The cases

The article examines why and how institutions are transplanted to the EU by concentrating on two cases. The first concerns the transfer of the public audit institution, which led to the creation of the European Court of Auditors (ECAs), while the second pertains to the transfer of the ombudsman institution, which produced the office of the European Ombudsman (EO).

Why these two cases? First, both the EO and the ECA belong to the class of major EU institutions.Footnote 1 They have the status of ‘institutions of the union’ in the EU Treaties, which serve as the constitutional documents of the EU. This title makes them legally equivalent to better-known bodies, such as the European Commission, the EP or the European Court of Justice. It is therefore interesting to explore how these significant structures were put in place. Although the existing literature has paid attention to the establishment of the two institutions (Marias, Reference Marias1994; Laffan, Reference Laffan1997, Reference Laffan2003; Magnette, Reference Magnette2003), past studies lack a rigorous theory-driven explanation. This study aims not only at explaining why and how the ECA and the EO were established, but also, through theory, at drawing general lessons from these two cases that apply beyond the particular context that gave birth to those two organizations.

Second, and most importantly, the ECA and the EO are the only principal institutions of the EU, which were not put in place by the Treaties of Paris and Rome, the original Treaties, which were signed in the 1950s and first gave life to the European Communities, the forbearers of the EU. Instead, they were added to the European Communities (and eventually bequeathed to the EU) by separate treaties many years later: the ECA was added in the mid-1970s, while the EO in the early 1990s. This time lapse is very significant because the conditions under which the original institutions were created differ considerably from the conditions under which new ones were established. More specifically, the original institutions of the Communities were created in a purely intergovernmental environment, namely, one in which only the preferences of national governments mattered. By contrast, the EO and the ECA were established in an environment in which supranational actors, such as the European Commission and the EP, operated with a significant degree of independence from member states. In fact, they were using the powers available to them by the Treaties in order to act on their own preferences and modify their relative position in the EU structure. The conditions under which the EO and the ECA were created are more likely to be replicated in the future compared to the purely intergovernmental environment in which the original institutions were created. Therefore, by studying how these two institutions were established, we can draw lessons about the transfer of institutions in the future and the evolution in the EU more generally.

A historical institutionalist explanation

This article addresses two principal questions about institutional transfers to the EU. First, why are institutions transferred? I answer by examining the preferences of supranational and national actors who are active in the institutional environment of the EU. The paradigm that informs actor preferences for this study is HI (Hall and Taylor, Reference Hall and Taylor1996). For HI, actors situated in an institutional setting have preferences endogenous to that environment. In this particular study, for example, what drives actor behavior is to maintain, at a minimum, or advance, if possible, their position against each other in the setting of the EU. In other words, actors seek to shape the EU environment to their advantage.

We infer then that actor preferences on a potential institutional transfer may very well be in conflict. In those instances, it should be very difficult to change the status quo by introducing a new institution into the system. As a result, the status quo will tend to prevail and the institutional environment can be described as ‘path-dependent’ (Pierson, Reference Pierson2000). Nevertheless, the resilience of the status quo does not mean it cannot change. For HI change occurs at ‘critical junctures’ (Pierson, Reference Pierson2000). During those extraordinary periods, either actor preferences converge in favor of change or actors who favor change manage to overpower those who do not. In this article critical junctures, that is, opportunities for institutional transfer appear when new treaty texts are negotiated. At the same time, the article illustrates an additional historical institutionalist claim (Streeck and Thelen, Reference Streeck and Thelen2005) that change is not abrupt and unexpected. Events that take place between critical junctures are also important because gradually they set the foundations for status quo reversal. For example, in the cases studied here, the actions of the European Commission and the EP before treaty negotiations begun were key in building up momentum for institutional transfer.

Isomorphic theory

The second principal question this article addresses is: how does institutional transfer occur? Put differently, what are the processes that make institutional change possible during critical junctures? To understand the mechanisms that translate actor preferences into specific outcomes this study tests the theory of institutional isomorphism (DiMaggio and Powell, Reference DiMaggio and Powell1983). While HI is a paradigm, which includes general principles about institutional nature and change, isomorphic theory provides the conceptual tools to describe and understand the particular dynamics of institutional transfer. Furthermore, as a set of general theoretical propositions, it allows us to generalize beyond the particular cases examined in this study and understand the process of institutional transfer more generally.

The theory originates in organizational sociology. In their seminar article on organizational behavior, DiMaggio and Powell (Reference DiMaggio and Powell1983) sought to explain why so many organizations look so similar, especially when they are tied together with relations of interdependence. DiMaggio and Powell perceived organizations as systems consisting of various component structures and came to the conclusion that organizational homogeneity is caused by three processes, each one of which relies on the exercise of coercive, mimetic and normative pressures, respectively. The outcome of these processes is that some organizations adopt structural components from their peers. In other words, organizations look similar because they copy each other – willingly or not. Moreover, the reason particular structural components are singled out for copying is because they are perceived as appropriate, that is, the new structures allow the recipient organization to conform better to normative demands in its environment. Despite its sociological origins, isomorphic theory holds explanatory potential for the study of political systems. In fact, research indicates that governmental institutions are more susceptible to isomorphic pressures compared to other types of institutions because of their increased need to appear as espousing accepted norms (Frumkin and Galaskievicz, Reference Frumkin and Galaskievicz2004).

