Introduction

In an earlier contribution in this journal Bleiklie and Michelsen (2013) have convincingly pointed to the need of a more in-depth analysis of higher education policy making. Assuming that countries have put in place policies with similar objectives, they indicate that the differences in policy content can be traced to differences in national political organization. This article contributes to their debate, but takes a somewhat different approach to the issue by focusing on the role of actors, specifically political parties, and introducing a two-dimensional analytical framework to capture their preferences in relation to higher education. In doing so, it questions their assumption of converging policy objectives and puts a stronger emphasis on diverging partisan positions and their impact on higher education policy.

Change processes connected to higher education have been an important part of the research agenda for some time now and multiple conceptual lenses have been used to study them (Gornitzka et al. 2005; Vukasovic et al. 2012). However, the party political aspects of such reforms have been rather neglected so far. This is particularly surprising since higher education policy has gained more political saliency in the last years. First, massification has led to a growing amount of public resources spent in higher education, and second the discussions around the knowledge economy as well as processes of European integration and coordination have amplified the importance of higher education (Maassen et al. 2012; Maassen and Stensaker 2011). This growth in political saliency has led to a situation where higher education is treated as less special, yet more important and is expected to deliver solutions in other policy areas (Gornitzka and Maassen 2011). The growing saliency has been accompanied by the emergence of a plurality of belief systems linked to higher education, and a renegotiation of the contract between higher education and society (Gornitzka et al. 2007). This renegotiation takes place to a large extent within the political arena and between different political parties with their respective platforms and visions concerning higher education (Busemeyer et al. 2013, 533ff). As higher education is not a public good accessed by all citizens in all countries, it also has the potential to be used as a tool for re-distribution in the context of welfare state policies (Ansell 2010). Taken together, this leads to a growing politicization of higher education policy and the incorporation of higher education in debates addressing for example the relationship between the state and the public sector.

There is a rich body of literature in political science dedicated to the influence of the ideological background of a party on its policy positions in different fields and the resulting variations in policy outputs (see for example Hibbs 1977). Even though the number of existing studies on education and especially higher education policy is much more limited (for an overview see: Busemeyer and Trampusch 2011), it is surprising that the conceptual approaches have so far not been used more often in higher education policy research. It is beyond the scope of this article to present this literature in detail. However, the aim is to fill the existing void by taking analytical tools used in political science and to provide a framework for their use in higher education research.

The existing studies addressing partisan dynamics in higher education policy only focus on re-distributive conflicts linked to higher education. This article goes beyond the current empirical studies and presents a more complex analytical approach. Therefore, it will use their findings as well as research on the changing relation between the state and the public sector as a basis to argue for an expansion of the analytical frameworks used so far. This expansion introduces a second dimension of political conflict, which focuses on the question whether the control of the higher education sector is centralized or de-centralized. This new two-dimensional framework will be illustrated by presenting both “ideal” higher education systems as well as political parties that can be expected to favor such a system. In doing so, the article also proposes a research agenda outlining further steps towards a better understanding of the relation between political parties and higher education policy change. The questions addressed in this article are: How can political parties’ preferences in higher education policy be conceptualized? What could possibly constrain partisan competition in higher education policy?

The following section of the article will address the main existing empirical studies and synthesize their results and the political conflicts represented through them along one dimension that focuses on the re-distributive characteristics of higher education. The next section uses findings from studies of the relationship between the state and the public sector to argue for a second analytical dimension addressing the political conflicts in relation to the control of the higher education sector. Afterwards, the two-dimensional analytical framework and hypotheses for ideal party positions in relation to it will be presented. The subsequent section will focus on possible path dependencies and constrains on party competition in higher education policy. Finally, in the conclusion the findings will be summarized, open problems discussed and avenues for further research highlighted.

What do we know so far about partisan politics in higher education?

