplementEnergy Sector Regulation in EU, Available at SSRN:https://ssrn.com/abstract=1639883orhttp://dx.doi.org/10.2139/ssrn.1639883.

Dietrich,S.(2013)."BypassorEngage?ExplainingDonorDeliveryTacticsinForeignAidAllocation*."InternationalStudiesQuarterly57(4):698-712.

244

Dietrich,S.(2015)."DonorPoliticalEconomiesandthePursuitofAidEffectiveness."InternationalOrganization70(1):65-102.

DiMaggio, P. J. (1988). "Interest and agency in institutional theory." Institutionalpatternsandorganizations:Cultureandenvironment1:3-22.

Djelic,M.-L. (2004). "Social networks and country-to-country transfer: dense andweaktiesinthediffusionofknowledge."Socio-EconomicReview2(3):341-370.

Djelic,M.L.andK.Sahlin-Andersson(2006).TransnationalGovernance:InstitutionalDynamicsofRegulation,CambridgeUniversityPress.

Dobbin, F. (1995). Railroads. Organizations in Industry. Strategy, Structure andSelection.G.R.CarrolandM.T.Hannan.NewYork,OxfordUniversityPress:59-86.

Dollar,D.andV.Levin(2006)."TheIncreasingSelectivityofForeignAid,1984–2003."WorldDevelopment34(12):2034-2046.

Dowd,T.J.andF.Dobbin(2001).Originsofthemythofneoliberalism:regulationinthe first century of US railroading. The State, Regulation and the Economy. AnHistoricalPerspective.L.MagnussonandJ.Ottosson.Cheltenham,UK,EdwardElgar:61-88.

Dowding,K.(1995)."ModelorMetaphor?ACriticalReviewofthePolicyNetworkApproach."PoliticalStudiesXLIII:23.

Dowding,K.(2001)."ThereMustBeEndtoConfusion:PolicyNetworks,IntellectualFatigue,andtheNeedforPoliticalScienceMethodsCoursesinBritishUniversities."PoliticalStudies49(1):89-105.

Dowdle,M.W.,J.GillespieandI.Maher(2013).AsianCapitalismandtheRegulationof Competition: Towards a Regulatory Geography of Global Competition Law,CambridgeUniversityPress.

Downs,A.(1967).Insidebureaucracy,Little,BrownBoston.

Drezner, D. W. (2007). All Politics Is Global: Explaining International RegulatoryRegimes,PrincetonUniversityPress.

Dubash,N.K.andB.Morgan(2012)."UnderstandingtheriseoftheregulatorystateoftheSouth*."Regulation&Governance6(3):261-281.

Dunlop,C.A.(2009)."Policytransferaslearning:capturingvariationinwhatdecision-makerslearnfromepistemiccommunities."Policystudies30(3):289-311.

245

Eastman, J. B. (1919). "The Advantages of National Operation." Annals of theAmericanAcademyofPoliticalandSocialScience86:77-90.

Eberlein,B. (2008)."TheMakingoftheEuropeanEnergyMarket:The InterplayofGovernanceandGovernment."JournalofPublicPolicy28(01):73-92.

Eberlein, B. and E. Grande (2005). "Beyond delegation: transnational regulatoryregimesandtheEUregulatorystate."JournalofEuropeanPublicPolicy12(1):89-112.

Eberlein, B. and A. Newman (2008). "Escaping the International GovernanceDilemma? Incorporated Transgovernmental Networks in the European Union."Governance21(1):25-52.

Egeberg,M.,J.TrondalandN.M.Vestlund(2014)."Thequestfororder:unravellingtherelationshipbetweentheEuropeanCommissionandEuropeanUnionagencies."JournalofEuropeanPublicPolicy22(5):1-21.

Eichenauer,V.Z.andB.Reinsberg(2017)."Whatdeterminesearmarkedfundingtointernationaldevelopmentorganizations?Evidencefromthenewmulti-biaiddata."TheReviewofInternationalOrganizations12(2):171-197.

Eising,R.(2008)"InterestgroupsinEUpolicy-making."LivingReviewsinEuropeanGovernance(LREG)3.

Eising, R. and B. Kohler-Koch (2003). The Transformation of Governance in theEuropeanUnion,Taylor&Francis.

Elkins, Z., A. T. Guzman and B. A. Simmons (2006). "Competing for Capital: TheDiffusionofBilateral Investment Treaties, 1960–2000." InternationalOrganization60(04):811-846.

Escribano, G. (2010). "Convergence towards differentiation: the case ofMediterraneanenergycorridors."MediterraneanPolitics15(2):211-229.

Escribano,G. (2011)."Marketsorgeopolitics?TheEuropeanizationofEU'senergycorridors."InternationalJournalofEnergySectorManagement5(1):39-59.

Escribano,G.andJ.M.Jordán(1999)."Sub�regionalintegrationintheMENAregionand the euro�mediterranean free trade area."Mediterranean Politics4(2): 133-148.

Escribano,G.andE.SanMartínGonzález(2014)."ManagingEnergyInterdependencyintheWesternMediterranean."RevueMaroco-EspagnoledeDroitInternationaletRelationsInternationales2:81-102.

246

Essex,J.(2013).Development,Security,andAid:GeopoliticsandGeoeconomicsattheU.S.AgencyforInternationalDevelopment,UniversityofGeorgiaPress.

European Commission (1996). Communication from the Commission to theEuropean Parliament and the Council Concerning the Euro- MediterraneanPartnershipintheEnergySector.COM(96)0149-C4-0238/96.Brussels.

Fainsod,M.andL.Gordon(1948).GovernmentandtheAmericaneconomy,W.W.Norton&company,inc.

Falkner,G.(2000)."Policynetworksinamulti�levelsystem:Convergencetowardsmoderatediversity?"WestEuropeanPolitics23(4):94-120.

Farrell, H. and A. Newman (2015). "The New Politics of Interdependence: Cross-National Layering in Trans-Atlantic Regulatory Disputes." Comparative PoliticalStudies48(4):497-526.

Farrell, H. and A. L. Newman (2010). "Making global markets: Historicalinstitutionalismininternationalpoliticaleconomy."ReviewofInternationalPoliticalEconomy17(4):609-638.

Farrell,H.andA.L.Newman(2014)."DomesticInstitutionsbeyondtheNation-State:ChartingtheNewInterdependenceApproach."WorldPolitics66(02):331-363.

Fernández-i-Marín, X. and J. Jordana (2015). "The emergence of regulatoryregionalism:transnationalnetworksandthediffusionofregulatoryagencieswithinregions."ContemporaryPolitics21(4):417-434.

Findley,M.G.,H.V.MilnerandD.L.Nielson(2017)."Thechoiceamongaiddonors:The effects of multilateral vs. bilateral aid on recipient behavioral support." TheReviewofInternationalOrganizations12(2):307-334.

Finger, M., J. Groenewegen and R. Künneke (2005). "The Quest for Coherencebetween Institutions and Technologies in Infrastructures." Journal of NetworkIndustriesos-6(4):227-259.

Fischer, M., K. Ingold and S. Ivanova (2017). "Information exchange underuncertainty:ThecaseofunconventionalgasdevelopmentintheUnitedKingdom."LandUsePolicy67(SupplementC):200-211.

Fischer, M. and P. Sciarini (2015). "Unpacking reputational power: Intended andunintendeddeterminantsoftheassessmentofactors’power."SocialNetworks42:60-71.

247

Freeman,J.L.andJ.P.Stevens(1987)."ATheoreticalandConceptualReexaminationofSubsystemPolitics."PublicPolicyandAdministration2(1):9-24.

Freyburg, T. (2015). "Transgovernmental Networks as an Apprenticeship inDemocracy? Socialization into Democratic Governance through Cross-nationalActivities."InternationalStudiesQuarterly59(1):59-72.

Gerber,B.J.andP.Teske(2000)."RegulatoryPolicymakingintheAmericanStates:AReviewofTheoriesandEvidence."PoliticalResearchQuarterly53(4):849-886.

Gilardi, F. (2002). "Policy credibility and delegation to independent regulatoryagencies:acomparativeempiricalanalysis."JournalofEuropeanPublicPolicy9(6):873-893.

Gilardi,F.(2007)."Thesame,butdifferent:Centralbanks,regulatoryagencies,andthepoliticsofdelegationtoindependentauthorities."ComparativeEuropeanPolitics5(3):303-327.

Gilbert,R.J.andE.P.Kahn(1996).Internationalcomparisonsofelectricityregulation,CambridgeUniversityPress.

Glachant,J.M.,H.Khalfallah,Y.Perez,V.RiousandM.Saguan(2013)."ImplementingIncentiveRegulationandRegulatoryAlignmentwithResourceBoundedRegulators."CompetitionandRegulationinNetworkIndustries(CRNI)14(3):265-291.

Goetz,K.H.(2010).TheTemporalDimension.WhichEurope?,Springer:67-80.

Goldbach,R.(2015)."Asymmetricinfluenceinglobalbankingregulation."ReviewofInternationalPoliticalEconomy22(6):1087-1127.

Goldin,C.andG.D.Libecap(2008).TheRegulatedEconomy:AHistoricalApproachtoPoliticalEconomy,UniversityofChicagoPress.

Goldthau,A.andN.Sitter(2014)."Aliberalactorinarealistworld?TheCommissionandtheexternaldimensionof thesinglemarket forenergy." JournalofEuropeanPublicPolicy21(10):1452-1472.

Gormley Jr, W. T. (1983). The politics of public utility regulation, University ofPittsburghPress.

Gould, R. V. (2003). Uses of Network Tools in Comparative Historical Research.Comparative Historical Research in the Social Sciences. J. Mahoney and D.Rueschemeyer.Cambridge,CambridgeUniversityPress.

248

Granovetter, M. S. (1973). "The Strength of Weak Ties." American Journal ofSociology78(6):1360-1380.

Guardiancich,I.andM.Guidi(2016)."FormalindependenceofregulatoryagenciesandVarietiesofCapitalism:Acaseofinstitutionalcomplementarity?"Regulation&Governance10(3):211-229.

Guerriero,C.(2010).ThePoliticalEconomyof(De)Regulation:TheoryandEvidencefromtheUSElectricityMarket.WorkingPaperSeries.Milan,IEFE.