The combination of HI and isomorphic theory allows us to make the argument that actors seek to advance or maintain their relative position in their environment by exercising isomorphic pressures for the transfer of normatively appropriate institutional models. In other words, isomorphism allows us to understand how actor preferences are bounded by normative considerations. In this study, coercive, mimetic and normative pressures are perceived as vehicles for the transfer of institutional structures from one political system to another. The specific actors that initiate and advocate these transfers justify them as desirable. The main condition for the three forces to yield results is that the political systems among which they are exercised need to maintain relations and depend on each other. Put differently, the theory cannot explain similarities between political systems, which are independent of each other.

For this study more specifically isomorphic pressures are understood in the following way. First, coercive pressure is applied when political actors, such as a national government or the military, in one political system compel another political system to adopt a new institution. Second, a political system succumbs to mimetic pressure when actors that operate inside that system identify institutions, which have been developed elsewhere as desirable. In turn, those actors seek to adopt the ‘foreign’ structures into their own political system. It is important to stress the distinction that while mimetic pressures are exercised by actors which are located inside the political system that accepts an institutional transfer, coercive pressures originate in other systems. Third, normative pressures are exercised by the members of elite groups, such as ‘epistemic communities’ (Haas, Reference Haas1992), the members of which are dispersed in many different political systems. The members of those groups put pressure simultaneously on the different polities in which they operate to adopt similar structures, which conform to the group’s shared background of norms and generally accepted practices.

The preceding general formulation of the theory can be operationalized in more specific terms for the cases of the EO and the ECA. First, ombudsmen and public audit institutions are widely accepted as agents of government accountability (Mulgan, Reference Mulgan2003). They reflect the normative expectation that in a democracy the members of the government apparatus must be held accountable for their actions. At the same time, they constitute agents of a narrow set of practices by which the normative expectation of accountability is realized. More specifically, a public audit body holds government officials accountable by having developed a series of methods and ethical standards, which it applies to control public finances. In the same vein, an ombudsman’s duties represent a different set of practices, which also aim at holding government officials accountable by investigating citizen complaints on their behavior.

Having addressed the normative appeal of these two institutions, how do we know that isomorphic pressures have been applied for their transfer to the EU? For the purpose of this study, evidence of coercive force is observed when national governments exercise pressure for the transfer of a new institution to the EU during treaty negotiations. Although national governments are participants in the EU decision-making, they are not supranational in nature, and therefore I consider them as external to the EU supranational structure. Second, evidence of mimetic pressure can be found in the behavior of supranational actors, and more specifically when these actors push for the import of the institution of the ombudsman or the public audit body into the EU system. The particular supranational actors this article will focus on are the European Commission and the EP. Finally, evidence of normative pressure can be found in the behavior of national ombudsmen and public auditors, which operate in different member states of the EU and are characterized by common professional norms and similar professional practices. When these bodies push for the adoption of the EO and the ECA, respectively, they exercise normative pressures.

The EU literature already includes several studies that utilize isomorphic theory, mostly within the context of ‘Europeanization’, the notion that the EU has a homogenizing effect on member state policies (Featherstone and Radaelli, Reference Featherstone and Radaelli2003) and administrations (Knill, Reference Knill2001). Hence, Radaelli (Reference Radaelli2000) interprets the role of EU institutions in encouraging policy transfers between member states as that of agents of isomorphic pressures, while Pedersen (Reference Pedersen2006) examines how the independent regulatory agency model is spread by isomorphic forces among EU member states as a ‘cognitive script’. In addition, Lodge (Reference Lodge2000) employs isomorphism to explain the ‘Europeanization’ effect on German competition and public procurement policies. What all these studies have in common is their interest in transfers from the EU to its member states. Not only is this article the first to utilize isomorphic theory to understand institutional transfers from member states to the EU but it also articulates isomorphic theory more precisely compared to these earlier studies. Moreover, with its use of theory, the article aims at reaching a more generalizable set of conclusions compared to past studies on similar transfers to the EU (Héritier, Reference Héritier1996; Mamadouh, Reference Mamadouh2002).

In the next two sections, I present the events that led to the establishment of the ECA and the EO, respectively. These sections include the historical circumstances that lead to the establishment of the two institutions. The analysis of the facts is done in a separate section that follows. Patterns that are common in the two cases are analyzed in light of HI and isomorphic theory. The article concludes with the main findings and discusses the more general implications of those findings.

