does the organizational autonomy of regulatory agencies affect their position … · 2017. 5....
TRANSCRIPT
Does the organizational autonomy of regulatory agencies
affect their position and power in multilevel regulatory arrangements?
Comparative Case Study of the Belgian Sector Regulators in Energy and Telecommunication1
Panel 14: “New Directions in the Study of Regulatory Administration: Regulatory Networks and Regulatory Autonomy
7nd NIG Annual Work Conference, 25‐26 November 2010,
Maastricht
Joery MATTHYS*, Koen VERHOEST, David AUBIN, Emmanuelle MATHIEU, and Karin INGOLD
*Corresponding author:
Public Management Institute
K.U. Leuven
B‐3000 Leuven (Belgium)
Tel. +32 16 32 32 15
Email: [email protected]
Work in progress: do not cite without permission
1 This paper has been prepared within the framework of the ongoing REGUNET‐project. This research project of the Belgian Federal Science Office is realised in execution of the Research Programme “Society and Future”. Several public and private organizations, among which the Belgian sector regulators, are participating willingly in the REGUNET‐project, and have been quite forthcoming with information. However, the authors would like to stress that all data gathering and conclusions reached in this paper are their sole responsibility. The aforementioned organizations have never authorized or have been asked to authorize in any way the content of this paper.
Introduction In the post‐liberalization era, regulation of network utilities has often been entrusted to independent
regulatory agencies (or IRA’s), which have since become one of the main institutional features of the
rising regulatory state (Gilardi 2002). These institutions, however, do not exist in a regulatory
vacuum: they operate in a context of a wider set of authorities with general or sector‐based
regulatory competencies. As in most sectors, the regulatory tasks – comprised of rule
implementation and standard setting, licensing, monitoring and enforcement (Hood et al. 2001) – are
distributed or shared among several public actors. This network of public actors can be seen to form
a regulatory constellation or arrangement. Sector‐specific regulatory agencies are but one
component of this regulatory arrangement, which can have different features depending on the
country and/or sector analyzed.
This paper explores to what extent and in what way the autonomy of independent regulatory
agencies towards government relates to the regulatory power and position of these agencies in the
broader regulatory arrangement. A sector‐specific IRA has a strong regulatory position and power
when (1) it can take most or all decisions on sector‐related regulatory issues, independently from
influence from external actors (i.e. formal position and regulatory power), and (2) when the IRA is
perceived by all relevant actors (regulatees and other regulatory bodies involved in the regulation of
a sector) to be the most central, important and influential actor for the regulation of the market at
stake (i.e. perceived position and regulatory power).
Since the complexity of the broader regulatory arrangement can have an impact, we analyze and
compare the autonomy and regulatory power of the sector regulators within two markets that we
expect to differ in terms of the complexity of their regulatory arrangement: the energy and
telecommunication markets in Belgium. Both sectors come from a common background: they both
concern network utilities with former monopolies that were liberalized recently. The divergence in
regulatory arrangements is significant, however, most notably because telecommunication
regulation has remained Federal, while energy regulation is a shared competency of both the Federal
and Regional governments.
After a conceptual part and a theoretical section, which theorizes the relation between the
autonomy of the IRA towards government on the one hand and the regulatory power and position of
the sector regulator in the regulatory arrangement on the other hand, the paper discusses what
operationalization and methodologies were used in the multiple case study. For the data gathering
and analysis for each of the main variables (autonomy towards government, regulatory power, and
position) we rely on methodological triangulation. The empirical section then discusses each case,
before comparing across cases. Each case study first introduces the broader regulatory arrangement
that regulates the involved market. Next to that, both the formal and de facto autonomy of the
sector regulator is discussed, followed by an assessment of the formal and perceived regulatory
power and position of the sector regulator within the broader regulatory arrangement. In the
remainder of the paper, the relations between autonomy and regulatory power/position are
compared across both case studies and some implications at the empirical, theoretical‐conceptual,
and methodological level are made.
1. Independent regulatory agencies, autonomy, and regulatory power 1.1. IRA’s and their formal autonomy towards government Independent regulatory agencies are a major form of regulatory governance. Being a long practiced
governance form in the US (Carpenter 2001), these independent, non‐majoritarian agencies – which
develop and implement rules and regulations – have been created throughout Europe during the rise
of the European regulatory state (Majone 1994). EU liberalization directives have forced many
national governments to separate regulatory activity from industrial activity, thus creating the need
for delegation of regulatory activities to new actors, i.e. national regulatory authorities (Coen and
Thatcher 2000; Eberlein 2000b; Gilardi 2002). Also, because of normative pressure from international
bodies like IMF and World Bank, and globalizing markets, this model of governance has spread
rapidly and massively throughout the developed and developing world (Jordana, Levi‐Faur and
Fernandez 2007; Gilardi 2008). More broadly, a lot of government functions have been agencified,
causing scholars to refer to an ‘agency fever’ (Pollitt et al. 2001). In a lot of countries, government
bureaucracies have been disaggregated in a myriad of semi‐autonomous bodies. However,
independent regulatory agencies are a very specific subset of this broader category of agencies (see
for a definition of agencies as a broad category of semi‐autonomous public sector organizations,
Pollitt and Talbot 2004 or Verhoest et al. 2010). They are public bodies that exercise a specific
delegated public authority to develop and implement regulation, ranging from goal formulation, rule
making, and standard setting, to monitoring, enforcement, adjudication, and application of sanctions
and rewards. They are structurally and organizationally disaggregated from other parts of the public
sector, but more specifically their enacting laws usually contain provisions to ensure formal
independence from elected politicians (cfr. Thatcher and Stone Sweet 2002). In the case of IRA’s the
scope of delegated public authority is usually so encompassing, and their separateness of democratic
institutions so atypical, that these features raise questions about democratic accountability and
technocratic dominance (Lodge 2004). However, the independence of IRA’s towards politicians is
believed to be an absolute requisite for a strong regulatory position and power towards regulatees
and the markets, because independent IRA’s are believed to yield more legitimacy and credibility for
the regulatory system.
Hence, the formal autonomy of IRA’s towards government in terms of its nature and degree has been
a major topic for study among scholars in regulation studies. This type of research tries to explain this
delegation using several institutionalist theories, most notably the rational choice theory, the
sociological theory, and the historical‐institutional theory (Gilardi 2004; Horn 1995). The first theory
especially seems to be the prevalent one, focusing on principal‐agent relationship theories and
credible commitment theories in order to explain the creation of these institutions (Majone 2001;
Levy and Spiller 1996; Gilardi 2002; Levi‐Faur 1999; Thatcher 2003). Within this research a
comparative approach developed, which allowed the acknowledgement of the specificities of each
country’s regulatory institutions, its traditions of regulatory practice, and its political and economic
legacies (Wilks 2001), as well as its corporate governance structure and general competition
authorities (Eyre and Lodge 2000). Separate from this evolution, cross‐sectoral economic studies
were conducted as well, since a regulatory regime can also differ from one sector to the other
(Eberlein 2000b; Thatcher 2001; Gilardi 2008). In this literature isomorphistic pressures, and to some
extent (national and sectoral) path‐dependent variety, prove to explain the differences in formal
autonomy of IRA’s across countries (see also Yesilkagit and Christensen 2010).
1.2. Formal autonomy towards government as the main prerequisite for regulatory power In the more recent literature at least an implicit link can be observed between the formal autonomy
a regulatory institution has and the regulatory power it wields (Gilardi 2008). Normative perspectives
on the beneficial effects of regulatory independence, as propagated by international organizations
(ITO, IMF, OECD), also emphasize that the main factor that determines the regulatory power of an
IRA is the formal independence it has towards its government.
As already stated, a sector IRA has a strong regulatory position and power when (1) it can take most
or all decisions on sector‐related regulatory issues, independently from influence from external
actors (i.e. formal position and regulatory power), and (2) when the IRA is perceived by all relevant
actors (regulatees and other regulatory bodies involved in the regulation of a market sector) to be
the most central, important, and influential actor for the regulation of the market at stake (i.e.
perceived position and regulatory power). Power refers to the ability to take decisions that affect the
choices and actions of other actors in order to reach certain goals (power “over” and power “to”).
Theoretical/normative arguments based on credibility, legitimacy, and administrative capacity
support the claim that regulatory agencies that are very autonomous towards government will have
a stronger position and power towards other actors, which are involved in the regulation of a specific
market, and towards the regulatees (Majone 2001; Maggetti 2009). Figure 1 models this simple
relation.
Formal organizational autonomy of the IRA towards
government
Regulatory position and power of the IRA
THEORY
Figure 1: a simple model of autonomy and regulatory power A first argument relates to credible commitment theory: shielding‐off IRA’s from political
intervention enhances the credibility of regulatory actors and actions vis‐à‐vis industries, investors,
customers, and citizens. Delegating regulatory authority to non‐majoritarian regulators helps to deal
with the time inconsistency problem, and helps to avoid policy change because of alternations of
cabinets. The extent to which IRA’s are considered to be formally independent from elected
politicians and the short‐term political cycle enhances their legitimacy, and therefore heighten their
influence towards regulatees and other regulatory bodies. Moreover, the insulation from politics in
recently liberalized markets may also be important from another perspective: in many of these
markets, the government still owns fully or partially the incumbent, i.e. former public monopolist.
Ownership interests within government may collide with the market regulator role of government.
Such conflict of interests can be avoided by delegating regulatory authority to a highly autonomous
IRA. A third argument is that delegation to a highly specialized and skilled agency will enhance
regulatory quality and regulatory efficiency. Specialized agencies are expected to possess the expert‐
based knowledge that politicians and bureaucrats lack (Maggetti 2009). Fourth, normative
arguments on agencification in general refer to the ability of such autonomous public bodies to act
swiftly and forcefully on new (regulatory) challenges and to make quick and timely decisions.
Agencies are believed to implement policies in a more efficient way (OECD 2002). Although empirical
studies on the performance of agencies point at mixed evidence, other studies show that agencies
use more performance‐oriented management techniques and innovative behavior, compared to
public organizations without autonomy (see for an overview of studies on the performance of
agencies, Verhoest et al 2004 and 2007; Laegreid and Verhoest 2010). Lastly, using their autonomy,
agencies may develop their own strategies for further enhancing their regulatory role or for gaining
support from their environment (Bendor et al. 2001).
However, we assess this reasoning, which equates formal autonomy towards government on the one
hand and regulatory power on the other hand, to be incomplete in at least two ways. First, a high
formal autonomy of the IRA does not necessarily lead to a high de facto autonomy towards
government in actual regulatory practices, and hence not necessarily to a strong regulatory position
and power. Secondly, an IRA, which is formally autonomous towards government, finds itself
increasingly embedded in a multi‐actor and multi‐level constellation of interdependent actors, with
whom it shares responsibilities in regulating industries on a specific market and with whom it has to
interact to perform its responsibilities.
1.3. De facto autonomy towards government Formal autonomy, granted to IRA’s by provisions in their enacting legislation, does not tell the
complete story. IRA’s are strategic actors that seek to maximize their regulatory power. From that
perspective they might deviate from the initial political delegation and seek to redefine their
functions (Huber and Shipan 2002; Bendor, Glazer and Hammond 2001). Moreover, both the elected
politicians and the regulated industries have incentives to influence the functioning of the IRA and its
decision making processes, despite formal autonomy (Maggetti 2009). This leads the literature to
turn its attention away from the formal functioning to the actual functioning of regulatory bodies,
that is the need to go beyond the formal institutional framework of delegation to IRA’s to uncover
the processes and mechanisms whereby the roles and relationships of IRA’s develop and affect the
governance of markets (Coen and Thatcher 2005). The disjunction between formal autonomy and de
facto autonomy has been noted by several researchers (Wilks and Bartle 2002; Thatcher 2002;
Maggetti 2009; see for the same argument for the wider population of agencies, Verhoest et al.
2004; Verhoest et al. 2010). However, whereas studies explaining formal autonomy of IRA’s are
numerous, empirically based explanations of the disjuncture of formal and de facto autonomy are
much more scarce (see e.g. Laegreid et al. 2009). In his recent study, Maggetti (2009) theoretically
and empirically shows that a combination of an old age and the presence of many veto players lead
to a high de facto independence for IRA’s. Focusing on the wider population of (regulatory and non‐
regulatory) agencies, Verhoest et al. (2010) in their three‐country study find that the age of the
agency positively influenced the extent of policy autonomy, but had a negative effect on the extent
of managerial autonomy. Furthermore, they found that de facto/perceived managerial autonomy is
positively influenced by, among others, the agency size, the governance structure, and the extent of
self‐generated income. Also factors related to administrative culture, polity and administrative
capacity at the centre of government proved to matter (Verhoest et al. 2010). Because of the
potential divergence of formal and de facto autonomy, a high formal autonomy of an IRA will not
necessarily translate itself in a strong regulatory position and power.
