networks as levers · 2018-07-23 · mediterranean energy regulators ... (medreg) and the energy...
TRANSCRIPT
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)
DependentMarketEconomies
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)
DependentMarketEconomies
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
2 Current Regulator CEER
3 Former Regulator CEER
4 Former Regulator CEER
5 Current Academic CEER
6 Former Consultant CEER
7 Former Regulator CEER
8 Former Executive CEER
9 Former Executive CEER
10 Current Executive CEER
11 Current Regulator CEER
12 Former Regulator CEER
13 Current Regulator CEER
14 Former Consultant CEER
15 Current Regulator CEER
16 Former Executive NARUC
17 Current Regulator CEER
18 Current Consultant CEER
19 Former Regulator CEER
20 Current Regulator CEER
21 Current Consultant NARUC
22 Current Regulator CEER
23 Former Regulator NARUC
24 Current Academic CEER
25 Current Regulator NARUC
26 Current Executive NARUC
270
27 Current Regulator NARUC
28 Former Executive NARUC
29 Former Regulator CEER
30 Former Regulator CEER
31 Current Executive NARUC
32 Current Regulator CEER
33 Current Executive NARUC
34 Current Academic NARUC
35 Current Regulator NARUC
36 Current Regulator NARUC
37 Former Executive NARUC
38 Current Regulator CEER
39 Current Academic NARUC
40 Current Executive NARUC
41 Current Executive 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
13 Current Executive MEDREG
14 Current Executive MEDREG
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
22 Current Regulator ERRA
23 26 Current Executive ERRA
24 Current Regulator ERRA
293
25 Current Executive ERRA
26 Former Regulator ERRA
27 Former Regulator ERRA
28 Former Regulator ERRA
29 28 Former Executive ERRA
30 Current Executive EnergyCommunity
31 31 Current Executive ERRA/Energy
Community
32 33 Current Executive ERRA/Energy
Community
33 Current Regulator EnergyCommunity
34 35 Current Regulator ERRA
35 34 Current Academic ERRA
36 37 Former Executive ERRA
37 38 Current Regulator ERRA
38 39 Current Academic ERRA
39 41 Current Executive ERRA
40 Current Regulator MEDREG
41 Current Regulator ERRA
42 19 Former Regulator EnergyCommunity