The ECAs: the fight for independent and rigorous audit

The EP made the first call for the creation of the ECA in the run-up to the ‘Merger Treaty’ of 1965, which consolidated the institutions of the three original European Communities. This Treaty created one Commission and one Council of Ministers and formed a single operating budget for the two newly minted institutions. In a resolution, the EP expressed the opinion that the two new powerful players needed to be checked more rigorously than the current system allowed (European Parliament, 1964). This proposal included provisions for a new external audit body, an ‘ECA’. However, the proposal remained merely a declaration and did not lead to any significant moves. The issue kept returning during parliamentary debates from 1967 to 1969 without any practical results (European Parliament, 1973a: 68, 69).

At the time, most of the public audit in the Communities was performed by a body called the Audit Board (Laffan, Reference Laffan1997), which was characterized by the absence of two very important features common among the supreme public audit institutions in the member states: independence from political masters and monitoring of the execution of the budget. First, although the relevant articles described the Board as an independent body, they included no specific provisions on how this independence would be maintained in practice. Hence, the conditions of appointment of its members were not clearly stipulated. As a result, some auditors were allowed to serve on the Board on a part-time basis, a practice that bred potential conflicts of interest. In the late 1960s, for example, some governments started nominating officials from their respective ministries of finance to assume positions as members of the Board, raising the fear that ‘the Communities’ Audit Board will gradually come to resemble the committees of the Council of Ministers’ (European Parliament, 1973a: 19). The concern was that the Board’s audit conclusions would evolve into the political outcome of interstate bargaining, instead of impartial assessments of fiscal performance.

Second, the audit performed by the Board was exclusively retrospective, namely, the Board could audit the budget only after its execution. Specifically, the Treaties gave the Board the mandate to submit an audit report to the Council and the EP, the competent budget authorities, only after the end of the fiscal year. Moreover, the Treaties gave the Commission the sole responsibility of budget implementation, which included monitoring the progress of its execution. In practical terms, this meant that the Council and the Parliament could not request reports and opinions from the Board to assess how the execution of the budget was progressing. The exclusively retrospective nature of the audit performed by the Board did not only mean that the Board itself ‘lacked teeth’, but it also undermined the authority of the Council and the Parliament as budgetary authorities.

The event that triggered the revision of this public audit regime was the signature of the first Treaty Amending Certain Financial Provisions in 1970. The First Budget Treaty, as it is also known, created one unified budget for all three Communities. It also marked a milestone for their financial independence, because it endowed them with their ‘own resources’ for the first time (Laffan, Reference Laffan1997). The Communities acquired the right to collect revenue in the form of agriculture levies, customs duties and value added tax. In terms of institutional dynamics, the Treaty elevated the role of the EP, so that it became equally responsible with the Council of Ministers for the adoption of the unified budget. It also made both institutions responsible for supervising and discharging the Commission for the implementation of the budget. Yet, the First Budget Treaty did not alter the way in which the control itself was conducted: it continued to be entirely retrospective. Hence, the Council of Ministers and the EP could only review the implementation of the budget only after it was completed, while the Commission maintained the privilege of exclusive responsibility for its execution. Moreover, the Treaty failed to address the issue of the Audit Board’s independence. The perceived shortcomings of the First Budget Treaty led to the process of its revision almost immediately after its signature. Additionally, the imminent assumption of full budgetary responsibilities by the Communities created a special sense of urgency for a revision of the EP’s budgetary powers at a time when a broader discussion on the role of the EP in the Communities was taking place. Therefore, the negotiation for a Second Budget Treaty began immediately.

As part of the more general debate on increasing the EP’s authority, the Commission requested the Vedel Report (European Commission, 1972), which contained some well-known and significant innovations for the role of the EP in the Communities, including the co-decision procedure and its direct election. Yet, in regard to budgetary supervision, the report merely re-iterated the status quo created by the Budget Treaty of 1970. The need to replace the Audit Board with a new body was not even mentioned. However, during the same period, the deficiencies in monitoring the budget implementation came up on several occasions in EP debates. These discussions were not related to the process of Treaty revision, but nevertheless reveal the momentum for change in the budget monitoring regime, which was building up in the Parliament. Hence, during annual discussions on the budget, there were sporadic complaints from Members of the EP (MEPs), concerning the absence of continuous communication between the Parliament and the Audit Board as well as the obstacles the Board faced in obtaining the information to which it was entitled by the Treaties (European Parliament, 1971, 1972a, b).