Haas, E. B. (1975). "Is there a hole in the whole? Knowledge, technology,interdependence, and the construction of international regimes." InternationalOrganization29(03):827-876.

Haas,E.B.(1975)."OnSystemsandInternationalRegimes."WorldPolitics27(02):147-174.

Haas, E. B. (1980). "Why Collaborate? Issue-Linkage and International Regimes."WorldPolitics32(03):357-405.

Haas,E.B.,M.P.WilliamsandD.Babai(1977).ScientistsandWorldOrder:TheUsesofTechnicalKnowledgeinInternationalOrganizations,UniversityofCaliforniaPress.

Haas,P.M. (1992). "Introduction:epistemiccommunitiesand internationalpolicycoordination."InternationalOrganization46(01):1-35.

Hall, P. A. (1993). "Policy Paradigms, Social Learning, and the State: The Case ofEconomicPolicymakinginBritain."ComparativePolitics25(3):275-296.

Hall, P. A. and D. W. Gingerich (2009). "Varieties of Capitalism and InstitutionalComplementaritiesinthePoliticalEconomy:AnEmpiricalAnalysis."BritishJournalofPoliticalScience39(03):449-482.

Hall,P.A.andD.Soskice(2001).Varietiesofcapitalism:Theinstitutionalfoundationsofcomparativeadvantage,OUPOxford.

Hamilton,M.andM. Lubell (2017). "CollaborativeGovernanceofClimateChangeAdaptation Across Spatial and Institutional Scales." Policy Studies Journal:10.1111/psj.12224.

Hancher,L.(1996).TheregulatoryroleoftheEuropeanUnion.TheEuropeanUnionandNationalIndustrialPolicy.H.KassimandA.Menon.London,Routledge.

Hancher, L. and M. Moran (1989). Organizing regulatory space. A reader onregulation.R.Baldwin,C.ScottandC.Hood,OxfordUniversityPress.

249

Hancké, B., M. Rhodes and M. Thatcher (2007). Beyond Varieties of Capitalism:Conflict,Contradictions,andComplementarities intheEuropeanEconomy,OxfordUniversityPress.

Handcock,M.,D.Hunter,C.T.Butts,S.Goodreau,P.KrivitskyandM.Morris(2017).ergm: Fit, Simulate and Diagnose Exponential-Family Models for Networks: Rpackage.

Hanretty,C.(2010)."ExplainingtheDeFactoindependenceofPublicBroadcasters."BritishJournalofPoliticalScience40(1):75-89.

Hausman,W.J.andJ.L.Neufeld(2011)."Howpolitics,economics,andinstitutionsshapedelectricutilityregulationintheUnitedStates:1879–2009."BusinessHistory53(5):723-746.

Heclo, H. (1978). Issue Networks and the Executive Establishment: GovernmentGrowth in an Age of Improvement. The New American Political System. A. Kind.Washington,DC,AmericanEnterpriseInstitute.

Heclo, H. and A. Wildavsky (1974). The private government of public money:communityandpolicyinsideBritishpoliticaladministration,London:Macmillan.

Hedge,D.M.,M.J.ScicchitanoandP.Metz(1991)."ThePrincipal-AgentModelandRegulatoryFederalism."TheWesternPoliticalQuarterly44(4):1055-1080.

Helleiner, E. and T. Porter (2010). "Making transnational networks moreaccountable."Economics,ManagementandFinancialMarkets5(2):158.

Helm,D.(2009)."Infrastructureinvestment,thecostofcapital,andregulation:anassessment."OxfordReviewofEconomicPolicy25(3):307-326.

Henisz,W. J., B. A. Zelner andM. F. Guillén (2005). "TheWorldwideDiffusion ofMarket-OrientedInfrastructureReform,1977–1999."AmericanSociologicalReview70(6):871-897.

Henriksen, L. F. and S. Ponte (2018). "Public orchestration, social networks, andtransnational environmental governance: Lessons from the aviation industry."Regulation&Governance12(1):23-45.

Herweg,N.(2015).AgainstAllOdds:TheLiberalisationoftheEuropeanNaturalGasMarket—AMultipleStreamsPerspective.EnergyPolicyMakingintheEU.J.Tosun,S.BiesenbenderandK.Schulze,SpringerLondon.28:87-105.

250

Heugens,P.P.M.A.R.andM.W.Lander (2009)."Structure!Agency! (AndOtherQuarrels):AMeta-AnalysisOf InstitutionalTheoriesOfOrganization."AcademyofManagementJournal52(1):61-85.

Himmelberg, R. F. (1994). Growth of the Regulatory State, 1900-1917: State andFederalRegulationofRailroadsandOtherEnterprises,GarlandPub.

Hooghe,L.andG.Marks(2001).Multi-levelGovernanceandEuropeanIntegration,Rowman&LittlefieldPublishers.

Hooghe,L.andG.Marks(2003)."UnravelingtheCentralState,butHow?TypesofMulti-LevelGovernance."TheAmericanPoliticalScienceReview97(2):233-243.

Hooper,E.andA.Medvedev(2009)."Electrifyingintegration:ElectricityproductionandtheSouthEastEuroperegionalenergymarket."UtilitiesPolicy17(1):24-33.

Horrall, J. (2008). Influencing Infrastructure Performance through Cross-BorderNetworksofRegulatoryAgencies.TelecommunicationsResearchTrends:91-102.

Humphreys,P.andS.Simpson(2008)."Globalization,the‘Competition’StateandtheRiseofthe‘Regulatory’StateinEuropeanTelecommunications*."JCMS:JournalofCommonMarketStudies46(4):849-874.

Ikenberry, G. J. and C. A. Kupchan (1990). "Socialization and hegemonic power."InternationalOrganization44(03):283-315.

Ingold, K. and P. Leifeld (2014). "Structural and Institutional Determinants ofInfluence Reputation: A Comparison of Collaborative and Adversarial PolicyNetworksinDecisionMakingandImplementation."JournalofPublicAdministrationResearchandTheory.

Ingold,K.,F.VaroneandF.Stokman(2013)."Asocialnetwork-basedapproachtoassessde facto independenceof regulatoryagencies." JournalofEuropeanPublicPolicy20(10):1464-1481.

International Competition Network (2003). Capacity Building and TechnicalAssistance. Building credible competition authorities in developing and transitioneconomies.,InternationalCompetitionNetwork.

Isett, K. R., I. A.Mergel, K. LeRoux, P. A.Mischen and R. K. Rethemeyer (2011)."NetworksinPublicAdministrationScholarship:UnderstandingWhereWeAreandWhere We Need to Go." Journal of Public Administration Research and Theory21(suppl1):i157-i173.

251

Jamasb,T.andM.Pollitt(2005)."ElectricityMarketReformintheEuropeanUnion:Review of Progress toward Liberalization & Integration." The Energy Journal26(EuropeanEnergyLiberalisationSpecialIssue):11-41.

Jevnaker, T. (2015). "Pushing administrative EU integration: the path towardsEuropeannetworkcodesforelectricity."JournalofEuropeanPublicPolicy22(7):1-21.

Joint Communication (2011a). Joint Communication of the European CommissionandtheHighRepresentativeoftheEuropeanUniontotheEuropeanParliament,theCouncil, theEuropeanEconomicandSocialCommitteeandtheCommitteeof theRegions. A Partnership for Democracy and Shared Prosperity with the SouthernMediterraneanCountries.COM(2011)200final.Brussels.

Joint Communication (2011b). Joint Communication of the EuropeanCommissionandtheHighRepresentativeoftheEuropeanUniontotheEuropeanParliament,theCouncil, theEuropeanEconomicandSocialCommitteeandtheCommitteeof theRegions.AnewresponsetoachangingNeighbourhood.AReviewoftheEuropeanNeighbouhoodPolicy.COM(2011)303final.Brussels.

JointPaper(2006).AnExternalPolicytoServeEurope'sEnergyInterests.JointPaperby the Commission and the Secretary General/High Representative. 9971/06.Brussels.

Jordana,J.(2011).TheInstitutionalizationofRegionalRegulatoryNetworks:EuropeandLatinAmericanCompared.SASEAnnualConference.Madrid.

Jordana,J.(2017)."TransgovernmentalNetworksasRegulatoryIntermediaries."TheANNALSoftheAmericanAcademyofPoliticalandSocialScience670(1):245-262.

Jordana, J. and D. Levi�Faur (2012). "Regional Integration and TransnationalRegulatoryRegimes:ThePolycentricArchitectureofGovernanceinLatinAmericanTelecommunications."

Judge, A. and T. Maltby (2017). "European Energy Union? Caught betweensecuritisation and ‘riskification’." European Journal of International Security 2(2):179-202.

Kaiser,W.(2009)."BringingHistoryBackInTotheStudyofTransnationalNetworksinEuropeanIntegration."JournalofPublicPolicy29(SpecialIssue02):223-239.

Kalaitzake,M.(2017)."DeathbyaThousandCuts?FinancialPoliticalPowerandtheCaseoftheEuropeanFinancialTransactionTax."NewPoliticalEconomy:1-18.

252

Kanazawa,M.T.andR.G.Noll(1994).Theoriginsofstaterailroadregulation:theIllinois constitution of 1870. The regulated economy: A historical approach topoliticaleconomy,UniversityofChicagoPress:13-54.

Kapstein,E.B.(1992)."Betweenpowerandpurpose:centralbankersandthepoliticsofregulatoryconvergence."InternationalOrganization46(01):265-287.

Karova,R.(2009).EnergyCommunityforSouthEastEurope:RationaleBehindandImplementation to Date. EUI Working Papers. E. Robert Schuman Centre forAdvancedStudies.Florence,RSCAS.

Keegan,S.a.,R.d.s.Craufurd-Smith,N.d.s.NicShuibhneandA.d.s.Scott(2013).Legitimacy in theEUsinglemarket : the roleofnormative regulatorygovernanceThesis(Ph.D.),UniversityofEdinburgh.

Kelemen,R.D.andA.D.Tarrant(2011)."ThePoliticalFoundationsoftheEurocracy."WestEuropeanPolitics34(5):922-947.

Kenis,P.andK.G.Provan(2009)."Towardsanexogenoustheoryofpublicnetworkperformance."PublicAdministration87(3):440-456.