1.4. Proliferation and interdependence within multi‐actor and multi‐level regulatory arrangements Another reason why a high formal independence of an IRA towards government does not necessarily
translates into a strong regulatory position and power is due to the proliferation and
interdependence of regulatory institutions in most sectors. While earlier country‐comparative
research on utility regulation focused solely on the sector specific regulatory institutions (Majone
1997b; Schneider 2001b), it is increasingly acknowledged that other institutions are involved, both at
the national level (especially the ministry and the competition authority) and at the European level,
and that there is interplay between them as well (Barbieri 2006b). Regulatory institutions interact
with other actors within a broad institutional environment, and relations in this environment are an
important aspect of research in the field, especially between IRA’s, ministries, and competition
authorities (Bölhoff 2005; Eberlein 2000a; Schneider 2001b; Wolf 2000). Regulatory regimes act thus
in dynamic institutional settings, and the power constellations vary across countries (Bölhoff 2005).
As a consequence authors acknowledge the evolving character of institutional arrangements and
seek to conceptualize it (Coen and Thatcher 2005).
The responsibilities in regulating industries is increasingly shared between ‐ and influenced by the
interplay of ‐ several public actors at different levels of government, leading to a multi‐actor and
multi‐level regulatory arrangement (Jordana and Levi‐Faur 2004). The proliferation and
interdependence can be seen along four main axes: horizontal, vertical, between sectors, and
between general competition and sector‐specific regulation.
(1) Horizontally, many IRA’s share regulatory responsibilities with other actors, like the
minister, the related ministerial department, but also advisory commissions, information
gathering bodies or even private regulators hold specific competencies and tasks and need to
interact with the IRA. Moreover, appeal courts constraint or influence the actions of IRA’s by
their decisions. Christensen and Laegreid (2006) point out that there are several ways in
which regulatory competencies can overlap between regulatory agencies, ministries,
legislative bodies, and courts, this in the context of a movement from a state‐centric
approach to a multi‐level perspective (see also Maggetti 2009: 40).
(2) Proliferation of regulatory actors also increasingly has a vertical or multi‐level component:
sector‐specific regulators are established on different governmental levels, be responsible for
different regulatory tasks, and increasingly interact with each other. Multi‐level regulatory
arrangements refer to arrangements that involve several levels of government, and in which
these levels of government and their affiliated public bodies (e.g. regulatory agencies)
interact, reinforce or restrict each other or collide in the process of regulation (Doern and
Johnson 2006). As private activity expands beyond the nation state, there is a pressure to
have a political debate and possibly regulate on an international and European level.
Increasingly, scholars in regulation and multi‐level governance target their research on the
extent to which the EU, and in particular, the Commission, influence the decision making of
national IRA’s, not only by drafting EU directives, but also by direct interaction during
decision making procedures and by indirect interaction through regulatory networks. These
networks of regulatory authorities c n either have a cross‐sector or sector‐specific nature,
and range from informal networks to highly formalized interaction platforms with explicit
assigned tasks and membership (Coen and Thatcher 2008; Eberlein and Grande 2005).
Certainly in the field of regulation, several scholars claim that the EU uses these forms of
interaction to bypass national governments, directly influencing national IRA’s in their
implementation tasks. Egeberg (2006) refers to the ‘double‐hattedness of national agencies
in the multi‐level Union administration’, to denote the emergence of the Commission as a
new executive centre delegating tasks towards the national IRA’s (Egeberg 2006; Coen and
Thatcher 2008). However, multi‐level regulatory arrangement may also have an intra‐state
component as modern nation states have been delegating tasks to lower levels of
government as well (Hooghe and Marks 2003). This is even more the case in federal states
where the regulation of economic sectors may be a shared responsibility between the
national and regional level of government, each establishing their own IRA’s.
(3) IRA’s increasingly need to interact and coordinate intensively with regulators in affiliated
economic sectors. As sector regulators focus each on a narrowly defined market, a need for
coordination becomes pertinent when enterprises start to develop activities that cut across
different markets (e.g. telecom companies offering media services). Moreover, when
regulating industries, IRA’s increasingly have to take into account actions from regulators on
other policy fields (like environmental and sustainable growth regulations).
(4) In order to avoid market failures, the regulation in liberalized markets comprises both ex
ante regulation by the sector regulators and ex post regulation by the general competition
authorities. As markets mature the role of the sector regulator changes, which means sector
regulators and general competition authorities need to interact in decision making
procedures (see e.g. the market analysis procedure in telecom) in ways of mutual advice or
co‐decision (Naert 2009).
Clearly, even IRA’s with a very high formal independence towards government increasingly find
themselves embedded within a larger constellation of regulatory actors with which they have to
coordinate intentions, decisions and actions. This interdependence might affect their regulatory
power and position in the broader regulatory arrangement and towards regulatees (Newman and
Posner 2005) negatively, not only because these regulatory actors share responsibilities, but also
because of the detrimental effects of complexity and intransparency of the regulatory arrangement.
Indeed, given the array of authorities with overlapping competencies (vertical and horizontal) the
regulatory structure gains complexity. This represents a risk for institutional conflicts to emerge
between regulators operating at different levels and national competition authorities. Drawing on
theories of public administration, Bauer advances that regulatory proliferation enhances
administrative costs via the emergence of infighting of power and competencies. The empirical
investigation on British and German telecommunications, electricity and railways sectors confirm
that “the more dispersed regulatory authority is, the higher coordination administrative costs
become” (Bauer 2005: 86).
1.5. Autonomy and regulatory power in multi‐actor and multi‐level regulatory arrangements A high formal autonomy of an IRA towards government will thus result in a strong regulatory position
and power, as perceived by relevant actors in the regulatory arrangement, on the condition (1) that
the IRA is also de facto autonomous from government in its regulatory actions and (2) that the
dependencies of other regulatory actors in taking regulatory decisions are limited. Figure 2 shows
this more elaborated theoretical model.
Formal organizational autonomy of the IRA towards
government
De facto/perceived organizational autonomy of
the IRA
Formal position and power of the IRA in the broader regulatory arrangement
Perceived position and power of the IRA in the broader regulatory arrangement
THEORY
Figure 2: a more refined model of autonomy and regulatory power As shown in figure 2, one can also hypothesize a feedback loop, meaning that a strong position and
power in the regulatory arrangement, as perceived by market actors, may enhance an increased
autonomy towards government over time. As Maggetti (2009:94) asserts, increased autonomy from
the principal allows the independent agency to “accumulate informational, material and symbolic
resources in the course of repeated interactions with regulatees”, but also with other regulators. As
IRA's gain more legitimacy from market actors and regulatory peers, it may pressure government for
more autonomy. Carpenter (2001) showed in his study that agencies’ autonomy is strongly
determined by its reputation, which in turn can be enhanced by coalition building and participation
in networks of peers. For example, the strong reputation of IRA’s within European networks may be
used as a strategic resource by the IRA in its interactions with government. Moreover, considering
national IRA’s as being double‐hattedness’, acting as agents for both the national governments and
the EU‐Commission (Egeberg 2006), these IRA’s may gain autonomy by “playing principals off against
each other”.
However, the reciprocal effect of formal autonomy of the IRA on the one hand and the regulatory
position and power of the IRA in the regulatory arrangement on the other hand can be assumed to
be different across market sectors, because of three factors:
(1) The time of liberalization and creation of the IRA: in very recently liberalized markets,
formal independent regulatory agencies have little time to build up legitimacy towards
regulatees and other regulatory actors. Moreover, as Maggetti (2009) shows, older IRA’s
tend to have higher de facto autonomy. Because of sector path dependencies and inertia of
institutions, recently established IRA's may find themselves to be controlled in the same
bureaucratic way as before their agencification (Maggetti 2009: 98). That implies that in case
of a recently established IRA in a recently liberalized market, a high formal autonomy
towards government does not necessarily translate in a strong position and power in the
regulatory arrangement.
(2) The complexity of the multi‐level regulatory arrangement in terms of number of involved
governmental levels: if the responsibility for regulating a market sector is shared among
more governmental levels, it will be harder for a national IRA to attain a strong regulatory
power and position.
(3) The presence or absence of a publicly owned incumbent: In case of absence of a public
incumbent, a high formal and de facto autonomy is needed less in order to gain a strong
position and power in the regulatory arrangement, while an IRA confronted with an
incumbent needs a high amount of autonomy in order to maintain its position in the
network. Therefore, the same amount of formal autonomy by two organizations can lead to
a lower power and position within the network for the organization that is confronted with
an incumbent, while the organization that doesn’t have to deal with the incumbent has a
higher position for the same amount of autonomy.
2. Central concepts In this section we aim to define the main concepts, which are used in this study. Furthermore, we
introduce the case studies, and provide a framework of the operationalization.
2.1. Definition of concepts To understand what actors are part of the regulatory arrangement, it is important to ask the question
what is regulation? The broadest possible definition of regulation can be found with Mitnick (1980:
30): “(Regulation is the) public administrative policing of a private activity with respect to a rule
prescribed in the public interest”. However, a more functional definition is needed to better
delineate the regulatory arrangement. Therefore, Hood’s definition is better suited (Hood et al.
2001), and differentiates between (1) goal formulation, rule‐making and standard setting; (2)
monitoring, information‐gathering, scrutiny, inspection, audits and evaluation; and (3) enforcement,
behaviour modification and the application of awards and sanctions. A separate category of licensing
can potentially be added as well, since making individual decisions on a specific case‐by‐case basis
can be labelled either as standard setting or as monitoring. Regulation can thus be defined as ‘a
specific form of governance that is operating through the promulgation of authoritative sets of rules,
while setting up mechanisms for monitoring, scrutinizing, and promoting compliance with these
rules’ (Maggetti 2009: 27). In our operationalization we exclude the elements of law making and
preparation of legal texts, but focus on the implementation and enforcement of regulations. A map
of the regulatory arrangement hence includes all public organizations in a sector that occupy
themselves with one or more of the aforementioned regulatory tasks.
Early research on utility regulation focused on the sector specific regulatory institutions (Majone
1997a; Schneider 2001a). However, it is clear that other institutions are involved, both at the
national level and at the European level, and that there is interplay between them (Barbieri 2006a).
Regulatory institutions interact with other actors within a broad institutional environment, and
relations in this environment are an important aspect (Bölhoff 2005; Eberlein 2000b; Schneider
2001a; Wolf 2000). Therefore, a regulatory arrangement consists of a constellation of public actors
(Jordana and Sancho 2004) involved in regulation of a sector. There are several levels on which these
actors can exist in relation to each other. Thus, the actors in the arrangement can be placed on the
four axes already distinguished before (Aubin et al., 2009):
(1) The horizontal axis: this includes all regulatory actors that operate on the same policy
level, though potentially in different hierarchical roles. A Minister and its Department for
example, or a sector regulator and the Minister
(2) The vertical axis: this includes all regulatory actors that operate on different policy levels.
EU organizations operate on a different level than federal organizations for example, and
likewise federal and regional organizations.
(3) Sector specific versus general competition axis: most regulatory actors will be engaged in
ex ante regulation, with ex post monitoring of course. General competition law on the
contrary can be considered ex post regulation, since it in general2 does not regulate the
influx of market activities, and concentrates on the outcome of them.
(4) The sector axis: most regulatory actors will perform the regulatory tasks specific to the
sector under scrutiny. However, it is possible that actors responsible for one or more
2 We state that this is the case in general because there are some activities in general competition law that should be considered ex ante regulation. Most notably this applies to potential mergers that have to be reported and approved by the relevant Competition Authority.
regulatory tasks in other sectors have an effect as well. It is necessary to take these other
actors into account, for whatever reason this interference exists3.
Organizational autonomy has been a subject for study in several streams of literature. A major part
of public organization research is focused on specialization and more specifically on semi‐
autonomous public service organizations (PSOs) or so‐called ‘agencies’: what, why, and how (see e.g.
Pollitt et al. 2005). Internationally, three groups of cumulative research efforts on agency autonomy
and control can be identified:
(1) The North‐American school on structural choice (see e.g. Moe 1990; Huber and Shipan
2002), which uses rational choice theories, formal analysis, and case studies.
(2) A cluster of political scientists and law scholars focusing on the emergence and
independence of regulatory agencies, using formal analysis, comparative mapping
techniques (e.g. Jordana and Levi‐Faur 2004), and to some extent surveys (see Gilardi 2008).
They use the rational choice and sociological institutional schools as theoretical frameworks.
(3) The COBRA network and the COST ISO601‐CRIPO network. This involves research teams in
23 countries, which are currently building a joint database with similar survey data from 15
countries and joint mapping efforts. These networks mainly use organization theories,
theories on control, and institutional theories to explain empirical observations.
