Sub-Constitutional Engineering:
Negotiation, Content, and Legal Value of
Interinstitutional Agreements in the EU
Isabella Eiselt and Peter Slominski*
Abstract: This article analyses the roles and impact of Interinstitutional Agreements
(IIAs) in the EU, taking into account their relationship to primary law. Concretely speaking,
these roles range from (a) explicitly authorised specifications of Treaty provisions
via (b) not explicitly authorised specifications of vague Treaty law to (c) pure political
undertaking. Based on the distinction between the constitutional and the operational level
of the political game, we challenge the assumption that IIAs usually strengthen the European
Parliament. As our case study, the 1993 interrelated package of IIAs on democracy,
transparency and subsidiarity, illustrates, the European Parliament is not the only institution
that benefits from IIAs, especially if they lack a sufficiently precise Treaty basis.
Furthermore, if Treaty provisions underlying IIAs are precise, they also tend to produce
precise and thus legally relevant content. Conversely, if IIAs deal primarily with elusive
concepts they are likely to be legally ambiguous or even irrelevant at all.
I Introduction
Constitutionalisation of the European Union has to be understood as a process that
started with the founding treaties in the 1950s and included several amendments by
Intergovernmental Conferences (IGCs) in 1986, 1992, 1997, and 2000. But the constitutional
reality is more than the `grand bargains' negotiated at IGCs. Important
constitutional developments have also occurred between amendments of primary law,
notably through legislative activity by the Community institutions and the rulings of
the European Court of Justice. Apart from legislative and judicial activities, numerous
institutional innovations have commenced outside formal Treaty law and have--
sometimes though not always--been included into the Treaties at a later stage.
Some of these phenomena have resulted from Interinstitutional Agreements (IIAs)
which have a long history in the European Union.1
Since the Single European Act
has strengthened the role of the European Parliament, the need for interinstitutional
European Law Journal, Vol. 12, No. 2, March 2006, pp. 209­225.
 2006 The Authors
Journal compilation  2006 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford, OX4 2DQ, UK
and 350 Main Street, Malden, MA 02148, USA
* Research Fellows at the Institute for European Integration Research, Austrian Academy of Sciences.
1
The so-called `Luns Procedure' of 25 February 1964, which ensured the involvement of the European
Parliament in the negotiation of association agreements is considered to be the first IIA, see C. Bobbert,
Interinstitutionelle Vereinbarungen im Europäischen Gemeinschaftsrecht (Peter Lang, 2001) p. 1.
cooperation has led to an increased importance of IIAs.2
Empirical evidence suggests
that there have been well over 100 IIAs adopted over the years.3
However, as yet we do
not know the exact number of IIAs, nor do we have comprehensive knowledge about
their political functions and legal status. One of the statements that can be found in
the sparse literature on this subject is that IIAs have strengthened the position of the
European Parliament within the EU's institutional arrangement.4
This article takes a
different stance: After providing a succinct overview of what we understand as IIAs as
well as of their legal status, we argue that IIAs per se do not strengthen the European
Parliament. In order to give a proper understanding of IIAs we have to take into
account their relationship to primary law which provides the framework for the negotiations
and the concluded content of IIAs. Or in the words of historical institutionalism:
Decisions at an earlier point of time (e.g. adoption of primary law) will structure
the negotiations as well as the possible outcome (e.g. IIAs) occurring at a later point
of time.5
By emphasising the impact of (formal) institutions on politics over time, we
can easily see that the negotiation success of any given Community institution depends
on its surrounding legal environment, i.e. Treaty law. To put some flesh to this argument
we take the 1993 interrelated package6
of IIAs on democracy, transparency, and
subsidiarity as an example to show that IIAs not only have to be analysed against the
backdrop of primary law, but also that IIAs fulfil three different roles depending on
their relationship to primary law. Concretely speaking, these roles range from (a) explicitly
authorised specification of Treaty provisions via (b) not explicitly authorised
specification of notoriously vague Treaty law to (c) pure political undertaking.
European Law Journal Volume 12
210  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
2
See P. Craig and G. de Burca, EU Law. Text, Cases, and Materials (Oxford University Press, 2003) p. 159;
J. Monar, `Interinstitutional Agreements: The Phenomenon and its New Dynamics after Maastrichť,
(1994) 31 CMLR 693; F. Snyder, `Interinstitutional Agreements: Forms and Constitutional Limitations,
in G. Winter (ed.), Sources and Categories of European Union Law (Nomos Verlag, 1996) p. 453.
3
In a recent publication, W. Hummer speaks of 161 IIAs he has collected over the years, but unfortunately
fails to provide detailed references of the relevant documents: see W. Hummer, `Interinstitutionelle Vereinbarungen
und "institutionelles Gleichgewicht"', in W. Hummer (ed.), Paradigmenwechsel im Europarecht
zur Jahrtausendwende. Ansichten österreichischer Integrationsexperten zu aktuellen Problemlagen. Forschung
und Lehre im Europarecht in Österreich (Springer Verlag, 2004) pp. 129­136. The authors are currently
involved in an international research project funded by the Austrian Ministry of Education, Science and
Culture, which inter alia aims to provide a complete list of all IIAs ever concluded.
4
Hummer, op. cit. note 3 supra, at 115; Monar, op. cit. note 2 supra, at 693; A. Maurer, D. Kietz and
C. Völkel, Interinstitutional Agreements in CFSP: Parliamentarisation through the Backdoor?, (2004) EIF
Working Paper No.5.
5
See P. Pierson, `Increasing Returns, Path Dependence, and the Study of Politics', (2000) 94 American
Political Science Review 251; M. A. Pollack, `The New Institutionalisms and European Integration', in
A. Wiener and T. Diez (eds), European Integration Theory (Oxford University Press, 2004) pp. 139­141.
6
We choose the term `package' because it highlights the fact that the adoption of the IIA on subsidiarity
was made conditional by the European Parliament on the conclusion of another IIA on democracy and
transparency. Their common political context is further confirmed by the fact that they were published in
the same section of the Official Journal and mentioned in the Interinstitutional Declaration of 25 October
1993 of the European Parliament, the Council, and the Commission on Democracy, Transparency and
Subsidiarity (number 4­6). Concretely, we are dealing with the following three documents: (a) The Interinstitutional
Declaration of 25 October 1993 of the European Parliament, the Council and the Commission
on Democracy, Transparency and Subsidiarity (OJ C 329/133 25/10/1993); (b) Draft Interinstitutional
Agreement on Procedures for Implementing the Principle of Subsidiarity (OJ C 329/135 1993); (c) Draft
Decision of the European Parliament laying down the regulation and general conditions governing
the performance of the Ombudsman's duties (OJ C 329/136 1993; the Decision was finally adopted on
9 March 1994 (OJ L 113/15 1994).
We will find that the European Parliament gets the most out of IIA negotiations if
the Treaty provisions are already sufficiently clear and explicitly authorise Community
institutions to determine only some `lower leveľ operational details. If, by contrast, the
Member States are unable to forge a viable and precise compromise at an IGC, resorting
to vague terms without real meaning, the European Parliament is far less successful
in overcoming these differences among the Member States through an IIA. While
the former results in sufficiently clear and enforceable IIA provisions, the latter tends
to lead to political declarations or vague provisions that closely resemble their corresponding
Treaty provisions.
II What are IIAs and What is their Legal Status?