Kenis,P.andV.Schneider(1991).Policynetworksandpolicyanalysis:scrutinizinganew analytical toolbox. Policy networks: Empirical evidence and theoreticalconsiderations,CampusVerlag:25-59.

Keohane, R. O. (1982). "The demand for international regimes." InternationalOrganization36(02):325-355.

Keohane,R.O. (1988). "International Institutions:TwoApproaches." InternationalStudiesQuarterly32(4):379-396.

Keohane, R. O. (1998). International Institutions: Can Interdependence Work?ForeignPolicy:82-96.

Keohane,R.O.andJ.S.Nye(1974)."TransgovernmentalRelationsandInternationalOrganizations."WorldPolitics27(1):25.

Kerr, K. A. (1968). American Railroad Politics, 1914–1920: Rates, Wages andEfficiency,UniversityofPittsburghPress.

Khalfallah, H. (2015). "Connecting Mediterranean countries through electricitycorridors:NewInstitutionalEconomicandregulatoryanalysis."UtilitiesPolicy32:45-54.

253

Klein, M. U. and T. Harford (2005). The Market for Aid, International FinanceCorporation.

Knoke,D.(1990)."NetworksofPoliticalAction:TowardTheoryConstruction."SocialForces68(4):1041-1063.

Knoke,D.(1996).Comparingpolicynetworks:laborpoliticsintheUS,Germany,andJapan,CambridgeUniversityPress.

Kohler-Koch,B.andB.Rittberger(2006)."ReviewArticle:The‘GovernanceTurn’inEUStudies*."JCMS:JournalofCommonMarketStudies44:27-49.

Kolko, G. (1965). Railroads and Regulation, 1877 - 1916. Princeton, New Jersey,PrincetonUniversityPress.

König,T.andT.Bräuninger(1998)."TheFormationofPolicyNetworks:Preferences,InstitutionsandActors'Choiceof InformationandExchangeRelations."JournalofTheoreticalPolitics10(4):445-471.

Koop, C. andC.Hanretty (2017). "Political Independence, Accountability, and theQuality of Regulatory Decision-Making." Comparative Political Studies:0010414017695329.

Koske,I.,I.Wanner,R.BitettiandO.Barbiero(2015).The2013updateoftheOECDproduct market regulation indicators: policy insights for OECD and non-OECDcountries.OECDEconomicsDepartmentWorkingPapers,OECD.

Krasner,S.D. (1981)."Transforming InternationalRegimes:What theThirdWorldWantsandWhy."InternationalStudiesQuarterly25(1):119-148.

Krasner, S. D. (1982). "Structural causes and regime consequences: regimes asinterveningvariables."InternationalOrganization36(02):185-205.

Krishnaswamy,V.andG.Stuggins(2003).PrivateSectorParticipationinthePowerSector inEuropeandCentralAsia:LessonsfromtheLastDecade.WashingtonDC,WorldBank.

Künneke,R.,J.GroenewegenandC.Ménard(2010)."Aligningmodesoforganizationwith technology: Critical transactions in the reformof infrastructures." Journal ofEconomicBehavior&Organization75(3):494-505.

Künneke,R.W.(2008)."Institutionalreformandtechnologicalpractice:thecaseofelectricity."IndustrialandCorporateChange17(2):233-265.

254

LaBelle, M. C. (2012). Diffusion of Regulatory Knowledge: A case study oftransnationalregulatorynetworks.C.E.U.B.School.Budapest.

Lagendijk,V.(2008).ElectrifyingEurope:ThePowerofEuropeintheConstructionofElectricityNetworks,Aksant.

Larner, W. and N. Laurie (2010). "Travelling technocrats, embodied knowledges:Globalisingprivatisationintelecomsandwater."Geoforum41(2):218-226.

Lavenex,S.(2004)."EUexternalgovernancein'widerEurope'."JournalofEuropeanPublicPolicy11(4):680-700.

Lavenex,S.andF.Schimmelfennig(2009)."EUrulesbeyondEUborders:theorizingexternalgovernanceinEuropeanpolitics."JournalofEuropeanPublicPolicy16(6):791-812.

Lavenex, S. and N. Wichmann (2008). "The External Governance of EU InternalSecurity."JournalofEuropeanIntegration31(1):83-102.

Lavrijssen,S.andL.Hancher(2009)."Networksontrack:fromEuropeanregulatorynetworks to European regulatory network agencies." Legal Issues of EconomicIntegration36(1):23-55.

Law,M.T.andS.Kim(2011).TheriseoftheAmericanregulatorystate:aviewfromtheProgressive Era.Handbookon thepolitics of regulation.D. Levi-Faur, EdwardElgar.

Law, M. T. and C. X. Long (2011). "Revolving door laws and state public utilitycommissioners."Regulation&Governance5(4):405-424.

Lawton, R. W. and B. Burns (1996). "Models of Cooperative Federalism forTelecommunications."Alb.LJSci.&Tech.6:71.

Lazega,E.,E.QuintaneandS.Casenaz(2017)."Collegialoligarchyandnetworksofnormativealignmentsintransnationalinstitutionbuilding."SocialNetworks48:10-22.

Lazer, D. (2001). "Regulatory interdependence and international governance."JournalofEuropeanPublicPolicy8(3):474-492.

Leaver, C. (2009). "BureaucraticMinimal Squawk Behavior: Theory and EvidencefromRegulatoryAgencies."AmericanEconomicReview99(3):572-607.

255

Lee,Y.,I.W.LeeandR.C.Feiock(2012)."InterorganizationalCollaborationNetworksinEconomicDevelopmentPolicy:AnExponentialRandomGraphModelAnalysis*."PolicyStudiesJournal40(3):547-573.

Leifeld,P.(2013)."texreg:ConversionofStatisticalModelOutputinRtoLATEXandHTMLTables."201355(8):24.

Levi-Faur,D.(2011)."Regulatorynetworksandregulatoryagencification:towardsaSingleEuropeanRegulatorySpace."JournalofEuropeanPublicPolicy18(6):810-829.

Levy, B. and P. T. Spiller (1994). "The Institutional Foundations of RegulatoryCommitment:AComparativeAnalysisofTelecommunicationsRegulation."JournalofLaw,Economics&Organization10(2):46.

Lodge, M. (2002). From Varieties of the Welfare State to Convergence of theRegulatory State? The'Europeanisation'of Regulatory Transparency, QueensUniversityBelfast.

Lodge,M. (2008). "Regulation, theRegulatory StateandEuropeanPolitics."WestEuropeanPolitics31(1-2):280-301.

Lodge,M.(2010)."RegulatingInfrastructures:Thelimitsoftheregulatorystate."dermodernestaat–ZeitschriftfürPublicPolicy,RechtundManagement3(1).

Lodge, M. and K. Wegrich (2014). Introduction: Governance Innovation,AdministrativeCapacities,andPolicyInstruments.TheProblem-solvingCapacityoftheModernState.Oxford,OxfordUniversityPress.

Maggetti,M.(2007)."Defactoindependenceafterdelegation:Afuzzy-setanalysis."Regulation&Governance1(4):271-294.

Maggetti,M.(2009)."Theroleofindependentregulatoryagenciesinpolicy-making:acomparativeanalysis."JournalofEuropeanPublicPolicy16(3):450-470.

Maggetti, M. (2010). "Legitimacy and Accountability of Independent RegulatoryAgencies:ACriticalReview."LivingReviewsinDemocracy2:1-9.

Maggetti,M. (2013). "The rewards of cooperation: The effects ofmembership inEuropeanregulatorynetworks."EuropeanJournalofPoliticalResearch53(3):480-499.

Maggetti,M. (2014). "ThePoliticsofNetworkGovernance inEurope:TheCaseofEnergyRegulation."WestEuropeanPolitics37(3):497-514.

256

Maggetti, M. and F. Gilardi (2011). "The policy-making structure of Europeanregulatorynetworksandthedomesticadoptionofstandards."JournalofEuropeanPublicPolicy18(6):830-847.

Maggetti,M.andF.Gilardi(2014)."Networkgovernanceandthedomesticadoptionofsoftrules."JournalofEuropeanPublicPolicy21(9):1293-1310.

Maggetti,M.,K.IngoldandF.Varone(2013)."HavingYourCakeandEatingIt,Too:Can Regulatory Agencies Be Both Independent and Accountable?" Swiss PoliticalScienceReview19(1):1-25.

Maggetti, M. and K. Verhoest (2014). "Unexplored aspects of bureaucraticautonomy: a state of the field and ways forward." International Review ofAdministrativeSciences80(2):239-256.

Magone, J.M.,B.LaffanandC.Schweiger (2016).Core-peripheryRelations in theEuropeanUnion:PowerandConflictinaDualistPoliticalEconomy,Taylor&Francis.

Mahoney, J. and D. Rueschemeyer (2003). Comparative Historical Analysis in theSocialSciences,CambridgeUniversityPress.

Maitland,C.F.andA.vanGorp(2009)."BeyondHarmonization:ICTPolicymakinginRegionalEconomicCommunities."TheInformationSociety25(1):23-37.

Majone, G. (1994). "The rise of the regulatory state in Europe."West EuropeanPolitics17(3):77-101.

Majone, G. (1997). "From the Positive to the Regulatory State: Causes andConsequences of Changes in the Mode of Governance." Journal of Public Policy17(02):139-167.

Majone,G.(1997)."ThenewEuropeanagencies:regulationbyinformation."JournalofEuropeanPublicPolicy4(2):262-275.

Majone, G. (1999). "The regulatory state and its legitimacy problems." WestEuropeanPolitics22(1):1-24.

Maltby,T. (2013)."EuropeanUnionenergypolicy integration:AcaseofEuropeanCommission policy entrepreneurship and increasing supranationalism." EnergyPolicy55:435-444.

Manners,I.(2015)."SociologyofKnowledgeandProductionofNormativePowerintheEuropeanUnion’sExternalActions."JournalofEuropeanIntegration37(2):299.

257

Maoz, Z. (2012). "Preferential Attachment, Homophily, and the Structure ofInternationalNetworks,1816–2003."ConflictManagementandPeaceScience29(3):341-369.

Marcoux,C.andJ.Urpelainen(2012)."Capacity,notconstraints:AtheoryofNorth-Southregulatorycooperation."TheReviewofInternationalOrganizations7(4):399-424.