2.2. Case studies In this paper we explore the autonomy and regulatory power of the national IRA in two case studies:
the regulation of the telecommunication market and the energy market in Belgium. Both cases refer
to liberalized utility sectors. Regulation of utilities is just one possible focus out of many, but
nevertheless a very interesting one because it deals with such issues as the provision of essential
services (Scott 2000), and the transition from the old governmental involvement in the sector to the
new style of regulating private and semi‐private companies (Coen and Héritier 2005; Coen 2008).
Most importantly, however, the utility sectors we are looking at already have a very complicated
structure, with submarkets in infrastructure, distribution, and commercialization, some of which still
are pure natural monopolies. When the companies were still state‐owned, this did not represent
many problems from an organizational/coordination point of view (Horn 1995), but with privatization
and the accompanying liberalization a certain amount of reregulation can now be observed to take
into account these new realities (Tenbücken 2006).
3 It could be the case that two sectors grow towards each other on the market, for example because the same companies are involved in offering both products/services. Provision of digital television and provision of telecom services come to mind here. Alternatively, it is possible that the public organizations in one sector are directly responsible for a specific set of decisions that greatly impact another economic sector. An example of this is the importance of decisions by spatial planning organizations on the energy sector; because of the necessity for permits for new production facilities, and potentially the planting of new infrastructure.
These two cases seem to differ in terms of the factors that we identified above as being intervening
factors: the time of liberalization and creation of the IRA; the number of governmental levels
involved; and the presence of a still dominant public incumbent. According to our model, an equal
level of formal autonomy of the IRA in both telecommunication and energy would nevertheless
relate to a stronger regulatory position and power of the IRA in telecommunications, because:
(1) The liberalization of telecommunication and the creation of the involved IRA happened
earlier in time, compared to energy
(2) Telecommunication regulation remained a federal competency, while energy regulation is
a shared competency between the Federal and Regional levels of government
(3) As a counterargument, the telecommunication market is still dominated by the public
incumbent, which is also for a large part still the owner of the network, whereas in the
energy market the main actor, Electrabel, is a former private monopolist with strong links to
local governments (but not directly to the national government). Electrabel also does no
longer own the network. Therefore, a similar level of autonomy would lead to a higher
position and power for CREG, since it is no longer confronted with the same dominant
incumbent.
2.3. Methodological framework We operationalize and measure the different variables by using methodologies based both on
researcher coding and document analysis, and based on using subjective, perceptual data gathering
strategies, as seen in figure 3. This figure integrates the previous figures 1 and 2.
Figure 3: the theoretical model and methodological triangulation
3. Research methodology In this chapter, we provide a more in‐depth view of the methodology we use to analyze both formal
and perceived autonomy of an agency towards government, and the formal and perceived position
and power of the agency in the broader regulatory arrangement.
3.1. Organizational autonomy towards government 3.1.1. Formal autonomy towards government The formal analysis uses the Gilardi index (Gilardi 2002). Starting from the research of Cukierman et
al (1992), the Gilardi index distinguishes between five dimensions that influence formal autonomy:
agency head status, management board members’ status, the general frame of the relationships with
the government and the parliament, financial and organizational autonomy, and the extent of
delegated regulatory competencies. The indicators that operationalize these dimensions can be
found in the case analysis. Each indicator is numerically coded on a scale of 0 (lowest level of
autonomy) to 1 (highest level of autonomy). Consequently, the results are aggregated in two steps,
first per dimension and then into a single index. Each dimension is given the same weight (Gilardi
2002).
Data collection is obtained by examining the enacting legislation and start‐up documents of the
agency.
3.1.2. Perceived/de facto autonomy towards government The subjective analysis is based on a survey used in COBRA research (COBRA 2010). The survey
encompassed more questions than only on perception of autonomy, but for the purpose of this
paper only the questions regarding financial management autonomy, personnel management
autonomy, and policy autonomy were used. It was possible to give scores to some questions, to
allow for a better comparison between sectors. For other questions this was not possible, but we
included them nevertheless to provide a context so that the answers to these are represented in a
text, to provide context to the scores.
Data gathering comes exclusively from the survey, presented to an upper level member of the
regulatory actor.
3.2. Regulatory power and position of the IRA in the regulatory arrangement 3.2.1. Formal power and position of the IRA in the broader regulatory arrangement The formal analysis is based on the method of regulatory coordination indices, first brought forward
in the paper by Mathieu et al. (2009). A distinction is made on two types of coordination between
regulatory agencies that have an impact on the decision making power and position of the agency in
a regulatory arrangement: coordination in decision making procedures and coordination through
structures (coordination platforms or functions). On the one hand, procedural interaction
encompasses interactions between the actors in the regulatory arrangement in legally defined
decision making procedures. Regulatory decision making procedures as defined in laws and
regulations define the sequential steps taken in order to produce a final decision. Structural
coordination on the other hand refers to interactions that are not handled in legally defined decision
making procedures, but are mainly occurring through coordinating structures, like platforms for
advice and consultation, or through information sharing systems. As regulatory actions of IRA’s needs
a legal basis, most coordination between regulatory actors is organized through involvement of the
different regulatory actors in legally defined decision making procedures (Aubin et al. 2009).
Therefore, in calculating the regulatory arrangement indices, we focus solely on procedural
coordination between actors in the regulatory arrangement.
Three indices are calculated, and each of them gives a better understanding of the position of the
different actors in the regulatory arrangement. The calculation method for each of these indices can
be found in annex 1 of this paper, as explained by Mathieu et al. (2009). However, the terminology
has been altered, to better cover what the indices are actually measuring.
(1) The procedural participation index gives an overall picture of the degree of involvement
of the different actors in the different decision making procedures. The index is low when
regulatory actors tend to not interact in the different procedures.. The index will be high
when there is a lot of involvement of all actors in the regulatory arrangement in each of the
regulatory decision making procedures. In that case, decision making procedures are more
open and require the participation of a greater number of actors for each procedure.
(2) The actor influence index measures the involvement of an actor in the decision making
procedures within a regulatory arrangement. It also takes into account the exact influence
the actor has in each of the decisions, by assigning a weight to the influence4. The higher the
index, the more influence an actor has on procedures.
(3) The influence arrangement index provides a way to evaluate the position of influence of a
specific actor, taking into account the influence of all the other actors within the overall
arrangement. The higher the index, the more dominating the actor is. It uses the actor
influence scores of all the actors in the regulatory constellation, and looks at the differences
between them.
4 A weight is given to the influence on the decision making procedure in the following way: 0: the actor is not involved 0.2: the actor is informed on a draft of a decision, and can voluntarily provide input
0.4: the actor has to legally be consulted, and has to provide an advice, but that advice is not binding. 0.6: the actor has the right to make proposals of its own, or has to provide a binding advice
0.8: the actor is a co‐decider 1: the actor is the main decider, with no other actors holding equal decision rights
The data gathered for this research method is based on an analysis of the different decision making
procedures, as defined in primary and secondary legislation, which occur in a regulatory
arrangement5. The data itself can be found in legal acts on both the Regional, Federal and European
levels. These documents describe both the involvement of the different regulatory actors in the
decision making procedures, and the extent of their formal influence in each of these.
3.2.2. Perceived power and position Perceived power and position is analyzed through the use of a social network analysis (or SNA as it
will be referred to from now on). This has been described as a toolbox for describing and measuring
relational configurations and their structural characteristics (Kenis and Schneider 1991: 44).Though
SNA is a methodological tool, it builds on previous theories that state that individuals and
organizations interact within networks. Government as well can, in this regard, be seen as a network
of actors involved in policy processes, where the different actors are interdependent and limit the
activity each one of them can engage in (Hanf&Sharpf 1978: 12). Hence, policy processes have been
the subject of SNA analyses in the past (Knoke et al. 1996; Krackhardt 1990), proving that a formal
SNA is very useful for empirical analysis of policy network structures (Ingold&Varone 2009). An SNA
points at several concepts and related quantified indices. In the research we make use of the
concepts of reputational power, degree and betweenness centrality, ego‐networks, and cliques.
(1) Reputational power shows how many actors in the network consider another actor as
very important (Ingold et al. 2010).
(2) Degree and betweenness centrality (Freeman 1979) both measure the empowerment of
an actor in a given network, in this case the regulatory arrangement (Ingold et al. 2010).
Degree centrality on the one hand is measured by looking at direct connections between
actors, and in networks were it is important to know the direction of the connection (e.g.: the
direction of the information flow), such as the ones used in the case studies, a difference is
made between in‐degrees and out‐degrees, where in‐degree relates to the amount of
information that is received by an actor, and out‐degree to the amount of information an
actor sends out. In this paper we focus solely on out‐degree relations of information and
5 Also, a territorial approach has to be taken, meaning that the multi‐level regulatory arrangement is described from the perspective of a private actor in one particular place. Arguments both for and against can be given for this. On the one hand, regulatory arrangements are always seen from a certain perspective. For example, a firm operating in the Flemish Region will consider the regulatory arrangement on the Flemish, Federal and European level, without worrying too much about coordination between Flanders and Wallonia. On the other hand, a regulatory arrangement also transcends the local boundaries, and coordination between territories is important for international firms. However, according to this logic the regulatory arrangement can potentially include all regulatory actors in Europe. Thus, the research takes the first approach. he data itself is found in legal acts on both the Regional, Federal and European levels. These documents describe both the involvement of the different regulatory actors in the procedures, and the importance of their involvement in each of them.
influence, as we want to measure the perceived regulatory position and power. Furthermore,
we look at general degree centrality for contacts. Betweenness centrality on the other hand
is done by measuring in how far the neighbours of an actor, which are the actors it has a
connection with, are connected themselves to other actors. This gives an indication on how
well established the actor is within the whole network: an actor with a high degree centrality
but low betweenness centrality is connected to a lot of outlying actors in the network, which
means its influence is lower than an actor that is connected to other well connected actors.
Both centrality indices are thus a tool to identify strategic positions and powerful actors in
the network (Christopoulos&Ingold 2009; Ingold&Varone 2009; Christopoulos 2006; Ingold et
al. 2010). Betweenness centrality is only measured for information flows, not influence.
(3) The position of an actor in a network can also be measured by analyzing different
relational types identifying subgroups of strongly related actors. The tools for this are
measurement from an ego‐centric perspective of an actor (Mitchell 1969: 8‐50) and its
participation in cliques (Wassermann&Faust 1994). We will disregard the participation in
cliques for this paper, and concentrate on the ego‐network of strong contacts. The degree
and betweenness centrality scores of this ego network give an indication of how well
embedded the actor is in the overall regulatory arrangement. If strongly embedded, the
actor’s reputation amongst the other organizations will be quite strong.
The acquisition of data for an SNA starts with a study of both primary and secondary legislation, this
on all relevant levels. This is necessary to be able to identify all the relevant actors in a regulatory
arrangement. This is complemented by looking at secondary sources, such as academic work and
information provided on the websites of the different actors. Interviews were also conducted with
companies active in both the liberalized and the regulated markets, and the question was posed
what public actors they came in contact with. Each of these actors was then contacted to answer a
survey. The survey posed questions concerning with what other actors information was shared and in
what direction, which actors influenced decision making, which actors were in turn influenced by
their own decisions, and where there were conflictual or co‐operational relationships. The actors
contacted and mentioned in the survey included both regulatory actors and private and public
companies.
4. Case studies 4.1. The energy market in Belgium 4.1.1. The regulatory arrangement in the energy market There are four independent sector regulators for energy in Belgium, one on the Federal level (CREG)
and three on the Regional levels (VREG, CWAPE, and Brugel). The task of regulating the sector is
primarily centred on these two levels, while the European Commission is mostly involved in the
production of rules, not in the everyday regulatory process. For the purpose of this paper, we will
look at the autonomy of the Federal sector regulator.
In regard to the regulated markets (transmission/transport and distribution), the transmission
submarket is regulated on the Federal level, together with the setting of the social tariffs, and tariffs
for the different distribution networks. Most other regulation in regard to the distribution network is
situated on the Regional level. Regulation concerning the liberalized markets (production and supply)
has also been divided amongst the two levels along the same lines. Both levels have competencies
regarding production settlement, and both can impose public service obligations. There are several
other private and public actors involved in the regulation of the energy sector, and figure 4 gives an
overview of these actors and the relations they have with each other.
Figure 4: regulatory arrangement in the Belgian energy sector (source: Aubin et al. 2009) The different actors can be put on four axes, as explained before:
(1) With respect to the vertical axis, most coordination happens through structural links, both
between the European level and the national level on the one hand, and the Federal and
Regional levels on the other hand. This is evidenced on the European level by national
representatives working together in comitology, and ERGEG and CEER bringing together the
national sector regulators. These actors are, however, not present in procedures at this point
in time. Between the Federal and Regional levels, there is a distinct delineation of
competencies. Structural coordination instruments exist in order to coordinate common
matters: the Federal level consults the Regions within ENOVER, a platform for the
coordination of policy‐making, and another platform for concertation, FORBEG, groups the
sector regulators on both the Federal and Regional levels and focuses on implementation.