This article defines an IIA as an agreement between institutions of the EU.7
This broad
definition allows us to include a wide variety of existing agreements regardless of their
denomination, form and content and to consider their potential common features. As
mentioned above, there is as yet no comprehensive list of adopted IIAs.8
This can at
least partly be explained by the fact that IIAs are often difficult to detect.9
While the
term `Inter-institutional Agreemenť was first used in 1988,10
IIAs come in many denominations
ranging from joint declaration, exchange of letters, Council Note, or Decision
of the European Parliament.11
Second, there is no Treaty provision regarding the publication
of IIAs. As a consequence, numerous IIAs, especially the early ones, are not
published in the Official Journal.12
Even within the group of IIAs published in the Official
Journal, we can discern differences: while most IIAs are published in the C series
of the OJ, some are published in its L series. Third, IIAs are used in a huge number of
March 2006 Sub-Constitutional Engineering
 2006 The Authors 211
Journal compilation  Blackwell Publishing Ltd. 2006
7
A different more political science oriented conceptualisation of IIAs can be found in J. Stacey, Constitutional
Re-engineering in the European Union: The Impact of Informal Interinstitutional Dynamics.
Paper presented at ECSA Seventh Biennial International Conference, 31 May­2 June 2001, Madison,
Wisconsin, at 36. In most cases, IIAs are concluded between two or three Community institutions, namely
the European Parliament, the Council, and the Commission. However, there are numerous other IIAs
which also involve other institutions such as the Court of Auditors, the European Court of Justice, the
Economic and Social Committee, and so on.
8
For a preliminary number of IIAs see Hummer, op. cit. note 3 supra; in 1973 the Commission published
a Communication which seems to be the only official document on `practical measures' bilaterally agreed
upon between the Commission and the Parliament, see Commission Communication to the European Parliament
on `Practical Measures to Strengthen the Powers of Control of the Parliament and to Improve
Relations Between the Parliament and the Commission', (COM(73)999). Some examples of IIAs can also
be found in Snyder, op. cit. note 2 supra, at 455­458 and Bobbert, op. cit. note 1 supra, at 13­40.
9
While English and French scholarly literature seems to stick to one term (inter-institutional agreement;
accords interinstitutionnels), German legal literature uses various terms to deal with IIAs, namely `interinstitutionelle
Vereinbarung', `Interorganvereinbarung', or `Inter-Organ-Verträge', see Bobbert, op. cit. note 1
supra, at 6.
10
Interinstitutional Agreement of 29 June 1988 between the European Parliament, the Council and the Commission
on Budgetary Discipline and Improvement of the Budgetary Procedure, OJ L 185/33 1988.
11
For an overview see Hummer, op. cit. note 3 supra, at 133.
12
The first IIA was published in 1975, see Joint Declaration of 4 March 1975 of the European Parliament,
the Council, and the Commission on the Conciliation Procedure in Budgetary Matters, OJ C 89/1 1975.
For example the so-called `Luns' procedure (see note 1 supra) was not published and can only be found
in the academic literature, see J-V Louis, `Le rôle du Parlement européen dans ľélaboration et la conclusion
des accords internationaux et des traités ďadhésion', Liber Amicorum Frédérique Dumon (Kluwer,
1983) 1153; H-W. Rengeling, `Zu den Befugnissen des Europäischen Parlaments beim Abschluß völkerrechtlicher
Verträge im Rahmen der Gemeinschaftsverfassung', in I. von Münch (ed.), StaatsrechtVölkerrecht-Europarecht,
Festschrift für Hans-Jürgen Schlochauer (Walter de Gruyter, 1981) p. 879.
policy fields including budgetary procedure and comitology, as well as legislative
procedures.13
Given the sheer number and complexity of IIAs, it seems plausible
to advocate first and foremost an official compilation of all currently existing IIAs.14
Additionally, the IIAs should have an explicit legal basis in the Treaty ensuring a standardised
denomination as well as mandatory publication in the Official Journal.15
Against this background of pluralism, we turn our attention to the legal status of
IIAs. While it is beyond the scope of this article to provide a detailed legal analysis
of IIAs as such it seems appropriate to focus on legal effects of IIAs in general and of
our selected IIAs in particular. Although the Treaties do not contain an explicit provision
allowing for the adoption of IIAs per se, Community institutions are entitled
to adopt IIAs on the basis of either an explicit authorisation16
or the duty of loyal
cooperation (Article 10 EC).17
It remains undisputed that IIAs cannot modify primary
or secondary law.18
But within these legal boundaries, IIAs may eventually have legal
effects19
deriving either from their Treaty basis or the intention of the drafting parties.
Although there is no comprehensive jurisprudence of the Court of Justice on IIAs, we
may assume that contracting parties expressly intend to bind themselves if the wording
is `clear'20
or `sufficiently precise and unconditionaľ.21
In these cases, IIA provisions
tend to be binding at least among the parties involved. Conversely, IIA provisions that
are vague or ambiguous may imply that the parties involved do not intend to bind them-
selves.22
Another indicator of the legally binding intention is a provision that an IIA
can only be amended by common agreement of the institutions involved.23
All other
European Law Journal Volume 12
212  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
13
For a good list of IIA affected policy fields see Hummer, op. cit. note 3 supra, 134 et seq.
14
The role model could be the so-called register of Comitology of the Commission, see
<http://europa.eu.int/comm/secretariat_general/regcomito/registre.cfm?CL=en>.
15
See Art III-397 of the Treaty establishing a Constitution for Europe, which states that the Parliament,
Commission and Council may conclude inter-institutional agreements--which may be of a binding
nature--in order to arrange their cooperation. While this may be regarded as a step in the right direction,
the provision fails to ensure a uniform denomination and mandatory publication.
16
See Art 218(1) EC; Art 193(3) EC; Art 195(4); Art 248(3) EC; Art 272(9) EC.
17
Hummer, op. cit. note 3 supra, at 253.
18
Snyder, op. cit. note 2 supra, at 464, referring to Advocate General Manchini's often-cited statement:
`[I]t remains nevertheless undeniable that joint declarations and similar measures merely constitute "droit
de complement" which may not derogate from primary law on pain of invalidity.' (Case C-204/86 Greece
v Council [1988] ECR 5323, para 9).
19
Snyder has pointed out that the legal effect of IIAs `does not necessarily mean to be legally binding erga
omnes', but may in fact appear in many forms ranging from expressing general principles, creating expectations
of conduct to serving as an aid in judicial interpretation. Moreover, IIAs may be legally binding
among the contracting Community institutions (e.g. Case C-25/94 Commission v Council [1996] ECR I1469,
para 49); in contrast to contracting agreements, Community institutions are also entitled to commit
themselves unilaterally, i.e. they may adopt rules and thus be `bound to observe those rules, even if [there]
was under no legal obligation to adopt them' (see Joined Opinions by Advocate General Vesterdorf
on T-1-4/89, T-6-15/89 ECR [1991] II-867; see also Case 81/72 Commission v Council ECR [1973] ECR
para 10). They may also impact on third parties (as will become obvious with regard to the Decision of
the Ombudsman, see below), for details see Snyder, op. cit. note 2 supra, at 461­463.
20
Case C-25/94 Commission v Council [1996] ECR I-1469, para 49; Case C-58/94 Netherlands v Council [1996]
ECR I-2169, para 25; Case C-106/96 UK v Commission [1998] ECR I-2729 para 25.
21
See Advocate General Mancini in Case 204/86 Greece v Council [1988] ECR 5323 para 9.
22
Snyder, op. cit. note 2 supra, at 464.
23
See e.g. Joint Declaration of 30 June 1982 by the European Parliament, the Council, and the Commission
on various measures to improve the budgetary procedure (OJ C 194/1 1982) and the pertinent Court
of Justice ruling in Case 204/86 Greece v Council [1988] ECR 5323 para 16 et seq.; see Bobbert, op. cit.
note 1 supra, at 110.
