at UCL LAWS
Promoting Accountability in Multi-Level Governance: A Network Approach
By Carol Harlow* and Richard Rawlings**
Abstract
This paper addresses problems of accountability in the system of multi-level governance, organized
around networks, which obtains in the European Union. An ‘accountability deficit’ arises when gaps
are left by the accountability machinery of the several levels of government, supranational and
national. This paper suggests a new evaluative framework based on the concept of 'accountability
network', questioning the hierarchical and pyramidal assumptions that presently underpin
accountability theory in the EU context. Using case studies of the Community Courts and European
Ombudsman, the paper suggests that ‘accountability networks’ may be emerging, composed of
agencies specializing in a specific mode of accountability, which come together or coalesce in a
relationship of mutual dependency, fortified by shared professional expertise and ethos. At present
fragmentary and imperfect, these might ultimately be capable of providing effective machinery for
accountability in network governance systems.
Keywords: governance, accountability, network, accountability network, European Union, Court of
Justice, ombudsman
I. The Challenge of Network Governance
Whether or not the terms ‘governance’ and ‘network’ are synonymous, there is at least substantial
overlap; Rhodes, for example, defines a socio-cybernetic system of governance as one composed of
‘self-organising networks’. 1 However this may be, as network systems or systems of governance
replace hierarchical and centred structures of government, the traditional control systems and
machinery for accountability are undermined. Kickert asserts that these ‘networks of accountability’
are left largely to control themselves because ‘government does not have enough power to exert its
will on other actors.’ A vacuum left by the decline of hierarchy is filled by 'self-responsibility’. 2
In EU governance, these characteristics are immediately apparent. Even in the First or Community
Pillar, where the more regular and institutional ‘Community method’ obtains,3 the task of the
Commission involves networking. 4 Long chains of actors need to be co-ordinated. Composed in
matters of policy-formation or rulemaking largely of national government representatives, the structure
* Emeritus Professor of Public Law, London School of Economics and Political Science; ** Professor of Public Law,
University College London. Our thanks are due to members of the Connex accountability workshop and to Renaud
Dehousse and Christian Joerges for constructive comments.
1
R. Rhodes, 'The New Governance: Governing Without Government' (1996) 44 Political Studies 652, 656-658; A. NiemiIilahti, 'Will Networks and Hierarchies ever Meet?' in A. Salminen (ed), Governing Networks (Amsterdam: IOS Press)
2003, 59; B. Jessop, 'Multi-Level Governance and 'Multi-Level Meta Governance', in I. Bache and M. Flinders (eds),
Multi-Level Governance (Oxford: Oxford University Press) 2004, 49.
2
W. Kickert, 'Complexity, Governance and Dynamics: Conceptual Explorations of Network Management' in J. Kooiman
(ed), Modern Governance: new government-society interactions (London: Sage) 1993, 275.
3
For further explanation, see K. Lenaerts and A. Verhoeven, ‘Institutional Balance as a Guarantee for Democracy in EU
Governance’ in C. Joerges and R. Dehousse (eds), Good Governance in Europe’s Integrated Market (Oxford: Oxford
University Press) 2002.
4
L. Metcalfe, 'The European Commission as a Network Organization' (1996) Publius: The Journal of Federalism 43.
is typically extended for implementation purposes by the addition of regional and local actors. On
many occasions too the networks include the private sector: corporate actors, interest group
representatives, voluntary and civil society organisations. In the field of structural funding, used by
Rhodes as a paradigm of ‘governance’, 5 this type of network proliferates. 6 In areas where ‘new
governance’ methods (such as the ‘Open Method of Co-ordination’) are in use, the Commission’s coordinating and harmonising functions are more striking still. Defined by the White Paper on European
Governance (2001) as a way of encouraging co-operation, the exchange of best practice, and
agreeing common targets and guidelines for Member States, OMC operates through ‘soft law’ without
formal enforcement mechanisms; instead, it ‘relies on regular monitoring of progress to meet those
targets, allowing Member States to compare their efforts and learn from the experience of others’. 7
This is Kickert's 'network of accountability', where accountability is largely left to participants in the
network through ‘report back’ mechanisms and ‘peer review’, the only sanction lying in
recommendations for improvement issued by the Commission.
To date three main strategies have been suggested to alleviate these problems, the first and most
conventional being the replication of state structures at transnational level: as might be said, the
transmutation of 'governance' into 'government'. 8 This process is exemplified in the progression
towards a European Constitution with federal tendencies. 9 The tacit assumption must be that any
'accountability gaps' left by the transfer of functions from national to supra-national institutions would
be filled by the creation of new institutions and a process of gradual integration, more particularly
integration through law. 10 But although an extensive literature was devoted to issues of sovereignty
and legitimacy, accountability at first attracted less attention. Exceptionally, Lodge noted an
accountability deficit, remarking on the 'extent to which the European Parliament had 'not won powers
forfeited to national governments by national parliaments'. She blamed national governments for
deliberately engineering 'a situation whereby the national parliaments were denied effective controls over
national executives. This made it easier for national governments, working within the Council, to escape
national as well as European parliamentary scrutiny and control.' 11 Third Pillar procedures accentuated
the problem, as did the stealthy transfer of European operations to EU agencies distant from their
publics. 12 In similar vein, Scott and Trubek 13 later warned of the lack of machinery for effective
control of the emerging practices of 'new governance'. Serious accountability gaps were developing.
The second strategy, of recourse to techniques of participatory decision-making, is here discounted.
This paper focuses on ex post facto accountability. We recognise that accountability may encompass
prior participation in policy-making, comprising also a standard-setting function, 14 but emphatically
reject the idea that these can provide an adequate substitute for ex post facto political and legal
accountability. In addition to the danger that participants in policy-making may be sucked into the
network and rendered complicit in decisions, Bignami highlights the weakness of EU rulemaking
5
R. Rhodes, Understanding Governance (Buckingham: Open University Press) 1997, 4.
J. Scott, 'Law, legitimacy and EC governance: prospects for “Partnership”.' (1998) 36 Journal of Common Market
Studies 175.
7
S. Regent, ‘The Open Method of Coordination: A New Supranational Form of Governance’ (2003) 9 ELJ 190; C. de la
Porte, ‘Is the open method of coordination appropriate for organizing activities at European level in sensitive policy areas?’
(2002) 8 ELJ 38.
8
C. Harlow, ‘Deconstructing Governance’ (2004) 23 YEL 57.
9
T. Christiansen, 'Towards Statehood? The EU's move towards Constitutionalisation and Territorialisation' Arena Working
Paper No 21 (2005), available at www.arena.uio.no
10
J. Weiler, The Constitution of Europe: “Do the new clothes have an Emperor?” and other essays on European
Integration (New York: Cambridge University Press) 1999, 10-96.
11
J. Lodge, 'The European Parliament', in S. Andersen and K. Eliassen (eds), The European Union: How Democratic Is It?
(London: Sage) 1996, 188.
12
D. Curtin, 'Delegation to EU non-majoritarian agencies and emerging practices of public accountability' in D. Gerardin,
R. Munoz and N. Petit (eds) Regulation through Agencies: A New Paradigm of European Governance (Cheltenham:
Edward Elgar) 2005.
13
J. Scott and D. Trubek, ‘Mind the Gap: Law and New Approaches to Governance in the European Union’ (2002) 8 ELJ
1.
14
C. Scott, 'The Governance of the European Union: The Potential for Multi-Level Control' (2002) 8 ELJ 59.
6
procedures, contrasting them unfavourably with the robust protections of American administrative law.
15
The third strategy involves reinforcement of network checks and balances. Regulatory theory is
generally somewhat weak on accountability; indeed, regulators frequently conflate the two concepts,
conceptualising accountability as a bipolar dialogue between regulators and regulated and showing a
marked preference for negotiated methods of problem solving, which are not necessarily exercises in
public accountability.16 In this respect Scott’s extended study of regulatory accountability is somewhat
exceptional. Convinced of the potential for harnessing ‘dense networks of accountability within which
public power is exercised… for the purpose of achieving effective accountability or control’, Scott
identifies two alternative models. In his interdependence model, members of a network, who are
'dependent on each other in their actions because of the dispersal of key resources of authority
(formal and informal), information, expertise, and capacity to bestow legitimacy' form a 'mutual
accountability network'. 17 This autonomous self-responsibility may (as Kickert implies), be a substitute
for the formal accountability to public law institutions eroded by network governance or (as Scott
suggests), be supplemented by formal accountability to public law institutions.