Marks,G., L.HoogheandK.Blank (1996). "European Integration from the1980s:State-Centricv.Multi-levelGovernance*."JCMS:JournalofCommonMarketStudies34(3):341-378.

Marks,G.,F.W.Scharpf,P.C.SchmitterandW.Streeck(1996).GovernanceintheEuropeanUnion,Sage.

Marsh, D. and R. A. W. Rhodes (1992). Policy networks in British government,ClarendonPress.

Marsh, D. and M. Smith (2000). "Understanding Policy Networks: towards aDialecticalApproach."PoliticalStudies48(1):4-21.

Mastenbroek, E. and D. S. Martinsen (2018). "Filling the gap in the Europeanadministrativespace:theroleofadministrativenetworksinEUimplementationandenforcement."JournalofEuropeanPublicPolicy25(3):422-435.

Mathieu, E. (2016). "When Europeanization Feeds Back into EU Governance: EULegislation, National Regulatory Agencies, and EU Regulatory Networks." PublicAdministration94(1):25-39.

Mattli,W.(1999).TheLogicofRegionalIntegration:EuropeandBeyond,CambridgeUniversityPress.

Mayer,S.(2008)."PathdependenceandCommissionactivismintheevolutionoftheEuropeanUnion'sexternalenergypolicy."JIntRelatDev11(3):251-278.

McGuire,M. (2006). "IntergovernmentalManagement:AViewfromtheBottom."PublicAdministrationReview66(5):677-679.

McNamara,K.(2002)."RationalFictions:CentralBankIndependenceandtheSocialLogicofDelegation."WestEuropeanPolitics25(1):47-76.

McPherson,M.,L.Smith-LovinandJ.M.Cook(2001)."Birdsofafeather:Homophilyinsocialnetworks."Annualreviewofsociology27(1):415-444.

258

Meseguer,C.(2004)."WhatRoleforLearning?TheDiffusionofPrivatisationinOECDandLatinAmericanCountries."JournalofPublicPolicy24(03):299-325.

Metre,T.W.v.(1918)."FailuresandPossibilitiesinRailroadRegulation."AnnalsoftheAmericanAcademyofPoliticalandSocialScience76:1-13.

Meyer,J.W.andB.Rowan(1977)."InstitutionalizedOrganizations:FormalStructureasMythandCeremony."AmericanJournalofSociology83(2):340-363.

Miller,E.S.andW.J.Samuels(2002).TheInstitutionalistApproachtoPublicUtilitiesRegulation,MichiganStateUniversityPress.

Milner,H. V. andD.H. Tingley (2010). "The Political EconomyofUS ForeignAid:AmericanLegislatorsandtheDomesticPoliticsofAid"Economics&Politics22(2):200-232.

Mintrom,M.andS.Vergari(1998)."PolicyNetworksandInnovationDiffusion:TheCaseofStateEducationReforms."TheJournalofPolitics60(01):126-148.

Monti, G. (2014). Independence, Interdepedence and Legitimacy: The EUCommission, National Competition Authorities and the European CompetitionNetwork.EUIWorkingPapersLaw.E.U.Institute.

Moran,M.(2002)."Understandingtheregulatorystate."BritishJournalofPoliticalScience32(2):391-413.

Murillo,M. V. and C.Martínez-Gallardo (2007). "Political Competition and PolicyAdoption:MarketReforms inLatinAmericanPublicUtilities."AmericanJournalofPoliticalScience51(1):120-139.

Natsios, A. S. (2006). "Five Debates on International Development: The USPerspective."DevelopmentPolicyReview24(2):131-139.

Neumayer,E.(2002)."IsGoodGovernanceRewarded?ACross-nationalAnalysisofDebtForgiveness."WorldDevelopment30(6):913-930.

Neumayer, E. (2003). "Do Human Rights Matter in Bilateral Aid Allocation? AQuantitativeAnalysisof21DonorCountries*."SocialScienceQuarterly84(3):650-666.

Neumayer,E.(2005).ThePatternofAidGiving:TheImpactofGoodGovernanceonDevelopmentAssistance,Taylor&Francis.

259

Newman,A.andE.Posner(2016)."Transnationalfeedback,softlaw,andpreferencesinglobalfinancialregulation."ReviewofInternationalPoliticalEconomy23(1):123-152.

Newman, A. L. (2008). "Building Transnational Civil Liberties: TransgovernmentalEntrepreneursandtheEuropeanDataPrivacyDirective."InternationalOrganization62(01):103-130.

Newman,A.L.andD.Zaring (2013).RegulatoryNetworks:Power,LegitimacyandCompliance. Interdisciplinary Perspectives on International Law and InternationalRelations.TheStateoftheArt.J.L.DunoffandM.A.Pollack.NewYork,CambridgeUniversityPress.

Nice,D.C. (1987)."Stateregulationofrailroads."TransportationResearchPartA:General21(6):411-420.

Nicolaides,P.(2004)."ThePoliticalEconomyofMulti-tieredRegulationinEurope."JCMS:JournalofCommonMarketStudies42(3):599-618.

Nölke, A. and A. Vliegenthart (2009). "Enlarging the varieties of capitalism: TheemergenceofdependentmarketeconomiesinEastCentralEurope."Worldpolitics61(04):670-702.

O'Toole Jr,L. J. (1997). "Treatingnetworksseriously:Practicalandresearch-basedagendasinpublicadministration."Publicadministrationreview:45-52.

Ogus, A. (2002). "Regulatory Institutions And Structures." Annals of Public andCooperativeEconomics73(4):627-648.

OlmosCamacho,L.andI.J.Pérez-Arriaga(2007)."Comparisonofseveralinter-TSOcompensation methods in the context of the internal electricity market of theEuropeanUnion."EnergyPolicy35(4):2379-2389.

Osborne, D. and T. Gaebler (1993). Reinventing Government: How theEntrepreneurialSpiritisTransformingthePublicSector,Plume.

Padgett,J.F.andC.K.Ansell(1993)."RobustActionandtheRiseoftheMedici,1400-1434."AmericanJournalofSociology98(6):1259-1319.

Padgett, S. (2012). "Multilateral institutions, accession conditionality and ruletransfer in the European Union: the Energy Community in South East Europe."JournalofPublicPolicy32(03):261-282.

Papadopoulos, Y. (2005). "Taking stock of multi-level governance networks."EuropeanPoliticalScience4(3):316-327.

260

Papadopoulos, Y. (2007). "Problems of democratic accountability in network andmultilevelgovernance."Europeanlawjournal13(4):469-486.

Papadopoulos,Y.(2017)."Howdoesknowledgecirculateinaregulatorynetwork?Observing a European Platformof Regulatory Authoritiesmeeting." Regulation&Governance:10.1111/rego.12171.

Peltzman,S.(1971)."Pricinginpublicandprivateenterprises:electricutilitiesintheUnitedStates."JL&Econ.14:109.

Peltzman,S.(1976)."TowardaMoreGeneralTheoryofRegulation."JournalofLawandEconomics19(2):211-240.

Pérez-Arriaga, I. J. (2014). Regulation of the power sector, Springer Science &BusinessMedia.

Pfeffer,J.andG.R.Salancik(1978).TheExternalControlofOrganizations.AResourceDependencePerspective.NewYork,Harper&Row.

Piattoni,S.(2010).TheTheoryofMulti-levelGovernance:Conceptual,Empirical,andNormativeChallenges.Oxford,OxfordUniversityPress.

PierceJr,R.J.(1984)."Regulation,Deregulation,Federalism,andAdministrativeLaw:AgencyPowertoPreemptStateRegulation."U.Pitt.L.Rev.46:607.

Pierson, P. (2003). Big, Slow-Moving and... Invisible:Macrosocial Processes in theStudyofComparativePolitics.ComparativeHistoricalAnalysisintheSocialSciences.J.MahoneyandD.Rueschemeyer.Cambridge,CambridgeUniversityPress.

Pollitt,M. (2009). "Evaluating the evidence on electricity reform: Lessons for theSouthEastEurope(SEE)market."UtilitiesPolicy17(1):13-23.

Pollitt, M. G. and J. Stern (2009). Human Resource Constraints for ElectricityRegulation inDevelopingCountries:HasAnythingChanged?EPRGWorkingPaperEPRG0910-CambridgeWorkingPaperinEconomicsCWPE0914.Cambridge.

Powell, W. (1995). Trust-Based Forms of Governance. Trust in organizations:Frontiersoftheoryandresearch.R.M.KramerandT.R.Tyler.USA,SagePublicaions:51-68.

Powell, W. W. and P. DiMaggio (1983). "The iron cage revisited: Institutionalisomorphism and collective rationality in organizational fields." AmericanSociologicalReview48(2):147-160.

261

Prosser, T. (2010). The regulatory enterprise: government, regulation, andlegitimacy,OxfordUniversityPress.

Provan, K. G. and P. Kenis (2008). "Modes of network governance: Structure,management, and effectiveness." Journal of Public Administration Research andTheory18(2):229-252.

Raab,J.andP.Kenis(2006)."14TakingStockofPolicyNetworks:DoTheyMatter?".

Radaelli, C. M. and A. Meuwese (2012). "How the regulatory state differs. Theconstitutional dimensions of rulemaking in the European Union and the UnitedStates."Rivistaitalianadiscienzapolitica42(2):177-196.

Raustiala, K. (1997). "Domestic Institutions and International RegulatoryCooperation: Comparative Responses to the Convention on Biological Diversity."WorldPolitics49(04):482-509.

Raustiala,K.(2002)."ArchitectureofInternationalCooperation:TransgovernmentalNetworksandtheFutureofInternationalLaw,The."VirginiaJournalofInternationalLaw43(1):1-92.

Reinsberg, B. (2015). "Foreign Aid Responses to Political Liberalization." WorldDevelopment75:46-61.

Reisenbichler,A.(2015)."Thedomesticsourcesandpowerdynamicsofregulatorynetworks: evidence from the financial stability forum." Review of InternationalPoliticalEconomy22(5):996-1024.

Renner,S.(2009)."TheEnergyCommunityofSoutheastEurope:Aneo-functionalistprojectofregionalintegration."EuropeanIntegrationonlinePapers(EIoP).

Richardson, J. J. andA.G. Jordan (1979).Governingunderpressure.Oxford (UK),MartinRobertson.