IIAs whose binding quality can neither be based on explicit Treaty law nor on the intention
of the drafters can be regarded as mere political declarations without any legal
impact whatsoever.
III Treaty Law and IIAs: A Crucial Relationship
Before we deal with the relationship between primary law and IIAs in greater detail,
we would like to recall that EU studies have undergone a `constructivist turn' in the
recent past.24
Hence, political choices are not only driven by material interests, as rational-oriented
theories would suggest, but also by ideas.25
We argue, however, that interests
and ideas are involved in different ways depending on the level of institutional
interaction. Rittberger rightly points out that we have to distinguish between the
constitutional and the operational level of the political game.26
As to constitutional
issues, policy makers are concerned with `higher leveľ concerns such as the basic
options of institutional design. Whenever `higher leveľ issues are negotiated, there is
uncertainty about interpretation of fundamental principles and norms guiding the
political process.27
As a consequence, debates on `higher leveľ issues tend to revolve not
only around material interests but also around general ideas including `philosophical
assumptions concerning the interplay among human nature, political institutions, and
the good society'.28
Once `higher leveľ decisions have been taken, policy makers turn
to `lower leveľ issues specifying provisions which have been agreed upon beforehand.
`Lower leveľ negotiations operating `within a given set of rules' are mainly `concerned
with who gets what, when, and how',29
which can be called distributive bargaining. In
the following we will discuss the three--already mentioned--roles of IIAs emphasising
on their impact on negotiation style, content, and legal value.
A Explicit Authorisation
Neither the TEU nor the TEC have a general provision that explicitly authorises
the Community institutions to conclude IIAs whenever they want to. But there are
numerous Treaty provisions that authorise either two or three Community institutions
to adopt specific IIAs in certain policy fields.30
Additionally, both the IGC in
March 2006 Sub-Constitutional Engineering
 2006 The Authors 213
Journal compilation  Blackwell Publishing Ltd. 2006
24
T. Christiansen, K. E. Joergensen and A. Wiener, `The Social Construction of Europe', (1999) 6 Journal
of European Public Policy 528 at 543.
25
V. Vanberg and J. M. Buchanan, `Interests and Theories in Constitutional Choice', (1989) 1 Journal of
Theoretical Politics 49 at 51 (quoted in J. Lindner and B. Rittberger, `The Creation, Interpretation and
Contestation of Institutions--Revisiting Historical Institutionalism', (2003) 41 Journal of Common
Market Studies 445 at 449).
26
B. Rittberger, `Which institutions for post-war Europe? Explaining the institutional design of Europe's
first community', (2001) 8 Journal of European Public Policy 673 at 677; see also J. M. Buchanan and
G. Tullock, The Calculus of Consent. Logical Foundations of Constitutional Democracy (University of
Michigan Press, 1962).
27
Lindner and Rittberger, op. cit. note 25 supra, at 450.
28
C. Jillson and C. Eubanks, `The political structure of constitution making: The Federal Convention of
1787', 28 (1984) American Journal of Political Science 435, at 438.
29
See V. Ostrom, Constitutional Level of Analysis: Problems and Prospects. Convention Paper presented at
the Meeting of the Western Political Science Association, Portland on 22­24 October 1979 (quoted in
Jillson and Eubanks, op. cit. note 28 supra, 438).
30
With regard to Structural Funds (Art 161(3) EC); the European Parliamenťs Committee of Inquiry
(Art 193(3) EC); Ombudsman (Art 195(4) EC); Treaty basis for bilateral IIAs between Commission and
Council (Art 218(1) EC); Court of Auditors (Art 248(3) EC); budgetary procedure (Art 272(9) EC).
Amsterdam31
and Nice32
adopted Protocols or Declarations that referred to certain IIAs
or stipulated that IIAs may have to be concluded in certain policy fields if deemed
necessary.
If primary law explicitly authorises the Community institutions to adopt an IIA
the institutions involved usually have to operate within rather tight legal boundaries.
Generally speaking most `higher leveľ issues would have been solved already during
the IGC, thus leaving `only' technical specification to the IIA negotiators. As a corollary,
negotiations of that kind tend to revolve around `lower leveľ issues,33
i.e. making
procedures work or governing the performance of institutions, such as the Ombudsman.
However, this does not mean, that this kind of situation is less adversarial or
competitive. On the contrary, negotiations of `lower leveľ issues are dominated by
distributive bargaining, which implies that the parties involved try to get as much as
they can for themselves. Preferences of an actor/institution engaging in these negotiations
are fixed during the whole communicative process and negotiation is only about
costs and benefits of a certain negotiation outcome. Usually, the agreed content of this
zero-sum game is clear and precise, which indicates that the drafting parties intend to
commit themselves.
B Specification of Treaty Provision Without Explicit Authorisation
While vagueness can be regarded as an inescapable trait of language, it is especially
pertinent for legal language.34
Generally speaking, legislators are faced with the `impossibility
of foreseeing all possible combinations of circumstances that the future may
bring. . . . This means that all legal rules and concepts are "open"'.35
Because the EU
founding Treaties `lack . . . detailed rules governing relations among institutions',36
IIAs
are viewed as a `pragmatic answer' to arising coordination problems and conflicts stemming
from this institutional arrangement, as they allow circumvention of the timeconsuming
procedure of formal Treaty amendment.37
In absence of sufficiently precise
Treaty law, IIA negotiators tend to operate under lesser constraints than those who are
explicitly authorised by primary law. But lesser legal constraints could mean that fundamental
constitutional choices have not been made thus turning supposedly `lower
leveľ negotiations into `higher leveľ ones characterised by conflicting interests and
ideas. Consequently, negotiations of that kind are a mix of both distributive bargaining
and arguing on conflicting ideas. Similarly, the content of such IIAs in terms of
precision is difficult to predict and may vary. Because these IIAs do not derive directly
from an explicit Treaty provision, one has to analyse the intention of the concluding
parties in order to find out whether such an IIA can be considered legally binding
or not.
European Law Journal Volume 12
214  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
31
See Protocol No 30 on subsidiarity; Declaration No 33 on Art 188c(3) EC.
32
See Declaration No 3 on Art 10 EC; Declaration No 6 on Art 100 EC.
33
See Jillson and Eubanks, op. cit. note 28 supra, 438 (as quoted by Rittberger, op. cit. note 26 supra, at
677­680).
34
See G. C. Christie, `Vagueness and Legal Language', (1964) 48 Minnesota Law Review 885.
35
H. L. A. Hart, `Jhering's Heaven of Concepts and Modern Analytical Jurisprudence, in: H. L. A. Hart
(ed.), Essays in Jurisprudence and Philosophy (Oxford University Press, 1983) 265 at 269.
36
Snyder, op. cit. note 2 supra, 463.
37
Monar, op. cit. note 2 supra, 695­696.
C Pursuit of General Political Interests
As already mentioned, Community institutions are entitled to adopt IIAs even without
referring to a specific Treaty provision in order to pursue a given political goal. They
only have to take into account the rule that they are not allowed to amend Treaty law
through IIAs. These are the only legal constraints they face, which means that the negotiations
as well as the content of the respective IIA will deal with issues that have not
been tackled by Community or Union law before. Consequently, negotiations tend to
revolve around `unresolveď or `higher leveľ issues, which means that we can expect
conflict of interests and ideas among the institutions concerned. To put it differently:
Negotiations of that kind can be conceived as negotiations that happen typically before
parties are ready to legalise their commitment. This means that Community institutions
are not necessarily involved in zero-sum bargaining, but may find themselves
arguing about `higher leveľ issues such as general principles of democracy and the like.