Applied to EU multi-level governance, the interdependence model suggests that the supervisory
powers vested in EU institutions (typically the Commission or the EU audit machinery) 18 can shore up
and fill gaps created by network governance or left by decreased accountability at national level – in
our view, a dubious assumption. We agree that the presence of state agents in a network can operate
as a control device to limit opportunistic behaviour by private parties and ensure respect for the public
interest – a process, we would add, that could work in reverse. We cannot, however, accept such
behavioural pressures as a substitute for accountability properly so called. Mutual accountability
networks tend to be more concerned with policy input and long-term relationships than retrospective
evaluation, rendering accountability difficult. There is a very real risk that they will degenerate into a
complacent ‘old boy network’, their accountability function blunted by mutual interest. There are
obvious problems of transparency, as Scott notes. 19 It is therefore questionable whether a mutual
accountability network can be shored up so as to add the requisite element of legitimacy to the
accountability process.
In Scott’s second, redundancy model, ‘overlapping (and ostensibly superfluous) accountability
mechanisms reduce the centrality of any one of them’. 20 Scott, instancing joint national and EU
expenditure programmes in the area of structural funding, describes this as a ‘belt-and-braces’ or
‘failsafe’ model of accountability, in which two or more independent mechanisms, each capable of
working on its own, are deployed to ensure the system does not fail. The requirement of joint funding
helps to ensure that both domestic and EU audit institutions take an interest in expenditure
programmes within Member States. 21 Even leaving aside the chance of ‘simultaneous failure’ of both
accountability systems, 22 this model is problematic, in that gaps may be left, which are not covered by
either. 23 In short, coverage by national and supranational accountability machinery of multi-level
decision-making processes is erratic and requires something stronger. We suggest that a fourth
strategy, the 'accountability network' as defined below, might prove more effective.
15
F. Bignami, ‘The Democratic Deficit in European Community Rulemaking: A Call for Notice and Comment in
Comitology’ 40 Harv Int’l LJ 451 (1999) and ‘Three Generations of Participation Rights in European Administrative
Proceedings’ (2004) 68 Law and Contemporary Problems 61.
16
J. Braithwaite, 'On Speaking Softly and Carrying Sticks: Neglected Dimensions of Republican Separation of Powers'
(1997) 47 UTLJ 47; R. Baldwin, Rules and Government (Oxford: Clarendon) 1995, 273-283. The slight change in attitude
noticed by R. Baldwin, 'The New Punitive Regulation' (2004) 67 MLR 351, does not affect our argument.
17
C. Scott, ‘Accountability in the Regulatory State’ (2000) 27 JLS 38.
18
On which see C Harlow, Accountability in the European Union (Oxford: Oxford University Press) 2002, ch. 5.
19
Scott, above note 17, pp. 57-58
20
ibid, 52
21
ibid, 53-54.
22
ibid, 60.
23
As in the 'Santer affair', below n. 54 and the Stichting Greenpeace case, below n. 48.
II. Analysing accountability
For Bovens, 24 whose definition we adopt as a convenient framework for this paper, accountability
consists of three main elements: (i) giving an account, in the attenuated sense of narration; (ii)
questioning or debating the issues; and (iii) evaluation or passing judgment. The implication, accepted
in this paper, is that accountability is primarily retrospective. Again in common with Bovens, we take
accountability to be essentially a public procedure, sited in an open forum or at least accessible to
citizens.
Bovens adds (iv) the possibility of sanction. Here we would inject a note of caution. It is in our view by
no means clear that sanction is an essential element in accountability, though we accept that lawyers
may be predisposed to think that it is. Sanction is very often illusory and may, rather than ‘thickening’
accountability, act as a deterrent by creating incentives to deny responsibility. We do not see this
difficulty as overcome by stretching the term (as Bovens does) to embrace informal ‘sanctions’ of
publicity or apology and negative consequences such as 'disintegration of reputation or career'. More
positively, we see reparation and effective redress as key factors in legitimation through
accountability; the machinery should, in other words, operate so as ‘to put matters right if it should
appear that errors have been made’. 25 We would thus accept, e.g., recommendations for
improvement issued by the Commission in OMC procedures as a form of accountability and, in our
second case study of the European Ombudsman (EO), we argue that the office is properly classified
as machinery for accountability.
Since ‘accountability network’ is a term used in several senses, it is important to establish the sense
in which we use it. For Stone, 26 a ‘mutual accountability’ network is composed of actors concerned
with the planning and execution of any specific area of activity (environmental policy or delivery of
policing and immigration services.) For Scott (as already indicated), an ‘extended accountability
network’ is created when standard organs of accountability, such as a parliamentary committee
entrusted with supervision of a given activity area like environment, or a regulator supervising
provision of a public service, are grafted on to a 'mutual accountability network' formed by participants
in a functional or policy network. 27 By linking the two disparate concepts of ‘policy network’ and
‘accountability network’, the accountability process is largely internalised.
This is not the usage we adopt in this paper. We reserve the term ‘accountability network’ for (i) a
network of agencies specialising in a specific method of accountability, such as investigation,
adjudication or audit, which (ii) come together or coalesce in a relationship of mutual dependency, (iii)
fortified by shared professional expertise and ethos. The two systems singled out for consideration in
the next two sections - the judicial system, best known and best developed of EU accountability
machinery and the European ombudsman network, which we see as coalescing into an accountability
network – have these characteristics in common. But as we shall suggest in our conclusions, a further
element is essential if an effective accountability network is to develop: (iv) a sense of a common
purpose. This is less easily discerned.
24
Bovens, in this Issue, 0.
D. Oliver, Government in the United Kingdom: The Search for Accountability, Effectiveness and Citizenship (Milton
Keynes: Open University Press) 1991, 22.
26
B. Stone, 'Administrative Accountability in the ‘Westminster’ Democracies: Towards a New Conceptual Framework.'
(1995) 8 Governance 505.
26 Scott, above n 17, 50
25
III. Courts as an Accountability Network
A. Legal Process and Accountability
Both lawyers and political scientists would subscribe to Mulgan’s assertion that an effective, independent
judicial system is a ‘fundamental prerequisite for effective executive accountability’. 28 'No society', argues
Judge Mancini, 29 can be considered truly democratic if its citizens are denied the possibility of vindicating
their legal rights in judicial proceedings, whether against the oppressive acts of a powerful legislature –
even a democratically elected one – or against the unlawful practices of an overweening administration.'
There would too be wide agreement round the view of courts as machinery for enforcement, and of
judicial process as 'thickening' accountability by providing sanction, reparation and redress. Although the
preferred terminology of lawyers remains that of the rule of law, 30 Oliver helps to equate the two notions,
defining accountability as 'a framework for the exercise of state power in a liberal-democratic system', 31 a
function ascribed by lawyers to the rule of law.
In the quest for public accountability, courts play a composite role, constituting machinery for
accountability (an accountability forum) but contributing also to public accountability indirectly – e.g.,
by buttressing transparency. 32 In Ely's influential American theory of constitutional review, the primary
role of constitutional courts is 'representation-reinforcement'; 33 courts shore up the institutions and
processes of democracy, protect political rights, underpin machinery for citizen participation and
guard the democratic process against 'malfunctions'. A somewhat similar claim has been made for the
European Court of Justice, on the basis of its doctrine of 'institutional balance' 34 and a
'representation-reinforcing' decision where the Treaties were read so as to open the doors of the court
to the European Parliament. 35
Courts can also contribute to transparency. 'The fact that citizens are aware of what the
administration is doing is a guarantee that it will operate properly. Supervision by those who confer
legitimacy on the public authorities encourages them to be effective in adhering to their [the citizens']
initial will and can thereby inspire their confidence, which is a guarantee of public content as well as
the proper functioning of the democratic system.' 36 Whether the two Community Courts have lived up
to these high expectations is doubtful. By holding the institutions strictly to the letter of the procedural
law (as they have usually done) they provide the occasion for a change of heart but they have not, as
urged by academic commentators, 37agreed to characterise transparency as a 'right' of democratic
citizenship, so transmuting into an accountability forum. 38 Occasionally courts have proved
28
R. Mulgan, Holding Power to Account, Accountability in Modern Democracies (Basingstoke: Palgrave Macmillan)
2003, 75-6
29
F. Mancini and D. Keeling, 'Democracy and the European Court of Justice' (1994) 57 MLR 175, 181.
30
Not so in the United States: see M. Dowdle, 'Public Accountability: Conceptual, Historical, and Epistemic Mappings', in
M. Dowdle (ed), Public Accountability, Designs, Dilemmas and Experiences (Cambridge: Cambridge University Press)
2006, 5.
