University of Bradford (UK) 

Policy Transfer in The European Union: Institutional Isomorphism as a Source of Legitimacy 

Abstract: The paper examines public policy in the European Union (EU) by drawing upon the conceptual framework of policy transfer, which has been recently refined by comparativists, and the concept of isomorphism developed within organizational theory. Three case studies of EU policy transfer - namely monetary policy, tax policy, and media ownership policy - are discussed and compared for assessing the potential of isomorphism for the analysis of policy diffusion. The author argues that European institutions, which have a serious political limitation in terms of legitimacy, stimulate policy transfer by catalysing isomorphic processes which diffuse throughout the EU national policy solutions to collective problems. By contrast, policy transfer is severely constrained when there are no national cases to be imitated. In this circumstance, however, European institutions, most notably the European Commission, can overcome the problem by ‘inseminating’ solutions into national political systems.

The point of departure of this paper is the search for new conceptual lenses in the study of European Union (EU) public policy. As observed by Caporaso and Keeler (1995) in their systematic review of EU integration theories, ‘grand’ theories (such as, for example, neo-functionalism, neo-realism and liberal inter-governmentalism) contain elements of rigidity that obstacle the analysis of concrete policy dynamics in the EU. In addition, Caporaso has argued in a recent article that:

The study of European integration is moving into a post-ontological state; scholars are less concerned with how to categorise than how to explain process and outcome, paying less attention to the ‘nature of the beast’ (Caporaso 1996:30).

This paper proceeds from the acknowledgement of a post-ontological state in EU integration theory and accordingly turns to models and conceptual frameworks developed within the disciplines of comparative politics (1) and organizational analysis for inspiration. More precisely, it employs the framework of policy transfer suggested by scholars of comparative politics and the concept of institutional isomorphism which has been fleshed out in studies of organizational behaviour.

Let us begin with policy transfer (2). There is no need to illustrate the details of policy transfer (see Dolowitz and Marsh in this issue), but it is important to observe that it is a conceptual framework (rather than an explanatory model, see below) broad enough as to account for voluntary and coercive transfer, positive and negative lessons (i.e. lessons about how not to proceed), transfer of wholesale policies and more limited transfer of instruments. For our purposes, the essential question, however, is how policy transfer occurs in the EU. Although mentioned by authors such as Rose (1993:105) and Dolowitz and Marsh (1996:346), policy transfer in the EU has received scant attention within this body of literature.

Arguably, the single most striking element of differentiation between policy transfer at the level of democratic states (i.e. poliarchies) and supra-national policy transfer in the EU concerns legitimacy. When a poliarchy decides to ‘copy’ one policy programme from another political system, it is legitimised in doing so by democratic rules and procedures, most notably elections. By contrast, EU institutions do not possess a similar degree of legitimacy (3) . Unfortunately, the debate on European integration - as already mentioned - has indeed focused upon ontology (Caporaso 1996), thus - as Bufacchi observes - ‘chasing the shadow of the real issue: political legitimacy (...) Is the power of the European Community legitimate? and if so, how’ (Bufacchi 1994:229).

The study of legitimacy is stimulating as scholars engaged in the explanation of regulatory regimes have shown that political legitimacy can be secured independently from governmental action and the ensuing legitimation of power. An example is provided by Underhill (1995), who has illustrated how the regulation of the transnational securities market proceeded by keeping governments out of politics. In the context of transnational securities - Underhill notes - governments were kept ‘at bay’, but legitimacy was nevertheless achieved.

Underhill’s paper, although not concerned with policy transfer, is useful as sheds light on the construction of political legitimacy in political domains lacking the traditional ‘device’ for producing legitimacy, i.e. a democratic government. With the EU in mind, the following question arises: how does the EU - most notably, the European Commission, which is the main policy initiator in the EU - ‘produce’ legitimation of supra-national policy transfer? This paper suggests that the search for the appropriate answer should begin by an exploration of the literature on new institutionalism in organizational analysis ( Powell and DiMaggio 1991; Scott 1995). Organizational theorists have in fact explained a phenomenon which is inherently similar to policy transfer, i.e. the homogenisation of organizational fields (4).

There is a strong rationale for combining policy transfer and institutional analysis. Indeed, as mentioned above policy transfer is a conceptual framework, not a model. In order to provide an explanation of specific cases of policy transfer, a conceptual framework must be supplemented by models which shed light on policy processes. By supplying explanations of how actors legitimise public policy and institutions emerge, new institutionalism provides models which add flesh to the skeleton of the ‘policy transfer’ conceptual framework.

To recapitulate, this paper is an exploratory exercise in the analysis of supra-national policy transfer in the EU. The aim is to examine three EU policy areas in order to shed light on the relationship between isomorphism and legitimacy. Section 2 introduces the concept of isomorphism and discusses its ‘transferability’ to the EU. The succeeding sections are dedicated to the examination of EU cases of policy transfer. Section 3 analyses the case of EU monetary policy, section 4 deals with tax policy, and section 5 investigates media ownership policy. Finally, section 6 presents some concluding remarks.

Organizational theorists dealing with isomorphism are more interested in explaining homogenisation than variation. Similarity rather than differentiation is the key area of investigation in this area of research (5) The emergence and structuration of an organizational field - DiMaggio and Powell (1991) note - result in the homogenisation of organizations included in the field as well as of new entrants. Two processes lead to greater homogeneity, namely institutional and competitive isomorphism. The latter can be hypothesised only if the assumption of competitive markets is tenable. Given certain environmental conditions that a population of organizations needs to face, open and competitive markets will favour the emergence and diffusion of the most rational and efficient solutions to environmental challenges. For political scientists, however, it is more realistic to relax the assumption of open and competitive markets; for example, in the EU there are complex multi-layered political processes and formidable barriers to the adoption of efficient solutions (Scharpf 1988). Turning to institutional isomorphism, this concept, instead of assuming free and open competition, posits that ‘organizations compete not just for resources and customers, but for political power and institutional legitimacy’ (Aldrich 1979:265 quoted in DiMaggio and Powell 1991:66). Copying organizational structures is not a process driven by efficiency considerations, but a way of securing legitimacy in institutional life. The adopted models may not enhance efficiency, but secure legitimacy. In the words of DiMaggio and Powell (1991:73):

Institutional isomorphic processes can be expected to proceed in the absence of evidence that it increases internal organizational efficiency (...) Similarity can make it easier for organizations to transact with other organizations, to attract career-minded staff, to be acknowledged as legitimate and reputable, and to fit into administrative categories that define eligibility for public and private grants and contracts. None of this, however, ensures that conformist organizations do what they do more efficiently than do their more deviant peers (DiMaggio and Powell 1991:73).

If it is not efficiency the main rationale of institutional isomorphism, the diffusion of organizational structures and policy programmes must proceed through other mechanisms. DiMaggio and Powell present three sources of institutional isomorphic change: coercive, mimetic, and normative. Coercive isomorphism is the response to such pressures as a government mandate or dependence on key organizations. For example, an organization will tend to become similar to those organizations on which is dependent.

However, imitation can be encouraged by forces other than power and dependence. This is the case of mimetic isomorphism, which is triggered by uncertainty: ‘when goals are ambiguous, or when the environment creates symbolic uncertainty, organizations may model themselves on other organizations’ (DiMaggio and Powell 1991:69) (6). Mimetism stems from the need to cope with uncertainty by imitating organizations which are perceived to be more legitimate or more successful. In this case again, imitation of models may not assure efficiency, but is nonetheless extremely effective is generating legitimacy.

In a paper on institutional design and democratic transitions, Offe has extended the idea of mimetism from organisations to institutions. Perceptively, he observes that ‘institutional designs are typically copies, and are frequently advocated as such’ (Offe 1992:21). There is in fact a political advantage in claiming that one is copying rather than creating:

The designer, if seen as such, will unavoidably come under the suspicion of trying to impose his particular interest or normative point of view upon the broader community, and that suspicion alone, unjustified though it may be in some cases, may invalidate the recognition and respect of the new institution’ (Offe 1992:24).

Thus we should expect mimetism both under conditions of genuine uncertainty and as a political strategy. In both cases, the goal is to achieve legitimacy. Finally, the third ‘ideal-type’ of isomorphic processes is represented by normative pressures induced by professionalisation. Professionals, their associations and the mechanisms of formal education, socialisation and recruitment produce a common cognitive base and a shared legitimation of occupational autonomy which make organizational structures similar one to another.

Isomorphic processes have been empirically analysed in a number of areas, from college textbooks to administrative reforms (see the literature reviewed by DiMaggio and Powell 1991). In addition, the original focus on organizations has been widened to include the imitation of policy programs and institutional design (Offe 1992). Briefly, this method of looking at structured organizational fields with poor competitive components has potential also for structured political interactions: the EU political system appears to be a good example, at least prima facie.

Turning to the EU, it can be observed that the first type of isomorphism, i.e. coercion, is not likely to dominate EU policy transfer as there is no ‘centralisation’ of the EU policy process around a key organization. Policy-making in the EU is non-hierarchical (Richardson 1996) and the pivotal role of the Commission is more a consequence of its role as an active bourse of interests and ideas than the result of member states’ dependence on resources controlled by Brussels (Mazey and Richardson 1995). However, it should be noted that each member state is dependent on key resources produced by other member states: for example, within the EU, economic growth is dependent on the level of demand in key countries such as Germany, and interest rates in Paris or Rome cannot be set without taking Bundesbank’s policy decisions into serious consideration. Given the level of uncertainty that pervades EU policy-making (Peters 1994; Richardson 1996), it can be argued that policy transfer in the EU should follow the path of mimetic isomorphism in many circumstances. Finally, a variant of normative isomorphism can occur in those EU policy-making processes dominated by cognitive resources, expertise and even technocratic aspects of political life. An example is provided by the Delors Committee (Verdun 1996) which was instrumental in transmitting the beliefs of central bankers into the EU project for a single currency. Arguably, the political action of epistemic communities in the EU, recently drawn attention to (Radaelli 1997; Richardson 1996; Verdun 1996), albeit not equivalent to it, chimes with the consequences of normative isomorphism described by organizational theorists.

Following Dolowitz and Marsh (1996), the analysis of policy transfer can be organised around a series of questions which are reproduced in table 1: who transfers policy? What is transferred? Why and how? Table 1 covers the three policy areas under investigation here (i.e. monetary policy, corporate tax policy, and media ownership policy). These three policy areas are somewhat connected to the first pillar of the EU, i.e. the ‘community’ or the single market. Yet the degree of policy transfer achieved within the three areas varies to a great degree, with a maximum in monetary policy to a limited policy transfer in corporate tax policy. Media ownership policy is an area in which the EU policy stage is in its early stage: at the moment of writing this paper, a formal proposal of the Commission (i.e. a proposed directive) has been announced as imminent but has not as yet materialised.

Table 1 - Supra-national policy transfer in the EU: monetary policy, corporate tax policy and media ownership policy

Let us proceed with order and begin with monetary policy, which is the subject of this section. The Economic and Monetary Union (EMU) provisions enshrined in the Treaty of Maastricht represent a striking case of successful supra-national policy transfer (7). Member states decided to surrender sovereignty in monetary policy, gave constitutional status to the economic paradigm of policy credibility, and engaged in an institution-building exercise by devising the future functions to be performed by the European Central Bank and the Council. Why did they do it? Uni-lateral explanations are probably inadequate, given the complexity and the political importance of this episode. Of course, the lessons of learning both from past failures (i.e. the Weimar crisis or the more recent demise of the Bretton Woods regime) and successes (i.e. the performance of the European Monetary System in the 1980s) appeared to show, at the end of the 1908s, that the long-standing quest for a monetary union in Europe (dating back since the Werner report, 1970) was both economically desirable and politically achievable.

Interestingly, the single currency project was compatible with more than one policy choice. In other terms, once secured the political acceptability of EMU (‘should we do it?’), the problem of technical feasibility (‘how will we do it?’) had to be solved (8). A monetary union can be implemented either by, say, building up a system of fiscal federalism or by a monetary regime based on ‘policy credibility’ and devoid of fiscal instruments. Both positions had their supporters in the European policy process (on fiscal federalism see the Mac Dougall report, Commission 1977) and reflected different monetary orthodoxies, but the ‘policy credibility’ approach prevailed. Presumably, this choice was yet another example of learning. The inflationary crisis of the 1970s had illustrated that governments could not easily resist the demand of inflation. Consequently, the supply of inflation had to be curbed by removing monetary policy discretion from the political system. Tying governments’ hands on monetary policy and empowering non-majoritarian institutions with the mission to fight inflation at any cost - i.e. the paradigm of policy credibility - was the option eventually selected at Maastricht.

It is difficult to locate a single actor steering the EMU policy process. The Treaty provisions on EMU were the product of at least three categories of actors. First, supra-national institutions, with the Commission as pivot. Secondly, national policy-makers, with the Franco-German axis in a prominent position. Thirdly, monetary policy experts, with the Delors Committee playing the role of a sui-generi epistemic community (Verdun 1996). And a fourth powerful source, i.e. the ‘structural power of financial markets’ moulded the decisions taken at Maastricht (Dyson 1994). Therefore an analysis in terms of power of the EMU policy transfer should include the dynamics of the ‘two-level’ game (Putnam 1988) played by national policy makers at home and in the European Community, the power of economic ideas, and the structural power of financial markets (Dyson 1994).

Dyson (1994) adds another important element to the analysis of power relations within EMU by shedding light on the structural power of the ‘anchor’ currency, i.e. the Deutschmark. This is an important point of departure for understanding why the EMU policy transfer gained legitimacy in the eyes of national policy makers. Policy transfer was possible because the EMU policy process had a clear centre of gravity, represented by the Deutschmark. Not only was the Deutschmark a centripetal force in the EMU policy process, but the whole institutional set-up of German monetary policy appeared a good example to be imitated and transferred into the Treaty of Maastricht. From the mist of the economic chaos of the 1970s, the glaring example of the German model of monetary policy singled out. It is true that history does not provide lessons, but a stock of ambiguous evidence in search of interpretation (Furner and Supple 1990). But the shared interpretation throughout Europe was that the performance of the German model had been superior. Thus the EMU policy transfer contained an important element of isomorphism which enhanced EMU legitimacy. For certain important aspects the Treaty on the European Union is the German model writ large (Dyson 1994).

In conclusion, the legitimacy of the EMU transfer was achieved thanks to several elements: history and learning, bargaining, the anchor power of the Deutschmark, and the consensus on the paradigm of policy credibility. In terms of isomorphic processes, a certain degree of coercive isomorphism can be detected (i.e. the anchor power of the Deutschmark), but presumably the pervasive presence of the policy credibility paradigm in the EMU policy process is the result of mimetic isomorphism: at Maastricht, uncertainty found a response in the imitation of the (perceived) most successful national model of monetary policy. By the same token, mimetic isomorphism secured legitimacy during this ‘technocratic’ stage of the EMU policy process. Perhaps a case can be made for normative isomorphism too, when considering to what extent the Delors committee was instrumental in shaping the EMU policy process in accordance with the shared beliefs and values of the central bankers profession (Verdun 1996).

The fact that legitimacy was the crucial element in this episode of policy transfer was demonstrated by the fragility of the EMU success. Supra-national policy transfer in the EU is inherently vulnerable in terms of democratic legitimacy, given the existence of a serious problem of democratic deficit. In the case of EMU, the eminently technocratic policy process had completely overlooked the cultural dimension of money (Dyson 1994). As soon as 1992, the road to monetary union became rough and steep (Sandholtz 1996).

In addition, the effort to respect the convergence criteria of Maastricht and qualify for the single currency has now produced in countries such a France, Germany, Italy and Spain a coercion to restructure public finance and reform the welfare state. ‘When the Maastricht Treaty was drafted’ - Bovenberg and de Jong note - ‘the criteria did not seem too stringent. In fact, at that time eight or nine countries satisfied each of the criteria’ (Bovenberg and de Jong 1997:94). However, it has now become more difficult to be in line with the convergence criteria and for the future the stability pact will make the path of economic policy even narrower. During the years, the coercive aspects of EMU have gained importance, so much so that they are now prevailing over mimetism.

A final remark on legitimacy. The EMU policy transfer is at risk precisely because its democratic legitimacy is in doubt. Isomorphism, learning and economic policy paradigms secured technocratic legitimacy at the Maastricht table. But the other side of legitimacy, i.e. wide-spread democratic consensus over policy transfer, represents the major challenge for EMU: it is indeed the very Damocles’ sword hanging over the completion of the single currency project. European institutions, most notably the Commission, are currently at pains to show EU citizens that the single currency is part and parcel of European citizenship.

The colossal scope for monetary policy transfer in the EU dwarfs the very limited results achieved in EU direct tax policy harmonisation. Until now only two directives aimed at mitigating double taxation in the single market and one convention (on arbitration in international disputes concerning transfer pricing) were agreed upon by the Council in 1990 (9). The 1990 directives do not contain elements of convergence that could be compared to the ones achieved in monetary policy under the Treaty on the European Union. For example, no compulsory range of tax rates (or any other element that could be barely suggest the drive towards s single European tax system) appears in the corporate tax directives, although the Commission had recommended it in the 1970s and the 1980s. Rather, the 1990 tax directives dispose of cross-border withholding taxes that distort business in the single market.

In terms of ‘what has been transferred’, therefore, it can be argued that the paradigm of tax neutrality represents the rationale around which EU tax policy was built. This paradigm states that in a single market taxes should not affect decisions in terms of trade, investment, and localisation of firms. The difference with monetary policy, in which both a paradigm and institutions were transferred, is striking. In addition, only a limited degree of tax neutrality has been achieved. As observed by the committee of independent experts set up by the Commission for analysing the European tax system (Commission 1992), the single market still contains a number of tax distortions which have not been eliminated so far.

Given this frustrating state of affairs, the Commission has recently sought to promote a more political debate on tax coordination. The argument is that not only does the lack of tax policy co-ordination threat the single market, but it also triggers damaging tax competition. State compete in a race-to-bottom for attracting capital and in doing so they place an increasing share of the tax burden on labor. In turn, the heavy tax burden on labor is a cause of mass unemployment (Commission 1996). This new political discourse on tax coordination has been accepted by the Council, lead by the concern of states such as Germany and Belgium that tax competition could result in a loss of welfare for everybody. In conclusion, policy-makers in Europe have now accepted two important ideas: tax neutrality and the need to harness tax competition. Yet the picture has not improved much in terms of concrete action: after the 1990 tax measures, the only innovation is a code of conduct on tax competition drafted in 1997. The code is eminently an instrument of moral suasion: after all, it is almost impossible to stop a state which wishes to change its own domestic tax system in order to attract companies and savings.

Why has policy transfer been so limited in this area? The Commission has pushed for tax co-ordination in the Community since the 1960s (Easson 1992), but direct tax policy transfer materialised only in 1990. In that event, three elements facilitated policy transfer: the implications of the Single European Act for tax harmonisation, the shift of the Commission toward a pragmatic approach, and the intellectual activity of a group of experts sharing the pragmatic aim of the Commission (Radaelli 1997).

Member states have always been reluctant in accepting corporate tax policy transfer. Is this reluctance originated by political and cultural concerns, or, alternatively, by the lack of strong economic rationales for tax policy transfer in the EU? The cultural and political dimensions of direct taxation (suffice it to mention the link between taxation and representation, or the historical role of taxation in the formation of nation-states) have been powerful brakes on policy transfer. However, a similar constraint exists for monetary policy in that national currencies are both bastions of national sovereignty and elements of identification with a community of citizens (although the Germans are certainly more ‘identified’ with the Deutschmark than the Italians with their Lira!). Briefly, culture and symbols of statehood are as important in tax policy as they are in monetary policy. Political considerations should favour rather than hamper tax policy coordination. In fact unbridled tax competition can destroy the welfare state (Sinn 1990): in an extreme scenario, capital is not taxed, highly skilled labor emigrates to favourable tax regimes, and revenue has to be extracted from the least well-off. This system would be unfair and would not provide enough income to support the welfare state.

As far as economic rationales are considered, if the case for monetary policy transfer in the EU draws upon the anti-inflationary paradigm of policy credibility, there is also a strong economic rationale - based upon the concepts of capital export neutrality and capital import neutrality - for eliminating tax distortions in the single market (Devereux and Pearson 1989). Additionally, there is an economic rationale for arguing that tax co-ordination becomes indispensable if policy transfer is accepted in monetary policy. Indeed, the theory of second-best states that removing monetary barriers in a market in which other barriers are still present (cross-border withholding taxes are a case in point) can decrease the degree of economic welfare (Frenkel, Razin and Sadka 1991). Therefore monetary union should be implemented only in conjunction with the removal of tax distortions; otherwise the efficiency of the single market would be put in jeopardy.

To sum up, the legitimacy of EU policy transfer is poor in direct taxation, but not (or, more accurately, not only) because political, cultural and economic arguments against tax co-ordination in the EU are overwhelming. Accordingly, there must be ‘something else’ that explains the fundamental difference between monetary and tax policy in the EU. Isomorphism illuminates this difference. It has already been shown that the Treaty of the European Union has transferred elements of the German model to the whole of the EU. The EMU policy process has not created solutions for policy transfer from scratch, but has found in the German model of monetary policy and in the structural power of the Deutschmark a crucial ‘anchor’.

Nothing similar has happened in the direct tax policy process. There is no ‘anchor’ as far as domestic tax institutions are concerned. No one has as yet argued that the ‘policy credibility’ of, say, the German tax model is greater than that of the UK institutions for tax policy. The very idea of a national tax system that could be considered optimal (or most desirable) for tax harmonisation does not exist (Tanzi and Bovenberg 1990). In this situation, it could be argued that the only realistic way of tackling tax policy in the EU is to harmonise around the average of existing systems. This is precisely what the Commission had suggested in its 1975 proposed directive, later withdrawn. But the case for harmonising around existing tax systems is weak and ultimately flawed. Indeed, if all national tax systems are more or less sub-optimal, picking up the average of existing tax systems as the value around which tax harmonisation should proceed can produce undesirable results. Indeed, the average of non-optimal national solutions is itself sub-optimal and can decrease the overall efficiency of EU taxation. Not surprisingly then, the 1975 proposed directive did not make much progress in the EU policy process and was eventually withdrawn by the Commission.

A solution to the tax impasse could be articulated as follows. If national tax systems do not provide ‘anchors’ capable of triggering isomorphic process, perhaps solutions elaborated through the rigorous application of optimality criteria could be acceptable to the ‘inter-governmental’ solutions. Economists have proposed not one, but a whole ‘menu’ of optimal taxes from which actors engaged in policy transfer could choose. The names of Kaldor, Meade, Lodin, and Bradford are associated with a specific proposal for optimal structures of corporation tax. More recently, the flat-rate tax concocted by Hall and Rabushka (1995) has become a theme of political debate in the US presidential campaign. Finally, a number of British experts have prepared a proposal for a ‘neutral’ corporation tax which, according to its proponents (Gammie 1991; IFS 1994), is particularly suitable for EU tax harmonisation.

However, the ‘optimal taxation’ approach, although procedurally impeccable, has not produced results: ‘optimal’ taxes are not on the agenda of the EU direct tax policy process. The problem is that rational-synoptic approaches crash against the wall of legitimacy. EU institutions have limited legitimacy in the EU policy process. They are not Demiurges which can impose optimal models on different national tax institutions crafted by history and politics (10). To paraphrase Aaron Wildavsky (1987) intellectual cogitation (in this case, the elaboration of optimal solutions external to the policy process) is never a substitute for social interaction.

The previous remarks illustrate that the legitimacy of EU institutions in promoting policy transfer in the EU is inherently constrained. In the two examples analysed so far, the Commission is more a catalyst or a broker in the diffusion of existing models than the creator of solutions that are transferred. Put differently, given the limited legitimacy of EU institutions in imposing models upon member states, the role of EU institutions is to catalyse isomorphic processes. Media concentration policy confirms this assertion, but shows how the Commission to a certain extent is capable of devising instruments which are first purposefully ‘inseminated’ in member states and then proposed for isomorphic diffusion to the whole of the EU.

The regulation of media concentration has become an important issue in both national and EU policy agendas. At the EU level, the Commission is about to launch a draft directive (Beltrame 1996; Harcourt 1996; 1998 forthcoming). At the national level,

in the UK, new legislation on media concentration, as initiated by the Department of National Heritage, were presented in the new Broadcasting Act of July 1996 and will come into force on November 1 1996. Similarly, in January 1996, the German Länder proposed centralisation of media concentration rules which resulted in a cross-Länder treaty also agreed upon in July 1996. In Italy, one of the earliest acts of the Prodi government was the drafting of a bill for the creation of an authority which will deal with media and telecommunications.

The EU media concentration policy process has witnessed a persistent attempt of the Commission to finalise a draft proposal. The European Parliament has endorsed the issue of EU rules. Yet European institutions have had huge difficulties in making progress in this area so far. Media ownership regulation touches upon fundamental issues of pluralism and cultural identity and therefore the legitimacy of EU institutions is severely constrained. Suffice it to say that the EU is bereft of a right (with constitutional status) to free communication, which is the most important foundation for national media concentration legislation. However, as mentioned earlier, the Commission is about to present a proposal which is deemed to represent the most important proposed directive in the single market domain for the current year. Isomorphism has been instrumental in this advancement of the EU media policy process, but in a very unconventional way. Indeed, it is argued here that the Commission, by suggesting the adoption of certain policy instruments in national legislation, created the pre-conditions for isomorphic processes.

The aspect of this policy process most useful for the analysis of supra-national policy transfer and isomorphism is the choice of policy instruments. Indeed, the most innovative component of the soon to be announced EU draft directive is its stress on audience share (11) as the key instrument for measuring the strength of media companies. The Commission, in its two green papers on media concentration regulation (12), was the first to suggest this utilisation of audience share measurement as the most appropriate policy instrument. Interestingly, both in the UK and in the Germany the Commission’s idea of audience share measurement for media concentration found favourable reception amongst national policy-makers who were revising domestic rules.

Why were national policy-makers so keen on taking on board the Commission’s suggestion? Presumably, policy instruments affect the position of different actors in the policy process and consequently attention must be turned to the relationship between instruments and actors. In the UK, the utilisation of audience share - a ‘cultural’ measure, as opposed to more ‘economic’ indicators and measures - has strengthened the grip of the Department of National Heritage on media ownership regulation. Other measures would have shifted the control of agenda from this Department to the Department of Trade and Industry. For example, regulating media through market shares (or revenue shares) assigns media concentration policy the status of a sui generi competition policy. This is why the role of national bureaucracies is highlighted in table one. More pertinently, inter-bureaucratic competition for agenda control (between the Department of National Heritage and the Department of Trade and Industry) is the crucial factor which brought about the adoption of this apparently innocuous instruments of regulation in the UK. By ‘buying’ suggestions from Brussels, national bureaucracies hence became implicitly involved in the early stage of supra-national policy transfer. Similarly, in Germany the choice of ‘audience share’ ensured that the jurisdiction of media concentration policy remained with the Länder (state) governments, which are constitutionally responsible for cultural policy concerns, and was not transferred to the federal government’s cartel office (Bundeskartellamt).

Once audience share measure entered the UK and German political debate, the Commission turned to its congenial role of catalysing isomorphic processes. Indeed, the Commission is now arguing that its announced directive is nothing but the extension of national regulatory instruments to the EU level. The UK and German governments, which are the only EU member states not to have endorsed the competence of the Commission in this area, may therefore be shooting themselves in the foot by adopting precisely the policy instrument suggested by the European Commission in its two green papers. A diffusion of the audience share model in national legislation will make it easier for the Commission to propose EU legislation based upon this policy instrument. The case of media ownership regulation should not be over-emphasised: all in all, this EU policy process is at its initial stage, with a directive ‘announced’ but not formally submitted. However, media regulation shows that legitimacy in the EU politics is not a constant, but a variable resource that can be created with ad-hoc ‘insemination’ of instruments to be transferred.

This paper has explored the relationship between policy transfer and legitimacy in the EU by using a concept, isomorphism, developed within the new institutional approach to organizational behaviour. To sum up the empirical results, EU policy transfer contains elements of isomorphism, particularly mimetic isomorphism. Under conditions of uncertainty, EU institutions catalyse mimetic processes of isomorphism. These processes are strengthened by the existence of additional aspects of normative isomorphism, as shown by the political role of monetary professionals in EMU (13). Finally, recent EMU-induced reforms of the welfare state and the pressure of convergence criteria highlight the role of coercive elements.

The three cases examined in this article do not allow for broad generalizations. However, within the tight limits of the empirical evidence provided here, some conclusions can be drawn. To begin with, the presence of national examples to be transferred is an important pre-requisite for isomorphism. EU politics is perhaps less concerned with dramatic innovation than with systematic imitation. As observed long ago, ‘there is no new thing under the sun’ (Ecclesiastes, quoted by Kingdon 1984:148). Judging from the cases of EMU and tax policy, the conclusion would be that the presence of desirable national models is a necessary and sufficient condition. But this conclusion is flawed: the picture changes as soon as one considers the third case study (media ownership policy). The third case study provides two lessons. First, however important a national model may be, it does not represent a necessary condition for policy transfer. The Commission has been able to set the agenda for EU media ownership policy without drawing upon lessons provided endogenously by member states. Second, when there are no models, isomorphic processes can be triggered by the strategic construction of national examples (14), as shown by ‘in utero’ insemination of policy instruments in media ownership regulation. Bearing in mind Offe’s observation that design is often presented as imitation even when it is creation (Offe 1992), the strategy pursued by the Commission makes political sense. For a policy entrepreneur acting in the constrained EU political system, the dissemination of best practices and models is an efficient strategy for building legitimacy. The Commission cannot be accused of trying to impose ‘the view of Brussels’ if it follows a policy design that is already in place somewhere in the EU... no matter if the national policy design was originally inspired by Brussels!

In terms of theoretical progress, this article has provided preliminary evidence that the concept of isomorphism has potential for explaining ‘process and outcome’ (to paraphrase Caporaso 1996:30) of policy transfer. This strengthens the case for combining a conceptual framework (policy transfer) with models of the policy process (in the cases explored by this paper, models of the legitimation process). There is a problem with the use of institutional isomorphism in the context of policy transfer, however. Whilst the policy transfer literature is part of an explanation of political life which is based upon the role of political actors as rational decision-makers, new institutionalism underlines the ‘taken-for-granted’ aspects of political life (March and Olsen 1989;1995). Policy transfer, and more generally the whole literature on policy diffusion, assumes that policy diffusion is a rational process wherein imitation, copying and adaptation are the consequences of rational decisions by policy-makers. This stark contrast should not be over-emphasised: interestingly, authors working on policy diffusion have challenged the logic of choice and have adopted a ‘processual perspective’ which goes beyond the ‘mechanical transfer model’ (Mörth 1997:6; see also Egeberg 1980). But the fact remains that the majority of studies on policy transfer downplay the logic of appropriateness and put emphasis on the logic of choice.

The question then arises whether the logic of choice and the logic of appropriateness should be treated as complementary of not. Although a case has been made for complementarity between choice and appropriateness (cf. Ostrom 1991 and Dowding 1994), further research on policy transfer has potential for providing evidence and theoretical insights for addressing this question, which is fundamental for the development of political science. This is yet another reason - and presumably the most important one - for intensifying research on policy transfer.

(1.) On the benefits of a ‘comparative politics’ perspective for the study of the EU see Hix’s article in West European Politics (1994) and the lively debate which followed (Hurrell and Menon 1996; Hix 1996). Back to (1)

(2.) See Dolowitz and Marsh (1996) and their discussion of similarities and differences between policy transfer and lesson-drawing as introduced by Rose (1991; 1993). Back to (2)

(3) For different approaches to the concept of political legitimacy see Dogan (1992), Barker (1990), and Beetham (1991). On legitimacy in the EU see Obradovic (1996). Back to (3)

(4) Following DiMaggio and Powell (1991:64-65) an organisational field denotes ‘those organizations that, in the aggregate, constitute a recognised area of institutional life: key suppliers, resource and product consumers, regulatory agencies, and other organizations that produce similar services or products’. Back to (4)

(5) It should be observed that this body of research belongs to the so-called new institutionalism in organisational analysis, which shares many themes and arguments with the new institutionalism in political science (March and Olsen 1989; 1995). Currently an important debate within organizational studies involves the advocates of new institutionalism, such as Powell and DiMaggio, on the one hand, and proponents of the ‘narrative approach’ (Czarniawska 1997), on the other. Back to (5)

(6) In a recent critique of DiMaggio and Powell, Czarniawska (1997:186 emphasis in original) argues that mimetic processes should be ‘interpreted in terms of fashion (a social phenomenon) rather than of imitation (a cognitive process). Back to (6)

(7) The literature on EMU is vast. Here I draw upon Bovenberg and de Jong (1997), Dyson (1994), Cameron (1995), Sandholtz (1993),Verdun (1996), and, for the post-Maastricht crisis, Sandholtz (1996). Back to (7)

(8) I borrow this distinction between ‘political acceptability’ and ‘technical feasibility’ from a conversation with Kenneth Dyson. Back to (8)

(9) Directive 23 July 1990, 90/435/EEC (parent-subsidiary directive); directive 23 July 1990 90/434/EEC (merger directive); convention 23 July 1990 90/436/EEC. Back to (9)

(10) See Steinmo (1993) for an analysis of tax systems in terms of historical institutionalism. Back to (10)

(11) As opposed to other instruments, such as for example market share or traditional methods of measurement such as licensing. Back to (11)

(12) COMM (92) 480 final "Pluralism and Media Concentration in the Internal Market" 23.12.92.; COM (94) 353 final "Follow Up to the Consultation Process Relating to the Green Paper on ‘Pluralism and Media Concentration in the Internal Market- an Assessment of the Need for Community Action’", 05.10.1994. Back to (12)

(13) The idea of normative isomorphism can perhaps be extended to cover the presence of copinage technocratique in the EU policy process, i.e. the interaction between the Commission and elites of policy experts in national public administrations and pressure groups. Further, as already mentioned, EU politics is characterised by the active presence of epistemic communities which favour normative isomorphic processes. Back to (13)

(14) Policy analysts have observed that the Commission plays the role of a purposeful opportunist in the EU policy process (Cram 1993). Back to (14)

BARKER R. (1990) Political Legitimacy and the State, Oxford: Clarendon.

BEETHAM (1991) The Legitimation of Power, London: Macmillan.

BELTRAME F. (1996) ‘Creating a directive on pluralism and media concentration: a case study of the EU legislative process, Paper prepared for the W.G. Hart Legal Workshop on Law-making in the European Union, London: Institute of Advanced Legal Studies.

BOVENBERG A.L. and A.H.M. DE JONG (1997) ‘The road to economic and monetary union’, Kyklos, 50(1): 83-109.

BUFACCHI V. (1994) ‘Is European integration politically legitimate?’, History of European Ideas, 19(1-3): 229-235.

CAMERON D. (1995) ‘Transnational relations and the development of European Economic and Monetary Union’, in T. Risse-Kappen (ed) Bringing Transnational Relations Back In, Cambridge: Cambridge University Press. PAGE DETAILS

CAPORASO J. (1996) ‘The European Union and Forms of State: Westphalian, Regulatory or Post-Modern?, Journal of Common Market Studies, 34(1) March: 29-52.

------ and J. KEELER (1995) The European Union and Regional Integration Theory, in C. Rhodes and S. Mazey (ed) The State of the European Union, vol.3: Building a European Polity?, London: Longman, 29-62.

Commission (1977) Report of the Study Group on the Role of Public Finance in European Integration [MacDougall Report], Brussels.

------ (1992) Report of the Committee of Independent Experts on Company Taxation, [Ruding Report], Luxembourg: Office for Official Publications.

------ (1996) ‘Taxation in the European Union: report on the development of tax systems’, Brussels, 22 October, COM (96) 546 final.

CRAM L. (1993) ‘Calling the tune without paying the piper? Social policy regulation: the role of the Commission in European Community social policy’, Policy and Politics, 21(2): 135-146.

CZARNIAWSKA B. (1997) Narrating the Organization. Dramas of Institutional Identity, Chicago and London: University of Chicago Press.

EASSON A.J. (1992) 'Harmonization of direct taxation in the European Community: from Neumark to Ruding', Canadian Tax Journal, 40(3): 600-638.

DIMAGGIO P. and W.W. POWELL (1991) ‘The iron cage revisited: institutional isomorphism and collective rationality in organizational fields’, in W.W. Powell and P. DIMAGGIO (eds) The New Institutionalism in Organizational Analysis, Chicago: University of Chicago Press, 63-82.

DOGAN, M. (1992) ‘Conceptions of legitimacy’, in M. Hawkesworth and M. Kogan (eds) Encyclopaedia of Government and Politics, London: Routledge: 116-26.

DOLOWITZ, D. and D. MARSH (1996) ‘Who learns from whom: a review of the policy transfer literature, Political Studies, XLIV: 343-57.

DOWDING K. (1994) ‘The compatibility of behaviouralism, rational choice and new institutionalism’, Journal of Theoretical Politics, 6(1): 105-17.

DYSON K. (1994) Elusive Union, London: Longman.

EGEBERG, M. (1980) ‘The fourth level of government: on the standardisation of public policy within international regions’ Scandinavian Political Studies, 3: 235-248..

FURNER, M.O. and B. SUPPLE (1990) (eds) The State and Economic Knowledge. The American and British Experiences, Cambridge (MA): Woodrow Wilson International Center for Scholars and Cambridge University Press.

HARCOURT, A. J. (1996) ‘Regulating for media concentration: the emerging policy of the European Union’, Utilities Law Review October: 202-210.

HARCOURT, A.J. (1998, forthcoming) ‘EU media concentration. The conflict over the definition of alternatives’ Journal of Common Market Studies March.

HIX, S. (1994) ‘The study of the EC: the challenge to comparative politics’, West European Politics 17(1): 1-30.

HIX S. (1996) ‘CP, IR and the EU! A rejoinder to Hurrell and Menon’, West European Politics, 19(4) October: 802-804.

HURRELL and A. MENON (1996) ‘Politics like any other? Comparative politics, international relations and the study of the European Union’ West European Politics 19(2) April: 386-402.

KINGDON J.W. (1984) Agendas, Alternatives, and Public Policies, HarperCollins.

MARCH, J.C. and J.P. OLSEN (1989) Rediscovering Institutions. The Organizational Basis of Political Life, New York: Free Press.

------ (1995) Democratic Governance, New York: The Free Press.

MAZEY, S. and J.J. RICHARDSON (1995) ‘Promiscuous policy-making: the European policy style?’, in C. Rhodes and S. Mazey (eds) The State of the European Union Vol.3: Building a European Polity?, Boulder: Lynne Rienner, 337-359.

MORAVSICK, A. (1993) ‘Preferences and power in the European Community: a liberal intergovernmentalist approach’, Journal of Common Market Studies, 31(4): 473-524.

MÖRTH, U. (1997) ‘Policy diffusion in research and technological development: no government is an island’, Paper prepared for the ECSA conference, Seattle, 29 May-1 June.

OBRADOVIC, D. (1996) ‘Policy legitimacy and the European Union’, Journal of Common Market Studies 34(2) June: 191-221.

OFFE C. (1992) ‘Designing institutions for East European transitions’, Paper prepared for the institutional design conference, The Australian national University, Research School of Social Science, December, 7-8.

OSTROM E. (1991) ‘Rational choice theory and institutional analysis: toward complementarity’, American Political Science Review 85:237-43.

PETERS G.B. (1994) ‘Agenda-setting in the European Community’, Journal of European Public Policy, 1(1): 9-26.

POWELL W.W. and P.J. DIMAGGIO(1991) (eds) The New Institutionalism in Organizational Analysis, Chicago: University of Chicago Press.

PUTNAM R. (1988) ‘Diplomacy and domestic politics: the logic of two-level games’, International Organization, 42 (2): 427-460.

RADAELLI C. (1997) The Politics of Corporate Taxation in the European Union. Knowledge and International Policy Agendas, London: Routledge.

RICHARDSON J.J. (1996) ‘Policy-making in the EU: ideas and garbage-cans of primeval soup’, in J.J. Richardson (ed.) The European Union: Power and Policy-Making, London: Routledge, 3-23.

ROSE R. (1991) ‘What is lesson-drawing?’ Journal of Public Policy, 11: 3-30.

------ (1993) Lesson-Drawing in Public Policy, Chatham (NJ): Chatham House Publishers.

SANDOLTZ, W. (1993) ‘Choosing union: monetary politics and Maastricht’, International Organization, 47 (1): 1-39.

------ (1996) ‘Money troubles: Europe’s rough road to monetary union’, Journal of European Public Policy, GET DETAILS

SCHARPF F.W. (1988) ‘The joint-decision trap: lessons from German federalism’, Public Administration, 66(3): 239-278.

SCOTT R.W. (1995) Institutions and Organizations, London: Sage.

SINN H.W. (1990) ‘Tax harmonisation and tax competition in Europe’, European Economic Review, 34: 489-504.

STEINMO S. (1993) Taxation and Democracy, New Haven: Yale University Press.

UNDERHILL G. (1995) ‘Keeping governments out of politics: transnational securities markets, regulatory cooperation, and political legitimacy’, Review of International Studies, 21: 251-278.

VERDUN, A. (1996) ‘EMU-The product of learning and consensus amongst monetary experts’ Human Capital Mobility Network Occasional Papers, no.7.

WILDAVSKY A. (1987) The Art and Craft of Policy Analysis, New Brunswick, NJ: Transaction Books. First edition: 1979.

ã Copyright 1997. Jean Monnet Chair of European Comparative Politics.

Claudio M. Radaelli.

Lecturer in comparative public policy. Department of European Studies. University of Bradford.

Bradford BD7 1DP West Yorkshire, England.

c.radaelli@bradford.ac.uk