Introduction

Since 2000, various initiatives aimed at stimulating European integration in higher education have attracted researchers’ attention to the changes of policy capacity at the European level (Corbett 2005; Neave and Maassen 2007), as well as the changes these European initiatives bring about at the national (macro), institutional (meso) or individual (micro) level (Dobbins and Knill 2009; Gornitzka 2006; Huisman 2009; Maassen and Musselin 2009; Välimaa et al. 2006; Witte 2006).

To avoid the trap of concept stretching and overestimating the importance of European initiatives, the first aim of this article is to clarify what “Europeanization of higher education” actually entails. Referring to the comparative politics and international relations literature (Börzel and Risse 2003; Radaelli 2003; Sedelmeier 2011) as well as recent higher education studies (Musselin 2009), the processes through which higher education entities (systems and organizations) adapt to particular environmental pressures coming from European higher education initiatives will be understood as Europeanization. The second aim is to present a more robust analytical framework for accounting for adaptation of higher education to European pressures for change, by building on two theoretical perspectives on Europeanization—an external incentives and a social learning perspective—the first based on rationalist and the second on sociological institutionalism (Börzel and Risse 2003; Sedelmeier 2011). Taking these two aims as its starting point the article addresses the following question: how do higher education systems and organizations respond to environmental pressures to change, in particular when these pressures are coming from the European level?

The explanatory potential of as well as challenges in using the two overarching theoretical perspectives on Europeanization are explored by analysing changes in quality assurance (QA) as an illustration. The analysis is based on a number of reports and studies of changes in QA, as well as official EU and the Bologna Process documents. It includes a discussion on what aspects of these patterns can be accounted for by the two Europeanization perspectives and concludes with an assessment of their explanatory potential, as well as a discussion of some analytical challenges and topics for further research.

Conceptualizing and operationalizing Europeanization of higher education

Based on Radaelli (2003) and Musselin (2009), Europeanization is here understood as: the institutionalization of formal and informal rules developed in a process that involves a supranational or an intergovernmental body, e.g. the EU, the Council of Europe—CoE or the Bologna Follow Up Group—BFUG. The process of institutionalization of said rules, i.e. adaptation to the European rules, can take place at the macro level (higher education system) and meso level (universities, QA agencies etc.). At the micro level, institutionalization of said rules is expected to bring about a change in individuals’ behaviour.

As the proponents of such a top-down understanding of Europeanization acknowledge, this definition disregards processes in which countries are uploading their policy preferences to the European level (Börzel and Risse 2003). In such cases there is no simple adaptation, but rather a mutual transformation of the European and macro level, or European and meso level. This may be of particular relevance for intergovernmental processes which, given the lack of a supranational structure, provide room for uploading. To what extent this is indeed the case in higher education depends (a) on the system and higher education institution in question and (b) on the issue at hand. Countries and higher education institutions that have been on the periphery of European initiatives (e.g. Balkans or Caucasus) had previously neither the opportunities nor (arguably) the capacity to upload policy preferences. Furthermore, countries which were pioneers in some issues may not have wanted or managed to upload all of their preferences to the European level, indicating that European HE initiatives may include issues that are novel for the majority of countries or institutions involved. Finally, it is important to stress that the pressure to adapt can also be in the eye of the beholder; the top-down approach is also justified in cases where the pressure is simply perceived as distinctly European and independent from domestic preferences (Trondal 2002).

Two main theoretical perspectives on how Europeanization unfolds can be identified (Börzel and Risse 2003; Sedelmeier 2011). The first—external incentives perspective—aligns with rationalist institutionalism. It focuses on changes in opportunity structures for different actors and expects that the institutionalization of rules will follow the logic of consequence. The second—social learning perspective—subscribes more to sociological-constructivist institutionalism and claims that actors need to be persuaded into following the European rules, in line with the logic of appropriateness.

