The legitimation of funding decisions in higher education: the role of policy framing

Transcription

1 The legitimation of funding decisions in higher education: the role of policy framing Jelle Mampaey, Martina Vukasovic Centre for Higher Education Governance Ghent (CHEGG) Department of Sociology, Ghent University Corresponding author: Martina Vukasovic, IPSA, July, Poznan, Poland WORK IN PROGRESS. PLEASE DO NOT CITE WITHOUT PERMISSION! Abstract Our study analyses how different policy actors engage in policy frames in a debate about a contentious policy issue. In particular we explore three elements of policy frames cognitive, normative and causal and how these frames vary across the different actors and across time. The empirical focus is on Flemish debates about increasing tuition fees that unfolded in By analysing media sources reporting on this issue (communicative discourse), we find that the diagnosis of the policy issue at hand is not really disputed (cognitive element), while the effects of the decision are (causal element). Those arguing against the proposal overall use more normative elements in their framing, but it should be noted that the normative framing in general increases as the debate progresses. Keywords: policy analysis, policy frames, legitimation, higher education, funding, Flanders 1

2 Introduction Analysis of policy processes has been one of the hallmarks of higher education research (Tight, 2012). The bulk of this research is primarily concerned with policy implementation and the extent to which reform processes bring about change in higher education (see e.g. Cerych & Sabatier, 1986; Gornitzka, Kogan, & Amaral, 2005; Kohoutek, 2013). Comparatively, the early stages of a policy process, in particular agenda-setting, are somewhat less studied and focus either on how more encompassing changes concerning steering approaches and understandings of the role of higher education are legitimated (e.g. Clegg, Hudson, & Steel, 2003; Gidley, Hampson, Wheeler, & Bereded-Samuel, 2010; Pick, 2006; see also Verger, 2012 for an example concerning pre-higher education) or how EU institutions legitimize their ideas concerning higher education as well as the push for expanding their competences in the area of higher education (e.g. Capano & Piattoni, 2011; Cerna & Chou, 2013; Keeling, 2006; Walkenhorst, 2008). In both cases, the main focus is on one specific actor (government or the EU institutions) and its legitimating strategies. This is problematic because it effectively neglects the role of other actors in the policy arena, i.e. stakeholders, such as universities, student and staff unions, business associations, which increasingly play a role in higher education policy making, on both the national and the European level (Amaral & Magalhães, 2002; Gornitzka & Maassen, 2000; Klemenčič, 2012). These actors seek to influence policy-making by putting forward specific issues on the policy agenda and advocating for specific approaches to these policy issues. For instance, a specific issue could be the (perceived) underfunding of higher education and the (perceived) solution could be higher tuition fees. In doing so, they get involved in policy framing selecting and emphasizing a specific aspect of a policy issue that would benefit their constituents. Given the diversity of stakeholders, they are likely not to see entirely eyeto-eye on all policy issues. Thus, an approach that focuses only on one actor effectively misses another important feature of the process contestation over policy issues and different ways in which the stakeholders argue for their view of a particular issue. With this in mind, stakeholder legitimation of different approaches to one policy issue is the key focus of this study. We focus on recent decisions to increase tuition fees for higher education in Flanders. Funding of higher education is particularly suitable for exploring legitimation of policy decisions and policy framing, given that it is one of the key instruments policy-makers have at their disposal for shaping the outcomes in a particular policy sector, including higher education (Hood & Margetts, 2007). In particular, decisions concerning whether students and/or their parents need to pay for higher 2

3 education can have a profound impact on both institutional and student behaviour (Orr, Wespel, & Usher, 2014), as well as higher education teaching and learning more generally (Naidoo & Jamieson, 2007). Thus, funding decisions are likely to induce significant controversy. In particular, we explore what are the elements comprising policy frames and how do they change over time, as well as map the positions of different actors involved, in particular juxtaposing those in favour and those against a decision (in this case increasing the tuition fee). Conceptually, the study relies on interest groups and public policy literature focusing on the role of stakeholder organizations, policy frames and framing approaches in policy formation processes. Empirically, we analyse major Flemish media entries published during the 2014 debates concerning increasing tuition fees in Flanders. We start off with outlining our theoretical approach, followed by brief information on the Flemish policy context, in particular with regards to actors involved in higher education policy processes. Then we describe the data used and the methodological approach, followed by presentation of results. In the final section, we reflect on our findings, the methodological and theoretical approach and suggest some points for further studies. Theoretical considerations On actors and their interests We build our analysis on the understanding that a policy is a public statement of an objective and the kind of instruments that will be used to achieve it (Gornitzka, 1999, p. 14). This implies that both state and non-state actors can have policies, the key difference here being the steering capacity. Namely, state actors (e.g. parliament, government) can make decisions that will steer other actors behaviour by way of, for example, regulation or funding allocation, while non-state actors are limited to advocating for their specific policies and influencing (or trying to) the state actors. In democratic societies interests of various actors are expected to be taken into account when making policies (March & Olsen, 1989; Peters, 2015), in particular when these policies are expected to have an impact on those actors (the reason why often such actors are referred to as stakeholders ). The specific mode in which this is done depends on the overall approach to interest intermediation (Eising 2004). Thus, a distinction can be made between (1) statism, in which the state is the dominant actor 3