(2) Horizontal coordination is discussed on the Federal and the Regional level. At the Federal
level, economic regulation is mostly shared between CREG and the Federal Minister of
Energy. This is mirrored on the Regional level, with a distribution of competencies mostly
between VREG and the Flemish Minister of the Energy. Within the Flemish Region, there is
further specialisation because other agencies also have competencies related to the energy
sector, for instance the Flemish Energy Agency (VEA) which is competent in energy savings.
This is less so on the Federal level. In general, one can conclude that there is a clearer
emphasis on procedural coordination on the horizontal level. Structural coordination is not
absent, for example, on the Flemish level there are regular meetings between several actors
under the auspices of the Department for the Environment, Nature and Energy.
(3) The axis regarding coordination between energy regulation and general competition law
is mostly procedural in nature again, with a clear delineation of competencies on the Federal
level: CREG is charged to collaborate with the Competition Authority to eliminate anti‐
competitive behaviour, namely by reporting this behaviour to the Federal Minister and the
Competition Council when observed in the course of the monitoring competencies of CREG.
Until recently, the exchange of confidential information from CREG to the Competition
Council was problematic, since such an exchange was not foreseen in the different Acts.
However, this situation has recently been remedied. There are no competencies regarding
general competition law on the Regional levels.
(4) Lastly, in terms of inter‐sectoral coordination there is a strong coordination, or even
integration, regarding regulation of natural gas on the one hand and electricity on the other
hand: the same organizations are responsible for regulating both sectors. There is no other
sector inherently related to the electricity and gas sectors. Nevertheless, energy issues have
impacts on closely related policy sectors, such as the environment. Usually, coordination is
organized structurally at the Cabinet and administration levels during the policy‐making
stage: the different ministries make comments on bills in inter‐cabinet and inter‐department
meetings.
4.1.2. Formal and perceived autonomy towards government of the Federal energy regulator CREG was established on January 10th 2000, under the auspices of the Act of April 29th 1999 for
electricity, and the Act of April 12th 1965 for gas. Its tasks and competencies are enumerated in these
laws as well. CREG’s internal structure consists of a General Council (in Dutch, “Algemene Raad”) and
a Management Board (in Dutch, “Directiecomité”). The General Council is made up of
representatives of various organizations. Its actual members are put forward by these organizations,
and then accepted by the Minister for Energy for a renewable period of three years. This includes
representatives from the government itself, who only have an advisory vote. The areas of
competency of the General Council were fundamentally altered in 2007. It no longer has the
authority to assess the way in which the Management Board carries out its tasks. It is required to put
forward opinions requested by the Management Board within forty days of the request. These
opinions carry some moral weight, but are non‐binding nevertheless.
The Management Board Chair and the members of the Management Board of CREG are elected for a
renewable term of six years on the basis of expertise: the members for the divisions they are leading,
the Chair for general management. Besides the Chair, there are three Directors in the Management
Board, each one leading a Directorate in CREG: the Directorate for the technical operation of the
electricity and natural gas markets, the Directorate for price and accounts monitoring, and the
administrative Directorate. The Management Board is a collegial body that makes joint decisions. It is
tasked with the operational management of the organization and can perform any actions necessary
to carry out the tasks of CREG.
4.1.2.1. Formal autonomy The Gilardi index of CREG yielded the following results:
Dimension
Indicator
Numerical status
Agency head status Score: 0.56
1.) Term of office? § over 8 years § 6 to 8 years § 5 years § 4 years § fixed term under 4 years or at discretion of appointer § no fixed term 2.) Who appoints the agency head? § management Board members § complex mix of executive and legislative branches § legislative branch § executive branch collectively § one or two ministers 3.) Who dismisses the agency head? § dismissal is impossible § only for reasons not related to policy § no specific provisions for dismissal § at the appointer’s discretion 4.) May the agency head hold other offices in government? § no § only with permission of the executive branch § no specific provisions 5.) Is the appointment renewable? § no § yes, once § yes, more than once 6.) Is independence a formal requirement for the appointment? § yes § no
1.00 0.80 0.60 0.40 0.20 0.00 1.00 0.75 0.50 0.25 0.00 1.00 0.67 0.33 0.00 1.00 0.50 0.00 1.00 0.50 0.00 1.00 0.00
Management board member’s status Score: 0.56
1.) Term of office? § over 8 years § 6 to 8 years § 5 years § 4 years § fixed term under 4 years or at discretion of appointer § no fixed term 2.) Who appoints the management board members? § the agency head § complex mix of executive and legislative branches § legislative branch § executive branch collectively § one or two ministers 3.) Who dismisses the management board members? § dismissal is impossible § only for reasons not related to policy § no specific provisions for dismissal § at the appointer’s discretion 4.) May management board members ad hold other offices in government? § no § only with permission of the executive branch § no specific provisions 5.) Is the appointment renewable? § no § yes, once § yes, more than once 6.) Is independence a formal requirement for the appointment? § yes § no
1.00 0.80 0.60 0.40 0.20 0.00 1.00 0.75 0.50 0.25 0.00 1.00 0.67 0.33 0.00 1.00 0.50 0.00 1.00 0.50 0.00 1.00 0.00
Relationship with government and parliament Score: 0.5
1.) Is the independence of the agency formally stated? § yes § no 2.) Which are the formal obligations of the agency vis‐à‐vis the government? § none § presentation of an annual report for information only § presentation of an annual report that must be approved § the agency is fully accountable 3.) Which are the formal obligations of the agency vis‐à‐vis the parliament? § none § presentation of an annual report for information only § presentation of an annual report that must be approved § the agency is fully accountable 4.) Who, other than the court, can overturn the agency’s decision where it has exclusive competency? § none § a specialized body § the government, with qualifications § the government, unconditionally
1.00 0.00 1.00 0.67 0.33 0.00 1.00 0.67 0.33 0.00 1.00 0.67 0.33 0.00
Financial and organizational autonomy Score: 0.33
1.) What is the source of the agency’s budget? § external funding § government and external funding § government
1.00 0.50 0.00
2.) How is the budget controlled? § by the agency § by the accounting office or court § by both the government and the agency § by the government 3.) Who decides on the agency’s internal organization? § the agency § both the agency and the government § the government 4.) Who is in charge of the agency’s personnel policy? § the agency § both the agency and the government § the government
1.00 0.67 0.33 0.00 1.00 0.50 0.00 1.00 0.50 0.00
Regulatory competencies Score: 0,25
1.) Who is competent for regulation in the sector? § the agency only § the agency and other independent authority § the agency and the parliament § the agency and the government § the agency has consultative competencies
1.00 0.75 0.50 0.25 0.00
Table 1: Gilardi index CREG The overall score for formal autonomy of CREG is thus 0.44. The score is dragged down primarily by
regulatory competencies of the agency, since the agency and government regulate the energy sector
together. Also financial and organizational autonomy scores lower. In this category, only personnel
management autonomy scores very high. Financially, though the budget is collected from the energy
consumers, it is set in a Royal Decree, so it is scored as coming from the government. The agency’s
structure is also decided on in legislation, which also leads to a lower score.
4.1.2.2. Perceived autonomy The questions concerning financial management autonomy, personnel management autonomy, and
policy autonomy can be found in tables 2 through 4. Other questions regarding these topics were
also included in the survey, but are not explicitly listed here. Nevertheless, the answers have been
used to provide additional context in the text.
Possible Possible, but
constricted by rules set on a higher level
Not possible
Score
Loan funds X 0.5
Set tariffs for services X 1
Participate/buy shares in private companies 0
Shift between budgets within the year X 1
Shift between budgets across years X 0
Final score (corrected) 0.5
Table 2 : perceived financial management autonomy Possible, both
general rules and decisions concerning individuals
Possible, but only on decisions concerning individuals. General rules set from above
Not possible
Score
Salaries X 1
Promotions X 1
Evaluation X 1
Hire X 1
Fire X 1
Final score (corrected) 1
Table 3: perceived personnel management autonomy Score
x Actor sets its own goals 1 Actor sets its own goals after consultation with Minister or administration 0.75 Actor negotiates with Minister or administration regarding goals 0.5 Goals are set by Minister or administration, after consultation with actor 0.25 Goals are set by Minister or administration 0 Table 4: formulation of goals as an indicator of policy autonomy The perception of financial autonomy by the agency scores 0.5, due to the ability to set tariffs and to
shift budgets within a year.
Personnel management autonomy of CREG is very high. It basically acts like a private company in its
hiring and firing of personnel. Employees are therefore not considered civil servants, so rules specific
to civil servants do not have to be taken into account.
CREG scores very high on the only question that can be weighed regarding policy autonomy. In
conjunction with other answers on policy autonomy questions, this should be mediated a bit: the
Minister certainly can have an impact on the decisions set by CREG, since it can suspend certain
decisions, and even acts as an appeal body for some specific decisions by CREG.
4.1.3. The regulatory power and position of the energy regulator in the regulatory arrangement The autonomy indicators revealed that there is some discrepancy between formal autonomy and
perceived/de facto autonomy. While the de facto autonomy is only mediocre, the perception of
autonomy is very high: only the financial management autonomy has not been given the highest
score. This means that if we only take into account formal autonomy, power and position within the
regulatory arrangement should not be very high either. Perceived autonomy somewhat contradicts
this.
Starting from our original three assumptions, we also anticipate that the power and position of CREG
is not very high: though it is not confronted with an incumbent as powerful as in the telecom sector,
it is a fairly young IRA, and it operates within a dual Federal structure, inhibiting an overall strong
position.
4.1.3.1. Formal power and position
A calculation of the procedural indices can be found in annex, as well as a list of all procedures, and
the influence the actors have regarding each procedure. Figure 5 gives an example of such a
procedure, and table 5 gives the scores of the level of influence for this procedure.
All procedures relate to the electricity market, rather than the complete energy market. For the
purpose of this paper, the electricity regulatory arrangement can be equated to the energy
regulatory arrangement as a whole, since the differences between the electricity and natural gas
markets are minute.
Figure 5: procedure of licensing suppliers through transmission networks
Actor Level of influence
Federal Cabinet (and Minister for Energy) 1 CREG 0.6
Federal Energy Administration 0.4
Table 5: procedure of licensing (electricity) suppliers supplying through the transmission network Tables 6 and 7 give the different indices that were calculated. The actor influence indices are shown
for the four actors that scored the highest.
Actor Actor Influence
CREG 0.31
Federal Cabinet (and Minister for Energy) 0.43
VREG 0.3
Regional Cabinet 0.23
Table 6: actor influence of most dominant actors
Index Result
Procedural participation index 0.15
Arrangement concentration index (with benchmark actor CREG) 0.21
Table 7: procedural participation index and arrangement concentration index As was expected, the actor influence indices give CREG higher values than other actors, including the
Regional sector regulator. The only exception is the actor influence index of the Federal Minister and
Cabinet. That is contrary to European regulation, which requires the sector regulator to play the most
important role in energy regulation. The Flemish Minister and Cabinet also play a significant role in
regulatory decision making, although less prominent than the Regional sector regulator. In line with
the dual federalist state structure in Belgium, most decision making procedures on the Federal level
only involve Federal actors, while all decision making procedures on the Regional level only involve
Regional actors. Nevertheless, the actor influence index of CREG is slightly higher than the actor
influence index of the Regional sector regulator.
The procedural participation index shows a very low score, indicating that a significant number of
decisions are made by a very limited set of actors. This may imply that the same actors are involved
every time, but it could also mean that there are always different actors making the decision. Again,
the differentiation of regulation on two different governmental levels, each with their own actors
involved in their own decision making procedures, plays a significant role in the low value of the
index. When only considering one governmental level, the index will most likely be higher.
Lastly, the influence arrangement index with CREG as the benchmark actor gives a low score,
meaning that CREG is not the dominant actor in the regulatory arrangement, even though its actor
influence is higher than most other actors. This suggests that overall there is not one dominant actor
in the regulatory arrangement. However, again the Federal/Regional split in electricity regulation
may explain this low value of the influence arrangement index, since CREG is largely absent in regard
to decision making procedures on the Regional level. It may also be interesting to see what the
influence arrangement index is of CREG and Federal Cabinet and Minister taken together. If this has a
substantially higher value, this may indicate that the two actors together have a dominant position in
the network. However, more research into the implications of taking actors together is necessary
first.
4.1.3.2. Perceived power and position Fifteen out of 24 actors active in the Belgian regulatory arrangement responded to the survey.
Figure 6: Ego network of strong contacts CREG
Social network analysis index Score
Reputational power 0.67 (position: 2nd
shared)
Out degree centrality influence 14 (mean 7, position
5th shared)
Out degree centrality information 20 (mean: 7.97,
position: 4th)
Betweenness centrality information network 40.33 (mean: 10.5,
position: 3th)
Degree centrality strong contacts 12 (mean: 4.75,
position: 1st shared)
Betweenness centrality strong contacts 89.17 (mean: 10.67,
position: 1st)
Table 8: Scores for several SNA indicators Reputational power of CREG is fairly high, with nine out of 15 respondents considering it a very
important actor, while the other actors rated it as important. This is the second highest reputational
power in the regulatory network: transmission and distribution companies occupy a shared first
space. However, the second position is shared as well, this with the Federal Minister for Energy and
the Regional sector regulator.