Higher education studies have witnessed in the last years a growing focus on investigating policy shifts and change processes in higher education as well as the links to its political environment (see for example Bleiklie and Michelsen 2013). However, the role of political parties in these processes has only been addressed in a very limited number of studies (for an example on US higher education see McLendon et al. 2009). At the same time, education policy has traditionally been somewhat of a blind spot in political science studies on parties and their policy profiles, though, recent years have seen a resurgence of interest in this area, which to a limited extent also included higher education policy (Busemeyer and Trampusch 2011). However, the existing literature on partisan influence on higher education policy focuses almost exclusively on party competition on a re-distributive or socio-economic dimension.Footnote 1

Socio-economic conflicts are essentially about whether a party favors economic redistribution or not. Most existing studies on partisan effects in higher education policy capture the level of re-distribution through higher education policy by linking the ideological composition of the government to changes in public spending (e.g. Boix 1997; Busemeyer 2007; McLendon et al. 2009; Schmidt 2007). One part of these studies finds strong empirical links between the participation of left partiesFootnote 2 in government and increased public spending on (higher) education (Boix 1997; Castles and Obinger 2007). They conclude from this that by spending more public money and thus expanding publicly subsidized higher education, parties of the left can increase participation of their own electorate in higher education and offer them social upward mobility, thus using higher education as a tool for re-distribution.

Contrary to these findings, some studies on party effects in higher education policy come to contradicting conclusions. Their empirical results show that parties of the political right, once they are in government, lead to increased public higher education spending (Rauh et al. 2011). Their offered explanation is that, contrary to other instruments of the welfare state, access to higher education is skewed towards the wealthy part of the population, therefore the socio-economic background of students is related to their likelihood of attending university (Lucas 2001; Raftery and Hout 1993). This means that up to a certain rate of participation in higher education, additional public spending is mainly to the benefit of the wealthier electorate of political right parties. Thus, increased public higher education spending, through low private investment for attending higher education, can be seen as a tool of reverse re-distribution. This effect makes it more likely for parties of the political right to focus on higher education, as it shields their electorate and preserves their socially advantageous position (Ansell 2010).

Based on the structure of the higher education system the approaches to higher education policy, using it as a re-distributive or a reverse re-distributive tool, will manifest in diverse ways. Ansell (2010, 166f) finds a conditional effect of partisanship and enrolment levels in higher education with right-wing parties favoring public spending in higher education systems with less than 33 % of gross enrollment rates and left-wing parties favoring it in higher education systems with a gross enrollment rate over 50 %. Therefore, if a left party faces a higher education system that ensures a high level of enrollment, it is more likely to expand public funding for higher education, than a left party facing a low level.

One final factor determining the level of re-distribution in higher education policy is the amount of private higher education spending captured for example in tuition fees. Preference for such a form of shifting the costs for higher education to the users also depends on the level of enrollment. Left parties might favor private higher education spending as a way to finance expansion of higher education in a phase when participation levels are low (Ansell 2010). As access to higher education widens, the preference for further expansion of tuition fees shifts from left to right parties, as the latter use it to deter participation of a greater part of the population and protect their core electorate’s labor market advantage (Wolf and Zohlnhöfer 2009).

It has been shown that the existing studies on partisan conflicts in higher education policy address only one conflict dimension, the socio-economic or re-distributive one. It spans between two ideal points. On the one end, a restrictive higher education system is characterized by high personal costs for attending higher education, a limited access to higher education and often rather low public costs. On the other end, an expansive higher education system is characterized by low personal costs for attending higher education, a wide access to higher education and often rather high public costs. Parties of the left that favor re-distribution tend to be more on the expansive side of the continuum, while parties of the right that try to avoid re-distribution tend to be more on the restrictive side.

Who controls what? A rising conflict in higher education policy

The previous section presented the results of existing empirical studies on the partisan aspects of higher education policy and showed that these studies focus exclusively on a re-distributive conflict dimension. However, the recent literature on party competition very often conceptualizes the conflicts between parties to be organized along two dimensions: a re-distributive dimension and a dimension capturing conflicts about questions of control between actors with different sets of values and differing approaches to authority, power, autonomy and steering (Bornschier 2010; Kriesi 1998). In essence this resonates with the classic works that structure the political space along cleavage lines (Lipset and Rokkan 1967; Rokkan 2009). In the area of education policy, Ansell and Lindvall (2013) have shown that the political conflicts around the construction of primary education systems not only focus on the distributive questions but also on the amount of direct control the state holds vis-à-vis the educational institutions. They find that different political parties prefer different levels of centralized control and that this political conflict significantly shapes the structure of the educational sector. If political parties influence the centralization of control in primary education, which is much more locally governed than higher education, it can also be expected that they influence the way in which control over higher education is exercised. Furthermore, recent studies in higher education research show a growing interest in questions linked to different forms of steering of the higher education sector as well as the level of autonomy of universities (for example Enders et al. 2013; Jungblut and Vukasovic 2013). As differences in political preferences of parties can be expected also in relation to the level of centralization of control over the higher education sector, it is necessary add a second analytical dimension to the one presented earlier.