Eventually, the first official draft for the Second Budget Treaty (European Commission, 1973a) was presented by the Commission to the Council of Ministers in June 1973. The main bulk of the draft concerned proposals on increasing the involvement of the EP in deciding the budget. Nevertheless, the issue of budget supervision occupied a prominent place in it as well. In fact, the draft broke new ground by recommending that the task of discharging the Commission for the implementation of the budget should rest with the Parliament alone. A proposal to replace the Audit Board with an ECA was included in the draft, and it was tied to the new budget discharge duties of the assembly. The ECA outlined in the Commission proposal was a major departure from the Audit Board. The independent status of the proposed supervisory body would be subject to strong guarantees. The draft included specific language on the conditions of appointment of the members of the ECA, as well as the procedures for their election and dismissal. In addition, the Commission proposed measures, which guaranteed the body’s self-governance. Thus, the potential for external influence, political or otherwise, and conflicts of interest would be constitutionally eliminated. Details of the EP discussions on the Commission draft (European Commission, 1973c, d) reveal that representatives of all the main political groups and the Commission agreed on the creation of an independent audit institution. In order to justify their preference, several speakers made reference to the reality that legislative oversight of the budget was a regular activity in the member states and appeared to recognize not only its value in general but also the lack of it at the Community level. Yet, the Commission draft failed to address the second major shortfall in the Communities’ external audit regime. Regarding the question of who would supervise of the budget and how, the Commission again reiterated the status quo. The draft assigned to the new auditing body the role of conducting exclusively retrospective control, while preserving for the Commission the sole responsibility of controlling the execution of the budget. Specifically, the ECA would submit a report to the EP only after the completion of the budget. Hence, the Commission’s apparent self-interest prevented it from making bold recommendations for making the discharge itself the outcome of more rigorous external control.

The issue of revising drastically the external audit regime of the Communities was not put decisively on the table until the EP’s Subcommittee on Budget Control and Implementation published an extensive report titled The Case for a European Audit Office (European Parliament, 1973a). This document included a comprehensive analysis of the existing audit regime in the Communities, which included an extensive discussion of the problems posed by the current state of affairs and how these problems could be confronted by drawing on the experiences of the member states. Its main argument was that the tasks of the new auditing body should be extended beyond retrospective control. Just like many national audit bodies, it was argued, the new institutions should have the authority to conduct reports on the progress of the budget at any time, and not only at the request of the budgetary authorities, but also on its own initiative. At the conclusion of the second parliamentary debate on the draft Treaty, in October 1973, the EP issued a Resolution (European Parliament, 1973b), which echoed this demand. In it the EP insisted that ‘the Court must report to Parliament and be ready at all times to assist and advise it in the exercise of its auditing rights’ (emphasis added). Yet, in the amended version of the draft Treaty (European Commission, 1973b), the Commission left the relevant provision unchanged from its earlier draft. It did not heed the EP’s request for constant control during the process of implementation of the budget.

At the next stage, it was the turn of the Council of Ministers, which was in constant communication with the EP, to examine the draft. Between June and October 1974 there was a series of meetings, which included officials from the two bodies, as well as number of written exchanges, the purpose of which was to discuss amendments to the draft. As before, throughout this process of negotiation between the Parliament and the Council of Ministers, the main point of contention was the role of the EP in deciding the budget. However, the problem of budgetary control was raised in relation to the EP’s enhanced supervisory powers. In February 1975, the President of the Council officially presented a new draft to the EP and requested its final opinion. Unlike the earlier text proposed by the Commission, the Council draft not only made the ECA responsible for compiling an annual report at the end of the fiscal year, but in addition it stipulated that ‘the Court of Auditors may also, at any time, submit observations on specific questions and deliver opinions at the request of the one of the institutions of the Community’. In other words, the new draft extended the ECA’s mandate beyond retrospective control to audit of the execution of the budget. In so doing, the new text did not only enhance the ECA’s role but also provided extended powers of supervision over the budget to the EP and the Council of Ministers, which would become the final recipients of the ECA’s reports on the execution of the budget. The final report by the Parliament (European Parliament, 1975a) confirms the EP’s satisfaction with the new draft. On the recommendation of the Committee on Budgets, the plenary adopted the new Treaty text on 11 July 1975 (European Parliament, 1975b). In the end, the second Treaty Amending Certain Financial Provisions was signed in Brussels on July 22, 1975 by representatives of the member states, who gave life to the new institution of the ECA.

What was the role of member state public auditing authorities in this process? The evidence does not allow us to draw firm conclusions about their role. All member states had long established and effective public audit institutions, which reported their views on the establishment of the ECA at the request of the EP (European Parliament, 1973a). In addition, the presidents of national audit authorities were consulted on numerous occasions by the EP before extensive consultation between the EP and the Council began. In both cases, the point of contact in the EP for member state auditors was the Subcommittee on Budget Implementation and Control. These discussions centered on delineating the jurisdictions of the new supranational institution and its national peers, and they were instrumental in determining the future relations among the ECA and member state public auditors. Indeed, the chairman of the subcommittee praised the member state institutions for their contributions in his final report on the ECA (European Parliament, 1975a). Yet, valuable as their contributions might have been, the written evidence suggests the tentative conclusion that national auditors helped determine the shape of the new institution, but they were not the driving forces behind its transfer.