What is common is that both see the process of Europeanization leading to partial convergence at the most. This is a consequence of: (1) the complexity of pressures, (2) the differences between entities in the nature and perception of these pressures, (3) the presence (or lack thereof) of mediating factors (Börzel and Risse 2003), as well as (4) the differences between different entities at the starting point of the process. Thus, Europeanization in some entities will lead to higher levels of Europeanization (more profound change); while in others it will lead to lower levels of Europeanization (superficial change). This leads to a continuum of outcomes: from deep change—transformation, through accommodation and absorption, to no change—inertia or even retrenchment (Börzel and Risse 2003). The two Europeanization perspectives are to some extent seeing the entities as relatively passive in this process, an issue that will be addressed later. The complexity of outcomes is further extended by the fact that European rules are open for different interpretations by different actors, implying that layering can be expected in the course of their institutionalization. This also leaves room for re-nationalization, the situation in which actors take the opportunity to promote their particular preferences under the European umbrella (Gornitzka 2006; Musselin 2009).

The following section provides an adjustment of Europeanization perspectives (Börzel and Risse 2003; Sedelmeier 2011) to fit the specificities of higher education in terms of ambiguity of purpose, professional and institutional autonomy, bottom-heaviness and loose-coupling (Clark 1983; Weick 1976).

Europeanization of higher education through external incentives

Within the external incentives perspective the overall hypothesis adapted for higher education states that a higher education entity will adopt European rules if, from its own point of view, the benefits of rewards exceed adoption costs. This implies that the rules should be clear and should be set as conditions for benefits: the larger the benefits are and the faster they can be expected to be realized, the more likely the institutionalization of rules will be. Benefits can be in the form of European funding, access to cooperation partners, or prestige connected with participation in networks (e.g. Jean Monnet chairs).

Furthermore, the promise of benefits or threats related to non-complianceFootnote 1 needs to be credible. Herein lies the ambiguity: consequences of non-compliance in higher education are rather vague and European structures have relatively weak competences in higher education (Neave and Maassen 2007). An additional question is how open for interpretation European rules are. For clearer rules, e.g. calculating student workload and ECTS, cases of non-compliance are more obvious, while for less clear rules, e.g. “promotion of a European dimension in higher education”, whether there is compliance or not is open for interpretation.

In order for a European structure to exert the adaptation pressure it must have sufficient power and information on the adaptation process to decide whether to award the benefits. Thus, the power asymmetry in favour of the structures developing European rules needs to be higher than the information asymmetry about adaptation in favour of the entity. As already indicated, the European structures have limited power (Neave and Maassen 2007). This, coupled with bottom-heavy and loosely-coupled higher education institutions (Clark 1983; Weick 1976), makes it rather difficult to monitor adaptation. Therefore, in higher education the information asymmetry in favour of the entity is higher than the power asymmetry in favour of the relevant European structure.

The external incentives perspective also foresees that the less veto players incur adoption costs, the more likely the adoption of European rules. A high density of veto players, actors who can block decisions and changes at various levels of governance (Tsebelis 2002), implies strong vested interests and therefore high adoption costs. Veto players can be small parties supporting minority governments, buffer bodies whose support is necessary for policy development, academic staff in institutional governing bodies or individual academics when it comes to changes within the classroom.

Furthermore, if there are strong institutional legacies or if the administrative capacity is not high, the adoption of European rules will be less likely. The strength of institutional legacies refers to the extent to which existing rules about particular aspects of higher education are deeply rooted and constrain actors interested in change. For example, legacies concerning core activities of higher education, such as education and research, may be strong due to bottom-heaviness and loose-coupling, and thus likely to lead to lower levels of Europeanization. The presence of administrative capacity may be crucial for adaptation, in particular to more technical and more complex European rules. For example, success in EU funding competitions demands significant administrative capacity dedicated to application, and subsequent success also requires capacity for project coordination and reporting on funds already allocated.

Europeanization of higher education through social learning

The social learning perspective puts focus on actors being socialised to accept the appropriateness of European rules, or to be induced “into the norms and rules of a given community” (Checkel 2005: 804). This is related to the importance of belief systems (Clark 1983) and the claim that normative compliance is the main form of compliance in higher education (Etzioni 1961).