The relationship between the state and the public sector, including higher education, was rather stable in the decades after the Second World War. However, since the early 1980s new forms of organization have emerged and the relationship between the state and the public sector has become more volatile and open to changes (Gingrich 2011; Olsen 1988). One clearly identifiable set of changes is linked to new public management reforms and the move of governments from classical state structures to more managerial approaches (Pollitt 2001). These changes include the increasing relevance of markets as steering tools as well as the implementation of instruments such as increased accountability measures, output orientation and performance indicators, which also affected higher education (Christensen 2011; Paradeise et al. 2009). While these reforms show a strong discursive convergence, there is considerable divergence when it comes to the actually implemented changes. This can be explained both by the influence of the existing institutional context as well as decisions by actors (Pollitt 2001). Due to the adaptability of new public management reforms, governments can shape them in various ways to fit to their local context as well as ideological preferences. Thus, shifts in the composition of the government can cause shifts in the objectives of reforms (Green-Pedersen 2002; Hood 1996; Pollitt et al. 2007). Furthermore, governments composed of different parties can use similar reform strategies for differing reasons and with diverging outcomes (Gingrich 2011). Opening up the state–public sector relationship thus has led to conflicts between the state, the producers and the consumers of public goods. These were related to the question, who holds the effective control over a certain public sector. This in turn has encouraged political parties to take diverging positions on this issue (Gingrich 2011).

An early adaptation of the idea of conflicts around steering and governance to higher education is Burton Clark’s triangle of coordination, with which he conceptualizes the dynamics between the state, the market and the academic oligarchy in governing the higher education sector (Clark 1983). A more recent and elaborate concept is presented by Gornitzka and Maassen (2000) who, based on work of Olsen (1988), distinguish four steering modes: (1) the sovereign rationality-bounded mode, where steering lies with the government, (2) the institutional mode, where steering decisions are taken within the universities, (3) the corporate-pluralist mode, where different stakeholders negotiate the steering, and (4) the supermarket mode, which uses market mechanisms to steer the sector. These modes differ mainly with regard to: (a) the role of the state, and (b) the nature of professional autonomy in higher education, or using the terms of Van Vught (1997), whether a state control model or a state supervising model is used.

The concepts presented above show that there are different forms of political control over higher education that lead to differing levels of centralisation of power between the state and the higher education sector. As political parties have distinct preferences how to steer a public sector and how much autonomy professional communities should enjoy, it can be expected that these also play a role in their positions in higher education policy. These partisan conflicts can be captured through an analytical dimension, which focuses on the control over the higher education sector. It spans between the ideal points of centralised control, where the state and its bureaucracy are the main actors steering the higher education system, and de-centralised control, where higher education institutions are only supervised at arms-length and enjoy high levels of autonomy.

Positioning parties on this dimension that addresses different levels of centralisation of control over higher education, it can in general be expected that parties of the political left due to their preference for a strong and interventionist state will position themselves towards the centralised-control side (Jungblut 2014). Parties of the political right however, due to their preference for a weak state and focus on individual rather than collective solutions, will prefer de-centralised-control. These positions will be elaborated in the following section.

The two-dimensional analytical framework, exemplary higher education systems and possible party positions

After presenting both analytical dimensions separate from one another this section will bring them together, creating a two-dimensional analytical frame for partisan positions in higher education policy. The ideas behind this are exemplified by presenting illustrative higher education systems as well as “ideal” party positions linked to the analytical frame.

The two dimensions that form the basis for the framework are the re-distributive dimension and the control dimension. The former spans the ideal points of a restrictive and an expansive higher education system, while the latter spans the ideal points of centralised control and de-centralised control. Figure 1 illustrates this relation.