The EO: the fight for the protection of European rights

The original idea for the creation of an ombudsman of the European Communities came in 1974, when the Conservative Group in the EP mentioned the issue briefly in a policy document titled The European Community: Our Common Cause. For several years afterwards, MEPs from political groups of both the Left and the Right, submitted written and oral questions, primarily to the Commission but also to the Council of Ministers (European Parliament, 1974, 1975a, 1977a, b, c) in regard to this issue. The inquiries revealed doubts about the adequacy of mechanisms available to the citizens of the member states to claim rights they were entitled to by Community legislation, when these rights were not sufficiently respected by Community institutions. The increasing role of the Community in decisions affecting the lives of citizens exacerbated those concerns. At the same time, MEPs identified the institution of the ombudsman as a desirable remedy for this problem and recommended the creation of such a position, the occupant of which would be entrusted with the responsibility to address citizen concerns particularly in relation to the actions of the European Commission. In its responses, however, the Commission stated that it considered its internal redress mechanisms and the EP petition procedure as adequate for dealing with citizen complaints, and dismissed the calls for additional measures. The Council of Ministers agreed with the Commission.

By March 1978 enough momentum had been built in the EP in favor of appointing an ombudsman of the European Communities. The Walker–Smith report (European Parliament, 1979a) was commissioned by the EP to examine the need for an ombudsman. The report recommended the appointment of a ‘Parliamentary Commissioner’, who would advise citizens which body, either at the national or the supranational level, would be the most appropriate to address for problems that stem from the application of Community legislation. The report made explicit references to the practices followed by ombudsmen already operating in the member states, and sought to demonstrate the usefulness of such practices for the Communities. Yet, the drafters of the report refrained from recommending that the ‘Commissioner’ be assigned any investigatory power, which typically constitutes the core of an ombudsman’s work, because they felt that such a move would disrupt the balance among Community institutions. Therefore, the drafters indicated, an ombudsman with more extensive powers could only be the outcome of Treaty revision. The proposed ‘Commissioner’ would be an arm of the EP, from which it would derive authority. The EP adopted the report in a Resolution (European Parliament, 1979b). However, at this stage the idea of an ombudsman did not find any supporters outside the walls of the EP.

The second serious attempt for the creation of an ombudsman of the Communities was not made until 1984 at the European Council at Fontainebleau. Recognizing the need to make the European Community more approachable to the citizens of the member states, the Council mandated the creation of the ad hoc committee on ‘A People’s Europe’. This decision echoed concerns already expressed at the Copenhagen and Paris summits in 1973 and 1974, regarding the ‘special rights’ of the citizens of the member states emanating from European Community legislation. The ad hoc committee produced two reports, known as the Adonnino Reports, which were presented to the Council in 1985 (European Commission, 1985).

The second Adonnino Report is of particular significance for the EO. Its drafters recommended that moves be made by the EP to establish an independent ombudsman at the European level, in addition to strengthening the right of petition by the citizens to the EP itself. National representatives saw such a move as improving the legitimacy of the Communities in the eyes of the citizens. Unlike the Walker–Smith report, the second Adonnino Report made more assertive suggestions for the basic functions of the proposed institution. Hence, the ombudsman would not only co-ordinate already existing mechanisms of citizen redress, but it would also investigate complaints regarding the implementation of Community legislation. It was also recommended that the ombudsman submit regular reports to the EP ‘on his investigations, conclusions and recommendations’. Yet, just as the Walker–Smith report, the Adonnino Report recommended that that the ombudsman would be a ‘complement’ to the EP. The rationale behind this suggestion was not to disrupt the existing balance among the major Community institutions by creating a completely new one.

The member state governments approved the reports and urged the Commission, the EP, as well as national authorities, to proceed with the implementation of their recommendations as soon as possible. Hence, the Adonnino Report marks the first time that the issue of the creation of an ‘EO’ gains the backing by an intergovernmental forum. However, just like in the case of the Walker–Smith report, the Adonnino Committee recommendations for the ombudsman never came to fruition. As an EP resolution reveals (European Parliament, 1985), the implementation of the specific provisions regarding the ombudsman was abandoned, because it faced insurmountable legal problems. As a more feasible alternative, the EP committed to strengthening the citizens’ right of petition to the EP.

The third and successful attempt for establishing an ombudsman of the European Communities took place in 1991 and 1992, during the intergovernmental deliberations for the Treaty of Maastricht. Unlike the first two, the third attempt has already been documented in the literature (Marias, Reference Marias1994; Magnette, Reference Magnette2003), perhaps because it was the one which was successful. The current study, however, puts the events of this attempt in light of earlier developments and thus in a broader perspective. The idea of an ombudsman was introduced to the intergovernmental conference (IGC) with a letter by Felipe González, the Spanish Prime Minister. A Spanish ombudsman had been established in 1981, and it had proven to be quite an effective and popular institution (Retuerto Buades, Reference Retuerto Buades1994). Therefore, the idea of creating an ombudsman at the Community level appeared particularly appealing to the Spanish government. In September 1990, four months after the original letter, the Spanish delegation submitted a formal proposal to the IGC, in which it was suggested that the ‘special rights’, which were afforded to the citizens of the member states by the Treaties and revolved primarily around their participation in the Single Market, should be consolidated into a ‘European citizenship’. The Spanish proposal recommended furthermore that a European citizenship should be accompanied by extra guarantees for the special rights of the European citizen, in addition to already existing mechanisms, namely recourse to the European Court of Justice and the EP Petitions Committee. An ombudsman who would accept complaints from citizens and defend their rights before EU institutions, it was argued, could function as this additional guarantee.