Ripley,R.B.andG.A.Franklin(1980).Congress,theBureaucracy,andPublicPolicy,DorseyPress.

Robins,G.,J.M.LewisandP.Wang(2012)."StatisticalNetworkAnalysisforAnalyzingPolicyNetworks."PolicyStudiesJournal40(3):375-401.

Rodgers,P.(1979).TheNARUCwasthere:ahistoryoftheNationalAssociationofRegulatoryUtilityCommissioners,TheAssociation.

Rodine Hardy, K. (2008). Global Regulatory Reform in Telecommunications - theImportanceofInternationalOrganizations.ECPRRegulation&Governance

262

SecondBiennialConferenceofthe

StandingGrouponRegulationandGovernance.Utrecht,TheNetherlands.

Rodine-Hardy, K. (2015). "Globalization, International Organizations, andTelecommunications."ReviewofPolicyResearch32(5):517-537.

Ruffing, E. (2014). "How to Become an IndependentAgency: The Creation of theGermanFederalNetworkAgency."GermanPolitics23(1-2):43-58.

Ruffing, E. (2015). "Agencies between two worlds: information asymmetry inmultilevelpolicy-making."JournalofEuropeanPublicPolicy22(8):1109-1126.

Sabatier,P.A.(1991)."TowardBetterTheoriesofthePolicyProcess."PS:PoliticalScience&Politics24(02):147-156.

Sabel,C.F.andJ.Zeitlin(2008)."LearningfromDifference:TheNewArchitectureofExperimentalistGovernanceintheEU."EuropeanLawJournal14(3):271-327.

Sabel,C.F.andJ.Zeitlin(2012)."ExperimentalismintheEU:Commongroundandpersistentdifferences."Regulation&Governance6(3):410-426.

Saz-Carranza,A.,S.SalvadorIborraandA.Albareda(2016)."ThePowerDynamicsofMandated Network Administrative Organizations." Public Administration Review76(3):449-462.

Scharpf, F. W. (1997). "Introduction: the problem-solving capacity of multi-levelgovernance."JournalofEuropeanPublicPolicy4(4):520-538.

Schimmelfennig, F. (2012). "Europeanization beyond Europe." Living Reviews inEuropeanGovernance7(1).

Schmidt, S. K. (1998). "Commission activism: subsuming telecommunications andelectricityunderEuropeancompetitionlaw."JournalofEuropeanPublicPolicy5(1):169-184.

Schneider,V.(1992)."Thestructureofpolicynetworks."EuropeanJournalofPoliticalResearch21(1-2):109-129.

Schneider, V. (2001). "Institutional reform in telecommunications: The EuropeanUnionintransnationalpolicydiffusion."TransformingEurope.EuropeanizationandDomesticChange:60-78.

263

Schrefler, L. (2012). Expertise and Independent Regulatory Agencies: EmpiricalevidencefromtheUKOfficeofCommunications.4thBiennialECPRStandingGrouponRegulatoryGovernance.Exeter(UK).

Schulze,K.(2015).ExportingtheEnergyAcquis:TheExternalAgendaShapingPoweroftheEU.EnergyPolicyMakingintheEU.J.Tosun,S.BiesenbenderandK.Schulze.London,Springer.28:63-83.

Sebenius, J. K. (1992). "Challenging conventional explanations of internationalcooperation: negotiation analysis and the case of epistemic communities."InternationalOrganization46(01):323-365.

Shipan,C.R.andC.Volden(2014)."Whenthesmokeclears:expertise,learningandpolicydiffusion."JournalofPublicPolicy34(03):357-387.

SimonePolilloandMauroF.Guillén(2005)."GlobalizationPressuresandtheState:The Worldwide Spread of Central Bank Independence." American Journal ofSociology110(6):1764-1802.

Slaughter,A.-M.(1997)."Therealnewworldorder."ForeignAffairs:183-197.

Slaughter, A.-M. (2000a). "Accountability of Government Networks, The." Ind. J.GlobalLegalStud.8:347.

Slaughter,A.-M.(2002)."Globalgovernmentnetworks,globalinformationagencies,anddisaggregateddemocracy."Mich.J.Int'lL.24:1041.

Slaughter,A.-M.(2004).Anewworldorder,Princeton:PrincetonUniversityPress.

Slaughter, A.-M. (2004a). "Disaggregated Sovereignty: Towards the PublicAccountabilityofGlobalGovernmentNetworks."GovernmentandOpposition39(2):159-190.

Slaughter, A.-M. (2004b). "Sovereignty and power in a networked world order."StanfordJournalofInternationalLaw40:283-327.

Slaughter,A.-M.(2017).TheChessboardandtheWeb.StrategiesofConnectioninaNetworkedWorld.,YaleUniversityPress.

Slaughter,A.-M.andD.Zaring(2006)."NetworkingGoesInternational:AnUpdate."AnnualReviewofLawandSocialScience2(1):211-229.

Smith, K. E. (2005). "The outsiders: the European neighbourhood policy."Internationalaffairs81(4):757-773.

264

Spiller,P.T.(1996)."InstitutionsandCommitment."IndustrialandCorporateChange5(2):421-452.

Spiller,P.T.andM.Tommasi(2005).Theinstitutionsofregulation:Anapplicationtopublicutilities.Handbookofnewinstitutionaleconomics,Springer:515-543.

Splawn, W. M. W. (1939). "Railroad Regulation by the Interstate CommerceCommission."AnnalsoftheAmericanAcademyofPoliticalandSocialScience201:152-164.

Stephenson, P. (2013). "Twenty years of multi-level governance: ‘Where Does ItComeFrom?WhatIsIt?WhereIsItGoing?’."JournalofEuropeanPublicPolicy20(6):817-837.

Stern,J.(2012)."Therelationshipbetweenregulationandcontractsininfrastructureindustries: Regulation as ordered renegotiation." Regulation & Governance 6(4):474-498.

Stern, J. (2014)."TheBritishutility regulationmodel: Its recenthistoryand futureprospects."UtilitiesPolicy31:162-172.

Stern,J.andF.Trillas(2003)."Independenceanddiscretionintelecommunicationsregulation:lessonsfromindependentcentralbanks."UtilitiesPolicy11(4):191-201.

Stigler, G. J. (1971). "The theory of economic regulation." The Bell Journal ofEconomicsandManagementScience2(1):3-21.

Stigler,G. J. andC. Friedland (1962). "What can regulators regulate? The case ofelectricity."JournalofLawandEconomics5:16.

Stonex,W.L.(1919)."TheCaseforGovernmentOperationandOwnership."AnnalsoftheAmericanAcademyofPoliticalandSocialScience86:67-76.

Sunstein,C.R.(1989)."InterpretingStatutesintheRegulatoryState."HarvardLawReview103(2):405-508.

Sunstein, C. R. (2014). Valuing life : humanizing the regulatory state. Chicago ;London,TheUniversityofChicagoPress.

Sutherland, P. (1992). The Internal Market after 1992. Meeting the challenge.EuropeanReport,EEC.

Tarrant, A. and R. D. Kelemen (2007). Building the Eurocracy: The Politics of EUAgenciesandNetworks.BiennialEuropeanUnionStudiesAssociationConvention.Montreal,Canada.

265

Tarrant,A.andR.D.Kelemen(2017)."ReconceptualizingEuropeanUnionregulatorynetworks:AresponsetoBlaubergerandRittberger."Regulation&Governance11(2):213-222.

Thatcher,M.(2007).ReformingNationalRegulatoryInstitutions:theEUandCross-National Variety in EuropeanNetwork Industries. Beyond Varieties of Capitalism:Conflict,ContradictionandComplementaritiesintheEuropeanEconomy.B.Hancke,Rhodes,M.,Thatcher,M.Oxford,OUP:147-172.

Thatcher,M.(2011)."ThecreationofEuropeanregulatoryagenciesanditslimits:acomparative analysis of European delegation." Journal of European Public Policy18(6):790-809.

Thatcher, M. and D. Coen (2008). "Reshaping european regulatory space: Anevolutionaryanalysis."WestEuropeanPolitics31(4):806-836.

Thelen, K. (2003). How Institutions Evolve: Insights from Comparative HistoricalAnalysis.ComparativeHistoricalAnalysisintheSocialSciences.J.MahoneyandD.Rueschemeyer.Cambridge(UK),CambridgeUniversityPress:208-241.

Tholens, S. (2014). "An EU–South Mediterranean Energy Community: The RightPolicyfortheRightRegion?"TheInternationalSpectator49(2):34-49.

Torriti, J. (2010). "Impact Assessment and the Liberalization of the EU EnergyMarkets: Evidence-Based Policy-Making or Policy-Based Evidence-Making?" JCMS:JournalofCommonMarketStudies48(4):1065-1081.

Trein,P.(2017)."Coevolutionofpolicysectors:Acomparativeanalysisofhealthcareandpublichealth."PublicAdministration95(3):744-758.

Urpelainen, J. (2010). "Enforcement and capacity building in internationalcooperation."InternationalTheory2(1):32-49.

USA Congress, O. o. T. A. (1993). Energy Efficiency Technologies for Central andEasternEurope.WashingtonDC,USCongress.

USAID(2007).TransformationalDevelopment:LessonsfromtheEnergyTransitioninEuropeandEurasia.EnergyUpdate,USAID.2.

Uzzi,B.(1997)."SocialStructureandCompetitioninInterfirmNetworks:TheParadoxofEmbeddedness."AdministrativeScienceQuarterly42(1):35-67.

Vagliasindi, M. (2004). "The role of investment and regulatory reforms in thedevelopmentof infrastructure across transition economies."Utilities Policy12(4):303-314.

266

Vagliasindi,M.andA.K. Izaguirre (2007)"Privateparticipation in infrastructure inEuropeandCentralAsia.Alookatrecenttrends."GRIDLines.

VanBoetzelaer,K.andS.Princen(2012)."TheQuestforCo-ordinationinEuropeanRegulatoryNetworks*."JCMS:JournalofCommonMarketStudies50(5):819-836.

Vandeveer, S. D. and G. D. Dabelko (2001). "It's Capacity, Stupid: InternationalAssistanceandNationalImplementation."GlobalEnvironmentalPolitics1(2):18-29.

Vasconcelos, J. (2005). "Towards the Internal Energy Market: how to bridge aregulatory gap and build a regulatory framework." European Review of EnergyMarkets1(1):1-17.