This implies that the various positions may not always be fully developed or may still
be subject to intra-institutional debates. Negotiation outcomes therefore tend to be
rather general and vague, leaving a great margin of interpretation and discretion, and--
in combination with the lack of pertinent Treaty law--are thus closer to mere political
declarations that cannot be regarded as legally binding.38
IV Case Study: The 1993 Package of IIAs on Democracy, Transparency,
and Subsidiarity
In the course of the negotiations of the Maastricht Treaty, an `inter-institutional conference'
was set up enabling representatives of the Member States to discuss the Treaty
reform with the Parliament and the Commission. After the Maastricht Treaty had been
signed, the inter-institutional conference was recreated in order to tackle various
interrelated issues stemming from the new Treaty.39
Intense negotiations between
November 1992 and October 1993 led the European Parliament, the Commission and
the Council to adopt a package of IIAs revolving around `democracy, transparency,
and subsidiarity'. As said, this package is an excellent example in order to highlight
that IIAs fulfil three roles depending on their relationship to primary law and their
consequences for the negotiation style, content, and legal value. For each function we
have selected one IIA,40
namely (a) specification with explicit authorisation: the
Ombudsman, (b) specification without such an authorisation: subsidiarity and (c)
pursuit of general political need: democracy and transparency and focused on those
issues that have proved to be most controversial during the negotiations at the interinstitutional
conference.
A Ombudsman
The institutional design of the European Ombudsman is derived from Nordic constitutional
and administrative tradition. In the course of the second half of the twentieth
March 2006 Sub-Constitutional Engineering
 2006 The Authors 215
Journal compilation  Blackwell Publishing Ltd. 2006
38
However, it is--once more--important to recall Snyder's remarks that `what counts as "legally binding"
or "fully binding under Community law" is a highly complex concepť and that IIAs may also be a source
of information or an aid of interpretation of legally binding acts, see Snyder, op. cit. note 2 supra, 460­462.
39
R. Corbett, F. Jacobs and M. Shackleton, The European Parliament (John Harper, 2003), at 277.
40
For a detailed overview of the IIAs concerned see R. Corbett, The European Parliamenťs Role in Closer
EU Integration (Macmillan, 1998), 344­347; Bobbert, op. cit. note 1 supra, 24­34.
century Ombudsman offices have been installed in the majority of European nation
states.41
However, there are differences in how the office was adapted to the various political
systems. In general, Ombudsmen are non-judicial bodies established to safeguard
citizens' political, civil and social rights vis--vis national public administrations. As
parliamentary institution it acts as supplementary to parliamentary control.42
Thanks to a proposal by the Spanish government43
supported by Denmark44
the
European Ombudsman was--despite a widespread reluctance among most of the other
Member States--finally set up by the Maastricht Treaty. The position of the European
Parliament was also rather ambivalent. On the one hand, it has advocated the creation
of such an institution since the end of the 1970s45
but feared, on the other hand, that
the Ombudsman may rival its claim as the proper institution representing citizens' con-
cerns.46
The institutionalisation of the Ombudsman was part of a campaign trying to
persuade sceptics of political union, and was closely linked to the introduction of European
citizenship.47
Furthermore, it was designed to complement parliamentary control
over the EU administration, thus strengthening the democratic quality of the European
polity.48
The long-lasting `higher leveľ debates within the IGC led to a `very precise
definition of the role of the Ombudsman'49
in the Treaties, which left only comparably
`lower leveľ issues to be dealt with on the IIA level. The European Parliament was
explicitly authorised to take the initiative to draft and adopt a Decision with prior
approval of the Council that lays down the regulations and general conditions governing
the performance of the Ombudsman's duties.50
The inter-institutional negotiations came close to conclusion after only one session.
However, two issues remained unresolved and heavily debated:51
(a) the time limit for
referring complaints to the Ombudsman and (b) his/her access to documents of the
Member States in possession of the Commission. Each institution tried to secure its
European Law Journal Volume 12
216  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
41
P. Bonnor, `Ombudsmen and the Development of Public Law' (2003) 9 European Public Law, 237 at 237.
42
T. Läufer, in E. Grabitz and M. Hilf (eds.), Kommentar zur Europäischen Union, Art 138e EGV at 1.
43
Letter by the then Spanish Prime Minister, F. Gonzales, to the Members of the European Council,
4 May 1990, see A. G. Ibanez, `National Positions: Spain', in: F. Laursen and S. Vanhoonacker, The
Intergovernmental Conference on Political Union, (EIPA and Nijhoff, 1992) at 106; J. Söderman, `The
European Ombudsman, Report for the year 1995', available at
<http://www.euro-ombudsman.eu.int/report95/en/default.htm>.
44
The Danish delegation at the IGC issued a memorandum including Draft Treaty Articles on the Appointment
of an Ombudsman on 4 March 1991, see Laursen and Vanhoonacker, op. cit. note 43 supra, at 263.
45
European Parliament resolution OJ C 140/153 1979; see also the 1985 Report by the Adonnino Committee,
`A People's Europe', Bulletin EC, Supplement 7/85 at 21.
46
See Resolution of 7 April 1992 on the results of the Intergovernmental Conferences, OJ C 125/81 1992,
Doc. A3­123/92. The European Parliament saw the Ombudsman as a potential source of competition in
its function as guardian of citizens' rights. V. Reding (EPP, Luxemburg), the then spokeswomen of the
European Parliament Petitions Committee, declared that it was `a political manoeuvre which deprives
citizens of some of their rights' (Agence Europe, 16 May 1991; quoted in P. Magnette, `Between parliamentary
control and the rule of law: the political role of the Ombudsman in the European Union', (2003)
10 Journal of European Public Policy 677 at 680).
47
P. Leino, `The Wind is in the North, the First European Ombudsman (1995­2003)', (2004) 10 European
Public Law 333 at 334; W. Wessels and U. Diedrichs, A New Kind of Legitimacy for a New Kind of
Parliament--The Evolution of the European Parliament, European Integration online Papers, 6/1997, 7.
48
Magnette, op. cit. note 46 supra, at 678; Söderman, op. cit. note 43 supra.
49
Magnette, op. cit. note 46 supra, at 680.
50
See Art 195(4) EC.
51
C. Reich, `La mise en oeuvre du Traité sur ľUnion européenne par les accords interinstitutionnels', (1994)
375 Revue du Marché Commun et de ľUnion européenne 81 at 84 (note 7).
distributive gains based on a set of unitary and predefined preferences. The agreed procedural
details are consequently precise and will most probably prove rather resistant
to future attempts of `creative' interpretation. They thus provide a clear and enforceable
set of rules representing high legal value.
a) Time Limit
Besides a number of minor procedural issues,52
a major controversy revolved around
the time limit for cases to be brought to the Ombudsman. Obviously, it was in the interest
of the European Parliament not to have any time limitation on the filing of complaints
or--if this could not be achieved--to come as close as possible to this ideal.
Conversely, the Council, and to a lesser extent the Commission, pleaded for a preferably
short time limit because they assumed that they would be the subject of most of
the expected complaints.
The first parliamentary draft53
reflected its maximum demand and received fierce
opposition from the two other institutions that demanded a time limit of one year,54
pointing to the extent of workload and practicality. In response, the European Parliament
proposed a time limit of five years, arguing that this would correlate with the
Commission's term of office, as well as with the time limit for actions before the Court
of Justice on the grounds of non-contractual liability. Both Commission and Council
rejected the European Parliamenťs arguments and insisted on their initial demands.55
During the negotiations, the Commission and the European Parliament became more
flexible and were ready to agree on a three-year time limit.56
But the Council refused
to join the offered compromise, fearing that it would be the institution most affected
by the Ombudsman's activities. Nonetheless, the Council also showed willingness to
forge a compromise and extended its proposal to a two-year time limit for issuing
complaints to the Ombudsman.57
Eventually, the Counciľs last offer proved to be
March 2006 Sub-Constitutional Engineering
 2006 The Authors 217
Journal compilation  Blackwell Publishing Ltd. 2006
52
E.g. should the Ombudsman take his/her oath of office in front of the European Parliament or the Court
of Justice?, see Corbett, op. cit. note 40 supra, at 345.