Both out degree centrality for influence and information score substantially above the mean, though
in both cases there are several actors, including private actors, with higher scores. In all cases, the
Federal Minister for Energy scores higher, and even scores highest for out degree information
centrality. The Regional sector regulator always scores close to the Federal sector regulator. CREG
does have a better position for out degree centrality in the information network. Scores are also
consistent: the same actors score high. A similar picture is shown for betweenness centrality of the
information network, though in this case CREG does score higher than the Minister, occupying a third
position after energy producers and energy suppliers. Lastly, CREG has the highest scores for both
degree and betweenness centrality on strong contacts.
From these results we can portray the perceived position and power of CREG within the regulatory
arrangement. Though fairly high, it is in fact the Federal Minister for Energy, presumably with the
Cabinet, that can disseminate information more, and the high betweenness centrality for
information of the Minister makes sure that there is a very broad dissemination. The Minister is also
more influential. Nevertheless, the contact ego network of CREG shows that the organization is
strongly embedded in the regulatory arrangement.
4.1.4. Autonomy and regulatory power in the energy market
Our expectations of the case study were the following: CREG is a fairly recent IRA, which operates in
a dual Federalist system. These elements suggest that the actor will not occupy a very high position
in the network. At the same time, the absence of an incumbent with strong ties to the national
government increases the likelihood of a dominant position, even in the absence of a high autonomy.
Initial findings showed that formal autonomy was not particularly high. We therefore expected that
perceived autonomy would reflect this, and that the power and position within the network would
not be very high either, both formal and perceived/de facto.
The findings do not completely reflect these initial assumptions. First, there is some discrepancy
between the formal and perceived autonomy of CREG, with perceived autonomy being substantially
higher than formal autonomy. Especially the personnel management policy and the policy autonomy
score very high, while the Gilardi index is moderate at best. Second, and contrary to this, the formal
position and power indices are more in line with our expectations, since they are not particularly
high. Third, the perceived/de facto power and position is again moderate, though CREG scores
consistently higher than the mean. Nevertheless, except for the ego network for contacts, there are
several other organizations with higher scores than CREG in each of the different results.
To summarize, we are confronted with mediocre formal autonomy of CREG, high perceived
autonomy, relatively low formal position and power, but higher than average perceived/de facto
position and power. We note that the formal autonomy follows the formal power and position, while
the perceived autonomy follows the perceived power and position.
4.2. The telecom market in Belgium 4.2.1. The regulatory arrangement in the telecom market The regulatory arrangement in the telecom sector in Belgium is organized around the European
Commission, and the Belgian Institute for Post and Telecommunications (BIPT), the sector regulator.
The European level is hereby also actively involved in some procedures. The Belgian Federal level is
mainly involved in the implementation of the EU framework on telecommunication. The Regions on
the other hand manage broadcasting, a sector with increasingly close ties to the telecommunication
sector. This means that the media regulators become involved (VRM in the Flemish Region, CSA in
the French Region, and Medienrat in the German Region). Other actors are involved as well, which
can be found in figure 7.
Figure 7: regulatory arrangement in the telecommunications sector in Belgium (source: Aubin et al. 2009) Again, we put the different actors on four axes:
(1) On the vertical axis, interactions between BIPT and the European Commission are mainly
hierarchical in nature. Other channels were, however, also developed either through
voluntary cooperation or joint decision (e.g. the non‐binding benchmarks of the ERG, which
have a high moral value, or the binding technical implementing decisions taken under the
comitology procedure with the national representatives). Thus, vertical coordination is
widespread with a mix of hierarchy and cooperation between the Commission and the sector
regulators. Coordination is often based on a hierarchical power of the Commission, although
hierarchy is very rarely visible in practice while cooperation seems to be the main mode of
relationships between the Commission and national authorities.
(2) On the horizontal axis, we only have to consider the Federal level. Here, law‐making in the
telecommunications is a competency of the Federal Cabinet and Parliament, while regulation
is delegated to BIPT. BIPT is also involved in drafting legislation, and can issue non‐binding
opinions on these drafts. The role of the Minister of Telecom in regulation is relatively small.
At the Regional level, the sector regulators are only involved in broadcasting (CSA, VRM, and
Medienrat).
(3) Between sector specific regulation and general competition law, the European and
horizontal Federal levels are taken into account. Regional actors have no competencies
regarding general competition issues. At the European level, a joint task force is organized
inside the European Commission between DG COMP and DG INFSO in order to coordinate
the telecom regulation with the general competition law. At the Federal level, the division of
tasks between the Competition Authority (consisting of the Federal DG Competition and the
Competition Council) and the telecom sector regulator is rather clear: sector‐specific
regulation is applied by BIPT, while the Competition Authority enforces the general
competition law. There is, however, an exception to this principle: the Competition Council is
competent for applying telecom regulation relative to interconnection in case of litigation.
But in practice this competency is almost never solicited by operators who prefer to have
conflicts dealt with by tribunals. The competency distribution is further complemented by a
dense cooperation between the two regulatory actors. In the market analysis, the
Competition Council is asked to give BIPT advices, most of the times non‐binding. BIPT is also
regularly solicited by the Competition Council as an experts when the latter is judging cases
related to the telecommunications sector.
(4) In regard to the inter‐sectoral axis, the media and broadband sectors are becoming more
and more closely related to the telecom sector, as a consequence of both technical
innovation and the desire of companies in both sectors to diversify their products. In
Belgium, the implementation of the EU telecom framework initially produced a conflict of
competency between BIPT and the media regulators, as well as between the Regional level
and the Federal level regarding law‐making. The Constitutional Court set an ultimatum and
obliged the Regional level and the Federal level to reach an agreement before taking further
measures. Long negotiations followed, ending in 2006 with an agreement on a cooperation
mechanism based on a network of regulators (for regulatory issues) and a network of
Ministers (for legislative issues). It is likely that in the future similar issues will surface, given
the increasing crossovers between the two sectors.
4.2.2. Formal and perceived autonomy towards government of the telecom regulator The Belgian Institute for post and telecommunications (BIPT) was created by the Act of March 21st
1991, which provided them with juridical personality. The organization has been deeply reformed by
the Act of January 17th 2003, which implements the framework. BIPT is competent for the postal
sector and the telecommunications sector, also called today electronic communications. BIPT is run
by a Board, which is composed of a chairman with a casting vote and three other Members. It has the
authority to undertake all actions that may be useful to the fulfilment of the actor’s powers. It
represents the Institute in court and towards third parties, and may enter into contracts on behalf of
the Institute. BIPT is further divided into nine services, and has two consultative committees, one for
telecommunications, and the other for postal services. Both committees give annual
recommendations about the actor’s activities. The committees have three main tasks: statutory
tasks, a general advisory task, and an international task (analysis of the international aspects of
telecom and postal services)6.
4.2.2.1. Formal autonomy Calculation of the Gilardi index for BIPT yielded the following results:
Dimension
Indicator
Numerical status
Agency head status Score: 0.56
1.) Term of office? § over 8 years § 6 to 8 years § 5 years § 4 years § fixed term under 4 years or at discretion of appointer § no fixed term 2.) Who appoints the agency head? § management Board members § complex mix of executive and legislative branches § legislative branch § executive branch collectively § one or two ministers 3.) Who dismisses the agency head? § dismissal is impossible § only for reasons not related to policy § no specific provisions for dismissal § at the appointer’s discretion 4.) May the agency head hold other offices in government? § no § only with permission of the executive branch § no specific provisions 5.) Is the appointment renewable? § no § yes, once § yes, more than once 6.) Is independence a formal requirement for the appointment? § yes § no
1.00 0.80 0.60 0.40 0.20 0.00 1.00 0.75 0.50 0.25 0.00 1.00 0.67 0.33 0.00 1.00 0.50 0.00 1.00 0.50 0.00 1.00 0.00
Management board member’s status Score: 0.56
1.) Term of office? § over 8 years § 6 to 8 years § 5 years § 4 years
1.00 0.80 0.60 0.40
6 www.bipt.be
§ fixed term under 4 years or at discretion of appointer § no fixed term 2.) Who appoints the management board members? § the agency head § complex mix of executive and legislative branches § legislative branch § executive branch collectively § one or two ministers 3.) Who dismisses the management board members? § dismissal is impossible § only for reasons not related to policy § no specific provisions for dismissal § at the appointer’s discretion 4.) May management board members ad hold other offices in government? § no § only with permission of the executive branch § no specific provisions 5.) Is the appointment renewable? § no § yes, once § yes, more than once 6.) Is independence a formal requirement for the appointment? § yes § no
0.20 0.00 1.00 0.75 0.50 0.25 0.00 1.00 0.67 0.33 0.00 1.00 0.50 0.00 1.00 0.50 0.00 1.00 0.00
Relationship with government and parliament Score: 0.42
1.) Is the independence of the agency formally stated? § yes § no 2.) Which are the formal obligations of the agency vis‐à‐vis the government? § none § presentation of an annual report for information only § presentation of an annual report that must be approved § the agency is fully accountable 3.) Which are the formal obligations of the agency vis‐à‐vis the parliament? § none § presentation of an annual report for information only § presentation of an annual report that must be approved § the agency is fully accountable 4.) Who, other than the court, can overturn the agency’s decision where it has exclusive competency? § none § a specialized body § the government, with qualifications § the government, unconditionally
1.00 0.00 1.00 0.67 0.33 0.00 1.00 0.67 0.33 0.00 1.00 0.67 0.33 0.00
Financial and organizational autonomy Score: 0.58
1.) What is the source of the agency’s budget? § external funding § government and external funding § government 2.) How is the budget controlled? § by the agency § by the accounting office or court § by both the government and the agency § by the government
1.00 0.50 0.00 1.00 0.67 0.33 0.00
3.) Who decides on the agency’s internal organization? § the agency § both the agency and the government § the government 4.) Who is in charge of the agency’s personnel policy? § the agency § both the agency and the government § the government
1.00 0.50 0.00 1.00 0.50 0.00
Regulatory competencies Score: 0.25
1.) Who is competent for regulation in the sector? § the agency only § the agency and other independent authority § the agency and the parliament § the agency and the government § the agency has consultative competencies
1.00 0.75 0.50 0.25 0.00
Table 9: Gilardi index BIPT The overall score for formal autonomy of BIPT is 0.47. The different scores are pretty consistent
around the 0.5 mark, except for regulatory competencies, because these are shared between the
independent agency and the government. Compared to the energy sector, it has a particularly high
score in its financial and organizational autonomy, stemming from its use of external funding for the
budget. Only the “regulatory competencies” dimension scores equally low, and that is the reason
why the final score is under the 0.5 mark.
4.2.2.2. Perceived autonomy
The same questions regarding financial management autonomy, personnel management autonomy,
and policy autonomy that were asked to a member of CREG were also asked to a member of BIPT.
The results of these are represented in tables 10 through 12.
Possible Possible, but
constricted by rules set on a higher level
Not possible
Score
Loan funds X 0
Set tariffs for services X 0
Participate/buy shares in private companies X 0.5
Shift between budgets within the year X 0.5
Shift between budgets across years X 0
Final score (corrected) 0.2
Table 10: perceived financial management autonomy
Possible, both
general rules and decisions concerning individuals
Possible, but only on decisions concerning individuals. General rules set from above
Not possible
Score
Salaries X 0
Promotions X 0.5
Evaluation X 0.5
Hire X 0.5
Fire X 0.5
Final score (corrected) 0.4
Table 11: perceived personnel management autonomy Score
X Actor sets its own goals 1
Actor sets its own goals after consultation with Minister or administration 0.75
Actor negotiates with Minister or administration regarding goals 0.5
Goals are set by Minister or administration, after consultation with actor 0.25
Goals are set by Minister or administration 0
Table 12: formulation of goals as indicator of perceived policy autonomy The perception of financial autonomy by the agency scores 0.2, a low score because of the limited
amount of ways the budget can be manipulated by the agency itself.
The perceived personnel management autonomy scores only 0.4 for BIPT. This is normal since
employees of BIPT are considered civil servants, not private employees (as is the case with CREG).
Therefore, the organization always has to take into account rules set for civil servants when
managing its personnel.
Much like with CREG, BIPT scores very high on the only question on policy autonomy that can be
weighed. Nevertheless, the other questions indicate that there is some influence by the Minister, but
not by the Department. Just like with CREG, the Minister has an appeal possibility, and can suspend
certain decisions.