Fig. 1
figure 1

The two-dimensional analytical frame

Full size image

To make the ideas behind this analytical frame more tangible one can provide empirical examples of higher education systems for each of the four quadrants of the figure. Starting in the upper right corner a higher education system that is characterised by centralised control and restrictive re-distribution is the German system. Germany has a long tradition of a strong vocational education system that provides attractive employment opportunities without the need to access higher education as well as a matching selective secondary education system. These are two of the main reasons for its rather low participation rate in higher education, which is below the OECD average (Andres and Pechar 2013). Even though there are barely any private costs linked to attending higher education in Germany, as tuitions fees are either absent or comparatively low, due to its low participation rate Germany can be characterised as having a restrictive higher education system. This also fits to its conservative continental welfare state model (Busemeyer et al. 2013; Esping-Andersen 1985).

One might argue that the control over the German higher education system is rather de-centralised due to the fact that it lies mainly with the governments of the Bundesländer as the federal level only has very limited authority concerning education. However, in reference to the analytical framework proposed in this article, Germany should rather be considered as being centralised. The reason for this is that the main decisions concerning higher education are taken by the governments and bureaucracies of the Bundesländer and not the higher education institutions themselves (Van Vught 1997). As the decisive criterion for the control dimension of the framework is the relation between state structures, on whatever level they may be situated, and the higher education sector, even in federal countries where higher education is steered by subsidiary entities one finds centralised control. This also resonates with the classical model of the Humboldtian university where the state owned and controlled the higher education institutions (Kogan and Marton 2006).

In the upper left corner Norway can serve as an exemplary case. Like all of the Scandinavian countries its higher education system is characterised by being very accessible and expansive and has a participation rate in higher education that is significantly higher than the OECD average (Andres and Pechar 2013). Additionally, public higher education in Norway is completely tuition free both for Norwegians and EU citizens but also for students from non-EU countries. The Norwegian higher education system thus also reflects the nature of the Nordic welfare state model (Esping-Andersen 1985). At the same time, the dominant actor in relation to the control of the Norwegian higher education system is the state. Even though there have been several reforms that introduced for example intermediary bodies like a quality assurance agency and universities have more freedom in applying variations in their internal governance structures, the higher education institutions still have less room to manoeuvre than in other Scandinavian countries like for example Denmark (Gornitzka and Maassen 2011). Due to the fact that the state still controls the higher education sector in Norway, it can be considered as a centralised system. Therefore, both examples in the upper half of the figure resemble what Clark referred to as the “continental mode” of authority distribution (Clark 1983, 125f) or what Kogan and Marton labelled the “State model” (2006, 77f), where the relation between the universities and the state is characterised by high bureaucratic authority on the side of the government.

It is hard to find a contemporary example for the lower right corner of the framework, however historically the situation of the British universities, illustrated for example by Oxford and Cambridge up until the early twentieth century can serve as an example. Both institutions were characterised until then by being rather exclusive in their admissions, socially selective and therefore fitting well to the restrictive side of the re-distributive dimension. At the same time, both institutions were governed by what Clark referred to as the “British mode” (Clark 1983, 127) and enjoyed a very high amount of professional independence (Clark 1983, 127ff). As the state kept its interference in their activities to a minimum and both universities could govern themselves autonomously, they provide a good example for a situation where the control over the higher education system is de-centralized.

Finally, in the lower left corner the higher education system of the United States serves as an example. Due to its many different types of higher education institutions catering to different groups of students and a participation rate that is above OECD average (Andres and Pechar 2013), US higher education can be seen as expansive. However, the wide access is combined with sometimes high personal costs for attending higher education, which vary greatly between different types of higher education institutions and are most of the time covered through student loans. This is also in line with the liberal welfare state model present in the United States (Esping-Andersen 1985). At the same time, the higher education sector in the US enjoys a high level of autonomy. Both private and public institutions can freely decide on most substantial issues without state interference and the control of the sector is mainly regulated through market mechanisms and academic self-governance. Clark referred to this as the “American Mode” (Clark 1983, 129). Therefore, the US system can be characterised as an example for de-centralised control. Both systems in the lower half of the figure resemble what Kogan and Marton called the “Liberal model”, where the university is kept out of reach of the public authority (2006, 77f).