53
Résolution A3-0298/92, Annex: `Décision du Parlement Européen concernant le statut et les conditions
générales ďexercice de ses fonctions du médiateur européen', Jeudi 17 décembre 1992, OJ C 21/142­
147 1992.
54
Monar, op. cit. note 2 supra, at 707; see also J. Delors, the then President of the Commission, who
`rappelle ľimportance que revt la discussion pour la Commission en tant qu'institution devant tre la
plus souvent sollicitée par le médiateur. . . . Il faut établir un délai fixe ďun maximum ďun an pour saisir le
médiateur. . . . M. Delors attire également ľattention sur la nécessité de tenir compte des problmes
budgétaires afférents  la charge de travail du médiateur.' (Conférence Interinstitutionnelle, Compte Rendu
Analytique des Débats, 7 June 1993, PE 205.271).
55
The then Counciľs President N. Helveg Petersen `estime que la comparaison avec la Cour de Justice n'est
pas pertinente, ďautant que si le délai ďun ans pour saisir le médiateur, est dépassé, il est toujours possible
de saisir la Cour de justice'. (Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
7 June 1993, PE 205.27 1)
56
Comment by MEP Baron: `Pour ce qui est du délai du saisine, le délai de trois ans suggéré par Mme Vayssade
a déj convaincu M. Delors'. (Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
7 June 1993, PE 205.27 1)
57
`M. Helveg Petersen `déclare que beaucoup ďÉtats membres jugent excessif ce délai de trois ans et qu'en
définitive, il s'agit de réflexions sur la méthode de travail la plus adéquate pour le médiateur', whereas MEP
Baron `considre que la proposition du Conseil s'apparente  du marchandage et qu'il ne motive pas sa proposition
ďun délai de deux ans'. The then Council President N. Helveg Petersen put pressure on concluding
this issue by pointing at `la nécessité de finir le plus rapidement possible vu ľordre du jour trs chargé du
Conseil pour les deux journées qui viennenť. (Conférence Interinstitutionnelle, Compte Rendu Analytique
des Débats, 7 June 1993, PE 205.271).
acceptable for all parties involved.58
This successive exchange of demands confirms that
these procedures followed the logic of distributive bargaining typical for `lower leveľ
issues in IIA negotiations. The content of the outcome can be regarded as precise and
thus easily enforceable.59,60
b) Confidentiality/Secrecy
Negotiations on the Ombudsman's access to Member States' documents in the possession
of the Commission were even more controversial. Similar to the first case, the
Community institutions had clear and fixed preferences. In order to strengthen its grip
on the other two institutions, especially on the Council, the European Parliament
wanted the Ombudsman to have access to as many documents as possible expressly
excluding only secret documents.61
The Council, as the institution whose politically
sensitive documents were at stake, held a more restrictive view. It advocated that even
confidential documents should not be handed to the Ombudsman.62
The Commission
did not engage actively in this dispute, but may be interpreted as having an interest in
shedding light on the Counciľs internal documents, and thus to support the European
Parliamenťs position.63
Finally, the institutions reached a political compromise obliging the Member States
`to provide the Ombudsman, whenever he may so request . . . with any information
European Law Journal Volume 12
218  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
58
See Art 2(4) Decision of the European Parliament of 9 March 1994 on the regulations and general
conditions governing the performance of the Ombudsman's duties (OJ L 113/15 1994).
59
It has to be noted that the term `enforceable' must not be confused with `legally binding'. It just points to
the fact that the precise content of the negotiation outcome, namely the two year time limit, allows the
Ombudsman to declare complaints which exceed this time limit inadmissible (in 1996: 12 complaints were
declared inadmissible on this ground; 1997: 4; 1998: 6; 1999: 5; 2000: 2; 2001: 7; 2002: 10).
60
However, the first annual report of the European Ombudsman highlights problems in strictly applying the
two-year time limit. It argues that `it would be harsh to apply strictly the provision in Article 2(4) of the
Statute' and underlined its argument with a reference to national Ombudsman systems advocating the possibility
to `waive such time limits where it is necessary to do so in the interests of justice', see Söderman,
op. cit. note 43 supra. If a complaint is made after the date on which the facts underlying it came to the
attention of the person lodging the complaint, the Ombudsman can decide to conduct further inquiries
on his/her own initiative, see e.g. J. Söderman, `The European Ombudsman, Report for the year 1996', at
68, available at <http://www.euro-ombudsman.eu.int/report96/pdf/en/rap96_en.pdf>.
61
The then European Parliament President E. Klepsch referred to `la différence entre la notion de confidentialité,
utilisée par les fonctionnaires et ne figurant nulle part dans les traités, et celle de secret, qui y figure.
Le pouvoir ďun fonctionnaire de qualifier de confidentielle une information ne peut pas constituer une position
politique et ne peut pas tre acceptée par les citoyens comme moyen de priver le médiateur ďune telle
information'. (Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 7 June 1993,
PE 205.271.); see also Résolution A3-0298/92, Annex: `Décision du Parlement Européen concernant le
statut et les conditions générales ďexercice de ses fonctions du médiateur européen', 17 December 1992,
OJ 1992 C 21/142­147, at 144, Art 3(2): `Les institutions et organes communautaires sont tenues de
fournier au médiateur les renseignements demandés et lui donner accs aux dossier concernés. Ils ne
peuvent s'y refuser en opposant le secret.'; see also Art 4 (1): `Le médiateur . . . est tenu de ne pas divulguer
les informations et pices confidentielles dont il a eu connaissance dans le cadre de ses enqutes'.
62
The then Council President N. Helveg Petersen pointed to the `position commune du Conseiľ and `précise
que le Conseil voudrait voir rétablis les deux concepts de secret et de confidentialité seulement lorsqu'il s'agit
de documents émanent ďun Etat membre' and makes clear that `[l]e Conseil n'est pas en mesure de faire
des concessions sur ce poinť (Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
7 June 1993, PE 205.271).
63
`J. Delors estime que la Commission peut trs bien travailler avec la notion du "secret" car effectivement la
"confidentialité" est une notion subjective' (Conférence Interinstitutionnelle, Compte Rendu Analytique des
Débats, 7 June 1993, PE 205.271).
. . . unless such information is covered by laws or regulations on secrecy or by provisions
preventing its being communicateď.64
While a first superficial reading of this
result might suggest that the European Parliament has succeeded in the IIA negotiations,
it is in fact the Council that accomplished a victory over the European Parliament,
because it is left to the Member States' discretion to refuse the communication
of a given document.65
The two controversial issues with regard to the Ombudsman showed that the main
adversaries were the Council and the European Parliament. Both institutions came up
with fixed and precise preferences reflecting the institutional roles as laid down in the
Treaties. The Commission was more ready to compromise as to the time limit and had
no clear and strong interest with regard to the issue of confidentiality/secrecy. The
results of the negotiations showed that--contrary to widespread expectation in the pertinent
literature--it was not the European Parliament but the Council that succeeded
in the issues discussed. That does not mean that the European Parliament was not in
a better position vis--vis the two other institutions, especially vis--vis the Council,
than before the Maastricht Treaty. Far from it! But it primarily benefited from the
favourable legal framework as set out in the Treaty and less from the IIA itself. With
regard to the legal consequences, we argue that the IIA on the Ombudsman is legally
binding among the institutions involved as well as with regard to third parties, not only
because it was explicitly authorised by Treaty law, but also because of its precise
content. Based on the latter we conclude that the institutions involved intended the IIA
to have a legal effect.