The Commission gave its endorsement to the Spanish proposal for a European citizenship in a paper published in October 1990 (European Commission, 1990). However, it did not make any reference to the need for an ombudsman at the Community level. This is not surprising in light of the Commission’s reluctance to encourage the creation of the institution in the past. However, during the same month, a Danish memorandum was submitted to the IGC, which endorsed the Spanish proposal and asked for an ombudsman to be created under the auspices of the EP. Denmark already had a national ombudsman since 1952, which had been enjoying great success.

In February and March 1991, Spain and Denmark submitted more detailed proposals to the IGC. These two proposals adopted quite different views on the position and the role of an EO. The Spanish proposal was quite ambitious. Not only did it contain a recommendation for an EO, but the proposed institution would investigate complaints concerning the application of EC rules by both EU and national authorities. In addition, it recommended that all member states appoint an ombudsman or a body with equivalent functions. The EO would cooperate on its investigations with national counterparts and coordinate their actions. As a result, the new institution would be in charge of a network of national authorities. Although it was not stated explicitly, a hierarchy was implied with the EO at the apex of the pyramid. In comparison, the Danish proposal was much more attuned to the realities of the cohabitation of national and supranational institutions. It called for the creation of an EO as well, but restricted its investigative powers by limiting his jurisdiction only to the institutions of the EC. The proposal outlined the terms of the relationship with the EP, as well. The Danes, in line with their national experience, recommended an independent institution, which however, would work closely with the EP.

A third proposal was submitted by the Commission in May 1991 (European Commission, 1991). The proposal made no reference to an EO. In contrast, the Commission, apparently still doubtful, contained a recommendation for the creation of national ombudsmen, who would assist citizens defend the rights that derived from the Treaties. Eventually, the Luxembourg presidency was able to find a compromise among the different views (Marias, Reference Marias1994) and the Danish proposal prevailed. Indeed, the text that was finally adopted by the IGC and made it into the final Treaty resembles the text in the Danish proposal very closely. The new institution would be an independent parliamentary ombudsman with authority to investigate complaints of maladministration against Community institutions, excluding the courts, a main feature of the Danish ombudsman model.

Regarding the role of professional groups in the events that led to the transfer of the ombudsman institution to the EU, just like in the case of the ECA, the evidence does not allow us to draw safe conclusions. The views of the national ombudsmen of Denmark, France and the UK, the only ombudsmen that existed in member states at the time, were solicited by the EP and they were included in the Walker–Smith report. In that instance, member state ombudsmen were asked to express their views on the future relations with the new supranational institution. However, there is no evidence that they were involved in the negotiations that followed. Based on this observation, we may conclude tentatively that their role on determining the final shape of the EO was even more limited than in the case of the ECA.

Analyzing the events

How can HI and isomorphic theory help us make sense of the preceding events? Adhering to Process Tracing principles, here observations are organized into concrete causal steps that link cause (isomorphic pressures) to effect (institutional transfer). Hence, the events leading to the establishment of the EO and the ECA followed a similar pattern. First, the EP identified an organizational model as desirable for transfer. Both in the case of the EO and the ECA, MEPs presented those two institutions as solutions to accountability problems faced by the Communities. In the case of the EO, the justification for transfer was the protection of citizen rights. The new institution would hold other Community institutions accountable by increasing their transparency and by requiring that they address directly any complaints by citizens of the member states. In the case of the ECA, the need for more accountability of those who managed Community funds was tied to the newly acquired ‘own resources’ of the Communities. Because Community institutions would begin handling funds that were previously managed by national governments, so the argument went, the deficient external audit regime was no longer acceptable. Instead, it should conform to the standards of public audit in the member states. Hence, both EP initiatives had normative justifications.

Second, the Parliament began the process of transfer by applying pressures for the adoption of the new institution. We can label pressures as mimetic, because they originate in the EP, an actor who is part of the supranational structure of the EU. Moreover, the intention of that actor was to import that institution to the EU system of which the EP is part. In the beginning, the pressures were weak. The first indication of a move for the adoption of a new organization was usually an isolated declaration. However, in time, those pressures increased as the issue gained momentum. Hence, for example, in the case of the ombudsman, the EP actively solicited the Walker–Smith report. Despite the Treaty-based principle of institutional balance, which does not permit the increase of the power of a European institution without a Treaty revision, the EP went so far as to propose the creation of the EO even outside of the framework of this process. Similarly, in the case of the ECA, during the process of revision of the Budget Treaty, the EP exerted continuous pressure on the Commission and the Council to alter their Treaty drafts.