In this perspective, the rules need to be seen as legitimate by relevant actors, either in terms of their content or in terms of the process through which they were developed. The legitimacy of content is connected to stronger resonance with existing entity rules. For example, if the new rules on how to develop study programmes are too different from the rules the entity currently employs, it will be difficult to socialise relevant actors into adopting them. Furthermore, if the entity representatives took part in developing these rules the resonance will be higher (see also earlier comments on uploading). The legitimacy of the process by which European rules are developed points to participation of relevant stakeholders. Participation of national ministry representatives (or representatives of universities or QA agencies for that matter), is an important factor for perception of legitimacy of process.Footnote 2 However, there may be differences in terms of legitimacy of ministry positions on particular issues in higher education, especially if they are seen to use the European processes for furthering particular domestic agendas (Musselin 2009). Legitimacy of process may also be dependent on participation of European stakeholder organisations, such as the European University Association (EUA), European Association of Institutions of Higher Education (EURASHE), Education International (EI), European Association for Quality Assurance in Higher Education (ENQA) or European Students’ Union (ESU) and, naturally, the extent to which they are seen as legitimate representatives of their respective constituencies on the European level. For example, if a European organisation does not have members from a particular country (e.g. ESU does not have members in Greece), this may lead to a situation in which European rules are not seen as legitimate by relevant actors and are thus openly rejected.

A further element of the social learning perspective, connected to the aforementioned legitimacy and resonance, is identity. The more the entity that is expected to adopt European rules can identify itself with the European structure that established the rule, the more likely the adoption is. Therefore, it becomes relevant whether the entity actually participated in the European structures developing these rules and as a consequence can identify itself and its preferences with the European structures and preferences. Identification can also be achieved through participation of key actors in European (or related transnational) epistemic communities or issue networks (Haas 1992). Such communities or networks may be formed around joint projects, cooperation programmes, thematic networks in higher education and European stakeholder organisations, or they can be organized by the European Commission itself (Gornitzka and Sverdrup 2008). There the actors can learn about the proposed rules, gradually become persuaded in their appropriateness, and become promoters of these rules within the entity. Once the rules are not seen entirely as external their adoption at different levels within the entity becomes easier.

Finally, Europeanization is facilitated by increasing internationalization of the sector, in terms of participation of foreign experts in the decision-making processes or reference to foreign models. Higher education has had an international outlook through various forms of cooperation long before the European policy developments. This was in particular the case on the meso level (higher education institutions—HEI), through student mobility and joint research projects. Thus, higher education has been a relatively internationalized sector prior to the European initiatives.

Summary of the two perspectives

The two perspectives are summarized in Table 1. Though the analytical distinction between the two perspectives is potentially useful for testing the explanatory potential of each of the two strands of institutionalism (hence the distinction in the table), they are often seen as complementary (Börzel and Risse 2003) and will be treated as such within this article.

Table 1 Europeanization of higher education—factors and outcomes
Full size table

As indicated earlier, both perspectives on Europeanization lead to an expectation of partial convergence in terms of outcomes of the process. This is related to differences between countries when it comes to presence of mediating factors listed above, but also to differences between various higher education aspects: e.g. in some aspects there may be more veto players than in others. Furthermore, the two Europeanization perspectives implicitly treat all factors as equally important which may not reflect the empirical situation. With an aim to explore some of these challenges as well as illustrating the overall explanatory potential of the two perspectives, the next section will focus on changes in quality assurance (QA).

Focus on quality assurance

Methodological remarks

The choice of QA is motivated by it being one of the central issues in European discussions and a priority topic of the Bologna Process. Various European projects, development of the European Standards and Guidelines for Quality Assurance in Higher Education (ESG), establishment of the European Quality Assurance Register (EQAR), and use of ESG for external review of quality assurance agencies all indicate that the process of change has gone beyond mere rhetoric. In addition, while there were European developments in the area of QA before (see below), the Bologna Process and the adoption of the ESG provided new impetus to the topic, increased the number of systems and organisations involved and introduced a number of new “issues” (e.g. student participation and external evaluation of QA agencies).

The illustration starts with a historical overview of key European developments in the area of QA in the last 15 years, based on an analysis of EU documents, the Bologna Declaration, the subsequent communiqués, the actual text of the ESG, as well as secondary literature on European QA developments prior to and during the Bologna Process. This will be followed by a presentation of what we know so far about adaptation of national QA approaches and the role European developments played in these processes, based on a number of reports and studies ranging from the late 1990s and early 2000s until the early 2010s, including but not limited to reports from different European projects, data collected by Eurydice, the so-called Bologna Independent Assessment (Westerheijden et al. 2010) etc. The final part of this illustration brings together the “what we know so far” and the Europeanization perspectives. It discusses how the two Europeanization perspectives help in accounting for “what we know”. This is also used to identify possible topics for further research in the concluding section.

A brief history of QA developments

In the 1980s QA came into the focus as part and parcel of the new steering mechanisms in Western Europe (in particular in Denmark, the Netherlands and the UK) and, in the early 1990s in response to massification, curricula reform and mushrooming of private HEI in Central and Eastern Europe (CEE, in particular Poland) (Schwarz and Westerheijden 2004). Several EU Pilot Projects on QA tied quality to the issue of mobility (ENQA 2010) and were instrumental in promoting QA in countries where it was non-existent (Westerheijden et al. 1994). In 1998 the European Council published a “Recommendation on European cooperation in quality assurance in higher education” (EU Council 1998), which stated that QA approaches (where available) had already some common points, e.g. an independent QA agency, a fitness-for-purpose principle and a self-evaluation component. The recommendation also suggested an establishment of a European network focusing on QA and in 2000 ENQA was indeed established with the Commission support. In addition, several European QA agencies were already cooperating in what will later be labelled as the Joint Quality Initiative (JQI)(ENQA 2000).

The existence of similarities between QA approaches indicates that some policy borrowing and learning took place even prior to the emergence of an explicitly European QA initiative. Linked with the key elements of the QA approaches promoted through the Bologna Process (see below) and in particular to the ESG, it also indicates uploading of some of the policy preferences from the pioneering countries to the European level, though, as will be demonstrated below, the European QA initiatives also included some aspects that were rather novel.

Promotion of European cooperation in QA was one of the original Bologna action lines, further elaborated by the Prague Communiqué calling upon ENQA, national QA agencies and HEIs to “collaborate in establishing a common framework of reference” (Prague Communiqué 2001: 2). 2 years later, ministers gathered in Berlin gave the mandate to the so-called E4 groupFootnote 3 to develop “an agreed set of standards, procedures and guidelines for quality assurance” and to “explore ways of ensuring an adequate peer review system for quality assurance and/or accreditation agencies or bodies” (Berlin Communiqué 2003: 3).

The ESG were adopted in Bergen in 2005 and consisted of a total of 23 standards (and related guidelines) organised in three parts: (1) internal and (2) external QA of HEIs, and (3) external QA of QA agencies. The common points identified earlier constituted the foundation of the ESG (ENQA 2010), which were envisaged to promote “joint understanding of how processes of quality assurance should be conducted, while allowing for national and institutional particularities and essential characteristics” (Stensaker et al. 2010: 577). The ESG also foresaw a European register of QA agencies, for which the model was adopted in London in 2007. The European Quality Assurance Register Association (EQAR) was founded in March 2008. Interestingly, the idea of EQAR was not a novel one, given that a similar initiative (Worldwide Quality Label) failed in the early 2000s due to resistance of university rectors (Jeliazkova and Westerheijden 2002) as well as the fact that it was not developed by the BFUG but was rather promoted by a limited number of QA agencies gathered around the JQI.

What we know about changes in QA

As the 1998 EU Council Recommendation states, the national approaches to QA in the late 1990s already had a number of common features (EU Council 1998), in particular between pioneering countries which learned and borrowed from each other (Westerheijden et al. 1994). However, the impression of similarities of QA approaches may have been superficial: while there was convergence of regulation and broad features of the system, there were significant differences in the basic nature of the evaluation exercise (summative or formative), use of peers (selection, experience, training, tasks) and study visits organisation and duration (Billing 2004).

Furthermore, a number of aspects was lacking, such as use of “truly international reviewers, [and] application of explicitly international standards and criteria” (Schwarz and Westerheijden 2004: 18, author’s emphasis), and participation of students in different QA procedures (Pataraia 2010). According to the Eurydice report for 2004/05 (EURYDICE 2005, i.e. before ESG), in almost ¾ of the 41 countries covered students were not represented in national QA agencies. When it comes to external evaluation of institutions, approximately 1/6 included neither foreign experts nor students and 1/6 included foreign experts but not students, with the issue of student participation being problematic in particular in internal evaluation of institutions and due to differences in interpretation what it actually entails. The ESG is rather explicit about both of these issues; standards and guidelines for selection, competence and nature of participation of students and foreign experts allow for closer monitoring of Europeanization in this aspect of the QA approaches.

Thus, given the status of the ESG, since 2007 various Bologna reports have explicitly linked the stocktaking in the area of QA to compliance with the ESG, although the standards and procedures for evaluation of this compliance have been changed and challenged over the years. The BFUG itself admitted that the criteria were not sufficiently robust in 2007, and defined stricter criteria for the 2009 Stocktaking exercise (Rauhvargers et al. 2009), the 2012 report (EACEA 2012), and the Bologna Independent Assessment (Westerheijden et al. 2010). The latter claims that while there seemed to be substantial compliance on the surface (47 out of 48 countries had internal and external QA), this translated neither into evaluations of HEIs being done in accordance with the ESG at least once thus far (this was by that time done in 25 systems) nor into having a regular internal QA (by that time done in 18 systems). The assessment further indicated that “in quality assurance … national histories and national agendas are strong drivers of the actual changes made”(Westerheijden et al. 2010: 37). It also stressed that more developed QA approaches can adapt more easily to the European initiatives (ibid.), indicating that (1) uploading and subsequent negotiations about policy preferences on the European level, (2) learning facilitated by the existence of European initiatives and (3) strengthening of a European dimension and identity through participation in these initiatives may be supportive of Europeanization.

Given the strengthening of Stocktaking criteria, the 2012 Bologna implementation report (substituting the usual Stoctaking report in 2012, EACEA 2012) demonstrated less than complete compliance with the Bologna Process in general, and the ESG in particular. This is especially the case for student and international experts’ participation in QA processes (ibid.), an issue essentially introduced explicitly and in more detail by the ESG. However, the 2012 report also indicated that external QA (part 2 of the ESG) was no longer a resisted issue, which was linked to (1) accordance between European and national QA policies, (2) resonance of that part of the ESG with the priorities of main national actors, as well as (3) the fit between the European initiatives and national legal setting and cultural traditions (Hopbach 2012). As far as part three of ESG (external evaluation of QA agencies) was concerned, there has been mixed progress: while 28 agencies from 13 countries have been registered in EQAR, only in 15 systems is it possible (by law) for HEIs to be evaluated by foreign agencies (EACEA 2012).

In essence, the ESG is used both as a reference point and as a “box-ticking” compliance tool, due to the linkage between compliance to the ESG and the EQAR and ENQA membership criteria (ENQA 2011). The immaturity of agencies or restrictive national legislation seem to lead to lack of compliance (Hopbach 2012). The former was also addressed in an EUA study on internal QA (part 1 of the ESG), which stated that the internal QA “as it is understood nowadays, is a recent phenomenon” (Loukkola 2012: 306).

However, the focus on external QA of HEI and external QA of agencies (parts 2 and 3 of the ESG) may be driving attention away from part 1 (internal QA of HEI). This could be reducing compliance to this aspect of the ESG to writing self-evaluation reports, an issue that has indeed been picked up by the Stocktaking exercise in 2009 (Rauhvargers et al. 2009). This led to calls for further work on awareness and ownership of the ESG at the grass root level (ENQA 2011), indicating weak identification of shop-floor actors with the European approach to QA.

As previously indicated, the issue of student participation in QA is particularly interesting, given that it was explicitly introduced by the ESG. As indicated by several reports and studies (e.g. EACEA 2012; Pataraia 2010), student participation seems to be a rather difficult part of ESG to comply to, due to a lack of prior traditions of student involvement in general and ceremonial compliance in terms of formally allowing students in but simultaneously limiting their role (Pataraia 2010).

Many of these reports and studies do not pay particular attention to the how and why behind these changes. Those that do, do it in a somewhat fragmented way. They highlight the ambiguity of the European QA initiatives in general, the related inconsistencies in interpretation of the ESG (ENQA 2011), as well as the fact that some parts of the ESG are less ambiguous than others (Witte 2009). They also point to the different paths of adaptation: adapting the label but not the practice, adapting the practice but not the label and adapting both (Stensaker et al. 2010). However, a number of interlinked factors critical for the success of European QA initiatives have been suggested: national agenda in line with the Bologna agenda, actors ready to push the process despite possible opposition and common understanding of the issues in the higher education and policy community (Westerheijden et al. 2010), i.e. some of the elements of the two Europeanization perspectives.

Enter Europeanization

Europeanization of QA through external incentives?

Within the external incentives perspective, the first factor relates to the clarity of rules (Table 1). That lack of complete clarity provides domestic actors room for interpretation and therefore also re-nationalization was in particular demonstrated by the rather limited adaptation to the ESG standards related to student participation in QA. Regarding consequences of compliance in terms of funding and status, it was shown that part 3 of the ESG has already affected the QA agencies, given membership criteria of EQAR and ENQA. In addition, though previously QA agencies were mostly (if at all) evaluated by national authorities, this is more and more done by European structures (Stensaker et al. 2010). Moreover, lack of compliance with part 1 has less visible consequences than lack of compliance with part 2, let alone part 3, and this difference did indeed lead to differences in compliance with different aspects of ESG, reflecting also the importance of information asymmetry.

Furthermore, as suggested in previous studies, for successful adaptation of QA to European initiatives actors who can potentially act as veto players need to be brought on board. The story of EQAR is particularly illustrative: a similar initiative failed earlier due to resistance of the university rectors, but was accepted later and seems to have had a relatively high impact on the national QA agencies during its short existence. The importance of administrative capacity as a factor of Europeanization was also picked up by the studies presented in the previous section: younger agencies do seem to have problems with compliance with the ESG.

However, the empirical material with regard to the institutional legacy factor provides a somewhat contradictory picture. On the one hand, more developed QA agencies seem to adapt more easily than less developed ones, potentially refuting the idea behind the institutional legacy factor. On the other hand, the fact that the idea of HEIs being evaluated by foreign QA agencies is acceptable in less than a third of Bologna countries or that there are difficulties with student participation in QA in countries where student participation has not been an acceptable practice before are in line with the institutional legacy factor. One could argue that the latter two instances are more a case in favour of the administrative capacity factor than against the institutional legacy factor, but there seems to be a potential conflict between these two factors, an issue addressed in the conclusion.

Europeanization of QA through social learning?

The first factor of importance for the social learning perspective is legitimacy (Table 1). This is most evident in the case of the EQAR story: while the first initiative developed and pushed by a limited number of QA agencies did not take off, the later one, supported by the entire BFUG (including its consultative members) did. As far as the resonance between European initiatives and national agendas is concerned, previous studies clearly indicate that resonance is important: where there is resonance there is more adaptation (Hopbach 2012), and where resonance is lacking so is adaptation (Pataraia 2010, on the issue of student participation).

The challenges with regards to adaptation to Part 1 of the ESG and the related call for raising awareness and ownership of the ESG amongst the shop-floor level actors (see above) suggests that where identity is lacking Europeanization is stalled. In addition, a number of projects and networks (formal and informal) were important for adaptation of national QA approaches (ENQA 2010), also pointing to the importance of epistemic communities for the Europeanization process. The learning that these epistemic communities may be conducive of happens on the European level, but may also support learning within the national context: those already socialised in European QA initiatives can work with those who need more persuasion, in particular if they are in the position of veto players.

Finally, the fact that more concrete European QA initiatives (such as the ESG and later the EQAR) came to a sector already marked with high internationalisation (given international cooperation in this area during the late 1980s and early 1990s) provides support for the internationalisation factor indicated in Table 1. In sum, the social learning perspective on Europeanization seem to fit when it comes to accounting for different patterns of adaptation of national QA systems so far identified, even more so than the external incentives perspective.

Conclusion

As has been illustrated with changes in QA approaches, the demonstrated fit between Europeanization perspectives and already identified patterns of adaptation of QA indicates that the perspectives do provide a more robust theoretical backbone for studies of higher education change in response to European pressures. Though there are studies that have referred to some of the factors used by the two Europeanization perspectives, the approach presented here provides a systematized and richer explanatory toolbox. The robustness and applicability of Europeanization perspectives should be further assessed through studies on other aspects of higher education that are in the focus of European initiatives, e.g. curricular reform.

In addition, while social learning takes place in epistemic communities and legitimacy and resonance seem to be important factors of Europeanization, research focusing on change at the micro (individual) level is necessary to fully assess the usability of the social learning perspective on Europeanization. This is particularly important given the nature of European level initiatives with respect to higher education (lack of strong legal competence), and due to the nature of higher education, including loose-coupling, bottom-heaviness and importance of belief systems.

However, several analytical challenges need to be addressed in order to refine the two perspectives so they can be even more useful for building causal mechanisms able to explain the complexities of change in higher education.

The first relates to an understanding implicit in the two perspectives that entities adapting to European pressures are relatively passive. Neither higher education systems nor higher education organizations (e.g. universities, QA agencies) can be considered as completely passive and change processes are more complex than simple environmental determinism (Gornitzka et al. 2007). Therefore, it would be useful to allow strategic response to external pressures, following Oliver (1991) and her conceptualizations of defiance (cf. retrenchment), avoidance (cf. inertia), manipulation (cf. absorption), compromise (cf. accommodation) and acquiescence (cf. transformation).

Secondly, in order to address a potential inconsistency in the external incentives perspective vis-à-vis institutional legacy and administrative capacity factors, a better operationalization is necessary. Essentially: does a strong administrative capacity need strong institutions (and therefore strong institutional legacies) to develop? A “yes” turns the issue into an empirical one and potentially leads to dropping one of the factors or changing the direction of its influence. A “no” takes the discussion back to how institutions emerge in the first place, a topic on which the institutional approach provides only partial answers (Peters 1999), and a topic beyond the scope of this article.

An additional challenge is the implicit understanding that individual factors are all equally important. For example, is participation in epistemic communities necessary for identification with the European rules, i.e. are epistemic communities more important than other factors in the causal mechanism? To respond to this challenge, a more detailed empirical basis, focusing in particular on the presence or absence of each individual factor, would be necessary.

The next challenge is the lack of explicit reference to time. While the external incentives perspective relies on more immediate factors and therefore has a shorter time horizon, the social learning perspective does highlight the process aspect, and therefore the temporal dimension more. However, how the empirical material is cut in terms of time periods analysed may have an impact on the assessment of the outcomes and mechanisms of Europeanization. If cutting too narrow, lower levels of Europeanization may be identified and an external incentives perspective may seem more relevant, and vice versa.

The final challenge is related to situations in which some convergence exists prior to European initiatives, also bringing forward a problem of disentangling (horizontal) policy transfer processes (Dolowitz and Marsh 1996) from (vertical) Europeanization processes. Furthermore, European initiatives may be a result of uploading policy preferences from the national or organizational to the European level, which decreases the clear top-down character of the adaptation process, essentially pushing it outside of the concept of Europeanization accepted in this article. Again, the answer lies in a more robust empirical base, for example, in terms of a longitudinal national case study that would trace the process of changes of the national QA system with sufficient time-frame and detail to separate uploading, policy transfer and Europeanization. Furthermore, comparative studies of (a) adaptation to European initiatives where uploading can be identified and issues where this is not the case (e.g. student participation) in (b) a group of countries some of which are successful in uploading and some of which are not, would provide an opportunity to assess more clearly what is the role of uploading in the adaptation process and what are the implications of this for the conceptualization and operationalization of Europeanization.

These challenges notwithstanding, or perhaps precisely for being aware of these challenges, the two Europeanization perspectives presented can contribute to strengthening the theoretical backbone of many studies on Europeanization of higher education. In addition, the increase in clarity of the concept of Europeanization demonstrated in this article is important, particularly at times of frenzy, both on the policy and research side, about “all things European”.