Figure 2 now visualises all examples.

Fig. 2
figure 2

Illustrative higher education systems

Full size image

Following these examples, one can link the systems presented to political parties that would prefer such an “ideal” higher education system. This gives an indication of how different families of political parties can be expected to position themselves in their higher education policy. While single representatives of these families in specific countries might be more or less in line with the general argument presented here, the overall relation of the different families to one another nevertheless holds.

As a party of the political left, a Social Democratic Party can be expected to support an expansive higher education system that provides strong re-distributive possibilities to its electorate and limits individual costs for higher education (Ansell 2010, 137). This is in line with the results from multiple studies showing that left parties lead to higher public spending in areas that create the potential for societal re-distribution (Boix 1997; Busemeyer 2007; Dalton et al. 2013). At the same time, due to the preference of the Social Democrats for strong and active states that shape the life of their citizens and their critical distance to market mechanisms, they will prefer a more centralised control of higher education (Jungblut 2014). This is rooted in the neo-Weberian assumption of a positive role of the state (Hood 1996; Paradeise et al. 2009; Pollitt et al. 2007) and puts the Social Democrats in the upper left quadrant of the framework. One example for such a Social Democratic Party is the Norwegian Arbeiderpartiet.

A Christian Democratic Party can be expected to be in favour of a more restrictive higher education system, on the one hand because of its desire to limit re-distribution and protect the wealthier part of its electorate (Ansell 2010, 137), but also on the other hand because of its support for vocational professions (Busemeyer et al. 2013, 6f; Iversen and Stephens 2008, 611f). While Christian Democratic Parties are generally in favour of giving competences to local institutions, they put a high emphasis on the qualitative homogeneity of public services including education (Busemeyer et al. 2013). Thus, for Christian Democrats the state has an important role in assuring an equal level of quality also in higher education (Jungblut 2014). This demands an active state and more state control. Therefore, Christian Democrats can be expected to opt for more centralised control of the higher education sector. This positions them in the upper right quadrant of the framework. One example for a Christian Democratic Party supporting these positions is the German CDU.

Contrary, a Conservative party would support more de-centralised control. This is due to their preference for smaller states and more streamlined public services, which are often combined with forms of market competition and also reflected in their support for new public management reforms (Green-Pedersen 2002, 272, 274; Jungblut 2014; Paradeise et al. 2009, 89f; Pollitt et al. 2007). Furthermore, Conservatives are not opposed to heterogeneity in the quality of higher education as this is an expression of and regulated through market competition. As Conservative Parties are expected to shield their wealthier electorate from re-distribution, they prefer a restrictive higher education system. Such a system would on the one hand limit public spending and on the other hand protect the labour market advantage of the more privileged electorate of Conservative parties that profits from the skewed-access to higher education (Ansell 2010, 137; Iversen and Stephens 2008; Rauh et al. 2011). This puts the Conservatives in the lower right quadrant of the framework. One example for such a party is the British Conservative Party.

Finally, a Liberal Party supports a small public sector, a more streamlined state structure and market competition. Just like the Conservatives, Liberal Parties see heterogeneity in the quality of different higher education institutions not as a problem for the state but rather as an issue which is regulated through market mechanisms. They support new public management reforms and an increase in the autonomy of higher education institutions to enable them to compete freely with others and find their niche in the (inter-)national higher education market (Green-Pedersen 2002, 272, 274; Hood 1996; Jungblut 2014; Pollitt et al. 2007). Thus, Liberal Parties can be expected to favour de-centralised control. One of the core ideas of Liberal Parties is that individuals should be able to use their skills to improve their socio-economic status by themselves without state interference or support. Furthermore, Liberal Parties also want that the education system caters to the labour market’s needs to ensure the future well-being of the economy. With the growing need for a highly skilled workforce in today’s knowledge economies, Liberal Parties can be expected to support an expansive higher education system to ensure that the economy is well supplied with qualified graduates (Ansell 2010, 137; Iversen and Stephens 2008; Rauh et al. 2011). However, due to the Liberals’ focus on the individual’s ability to shape his/her future, it can be expected that they favour some form of individual costs for attending higher education. This would have two positive effects for them: First, it would ensure that public spending for higher education would remain within certain limits and second it would favour the more wealthy electorate of the Liberal Party in comparison to the electorate of Social Democrats or other left parties, as it would be less deterring for the Liberal Party’s electorate to participate in higher education even though there would be individual costs linked to it. Therefore, Liberal Parties can be expected to position themselves in the lower left quadrant of the framework. One example for such a party is the Dutch party VVD. (Fig. 3)

Fig. 3
figure 3

Preferences of political parties in relation to the framework

Full size image

Based on the analytical framework one can propose four hypotheses for further research:

If a political party belongs to the group of Social Democratic Parties, then it is more likely to favor an expansive and centrally controlled higher education system.

If a political party belongs to the group of Christian Democratic Parties, then it is more likely to favor a restrictive and centrally controlled higher education system.

If a political party belongs to the group of Conservative Parties, then it is more likely to favor a restrictive and de-centralized higher education system.

If a political party belongs to the group of Liberal Parties, then it is more likely to favor an expansive and de-centralized higher education system.

Such hypotheses are harder to formulate for other party families. On the one hand, anti-establishment parties at the left and right fringe of the political spectrum are known for having only very limited positions concerning higher education policy, making it hard to match them to the two-dimensional framework (for details see Ansell 2010, 137ff). On the other hand, Green parties are harder to pin point because their electorate is more and more drawn from high income and highly educated strata, while their election manifestos hold diverse positions having both politically left and right characteristics (Rauh et al. 2011). This diversity makes it hard to predict where their preferences concerning higher education policy might be and therefore this article refrains from formulating clear hypotheses in relation to Green Parties.

Path dependencies and constrained partisan competition

The positions of political parties presented in the previous section are based on typical higher education systems that can be expected to be preferred by the respective party. Whether a political party would formulate higher education policies, for example in their election manifesto, which are completely in line with this ideal depends on more complex circumstances. The reason for this is that these decisions are not taken in a vacuum, but rather in relation to existing institutional contexts that influence the parties’ actions.

Political parties, both in the process of forming their positions but also when governing, cannot ignore prior policy decisions or the structural setting of a country or a certain public sector (Gingrich 2011). In higher education the structure of the existing higher education system is such a factor, since the existing institutional arrangement presents the starting point for any political discussion of reforms. Therefore, change in higher education policy can be seen as an interplay between institutional context and agency (Bleiklie and Kogan 2006, 12f, 17) and dynamics in higher education policy also depend on the political situation and political possibilities linked to it (Kauko 2013). A good conceptualization of this can be found in the work of Ben Ansell (2010). He expects governments to face a trilemma in higher education policy, as they can only achieve two out of three possibly desirable policy objectives: mass enrollment, full public subsidization, and low total public costs. If this is the case, then a party that would like to shift the focus to the neglected objective has to sacrifice one of the other two, leading to path dependencies for policy decisions and limited room to manoeuver for the party. An additional way out of the trilemma would be the option of increasing access without any additional funding, leading to a deterioration in quality (Plümper and Schneider 2007). All in all, these examples are in line with the findings of Bauer et al. (2006, 172) that differences in the starting points of higher education reforms lead to differing reforms and outcomes.

Another possible source for path dependencies is the policy legacy stemming from the question how the conflict between church and state over education has been settled. Countries with a dominant Catholic heritage and a strong Christian Democratic party in government, experienced the creation of conservative welfare states, segregated educational systems, and slower expansion of the educational sector with only a limited access to higher education (Busemeyer et al. 2013, 6f). One example for this can be Germany. In such a country one can expect parties of the political left to be the main proponents of educational expansion. Countries with a strong Protestant tradition are expected to be more supportive of educational expansion, characterized by comprehensive and state-centered education systems. An example here could be Norway. In these countries the political right is expected to be the main issue-owner in education (Busemeyer et al. 2013, 6f).

The two examples presented above show that when analyzing the impact of political parties and their preferences concerning change in higher education policy, the main explanatory power does not lie with pure partisan preferences but rather with constrained partisan preferences (Gingrich 2011). These constrains stem from the fact that the existing structure of the higher education system shapes any political debate around possible reforms in higher education. Changes in the system create winners and losers and existing institutions pre-structure political opportunities and therefore limit the options of parties to position themselves freely. Thus, if for example a Social Democratic Party finds itself in a higher education system that is restrictive and whose control is de-centralized, it is rather unlikely that they propose a reform agenda that would transform the system directly into an expansive and centrally controlled system. Rather, the party first proposes a shift along one of the two dimensions and only after this was successfully implemented proposes the next step. Such a strategy would also help to limit the possible resistance from the higher education sector by preventing too much change at once. Therefore, depending on the higher education system and the chosen reform strategy inter-country differences in the political positions of parties from the same party family can be possible.

Additionally, once a party enters the government the process of policy implementation presents even more constrains. Here questions linked to possible coalition negotiations and resulting programs for the government limit the transferability of a party’s program into action. Furthermore, the implementation of the governmental program depends on many intermediary factors such as the availability of resources, the cooperation of the bureaucracy or the strengths and support of professional organizations and stakeholder groups (Hupe 2011). All of these factors can hinder the implementation, interfere with the policy’s impact and limit its visible effects. Due to this possible difference between a party’s policy position and the subsequent implemented policy, it is advisable when studying the role of political parties in change in higher education policy to use two dependent variables and assess separately the difference in policy positions and the different policy outputs between different parties and governments composed by them.

Conclusion

This article offered an analytical framework that allows for a more in-depth analysis of higher education policy. By putting political parties in the center it introduced a new focus into the literature and connected work from political science with approaches from higher education studies. In doing so it complemented the existing, more structurally focused approaches (e.g. Bleiklie and Michelsen 2013) by focusing stronger on the role of parties as actors in change processes in higher education. This broadens the scope of analysis as it goes beyond the assumption that countries have put in place similar policy objectives with variations in reforms being mainly due to the countries’ politico-administrative system and rather focuses on the partisan characteristics of higher education policy.

The article raised two questions: How can political parties’ preferences in higher education policy be conceptualized; and what could possibly constrain partisan competition in higher education policy? To capture how political parties position themselves in higher education policy the article proposed a two-dimensional analytical framework. This framework is build up by one dimension capturing re-distributive conflicts and a second dimension capturing conflicts linked to the control over the higher education sector. The framework was exemplified by presenting one illustrative higher education system for each of its quadrants and linking these to political parties that would favor such a system. In this process four hypotheses for further research have been presented. Finally, the article argued, in line with Kauko (2013) that political parties when positioning themselves in higher education policy face a situation of constrained partisan preferences, as the existing institutional setup of the higher education system might limit their room to manoeuver when creating their policy positions. Furthermore, due to the complex process of transferring partisan positions into policy outputs it is advisable to study the role of political parties in higher education policy change by separately assessing the difference in policy positions and the different policy outputs.

The analytical framework presented offers multiple avenues for further empirical research. Applying it and testing the presented hypotheses would add to the so far only limited amount of literature on the partisan influences in higher education policy. It would also add to the general knowledge basis on partisan policy positions and the link between partisan input and policy output. Through the integration of political parties in analyses of higher education policy one can loosen the assumption used by the more structural approaches (e.g. Bleiklie and Michelsen 2013), that higher education policies follow similar objectives. One way of doing this would be through detailed analyses of election manifestos and political programs of different parties specifically concerning their position on higher education policy. Such an endeavour would be especially fruitful since most existing comparative analyses of election manifestos only cover the category of education policy as a whole, not paying attention to the differences between primary, secondary and higher education (see for example Ansell 2010; Busemeyer et al. 2013). Furthermore, such an empirical investigation would allow for the integration of anti-establishment and Green Parties into the presented framework.

Second, an analysis of partisan preferences and higher education policy outputs that goes beyond measuring the level of public spending and the partisan composition of government would offer a clearer picture how a change in government might also lead to a shift in higher education policy outputs on both dimensions presented above. Furthermore, the link between political parties and their electorate, specifically concerning the questions, (1) whether the parties orient themselves towards one another or their respective electorate when deciding on policy positions in higher education, and (2) whether the parties’ electorates have specific preferences concerning higher education policy, is not sufficiently studied so far.Footnote 3 Consequently, including political parties, their positions and policy output in the work on higher education policy analysis will add more analytical tools to the existing toolkit and offer fruitful connections between political science and higher education studies.