4.2.3. The regulatory power and position of the telecom regulator in the regulatory arrangement The formal and perceived autonomy indicators for BIPT are more in line with each other than was
the case with CREG, though this time the perceived autonomy is somewhat on the low side, while
formal autonomy remains at an average level. This means that, based on the autonomy, we expect a
fairly low power and position of BIPT in the overall arrangement.
Contrary to this, starting from our original three assumptions, we expect BIPT to score somewhat
higher: it is already an older IRA, and it operates in a sector that is not divided between Regions. On
the other hand, the structure of BIPT has been changed considerably relatively recently, and in the
market there is a very strong incumbent with network ownership.
4.2.3.1. Formal power and position
A calculation of both indices can be found in annex, as well as a list of all procedures, and the
influence the actors have regarding each procedure. Figure 8 shows a graphical representation of
such a procedure, namely the market analysis procedure, while table 13 shows the influence each of
the actors involved has.
Figure 8: Market analysis procedure
Actor Level of influence on
definition of market Level of influence on
analysis market Level of influence on choice of remedies
European Commission 1 0.8 0.4
BIPT 0.6 0.8 0.8
Conference of Regulators for Communications (CRC)
0.6 0.6 0.6
Competition Council 0.4 0.4 0.6
ERG 0.2 0.2 0.4 Cocom 0 0 0.4
Private operators 0.4 0.4 0.4
Table 13: Market analysis procedure
Actor Actor influence
BIPT 0.63
European Commission 0.47
CRC 0.36
Competition Council 0.28
Table 14: actor influence index of most dominant actors
Index Result
Procedural participation index 0.4
Arrangement concentration index (with BIPT as the benchmark actor) 0.42
Table 15: procedural participation index and arrangement concentration index The actor influence indices reveal a high influence for BIPT compared to the other actors, this in
contrast to the more mediocre autonomy scores. The Federal Minister is, however, completely
absent from the top four of the actors with the highest actor influence scores (see table 14). The high
score of the European Commission is equally noteworthy. Since the implementation of European
policies is a responsibility of individual Member States, intuitively one would have expected the
relevant Federal Minister to have a higher score. The reason behind the high score of the European
Commission must be found in the market analysis procedure, which is very important within the
whole regulatory framework. Specifically in this procedure the European Commission plays and
important role, with a veto right and notification procedures.
The procedural participation index is rather moderate, so for each procedure only a limited amount
of actors is involved. However, at the same time, the influence arrangement index with BIPT as the
benchmark actor is also at a moderate level, though higher than any other actor (see table 15),
indicating BIPT does not have a dominant position, and in fact no actor in the regulatory
arrangement does. This index value shows that multiple procedures involve different dominant
decision making actors or that several actors have equal importance within procedures. Most
probably, the rather high amount of procedures in which both BIPT and the European Commission
play a significant role might explain why the influence arrangement index is quite moderate.
4.2.3.1. Perceived power and position Sixteen out of 34 actors active in the Belgian regulatory arrangement for telecom responded to the
survey.
Figure 9: Ego network of strong contacts BIPT (Ingold et al. 2010)
Social network analysis index Score
Reputational power 0.94 (position: 2nd)
Out degree centrality influence 9 (mean: 6.66,
position: 11th)
Out degree centrality information 11 (mean: 6.98,
position 11th shared)
Betweenness centrality information network Not calculated
Degree centrality strong contacts Not calculated
Betweenness centrality strong contacts 23.3 (mean: 6.9,
position: 4th)
Table 16: Scores of BIPT for several SNA indicators (Ingold et al. 2010) Reputational power of BIPT is very high, with all except one respondent rating it as very important,
and the last respondent rating it as important. However, the European Commission was still rated
even more important, while the incumbent, Belgacom, occupied a very similar position as BIPT.
Both out degree centrality for influence and information score above the mean, but the ranking of
BIPT is not very high. Several actors score higher, including consistently the Federal Cabinet, the
incumbent, and the Competition Council. Strangely, the European Commission scores either the
same or lower than BIPT in these cases. The betweenness centrality for strong contacts on the other
hand shows a higher score, and a higher ranking. This indicates that BIPT is nevertheless quite
strongly embedded into the regulatory network.
From these results we can portray the perceived position and power of CREG within the regulatory
arrangement. This is fairly low, with other private actors, public Federal actors, and the European
Commission scoring either similarly or substantially higher than BIPT.
4.2.4. Autonomy and regulatory power in the telecom market Our expectations of the case study were the following: BIPT is an established independent regulatory
agency, operating largely on a single level, though there is some feedback to the European level. It is
therefore assumed it will have a fairly high position and power in the regulatory arrangement. At the
same time, there is an incumbent with ownership of the (fixed) network, meaning that for a similar
level of autonomy, relative power and position will be lower. Moreover, initial findings showed that
formal autonomy was neither very high nor very low. This would mean that perceived autonomy,
and formal and perceived power and positions would be average too.
Perceived/de facto autonomy is consistent with the formal autonomy, but is generally even lower:
only policy autonomy received a high score. The formal position and power indices are even more in
line with the formal autonomy. Lastly, the perceived/de facto power and position scores are
generally low, again in line with the perceived autonomy, more than with the formal autonomy or
formal position and power. Some caution is warranted regarding the findings of the SNA analysis.
The response rate of the survey was lower than ideal, so that some of the results may be
compromised. Though we believe it is still useful to analyze the date, we would nevertheless be
prudent with their interpretation.
4.3. Cross‐case analysis Table 17 provides an overview of the different results obtained for CREG and BIPT. CREG BIPT
Agency head status 0.56 0.56
Management board members status 0.56 0.56
Relationship with government and Parliament 0.5 0.42
Financial and organizational autonomy 0.33 0.58
Regulatory competencies 0.25 0.25
Gilardi index
Overall 0.44 0.47
Financial management autonomy 0.5 0.2
Personnel management autonomy 1 0.4
COBRA survey
Policy autonomy 1 1
Actor influence index independent agency 0.31 0.4 Procedural indices
Influence arrangement index with independent agency as the benchmark actor
0.21 0.42
Reputational power 0.67 (position: 2nd shared)
0.94 (position: 2nd)Social network analysis
Out degree centrality influence 14 (mean 7, position 5th shared)
9 (mean: 6.66, position: 11th)
Out degree centrality information 20 (mean: 7.97, position: 4th)
11 (mean: 6.98, position 11th shared)
Betweenness centrality information 40.33 (mean: 10.5, position: 3rd)
Not calculated
Degree centrality strong contacts 12 (mean: 4.75, position: 1st shared)
Not calculated
Betweenness centrality strong contacts 89.17 (mean: 10.67, position: 1st)
23.3 (mean: 6.9,
position: 4th)
Table 17: cross sector results The different results have already been discussed separately, but by putting them next to each other,
some additional findings can be highlighted. In general, BIPT scores better in formal autonomy and
formal position and power in the regulatory network than CREG. At the same time, the perceived
autonomy and the perceived power and position is higher for CREG. This means that formal
autonomy and formal position and power seem to be linked to each other, as well as perceived
autonomy and perceived position and power7.
In regard to our three original assumptions about the differences between levels, the following
conclusion can be made. Given the fact that formal autonomy of BIPT is slightly higher than the
formal autonomy of CREG, we would have expected a higher formal and perceived position and
power of BIPT as compared to CREG, since the former is an older IRA, and telecom regulation is
mostly done at the Federal level. At the same time, the telecom market has an incumbent with
network ownership, while the energy sector does not, and this can negatively impact the perceived
power and position of BIPT. It seems that primarily this last distinction has had the greatest effect,
since the perceived power and position of CREG is substantially lower than that of BIPT. The formal
power and position does confirm the first two arguments, since for a relatively similar formal
autonomy, the formal position of BIPT is higher.
7 Formal autonomy CREG< formal autonomy BIPT Formal power and position CREG < Formal power and position BIPT Perceived autonomy CREG > perceived autonomy BIPT Perceived power and position CREG > perceived power and position BIPT
5. Conclusion In this paper we have tried to theoretically construct a framework to tie the autonomy of an IRA in
regard to the government with its power and position in the broader regulatory arrangement. Both
formal and perceived/de facto autonomy was taken into account, as well as formal and de
facto/perceived position. A number of measurements were proposed to quantify these four different
elements, to allow for a comparison between sectors. The assumption we made was that for equal
formal autonomy, three other attributes of the IRA and regulatory arrangement impact the position
and power of the organization: age, single level or multilevel arrangement, and presence of an
incumbent.
These measurements and assumptions were then tested on two case studies: the energy sector and
the telecom sector in Belgium. The findings have not been completely in line with the original
assumptions. On the one hand, the link between formal autonomy and perceived autonomy is only
partially confirmed. Also the link between formal position in the regulatory arrangement and
perceived position has been brought in question. On the other hand, a link can be observed between
formal autonomy and formal position in a network, and perceived autonomy and perceived position
in the network. In regard to the three elements that impact position and power given equal levels of
autonomy, the cases brought forward that formal autonomy and formal position seem to be
impacted most by age and multilevel regulatory arrangements, while the presence of an incumbent
impacts the perceived autonomy and power. Further research is necessary to make sure there are no
other factors that have a substantial impact on these links, because in the current case studies this
has not been tested yet.