31
D. Oliver, Government in the United Kingdom: The Search for Accountability, Effectiveness and Citzenship (Milton
Keynes: Open University Press) 1991, 28.
32
P. Dyrberg, 'Accountability and Legitimacy: What is the Contribution of Transparency?' in A. Arnull and D. Wincott
(eds) Accountability and Legitimacy in the European Union (Oxford: Oxford University Press) 2002.
33
J. Hart Ely, Democracy and Distrust: A Theory of Judicial Review (Cambridge, Mass: Harvard University Press) 1980.
34
On which see K. Lenaerts and A. Verhoeven, ‘Institutional Balance as a Guarantee for Democracy in EU Governance’ in
C. Joerges and R. Dehousse (eds) Good Governance in Europe’s Integrated Market (Oxford: Oxford University Press )
2002.
35
Case C-70/88 Parliament v Council (Chernobyl) [1990] ECR I-204; G. Bebr, 'The Standing of the European Parliament
in the Community System of Legal Remedies: A Thorny Jurisprudential Development' (1990) 10 YEL 171.
36
Case C-353/99P Hautala v Council No 2 [2001] ECR I-9565, Opinion of A G Léger, para 52.
37
D. Curtin, 'Citizens’ Fundamental Right of Access to EU Information: An evolving digital passepartout?' (2000) 37 CML
Rev 7; H. Kranenborg, 'Is it Time to Revise the European Regulation on Public Access to Documents?' (2006) 12 EPL 251.
38
Case C-68/94 Netherlands v Council [1996] ECR I-2169, followed in Hautala (above).
obstructive, as in Turco, 39 where the Court of First Instance construed one of the statutory exceptions
so expansively as to block public access to advice provided by the Council legal service on (ironically)
the Commission proposal for the regulation on public access to documents. Notably, the European
Ombudsman had reacted very differently. 40 It must, however, be recognised that freedom of
information is a hotly contested political issue in the EU. The regulatory package agreed only recently
and with difficulty represents an uneasy trade off between Member States with very different traditions
and EU institutions with widely differing interests. 41 The courts can be portrayed as acting in a
'representation-reinforcing' fashion by standing on the letter of the written law; by stepping in to
correct perceived 'malfunctions', they would bring into question their own accountability as policymakers.
For Shapiro, the duty to explain and give reasons underlies all review by courts. In the EU, Article 253
makes this a foundational Treaty obligation, serving, according to the Court of Justice, three distinct
purposes: the opportunity for parties to defend their rights; for the Court to exercise its supervisory
functions; and 'to Member-states and to all interested nationals of ascertaining the circumstances in
which the Commission has applied the Treaty'. 42 Reasons enable court and decision-maker to
engage in a 'synoptic dialogue': the decision-maker has to supply evidence and sufficiently good
reasons to persuade a more or less sceptical judge that the decision-making process, and ultimately
the decision, is rational. 43 Accountability can be heightened by application of the proportionality test
standard in EC law, whereby administrative decisions must demonstrably be: (i) suitable to achieve
the administrative objective; (ii) necessary for the achievement of the administrative objective; and (iii)
proportionate in the burden imposed on individuals. 44
B. Judicial Networking
If, as Lord suggests, 45 an effective legal accountability system must provide for ‘any citizen on a basis
of equality’ to access the court ‘with a complaint that power-holders are seeking to evade or distort the
rules by which they are themselves brought to account’, access becomes a pivotal issue. The
complex division of responsibility among the European courts is reflected in the access (standing)
rules, which are not homogenous but variable in national legal systems. To add to the problems,
Article 230 limits direct access to the two Community Courts to those who have a 'direct and
individual interest' and the Court of Justice has, in the face of constant academic criticism, 46 resisted
opening up the Community legal process to individuals. This opens dangerous gaps in legal
accountability. In Stichting Greenpeace, 47 protestors sought, with the aid of Greenpeace, to challenge
the building of a power station in the Azores, a Spanish project rendered possible only by
Commission funding. By holding the action inadmissible, the Court of First Instance blocked the path
to Commission accountability, leaving open only possible recourse against Spain in the Spanish
courts. It is too virtually impossible to test the validity of a Community regulation in the Community
39
Case T-84/03 Maurizio Turco v Council [2004] ELR II- 4061. On Council transparency, see D. Stavasage, 'Does
Transparency Make A Difference?' in C. Hood and D. Heald (eds) Transparency – The Key to Better Governance?
(London: The British Academy) 2006.
40
See Special Report to the Council in complaint 1542/2000(PB)(SM). The EO closed the matter in light of the
proceedings in Turco: see Decision on complaint 1542/2000 (PB) (SM) IJH against the Council.
41
Regulation EC 1049/2001 of the European Parliament and the Council of 30 May 2001 regarding public access to
European Parliament, Council and Commission Documents, 2001 OJ L 145/43; J. Heliskosi and P. Leino, 'Darkness at the
Break of Noon: The case law on Reg. No 1049/2001 on access to documents (2006) 43 CML Rev 735; Kranenborg, above
n. 37.
42
Case C-350/88 Delacre v Commission [1990] ECR I-395, para 15.
43
M. Shapiro, ‘The Giving Reasons Requirement’ (1992) University of Chicago Legal Forum 179, 183.
44
See now P. Craig, EU Administrative Law (Oxford: Oxford University Press) 2006, ch 17.
45
C. Lord, Democracy in the European Union (Sheffield: Academic Press) 1988, 96.
46
A. Arnull, 'Private Applicants and the Action for Annulment since Cordorniu’ (2001) 38 CML Rev 7; A Ward, ‘Judicial
Architecture at the Cross-Roads: Private Parties and Challenge to EC Measures Post- Jégo Quéré’ (2001) 4 Cambridge
Yearbook of European Legal Studies 00; A. Arnull, The European Union and its Court of Justice (Oxford: Oxford
University Press) 2nd edn, 2006, 00.
47
Case C-321/95 P Stichting Greenpeace International v Commission [1998] ECR I-1651.
Courts without a long diversion around the national legal systems. In Unión de Pequenos, 48 Advocate
General Jacobs advised the Court of Justice to tackle this effective denial of judicial protection by
changing its standing rules, arguing that: the long delays and heavy cost involved in the dual process;
the fact that Community institutions would not be before national courts; disparity in national standing
rules; all pointed in this direction. Declining, the Court of Justice asserted that it was for ‘Member
States to establish a system of legal remedies and procedures which ensure respect for the right to
effective judicial protection'; any incursion on their terrain would be 'unacceptable' – a recipe for
‘network comity’ often as not ignored! 49
Every court operating within the territory can, in the widest sense, be described as a Community
Court 50 and potential actor in a judicial network, though not every court operates as a public
accountability forum. Civil courts, handling disputes between private parties, are best characterised as
machinery for dispute-resolution, though they may act as an accountability forum when handling state
liability cases. Criminal courts, most closely linked with sanction, target citizens rather than state
officials, though we should not discount the current trend to lift jurisdictional barriers and immunities
and subject public actors to criminal penalties as in the recent Berlusconi cases. 51 In Cresson, 52 the
Court of Justice entered the long-running saga of corruption and maladministration in President
Santer's Commission 53 with a ruling that Madame Cresson had acted in breach of her obligations as
a European Commissioner. Whether this ruling supplied the requisite element of sanction is
questionable, since the finding of breach was said in all the circumstances to constitute a sufficient
penalty and no further sanction was imposed.