Vasconcelos,J.(2009)."EnergyregulationinEurope:regulatorypoliciesandpoliticsofregulation."EuropeanReviewofEnergyMarkets3(3):1-12.

Vazquez, M. and M. Hallack (2017). The role of regulatory learning in energytransition: the case of solar PV in Brazil. Working Paper Series. IEFE. BocconiUniversity,Milan.

Verdier, P.-H. (2009). "Transnational Regulatory Networks and Their Limits." YaleJournalofInternationalLaw34(1):113-172.

Vestlund, N.M. (2015). "Pooling administrative resources through EU regulatorynetworks."JournalofEuropeanPublicPolicy24(1):61-80.

von Hirschhausen, C. and P. Opitz (2001). Power Utility Re-Regulation in EastEuropean and CIS Transformation Countries (1990 - 1999). An InstitutionalInterpretation. DIW Discussion Papers. Berlin, Deutsches Institut fürWirtschaftsforschung.

Wallsten,S.(2006).TelecommunicationRegulationinUSStates:ItsRiseandImpactsin the Early Twentieth Century. Working Paper. AEI. Brookings Joint Center forRegulatoryStudies.

Waterman, R. (1919). "Proposed Plans for Railroad Legislation." Annals of theAmericanAcademyofPoliticalandSocialScience86:91-102.

Wilson,J.Q.(1980).ThePoliticsofregulation,BasicBooks.

Winters,M.S.(2010)."ChoosingtoTarget:WhatTypesofCountriesGetDifferentTypesofWorldBankProjects."WorldPolitics62(3):422-458.

267

WorldBank(1995).MassPrivatizationinCentralandEasternEuropeandtheFormerSovietUnion.StudiesofEconomiesinTransformation.A.E.IraW.Lieberman,MichalMejstrik,JoyitaMukherjeeandPeterFidler.WashingtonDC,WorldBank.

WorldBank(2002).Transition.TheFirstTenYears.AnalysisandLessonsforEasternEuropeandtheFormerSovietUnion.Washington,DC,WorldBank.

WorldBank(2006).InfrastructureattheCrossroads.WashingtonDC,WorldBank.

Yesilkagit,K.(2011)."Institutionalcompliance,Europeannetworksofregulationandthebureaucraticautonomyofnationalregulatoryauthorities."JournalofEuropeanPublicPolicy18(7):962-979.

Younas,J.(2008)."Motivationforbilateralaidallocation:Altruismortradebenefits."EuropeanJournalofPoliticalEconomy24(3):661-674.

Youngs,R.(2014)"ANewGeopoliticsofEUEnergySecurity."

Zaring, D. (1998). "International law by other means: the twilight existence ofinternationalfinancialregulatoryorganizations."TexasInternationalLawJournal33:281-330.

Zaring, D. (2009). "Three challenges for regulatory networks." The InternationalLawyer49(1):211-217.

Zaring,D. (2012). "Finding Legal Principle inGlobal Financial Regulation." VirginiaJournalofInternationalLaw52(3).

268

APPENDICES

269

Appendix1–InterviewsinformingChapter2and3

Interviewnumber Currentlyorformerlyinvolved? Role Topicofinterview

1 Current Regulator CEER


3 Former Regulator CEER


5 Current Academic CEER

6 Former Consultant CEER


8 Former Executive CEER

9 Former Executive CEER

10 Current Executive CEER




14 Former Consultant CEER


16 Former Executive NARUC


18 Current Consultant CEER



21 Current Consultant NARUC


23 Former Regulator NARUC

24 Current Academic CEER

25 Current Regulator NARUC

26 Current Executive NARUC

270








34 Current Academic NARUC





39 Current Academic NARUC



271

Appendix2–DatausedinChapter4-ElectricitysectorstructureinEUcountries

Country Electricitysectorstructure Maincompaniesoperatingin

Generation Transmission Distribution Retail

Austria The Republic of Austria owns 51% of Verbund's share

capital.ThefederalprovinceLowerAustriaholds(througha

holdingcompany)51%ofthesharecapitalofEVN.Themain

AustrianTSOisAustrianPowerGridAGowning94%ofthe

Austrian high voltage electricity grid. Verbund AG holds

100%ofthesharesinAustrianPowerGridAG.Thereare14

DSOs anddistribution areas inAustria.WienEnergie, Linz

Strom,SalzburgNetzandKelagarethemainparticipantsin

thissector.Oftheroughly130retailers,manyoperateonly

at local or regional level. The three largest suppliers are

Verbund,EVNandWienEnergie.

Verbund;EVN Verbund;EVN Verbund;

EVN;

provinces

Verbund,

EVN, Wien

Energie

Belgium Traditional suppliers, notablyGDF Suez and its subsidiary

Electrabel, continue to hold dominant positions for both

generation and supply. Other key generating companies

operating inBelgium includeEDFandE.ON. In thesupply

sector,inadditiontoElectrabel,someofthelargeplayers

includeEDF,EniandLampiris.InBelgium,DSOsaremainly

Electrabel; EDF;

E.ON

Elia regional Electrabel;

EDF; ENI;

Lampiris

272

groupedtogetherinorganisationslikeORESinWalloniaand

Eandis in Flanders. For example, Eandis is comprised of

seven Flemish electricity and gas distribution system

operators,whicharealsoitsshareholders.Elia,theTSO,has

alegalmonopolyasBelgium’ssoleelectricityTSO.

Bulgaria The 100% state-owned company, BEH is the parent

companyofmostgeneratingcompaniesandownstheTSO.

Electricity distribution companies were fully privatised in

2012.ThusEVN(Austria)andEnergo-Pro(E.OnnowCzech

ENergoPro) own 100% of the shares of the respective

subsidiarycompaniesandCEZ(SofiaandWesternBulgaria)

ownsover90%.Captivecustomersandthosewhohavenot

decidedtochangetheirsupplierbuytheirelectricityfrom

NEK (if they are connected to a high voltage grid) or the

distributioncompanies(iftheyareconnectedtoamedium

orlowvoltagegrid).

BEH BEH Private: EVN

(Austria);

Energo-Pro

(CEZ);CEZ

Private:EVN

(Austria);

Energo-Pro

(CEZ); CEZ;

NEK

Croatia Thestate-ownedholdingcompanyHEPd.d.istheownerof

theinfrastructure,whichismanagedbyitssubsidiaries(e.g.

HEP-DSO, HEP Generation). The transmission grid is

operated by the state-owned TSO HOPS. There are 28

companiesactiveinthegenerationsector.Themajorityof

theseareprivatelyowned.Theirmarketshareisdwarfedby

HEP HOPS HEP-DSO HEP-

Opskrba

d.o.o. (part

of the HEP

group)

273

thegenerationcapacitiesofstate-ownedcompanies.There

are 18 companies covering the supply business. Three of

thesecompaniesarestate-ownedandholdthemajorityof

themarket share. Theprivatelyowned supply companies

with the highest market share in 2014 were: GEN-I

(approximately 6.07%), RWE Energy (former Energija 2

Sustavi with approximately 4.52%) and Proenergy

(approximately2.32%).

Cyprus The incumbentElectricityAuthorityofCyprus (EAC)owns

boththetransmissionandthedistributionsystem.TheTSO

islegallybutnotfunctionallyunbundledfromEAC,sinceall

itsstaffissecondedfromEAC.Theobligationofownership

unbundling of the TSO does not apply, since Cyprus has

obtained a derogation from Article 9 of the 2009/72/EC

Directive.TheDSO is responsible formanaging,operating

and developing the network, safeguarding access to the

distributionnetworkandequaltreatmentforallusers.EAC

has unbundled the accounts of the DSO. No wholesale

market is currently operating. EAC is the sole electricity

supplier.

EAC EAC EAC EAC

274

Czechrepublic A large part of the generation, distribution and supply

segmentsareintegratedbusinessesownedbyCEZ,a.s.and

its subsidiaries (CEZ Group). CEPS is the sole owner and

operatorofthetransmissiongrid,whichisfullyownedby

theCzechstate.Thedistributionsystem ispredominantly

ownedandoperatedby:EZDistribucea.s.,E.ONDistribuce

a.s.andPREdistribucea.s.(DSOs).TheCzechRepublichas

retainedacontrollingshareholdingofapproximately48%of

eachof thedistributioncompanies. In2014, the3 largest

Czech electricity suppliers were EZ Prodej s.r.o., E.ON

Energiea.sandPrasköenergetikaa.s.

CEZ CEZ EZ

Distribuce;

E.ON

Distribuce;

PREdistribuc

e

CEZ; E.ON

Energie

Denmark Two companies, DONG and Vattenfall, share 70% of

market'scapacity.BothgridsaremanagedbyEnerginet.dk,

an independent public enterprise owned by the Danish

State under the Ministry of Climate and Energy. DONG

Energycoversa leadingpositionholdinga21%ofmarket

shares,followedbyEnergiden(10%circa),EnergiFyn,SEAS-

NVE and Natur-Energi. The largest DSO DONG Energy

Distribution has more than a million customers which is

morethan20%oftheDanishpopulationwhilethesmallest

DSOsonlyhaveacoupleofthousandcustomers.

DONG;Vattenfall Energinet.dk DONG

Energy

Distribution

DONG

Energy

275

Estonia In Estonia the TSO Elering AS and the main electricity

producer and seller AS Eesti Energia along with its

subsidiary,thedistributionnetworkoperatorElektrileviOU,

belong100%totheState.ElectricityproductioninEstonia

isdominatedbythestate-ownedcompany,EestiEnergia.In

2011,itsshareofthewholesaleelectricitymarketwas90%,

whileitsshareoftheretailmarketwas76%.Transmission

wasunbundledfromproductionin2010.Thestate-owned

company, Elering, provides the transmission networking

service, but also acts as the single transmission system

operator.Thereare38distributionnetworks,thelargestof

which is owned by Eesti Energia, with 86% share of the

distributionmarket.