Reference List:
Aubin, D, Verhoest, K., Mathieu, E. and Matthys, J. (2009), “Limited Fragmentation through Coordination in the Regulation of the Energy and Telecommunications Sector in Belgium.”, Potsdam: ECPR General Conference Barbieri, D. (2006a), “Evaluating Delegated Modes of Governance Inside EU: towards a Model of Communitarian Agency?” In: C. Reichard, R. Mussari, R. and Sören, K. (Eds.), The Governance of Services of General Interest between State, Market and Society, Berlin: Wissenschaftlicher Verlag Barbieri, D. (2006b), “Transnational Networks Meet National Hierarchies: the Case of Italian Competition and Environment Agencies.” In: Egeberg M. (Ed.), Multilevel Union Administration, Hampshire: Palgrave Macmillan Bauer, M. W. (2005), “Administrative Costs of Reforming Utilities.” In: Coen, D. and Héritier, A. (Eds.), Refining Regulatory Regimes: Utilities in Europe, Cheltenham: Edward Elgar Bendor, J, Glazer, A. and Hammond, T. (2001), “Theories of Delegation.”, Annual Review of Political Science, 4: 235‐269 Bölhoff, D. (2005), “Developments in Regulatory Regimes: Comparison of Telecommunications, Energy and Rail.” In: Coen, D. and Héritier, A. (Eds.), Refining regulatory regimes: Utilities in Europe, Cheltenham: Edward Elgar Carpenter, D. (2001), The Forging of Bureaucratic Autonomy: Reputations, Networks and Policy Innovation in Executive Agencies, 1862‐1928, Princeton: Princeton University Press Christensen, T. and Laegreid, P. (2006), Autonomy and Regulation: Coping with Agencies in the Modern State, Cheltenham: Edward Elgar Christopoulos, D. (2006), “Relational attributes of Political Entrepreneurs: A Network Perspective.”, Journal of European Public Policy, 13 (5): 757 ‐ 778 Christopoulos, D. and Ingold, K. (2009), “Brokerage and Special Agency.”, London, UK Social Network Conference COBRA (2010), Common Data in the COBRA‐Research: an Outline, online: http://soc.kuleuven.be/io/cost/survey/surv_core.pdf Coen, D. (2008), EU Lobbying: Theoretical and Empirical Developments, London: Routledge Coen, D. and Héritier, A. (2005), Refining Regulatory Regimes: Utilities in Europe, Cheltenham: Edward Elgar Coen, D. and Thatcher, M. (2000), “Introduction: Reform of Utility Regulation in the EU.”, Current Politics and Economics of Europe, 9 (4): 1‐10 Coen, D. and Thatcher, M. (2005), “The New Governance of Markets and Non‐Majoritarian Regulators.”, Governance, 18 (3): 329‐346
Coen D. and Thatcher, M. (2008) “European Regulatory Networks: Reshaping Regulatory Space”, NewGov Policy Papers, 30: 1 Cukierman, A., Webb, S. B. and Neyapti, B. (1992) “Measuring the Independence of Central Banks and its Effects on Policy Outcomes.”, World Bank Economic Review, 6(3): 353–98 Doern, G. B. and Johnson, R. (2006), Rules, Rules, Rules, Rules: Multi‐level regulatory Governance, Toronto: University of Toronto Press Eberlein, B. (2000a), “Institutional Change and Continuity in German Infrastructure Management: the Case of Electricity.”, German Politics, 9 (3): 81‐104 Eberlein, B. (2000b), “Regulierung und die Konstitution von Märkten in Europa.“ In: Czada, R. and Lütz, S. (Eds.), Die Politische Konstitution von Märkten, Wiesbaden: Westdeutcher Verlag Eberlein, B. and Grande, E. (2005), “Beyond Delegation: Trans‐national Regulatory Regimes and the EU Regulatory State.”, Journal of European Public Policy, 12 (1): 89‐112 Egeberg, M. (2006), Multilevel Union Administration. The Transformation of Executive Politics in Europe, Hampshire: Palgrave Macmillan Eyre, S. and Lodge, M. (2000), “National Tunes and a European Melody? Competition Law Reform in the UK and Germany.”, Journal of European Public Policy, 7 (1): 63‐79 Freeman, L.C. (1979), “Centrality in Social Networks: Conceptual Clarification”, Social Networks, 1: 215‐239. Gilardi, F. (2002), “Policy Credibility and Delegation to Independent Regulatory Agencies: a Comparative Empirical Analysis.” Journal of European Public Policy, 9 (6): 873‐893. Gilardi, F. (2004), “Institutional Change in Regulatory Policies: Regulation through Independent Agencies and the Three New Institutionalisms.” In: Jordana, J. & D. Levi‐Faur, D. (Eds.), The Politics of Regulation: Institutions and Regulatory Reforms for the Age of Governance, Cheltenham: Edward Elgar Gilardi, F. (2008), Delegation in the Regulatory State: Independent Regulatory Agencies in Western Europe, Cheltenham: Edward Elgar Hanf, K. and Sharpf, F. W. (1978), Interorganisational Policy Making: Limits to Coordination and Central Control, London: Sage Hood, C., Rothstein, H. and Baldwin, R. (2001), The Government of Risk: Understanding Risk Regulation Regimes, Oxford: Oxford University Press Horn, M. J. (1995), The Political Economy of Public Administration: Institutional Choice in the Public Sector, Cambridge: Cambridge University Press Huber, J. D. and Shipan, C. R. (2002), Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy, Cambridge: Cambridge University Press Ingold, K. and Varone, F. (2009), “Bringing Policy Brokers Back in: Evidence from the Swiss Climate Policy, Potsdam, ECPR General Conference, Potsdam
Ingold, K., Varone, F., Aubin, D., Mathieu, E., Verhoest, K. and Matthys, J. (2010), “Governance Networks in Telecoms Regulation: how to Compare the Independence of the Sector‐Specific Regulator in Belgium and Switzerland?”, Münster, ECPR Joint Sessions of Workshops Jordana, J. and Levi‐Faur, D. (2004), The Politics of Regulation: Institutions and Regulatory Reforms for the Age of Governance, Cheltenham: Edward Elgar Jordana, J. and Sancho, D. (2004), “Regulatory Design, Institutional Constellations and the Study of the Regulatory State.” In: Jordana, J. and Levi‐Faur, D. (Eds.), The Politics of Regulation: Institutions and Regulatory Reforms for the Age of Governance, Cheltenham: Edward Elgar Jordana, J., D. Levi‐Faur, and X. Fernandez. (2007) "The Global Diffusion of Regulatory Agencies: Institutional Emulation and the Restructuring of Modern Bureaucracy.", Pisa: ECPR General Conference Kenis, P.N., & Schneider, V. (1991), “Policy Networks and Policy Analysis: Scrutinizing a New Analytical Toolbox” In Marin, B. and Mayntz, R. (Eds.), Policy networks: Empirical Evidence and Theoretical Considerations, Boulder: Westview Press Knoke, D. (1990), Political Networks: The Structural Perspective, Cambridge: Cambridge University Press Knoke, D., Pappi, F.U., Broadbent, J. and Tsujinaka, Y. (1996), Comparing Policy Networks: Labour Politics in the US, Germany and Japan, Cambridge: Cambridge University Press Krackhardt, D. (1990), “Assessing the Political Landscape: Structure, Cognition, and Power in Organisations.” Administrative Science Quarterly, 35: 342‐369 Laegreid, P., Roness, P. and Verhoest, K. (2009), “Explaining Innovative Culture and Behaviour of State Agencies.”, Malta: EGPA Annual Conference Laegreid, P. and Verhoest, K. (2010), "Reforming Public Sector Organizations" In: Laegreid, P. and Verhoest, K. (Eds.), Governance of Public Sector Organizations: Proliferation, Autonomy and Performance, Hampshire: Palgrave Macmillan Levi‐Faur, D. (1999), “Governing Dutch Telecommunications Reform: State‐business Interaction in the Transformation of the National Policy Regimes to European Embedded Policy Regimes.”, Journal of European Public Policy, 6 (1): 57‐78 Levy, B. & Spiller, P. T. (1996), Regulations, Institutions and Commitment: Comparative Studies of Telecommunications, Cambridge: Cambridge University Press. Maggetti, M. (2009), “Delegated Authority: Legitimizing Independent Regulatory Agencies.” In: Blühdorn, I. (Ed.), In Search of Legitimacy. Policy Making in Europe and the Challenge of Complexity, Leverkusen: Barbara Budrich Verlag Majone, G. (1994), “The Rise of the Regulatory State in Europe.”, West European Politics, 17 (3): 77‐101 Majone, G. (1997a), “From the Positive to the Regulatory State: Causes and Consequences of Changes in the Mode of Governance.” Journal of Public Policy, 17 (2): 139‐167.
Majone, G. (1997b). “The New European Agencies: Regulation by Information.” Journal of European Public Policy, 4 (2): 262‐275. Majone, G. (2001), “Two Logics of Delegation: Agency and Fiduciary Relations in EU Governance.” European Union Politics, 2 (1): 103‐122. Mathieu, E., Aubin, D, Verhoest, K and Matthys, J. (2009), “Assessing the Integration of Regulatory Arrangements: the Energy and Telecommunications Sectors in Belgium”, Brussels: Annual Conference on Competition and Regulation in Network Industries Mitchell, J.C. (1969), “The Concept and Use of Social Networks” In J.C. Mitchell, J. C. (Ed.), Social Networks in Urban Situations, Manchester: Manchester University Press Mitnick, B. (1980), The Political Economy of Regulation: Creating, Designing and Removing Regulatory Forms, New York: Columbia University Press. Moe, T.M. (1990), “Political Institutions: The Neglected Side of The Story”, Journal of Law, Economics and Organization, 6: 213‐253 Naert, F. (2009), “Competition Authorities and Regulators in Belgium: Hierarchy versus Cooperation”, Competition and Regulation in Network Industries, 10 (2): 139‐158 Newman, A. & Posner, E. (2005), “International Interdependence and Regulatory Power.”, Washington DC: Annual American Political Science Association Conference OECD (2002) Distributed Public Governance. Agencies, Authorities and Other Autonomous Bodies, Paris: OECD Pollitt, C., Caulfield, J., Smullen, A. and Talbot, C. (2001), “Agency Fever? Analysis of an International Fashion”, Journal of Comparative Policy Analysis, 3: 271‐290 Pollitt, C. and Talbot, C., Unbundled Government. A Critical Analysis of the Global Trend to Agencies, Quangos and Contractualisation, London: Routledge Pollitt, C., Caulfield, J., Smullen, A. and Talbot, C. (2005), How Governments Do Things Through Semi‐Autonomous Organizations, Hampshire: Palgrave MacMillan Schneider, J. P. (2001a), „Verwaltungskontrollen Und Kontrollmassstäbe in Komplexen Verwaltungsstrukturen“. In: Schmidt‐Assmann E. and Hoffmann‐Riem, W. (Eds.), Verwaltungskontrollen, Baden‐Baden: Nomos Schneider, V. (2001b), “Institutional Reform in Telecommunications: the European Union in Transnational Policy Diffusion.” In: Green‐Cowles, M., Caporaso, J. A., and Risse, T. (Eds.), Transforming Europe: Europeanization and Domestic Change, Ithaca: Cornell University Press Scott, C. (2000), “Services of General Interest in EC Law: Matching Values to Regulatory Technique in the Public and Privatised Sectors.”, European Law Journal, 6 (4): 310‐325 Tenbücken, M. (2006), The Regulation of Network Infrastructures in the New European Union: Regime Similarity after Two Decades of Reform?, Konstanz: Konstanz University
Thatcher, M. (2001), “The Commission and National Governments as Partners: EC Regulatory Expansion in Telecommunications 1979‐2000”, European Public Policy, 8 (4): 558‐584 Thatcher, M. (2002) “Regulation after Delegation: Independent Regulatory Agencies in Europe”, Journal of European Public Policy 9(6): 954–72 Thatcher, M. (2003). Independent Regulatory Agencies in Europe: Delegation and its Consequences, London: London School of Economics Thatcher, M. and Stone Sweet, A. (2002), “Theory and Practice of Delegation to Non‐Majoritarian Institutions”, West European Politics, 25 (1): 1‐22 Verhoest, K., Peters, G., Bouckaert and G., Verschuere, B. (2004), "The Study of Organisational Autonomy: a Conceptual Review", Public Administration and Development, 24(2): 101‐118 Verhoest, K., Verschuere, B. and Bouckaert G. (2007), "Pressure, legitimacy and innovative behaviour by public organisations", Governance, 20(3): 469‐498 Wasserman, S. and Faust, K. (1994), Social Network Analysis: Method and Applications. Cambridge: Cambridge University Press. Wilks, S. (2001), “The Politics of Regulation: a Triumph of Administration?” In: Vass, P.(Ed.) Regulation Regulatory Review Millenium Edition, Bath: Centre for the Study of Regulated Industries. Wolf, D. (2000), “Institutional Issues of Telecoms Regulation.” In: Ehlermann, C. D. and Gosling, L. (Eds.), European Competition Law Annual 1998: Regulating Communications Markets, Oxford: Hart Yesilkagit, K. and Christensen, J. G. (2010), “Institutional Design and Formal Autonomy: Political versus Historical and Cultural Explanations”, Journal of Public Administration Research and Theory, 20 (1): 53‐74
ANNEX Calculation of the procedural participation index
To calculate the procedural participation index, all decision making procedures, and the number of
actors per decision making procedure first have to be listed. The number of actors (a) per procedure
(d) have to be added to each other, which produces a rough idea of coordination: the higher the
number, the more actors are involved in each procedure, hence the more procedural participation
takes place in the regulatory arrangement. This is called the absolute procedural participation index
(PC(abs)).
However, comparisons are not possible this way, since both the total number of actors and the
amount of procedures can differ in a regulatory arrangement. It is therefore necessary to transform
this number to a range between zero and one, resulting in the relative procedural participation index
(PC(rel)). This is done first by subtracting from this amount the total number of actors in the
regulatory arrangement, making it possible for the index to have a measurement of zero, i.e. if all
actors decisions in the regulatory arrangement are made by only one actor, for each decision a
different one. The obtained amount is then divided by the maximum value the absolute index can
take, i.e. when all procedures have all actors involved, minus the total number of actors. The latter is
a correction necessary because otherwise it would be possible for the number to be greater than
one. In this case, a value of zero shows no procedural participation, while a value of one shows full
procedural participation. In mathematical terms, this would show as:
Equation 1: Formula for calculating the absolute value of the procedural participation index
Equation 2: Formula for calculating the relative value of the procedural participation index
Calculation of the actor influence index
To calculate the actor influence index, the number of procedures the actor is involved in has to be
known, as well as the exact influence (P) it has on the final decision per procedure. A weight is given
to the influence on the decision in the following way:
0: the actor is not involved
0.2: the actor is informed on a draft of a decision, and can voluntarily provide input
0.4: the actor has to legally be consulted, and has to provide an advice, but that advice is
not binding.