Well-developed networks of criminal and civil courts with considerable Commission backing exist in the
European Union, covering three separate fields of activity: the Brussels Convention negotiated under
TEC Article 220, which deals with recognition of jurisdiction and judgments; a Single Market initiative for
consumer protection, resulting in a Commission proposal, adopted by the Council in May 2001, to set up
a European Judicial Network (EJN) ‘to facilitate the life of people facing cross-border litigation’; and a
stream of activity involving criminal process after the Treaty of European Union formally inaugurated
cooperation in Justice and Home Affairs (JHA). The Commission’s Grotius programme (now replaced)
funded training programmes, exchanges, studies and research in the national and European courts with
a budget reaching EUR 3,750,000 by 2005. A network of civil judicial authorities became operational in
1998 with a sophisticated website, operated by the Commission, which provides information on, and links
to, judicial systems in Member States. 54 A dedicated server and information network aims to supply
judicial authorities with information necessary for sound judicial cooperation; to identify best practice; and
to resolve problems of judicial cooperation. Alongside the EJN, Eurojust, a transnational agency
describing itself as ‘the first permanent network of judicial authorities to be established anywhere in the
world’, 55 hosts meetings between investigators and prosecutors where good practice and ideas for selfimprovement are exchanged. These activities have gained in intensity since Enlargement and can be
expected to intensify further in the light of recent debates over the adequacy of the Bulgarian and
48
Case C-50/00 Unión de Pequenos Agricultores v Council [2002] ECR II-2365, with Opinion of Advocate General
Jacobs. See also Case C263/02P Commission v Jégo-Quéré et Cie SA [2004] ECR I-03425.
49
R. Crauford Smith, 'Remedies for Breaches of EU Law in National Courts: Legal Variation and Selection' in P. Craig and
G. de Burca (eds), The Evolution of EU Law (Oxford: Oxford University Press) 1998; T. Tridimas, ‘Enforcing Community
Rights in National Courts: Some Recent Developments’ in C. Kilpatrick, T. Novitz and P. Skidmore (eds), The Future of
Remedies in Europe (Oxford: Hart Publishing) 2000.
50
I. Maher, 'National Courts as EC Courts (1994) 14 Legal Studies 226.
51
C. Harlow, Accountability in the European Union (Oxford: Oxford University Press) 2002, 16-17. And see Joined cases
C-387/02, C-403/02 Berlusconi and Others (3 May 2005).
52
Case C-432/04 Commission v Edith Cresson (Judgement of 11 July 2006).
53
A. Tomkins, 'Responsibility and Resignation in the European Commission' (1999) 62 MLR 744; Committee of
Independent Experts, First Report on Allegations regarding fraud, mismanagement and nepotism in the European
Commission, Brussels (15 March 1999); Second Report on Reform of the Commission - analysis of current practice and
proposals for tackling mismanagement, irregularities and fraud (10 September 1999).
54
European Judicial Network in Civil and Commercial Matters at:
http://europa.eu.int/comm/justice_home/ejn/index_en.htm.
55
Eurojust, the European Union’s Judicial Cooperation Unit at: http://www.eurojust.eu.int
Rumanian justice systems. 56 But these are Stone's ‘networks of mutual accountability’, composed of
actors concerned with the planning and efficient execution of criminal and civil justice policies, tangential
to our theme.
In conformity with Mulgan, who describes judicial review as ‘in some respects the most powerful form of
external review of executive action’, 57 we would reserve the term ‘judicial accountability network’ for
those courts with jurisdiction to engage in judicial review of executive or legislative action - in practice,
usually administrative or constitutional courts. It is not always appreciated how greatly the twenty-five
national legal orders vary in this respect. Some systems possess specialised administrative jurisdictions,
modelled on the prestigious French Conseil d’Etat; some have separate constitutional courts; there may
here be no single, authoritative line of appellate authority; others, notably the United Kingdom, have a
single hierarchy of civil courts to which the state is subject. Judicial career arrangements, gender and
age of judges, recruitment, appointment and accountability are equally variable; as Bell puts it, judges in
different jurisdictions 'look different and behave differently'. 58These variants all affect the possibilities of
judicial networking.
The Court of Justice and Court of First Instance have competence ‘within their respective jurisdictions’ to
‘review the legality’ of the acts and omissions of the Community institutions and, where the Council so
provides, agencies and other entities. Because the EU possesses only limited competences, both courts
can also review the validity of secondary legislation. The Treaty of European Union, which set in place
the ‘Third Pillar’, left a significant gap in legal accountability, only partly filled at Amsterdam, when the
Council deliberately curtailed the jurisdiction of the Court of Justice in the JHA Title.59 (Predictions that
the Court would move to fill the glaring hole 60 receive some support from Pupino, 61 where the Court of
Justice, ruling that Member States are obliged to interpret national legislation in line with a JHA Council
framework decision, applied an established First Pillar obligation to the JHA, exposing the fiction that the
Court never intervenes in matters of national procedural law.
It is not open to national courts to pronounce on the validity of Community acts 62 but it falls to them to
hold national bodies accountable in respect of their conduct of EU affairs. The fragmentary nature of the
available information makes it hard to divine how successfully they do this. There are notable examples,
such as the Maastricht decision of the German Constitutional Court,63 which contains a significant
warning to the federal powers against 'hollowing out' national sovereignty; or the same Court's later
European arrest warrant decision, 64 which imposed substantial procedural restraints on the federal
German legislature. But there is variance between national legal orders, which differ not only with respect
to procedure, remedies and principles of judicial review but also with respect to its intensity and
effectiveness. Not only does the way in which judges approach their tasks vary but also the way they are
perceived: 'it is not just structures but political context (historical and contemporary) that ultimately
determines the level of judicialisation in any country.' 65 This helps to explain why some national
56
Report to European Parliament on the May 2006 Monitoring Reports on Bulgaria and Rumania, Memo 06/201, Brussels
(16 May 2006).
57
Mulgan, above note 27, 75.
58
J. Bell, Judiciaries within Europe, A Comparative Review (Oxford: Oxford University Press) 2006, 1.
59
TEC Art. 35 autorises preliminary rulings only where Member States have opted in and subject to the proviso of Art
35(5). For Title IV, see TEC Art 68.
60
E. Guild and S Peers, 'Deference or Defiance? The Court of Justice’s Jurisdiction in Immigration and Asylum', in E.
Guild and C. Harlow (eds), Implementing Amsterdam, Immigration and Asylum rights in EC Law (Oxford: Hart
Publishing) 2000.
61
Case C-105/03 Pupino [2005] ECR I-5285.
62
TEC Art. 234, as interpreted by the ECJ in Case 31/85 Foto-Frost (Firma) v Hauptzollamt Lübeck-Ost [1987] ECR
4199.
63
Bundesverfassungsgericht, 2nd chamber (Senat) Decisions BvR 2134/92 and 2 BvR 2159/92 (12 October 1993) BVerfGE
89, 155; in English, Brunner v The European Union Treaty [1994] 1 CMLR 57. And see P. Kirchhof, 'The Balance of
Powers Between National and European Constitutions' (1999) 5 ELJ 225; Hedergen, 'Maastricht and the German
Constitutional Court: Constitutional Restraints for an "Ever Closer Union"' (1994) 31 CML Rev 235.
64
Bundesverfassungsgericht, Decision 2BvR 2236/04 (18 July 2005).
65
C. Guarnieri and P.Pederzoli, The Power of Judges. A Comparative Study of Courts and Democracy (Oxford: Oxford
University Press) 2001, 3.
jurisdictions and some courts (the German Constitutional Court providing the prime example) have
emerged as stronger players than others.
Formally, the system of Community courts is held together by the preliminary reference procedure of
Article 234, which provides that any national ‘court or tribunal’ may pose questions concerning the
interpretation of EC law to the ECJ in the course of litigation, while those courts whose decision is final
must refer. It is for the national court to apply the ruling to the facts of the case before it. As drafted,
Article 234 seems to envisage a ‘flat’, non-hierarchical and co-operative judicial network, placing the
national courts in a position of technical equality, with the Court of Justice enjoying a consultative function
inter pares; it does not authorise the Court to over-rule or hear appeals from decisions of national courts.