4 and stakeholder are seldom, if at all, consulted; (2) pluralism in which multiple non-state actors are involved and (3) (neo-)corporatism in which only a privileged handful of non-state actors have access to state decision-makers. It should not be assumed a priori that the various actors involved (state and non-state) would necessarily see eye-to-eye with regards to the identified problems in a specific policy sector, objectives one should strive for, the ways in which these diverse objectives can be achieved or even what is the outcome of implementation of a specific set of policy decisions. Thus, the policy process can be conceived as a process in which different ideas about what should be done and which values should be pursued are developed and interact, and often come to conflict with each other (Capano, 2009, p. 18). Essentially, this approach highlights ideas and interests as key inter-linked aspects of a policy, and focuses the analysis on policy rhetoric and discourse (see e.g. Howlett & Ramesh, 1995; Zittoun, 2009). While the notion that ideas and interests are relevant is true for all stages of a policy process, we will in this study focus in particular on agenda-setting and policy formation. Here, it should be stressed that we do not subscribe fully to the so-called stage heuristic i.e. the idea that policy processes unfold in a linear manner in which different stages (agenda-setting, formation, implementation, evaluation) neatly follow one another and do not overlap. 1 We primarily use the labels agendasetting and policy formation to identify which aspects we focus on: (1) the process in which various ideas about problems to be addressed and solutions to be proposed are brought up by different actors to be put on the agenda and (2) the ensuing interaction between these ideas and interests, which results in a policy-decision by the relevant authority (e.g. parliament or government). On ideas and policy framing The point of departure of our approach is that policy actors perceive the world in general and specific problems that should be addressed by a (government) policy through different ideational prisms (Fischer, 2003; Howlett & Ramesh, 1995). These ideational prisms mean that actors highlight particular aspect of an issue in line with this ideational prism (see e.g. Daviter, 2007 for the case of the EU policies). These prisms are often labelled policy frames, used for both legitimation of a 1 For a discussion on strengths and weakness of the stage-heuristic see Sabatier (2007). 4

5 specific viewpoint and for persuasion of others into the applicability and relevance of such a viewpoint. Essentially, policy frames are rhetorical weapons of advocacy and consensus (Weiss, 1989, p. 117). Different weapons are used by different actors in the policy arena and a distinction can also be made between how policy actors discuss the policy issues amongst themselves coordinative discourse, and how they present these issues to the public communicative discourse (Schmidt, 2008), with the former usually being more technocratic and pragmatic than the latter. Both policy problems and policy solutions can be framed, i.e. viewed from a specific ideational prism. For example, whether one sees increasing social inequality as a problem depends on one s position towards inequality as a negative or a positive feature. The same applies for e.g. progressive taxation and other forms of wealth-redistribution as a desirable solution. Fischer (2003) goes even further and claims that apart from framing problems and solutions, evidence can also be framed, in the sense that what is chosen to be presented as worthy evidence also depends on actors norms, values and beliefs. Thus, given the distinction between normative and cognitive ideas (Schmidt, 2008), one can also distinguish between normative and cognitive elements of policy frames. Moreover, the coupling between problems and solutions proposed by the different actors is also an indication of policy framing of said actor. This is in particular the case in situations of high ambiguity and complexity in which actors cannot always demonstrate an objective / non-contentious causal mechanisms linking problems and solutions (Cohen, March, & Olsen, 1972). Moreover, it may be misleading to expect that such a causal mechanism exists. In essence, actors may have their preferred solutions and their pet-problems that they will try to push for whenever an opportunity opens up (Kingdon, 2003), and thus a causal mechanism they may identify as the linkage between problem and solution may be subject to policy framing as well. Our approach In all, this serves to highlight that policy frames can be understood as comprising the following elements: - cognitive element definition of a situation: this is what the issue is, this is the diagnosis of the situation ; - normative element value: this is why this issue is important, this is what should (not) be done ; and 5