0.6: the actor has the right to make proposals of its own, or has to provide a binding
advice
0.8: the actor is a co‐decider
1: the actor is the main decider, with no other actors holding equal decision rights
As with the procedural participation index, an absolute actor influence index (C(abs)Aj) and a relative
actor influence index (C(rel)Aj) can be built, the latter accounting for differences in number of actors
and number of procedures, bringing the value of the index between 0 and 1. This is done by dividing
the absolute actor influence index by the number of procedures, or more specifically the maximum
actor influence an actor can have in a given regulatory arrangement (the actor is given an influence
value of 1 in each procedure). Mathematically, this looks like:
Equation 3: Formula to calculate the absolute centrality of the Actor j
Equation 4: Formula to calculate the relative centrality of the Actor j
Calculation of the arrangement concentration index
The arrangement concentration index is derived from the actor influence index, this by looking at the
difference between the actor influence score of the most central actor (CAmax), and the actor
influence scores of each of the other actors (CAj), adding all the differences. This is the absolute
arrangement concentration index (C(abs)). The relative arrangement concentration index, similar to
the relative actor influence index, can be found by dividing the absolute index with the maximum
number the absolute index can have, so if all the other actors have an actor influence index of 0, and
the most central actor has an actor influence index of 1. This comes out to the number of actors
minus 1, or mathematically:
Equation 5: Formula to calculate the absolute arrangement coordination index of the regulatory
arrangement
Equation 6: Formula to calculate the relative arrangement coordination index of the regulatory
arrangement
Electricity procedures
Natural gas procedures
Telecommunication procedures
Social network analysis telecom
Information network degree centrality:
1 2 3 4 OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 22 Belgacom / Proximus 29.000 28.000 63.043 60.870 25 KPN Group 25.000 10.000 54.348 21.739 2 Federal Government 24.000 25.000 52.174 54.348 16 CSA 21.000 29.000 45.652 63.043 24 Telenet 17.000 20.000 36.957 43.478 13 VRM 17.000 17.000 36.957 36.957 5 Competition Council 17.000 18.000 36.957 39.130 29 Beltug 15.000 8.000 32.609 17.391 28 Test-achat / Test aankoop 15.000 13.000 32.609 28.261 6 FME CS 13.000 21.000 28.261 45.652 1 IBPT 11.000 11.000 23.913 23.913 4 FME TC 11.000 22.000 23.913 47.826 31 EC 10.000 10.000 21.739 21.739 27 Alt. Oper. 9.000 7.000 19.565 15.217 23 Mobistar 8.000 6.000 17.391 13.043 26 Voo 7.000 4.000 15.217 8.696 3 Federal Parlament 6.000 9.000 13.043 19.565 33 EP 6.000 5.000 13.043 10.870 9 BCA 5.000 5.000 10.870 10.870 10 CS 5.000 7.000 10.870 15.217 34 ERG 5.000 3.000 10.870 6.522 8 Belg. Rep EU 5.000 4.000 10.870 8.696 32 CM 4.000 2.000 8.696 4.348 19 Medienrat 4.000 4.000 8.696 8.696 14 FG 4.000 6.000 8.696 13.043 15 FP 3.000 4.000 6.522 8.696 30 FEB 3.000 3.000 6.522 6.522 35 IRG 3.000 1.000 6.522 2.174 47 ECJ 3.000 3.000 6.522 6.522 46 ECN 3.000 2.000 6.522 4.348 17 FCG 3.000 5.000 6.522 10.870 44 EPRA 2.000 2.000 4.348 4.348 18 FCP 2.000 3.000 4.348 6.522 20 GCG 2.000 2.000 4.348 4.348 7 CRC 2.000 3.000 4.348 6.522 12 Cass. Crt. 2.000 2.000 4.348 4.348 11 Com. Tr. 1.000 1.000 2.174 2.174 38 CEPT 1.000 0.000 2.174 0.000 36 Cocom 1.000 1.000 2.174 2.174 40 ERO 1.000 0.000 2.174 0.000 21 GCP 1.000 1.000 2.174 2.174 41 RSPG 1.000 1.000 2.174 2.174 45 ITU 1.000 0.000 2.174 0.000 37 RSC 0.000 0.000 0.000 0.000 43 Cenelec 0.000 0.000 0.000 0.000 39 ECC 0.000 0.000 0.000 0.000 42 ETSI 0.000 0.000 0.000 0.000 DESCRIPTIVE STATISTICS 1 2 3 4 OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 1 Mean 6.979 6.979 15.171 15.171 2 Std Dev 7.318 7.873 15.909 17.114 3 Sum 328.000 328.000 713.043 713.043 4 Variance 53.553 61.978 253.085 292.903 5 SSQ 4806.000 5202.000 22712.666 24584.121 6 MCSSQ 2516.979 2912.979 11894.984 13766.440 7 Euc Norm 69.325 72.125 150.707 156.793 8 Minimum 0.000 0.000 0.000 0.000 9 Maximum 29.000 29.000 63.043 63.043
Influence network degree centrality:
OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 24 Telenet 37.000 17.000 80.435 36.957 2 Federal Government 19.000 19.000 41.304 41.304 16 CSA 19.000 31.000 41.304 67.391 25 KPN Group 16.000 15.000 34.783 32.609 13 VRM 15.000 19.000 32.609 41.304 28 Test-achat / Test aankoop 14.000 3.000 30.435 6.522 3 Federal Parlament 12.000 33.000 26.087 71.739 5 Competition Council 11.000 18.000 23.913 39.130 29 Beltug 11.000 3.000 23.913 6.522 4 FME TC 10.000 8.000 21.739 17.391 1 IBPT 9.000 8.000 19.565 17.391 31 EC 9.000 4.000 19.565 8.696 22 Belgacom / Proximus 8.000 46.000 17.391 100.000 9 BCA 8.000 2.000 17.391 4.348 10 CS 7.000 3.000 15.217 6.522 23 Mobistar 7.000 7.000 15.217 15.217 6 FME CS 6.000 11.000 13.043 23.913 33 EP 6.000 2.000 13.043 4.348 26 Voo 5.000 6.000 10.870 13.043 7 CRC 5.000 1.000 10.870 2.174 32 CM 5.000 4.000 10.870 8.696 47 ECJ 5.000 2.000 10.870 4.348 12 Cass. Crt. 4.000 2.000 8.696 4.348 19 Medienrat 4.000 2.000 8.696 4.348 27 Alt. Oper. 4.000 4.000 8.696 8.696 14 FG 4.000 4.000 8.696 8.696 36 Cocom 4.000 2.000 8.696 4.348 34 ERG 4.000 2.000 8.696 4.348 44 EPRA 4.000 2.000 8.696 4.348 35 IRG 3.000 2.000 6.522 4.348 41 RSPG 3.000 1.000 6.522 2.174 40 ERO 3.000 1.000 6.522 2.174 15 FP 3.000 3.000 6.522 6.522 38 CEPT 3.000 1.000 6.522 2.174 17 FCG 3.000 2.000 6.522 4.348 37 RSC 3.000 1.000 6.522 2.174 20 GCG 2.000 3.000 4.348 6.522 11 Com. Tr. 2.000 2.000 4.348 4.348 39 ECC 2.000 2.000 4.348 4.348 8 Belg. Rep EU 2.000 3.000 4.348 6.522 46 ECN 2.000 2.000 4.348 4.348 18 FCP 2.000 1.000 4.348 2.174 43 Cenelec 2.000 2.000 4.348 4.348 45 ITU 2.000 1.000 4.348 2.174 42 ETSI 2.000 2.000 4.348 4.348 21 GCP 1.000 2.000 2.174 4.348 30 FEB 1.000 2.000 2.174 4.348 DESCRIPTIVE STATISTICS 1 2 3 4 OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 1 Mean 6.660 6.660 14.477 14.477 2 Std Dev 6.429 9.418 13.976 20.473 3 Sum 313.000 313.000 680.435 680.435 4 Variance 41.331 88.693 195.326 419.152 5 SSQ 4027.000 6253.000 19031.191 29551.039 6 MCSSQ 1942.553 4168.553 9180.309 19700.156 7 Euc Norm 63.459 79.076 137.954 171.904
Social network analysis energy
Information network degree centrality:
OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ Federale Minister Energie 22.000 14.000 75.862 48.276 Energieleveranciers 21.000 23.000 72.414 79.310 Energieproducenten 21.000 23.000 72.414 79.310 CREG 20.000 17.000 68.966 58.621 Transmissie 18.000 9.000 62.069 31.034 Distributie 17.000 9.000 58.621 31.034 Federale Minister Economie 13.000 10.000 44.828 34.483 VREG/CWAPE/BRUGEL 13.000 16.000 44.828 55.172 Raad voor de Mededinging 12.000 11.000 41.379 37.931 Hof van Beroep, Brussel 9.000 8.000 31.034 27.586 Europese DG Competitie 8.000 8.000 27.586 27.586 Europese DG Energie 7.000 10.000 24.138 34.483 Raad van State 7.000 6.000 24.138 20.690 Vlaams Departement Leefmilieu, afdeling milieuvergunnigen 6.000 10.000 20.690 34.483 FOD Economie, afdeling Energie 6.000 8.000 20.690 27.586 ACER 5.000 8.000 17.241 27.586 Federale Minister voor de Noordzee 5.000 5.000 17.241 17.241 VEA 5.000 9.000 17.241 31.034 Vlaamse Minister van Energie 5.000 6.000 17.241 20.690 BMM 4.000 4.000 13.793 13.793 Vlaams Departement Leefmilieu, afdeling energie 4.000 7.000 13.793 24.138 Vlaams Minister van Leefmilieu 4.000 7.000 13.793 24.138 FOD Economie, afdeling Mededinging 4.000 4.000 13.793 13.793 Vlaams Departement Ruimtelijke Ordening 3.000 7.000 10.345 24.138 DESCRIPTIVE STATISTICS 1 2 3 4 OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 1 Mean 7.967 7.967 27.471 27.471 2 Std Dev 6.940 6.030 23.932 20.794 3 Sum 239.000 239.000 824.138 824.138 4 Variance 48.166 36.366 572.718 432.409 5 SSQ 3349.000 2995.000 39821.641 35612.367 6 MCSSQ 1444.967 1090.967 17181.529 12972.256 7 Euc Norm 57.871 54.727 199.554 188.712 8 Minimum 0.000 0.000 0.000 0.000 9 Maximum 22.000 23.000 75.862 79.310
Information network betweenness centrality :
Betweenness nBetweenness ------------ ------------ Energieproducenten 83.104 10.234 Energieleveranciers 83.104 10.234 CREG 40.326 4.966 Federale Minister Energie 38.166 4.700 VREG/CWAPE/BRUGEL 21.165 2.607 Federale Minister Economie 11.105 1.368 Raad voor de Mededinging 8.031 0.989 Transmissie 7.774 0.957 Distributie 5.692 0.701 Vlaams Departement Leefmilieu, afdeling milieuvergunnigen 4.933 0.608 Europese DG Competitie 2.154 0.265 FOD Economie, afdeling Energie 1.979 0.244 Raad van State 1.850 0.228 Hof van Beroep, Brussel 1.595 0.196 VEA 1.549 0.191 Europese DG Energie 1.238 0.152 ACER 0.466 0.057 Vlaams Minister van Leefmilieu 0.295 0.036 Federale Minister voor de Noordzee 0.167 0.021 BMM 0.125 0.015 Vlaams Departement Leefmilieu, afdeling energie 0.091 0.011 Vlaamse Minister van Energie 0.091 0.011 FOD Economie, afdeling Mededinging 0.000 0.000 Vlaams Departement Ruimtelijke Ordening 0.000 0.000 DESCRIPTIVE STATISTICS FOR EACH MEASURE 1 2 Betweenness nBetweenness ------------ ------------ 1 Mean 10.500 1.293 2 Std Dev 21.878 2.694 3 Sum 315.000 38.793 4 Variance 478.654 7.260 5 SSQ 17667.135 267.950 6 MCSSQ 14359.635 217.787 7 Euc Norm 132.918 16.369 8 Minimum 0.000 0.000 9 Maximum 83.104 10.234
Influence network betweenness centrality :
OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ Energieproducenten 23.000 19.000 79.310 65.517 Energieleveranciers 23.000 19.000 79.310 65.517 Distributie 18.000 14.000 62.069 48.276 Federale Minister Energie 18.000 15.000 62.069 51.724 VREG/CWAPE/BRUGEL 14.000 12.000 48.276 41.379 CREG 14.000 15.000 48.276 51.724 Federale Minister Economie 11.000 6.000 37.931 20.690 Raad voor de Mededinging 10.000 9.000 34.483 31.034 Vlaams Departement Leefmilieu, afdeling milieuvergunningen 9.000 4.000 31.034 13.793 FOD Economie, afdeling Energie 9.000 6.000 31.034 20.690 Transmissie 7.000 5.000 24.138 17.241 Europese DG Competitie 6.000 11.000 20.690 37.931 Vlaamse Minister van Energie 6.000 7.000 20.690 24.138 Hof van Beroep, Brussel 5.000 9.000 17.241 31.034 Vlaams Departement Ruimtelijke Ordening 5.000 4.000 17.241 13.793 BMM 4.000 6.000 13.793 20.690 VEA 4.000 5.000 13.793 17.241 Federale Minister voor de Noordzee 4.000 4.000 13.793 13.793 Europese DG Energie 4.000 10.000 13.793 34.483 Vlaams Departement Leefmilieu, afdeling energie 4.000 4.000 13.793 13.793 Vlaams Minister van Leefmilieu 4.000 4.000 13.793 13.793 FOD Economie, afdeling Mededinging 4.000 4.000 13.793 13.793 Raad van State 2.000 10.000 6.897 34.483 ACER 2.000 8.000 6.897 27.586 DESCRIPTIVE STATISTICS 1 2 3 4 OutDegree InDegree NrmOutDeg NrmInDeg ------------ ------------ ------------ ------------ 1 Mean 7.000 7.000 24.138 24.138 2 Std Dev 6.588 5.453 22.717 18.803 3 Sum 210.000 210.000 724.138 724.138 4 Variance 43.400 29.733 516.052 353.547 5 SSQ 2772.000 2362.000 32960.762 28085.611 6 MCSSQ 1302.000 892.000 15481.570 10606.421 7 Euc Norm 52.650 48.600 181.551 167.588 8 Minimum 0.000 0.000 0.000 0.000 9 Maximum 23.000 19.000 79.310 65.517