Despite the professed desirability of the two organizations, the EP also had an ulterior motive: the increase of its own power vis-à-vis other supranational institutions, particularly the Commission. Thus, the addition of the new bodies would enhance the position of the EP in the supranational order of the Communities. Indeed, the choice of the two bodies does not appear to be a coincidence. At first glance, those two organizations may appear to be dissimilar. Each one specializes in a very different area: financial control and citizen rights. Furthermore, while the ECA has a very hierarchical structure and is characterized by strong reflexes for secrecy, the EO is almost its polar opposite, being a flexible and unusually transparent body. Yet, both organizations share a very important feature: in many countries their functions are traditionally associated with supporting the actions of the legislature. In fact, the common function of those two otherwise different types of organization is that they furnish information to the legislature on the behavior of the executive branch. By doing so, they contribute to increasing the executive’s accountability to the legislature.

Third, inevitably there was backlash from the executive, which counterweighed the mimetic pressures exerted by the EP. The Commission sought to avoid the inclusion of provisions in the drafts it produced which would allow the establishment of the new bodies. Although the European Commission is not a conventional executive, it nevertheless performs a function, which is traditionally associated with the executive branch of government: it is responsible for the implementation of EU policies and programs (Peterson, Reference Peterson2006). The two organizations under examination in this study were expected to control specific aspects of this function. In the case of the EO, the Commission, having the main responsibility for the implementation of Community policies, would be one of the primary subjects of the EO investigations of citizens’ rights violations. Moreover, the ECA would not only share its supervisory power over the execution of the budget, but it would also make the Commission the main subject of its audit. It is the Commission, which has the ultimate legal responsibility for budget implementation.

The prospect of establishing the ECA and the EO was perceived as a threat by Commission officials, because it bred the potential of deteriorating the position of the Commission in relation to the EP. Indeed, although both of the new institutions would be completely independent (and indeed they are), they would nevertheless be close allies of the EP. Hence, it is not surprising that the Commission reacted negatively to the pressure of the Parliament to transfer the EO and the ECA to the Communities. More importantly, the Treaty negotiation process provided the Commission with the opportunity to exert influence, by giving it the mandate to produce Treaty drafts that reflected its preferences. In the case of the EO, Commission officials constantly rejected parliamentary requests for an initiative that would explore the possibility of a parliamentary ombudsman, within or without the Treaty revision process. Similarly, in the case of the ECA the Commission only partially heeded the EP’s request for a public auditing body, fully equivalent to those in its member states. In each proposal, the Commission recommended that the ECA have strong guarantees for independence, but it did not provide the ECA with the role of monitoring the execution of the budget.

The fourth and last piece of the puzzle in each case was the role of member state governments. As entities external to the supranational infrastructure of the EU, member state national governments exerted coercive pressure on the EU. They used their privileged position as the official signatories to any Treaty that outlines the main rules and structures of the EU. Yet, a necessary condition for this occurrence was that they also shared the EP’s view on the desirability of the new institutions. As the competences of the Communities expanded, member state governments had an interest in enhancing the accountability mechanisms that checked the behavior of supranational institutions. This convergence of preferences between the Parliament and member states was quite evident in the case of the EO, as it was member state officials who drafted the Adonnino Reports, which affirmed the need for an ombudsman at the intergovernmental level. Furthermore two member states, Spain and Denmark, influenced developments directly. During the extensive negotiations for the Treaty of Maastricht, the Spanish and Danish delegations took the lead in making the EO a Community institution. Given the strong resistance of the Commission to include any reference to the ombudsman in its drafts, it was only member states that could introduce the institution in their own drafts. The evidence is less clear in the case of the ECA as member state willingness to heed the Parliament’s request for a rigorous audit regime can only be inferred, mainly by the fact that member state officials included the clause that allowed the ECA to perform audits throughout the process of the execution of the budget for the first time.

The pressures exercised by member state professional groups, which were labeled as normative pressures in the beginning of the article, were observed but their role is difficult to assess on the basis of the evidence. The evidence suggests that the EP approached both public auditing authorities and national ombudsmen, and that it played the role of conduit for their views. In both cases professional groups appeared to respond to the leadership exercised by the EP, rather than driving the events by themselves. However, the evidence is rather limited. One potential explanation for this scarcity of evidence could be that professional groups were not part of the EU order. As a result they did not have an interest to advance their position in that order. In other words, they did not have an interest in the negotiations. An alternative explanation could be this study’s focus on the process of Treaty revision, which, by its nature, excludes the direct participation of professional groups. Therefore, it is possible that professional groups played a more prominent role behind the scenes and this was not reflected in the documents relating to the treaty revision.