B Subsidiarity
Subsidiarity has been on the European Agenda since as early as the 1970s and 1980s.66
But only with the Treaty of Maastricht has this principle explicitly been recognised in
Community law.67
As the project of European integration became more political, some
Member States proposed subsidiarity as a tool to contain the perceived increase in
power of the Commission. Coming from opposite angles,68
Germany along with the
United Kingdom joined forces and spearheaded the effort to incorporate subsidiarity
into the Treaty. However, fundamental ideational differences between the Member
States, especially between the British and German governments, on the precise meaning
of subsidiarity led to a vague formulation of the relevant Treaty provision. In fact, it
lacked precise criteria for its application. The subsidiarity principle as determined by
the Treaty rather `tends to describe an abstract goal than a method of achieving iť.69
In the wake of the negative Danish referendum on the Maastricht Treaty in June
1992, the political context changed. While Treaty amendments were out of the question,
the political need to specify the vague concept of subsidiarity was so strong that
March 2006 Sub-Constitutional Engineering
 2006 The Authors 219
Journal compilation  Blackwell Publishing Ltd. 2006
64
See Art 2(3) Draft Decision of the European Parliament on the Regulation and general conditions
governing the Ombudsman's duties (OJ C 329/138 1993).
65
Ibid.
66
For an historical overview see K v. Kerksbergen and B. Verbeek, `The Politics of Subsidiarity in the
European Union', (1994) 32 JCMS 215, at 217­218.
67
See Art 5 EC.
68
A. L. Teasdale, `Subsidiarity in Post-Maastricht Europe', (1993) 64 The Political Quarterly 187, at 191.
69
G. A. Bermann, `Subsidiarity and the European Community', (1993) 17 Hastings International and Comparative
Law Review 97, at 103 (note 28).
the Member States along with the Community institutions looked for `extra-Treaty
safeguards against the growth of Community power'.70
Following a number of activities
by all relevant institutions,71
the Presidency Conclusions of the European Council
in Edinburgh (December 1992) included a declaration on subsidiarity that was meant
to serve as a basis for an IIA72
between the European Parliament, the Council, and the
Commission.73
The IIA negotiations revolving around subsidiarity are difficult to grasp. The following
examples serve to illustrate the negotiation style within the field of subsidiarity:
(a) the general debate on the principle of subsidiarity shows not only the `higher
leveľ negotiation style but also the insurmountable differences among the parties concerned;
(b) during the negotiations the question emerged what kind of institutions
should deal with difficulties possibly arising in the future. As these matters were by no
means ambitious and tackled the topic rather superficially, the institutions were soon
ready to reach a compromise. The rapid adoption of the IIA was primarily inhibited
by the parliamentary demand--backed by the Commission74
--to bind the IIA on subsidiarity
to another IIA on transparency and democracy.75
a) General Debate on the Concept of Subsidiarity
Because the Member States were not able to settle their ideational differences on subsidiarity
at the IGC, their `higher leveľ disagreements continued during IIA negotia-
tions.76
Or, to put it differently, deep ideational conflicts that could not be resolved by
the Member States at an IGC, could also not be solved during IIA negotiations.
European Law Journal Volume 12
220  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
70
Teasdale, op. cit. note 68 supra, at 193.
71
See e.g. the Memorandum presented by the then Commission President J. Delors shortly after the negative
Danish referendum, the Declaration of the European Council meeting in Lisbon (June 1992), the socalled
Birmingham Declaration on a `Community Close to Its Citizens' (October 1992), the European
Parliamenťs intensive preparatory work that formed the basis of the Commission's initiatives (see Conférence
Interinstitutionnelle, Compte Rendu Analytique des Débats, 10 November 1992, PE 202.939/CI/2),
e.g. a parliamentary resolution on the application of the subsidiarity principle including procedural details,
18 November 1992, PE 162.877/RC1/déf., 162.883/RC1/déf.
72
Monar, op. cit. note 2 supra, at 705.
73
Conclusions of the Presidency, Edinburgh, 12 December 1992, SN 456/92 at 4.
74
The then Commission President J. Delors was inclined to agree to a global version of an IIA. Furthermore,
he tried to link the negotiations on subsidiarity, transparency, and democracy with budgetary issues.
See Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 7 June 1993, PE 205.271.
75
In a note to the then Director of the EP Presidential Cabinet, J. Dohmes, it was noted that `la négociation
sur ľapplication du principe de subsidiarité soit élargie aux problmes de la transparence et á certains
aspects du dossier démocratie;  ce propos, le Conseil devrait accepter une négociation sur ľensemble du
dossier; ďailleurs, le Conseil européen ďEdimbourg s'est penché sur ces aspects (au moins sur deux, transparence
et subsidiarité)' (Commission institutionnelle, Sécretariat, `Note a ľattention de M. Dohmes', no
date). The Council Presidency was caught by surprise with this new parliamentary strategy: `Both Mr.
Helveg Petersen and Mr. strm Mller expressed regret that Parliament now seemed to be linking various
issues. This did not correspond to earlier indications from Mr. Klepsch according to which Parliament
was prepared to sign immediately the [EP] text on subsidiarity' (General Secretariat of the Council, Internal
Note on the meeting in Copenhagen on March 29 between the Presidency (Mr Helveg Petersen and
Mr strm Mller) and Mr Oreja, President of the European Parliamenťs Institutional Committee,
30 March 1993).
76
For example, the then Italian Under-secretary on Domestic Affairs, V. Spini, `souhaite que le document de
la présidence du Conseil soit approfondi et rediscuté. Le document de la Commission apparaît en revanche
comme un point de référence positif' (see Conférence Interinstitutionnelle, Compte Rendu Analytique des
Débats, 10 November 1992, PE 202.939/CI/2). On Member states ideological positions regarding the interpretation
of subsidiarity see van Kersbergen and Verbeek, op. cit. note 66 supra. The then British Council
The European Parliament as well as the Commission feared that subsidiarity may
entail an intergovernmentalist turn,77
and advocated concrete procedural measures,
including an annual public debate on a Commission report on the application of the
subsidiarity principle.78
Furthermore, the European Parliament linked its assent for the
IIA on subsidiarity to the signing of an additional IIA on transparency and democracy.
While the Council was ready to agree to rather general implementation measures regarding
the Community legislative process, it rejected the parliamentary demand for a global
agreement, including transparency and democracy measures.79
Though the Commission
supported the position of the European Parliament,80
it took an observer position in
further discussion.81
Despite the lack of agreement on subsidiarity itself, the Community
institutions managed to adopt some modest implementing measures. The outcome
can be summarised as an inter-institutional commitment to `take into account the principle
of subsidiarity and show that it has been observeď. Each institution `shall, under
their internal procedures, regularly check that action envisaged complies with the provisions
concerning subsidiarity'. Additionally, the Commission `shall draw up an annual
report for the EP and the Council on compliance with the principle of subsidiarity'.82
To
make a long story short: the European Parliament did not succeed with its claim.
b) Conference versus Meeting
As a result of the predominant focus on `higher leveľ issues, distributive bargaining
only occurred to a minor degree in inter-institutional negotiation on the subsidiarity
principle.83
The only controversial `lower leveľ issue revolved around the denomination
of the institutional forum designed to settle future difficulties and entitled to amend
March 2006 Sub-Constitutional Engineering
 2006 The Authors 221
Journal compilation  Blackwell Publishing Ltd. 2006
President Garel-Jones stated: `Le fonctionnement de ce principe est difficile  saisir dans des termes théorique
mais on peut reconnaîtré la nécessité de ľappliquer dans des cas concrets. . . . Ľapproche de la présidence est
ďexaminer ľArticle 3B paragraphe par paragraphe et mme mot par mot pour voir comment il peut devenir
réalité' (see Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 26 November 1992,
PE 202.943).