66
The view famously articulated by Rasmussen 67 is, however, that the Court of Justice from the outset
viewed itself as a constitutional court. The language and nature of the discourse, involving at its highest
level issues of primacy, supremacy and power relationships amongst the legal orders, does suggest a
hierarchical structure, with the Court of Justice determined to stand at the apex. Judge Lenaerts recently
staked out just such a position. He told the US Supreme Court that the Court of Justice had devised its
own position as 'umpire of the lines drawn by the Constitution', drawing the inference from disparate
Treaty provisions that the judicial system rested 'on the Court having the last word within that system.' 68
One view of Article 234 proceedings is of a dialogue or discourse, in which national courts, consensually
bound into the network by their participation in the process, play a role as important as the Court of
Justice. 69 Thus Dehousse talks of a mutually supportive network of ‘responsible partners’, or ‘form of
judicial partnership in which governmental intrusion is resented’. 70 In these accounts, the primacy rule 71
is read as an empowerment of national courts, supplying the network with 'common purpose': the
recognition and enforcement of EC law at national level.
The companionate model gains support from Judge Mancini's early account of the ‘unlimited patience’
shown by the Court of Justice towards national judges when making preliminary references, where
necessary reformulating their questions and ‘coaching them in the elements of Community law’ supports.
By following ‘this courteously didactic method', the Luxembourg judges 'won the confidence of their
colleagues.’ 72 Simultaneously, the Court's first public relations policy was set in place by President
Robert Lecourt; since 1965, the Court has hosted meetings of superior national judges, sponsored
judicial conferences and educational visits, and paid regular visits to national jurisdictions. 73 Judges and
Advocates General regularly lecture to the national judiciary on EC law with the declared objective of
promoting mutual knowledge and trust between national judges and the Court of Justice, and
guaranteeing smooth co-operation between the levels. A judicial culture was under construction.
Informal, professional relationships were calculated to appeal to judges as part of a professional
community of lawyers. Amongst national judges were surely many sympathetic to the European
enterprise and happy to number themselves among the ‘tight epistemic community’ that surrounds the
prestigious Court of Justice. 74 The notion of 'club government', familiar to students of public
administration, is pertinent.
66
T. de la Mare, 'Article 177 and Legal Integration' in Craig and. de Burca (eds),above note 49; K Alter, Establishing the
Supremacy of European Law, The Making of an International Rule of Law in Europe (Oxford: Oxford University Press)
2001, 9.
67
H. Rasmussen, 'Why is Art 173 Interpreted against Plaintiffs?' (1980) 5 EL Rev 114 and On Law and Policy in the Court
of Justice (The Hague: Martinus Nijhoff) 1986. See also E. Stein, 'Lawyers, Judges and the Making of a Transnational
Constitution' 75 American Journal of International Law 1 (1981).
68
Delegation of European Commission to the US Supreme Court, March 29 2005 available at
http://www.eurunion.org/News/speeches/2005/050329kl.htm.
69
A-M Slaughter, A. Stone Sweet, J. Weiler (eds), The European Courts and National Courts - Doctrine and
Jurisprudence (Oxford: Hart Publishing) 1998.
70
R. Dehousse, The European Court of Justice (Basingstoke: Macmillan) 1998, 135-147; J. Weiler, ‘A Quiet Revolution The European Court of Justice and its Interlocutors’ (1994) 26 Comparative Political Studies 510.
71
Established in Case 26/62 Van Gend en Loos v Nederlandse Administratie Belastingen [1963] ECR 1.
72
F. Mancini, 'The Making of a Constitution for Europe' (1989) 26 CML Rev 595, 605-606.
73
R. Lecourt, L’Europe des Juges (Brussels: Bruylant) 1976. Dehousse, above note 70, 139, contains a Table of study
visits; further information is available from the Court’s website: http://curia.eu.int/en/index.htm.
74
M. Shapiro, ‘Comparative Law and Comparative Politics’ (1980) 53 S. Calif. Law Rev 537, 537-538; H. Schepel and R.
Wesseling, 'The Legal Community: Judges, Lawyers, Officials and Clerks in the Writing of Europe' (1997) 3 ELJ 165.
But Alter in a careful empirical study dismisses as 'misleading myth' the idea of widespread consensus
behind the supremacy of European law: 'the silent majority of the national legal community was neither
convinced nor supportive of European legal supremacy.' 75 The primacy doctrine proved Janus-faced: it
empowered national courts against national institutions but reduced their standing vis-à-vis the Court of
Justice. 'Common purpose' unravelled with the realisation that, in the event of clash between legal
orders, EC law must prevail. Reference procedure undercut national legal hierarchies and constitutions.
76
National courts, concerned for the internal integrity of the national legal order, began to manifest
resistance, obliging the Court of Justice to draw on the ‘fidelity clause’ of TEC Article 10 (which calls on
the Member States ‘to take all appropriate measures’ to ensure fulfilment of obligations arising out of the
Treaty) to underscore their obligations. 77
Perhaps relationships are cyclical: at some points companionate, at others adversarial. Certainly, the
battle over primacy seems on the point of resurgence just when the network is being asked to absorb ten
new legal orders. 78 This may help to explain the remarkable case of Köbler, 79 where the Court of Justice
decreed that Member States incur liability to compensate an individual who suffers loss or injury flowing
from a 'manifestly incorrect’ judicial ruling on EC law. Whatever its logical force, the decision does not
suggest a network of ‘responsible partners’ - very much the reverse! Judicial comity is hardly assisted by
rendering network members legally liable at the behest of a member of the network, which is – or so we
suppose - subject to no such liability. Casting national courts as conscripts rather than willing
collaborators, the judgment underscores the claim of the Court of Justice to hierarchical superiority in the
new Enlargement network. There is too danger to the wider network that, in holding Member State
governments liable for judicial errors, judicial independence will be threatened, undercutting the vision of
a ‘judicial partnership in which governmental intrusion is resented’. 80
Behind the scenes, there is much soft networking. The Association of Councils of State and Supreme
Administrative Jurisdictions of the European Union (ACS) was initiated in 1963, during an informal visit
by the Belgian to the Italian Council of State. As they joined, the supreme administrative jurisdictions of
each Member State were added. Statutes officially adopted in 2000 transformed a ‘family reunion of
personally acquainted delegates’ into a ‘professional association with a formal legal framework.’ Although
it receives EU financial support, the ACS operates from the Belgian Council of State, with the German
Federal Administrative Court currently holding the presidency; it is trying to deepen relations with the
Court of Justice - at present merely a member represented on the Executive Board. Objectives are
primarily educational: the ACS promotes exchange of views and experience on matters concerning the
jurisprudence and functioning of its members in the performance of their duties, more particularly with
regard to EC law. By 1968, so much common ground had been identified that regular colloquiums were
initiated (there have now been nineteen) on matters of common interest.
Significantly, interest has shifted from comparative studies of different national approaches to ‘specific
questions of Community law’. 81 Article 234 proceedings have twice been the subject of study and
problems of delay are presently under discussion with the Court of Justice. Immigration procedures,
increasingly absorbing the time and resources of national courts, were the focus of ACS studies in a
75
Alter, above note 66, 24-5.
See on hierarchy, Hoffmann J in Stoke-on-Trent CC v B & Q plc [1991] Ch 48, 63-4; R. Rawlings, 'The Euro-law Game:
Some Deductions from a Saga' (1993) 20 JLS 309, 319.
77
The most sustained, but by no means the only, resistance came from the powerful German Constitutional Court in the
Solange I and II cases: (1974) 37 BverfGE 271; (1986) 73 BverfGE 339. On Italy, see L. Daniele, 'Après l'arrêt Granital:
Droit communautaire et droit national dans la jurisprudence récente de la cour constitutionnelle' (1992) 28 Cahiers de
droit européenne 1; A. Bondi, 'The Corte di Cassazione and the Proper Implementation of Community Law' (1996) 21 EL
Rev 485; and on Spain, E. Garcia de Enterria, 'The Extension of the Jurisdiction of National Administrative Courts by
Community Law: the Judgement of the Court of Justice in Borelli and Article 5 of the Treaty' (1993) 13 YEL 19.
78
Decision K 18/04 of the Polish Constitutional Court (11 May 2005), discussed by K. Kowalik-Banczyk, 'Should we
Polish it up? The Polish Constitutional Tribunal and the Idea of Supremacy of EU law' (2005) 6 German Law Journal No
10: www.germanlawjournal.com.
79
Case C-224/01 Köbler v Republic of Austria [2003] ELR I-10239.