EestiEnergia EleringAS Electrilevi

OU

Eesti

Energia

Finland The electricity market is dominated by Fortum, whose

market share in Finland’selectricitymarket is close to27

percent.FortumOyjisapubliclylistedenergycompany,in

whichthestateholds50.8%ofshares.PohjolanVoimaOy

(PVO) is the secondbiggest Finnishenergy company. The

FinnishtransmissiongridisownedbyFingrid,anotherstate-

owned company. Distribution companies are owned by

municipalitiesorprivatecompanies.TheDSOswithlargest

Fortum;PVO Fingrid Fortum;

Vattenfall;

municipaliti

es

Fortum

276

marketshareareFortumwithamarketshareof19%and

Vattenfallwithamarketshareof12%.

France In November 2004, the two incumbent monopoly

companies,ElectricitédeFrance (EDF)andGazdeFrance

(GDF), both of which were 100% state-owned, became

limitedcompanieswithaboardofdirectors.Thenextyear,

minoritystakesinthetwocompaniesweresoldtoprivate

investors.AsofDecember2014,thestateretainsan84.5%

stake in EDF, which dominates the sector - 90% of

generation;100%ofRTE(theTSO);100%ofERDF(DSO,95%

of the market); 91% of retail. The remaining 5% of the

distribution network are managed by local authorities

(collectivités territoriales), who also own the entire

network.

EDF RTE ERDF EDF

277

Germany TheGermanenergyindustryhastraditionallybeenprivately

owned, though there are still a large number of small

electricity and gas distribution companies that are either

wholly or partially owned by municipalities. Despite

reforms, the incumbent operators in the wholesale and

retailmarketshaveretainedlargemarketshares.E.ONand

RWEhavebeenamong thedominantplayers inboth the

natural-gas and the electricity markets. EnBW (Baden-

Wuttenberg),E.On,RWE,Vattenfall leadingingeneration,

distribution and supply. There are four TSOs: Transnet

(EnBW); Tennet (Holland govt); Amprion (Commerzbank,

RWE);50Hertz(EliaandIFM)

E.On,RWE,EnBW

(Baden-

Wuttenberg)

Transnet

(EnBW);

Tennet

(Holland govt);

Amprion

(Commerzbank

, RWE);

50Hertz (Elia

andIFM)

E.On, RWE,

EnBW

(Baden-

Wuttenberg

);

municipaliti

es

E.On, RWE,

EnBW

(Baden-

Wuttenberg

)

Greece The electricity sector remains dominated by the state-

controlled Public Power Corporation (PPC) and its

subsidiaries. The PPC’s generation market share has

declinedfrom98.6%in2003to65%in2013.Thecompany

continuestocontrolalmostallelectricitysupplyonthenon-

interconnected islands. In the electricity transmission

sector,thePPCownsalltransmissionlinesandholdsa49%

shareof assets in the transmission systemandwholesale

marketoperator(HTSO)withtherestownedbytheGreek

state.

PPC PPC PPC PPC

278

Hungary Thestate-ownedMVMisthebiggestplayerinthemarket,

controlling a considerable part of Hungary’s generation

capacity. MAVIR, a subsidiary of the state-owned MVM,

owns and operates the transmission system. The

distribution networks are owned and operated by six

privately owned DSOs (ÛDSZ (100% E.ON); DÛDSZ (100%

E.ON); TITSZ (100% E.ON); DÛMSZ (100% EDF); ÛMSZ

(54.3% RWE; 25% EnBW; 12%MVM; 18.7% others(; ELM

(55.3% RWE; 25% EnBW; 15.6% MVM; 4.1% others).

Outsidetheuniversalservicesector(DSOs),therecurrently

aremorethan50licensedtraderssellingelectricitytoend

usersatmarketbasedprices.

MVM MVM ÛDSZ;

DÛDSZ;

TITSZ;

DÛMSZ;

ÛMSZ;ELM

ÛDSZ;

DÛDSZ;

TITSZ;

DÛMSZ;

ÛMSZ;ELM.

Ireland State-owned companies dominate the electricity and

natural-gassectors.TheElectricitySupplyBoard(ESB)holds

two-thirdsofgeneratingcapacity,thoughitssharehasbeen

fallingasnewpowerproducershaveenteredthemarket.It

alsoownsthetransmissionsystem,theoperationofwhich

istheresponsibilityofanotherstate-ownedbody,EirGrid,

aswellasthedistributionnetwork.BordGaisEireann(BGE)

owns the gas transmission and distribution network,

operating the transmission system through a subsidiary

ESB EirGrid BGE BGE

279

company. Retail competition has developed to only a

relativelysmalldegree.

Italy ThemajorplayerintheelectricitygenerationmarketisEnel.

EneliscontrolledbytheItaliangovernment.Edison,ENIand

E.ONarealsokeymarketplayers.EnelDistribuzioneisthe

maindistributionnetworkoperator(DNO),with86%ofthe

distributedelectricityvolumes.OtherDNOs,significantby

marketshares,are:A2A(3.9%),AceaDistribuzione(3.2%)

andAem TorinoDistribuzione (1.4%). Enel is the primary

supplierwithabout37%oftheoverallsalesofelectricity.

The other major suppliers by market share are: Edison

group,withamarketshareof8.2%,followedbyAcea,with

a share of 4.6%, and Eni, with amarket share of almost

4.3%.

Enel, Edison, ENI,

E.ON

Terna Enel

Distribuzion

e; ACEA;

A2A

Enel;

Edison;

ACEA;Eni

Latvia Latvenergo enjoys a monopoly position as the largest

producer of electricity in Latvia and controls all of the

country’s public electricity distribution networks. The

Ministry of Economy, via Latvenergo, is also the ultimate

beneficiary shareholder of Latvenergo subsidiaries. The

Ministry of Finance owns 100% of the shares in the

Latvenergo Augstspriegum

atikli

Latvenergo Latvenergo

280

electricitytransmissioncompany,Augstspriegumatīkli.The

company isa formerLatvenergosubsidiarythatwasspun

offin2012inordertocomplywithDirective2009/72/ECof

theEuropeanParliamentandoftheCouncilof13July2009

concerning common rules for the internal market in

electricity.

Lithuania TheMinistryofEnergyexercisesstateownershiprights in

UAB EPSO-Gwhich has two listed subsidiaries: AB Litgrid

and AB Amber Grid, which operate respectively the

electricity and the natural gas transmission grid. State

ownership rights in Lithuanian Energy were previously

exercisedbytheMinistryofEnergybutweretransferredto

theMinistryofEconomyandthentotheMinistryofFinance

in 2012-13 in implementation of the 2009 Third Energy

Packageof theEuropeanParliament,which required that

theownershipofenergygenerationandsalebeseparate

fromtheownershipofenergytransmissionnetworks.

LietuvosEnergia ABLitgrid ABLesto ABLesto

Luxembourg Creos Luxembourg S.A. (formerly SOTEG) owns and

operates the transmission system, and it supplies the

majorityofthemarket.MostofCreos'ssharesareowned

by various private utilities, though the State maintains

minority ownership. Creos also operates one of the two

Enovos

International

Creos Municipaliti

es

Creos

281

main electricity-transmission systems in the country. The

other main electricity grid operator is the Societe de

Transport de l'Electricite (SOTEL).Most of the electricity-

distributioncompaniesareownedbymunicipalities.

Malta Therearenotransmissionsystemsortransmissionsystem

operators inMalta. The distribution system covering the

whole country remains under the responsibility of one

distributionsystemoperatorwhichformspartofavertically

integratedcompany,Enemaltaplc.Unbundlingisrequired

atinternalmanagementaccountslevelonly.Theelectricity

generationmarket isopentocompetitionandgenerators

mayproduceelectricityfortheirownconsumptionand/or

selltoEnemaltaplc.Theretailofelectricityisnotopento

competition.

Enemalta n/a Enemalta Enemalta

Netherlands Most electricity is generated by Essent (owned by RWE),

Nuon (owned by Vattenfall), Eneco, E.ON, Delta and

Electabel. The electricity distribution and transmission

networks are publicly managed and owned. The

Netherlands has one electricity transmission system

operator(TSO),TenneT,whichisownedbythestate.Since

theliberalisation,anumberofnewsupplycompanieshave

Essent (RWE);

Nuon(Vattenfall);

Eneco,E.ON.

TenneT Municipaliti

es

Nuon

(Vattenfall);

Essent

(RWE);

Eneco;Delta

282

joinedthemarketincludingDougEnergyandE.ONThefour

largestenergycompanies -Nuon(acquiredbyVattenfall),

Essent(acquiredbyRWE),EnecoandDELTA-stillhavethe

largestshareoftheretailmarket.

Norway TheNorwegianpowersectorcomprisesalargenumberof

mostly publicly owned participants. Around 90% of

generating capacity is in public ownership, with local

municipalities and county authorities alone owning just

overhalf. The state-ownedutility, Statkraft, is the largest

generator. There are more than 160 small distribution

systemoperators(DSOs)inNorway,mostofthempublicly

owned. The dominant supplier within a network area is

most often a vertically integrated supplier or a supplier

withinthesamecorporationastheDSO.

Statkraft Statnett regionaland

local

companies

regionaland

local

companies

Poland State-owned entities have been organised into four

vertically integratedgroupsandpartiallyprivatised.These

groups are PGE Polska Grupa Energetyczna S.A. (PGE),

TAURONPolskaEnergiaS.A. (TAURON),ENEAS.A. (ENEA)

and ENERGA S.A. (ENERGA). All of these companies

combine generation, distribution and trading (including

supply)activities. In2012PGE,TAURONandEDFhad the

largestmarketshareinthegenerationsector.ThesoleTSO,

PGE; TAURON;

EDF

PSE PGE

Dystrybucja;

TAURON

Dystrybucja;

ENERGA

Operator;

ENEA

Operator

PGE;

TAURON;

ENERG;

ENEA

283

PolskieSieciElektroenergetyczneS.A.(PSE),isafullystate-

owned joint stock company and owner of all the

transmissionassets.

Portugal During the first quarter of 2012, the Portuguese energy

sector saw the privatisation of a significant part of the

transmission operator's share capital (Redes Energeticas

Nacionais - REN) and of the incumbent's share capital

(EnergiasdePortugal-EDP),aformerverticallyintegrated

company, which now develops, through its subsidiaries,

generation, distribution and supply activities. Electricity

suppliers entail not only Portuguese companies (EDP

Comercial and Galp Power) but also several Spanish

companies (suchas Endesa, Iberdrola,Uniãn Fenosa, EGL

EnergêaIberiaandNexusEnergêa).