Conclusions and implications

From the preceding analysis, we can deduce that significant mimetic and coercive pressures explain the transfer of the EO and the ECA to the EU. It is however difficult to reach a firm conclusion on the importance normative pressures. Based on the historical evidence, we may conclude that the collusion of mimetic and coercive isomorphic pressures was necessary but not sufficient conditions for this transfer. In other words, the two institutions would not have been created, had it not been for the initiative taken by the EP and the assent of the member states. Yet, at the same time, the exercise of isomorphic pressures alone would not have produced the observed outcome of organizational transfer. The reason is that the actors who exercised those pressures operated in an institutional environment, which determined their options for action. Specifically, in the two cases presented in this study, the treaty negotiation process affected the final outcome by assigning specific roles to the each set of actors: it assigned a consultative role to the Parliament, the role of drafting the original proposals for reform to the Commission and the role of drafting reports and making the final decisions to the member states. Each one of these actors used its institutional position to encourage or discourage organizational transfer.

Moreover, the preferences of those actors were endogenous. On the one hand, whether supranational institutions were against or proreform depended on considerations about whether the reform would ameliorate or deteriorate their position in their institutional environment. On the other hand, the preferences of national governments were formed in response to an issue that was originally raised by supranational institutions. Thus, a historical institutionalist interpretation of endogenous actor preferences and path dependency (Hall and Taylor, Reference Hall and Taylor1996; Pierson, Reference Pierson2000) is supported by the evidence. At the same time, preferences were also culturally bounded. The ombudsman and the public audit institution were selected by the EP because of their desirability as accountability mechanisms, and their inclusion in a new treaty became possible only when the member states accountability concerns converged with those of the EP. Hence, the timing of the transfer was determined by the timing of preference convergence.

What are the implications of the preceding conclusions? First, the results of this study suggest a pattern of evolution for the EU, which is not fundamentally different from that of other institutional systems. To be sure, as a set of institutional structures, the EU is quite unusual. It has been characterized as a ‘hybrid’ (Rosamond, Reference Rosamond2000: 110) because it does not fit neatly into any pre-existing institutional classification. And yet, the EU develops institutionally by ‘borrowing’ structures from other systems in its immediate environment. In other words, the EU may appear to be unique if we take a static ‘snapshot’ of its structural configuration, but this uniqueness is less obvious when we think about the EU in terms of its evolutionary dynamic.

Second, although the findings do not allow us to make precise predictions about the future of the EU, they nevertheless allow us to make inferences about its future institutional trajectory. They suggest that future transfers to the EU will involve structures from the member states, which are perceived as successful and normatively appropriate for governance in modern democracies. The democratic legitimacy problems the EU has suffered in the eyes of the peoples of the EU are well documented (Beetham, Reference Beetham1998; Moravcsik, Reference Moravcsik2002; Lord and Magnette, Reference Lord and Magnette2004). By borrowing with which its citizens are already familiar with in the member states, the EU can increase its legitimacy. Hence, in its quest for legitimacy, increasingly the EU will model itself after its member states. Yet, at the same time, it is equally important that the institution garners enough support among supranational actors and national governments, in order for its transfer to be initiated and competed successfully.

Of course, even the casual observer would recognize that the EU today is very different from the European Communities of the 1970s and 80s when the events presented in this study transpired. Most notably the power relations among institutions have been shifting and numerous new member states have joined. How will this ever-changing reality affect the exercise of isomorphic pressures in the future? Although the number and roles of individual actors may change over time, the basic process of institutional transfer will remain the same so long as the treaty revision method remains the same. This method is inherently conservative and therefore contributes to institutional path-dependency in the EU. An institutional transfer through treaty revision requires the confluence of multiple interests, which is generally difficult to accomplish. The strengthening of the EP and the most recent addition of twelve new member states mean that now the interests of even more diverse and powerful actors must be in agreement. It is therefore reasonable to predict that any such transfers will become even more difficult in the future. The Simplified Revision Procedures of Article 48 in the Treaty of Lisbon aim at making the process of treaty revision more efficient. It remains to be seen if and how often the procedures will be used in the future, and how they may affect the creation of new institutions.

Finally, thinking beyond the EU, this study indicates that certain environments are particularly conducive to institutional change through transfer. These environments are characterized by a significant number of regular interactions among political systems at the subnational, national or supranational levels. Under these circumstances, an abundance of opportunities is available to actors who are active in those systems to exercise isomorphic pressures, thus increasing the frequency of institutional transfers. This explanation does not dismiss the possibility that a transfer may occur between systems that do not have long-term and regular interactions. Furthermore, the explanation advanced here is not meant to imply that a political system evolves only by accepting transfers from its peers. Indeed, even in an environment rich in regular interactions, such as that of the EU, political systems may evolve on the basis of their own internal dynamic. What this article argues for, however, is that the existence of regular interactions increases exponentially the likelihood that the systems involved in those interactions will evolve by borrowing institutions from each other.