P. Wattel, ‘Köbler, Cilfit and Welthgrove: We Can't Go on Meeting Like This’ (2004) 41 CML Rev 177.
81
Association of Councils of State and Supreme Administrative Jurisdictions of the European Union, Newsletter 2001
available at: www.juradmin.eu
76
2005 seminar aimed to ‘compare the solutions adopted or planned by the Association’s members.’ 82 And
ACS is collaborating with a French initiative to make an overall study of administrative justice in Europe.
Perhaps a reversion to earlier interests, this could come to form the basis for a harmonisation.
‘The real core' of ACS activity is, however, its information network. It maintains two vital databases: with
the Court of Justice, a record of national administrative decisions involving EC law (DEC-NAT), and
JuriFast, its own database of Article 234 references and other interesting cases involving EC law. In
addition, there is a regular Newsletter and a recently established interactive Forum, designed to promote
exchanges between members. Currently, improvements to this network are under consideration. 83 As
we shall see in the next section, electronic links of this kind have a real potential for grounding
accountability networks.
Set up in 1972, the Conference of European Constitutional Courts (CECC), exists to promote the
exchange of information on the working methods and constitutional case law of members, with the
exchange of opinions on ‘institutional, structural and operational issues as regards public-law and
constitutional jurisdiction’. CECC now numbers 34 European constitutional courts and similar institutions,
including courts in most Member States. CECC meets triennially. Its website, created by the Belgian
Court of Arbitration, affords electronic access to national reports plus reports from the Court of Human
Rights and Court of Justice. Perusal of its website 84suggests, however, that networking is so far less well
developed. This may be why at the 2004 conference held in Bled in conjunction with the Council of
Europe’s advisory commission on constitutional affairs (the Venice Commission on Democracy Through
Law), members laid great emphasis on ‘continuous communication between the Courts, the
intensification of a truly efficient network of data exchange.’ 85
Unlike the EJN, defined above as a policy or mutual accountability network, the ASC and CECC fit
more clearly inside our ideal-type. Made up of courts specialising in administrative and constitutional
review, both networks have come together at the volition of their members. Shared professional
interests, expertise and ethos hold them together; they are not inspired or organised by the Court of
Justice or Commission nor are they directly linked to Member State ministries. Neither network is
hierarchical, as they might be if dominated by the Court of Justice. Whether members subscribe to
common accountability objectives is, however, questionable. Certainly, the ACS brings Member State
judges together regularly to discuss matters of common concern. The move away from comparative
judicial review studies to, currently, the proper administration and implementation of EC law could be
read, however, as a divergence from the primary role of national administrative courts: to hold
national institutions accountable.
IV. Ombudsmen as an Accountability Network
A. Ombudsman Process and Accountability
On the tenth anniversary of his office, in 2005, 86 the European Ombudsman (EO) conveniently
outlined the basics of a pan-European understanding of what an ombudsman should be and do: 87



82
Personal dimension of the office, with a publicly-recognised office-holder
Independence
Free and easy access for the citizen
ACS Newsletter 2005.
ibid.
84
Available at http://www.confcoconsteu.org/
85
CECC, The position of Constitutional Courts following integration into the European Union available at
http://www.uvi.si/eng/media/accreditations/constitutional-courts-conference/
86
See for a useful overview, A. Peters, ‘The European Ombudsman and the European Constitution’ 42 (2005) Common
Market Law Review 697.
87
N. Diamandouros, speech to fifth seminar of the national ombudsmen of the EU member states, 12 September 2005.
83




Primary focus on the handling of complaints, whilst having the power to recommend not only
redress for individuals but also broader changes to laws and administrative practices
Use of proactive means, such as own-initiative inquiries and providing officials with guidance
on how to improve relations with the public
Effectiveness based on moral authority, cogency of reasoning and ability to persuade public
opinion, rather than power to issue binding decisions
Broad review function that can encompass legal rules and principles, the principles of good
administration, and human rights
Thus visualised, the ombudsman technique is a paradigm of Bovens’ ‘thin’ concept of accountability,
being closely engaged in ‘account giving’, ‘questioning’ and ‘passing judgement’, while lacking
‘sanction’. An institutional design feature that enables procedures and criteria of accessibility to be
more flexible, this may however be considered a strength. One way of looking at the contribution of
ombudsmen to accountability is through the lens of alternative dispute resolution (ADR). As compared
with courts, the EO rightly stresses the value of a complaints service that is not only free at the point
of delivery but also free from strict standing rules and relatively swift. Likewise, whereas legal process
is indelibly associated with the concept of a zero-sum game, promoting ‘positive-sum outcomes that
benefit both the complainant and public authority’ is very much part of the ombudsman’s job. 88 Yet it is
the ability to self-start – in the case of the EO, the so-called ‘own initiative inquiry’ – which so clearly
distinguishes the technique.
Highlighting the need for a rounded and interactive approach, one that further stresses the
inspectorial function or parallel with audit, the authors have long used the terms ‘fire-watching’ and
‘fire-fighting’ to describe the role of ombudsman. 89 In similar vein, reflecting the classical Nordic
tradition that originally informed the Office, a leading commentator was soon ascribing to the EO a
dual function: the first, very relevant for present purposes, of helping to make the EU more
accountable by ‘providing an independent critical appraisal of the quality of administration by
Community institutions and bodies and a stimulus towards improvement’; the second that of court
substitute. 90 Pride of place goes here to the EO’s European Code of Good Administrative Behaviour.
91
An active role in standard-setting, invocation of the code as a set of benchmark tests for
maladministration in individual cases, and public advocacy for moving on from ‘soft law’92 to a binding
codification, is all on offer.
Against the backdrop of increasing reliance on independent agencies etc., in EU governance and the
evident fragility of classic forms of political accountability in this system, the EO’s contribution to
accountability can be seen as helping to reconcile executive delegation with parliamentary
democracy. 93 Operating as a parallel system for redress of grievance more oriented to collective or
explicitly ‘political’ issues; and, via presentation of his annual report, as (thin) machinery for the EO’s
own accountability, the Petitions Committee of the European Parliament is seen offering only limited
support. A special premium has been placed on the need for ‘co-operation’ between the EO and
88
N. Diamandouros, 'The EU Constitution: The Best Way Forward?', Speech to the Asser Colloquium on European Law,
15 October 2004.
89
That is, in the Anglo-Saxon context: C. Harlow and Rawlings, Law and Administration (London: Weidenfeld and
Nicholson) 1st edn. 1984.
90
K. Heede, ‘Enhancing the Accountability of Community Institutions and Bodies: The Role of the European
Ombudsman’ (1997) 4 EPL 587, 588.
91
Originating in the Own Initiative Inquiry into the existence and the public accessibility, in the different Community
institutions and bodies, of a Code of Good Administrative Behaviour (OI/1/98/OV).
92
As discussed by P. Bonnor, ‘The European Ombudsman: A Novel Source of Soft Law in the European Community’
(2000) 25 ELR 39.
93
P. Magnette, ‘Between Parliamentary Control and the Rule of Law. The Political Role of the Ombudsman in the
European Union’, (2003) 10 JEPP 677.
those bodies and institutions, pre-eminently the Commission and Council, which it is his task to hold
accountable. 94
The EO Office is in fact a modest entity, with much to be modest about. In accordance with the
mandate to combat ‘maladministration’, and so ‘enhance relations’ between citizens and the EU level
of governance as contemplated in the Maastricht Treaty, it has so far dealt with more than 20,000
complaints. 95 Yet the great majority of these – some 70%96 - prove inadmissible, mostly because they
are against national, regional and local administrations in the Member States, over which the EO –
investigating Community bodies – has no jurisdiction. Not that the EO could hope to do a great deal
more, armed as he currently is with a budget of only EUR 8m and some sixty staff. 97 For our
purposes the jurisdictional limitation has special resonance, since the EO has been pressed from the
outset to promote an accountability network. Given the evident mismatch between the formal reach of
the Office and the common experience of citizens of implementation of Community law by domestic
authorities, the EO would otherwise be lacking in credibility. The Office regularly claims credit for
advising complainants how to go elsewhere: most often, to another (domestic) ombudsman. 98
The first EO, Jacob Soderman, displayed a cautious attitude, no doubt explained in part by concern to
win acceptance for the Office in the Brussels jungle.99 However, his well-known contribution to the
cause of transparency, typically in advance of the ECJ, illustrates the potential of an ombudsman in
the role of ‘repeat player’. An own initiative inquiry; followed by a special report on access to
documents;100 expansion of transparency requirements into secretive areas of Council activity via
individual complaints; 101 plus various interventions in the public discussion of a critical102 and/or
agenda-setting kind: 103 it all adds up. Nor should the infamous Parga case,104 which involved
allegations of bias in Commission enforcement proceedings, go unremarked. Demonstrating the need
for continuous vigilance and the important place for ombudsmen as an accountability forum, it saw the
EO reserve his toughest language for the potentially corrupting deployment of staff.
The arrival in 2003 of the second EO, Nikiforos Diamandouros, marks a step-change. Today, a more
pluralistic conception of the ombudsman role is apparent, one to which the foundational Nordic
tradition is still a major contributor, but which further reflects a more general re-orientation in ‘the
Ombudsman world’, associated especially with human rights concerns. 105 There is too in the
contemporary workings of the Office the sense of a changing ethos associated with the great
expansion of the Union eastwards.
94
See generally, N. Diamandouros and others (eds.), The European Ombudsman: Origins, Establishment, Evolution
(Luxembourg: European Communities) 2005.
95
For consideration by the authors of EO internal practice and procedure, see C. Harlow and R. Rawlings, ‘Accountability
and law enforcement: the centralised EU infringement procedure’ (2006) 31 ELR 447.
96
N. Diamandouros, 'The role of the European Ombudsman'. Speech to the British and Irish Ombudsman Association, 8
April 2005.
97
Figures that include substantially increased resources in the light of enlargement: see for details, AR 2005, 164-165.
98
During 2005, the EO directed some 950 complainants to a domestic ombudsman and transferred some 90 cases (AR
2005, 121).
99
See, for a self-assessment, J. Soderman, ‘The Early Years of the European Ombudsman’, in N. Diamandouros and
others, above n. 94.
100
Special Report and Decision by the European Ombudsman following the Own-Initiative Inquiry into Public Access to
Documents held by Community Institutions and Bodies (December 1997).
101
As in the ‘Statewatch cases’: Complaints Nos. 1053, 1056, 1057/25.11.96/STATEWATCH/UK/IJH against the Council.
102
As in controversy surrounding the drafting of the transparency regulation (Regulation No 1049/2001): see P. Leino, The
Wind is in the North’ (2004) 11 EPL 333, 351-352.
103
As in the debate on ‘the future of Europe’: see J. Soderman, the Functioning of the Institutions, CONV 505/03
CONTRIB 206.
104
Decision on complaint 1288/99/OV against the Commission.
105
K. Hossain, Human Rights Commissions and Ombudsman Offices: National experiences through the world (The Hague:
Kluwer) 2000; id, Ombudsmen and the Protection of Rights in the European Union (Athens: European Ombudsman and
Greek Ombudsman) 2003.
EO activities likewise show a new sense of vigour. Elements include a drive for increased visibility
(more targeted information and multiple visits); unequivocal recognition of the fact that
maladministration encompasses unlawfulness but is not confined to it; and a commitment to more
own initiative inquiries. 106 With reference to Bovens’ criteria of ‘account giving’ and ‘passing judgment’
a more robust and intensive style of review is in evidence: for example, greater questioning of
interpretations by the Commission’s legal service; 107 and elaboration of an administrative duty of ‘due
diligence’ in the operation of (Article 226) infringement procedure. 108 A bold attack on the Council, on
the basis of a stretched interpretation of Treaty provisions, for its habit of lawmaking in secret, also
stands out. 109 And high on the list is widening and deepening of an accountability network. For
Professor Diamandouros, a political scientist, ‘co-operation among Ombudsmen is the way
forward’. 110
B. Ombudsmen Networking
Putting in place the foundations of an accountability network was one of Jacob Soderman’s early
objectives, the first concrete step being the creation in 1996 of a system of liaison officers linking the
EO and national ombudsmen. 111 As well as redirecting complainants or transferring cases, the aims
naturally included promoting the flow of information about Community law and its implementation.
Today, however, a fresh discourse is on offer: ‘European ombudsmanship’, the European ‘family of
ombudsmen’, and even ‘ombudsmanship as part of the European legal and political tradition’.
Likewise, with a view ‘to establishing a clearer public identity for our co-operation’, the preferred term
of art is now ‘The European Network of Ombudsmen’ (ENO). 112
Comparative public administration has witnessed serial bouts of ‘ombudsmania’, not least in the
expanding landmass of the EU. 113 The establishment of ombudsmen is rightly seen as a significant
step in the transition from communist rule to democracy. Many of the Member States that previously
lacked ombudsman have also caught the fever. 114 A strong development of regional ombudsmen in
various Member States, and subsequent inclusion in the EO’s network, is especially noteworthy,
given the many internal responsibilities for implementation of EC policies at the sub-national level.
Whereas Mr. Söderman’s original liaison network was an intimate affair, with national ombudsmen in
only 7 out of 12 Member States, ENO currently boasts some 90 agencies in 30 countries.
The form and style of the network displays several key features. One concerns the existence of an
‘ombudsman world’. In a striking reversal of so much in the contemporary development of
transnational or global administrative law, 115 the ‘regional’ or European dimension is here being
superimposed on a highly developed form of networking on the international plane. 116 Closer to home,
there is a considerable history of ombudsman development and education and training under the
aegis of the Council of Europe. 117 Nor should such networking activity be visualised simply in terms of
bipolar relationships between the EO and ombudsmen in individual Member States: the British and
Irish Ombudsman Association (BIOA), for instance, is the model of a ‘sub-regional’ network. This
106
N. Diamandouros, ‘Reflections on the Future Role of the Ombudsman in a Changing Europe’, in N. Diamandouros and
others (eds.), above n. 94.
107
See especially Decision on complaint 1273/2004/G against the Commission.
108
See especially Special Report following the draft recommendation to the European Commission in complaint
289/2005/(WP)GG.
109
Special Report following the draft recommendation to the Council in complaint 2395/2003/GG.
110
Diamandouros, ‘Reflections’, above n.106, 229.
111
AR 1996, 92-93.
112
N. Diamandouros, The role of the Ombudsman in future Europe and the mandates of Ombudsmen in future Europe.
Speech to meeting of European Ombudspersons and the Council for Europe Commissioner for Human Rights, 31 March
2005.
113
R. Gregory and P. Giddings (eds.), Righting wrongs: the Ombudsman in six continents (Amsterdam: IOS Press) 2000.
114
Italy, with no national ombudsman or functional equivalent, is today the exception.
115
C. Harlow, 'Global Administrative Law: The Quest for Principles and Values (2006) 17 Eur J of Int Law 187.
116
Plaudits go to the International Ombudsman Institute (IOI), which in true pioneering spirit has operated to promote,
foster and validate the ombudsman technique around the globe.
117
Committee of Ministers’ Recommendation No. R (85) 13 on the institution of the ombudsman.
point has resonance in the case of the ‘new democracies’ in the EU and in existing and potential
‘candidate countries’. Ombudsman networking activity, sponsored under such major initiatives as the
EU’s Stability Pact and the joint EU and OECD Sigma programme, is particularly intense in and
among the countries of South- Eastern Europe, commonly focused on protection of minorities. 118
Secondly, as with courts, very considerable diversity exists within ‘the family of ombudsmen’. Even
inside a single state, let alone across the EU, substantial differences in terms of functions, powers
and orientations between public sector ombudsmen, may well occur. 119 The idea of ENO affording a
uniform and universal service thus remains something of a chimera. The EO, however, is actively
promoting an agreed statement, or skeletal ‘codification’, of the ombudsman role inside the network. 120
The third and related feature is the voluntary and flexible character of the co-operation, in what is a
paradigm of a self-organising network. There is no equivalent of Article 234 reference procedure or
legal ‘glue’ for ENO, nor does the EO have at his disposal anything resembling the legal doctrine of
‘effective remedy’ with which to frame and structure ombudsman practice and procedure in Member
States. The ‘dialogue’ in this way differs qualitatively from that within the judicial community, reflecting
an altogether looser form of pan-European arrangement. If the EO is to be characterised as primus
inter pares, the primacy is not - or not yet - that sought by the ECJ.
This point needs underlining given the EO’s contributions to the ‘Constitutional Convention’, in which
Mr Söderman urged explicit recognition of a multi-layered system of extra-judicial remedies, centred
round an obligation on the respective complaints-handlers to co-operate ‘in a spirit of trust while
maintaining their independence’. 121 Bound up with significant extensions to the EO’s own role and
jurisdiction, the network should be formalised, with domestic ombudsmen empowered to transfer any
case involving fundamental rights under EC law. 122 Echoing the practice of certain ombudsman
systems, the EO should also have a direct right of access to the ECJ in such cases. Not surprisingly,
these lofty ambitions failed to win favour, the implicit demand that each Member State be obliged to
ensure effective and appropriate machinery for ADR jarring both with respect for national
constitutional traditions and with the subsidiarity principle.
Two aspects of the network’s contribution deserve special mention. One is its nurturing and standardsetting function for the many ‘young’ ombudsman systems in the new democracies, so building up
machinery of accountability. The EO’s Code has been translated into many languages and several
Ombudsmen have used it as a resource to enhance the quality of public administration in their own
countries. 123 The second concerns the EO’s user-friendly website, illustrating the quest for a panEuropean complaints service. Currently under consideration is a shared point of access or electronic
‘one stop shop’ for directing citizens to the appropriate ombudsman, be it at EU, national or regional
level. 124 Meanwhile, around the network, rapid exchanges of information, shared analysis of problems
and dissemination of best practice are founded on increasingly sophisticated use of IT. 125
Finally, the potential exists for joint inquiries, building on instances of informal sharing of information in
particular cases. A first step has been taken with a ‘parallel inquiry’ in close cooperation with the EO’s
Spanish counterpart into how public library systems might be affected by a new Community directive.
Ten members of the network responded to the EO’s request for information on ways of implementing
Community law, against which the Commission’s decision to bring infringement proceedings against
118
See for details, www.stabilitypact.org/wt1/040607-ombudsman.asp
R. Lawson, General report to the fifth seminar of the National Ombudsmen of EU Member States (The Hague:
European Ombudsman) 2005.
120
More particularly, through ENO’s bi-annual seminars: AR 2005, pp 118-121.
121
J. Soderman, Proposals for Treaty Changes, CONV 221/02 CONTRIB 206. The proposed system would have
incorporated parliamentary petitions.
122
It was further envisaged, as a stopgap, that the EO would investigate such cases for maladministration in the application
of EC law more generally if there was no local complaints machinery with competence in the matter.
123
N. Diamandouros, ‘Reflections’, above n.106.
124
Ibid.
125
As through the network’s Ombudsman Daily News: see AR 2005, pp 121-122.
119
Spain could be measured.126 This might be a further way to fill the gap in areas of ‘shared’ EU and
national administration. ‘To shed light on exactly who is responsible for what and, if appropriate, use
our powers to ensure that the full range of legally possible solutions is adequately and duly examined’
127
would be the very model of an accountability network approach in such matters.
V. Conclusions
Earlier in this paper, reference was made to three main strategies for promoting accountability in
transnational or multi-level governance systems. The first was the replication of state structures at
transnational level; the second, recourse to techniques of participatory decision-making; the third,
reinforcement of network checks and balances with additional accountability machinery (Scott's
interdependence and redundancy models.) Focusing on courts and ombudsmen, this paper has
examined a fourth: the construction of accountability networks.
What purposes do accountability networks serve? And how can they help to bridge accountability
gaps? From an internal viewpoint, their primary function is one of institutional cohesion. They bring
together agencies specialising in accountability with shared professional expertise and ethos,
providing at the most rudimentary opportunities for members of the network to meet and make
contact. In time, a network may develop a 'collective identity', empowering the network and allowing it
to engage in self-promotion or defend itself collectively against attack. This was very much the
function of the judicial network in its early days, when Member States, unaccustomed to the idea of
EC law as overriding the domestic legal system, might have turned against their courts. Accountability
networks also possess important educational functions. Disseminating information amongst members,
promoting best practice and subjecting practices and procedures to scrutiny by peers, in much the
same fashion as does 'OMC'. Enlargement, adding new, and sometimes weak or undeveloped,
members to the networks, renders education particularly necessary at the present time. This need is
strongly met in the ombudsman network. The ombudsman is well ahead in technological terms, with a
well laid out website, user-friendly alike to network members and their clientele, which could in time
develop into a 'one stop shop' for citizens' complaints. This marks the network's external function as
an access point for citizens.
National courts do, as indicated, serve as a gateway to legal accountability. They do not, however,
provide a 'one stop shop' and their technology is far less advanced. And, although the working
relationship between the Court of Justice and national courts is older and its members share
professional expertise and ethos, there are ways in which the network does not conform to our idealtype. Yes, judges network and collaboration and interchange seem on the increase, albeit slowly. In
the ACS, we can see the germ of an accountability network in terms of our strict definition. This
informal network contrasts with the formal judicial arrangements for the implementation of Community
law as conceptualised by the Court of Justice. Created voluntarily by its members, its members
specialise in accountability and are involved in holding national governments accountable; the Court
of Justice, which fulfils the same function for the Community institutions, is an ordinary member
represented on the Council.
Earlier we suggested that common purpose is necessary for the effective functioning of an
accountability network. This we see as the primary reason why the ombudsman network seems,
somewhat to our surprise, further advanced towards an accountability network than the better-known
circle of Community Courts. A striking feature of the former is the extent to which developments are
happening on the initiative of those who deal in accountability and not at the instigation of
Commission, Council, or indeed the central governments of Member States. There is a strong sense
already of a self-organising and self-generating network of investigative officials, fortified by shared
professional expertise and ethos, who have come together to execute the common purpose of
fostering and encouraging good administration, and of holding administrators throughout the EU
126
127
Decision of the EO on joint complaints 3452/2004/JMA against the Commission).
N. Diamandouros, speech to fifth seminar of the national ombudsmen, above n. 87.
accountable for acts of maladministration. Experiments have begun with a ‘parallel inquiry’, the logical
conclusion of which would be a power to undertake joint investigations, favouring more effective
outcomes. This degree of cohesion assumes, however, that the network is able to agree upon a
common purpose.
Unlike the evolving ombudsman network, the judicial network is not based on ‘non-hierarchical method[s]
of mediating conflict’. 128 It is not a ‘flat hierarchy’ involving recognition of ‘co-ordinately valid legal
systems’.129 That ideal is undercut by both the primacy doctrine and the Court of Justice's ambition to
occupy the place of supreme constitutional court. Clearly this places national courts in an ambiguous
position, charged as they are with maintaining the integrity of the national legal order, while at the same
time upholding the primacy of EC law. The judicial pyramid generates internal tensions, creating doubt
over commitment to a common purpose. The network can combine against Member States but the
Community institutions are not treated in like fashion, since they are not accountable in national courts.
This is in one sense, the challenge of multilevel governance and the multiple loyalties it engenders. An
alternative reading of the situation is as a network malfunction impeding the development of a true
accountability network capable of meeting the challenge of network governance.
Fisher has expressed doubt whether accountability and accountability processes can operate outside
the context of a liberal-democratic government system. 130 We are cautiously optimistic. We see the
development of network governance as inviting a response in the shape of accountability networks.
We do not proffer accountability networks as the 'ultimate accountability principle’ nor, indeed, as
more than a partial answer to the acknowledged problems of accountability associated with the rise of
multi-level governance in Europe. What we do suggest is that the hierarchical and pyramidal
assumptions presently underpinning accountability theory in the EU context warrant re-examination.
The present EO, in a reference to ‘the increasing intensity of co-operation among administrations at
all levels of the European Union’, has spoken of the need for this to be ‘matched by co-operation
among ombudsmen’. 131 This is another way of articulating our thesis.
128
129
F. Scharpf, 'Community and autonomy: multi-level policy-making in the European Union' (1994) 1 JEPP 219, 225
N. MacCormick, 'Liberalism, Nationalism and the Post-sovereign State' in R. Bellamy and D. Castiglione,
Constitutionalism in Transformation: European and Theoretical Perspectives (Oxford: Blackwell) 1996 and 'Review Article
– Questioning “post-sovereignty”', (2004) 29 EL Rev 852.
130
131
E. Fisher, 'The European Union in the Age of Accountability' (2004) 24 OJLS 495, 496.
Diamandouros, above n. 87.