EDP;Galp REN Private EDP

Comercial;

GalpPower;

Endesa;

Iberdorla

284

Romania In Romania, 89% of the national electricity output is

generatedbystate-ownedgenerators.TheRomanianstate

ownsthemajorityofthesharesintheTSO(Transelectrica

SA).Fiveoftheeightdistributionoperatorsareprivatised:

CEZ Distribute SA; ENEL Distributie Banat SA; ENEL

DistributieDobrogeaSA;E.ONMoldovaDistributieSA;ENEL

Distributie Muntenia SUD SA; FDEE Electrica Distributie

MunteniaNord SA; FDEE ElectricaDistributie Transilvania

SudSA;andFDEEElectricaDistributieTransilvaniaNordSA.

Electrica SA (owner of the three companies (vi) to (viii)

above) holds the largest market share (39.27%) and was

majoritystateownedbytheMinistryofEconomy.

Hidroelectrica SA;

SN

Nuclearelectrica

SA; Complexul

Energetic Oltenia

SA;

Termoelectrica

DevaSA

Transelectrica

SA

Private: 6

DSOs

Electrica

Furnizare;

Enel

Energie;

EnelEnergie

Muntenia;

E.ON

Energie

Romania;

CEZVanzare

SlovakRepublic ThemainplayerintheSlovakelectricitygenerationmarket

is Slovensk elektrörne, a.s. (SE), a joint stock companyof

which66%isownedbyEnel-theItalianbasedmultinational

group,withtheother34%ownedbytheSlovakstate.The

entireelectricitytransmissionnetworkisownedbySEPS,a

wholly state-owned company. Three are regional

Distribution System Operators (DSO), co-owned by the

state(51%)andaprivate investor (49%andmanagement

control). E.ON indirectly holds 49% of the shares in ZSE

Distribucia a.s.; EDF holds 49% of the shares in SSE -

SE SEPS ZSE (E.ON);

SSE (EDF);

RWE

ZSE (E.ON);

SSE (EDF);

RWE

285

Distribuciaa.s.;andRWEEnergyAGholds49%oftheshares

inVchodoslovensködistribunöa.s.

Slovenia The key companies operating in the Slovenian electricity

market are: Elektro-Slovenija, d.o.o. (ELES), the

TransmissionSystemOperator(TSO);SODOd.o.o.(SODO),

the Distribution System Operator (DSO); 6 distribution

companies; Holding Slovenske Elektrarne d.o.o. (HSE), a

generation company and GEN energija d.o.o. (GEN), a

generation company. All the generators, distributors and

suppliers,aswellastheTSO,arepredominantlyorwholly

state-owned and no international investment is present.

Due to the levelof stateownership, thewholeelectricity

sectorisarguablyfullyverticallyintegrated.

HSE;GEN ELES SODO GEN-I

Spain Three-quarters of electricity is generated by three

companies:Iberdrola,Endesa(almost100%ownedbythe

ItalianutilityfirmENEL)andUnionFenosa(ownedbyGas

Natural).IberdrolaandEndesaaloneaccountforthebulkof

retailsales.REE(RedElectricadeEspana),inwhichthestate

holds a 20% stake, serves as the transmission system; it

ownsalmost theentire400kVgridand two-thirdsof the

220 kV grid. Iberdrola, Endesa andUnion Fenosa are the

Endesa (100%

ENEL); Iberdrola;

UnionFenosa

REE Private:

Iberdrola,

Endesa,

Union

Fenosa

Iberdrola,

Endesa

286

largestdistributors,althoughtherearemorethan300small

localdistributors.

Sweden The state-owned company Vattenfall is one of themajor

playersintheSwedishelectricitymarket.Mostofthesmall

local electricity distribution companies are owned by

municipalities. Three companies - Vattenfall, Fortum

(majority-owned by the Finnish government), and E.ON

Sverige-generatethebulkofpowerinSweden,ownmost

of the distribution assets and account for around half of

retail sales. The E.ON and Vattenfall groups have several

DSO areas. The market share for the three dominant

companies is51,4% (E.ON19,0%,FortumandVattenfall

16,2%each).ThestateownedutilitySvenskaKraftnätisthe

Transmission System Operator (TSO) for the electricity

market.

Vattenfall;

Fortum; E.ON

Sverige

Svenska

Kraftnät

Municipaliti

es;

Vattenfall;

Fortum;

E.ON

Sverige

Vattenfall;

Fortum;

E.ON

Sverige

287

UnitedKingdom The National Grid owns and operates the England and

Wales transmission system; the Scottish transmission

system is owned by Scottish Power and Scottish and

Southern Energy, and the Northern Ireland network by

Northern Ireland Electricity. Licences for 14 distribution

areas in Great Britain are currently held by six different

companies (Electricity North West Limited, Northern

Powergrid,SSE,ScottishPowerEnergyNetworks,UKPower

Networks,andWesternPowerDistribution).Retailsupply,

which isunbundledfromdistribution, isdominatedbysix

large companies (EDF Energy, E.ON, RWE,

Iberdrola/ScottishPower, Centrica and SSE) which supply

virtuallyallconsumers.

EDFEnergy;E.ON;

RWE;

Iberdrola/Scottish

Power; Centrica;

SSE

4TSOs Private: 14

DNOs are

ownedbysix

different

groups.

EDF Energy;

E.ON; RWE;

Iberdrola/Sc

ottishPower

; Centrica;

SSE

Compiledbytheauthorfromvarioussources,primarilyOECDFactsheets(2015)andCMSLawe-guides(datafrom2013/2014)

288

GovernmentownershipacrossdifferentsegmentsofelectricitysectorinEUcountries

Country Percentage of government ownership of the largest

companyactiveinelectricity…

Generation Transmission Distribution Retail

Austria 51 51 51 51

Belgium 0 45.4 79 0

Germany 0 0 0 0

Luxembourg 7.5 57.3 57.3 25.4

Netherlands 0 100 100 33

Bulgaria 60 100 0 0

Czech

Republic

69.8 100 69.8 69.8

Hungary 99.9 100 0 0

Poland 61.9 100 40.5 40.5

Romania 80 74 0 0

Slovak

Republic

34 100 51 51

Croatia 100 100 100 100

Cyprus 100 100 100 100

Estonia 100 100 100 100

Ireland 95 100 95 95

Latvia 100 100 100 100

Lithuania 96.1 97.5 82.6 82.6

Malta 100 100 100 100

Slovenia 100 100 100 50

United

Kingdom

0 0 0 0

289

France 84.4 84.4 84.4 84.4

Greece 51 51 51 51

Italy 31.2 29.9 31.2 31.2

Portugal 0 10.3 0 0

Spain 0 20 0 0

Denmark 80 100 80 80

Finland 50.8 53.1 50.8 50.8

Norway 100 100 0 100

Sweden 100 100 100 100

OECD Product Market Regulation Indicators -

http://www.oecd.org/eco/growth/indicatorsofproductmarketregulationhomepage.htm#i

ndicators

290

Appendix3–AdditionalmodelsforChapter5

Associationbetweenstaffcategories(ACER2016)andpopulationfigures(dependentvariable,Eurostat2016)

Model1(Intercept) 17.79***

(0.52)large -0.90

(0.59)

medium-large reference

medium -1.33* (0.60)

medium-small -2.08** (0.62)

small -2.79*** (0.62)

micro -4.28*** (0.64)

R2 0.78Adj.R2 0.73RMSE 0.74

Num.obs. 29

***p<0.001,**p<0.01,*p<0.05

Relationshipbetweennetworkactivism(dependentvariable)stafflevels(differentcategorization)andcovariates

Model1 Model2 Model3 Model4 Model5(Intercept) -2.00* 0.97 0.77 1.30 0.60 (0.79) (2.12) (3.00) (2.76) (2.19)Stafflevels large reference medium 3.33* 2.99* 3.02* 2.69 2.83* (1.25) (1.24) (1.29) (1.33) (1.32)medium-small 2.20 2.19 2.16 1.74 2.06 (1.32) (1.28) (1.32) (1.36) (1.34)small 3.60* 3.48* 3.17* 2.84 2.83

291

(1.32) (1.29) (1.39) (1.44) (1.46)micro 2.25 1.73 1.86 0.45 1.72 (1.42) (1.43) (1.48) (2.02) (1.88)Independence(electricity) -0.95 (0.63) Independence(gas) -0.88 (0.92) Liberalization(electricity) -0.75 (0.60) Liberalization(gas) -0.64 (0.51)R2 0.31 0.37 0.34 0.35 0.36Adj.R2 0.19 0.23 0.19 0.21 0.21Num.obs. 29 29 29 29 27RMSE 2.36 2.30 2.37 2.34 2.39

***p<0.001,**p<0.01,*p<0.05

292

Appendix4–InterviewsinformingChapter6and7

Interview

number

Previous

interview

number

Currentlyorformerly

involved?

Role Topicofinterview

1 1 Current Regulator MEDREG

2 2 Current Regulator MEDREG

3 3 Former Regulator MEDREG/ERRA

4 4 Former Regulator MEDREG

5 5 Current Academic MEDREG/ERRA

6 6 Former Consultant MEDREG/ERRA

7 7 Former Regulator MEDREG/ERRA

8 8 Former Executive ERRA

9 16 Former Executive NARUC/ERRA

10 Current Executive MEDREG

11 17 Current Regulator MEDREG/Energy

Community

12 18 Current Consultant MEDREG/ERRA



15 21 Current Consultant ERRA

16 23 Former Regulator ERRA

17 Current Executive ERRA

18 Former Regulator ERRA

19 Current Regulator ERRA

20 25 Current Regulator ERRA

21 Current Regulator MEDREG


23 26 Current Executive ERRA


293

25 Current Executive ERRA





30 Current Executive EnergyCommunity

31 31 Current Executive ERRA/Energy

Community

32 33 Current Executive ERRA/Energy

Community

33 Current Regulator EnergyCommunity


35 34 Current Academic ERRA



38 39 Current Academic ERRA

39 41 Current Executive ERRA

40 Current Regulator MEDREG


42 19 Former Regulator EnergyCommunity

Documents

NETWORKS AS LEVERS · 2018-07-23 · MEDITERRANEAN ENERGY REGULATORS ... (MedReg) and the Energy Regional Regulatory Association (ERRA) of Central and Eastern Europe. 8 Introduction: