The Side Effects of the Bologna Process on National Institutional Settings: The Case of France

Abstract

The successive declarations which punctuated the Bologna process since 1998 promote large ambitious goals – among them, competitiveness, quality, mobility, etc. – and rely on two main objectives to attain these goals. The first one has already been achieved in almost all signing countries and consists in the generalisation of a two-tier structure (bachelor and master) for all European curricula, most of them corresponding to a three year + two year cycle, even if variations are still to be found from one country to another or from one institutional sector to another in the same country (see, for instance, Witte 2006).

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

The successive declarations which punctuated the Bologna process since 1998 promote large ambitious goals – among them, competitiveness, quality, mobility, etc. – and rely on two main objectives to attain these goals. The first one has already been achieved in almost all signing countries and consists in the generalisation of a two-tier structure (bachelor and master) for all European curricula, most of them corresponding to a three year + two year cycle, even if variations are still to be found from one country to another or from one institutional sector to another in the same country (see, for instance, Witte 2006). The second objective (which is still being pursued) is the generalised implementation of compatible quality assurance systems guaranteeing the respect of a minimal standard of equivalent quality among all training programmes. In 2005, the European Association for Quality Assurance in Higher Education (ENQA 2005) published a document called Standards and Guidelines for Quality Assurance in the European Higher Education Area which aims at setting down some of the principles which could frame the development of such quality assurance systems.

The Bologna process is therefore not an ordinary reform as it does not directly address the transformation of the national higher education institutional settings within Europe. It does not try to modify the status of universities; it does not aim to transform the state–university relationships; it does not propose to introduce new tools to improve the management of the academic profession; it does not state how to allocate budgets; it does not intend to modify university governance, etc. While most reforms in the 1980s and 1990s in the EU countries (Eurydice 2000; Musselin 2005a) were expected to redesign the national institutional settings, the Bologna process on the contrary promotes another level of action. It thus aims, first to change the ‘products’ of higher education (i.e. the degrees) and to normalise them by recognising three main degrees: the bachelor, the master and the doctorate. Second, it intends to transform the higher education ‘production processes’ through the introduction of the European Credit Transfer System (ECTS),¹ modularisation, etc. and also through the promotion of coherent quality assurance processes among countries. The products and the way they are produced are the main points of impact chosen by the Bologna process (Musselin 2005b) when one looks at the different declarations and relies on the explicit goals of these reforms.

Nevertheless, does it have an impact on the national institutional settings? This is the question which will be discussed in this contribution. Two different ways could be used to address this question. The first one would be to reveal the implicit institutional models which can be ‘discovered’ when analysing the different texts, devices and reports accompanying, preparing and framing the different ‘Bologna conferences’. This will not be the way chosen here even if a quick analysis certainly suggests it would be relevant: it is clear that the documents pertaining to the Bologna process implicitly promote an ‘ideal’ institutional model based on autonomous, accountable and responsible higher education institutions.² But, rather than exploring this implicit ideal, I will take another perspective and question whether the concrete implementation of the Bologna process in the concerned countries has been disconnected from the reform of the institutional national settings, that is, whether implementing Bologna can be done without transforming the latter on the one hand and/or whether the Bologna reform has not been used by the national governments as a vehicle to transform the latter, on the other hand. Focusing on the specific case of France on which a study³ was led in 2005 by Mignot-Gérard and Musselin (2006) it will be argued that both mechanisms can be observed.

First, it will be argued that, as in many other countries engaged in this reform (see, for instance, Gornitzka 2006; Krücken et al. 2005; Witte 2006), a ‘re-nationalisation’ of this process happened because French public authorities used this reform to achieve other objectives. In particular, the French ministry grasped this opportunity to simultaneously promote university autonomy, standardise the degree offer among the different institutional sectors and transform the state–university relationships.⁴

But, second, we also observed some ‘side effects’ which have had an impact on the national institutional settings as a consequence of the implementation of the two-cycle structure of French universities. For instance, this reform fostered cooperation and joint development of curricula among higher education institutions located in the same city; it questioned the internal structure of universities and led to some reorganisation; it also reinforced the power of academics on the management of curricula to the detriment of the university level.

The issues addressed in this contribution are therefore very close to those discussed in the literature dealing with Europeanisation processes when considering the impact of European Union policies on domestic change. This is the reason why Europeanisation will be the analytical perspective used in this chapter to explain what happened in the French case. It is certainly not the only relevant approach which can be used to study Bologna;⁵ but, because the Bologna process is not an EU policy, the conditions by which it can be understood as a Europeanisation process should first be clarified. The Europeanisation issues which will be dealt with in this chapter should also be clearly defined. That will be the object of the following section.

The Bologna Process as A Europeanisation Process

In the case of France, it may be more difficult than in other countries to argue that the Bologna process is a Europeanisation process because the idea of the two-cycle structure has been developed by the French. Claude Allègre (French Minister of Education from 1997 to 2000) and his collaborators introduced this point in the first declaration signed by the Ministers of Great Britain, Germany and Italy at the Sorbonne Conference in May 1998.⁶ Nevertheless, helped by the fact that this reform finally took the name of the Italian city where the second conference was held, the introduction of what is called in France the ‘LMD’⁷ – licence (for bachelor), master and doctorate – is perceived by French faculty members as a non-escapable move ‘imposed’ by ‘Europe’. As a result, the difference between this non-constraining intergovernmental process and a directive of the European Commission is not always clearly understood or made by French interviewees.⁸

But this confusion is not the main reason why the Bologna process should be analysed as a Europeanisation process.

Conditions by Which the Bologna Process can be Studied as a Europeanisation Process

If one restricts Europeanisation to the third of the five⁹ ‘faces’ of this notion identified by Olsen (2002), that is, to what he calls the ‘domestic impacts of European-level institutions’, the definition proposed by Radaelli (2001) to qualify such impacts is probably the most complete as it includes institutional, cognitive and normative aspects. For Radaelli (2001: 110), Europeanisation is the ‘process of (a) construction; (b) diffusion; and (c) institutionalisation of formal and informal rules, procedures, policy paradigms, styles, ‘ways of doing things’ and shared beliefs and norms which are first defined and consolidated in the making of EU public policy and politics and then incorporated in the logic of domestic discourses, identities, political structures and public policies’ (my emphasis). This definition applies perfectly to the Bologna process (with the major exception being my emphasised words!). As mentioned above, the Bologna process is intergovernmental and does not only concern EU countries: therefore it is not ‘consolidated in the making of EU public policy and politics’. For Radaelli (2001), this would exclude the Bologna process from the realm of Europeanisation studies. Nevertheless, this position is not shared by all Europeanists. In their recent book, Palier et al. (2006) adopt and argue in favour of a less restrictive frame. They first suggest going beyond ‘EU policies’ strictly speaking and to look at sectors in which the European Commission has no specific competence and produces either few or no directives and public policies. They also include in their studies less coercive and legal types of actions which nevertheless allow for Europeanisation processes, through less restrictive ways.

But even this enlargement is insufficient if one wants to analyse the Bologna process as a Europeanisation process. Two supplementary conditions must be added. First, processes, instruments and decisions should be included even when not led by the European Commission; second, processes, instruments and decisions affecting EU countries but also non-EU state members, should also be considered. Rather than restricting the scope of study – as suggested by Radaelli (2001) in order to avoid naming everything ‘Europeanisation’ and thus weakening this concept – I suggest to avoid this risk by distinguishing and qualifying different kinds of Europeanisation processes. Some could be called neofunctionalist or EU-nisation when led by the commission, while others could be more intergovernmental, and among them a further distinction should be introduced between constraining and soft processes. A supplementary distinction could be introduced between those which are EU limited and those which are expected to have larger impacts. As a result, the Bologna process can be analysed as an intergovernmental, weakly constraining Europeanisation process with an impact on European countries, or, following Radaelli’s (2001) terms, as a ‘process of (a) construction; (b) diffusion; and (c) institutionalisation of formal and informal rules, procedures, policy paradigms, styles, ‘ways of doing things’ and shared beliefs and norms’ first defined and consolidated in the making of intergovernmental public policy and politics ‘and then incorporated in the logic of domestic discourses, identities, political structures and public policies’ of EU countries and other European countries.

A Focus on Re-Nationalisation Effects Linked to Europeanisation Processes

Within this framework, I will only focus on one specific aspect among those under scrutiny in Europeanisation studies. More precisely, two issues addressed by this approach will not be discussed here, that is, the degree of change and the mechanisms by which Europeanisation occurs. This chapter is thus aiming at neither measuring the degree of impact of this process¹⁰ on the French system nor assessing whether one should speak of inertia, absorption, accommodation or transformation.¹¹ It also does not intend to identify the mechanisms of Europeanisation involved in the Bologna process¹² or to qualify this process itself.¹³ In order to answer our questions about Bologna and national institutional settings, the chapter’s main focus will be on re-nationalisation processes linked to Europeanisation processes.

Relying on the existing literature on this point, two main meanings of re-nationalisation can be distinguished but, for each of them, further alternatives have to be suggested.

In most approaches, re-nationalisation characterises the transformation experienced by Europeanisation processes when confronted with national settings, structures, preferences and beliefs. Building on examples from Börzel (1999) and Caporaso et al. (2001), Radaelli (2001: 130), for instance, states that ‘in addition, institutions determine the distribution of resources among domestic actors affected by Europeanization. The result is that the impact of Europeanization is contingent on institutional factors’. But this conception is too overly top-down – it focuses on how measures emerging from European initiatives confront or accommodate national settings. Moreover it is one-sided: it only describes how national characteristics limit/allow the impact and transform/respect the content of European measures. But the other side of the coin (or another form of re-nationalisation having to do with national institutional settings) should also be considered. In particular, one should observe the side effects such measures have on the national¹⁴ level in order to be implemented or when being implemented. This means taking into account the country specific characteristics and institutions which are nevertheless affected by these measures and that experience specific national changes, although not directly.

In some other approaches, re-nationalisation refers to another phenomenon, that is, the capacity (or not) of national states to still play a role. It is then understood as the moment when ‘governmental actors fear to be ‘overwhelmed’, and start elaborating strategies to take control over [the European] policies’ (my translation) (Palier et al. 2006: 75). As outlined by these authors, ‘for the state to come back in, it first has to transform and thus to Europeanize itself’ because ‘it must control the new rules to become the master of the game or at least come back into the game’ (my translation) (Palier et al. 2006: 75). But even if I agree with this last assumption, I would argue that this conception only emphasises defensive reactions from the states. One should also pay attention to more offensive forms of re-nationalisation, by which governmental actors re-nationalise the process, not so much by taking control over the European measures but by using them to tackle domestic objectives or problems.

We thus now have four mechanisms of re-nationalisation, two linked to national institutional settings and two others to national governmental actors (see Table 1).

Table 1 Four forms of re-nationalisation linked to Europeanisation processes

Because of the intergovernmental, weakly constraining character of Bologna as a Europeanisation process, all four forms of re-nationalisation do not apply. In particular, there is no evidence, at least in the data collected with Mignot-Gérard (Mignot-Gérard and Musselin 2006), of the governmental actors taking over control of the Bologna process, probably because the latter is not a top-down compelling directive but a non-constraining commitment they accepted to sign and even participated in its design. If the members of the French ministry had to Europeanise themselves to develop the Bologna process (or at least become more aware of other European higher education systems and of the position of France among them), they never felt like having to ‘take over control’ of something that would have been imposed on them.

Below, I shall therefore focus only on the three other mechanisms. I shall start with the capacity of national institutions to affect the implementation of the Bologna process (form 1) because this mechanism did not impact the institutional national settings but rather showed their strength. I shall then turn to the way the French governmental actors used the LMD as an opportunity to tackle domestic issues (form 4) and in particular to transform some of the French institutional national settings, thus modifying the original goal of this reform. I shall conclude by identifying some of the side effects (form 2) the implementation of the LMD had on some institutional national settings.

The Impacts of the French Institutional Settings on the Implementation of the Bachelor–Master Scheme in France

Similar to many other European, international or global processes, the Bologna reform is confronted with specific institutional contexts in the country where it is implemented. There is therefore no automatic and similar declination of this reform in each country: when the same measures are ‘applied’ on different national settings, the latter incorporate the European measures and transform them into a specific national mixture. This is a well-known mechanism which has been studied many times by the tenants of the neo-institutionalist approach in comparative politics (Steinmo et al. 1992; Thelen 1999). When they compare different countries confronted with the same external phenomena (an oil crisis, for instance), with the same set of ideas or theory (for a study about Keynesian theories, cf. Hall 1989), or with the same policy instruments (as the creation of governmental agencies), they observe different ways of implementation, different impacts and different acceptances of these measures.

In the case of the bachelor–master scheme, Witte (2006) for the UK, the Netherlands, France and Germany clearly showed how the 3+2 structure, which finally emerged as a common reference, found specific declinations and led to many exceptions to the ‘rule’. The same has been observed by Krücken et al. (2005) for Germany, by Gornitzka (2006) for Norway, by Alesi et al. (2005) for the seven countries they studied and was also stressed by the Trend IV report (Reichert and Tauch 2005). Adaptations have been negotiated and accepted in order for the Bologna process to fit into the national specificities, constraints or preferences. In France, the grandes écoles constitute a first good example of such amendments: as access to the more reputed of them occurs after two years of special classes (held in lycées and finishing with a very selective exam), the grandes écoles did not introduce a bachelor degree and ‘only’ deliver masters. But exceptions are also observed within French universities: the job-oriented professional degrees delivered two years after the baccalauréat by the IUTs (University Institutes of Technology) have been maintained (and not expanded to a three-year degree). The same holds true for the BTS (higher degree for technicians) which is delivered two years after the baccalauréat, mostly offered by public or private lycées.

A further example of adaptation can be found with the division of master programmes into sub-levels, called mentions which can themselves be subdivided into spécialités. Trying to push for a reduction in the number of degrees and to rationalise the training offer, the ministry encouraged the creation of a large domain (possibly interdisciplinary) for each master. This was aimed at reshuffling the previous structure in which the fifth year at university was organised in various DEAs (research-oriented) and DESSs (job-oriented) degrees, each led by an academic in close interaction with some research teams or labs. But in fact it did not change very much. In many places, the master is not the relevant level of management. Some universities even decided not to name someone in charge of each master. The subdivision of masters into mentions led to a kind of replication of the former structure and the academic who chairs the mentions are the ‘real’ levels of decision and responsibility of the new training programmes. As a result, there is often confusion (voluntary or non-voluntary) between the different notions and it is frequent to hear the head of a mention (for instance, in molecular biology) in a master (for instance, in life sciences), speaking of ‘their master in molecular biology’. The small territories drawn before by the DEA and DESS have therefore been redesigned (if not recreated) in mentions or spécialités.

A frequent explanation of such phenomena is what Pierson (2000) called a broad¹⁵ version of path dependence which stresses the stability and inertia of already existing patterns, institutions, routines, etc. This of course applies in this case, but, in order to be more explicit than just recognise that ‘history matters’, it is necessary to identify where resistance came from. Two factors are important in this specific case. The first one is linked to the notion of bounded rationality first developed by Simon (1945): facing the need to find new solutions within a short time, academics often agree on solutions at hand rather than engage in a costly and lengthy search for innovative and new ones. The second factor is what I call the ‘paradoxical strength of formal structures’ within universities (Musselin 1990, 2006b). Unlike in other organisations, formal structures in universities fail to organise cooperation and coordination among individuals. Because teaching is a loosely coupled activity (Weick 1976) requiring few interactions with other colleagues and building on unclear technology (Cohen et al. 1972), belonging to the same department or teaching in the same programme does not imply strong cooperation with other members of the department or programme. But this does not mean that such formal structures do not play a role. They in fact act as defensive territories which become very visible each time they are threatened¹⁶ and which tend to reappear each time they are suppressed: the mutation of the DEA and DESS into mentions or spécialités seems a nice illustration of such a mechanism.

Whatever the explanations for it, this first re-nationalisation mechanism plays in favour of the stability of the already existing institutional settings, or to be more precise explains one aspect of the re-nationalisation processes affecting Europeanisation, by the strength of these settings. The next mechanism on the contrary affects the national settings and ‘nationalises’ at the same time.

A European Reform Which Also Serves Domestic Aims

This second mechanism is linked to the fact that each country rarely arrives exactly at the same European policy as its neighbours because national governments often aggregate other objectives or measures to a specific reform. The more spectacular example for this in the case of the Bologna process certainly is the ‘quality reform’ led in Norway in the name of Bologna but which affects many other aspects (Gornitzka 2006; Michelsen 2006) including university leadership and governance (Bleiklie 2006).

But similar phenomena can be observed in other countries as well. In the case of France, three supplementary objectives have been pushed by the ministry along with the LMD reform, which aimed at transforming some of the features of the French national institutional settings.

Trying to Blur Some of the Frontiers Between the Universities and the Grandes Écoles

First, the LMD has been used as an opportunity to standardise the degrees offered in France and thus to reshape the distinction between the university and the grandes écoles. To understand what happened it is necessary to recall, first, that in France the agreement to deliver national university degrees is given by the ministry and renegotiated every four years by the universities through a procedure called the habilitation and, second, that since the 1880 act,¹⁷ public universities have been granted the monopoly to deliver degrees such as the DEUG (delivered two years after the baccalauréat), then the licence (bac+3) and the maîtrise (bac+4). This was also true for the DEA and DESS (bac+5) and the PhD programmes, but with few exceptions: some grandes écoles (although rare) can also deliver PhDs. The institutional distinction between the two sectors was therefore accompanied by a clear difference in the type and name of the degrees offered. Today, the type of degrees remains different but the distinction is so subtle that few people really understand because the names are the same. What happened? In 1999, the division in charge of higher education at the ministry resurrected an old distinction between two terms: the notion of grade (which refers to a specific level of study achieved by obtaining a degree) and the notion of diplôme (which refers to the name and content of a specific degree).¹⁸ This subtle but legal nuance allows for the publication of decrees creating the grade of master¹⁹ to name all the degrees obtained in five years. As a result, each ‘bac+5’ degree delivered today by a grande école belongs to the grade of master.²⁰ As the same terminology (master) has been chosen to name the national degree (diplôme national) delivered by universities five years after the baccalauréat, grandes écoles and universities now both offer ‘products’ (degrees) having the same generic name: ‘master’. One is a grade and the other is a degree, but of course only a few people are able to understand the difference. My own institution, Sciences Po, for instance, delivers both a ‘diplôme de Sciences Po’ which is a gradeof master and is not submitted to a ministerial habilitationand a degree of master (national degree), habilité by the ministry and opening the door to the preparation of a PhD awarded by Sciences Po.

This may appear as a small change, and is often considered as such by external observers (Witte 2006, for instance), but it is in fact a supplementary step further experienced by two processes²¹ engaged for many years which brings universities and grandes écoles closer: on the one hand, universities have borrowed some practices from the grandes écoles (introduction of some selective programmes, creation of job-oriented curricula, etc.) for part of their training programmes while, on the other hand, the grandes écoles are transforming themselves into more ‘academic’ higher education institutions (fostering research activities, hiring professors with academic profiles rather than practitioners, etc.). During the LMD implementation, the benevolence of the ministry in favour of the development of national co-degrees between universities and grandes écoles also contributed to try to bring both sectors closer, more intertwined and thus blur the frontiers between them.

The grandes écoles want a master, because even if they are reputed in France, they are not internationally. I do not have a preference for the grandes écoles, I do not have a preference for the universities: I defend the French higher education system as a whole … Masters have to be based on research structures, but this is a problem for the grandes écoles. They are not very good in terms of research and that is a problem for them on the international scene. If we work correctly, partnerships between universities and grandes écoles will develop. It is already clear for the research masters: most grandes écoles created some with the universities (Ministry officer).

This process not only consists of concessions from the university side (loosening of their monopoly); it also involves efforts from the grandes écoles sector. They, for instance, accepted the introduction of specific processes granting accreditation of their degrees for a limited time period,²² that is, the introduction of procedures comparable to the ministerial habilitations existing for the universities. Although still notable, the existing horizontal divide between the two sectors tended to reduce with the LMD.

More Pedagogical Autonomy for the Universities

Second, the LMD has been an opportunity for the ministry to increase the autonomy of French universities in the management of their curricula. Three main facts document this assumption. On the one hand, the suppression of the national patterns (maquettes), which were used to define the content of each training programme leading to a national university diploma. This is more a symbolic than a ‘real’ and concrete modification; as a matter of fact, the national patterns were rather broad and did not define each programme in precise detail. Furthermore, the discrepancy between the project submitted to the ministerial habilitation and its concrete implementation was often rather important. These national patterns had little influence. But they were symbolically important as they illustrated the existence and the apparent maintenance of a nation-wide uniform definition of the training programmes and thus symbolically maintained one of the principles on which the French university system developed: a complete equivalence among all the delivered national degrees whatever the university which awards them, guaranteed by the equivalence in content assumed by the national patterns. Suppressing them therefore meant going a step further towards the acceptance and recognition of differentiation among French universities and a step further towards their pedagogical autonomy.

On the other hand, the next fact in favour of more institutional autonomy deals with the way in which the Bologna reform has been proposed rather than imposed on French universities. While in some other countries (Italy, Norway or the Netherlands, for instance) a date of passage has been set (with little negotiation with the institutions in the two first cases and on their own initiative in the third) (Witte 2006), the French ministry engaged in a progressive process, based on voluntary commitment, far away from its usual top-down directive style. It started at the end of the 1990s with decrees providing the opportunity to French higher education institutions to choose the ‘LMD’ and then (in November 2002) proposed (but did not oblige) to link the introduction of the LMD in each university with the negotiation of its four-year contract (Musselin 2001) with the ministry. The central administration thus relied on the progressive adhesion of the higher education institutions to the process – and it worked!

It’s exemplar. We changed all without changing anything. No text has been suppressed but we proposed something, only proposed, to the institutions. A reform has been proposed and was expected to go on progressively (Ministry officer).

One of the good intuitions for the LMD was to unbalance the system a bit. It created a dynamic which fed upon itself. It made the LMD desirable. Everybody wants the LMD, and especially the ‘M’ (master), not because the ministry said ‘do it’ but because the actors in the university want it (Ministry officer).

The universities whose four-year contracts had to be renegotiated²³ in 2003 had to decide whether they would at the same time introduce the LMD or wait for another four years. In fact, most if not all chose to be among the precursors rather than wait for the next contract and take the risk of being among the latecomers. Some even asked to initiate the LMD before they attained their period of contractualisation. One could of course speak of ‘constrained adhesion’, as many faculty members expressed the feeling that there was indeed no other way and that the introduction of the LMD could not be avoided. One could also speak of a ‘surface adhesion’ as few are really convinced of the real improvement of the curricula through this process but again developed conforming responses in order to appear to be doing what was expected (according to what Meyer and Rowan (1977) first described as a form of institutional isomorphism). Finally, one could speak of adhesion by manipulation, as the ministry succeeded in obtaining adhesion to the reform by consciously using a non-authoritarian method. But the fact is that this reform, contrary to many previous reforms having occasionally a more solid legal background, has been implemented, has incurred relatively few negative protests and is broadly accepted. Among the two pathways, resource redistribution and socialisation, identified by Börzel and Risse (2000: 8) as leading to domestic changes, the Bologna process certainly induces a socialisation process ‘by which actors learn to internalise new norms and rules in order to become members of (international) society’ and which relies on a logic of appropriateness (March and Olsen 1989).

Finally, the nature of the texts of the LMD reform is indicative of the will of the division in charge of higher education at the ministry to leave it to the higher education institutions and to incite them to be innovative. I will return later to the content of these texts but it is important to notice here that they were very ‘light’. This left the universities which first implemented the reform rather perplexed as they were accustomed to much more precise and restrictive notices from the ministry and thus did not know how to prepare their new curricula in order to have a chance to receive the habilitation from the ministry.

We received directives from the ministry but they were so vague that we in fact prepared our projects according to our own interpretations without any real knowledge of what was expected. We had no idea how the ministry would react (An academic).

Central administration finally produced some orientations or principles (for instance, it decided that the masters would be managed by the facultés and not by the doctoral schools, or that there could not be more than one master by domain, etc.) but even if part of the ‘doctrine’²⁴ has been written as the process was going on, this doctrine remains loosely prescriptive, thus leaving to the institutions the responsibility and autonomy to define their own rules, their own standards, their own structures, etc. As a result, the habilitation process has also been rather open. In the three universities we studied, very few projects were refused by the ministry.²⁵ Some transformations have been negotiated and introduced for some projects, but we also observed cases where certain academics decided not to take into account advice given by the experts²⁶ of the ministry assessing the projects and finally received the habilitation.

When I received the assessment from the experts at the ministry, I immediately understood they wanted to reduce expenses. For instance they asked us to merge the spécialité X with the spécialité Y. We refused and argued … and it worked (An academic).

As a matter of fact, the inflation in degrees which occurred with the introduction of the LMD has not been regulated by the ministry, which it potentially could have done.

A Transformed State–University Relationship

Along with the standardisation of degrees and the increasing pedagogical autonomy recognised by the universities, a third transformation of the national institutional settings consisted of modifying the state–university relationships. As mentioned above, these relationships with the LMD have not been directive or coercive compared to the usual French steering mechanisms. Their content was also rather different in nature as it was less substantive than procedural: in the texts, nothing was said about what a masters should consist of, which subjects should be taught and which classes should be included; but broad principles were defined, which did not apply to masters in a specific discipline but to all masters. It thus specified that masters should be linked to research teams, that they should be managed by facultés, that the students should obtain 120 ECTS at the end of the masters, etc. This may seem still rather restrictive to European academics accustomed to being completely autonomous in the design of their curricula, but it is much less formal than the regulations existing for the pre-Bologna degrees. Moreover, the nature of these directives is different as it frames the ‘production process’ of a masters but does not dictate its content (which subjects, how many hours for each discipline, etc.). The introduction of the LMD was thus part of the continuity of the transformation of the French university configuration (Musselin 2001) and from many points of view followed a dynamic change very similar to the introduction of the four-year contract more than 10 years earlier.²⁷ It is part of the ongoing evolution of the modes of steering developed by the ministry in enhancing the role and autonomy of higher education institutions, removing the ministry from the pure production and control of rules thus favouring more regulative behaviours, close to what Neave and Van Vught (1991) described as an ‘evaluative state’.

The LMD reform has been confronted by the French situation and has adapted to it. The larger scope of the reform and the parallel objectives it included not only intended changing the degree structure and the production processes of the degrees within the French universities, but also affected the national institutional settings. The French ministry used the LMD to achieve other specific domestic goals and thus re-nationalised the process.

The Institutional Side Effects Linked to the Implementation of the Bologna Reform

The point to argue in this last section is that the impact of the LMD on the national institutional settings is not only linked to the supplementary objectives the French ministry added to the LMD reform, but also to the consequences of this reform. In other words, transforming the degree structure and the production process attached to it, consequently affected some of the national institutional arrangements. This can be qualified as the institutional side effects of the LMD. Some of them are pure opportunistic effects, while others were not foreseen and popped up, and still others were unintended.

Opportunistic Effects: The Emergence of Higher Education ‘Sites’, or the Site Effects

By opportunistic effects, I refer to cases where some actors took the opportunity of the Bologna reform to launch, or accelerate, other changes which, for some of them at least, had an impact on the national settings. The ‘site effects’ are a good example of this. But in order to understand what a ‘site’ is, it should be remembered that one French institutional characteristic since 1968 is the development of several incomplete²⁸ public universities in big French cities (nine for Paris intra muros, four in Bordeaux, three in Grenoble, etc.). This was the result of the conflicts, political/ideological concerns and preferences, disciplinary-opposed interests which alimented the process by which French universities have been (re)created after the 1968 aftermath. This concentration of institutions (different universities and often different grandes écoles as well) on the same local territory is called a ‘site’. In many cases, the rivalries resulting from this period were still very active by the beginning of the third millennium, sometimes leading to an inflation of competing programmes offered by different institutions on the same site.²⁹ But in others, common efforts had already been developed to try and rationalise the situation, to propose common services (inter-libraries services, for instance) or to develop cooperation (‘Grenoble Universités’,³⁰ for instance, was created in 2001). In most cases, the Bologna process accelerated the trend in the second case and pacified the situation in the first one. None of the universities we studied in our research was concerned by these ‘site effects’ as they all are located in middle-sized towns with only one university, but the report on the LMD prepared by the IGAENR (2005) clearly identifies this phenomenon. In Marseille, for instance, where some classes in the sciences were offered by all three universities, the LMD reform gave rise to coordination efforts among the different science departments. They came to an agreement rationalising the offer on the Marseille area. Such cooperation sometimes occurred also between universities and grandes écoles located on the same ‘site’. It is certain that the LMD has not been the only driver in the development of institutions being brought together when located on the same territory. The Shanghai ranking and the relative absence of French universities in the top 50 have also been an incentive to join forces and attain a larger size; accumulate rather than disperse results and obtain greater visibility. The introduction of the LMD in a time when incomplete universities were beginning to see reasons to develop stronger relationships or even to merge (as the Strasbourg institutions announced) has been an opportunity to develop cooperation in teaching. The objective of rationalising the teaching offer thanks to the LMD provided a supplementary rationale to move in this direction. These ‘site effects’ should increase as the new research act³¹ of April 2004 created new structures, called PRES (pôle de recherche et d’enseignement supérieur, higher education and research poles), which provides a legal basis for several higher education institutions to build common entities locally.

In this second case, it is obvious that the introduction of the LMD accelerated and favoured the transformation of some aspects of the French institutional settings.

The Unforeseen Effects: The LMD Challenges the Current Organisation of French Universities

Another example of such a process is linked to the discussions (and in some cases the decisions) the LMD raised about internal university structures. Since the 1984 act, French universities are organised in UFRs (still often called facultés), responsible for both training (F for formation) and research (R for Recherche). The UFRs correspond to very diverse realities in size and scope (some can, for instance, regroup all the disciplines in humanities, while in other places there will be a UFR in history, a UFR in sociology, etc.). In the late 1990s, the creation of research federations (Instituts fédératifs de recherche) which could be transversal to several UFRs when they are interdisciplinary or constituted around a social problem, tended to deprive the UFRs of their capacity³² of intervention on research issues. The generalisation of doctoral schools within the same period further weakened the domain of action of the UFRs as they lost their responsibility on graduate studies. But the LMD had an even bigger impact. If the bachelor curricula tended to respect the frontiers of the UFRs, the masters often did not and they concerned more than one UFR. This quickly raised very practical issues: which UFR can account for the students registered in the master? Which part of the budget of the master should go to each UFR? Which expenses should be taken over by which UFR? In many universities, this led to lively debates about the relevance of the UFR structure, about the idea of distinguishing the structures for the licence programmes (still organised in UFRs) and the structure for the master curricula, about the possibility – even if the UFRs are maintained – to create specific structures for the management of the master, etc. According to the situation of the interviewee, their conceptions and preferences vary.³³

Today the training structure does not match the administrative structure. Either we suppress the UFRs and we redesign the university, which is an enormous task, or we go on and do what we can (bricole). But we meet real problems in terms of budgets, enrolments. Who does what? (A director of UFR).

Today, UFRs are in charge of the management of the masters and from my point of view it is nonsense. In our master, we will ask for a budget and pedagogical autonomy … When I say ‘we’, I mean those responsible for mentions. We face contradictory institutional logics and ideas: ‘we must ask for our budget autonomy’. Today, if I ask someone to teach in my mention, I have to ask the UFR. It is very surprising! We need transversal administrative units. We need budget autonomy as we already have the pedagogical autonomy. The current structure raises problems. Nobody had seen it before. We created the masters from an intellectual point of view, but now we face the day to day practice (An academic responsible for a mention in a master).

In one of the three universities under study, the presidential team decided very early on to reorganise the administrative staff and create specific administrative entities dedicated to the masters. In parallel, the president launched a reform to reduce the overall number of UFRs in his university. The two other universities discovered the problems while implementing the LMD and were still searching for solutions when we conducted our interviews.

Whatever the degree of consciousness or reactivity to the UFRs’ problem, the introduction of the LMD within universities not only affected the degrees and their production system, it also accentuated the non-appropriateness of the current structures, or, to put it another way, happened to be irrelevant with the existing organisation. It therefore reactivated the arguments and discussions³⁴ on the need for a new university act³⁵ but overall pointed at one of the major and unsolved problems of the French university system: its internal organisation and the desirable relationships to develop between the whole (the university level or the presidential team and its administrative staff) and its parts (the UFRs). Even if not specific to France, this issue is recurrent and especially acute in this country for historical reasons. The universities having been constructed against the facultés,³⁶ the relationships between them are rarely based on cooperation (Mignot-Gérard and Musselin 2002b), and none of the 1968, 1984 or 2007 acts described or prescribed the link or the nature of the link between the presidential team and the deans, or between the university and the UFR levels.

By further questioning the definition of the prerogatives of the UFRs and by creating new intermediary structures and leaders (those in charge of the masters or mentions of masters), the LMD pushed for a reconsideration of the pertinent territories and lines of cooperation within French universities. The initiators of the Bologna process, in the ministry as well as in the institutions, did not foresee this effect but at the present moment have to take it into account.

Unintended Effects: The Revival of the Academic Profession

A last category of effects concerns the unintended consequences of the Bologna process. The difference between these effects and the previous ones is that they contradict the expectations of the reformers.

As argued above, the ministry aimed to increase the autonomy of French universities through the implementation of the LMD. In this sense they pushed further the movement of transformation initiated by the contractual policy at the very end of the 1980s (Musselin 2001, 2006a) and expected the university leaders to use the opportunity of the LMD to take over responsibility of a domain in which they were still lacking legitimacy and on which they still rarely intervened (Mignot-Gérard and Musselin 1999, 2000, 2002a; Kletz and Pallez 2003): the pedagogical offer. This is clearly evident by the fact that the ministry avoided recourse to draconian rules, used soft rather than hard instruments (thus allowing rather than imposing) and preferred procedural rather than substantive orientations. This left initiatives to the higher education institutions but they were at the same time expected to filter the projects and develop a creative but at the same time rational and understandable offer.

As a matter of fact, it rarely happened. While the presidential teams often took the lead in the decision to implement the LMD as soon as possible rather than to wait and see, many (especially among the first to implement the new system) did not manage the process. The presidential teams often used the absence of clear instructions from the ministry to justify their distant steering. Having no ‘clear’ information from the central administration about the rules of the game, that is, how to present the projects in order to get the habilitation, they left it to the academics to develop their proposals and did not filter them.

Some presidential teams nevertheless tried to intervene in certain ways (in the definition of the domain covered by the masters, for instance, in the name and the design of the mentions, etc.). They intended to use this opportunity to improve the conditions of study (by decreasing the number of classes given in amphitheatres and increasing the number of classes in small groups, especially for the first-year students) and/or to normalise some rules (for instance, the organisation of examinations to occur concurrently for all training programmes, etc.). This strategy has not been very successful. On the one hand, they were not really able to go against the projects pushed by some academics: the latter developed arguments and refused to introduce changes; and this resistant strategy often succeeded. On the other hand, the presidential teams faced the reluctance of some faculty members to change their habits and ways of doing things. In one of the universities under study, for instance, we observed a strong discrepancy between the discourse of the presidential team, which claimed many changes and innovations and a strong control on the overall process, and the discourse of the interviewed academics who in many cases explained how they finally imposed their own views or simply refused to adopt the rules imposed by the presidential team.

There were some texts and directives from the Ministry, but we had to interpret them … We sometimes added some elements (about the examinations of the students, the president of a jury …) and sometimes colleagues told me ‘you are going further than the texts’. But as the texts did not say anything, we tried to go further. Some said ‘it is not in the texts’, for example the chart regulating the examinations and the curricula. But because there were many complaints and pleas about the exams in this institution, we tried to write down who is responsible for what in the organization of and during the exams (A member of the presidential team).

Let’s take the example of the university study schedules. The presidential team wanted them to be similar to the secondary schools schedules. But September is devoted to our own research projects. We could not accept such a change. They have to respect the specificity of each discipline … We did not accept everything. We refused the optional classes when they were not absolutely necessary (An academic).

Both cases led to the same consequences: an inflation in the number of existing degrees each university could deliver resulting from the creativity of academics but also from their opportunism. Many saw the LMD as a possibility for developing the training programme they dreamed of, or to extend their existing one-year³⁷ programme into a two-year one, or to be freed from pre-existing agreements.³⁸ They thus have been very reactive. Some figures estimate that the number of existing degrees doubled with the LMD. All did not open (for lack of students) and, in some cases, the presidential teams pushed for the opening of classes common to different programmes in order to avoid duplication. But, as a whole, despite the fact that the LMD was supposed to be implemented under a stable budget, many expected an increase in expenses as the few rationalisation processes which occurred could never cover the inflation in degrees or the improvements in pedagogy which were also attached to the reform (individual follow-ups of each student, foreign language courses, training in new technologies, etc.).

As a matter of fact, only a few presidential teams countered this intense academic reactivity. In our sample, from the beginning, one of the presidents imposed a limited number of masters and mentions and even chose the names of those who were to be responsible for the new programmes. But this is a rare case. Most of the time, the academic profession took control of the process and developed its projects in a quite isolated way, that is, without the university administration, without the students and mostly without the potential stakeholders (economic partners, local administrative or political elites, etc.).

As a result, by offering the academics an opportunity to take the upper hand again, these first years of the implementation of the LMD slowed down, rather than accelerated, the process of transformation of the French university system, which was intended to lead it from a profession-based pattern to a university-based configuration. From this point of view, too, the LMD affected, indirectly but concretely, the national institutional patterns.

This last assumption is common to the three categories of side effects analysed in this process. They all provide examples for the way in which the Bologna process, even if aiming at transforming the products and the production process within universities, had an impact on the national institutional settings. This is again a re-nationalisation effect as the affected issues are country specific (France specific in the case under study): the site effects, the problematic position and design of the UFRs and the increased role of the academic profession are all closely linked to characteristics of the French system.

Conclusions

The starting point of this chapter was to discuss whether and how the Bologna process has had an impact on the national institutional settings of the country where it was implemented, although this process does not explicitly aim at transforming the settings but at modifying the products and the production processes of higher education systems.

To address this question it was suggested to study the implementation of the bachelor–master scheme as a Europeanisation process and more specifically to focus on the re-nationalisation effects identified by the literature on Europeanisation. The case of France and the introduction of what is called the LMD in this country served as the main empirical sources. The Bologna process impacted on both the institutions constituting the French higher education system and the degrees, contents and structure of the curricula.

But which mechanisms can explain these larger effects of Bologna? Among the four re-nationalisation processes identified at the beginning of the chapter, three are especially useful in understanding the interplay between the implementation of the bachelor–master scheme and the transformation of the French institutional settings.

The first one focuses on the way in which the LMD has been adapted to some French characteristics and thus on how the national institutional settings played a role in limiting the automatic and systematic implementation of the LMD to all sectors, all programmes and all disciplines. This is the more classic form of re-nationalisation identified in Europeanisation studies and it is very close to phenomenon studied and outlined at length in comparative public policy analysis (e.g. Steinmo et al. 1992; Thelen 1999). Looking at how similar ideas (e.g. Hall 1989) or similar reforms (e.g. Pierson 1994 or Hacker 1998) had different effects in different countries, authors of this perspective highlight the role of past trajectories and of existing institutional settings to explain these variations.

But the ‘resistance’ of institutional settings to the Bologna process is only one aspect of the story. Despite their tendency for inertia they have also been affected. Two other re-nationalisation processes operated.

On the one hand, the French ministry used the Bologna reform to push domestic objectives that aimed to transform the national institutional settings. By contrast, with the re-nationalisation process identified by Palier et al. (2006), governmental actors did not take (or not only took) over control of Bologna but instrumentalised it to push further domestics issues.

On the other hand, the implementation of the LMD reform influenced the national institutional settings, not because they were aimed at the reform itself but because the introduction of the two-tier structure provoked side effects. These issues are closely linked to the existing specificities of the French system and were therefore addressed with ‘French’ solutions.

These two last mechanisms confirm some of the advantages that can be draw from soft-law processes such as the Bologna process, as identified by Abbott and Snidal (2000):³⁹ it does not deprive each country of its sovereignty (and even thus allows the addition of domestic issues to the global agreement) and it leaves way for learning mechanisms provoking side effects. If, as argued by the two authors, this kind of advantage can help us to understand why countries prefer soft-law agreement, it can be added that it also highlights why such agreement may have strong effects, although it does not explain why signing parties feel bound by it and implemented it.

Different conclusions can be drawn from the analysis and results developed in this chapter. The first one tackles the question of convergences in Europe and the capacity of Bologna to achieve them by the diffusion of the two-tier structure to all signing countries. In fact, the case of France and the different re-nationalisation processes involved in the implementation of the LMD lead us to wonder about the potential convergence effects of the Bologna process, unless the same aspects are addressed in the same way in each country, but there is poor evidence of that.⁴⁰ The various forms of re-nationalisation identified in this chapter in fact provoke an increased loose coupling between the institutional isomorphism supported by the Bologna reform and the actual national practices developed by the actors (Meyer and Rowan 1977). Even if the ritual signature of a new communiqué every two years regularly restates the commitment of each country to the implementation of the Bologna objectives, the local adaptations, national translations and side effects attached to each domestic implementation weaken the convergence potential of Bologna. These mechanisms certainly are not Bologna specific and could be extended to other European reforms to help explain why European reforms do not automatically imply more convergences.

At a more conceptual level, the French case under study in this chapter confirms the interest and relevance of looking at re-nationalisation processes not only in terms of direct and rather defensive reactions (form 1 and form 3 in Table 1), that is, at how actors or structures react or take control over a European process, but also in terms of indirect or more pro-active reactions (form 2 and form 4 in Table 1). In other words, there are more re-nationalisation mechanisms than identified up to now in the literature on Europeanisation and domestic side effects; the instrumentalisation of European processes in order to push domestic issues should be considered as well.

Finally, this case study also encourages the consideration of Europeanisation processes in a more interactive manner. In particular, it suggests that more attention should be paid to the national implementation of European reforms. In other words, it is clear from this case study that different re-nationalisation processes have an impact on the Bologna process within a single country. It also questions how such processes may affect the next steps of the Bologna process.

Notes

• 1. This system has been developed by the European Commission as a means to facilitate the mobility of European students. It is based on the convention that 60 credits measure the workload of a full-time student during one academic year. All the components of an educational programme provide a certain number of credits, which can be recognised by another European university.

• 2. In this perspective, one could argue and show that this ‘ideal institutional model’ tends to become more and more explicit. As stressed by Kehm and Teichler (2006), for instance, the Trends IV Report (Reichert and Tauch 2005) recommends giving more autonomy to universities. But the ENQA document (2005) mentioned above is also rather explicit about the model of the higher education system it supports. Thus, even if the declarations remain focused on the products and the production system, the documents and reflections accompanying the process are becoming more concerned with more ‘traditional’ institutional issues.

• 3. This research was funded by the ESEN (Ecole Supérieure de l’Education Nationale) and Sciences Po and consisted of an empirical study in three French universities and in the central ministerial services which implemented the reform. Based on the analysis of numerous documents and more than 100 interviews (see appendix), a detailed report has been written (Mignot-Gérard and Musselin 2006). A synthesis (in French) can be downloaded from: http://www.esen.education.fr/UserFiles/File/documentation/expertises/ensgt_sup/musselin_LMD.pdf.

• 4. In this chapter, the reasons why such objectives were pursued will not be tackled. I shall focus on what has been done and the consequences/impact this has had, not on why it has been led.

• 5. In a paper dealing with the specific dynamic of the Bologna process, the neo-institutionalist approach was preferred and Bologna described as a case for normative isomorphism (DiMaggio and Powell 1983). But in this chapter focusing on the implementation of Bologna, Europeanisation is a more relevant analytical framework.

• 6. Not only for a detailed and analytical description of the reasons why Allègre organised this meeting and why the three other ministers agreed to sign a declaration, but also for an analysis of how the two-cycle structure was included in the Sorbonne Declaration, see the research led by Ravinet (2005, 2007). Her cautious reconstruction of what happened tackles the usual story told about this first meeting but there is unfortunately not enough room here to go into it. As said in note 4, objectives and motives will not be discussed in this chapter.

• 7. In this chapter, the Bologna process will be mostly restricted to the implementation of the LMD. But, as argued in the introduction, the two-tier structure (LMD in France) is only one of the two transformations involved in the Bologna process. But it is the only one already really implemented (quality assurance and evaluation developed in many countries but without any common framework). This contribution will therefore focus on the LMD.

• 8. According to Cornelia Racké (2006), this confusion is not only French and can be explained by many factors, but it can also be explained by the fact that the European Commission is involved in this process and has used it as a vehicle to intervene more in a sector in which it has no official competencies.

• 9. The four others are: ‘changes in external territorial boundaries’, ‘development of institutions of governance at the European level’, ‘exporting forms of political organization and governance that are typical and distinct for Europe beyond the European Territory’ and ‘a political project aiming at a unified and politically stronger Europe’. For other reviews of the term, see Featherstone (2003).

• 10. For a very detailed and documented comparison of the degree of change introduced by the Bologna reform on four national higher education systems (France, Germany, Great Britain and the Netherlands), see Witte’s dissertation (2006).

• 11. These are the four possible impacts usually identified by the literature on Europeanisation (Radaelli 2001).

• 12. Radaelli (2001), for instance, distinguishes between vertical (top-down) and horizontal (market or mutual recognition-based mechanisms). Knill and Lehmkuhl (1999) distinguish three types of European policies, each corresponding to a specific mechanism: institutional change for policies of positive integration, change of domestic opportunity structures for policies of negative integration and cognitive evolution for policies of framing integration.

• 13. For a discussion of such mechanisms and an in-depth analysis of the Sorbonne, Bologna and Prague declarations, Ravinet’s research (2005, 2006) provides challenging results and stimulating discussions.

• 14. National also in the sense that they are observed in one country and not in the others affected by the same European process.

• 15. In an enlightening paper, Pierson (2000) distinguishes thus a narrow and a broad conception of path dependency. For him, the second conception ‘may entail the loose and not very helpful assertion that ‘history matters’, although it may also be presented with more rigor’. The narrower conception, which he discussed at length in his paper, by contrast focuses on situations where ‘preceding steps in a particular direction induce further movement in the same direction’ and is then ‘well captured by the idea of increasing return’ (Pierson 2000: 252).

• 16. Mergers of departments or labs in universities are typical examples of the activation of these defensive territories. The former structures are not defended because they were places of intense cooperation but because they draw a frontier protecting their members from others’ interventions.

• 17. See Renaut (1995: 166–179) to understand how, despite the 1875 act (recognising the freedom of the constitution to the French higher education and allowing the creation of non-public universities), the Republicans succeeded in voting the 1880 act which gave the monopoly of delivering national degrees to public universities.

• 18. Whether it be national or not.

• 19. Which was first spelled ‘mastaire’ before the Anglo Saxon spelling was finally adopted.

• 20. It can also be a diplôme national de master if the grande école accepts going through a specific accreditation process.

• 21. These processes are partly used by each sector. The ‘academisation’ of the grandes écoles is, for instance, favoured by the internationalisation and the normalisation of management studies all over the world. They can also be supported or favoured by successive French governments. For years, the gap between the grandes écoles and the universities regularly returns to the policy agenda as a problem to be solved (democratisation, access, curricula contents, equity, etc. being some of the multiple faces of this ‘problem’). The Bologna process has been one supplementary step in this long-term recurrent story.

• 22. Previously an agreement existed without a time limit.

• 23. French universities are organised in four groups, each one negotiating their four-year contracts one year after another. In other words, about one fourth of all French universities prepare a new contract each year.

• 24. This is the word used at the ministry.

• 25. This is a good illustration of the problem raised by the evaluation of public policy. As stressed by Majone and Wildavsky (1984) ‘implementation shapes policy’ and alters resources and objectives. Increasing institutional autonomy and reducing the number of degrees were both among the objectives of the ministry attached to Bologna. Rather than refusing to habilitate so many degrees (and thus reducing institutional autonomy), the ministry preferred to leave each institution to face its own choice and be responsible for managing the inflation of degrees while the budgets did not increase. This (conscious or not) strategy, built on the fact that universities will nevertheless be restricted by the budget at hand and on the conviction they will thus experience a learning process was probably more clever than any authoritarian rejection of degrees made by the ministry.

• 26. It is important to know that these experts are academics chosen by the ministry to assess the project.

• 27. For a comparison of the two processes, see Musselin (2006a).

• 28. By incomplete, I mean that they never welcome the four main families of disciplines (medicine, law, humanities and sciences) and either focus on only one of them (Grenoble 1 on Sciences) or present a limited spectrum of them (Lille 2, for instance, consists mainly of law and medicine).

• 29. For a good example, see the evaluation report of the CNE (Agency for the Evaluation of French Universities) on the ‘Aix-Marseille’ website at http://www.cne-evaluation.fr/WCNE_pdf/Aix-Marseille_Site.pdf.

• 30. For a presentation of Grenoble Universités, see http://www.grenoble-universites.fr/90633135/1/fiche_pagelibre/.

• 31. The ministry did not completely ‘follow’ this trend. It meets goals the ministry already had. In the ‘modernisation act’ which aimed at increasing the autonomy of universities but was withdrawn by Chirac in November 2003, some measures were already included which made possible the development of common services and even forms of mergers among universities.

• 32. Even if this capacity has always been more theoretical than real. The labs within the UFR have always been the concrete level for the management of research.

• 33. Those against distinctive structures for the licences and for the masters develop two counter-arguments. First, they fear it will lead to the constitution of undergraduate (teaching oriented) and graduate (research oriented) institutions and second, they argue that the masters being habilité for only four years will create potentially unstable structures (while the UFRs are stable).

• 34. In 2006, for instance, the yearly meeting of the deans of science dealt with this subject.

• 35. A new act was finally adopted in August 2007.

• 36. As shown in previous research (Musselin 2001), France had only facultés but no university (as a higher education institution). It was only in 1968 that today’s university had been recreated and confirmed by the 1984 act.

• 37. For instance, a professor in charge of a DEA (one-year programme) in a specific scientific speciality, became responsible for a mention of the master of science in this speciality … which is a two-year programme.

• 38. Some historians, who previously were associated with some social scientists in a DEA of history and social science, used the fact that the mention they jointly proposed with the social scientists had not been very well-assessed by the ministry experts to write their own project and to open a mention in history.

• 39. These authors identified four advantages of soft law over hard law: they are easier to reach, they better prevent the sovereignty of the signing countries, they allow to deal with uncertainty and for learning processes and finally they facilitate compromises.

• 40. At a distance, this seems to be the case, at least for some of the changes analysed in this chapter. For instance, the blurred boundaries between institutional sectors can be observed in other countries, as well as the organisational issues raised by the implementation of the bachelor–master scheme, etc. But a closer look reveals that the underlying rationales, the formulation of the problems and solutions agreed upon differ from one country to another. They are moreover never commonly discussed across countries.

Appendix

The study has been funded by the ESEN and Sciences Po. About 90 semi-structured interviews have been led between October 2004 and June 2005 in three universities and around 20 at the national level. Within the three higher education institutions, academic leaders and administrators have been interviewed. Then two faculties, one very involved in the LMD and the other less, have been chosen in order to lead further interviews. As a whole, 88 interviews were achieved.

Interviews led in the three universities

	Uni Pluri	Uni Multi	Uni Sciences
Faculty members	20	19	18
Academic leaders	6	4	5
Deans	3	2	3
Sub-total: Academics	29	25	26
Administrative staff in faculties	2	3	2
University administrative staff	1	0	0
Sub-total: Administrative staff	3	3	2
TOTAL	32	28	28

At the national level, 18 interviews were led within the ministry, the conference of university presidents, experts, etc. Many documents were also collected and exploited.

The research methodology relies on the approach developed by Crozier and Friedberg (Crozier 1964; Friedberg 1993, 1997; Crozier and Friedberg 1977, 1980).