6 - causal element mechanism linking the identified problem and the proposed solution: we need to do act in a particular manner (solution), because we expect that acting in such a manner will alleviate the problem. Not all elements need to be equally explicit. Some actors may choose to advocate for their position by highlighting primarily the definition of the situation and the value, and not necessarily suggest a solution to alleviate said problem. Others may focus primarily on the cognitive and causal element, without making their normative basis manifest, etc. Moreover, actors do not only develop and present their own policy frames (what can be termed as proactive policy framing), they can also engage with policy frames of other actors (reactive policy framing). They can support others policy frames, for example when forming advocacy coalitions (Sabatier & Weible, 2007) in order to make a specific policy problem or a specific policy solution more visible. Actors can also choose to attack policy frames of their opponents, arguing that the definition of a policy problem relies on inappropriate evidence, that the problem is actually not so significant or that the proposed instrument will not alleviate the identified policy problem (Boscarino, 2015). In this process, apart from an individual frame employed by one actor to influence decisionmakers, a dominant understanding of an issue can also emerge, reflecting the ideas of the most powerful actors (Baumgartner & Mahoney, 2008). The distinction between different elements, types, and levels of policy frames allows a more nuanced analysis of interactions between different actors during the agenda-setting and policy formation processes. For example, a dispute over the causal element of a frame only is indicative of a primarily instrumental difference between the different actors, a dispute over the normative element indicates an ideological difference, while a dispute over the cognitive element allows points to the politicization of information and evidence. The relative extent to which a specific actor engages in proactive or reactive policy framing is, in our view, indicative of their power within the policy arena more powerful actors engage primarily in proactive framing, while less powerful actors engage more in reactive framing and are also likely to provide framing that goes against the dominant understanding of an issue in the policy arena. 6

7 The Flemish policy making context In general, public policy making in Flanders is characterized by a corporatist approach to interest intermediation, in which a few actors have privileged access to policy-makers (Beyers, 2002). When it comes to higher education, since the federalisation in 1988, higher education is solely under the jurisdiction of the constituent regions Wallonia and Flanders (Verhoeven, 2008), making them effectively separate higher education systems. The policy development process often includes a public debate, both in the parliament and in the media, though in some cases key decisions may also be made by only a handful of actors, e.g. relevant ministry, current and former leadership of major higher education institutions (Huisman & Mampaey, forthcoming). The actors who are likely to engage in a public debate thus include, first and foremost, state actors prime minister and minister responsible for higher education and his/her deputies, as well as state agencies dealing with higher education. In particular when government members speak (or are referred to) in the public, both their portfolio as well as their party affiliation are referred to, e.g. for the current minister for higher education Hilde Crevits the reference to CD&V (Christian-Democrats) would be indicated as well. 2 Political opposition, primarily SPA (socialists) and Groen (Greens) is likely to get involved as well. On top of this, various stakeholder associations may get involved, namely: the Flemish University Council (Vlaamse Interuniversitaire Raad, VLIR), the Flemish Council of University Colleges (Vlaamse Hogeschoolenraad, VLHORA), the Flemish Union of Students (Vlaamse Vereniging van Studenten, VVS), business associations (e.g. VBO Federation of Enterprises in Belgium), trade unions (e.g. ACOD Onderwijs, branch of the trade union of the public sector focusing on education), etc. Given that in many cases these associations have other organisations as members, it may happen that both the association as a whole and individual members take part in the debate. To the best of our knowledge, there has not been a systematic study of these actors and their positions in policy processes so it is somewhat difficult to state ex ante who are the more powerful actors, apart from the obvious: the political parties forming the government. Arguably, given the dominant position of universities in the higher education system, one would also expect VLIR to be a more powerful actor than VLHORA. 2 The current Flemish government is a coalition between N-VA (New Flemish Alliance, centre-right Flemish nationalists), CD&V and Open Vld (Open Flemish Liberals and Democrats (conservative liberal). 7

8 Data and method Our methodology is a case study (Yin, 2009) of the (outcomes of the) debate on tuition fees in Flemish higher education, which is a suitable research design to explore political processes in real-life settings. We selected Flemish mass media as our main unit of analysis. In specific, we collected longitudinal and qualitative data, in particular 45 media texts from March 2014 (the moment when the university college sector raised higher tuition fees as a solution to solve the perceived problem of underfunding of Flemish higher education) till November 2014 (the moment that the decision was made to increase tuition fees from approximately 600 Euro to approximately 900 Euro). These texts were literally all texts we found that were including arguments concerning the perceived problem of the underfunding of Flemish higher education and/or the proposed solutions, collected in six relatively large Flemish newspapers (De Morgen, De Standaard, Het Nieuwsblad, Het Belang van Limburg, Het Laatste Nieuws, Gazet van Antwerpen). The actors making the argument ranged from actors in the university (college) sector (e.g. university rectors, university college representatives, professors, authorities in the field of education science, student unions), political actors (e.g. members of the right-wing Flemish government and the leftwing opposition) and business associations. Additionally, we collected other data sources including public documents and existing research reports and papers. The documents were used to gather more background information about the institutional context in which the debate took place (see above the section about the Flemish context). Multiple data sources were used to triangulate (Denzin, 1978). It should be noted that not all arguments were always verbatim quotes from the actors. In some media texts the argument was described by the journalist. It should also be noted that most of the media texts appeared in De Morgen, hence a comparison between journals (probably with a different political ideology) was impossible. These are obviously limitations of our study, but we still believe that our research significantly improves our understanding of the process of policy framing in higher education. We balanced between an inductive and a deductive approach to data analysis (Langley, 1999), iteratively moving back and forth between data and theory (Eisenhardt, 1989). In specific, our data analysis consisted of seven steps. First, we inductively coded 58 unique arguments in the 45 media texts. In case an argument was repeated by the same actor in the same media text, it was not coded as a unique argument. Otherwise, an argument was coded as a unique argument, hence also when e.g. the same argument is repeated by the same actor in different media texts or the same argument is 8

9 repeated in the same media text by a different actor. 3 Second, we inductively coded the type of actor that made the argument. Third, we inductively coded whether an argument was in favour of higher tuition fees, against higher tuition fees, or neither in favour nor against higher tuition fees (i.e. neutral). Fourth, we inductively grouped the arguments in three periods in that we found media silence in between: early debate (March 2014), mid-debate (late July early August 2014) and final debate (mid-september 2014). Fifth, we deductively coded explicit policy frames utilized in the 58 unique arguments. Sixth, we inductively identified an extra category or type of policy frame, namely whether an argument was new (i.e. mentioned for the first time) or repeated. Obviously, each argument may consist of multiple policy frame elements. To maximize reliability (Denzin, 1978), one of the authors carried out the analysis independently and then the result was discussed with the other author, as well as with other colleagues in the research group of both authors. Discussion took place till consensus was reached. Table 1 illustrates our coding with four examples in the early debate. Analogously, Table 2 is an illustration of the mid-debate and Table 3 contains four examples of the final debate. Table 1 Early debate Actor Argument 1: Higher education is underfunded and increasing tuition enhances funding of HE University college rectors Argument 2: Increasing tuition increases barriers to participate in higher education University rector UA Argument 3: The government should enhance funding of higher education University rector KUL Argument 4: Tuition is an investment of students in their futures Chairman of VLHORA In favour, against or neutral In favour Against Against In favour Period Early debate Early debate Early debate Early debate Cognitive frame X / / / Normative frame / / X / Causal frame X X X X Proactive frame X / / / New argument frame X X X X It should be noted that the former situation (same actor, same argument, different media source) appears in the data set only once. 9

10 Table 2 Mid-debate Actor Argument 1: The government cuts public funding of higher education and subsequently tuition will probably increase with 400 to 600 Euro, which is unacceptable University rectors Argument 2: HE is underfunded and firing staff is no solution to compensate for it would decrease quality University rector KUL Argument 3: Increasing tuition from 619,90 Euro to approximately 1000 Euro is acceptable Flemish government Argument 4: Tuition should not be increased for students from lower socioeconomic backgrounds University rector UH In favour, against or neutral Against Neutral In favour Against Period Mid-debate Mid-debate Mid-debate Mid-debate Cognitive frame X X X / Normative frame X / X X Causal frame X X / / Proactive frame / / / / New argument frame / X X X Table 3 Late debate Argument 1: Increasing tuition from 619,90 Euro to 1100 Euro is unacceptable in that it will harm the future of students and society Actor Student union VVS UG Argument 2: Increasing tuition is unacceptable unless compensation is provided for students from lower socioeconomic background Argument 3: Increasing tuition will not lead to social inequality in that presently, social inequality is already a reality Authority in the area of education science Argument 4: Increasing tuition is unacceptable unless compensation is provided for students from lower socioeconomic background University rector UG In favour, against or neutral Against Neutral In favour Neutral Period Final debate Final debate Final debate Final debate Cognitive frame X / / / Normative frame X X / X Causal frame X X X X Proactive frame / / / / New argument frame / / X / 10

11 Results Evolution of the debate as a whole The frame elements and frame characteristics, organised according to the stage of the debate, are presented in Table 4. The predominance of elements or types of frames, as well as usage of new arguments are indicated in relation to the total number of arguments given in that stage of the debate (e.g. 27 out of 42 arguments in mid-debate included a normative frame element). Table 4 Overview of frame elements and characteristics in different stages of the debate Cognitive Normative Causal Pro-active Re-active New argument Early debate 3/12 2/12 12/12 5/12 7/12 11/12 Mid-debate 20/42 27/42 36/42 0/42 42/42 24/42 Late debate 1/4 3/4 4/4 0/4 4/4 1/4 Overall 24/58 32/58 52/58 5/58 53/58 36/58 It can be argued that the findings related to the presence of new arguments, as well as the relative prominence of pro-active or re-active frames are as expected. As the debate proceeds, the framing becomes, by default, more re-active and fewer new arguments are presented. Moreover, what we identified as the late debate took place only a few days before the final decision was made to increase tuition fees and consisted of only four arguments, i.e. the bulk of the framing activity by actors took place in the early and the mid-debate. Concerning frame elements, it is interesting to observe that the cognitive elements of frames are actually not so dominant and are mostly used during the peak of the process, the mid-debate. This means that in many cases, in particular in the early and in the late debate, most of the actors did not provide an explicit diagnosis of the situation. Out of the total of 24 arguments which included a cognitive element, 20 arguments included the claim that higher education is under-funded and it was the first diagnosis put forward in our data set. Other cognitive elements used included: a claim related to SEB background of students in one of the Flemish universities, a claim that the fees are higher in 11

12 other systems, a claim concerning relative proportion of funding coming from the students (5%) and the tax payers (95%) and, related to this, a claim that higher education is a public good. This main claim of under-funding seems to be taken for granted and, even if mid-debate almost half of the arguments also use a cognitive frame element, there is no instance in which an actor actively disputed this (or any other) cognitive element of a frame. Contrary to this, there is a clear pattern of increasing prominence of normative frames, suggesting that as the debate proceeds it also becomes more ideologically laden. However, as will be presented later (see below), this is to a large extent driven by those arguing against the proposal to increase tuition fees. In the early debate, many arguments either did not have or had a rather implicit normative element. The only two arguments that included a normative frame concerned the desirability of increasing funding from the government. The normative elements in mid-debate were primarily concerned with the acceptability of increasing tuition fees, and if so for which groups of students, as well as where should additional income come from. As such, they were also mostly linked to the causal elements of frames. Related to this, almost everyone involved also used causal elements when framing their positions throughout the debate. This suggested that the effects of the decision to be taken in this case the increase of tuition fees were not taken for granted. Those arguing in favour claimed positive effects on quality and efficiency of the higher education system Tuition fees should be seen as an investment. People who have a diploma [of higher education] will earn more later [in their professional careers]. (VLHORA chairman, 17 March 2014, own translation) while those arguing against claimed negative effects on equity Increasing tuition can lead to the situation that [vulnerable] youth is not going to higher education anymore because they do not have the money. (Rector of KU Leuven, 25 July 2014, own translation). In response to this, some of in favour arguments started to dispute these claims 12

13 Barely anyone will stop going to higher education [after the introduction of higher tuition fees] in that higher education is already undemocratic. (individual professor, authority in the field of educational science, 19 September 2014, own translation). Analysis of actors With regards to mapping of actors in relation to the frames they were employing, those arguing in favour of higher tuition fees were: the Flemish government (including NVA, the largest right-wing party of Flanders), a local student union with strong network ties with the government (the chairman of the university student union is the son of an influential member of NVA), individual academic staff members, and a former KUL rector. Obviously, university colleges (VLHORA) were clearly in favour of higher tuition as they were the initiators if this idea in the first place. Actors arguing against higher tuition fees were: left-wing political parties that were not part of the Flemish government (SPA and Groen), the national association of student unions (VVS), rectors of most universities (KU Leuven, University of Hasselt, University of Antwerp), and the university council (VLIR). Some neutral actors could be identified that were neither consistently arguing in favour nor against higher tuition fees, including: business associations, a dean from the KU Leuven, and the Ghent University rector. The business associations entered the debate only when other actors (primarily the rector of Antwerp) put forward the idea that the necessary funding for higher education should be provided by the business sector, justifying this claim by stating that business sector pays rather low taxes while at the same time benefiting from highly educated employees. As could be expected, business associations and their members argued against this, but they did not argue in favour or against higher tuition fees as such. When it comes to the Ghent University rector, her position shifted during the debate: in the early debate she argued against the increase in the tuition fee [ ] the Ghent University rector believes that the government has to prioritize education. There are many good reasons to choose for additional investments in higher education. Politicians should dare to make that choice [ ] (17 March 2014, own translation) while in the final debate she argued (partly) in favour, after the envisaged increase was settled around approx. 900 Euros (instead of the initial suggestions that went as high as 2,500 Euros). 13

14 Increasing tuition fees is possible, but something should be said about social corrections: What does that mean? If this is only related to scholarships, I doubt if this is a sufficient social correction. Higher tuition fees also have implications for people not receiving scholarships. We believe that everyone should get the opportunity of access to highquality education. (Rector of Ghent University, 19 September 2014, own translation) The analysis of actors clearly indicated the politically loaded nature of the debate. Actors associated with the right-wing Flemish government clearly argued in favour of higher tuition fees, whereas actors associated with the left-wing opposition argued against. Hence, the most powerful political actors argued in favour of higher tuition fees. Contrary to this, arguably the more powerful actors within the higher education sector (in particular university rectors and the university council) mainly argued against higher tuition fees, whereas the less powerful actors (in particular university colleges and some academic staff members) mainly argued in favour. While the position of the universities and their rectors may not be necessarily surprising and could be interpreted as their attempt to maintain the status quo, they were not completely consistent and united in this position. Students were also not presenting a consistently united front, indicating that in general actors may not be as clear cut, independent and homogeneous as initially thought of. Moreover, the two groups of actors also differed with regards to frames they were employing. Figure 1 indicates the relative prominence of different frame elements in relation to all arguments given by a specific group of actors. The vertical axes indicate the proportion of arguments including a specific frame element or type, e.g. 50% of arguments employed by the in favour camp throughout the debate as a whole included a cognitive element. 14

15 Arguments in favor of tuition Arguments against tuition Cognitive Normative Causal Proactive Figure 1 Frame characteristics of the in favour and against arguments for the debate as a whole As can be expected, the in-favour actors engaged to some extent in pro-active framing, while all of the arguments put forward by the against actors were essentially reactions. Throughout the debate, the arguments against increasing tuition fee were more frequently employing normative frame elements than cognitive ones, and were also more normative overall than the in favour arguments. Evolution of arguments in favour or against In a similar way, we also investigated how frame elements of these two groups related to each other and how they changed during the debate (Figures 2-4), revealing starker distinctions between the two groups already early on in the debate (Figure 2). 15

16 Arguments in favor of tuition Arguments against tuition Cognitive Normative Causal Proactive Figure 2 Frame characteristics of the in favour and against arguments in the early debate The in favour actors presented their arguments employing solely the cognitive and causal elements, suggesting a rather technocratic approach to the issue. For example: The [ ] university colleges request an additional amount of 150 million Euro to facilitate research and development, and also more money for infrastructure and staff because this reduces work load and performance-related pressure (University college sector, 17 March 2014, own translation) Contrary to this, the actors arguing against tuition fees used normative elements of frames, sometimes linked with cognitive and rather often linked with the causal elements. However, this clearly technocratic nature of the arguments of the in favour actors is no longer there in the mid-debate. Both groups of actors use more normative elements in their arguments, with the in favour actors employing normative elements in half of their arguments while three quarters of arguments employed by the against camp had a normative element in them. Clearly, by this time, the debate took a markedly ideological turn. 16

17 Arguments in favor of tuition Arguments against tuition Cognitive Normative Causal Proactive Figure 3 Frame characteristics of the in favour and against arguments in mid-debate The final stage of the debate (Figure 4) was rather short and included four arguments only, with only one clearly in favour and only one clearly against (hence the 100% in both cases in the figure). 17

18 Arguments in favor of tuition Arguments against tuition Cognitive Normative Causal Proactive Figure 4 Frame characteristics of the in favour and against arguments in the late debate Here it is interesting that in these final days, the in favour actor actually uses a new argument, primarily relying on causal frame elements Many believe that increasing tuition to 1000 Euro or more goes against the democratization of higher education, but is this right? No, barely anyone will stop going to higher education in that higher education is already undemocratic. (individual professor, authority in the field of educational science, 19 September 2014, own translation) while the against actor repeats previous arguments, but fortifies the position using all three frame elements: Do not mess with our futures! It is also yours! [ ] Tuition will probably increase to 1100 Euro. [ ] This is more than a bridge too fare for VVS and the student unions. The proposals are concrete enough for the Flemish students to decide that it is time to resist. (VVS representative, 18 September 2014, own translation) 18

19 Figures 5-7 present the changes in the relative prominence of the different frame elements in the in favour and against groups Early debate Mid-debate Final debate In favour Against Figure Early debate Mid-debate Final debate In favour Against Figure 6 19

20 Early debate Mid-debate Final debate In favour Against Figure 7 Conclusions In this paper, we wanted to explore the phenomenon of policy framing in the context of higher education policy change and we were particularly interested in the role of different groups of actors. Building on the broader political science literature, we assumed that policy framing is one of the major aspects of policy change in higher education. Overall, our findings indicate that some policy frame elements are more prominent than others and these elements vary over time and between different groups of actors, suggesting that policy framing is a highly dynamic process characterized by contestation. In specific, a major finding was that the cognitive elements of frames are rather exceptional. This suggests that higher education policy framing is often based on taken for granted, not necessarily empirically grounded, assumptions about policy issues. Both the proponents and opponents of the proposed solution to the problem (in our case higher tuition fees) did not disagree about the underlying problem (the underfunding of higher education), although none of the actors provided empirical evidence for their assumptions. 20

21 We also found that normative elements of frames occur more frequently and these elements increase over time. Based on this finding, we suggest that higher education policy framing is mainly embedded within conflicting ideologies, where proponents and opponents of proposed solutions draw from different ideologies (in our case neoliberal versus egalitarian ideologies to either defend or oppose higher tuition fees). Moreover, the ideological nature of the policy debates are likely to increase over time, as the actors make continuous efforts to push for their specific policy preferences. Our analysis showed that most of the arguments were composed of causal elements. This implies that, contrary to the problem, the link between the problem, the proposed solution and its effects is not taken for granted, but rather highly contested. While these were some general tendencies, we also found that elements of frames vary between different groups of actors. Overall, we identified two camps of actors, a group of proponents of the proposed solution and a group of opponents. Most strikingly, throughout the debate, the opponents were more frequently employing normative frame elements than cognitive ones, and were also more normative overall than the proponents. Also, the proponents over time used less and less cognitive elements of frames, while the use of normative elements reached its peak mid-debate. The opponents increased their use of cognitive elements, most markedly so towards the end of the debate, and consistently increase their use of normative frame elements. This suggests that in policy debates proponents and opponents of a proposed solution (to a perceived policy problem) use different elements of frames and the differences get more prominent over time. Such political and ideological loading of the debate was also reflected in the fact that the proponents were part of the Flemish government or representing one of the higher education sectors (university colleges), whereas the opponents included both the political opposition and most of the other higher education sector (universities). In other words, policy debates and policy framing in higher education may expose conflicts between political as well as academic interests. Our study shed some light on the early stages of policy development, an aspect that has thus far been rather under-researched by the higher education research community. Moreover, it also showed that apart from the usual suspect the government other actors also engage in policy framing in order to push for their own preferences. The decision to distinguish between different elements of policy frames cognitive, normative and causal enabled us to analyse actors framing in a nuanced way, revealing differences between the actors involved. Moreover, the longitudinal approach we employed 21

22 also allowed us to map how actors framing changes over time, thus departing from the usual static (snapshot) view of policy framing. Given that our study focused primarily on communication towards the public what Schmidt (2008) refers to as communicative discourse, one avenue for further research would be to compare the communicative and coordinative (between policy actors) discourses employing the distinction between cognitive, normative and causal frame elements. This would allow the analysis to go beyond the relatively simple expectation that the coordinative discourse is likely to be more technocratic and pragmatic, and to analyse the ingredients of these discourses in a more detailed manner. Another promising direction is to employ the conceptual tool for studying what we refer to as framing of framing, i.e. to compare the official policy positions of the policy actors with how these policy positions are presented in the media. Using the conceptualization of the three ingredients of policy frames, such a study would be able to reveal not only the extent to which the media themselves engage in framing but also how they do this e.g. which frame elements are more likely to be enhanced or disputed as well as what is the relationship between the media framing and the ideological positioning of the media itself. Given the increasing use of media (including social media) by the policy actors themselves, it is essential to expose such processes and their outcomes in order to gain a deeper understanding of the intricacies of policy process, in particular its early stages. References Amaral, A., & Magalhães, A. (2002). The emergent role of external stakeholders in European higher education governance. In A. Amaral, G. A. Jones, & B. Karseth (Eds.), Governing Higher Education: National Perspectives on Institutional Governance (pp. 1-21). Dordrecht: Kluwer Academic Publishers. Baumgartner, F. R., & Mahoney, C. (2008). The Two Faces of Framing: Individual-Level Framing and Collective Issue Definition in the European Union. European Union Politics, 9(3), Beyers, J. (2002). Gaining and seeking access: The European adaptation of domestic interest associations. European Journal of Political Research, 41(5), doi: / Boscarino, J. E. (2015). Setting the Record Straight: Frame Contestation as an Advocacy Tactic. Policy Studies Journal, advance online publication. doi: /psj Capano, G. (2009). Understanding Policy Change as an Epistemological and Theoretical Problem. Journal of Comparative Policy Analysis: Research and Practice, 11(1), doi: /

23 Capano, G., & Piattoni, S. (2011). From Bologna to Lisbon: the political uses of the Lisbon 'script' in European higher education policy. Journal of European Public Policy, 18(4), doi: / Cerna, L., & Chou, M.-H. (2013). The regional dimension in the global competition for talent: Lessons from framing the European Scientific Visa and Blue Card. Journal of European Public Policy, doi: / Cerych, L., & Sabatier, P. A. (1986). Great expectations and mixed performance: the implementation of higher education reforms in Europe. Stoke-on-Trent: Trentham Books. Clegg, S., Hudson, A., & Steel, J. (2003). The Emperor's New Clothes: Globalisation and e-learning in Higher Education. British Journal of Sociology of Education, 24(1), Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A Garbage Can Model of Organizational Choice. Administrative Science Quarterly, 17(1), doi: / Daviter, F. (2007). POLICY FRAMING IN THE EUROPEAN UNION. Journal of European Public Policy, 14(4), doi: / Denzin, N. K. (1978). The research act : a theoretical introduction to sociological methods: New York (N.Y.) : McGraw- Hill. Eisenhardt, K. M. (1989). Building Theories from Case Study Research. The Academy of Management Review, 14(4), doi: / Fischer, F. (2003). Reframing public policy : discursive politics and deliberative practices: Oxford : Oxford university press. Gidley, J. M., Hampson, G. P., Wheeler, L., & Bereded-Samuel, E. (2010). From Access to Success: An Integrated Approach to Quality Higher Education Informed by Social Inclusion Theory and Practice. Higher Education Policy, 23(1), Gornitzka, Å. (1999). Governmental policies and organisational change in higher education. Higher education, 38(1), Gornitzka, Å., Kogan, M., & Amaral, A. (Eds.). (2005). Reform and Change in Higher Education. Analysing Policy Implementation. Dordrecht: Springer. Gornitzka, Å., & Maassen, P. (2000). Hybrid steering approaches with respect to European higher education. Higher Education Policy, 13(3), Hood, C. C., & Margetts, H. Z. (2007). The tools of government in the digital age. Basingstoke: Palgrave Macmillan. Howlett, M., & Ramesh, M. (1995). Studying public policy: Policy cycles and policy subsystems. Toronto etc.: Oxford University Press. Huisman, J., & Mampaey, J. (forthcoming). Setting up associations in Flemish higher education. In H. de Boer, J. File, J. Huisman, M. Seeber, M. Vukasovic, & D. Westerheijden (Eds.), Policy analysis of structural reforms in higher education: Process and outcomes. Basingstoke: Palgrave Macmillan. Keeling, R. (2006). The Bologna Process and the Lisbon Research Agenda: the European Commission s expanding role in higher education discourse. European Journal of Education, 41(2), doi: /j x Kingdon, J. W. (2003). Agendas, alternatives, and public policies. New York: Longman. 23

24 Klemenčič, M. (2012). The Changing Conceptions of Student Participation in HE Governance in the EHEA. In A. Curaj, P. Scott, L. Vlasceanu, & L. Wilson (Eds.), European Higher Education at the Crossroads: Between the Bologna Process and National Reforms (pp ). Dordrecht: Springer Netherlands. Kohoutek, J. (2013). Three Decades of Implementation Research in Higher Education: Limitations and Prospects of Theory Development. Higher Education Quarterly, 67(1), doi: /j x Langley, A. (1999). Strategies for Theorizing from Process Data. Academy of Management Review, 24(4), doi: /amr March, J. G., & Olsen, J. P. (1989). Rediscovering institutions: the organizational basis of politics. New York: Free Press. Naidoo, R., & Jamieson, I. (2007). Empowering participants or corroding learning? Towards a research agenda on the impact of student consumerism in higher education. Journal of Education Policy, 20(3), Orr, D., Wespel, J., & Usher, A. (2014). Do changes in cost-sharing have an impact on the behaviour of students and higher education institutions? Evidence from nine case studies. Retrieved from Brussels: Peters, B. G. (2015). Advanced introduction to public policy: Cheltenham : Elgar. Pick, D. (2006). The Re-Framing of Australian Higher Education. Higher Education Quarterly, 60(3), Sabatier, P. A. (2007). The Need for Better Theories. In P. Sabatier (Ed.), Theories of the Policy Process (pp. 3-17). Boulder: Westview Press. Sabatier, P. A., & Weible, C. (2007). The Advocacy Coalition Framework: Innovations and Clarifications. In P. Sabatier (Ed.), Theories of the Policy Process (pp ). Boulder: Westview Press. Schmidt, V. A. (2008). Discursive Institutionalism: The Explanatory Power of Ideas and Discourse. Annual Review of Political Science, 11(1), doi:doi: /annurev.polisci Tight, M. (2012). Researching higher education (second ed.). Maidenhead: Open University Press McGraw Hill Education. Verger, A. (2012). Framing and selling global education policy: The promotion of public-private partnerships for education in low-income contexts. Journal of Education Policy, 27(1), Verhoeven, J. C. (2008). Questioning the Binary Divide: Non-University Higher Education in Flanders (Belgium). In J. S. Taylor, J. B. Ferreira, M. d. L. Machado, & R. Santiago (Eds.), Non-University Higher Education in Europe (Vol. 23, pp ): Springer Netherlands. Walkenhorst, H. (2008). Explaining change in EU education policy. Journal of European Public Policy, 15(4), doi: / Weiss, J. A. (1989). The powers of problem definition: The case of government paperwork. Policy Sciences, 22(2), doi: /bf Yin, R. K. (2009). Case study research: design and methods. Los Angeles: Sage Publications. Zittoun, P. (2009). Understanding Policy Change as a Discursive Problem. Journal of Comparative Policy Analysis: Research and Practice, 11(1), doi: /