77
MEP Herman argued that `il semble que ľon assiste  un glissement vers ľintergouvernemental, alors que
celui-ci a fait preuve de son inefficacité' (see Conférence Interinstitutionnelle, Compte Rendu Analytique
des Débats, 26 November 1992, PE 202.943).
78
MEP J-P. Cot underlined that `En focalisant ľattention sur des questions de procédures, le Parlement
européen ne doit justement pas tre trop ambitieux mais plutôt modeste. Les questions de fonds seront
clarfiées dans la pratique' (see Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
10 November 1992, PE 202.939/CI/2).
79
The then German Minister for European Affairs, Mrs Seiler Albring, expressed her satisfaction that `les
négociations sur la subsidiarité soient sur le point de se conclure. . . . Les thmes de la démocratie et de la
transparence sont importants, mais les propositions soumises par le Parlement dans son projet vont trop
loin. La premier pas doit tre la conclusion ďun accord institutionnel sur la subsidiarité exclusivemenť (see
Conférence Interinstitutionelle, Compte Rendu Analytique des Débats, 7 June 1993, PE 205.27 1).
80
See Commission des Communautés Européennes, `Le Principe de la Subsidiarité, Communication de la
Commission au Conseil et au Parlement européen', 27 October 1992, SEC (92) 1990 final, and also Conférence
Interinstitutionnelle, Compte Rendu Analytique des Débats, 10 November 1992, PE 202.939/CI/2;
for a discussion on the IIA concerning democracy and transparency, see the following section.
81
The then Commission President J. Delors `annonce que la Commission n'a aucun problme ni divergence
avec les quatre principes énoncé par le Parlement. La Commission  ce stade des travaux préfre écouter les
échanges de vues qui vont tre exprimés par les différents intervenants' (see Conférence Interinstitutionnelle,
Compte Rendu Analytique des Débats, 26 November 1992, PE 202.943).
82
`Draft Interinstitutional Agreement between the European Parliament, the Council and the Commission
on procedures for implementing the principle of subsidiarity', OJ C 329/135­136 1993.
83
Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 10 November 1992, PE 202.939/
CI/2 and Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 26 November 1992,
PE 202.943.
the IIA. The European Parliament had a natural interest in adding weight to this interinstitutional
talk and preferred the term `interinstitutional conference'. Trying to downplay
the importance of the entire IIA negotiations, the Council advocated the more
low-key term `interinstitutional meeting'.84
The Commission was hardly engaged in the
debate. At the end of the day, the European Parliament succeeded and the term `interinstitutional
conference' was adopted. This was mainly because it was already used to
denominate the inter-institutional dialogue preceding IGC negotiations.85
Although the Treaty did not explicitly authorise the adoption of an IIA on subsidiarity,
the Community institutions felt the political as well as legal need to specify
this deliberately vague Treaty provision. Contrary to the Ombudsman, IIA negotiations
on subsidiarity revolved around conceptual core elements which is characteristic of
`higher leveľ debates. Because of fundamental disagreements among the Member
States, the Community institutions were not able to reach a substantial compromise,
i.e. to achieve a sufficiently precise and workable definition of subsidiarity (e.g. precise
criteria against which a legislative act has to be tested in order to fulfil the subsidiarity
provision). The confusion about the normative concept behind the Treaty provisions
prevented distributive bargaining on other `lower leveľ issues than the denomination
of the inter-institutional forum. This means that IIAs do not have the power to specify
Treaty provisions, which were deliberately kept vague because of fundamental differences.
To put it bluntly, IIA negotiations cannot act as a substitute for a missing
political consensus at the IGC.86
Consequently, the IIA on subsidiarity only includes
rather harmless and uncontroversial issues, thus circumventing the tricky task of
conceptual clarification. Moreover, it seems that even these minor measures do not have
legal effects. At least the text states that compliance with subsidiarity `shall be reviewed
under the normal Community process, in accordance with the rules laid down by the
Treaties'. It can therefore be argued that the Community institutions did not intend to
create new legal rights and duties.87
C Transparency and Democracy
In the aftermath of the first Danish referendum on the Treaty of Maastricht, the European
Parliament seized the opportunity opened up by the Counciľs preference to adopt
an IIA on subsidiarity, and successfully linked its assent to this IIA to the adoption of
another IIA on transparency and democracy.88
Because these elusive concepts were not
European Law Journal Volume 12
222  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
84
The then Foreign Minister of Luxemburg, Mr J. Poos, argued that `le terme de conférence, plus formel, doit
tre réservé, par exemple, pour la révision des traités' (see Conférence Interinstitutionnelle, Compte Rendu
Analytique des Débats, 7 June 1993, PE 205.271).
85
Corbett, Jacobs and Shackleton, op. cit. note 39 supra, 277; Corbett, op. cit. note 40 supra, 294­296;
M. Westlake, A Modern Guide to the European Parliament (Pinter, 1994), at 37.
86
This theoretical argument is confirmed by the then French Minster for European Affairs, Mrs Guigou,
stating that `il faut éviter de précipiter les choses si les idées ne sont pas claires' (see Conférence Interinstitutionnelle,
Compte Rendu Analytique des Débats, 10 November 1992, PE 202.939/CI/2).
87
The Protocol No 30 on subsidiarity added to the Amsterdam Treaty can be regarded as a specification as
well as proceduralisation of the subsidiarity principle as laid down in Art 5 EC and the IIA concerned.
However, it remains highly doubtful whether this attempt may substantially change the low legal value of
this principle; see e.g. A. G. Toth, `Is Subsidiarity Justiciable?', (1994) 19 European Law Review 268.
88
Reich, op. cit. note 51 supra, at 81; Monar, op. cit. note 2 supra, at 706; see also European Parliament,
`Résolution sur ľétat de ľUnion européenne et de la ratification du Traité de Maastrichť, par. 4,
OJ C 299/8­10 1992, at 9; `Avant-projet de Déclaration Solennelle du Parlement Européen, du Conseil et
de la Commission sur la Transparence et la Démocratie', Bruxelles, le 26 novembre 1992.
delimited on the level of primary law beforehand, they raised controversial `higher leveľ
debates stemming from a pot-pourri of divergent interests and ideas even within the
particular institutions. The major contested issues were found to be (a) the degree to
which Council debates should be opened to the general public and (b) a commitment
of the Council demanded by the European Parliament not to adopt legislative texts
previously rejected by the Parliament.89
In contrast to the IIAs on the Ombudsman and subsidiarity, the negotiations on
transparency and democracy were neither based on an explicit Treaty authorisation nor
on a perceived need to specify vague Treaty provisions. As a consequence, an IIA on
these issues ran the risk of creating primary law--which is legally not possible. Moreover,
such negotiations tend to be `higher leveľ ones dealing with issues that have never
been tackled or could not be agreed upon at an IGC.
Similar to the negotiations on subsidiarity, the debates on transparency and
democracy did not succeed in clarifying these concepts but concentrated on concrete
issues supposed to fall under the broad and elusive category of `transparency and
democracy'.
a) Publicly Held Council Debates
The European Parliament that insisted on publicity of Council debates on co-decision
matters was confronted with a Council that had difficulties to find a common view on
this issue which weakened its bargaining position vis--vis the European Parliament.
However, the Counciľs lowest common denominator was to publish its voting results.
While some Member States wanted to go further,90
others feared that publicly held
debates would block the decisional process thus leading to inefficiency.91
Eventually, the
Council agreed inter alia to `open some of its debates to the public'. An amendment
of the Counciľs rules of procedure provided for `retransmission by audiovisual means'
of policy debates on the six-monthly work programme of the Council Presidency and
the Commission's annual work programme. Furthermore, other debates on `important
issues' could be subject to public retransmission depending on a unanimity vote of the
Council.92
March 2006 Sub-Constitutional Engineering
 2006 The Authors 223
Journal compilation  Blackwell Publishing Ltd. 2006
89
See Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats, 10 November 1992,
PE 202.939/CI/2. Both demands had already been mentioned in a parliamentary resolution from
14 October 1992 which further included other parliamentary requests as regards transparency and democracy
such as adopting a legislative programme, the codification of Community law, a unitary voting
system for the European Parliament or the simplification of the comitology system see `État de ľUnion
européenne et Traité de Maastrichť, OJ C 299/8 1992.
90
The then Danish Minister for European Affairs, N. Helveg Petersen, declared that `certaines délégations,
en particulier la sienne, estiment elles aussi que les rgles sur la publicité des votes sont insuffisantes et qu'elles
doivent tre améliorées' (see Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
25 October 1993, PE 207.009).
91
The then Foreign Minster of Luxemburg, J. Poos, gave several reasons for the refusal of publicly held
Council debates `un danger de surenchre pouvant conduire a un report permanent des compromis nécessaires
 la décision;--une tendance  voir les problmes complexes traites en dehors du Conseil lui-mme,  ľimage
de ce qui se passe au Conseil de sécurité des Nations Unies, processus conduisant  une marginalisation
des petits pays.--un risque de transfert de la prise de décisions  un niveau inférieur tel que le Comite
monétaire et le COREPER', (see Conférence Interinstitutionnelle, Compte Rendu Analytique des Débats,
26 November 1992, PE 202.943).
92
`Council Decision of 6 December 1993 adopting the Counciľs Rules of Procedure', OJ L 304/2 1993.
b) Council Commitment not to Adopt a Legislative Proposal Previously Rejected
by the European Parliament
The European Parliament was aware of the fact that it could not amend the Treaties
by an IIA. It therefore tried to argue that its demand that the Council shall be no longer
entitled to adopt a legislative proposal that has already been rejected by the European
Parliament could be based on an extensive interpretation of existing Treaty provisions.93
However, this argument clearly transgressed the Treaties. As a consequence the Council
dismissed the European Parliamenťs demand during these IIA negotiations94
only to
accommodate it in the Amsterdam Treaty. This underlines, once again, our thesis that
the IIA per se does not necessarily strengthen the role of the European Parliament. It
can do so only if the European Parliament succeeds to base its claim raised within the
setting of IIA negotiations on pertinent Treaty law. That does not mean, however, that
an IIA cannot be utilised to press the Member States politically with the aim to get the
desired Treaty amendment later on.95
In sum, the outcome of the inter-institutional negotiations on transparency and
democracy confirm our theoretical assumption about IIA negotiations with regard to
controversial `higher leveľ issues in absence of primary law. Since it has no grounding
in Treaty provisions, the Interinstitutional Declaration on Democracy, Transparency
and Subsidiarity is largely vague, imprecise, and resembles more a mere political declaration
without any significant legal value.96
V Conclusion
The growing number of IIAs that have been adopted over the recent past indicates that
this instrument--though not entirely new--has gained more and more importance in
the EU's legal and political system. Depending on their relationship to Treaty law, IIAs
serve different roles in the EU's institutional architecture. Against this backdrop, we
challenge the assumption that IIAs usually strengthen the role of the European Parliament.
As our case study on the negotiation process and its outcome illustrates, the
European Parliament is by no means necessarily the only institution which benefits
from IIAs--irrespective of the roles IIAs fulfil. Generally speaking, the Maastricht
Treaty increased the power of the European Parliament. As IIAs operate within the
rules set by the Treaty, they reflect this increased power but are not primarily responsible
for it. Though the European Parliament may often initiate IIAs in order to
strengthen its power vis--vis the other Community institutions it has not necessarily a
privileged bargaining position--let alone the power to impose its interests and ideas
upon the other relevant institutions.
European Law Journal Volume 12
224  2006 The Authors
Journal compilation  Blackwell Publishing Ltd. 2006
93
The then European Parliament President, E. Klepsch, made clear that, `il ne s'agit pas de modifier le traité,
mais de ľappliquer ďune façon non restrictive et de ľinterpréter de façon positive' (see Conférence Interinstitutionnelle,
Compte Rendu Analytique des Débats, 7 June 1993, PE 205. 271).
94
The then Foreign Minister of Luxembourg, J. Poos, declared that `La proposition selon laquelle `le Conseil
s'engage  ne pas adopter de texte législatif qui aurait été rejeté auparavant par le Parlemenť: celle-ci est en
contradiction avec la procédure retenue par le Traite lui-mme' (see Conférence Interinstitutionnelle, Compte
Rendu Analytique des Débats, 26 November 1992, PE 202.943).
95
S. Hix, `Constitutional agenda-setting through discretion in rule-interpretation: why the European Parliament
won at Amsterdam', (2003) 32 British Journal of Political Science 259 and R. Corbett, `Academic Modeling
of the Codecision Procedure: A Practitioner's Puzzled Reaction', (2000) 1 European Union Studies 73.
96
Bobbert, op. cit. note 1 supra, at 148; see also Lindner and Rittberger, op. cit. note 25 supra, at 451.
This can be underlined by a closer look at the IIA's negotiation process and its
outcome. In case of those IIAs that are explicitly authorised by the Treaty, the European
Parliament, like the other institutions, is engaged in distributive bargaining. As a
corollary, the bargaining power of each institution is the crucial variable to explain why
one institution failed or succeeded in the negotiations concerned. Taking this into
account, it is impossible to argue that the European Parliament always wins IIA negotiations
or strengthens its role through IIAs. This is even more obvious if we turn our
attention to the two other roles that are performed by IIAs. Both the specification of
vague Treaty provisions without an explicit authorisation and the initiative to conclude
an IIA without basing the political claims on a Treaty provision illustrate that it is more
than unlikely that the European Parliament will succeed in these kinds of negotiations.
As the IIA on subsidiarity shows, the decisive differences among the Member States
resulting in a more than vague Treaty provision could not be substantially bridged by
arguments brought forward by the European Parliament. The same holds true for our
last category. While the European Parliament successfully linked its assent to the IIA
on subsidiarity to the adoption of another one on democracy and transparency, a closer
look at both IIAs reveals that their content was rather vague and imprecise.
But the different roles of IIAs do also influence the content of IIAs in terms of precision
and--as a consequence--its legal value. If Treaty provisions underlying the IIA
are already precise, the IIA negotiations tend to produce a precise and thus legally
relevant content. This point is supported by the IIA negotiations on the Ombudsman.
These tackled very concrete issues, such as the exact time limit of complaints and the
right to deny access to documents on the grounds of secrecy or even confidentiality.
Conversely, the negotiations on the other two IIAs revolved around more elusive
concepts such as subsidiarity, transparency, and democracy. The content of such IIAs
are thus likely to be legally ambiguous or even irrelevant at all. Nevertheless their
long-term political consequences may be considerable, and even lead to future Treaty
amendments.
First submitted June 2005
Final revision accepted November 2005
March 2006 Sub-Constitutional Engineering
 2006 The Authors 225
Journal compilation  Blackwell Publishing Ltd. 2006