Higher education governance and policy: an introduction to multi-issue, multi-level and multi-actor dynamics

Abstract This thematic issue introduces the multifaceted nature of contemporary public policy – its multi-level, multi-actor and multi-issue features – using the case of higher education policies from around the world. To do so, this introduction first describes how higher education as a policy sector should be garnering far more attention from scholars interested in political, economic and social transformation. A framework for identifying and accounting for how the ‘multi-s’ characteristics configure and re-configure public policy is then introduced. Next, this thematic issue’s contributions are summarized with highlights of how they bring to life the different ‘multi-s’ features. This introduction concludes with a discussion of what the proposed framework of the ‘multi-s’ offers to studies of higher education policy coordination. In so doing, the objectives of this thematic issue are to highlight what the case of higher education policy coordination offers to studies of public policy and to initiate a dialogue between all social scientists and practitioners interested in the increased complexity of governing, producing and using knowledge today.

This thematic issue of Policy and Society focuses on the increased multifaceted characteristic of contemporary public policy (Peters, 2015). Using the case of higher education policies from around the world, we highlight the multi-level, multi-actor and multi-issue -'multi-s'nature of public policy in areas of growing international and political attention.
The global shift towards knowledge-based economies and societies has placed 'knowledge' at the core of contemporary public policy and policy-making. The governance of knowledge, however, is not a neatly contained policy coordination exercise: it requires collaboration across multiple policy sectors that may have previously experienced very little or less interaction. A non-exhaustive list of relevant policy areas includes higher education, research, trade, foreign policy, development and home affairs (migration). Higher education policy coordination is thus permeated with respective sectoral concerns, with discussions taking place across distinct policy arenas, sometimes in silo, both inside and KEYWORDS Higher education policy; higher education politics; knowledge; multi-level; multi-actor; multi-issue; governance outside of formal government channels. While this brings forth the multi-issue aspect competing for attention in higher education policy coordination, it also points to the presence of multiple actors: state actors from different ministries and agencies, representatives from universities and businesses, other non-state actors (interest groups, stakeholder organizations), as well as consumers of such coordinative outputs (concerned parents and students). The multi-issue and multi-actor features of higher education policy coordination often result in duplication, competition, inconsistencies, clashing priorities and even potential bureaucratic and political conflict (Braun, 2008;Peters, 2015) -all symptoms of horizontal policy coordination challenges (Gornitzka, 2010). Adding to this is the fact that actors involved in the design, implementation, monitoring and evaluation of higher education policy (Gornitzka & Maassen, 2000;Olsen, 1988) often operate, and 'shop' for better policy solutions, across several governance levels. While the rise of the regions -both supranational and subnational -in the higher education policy domain has garnered some academic attention (Chou & Ravinet, 2015;Jayasuriya & Robertson, 2010), this multi-level dimension of policy coordination needs to be brought into sharper relief. Indeed, international knowledge policy coordination stretches across many levels, including the macro-regional (e.g. European Union [EU], Association of Southeast Asian Nations [ASEAN]), the meso-regional (Nordics, Baltics -bilateral or multilateral cooperation among states sharing specific geographical features), sub-regional (also bilateral or multilateral cross-border cooperation between distinct territories of different states), as well as the state/national (in federal systems), sub-national and organizational (see e.g. Piattoni, 2010 concerning multi-level governance (MLG) in the European context).
These observations of the many 'multi-s' reveal that the higher education sector provides an empirically rich context for acquiring further insights into contemporary public policy-making where the demarcation between the public and private is increasingly blurred. Indeed, similar to other sectors marked with significant salience and high professional autonomy (e.g. health, see Pinheiro, Geschwind, Ramirez, & Vrangbaek, 2016), the case of higher education policy coordination encourages the investigation of multi-issue, multi-level, and multi-actor facets of policy processes, both individually and when they interact. Following the interdisciplinary ethos of Policy and Society, the analyses of these characteristics and their interactions in this volume employ several theoretical lenses, drawing on policy analysis, public administration, comparative politics, international relations, political economy, party and interest group politics and political sociology. The contributions include both in-depth empirical case studies and application of the formal approach to offer a deeper understanding of the underlying mechanisms of complex policy dynamics as well as their generic features (Rohlfing, 2012). Furthermore, the contributions go beyond the usual cases of countries belonging to the Organisation for Economic Cooperation and Development (OECD) to include less researched countries in regions such as Central and Eastern Europe, East Asia and South-East Asia to provide additional comparative contexts and observations. These cases show how policy actors grapple with the coordination challenges and exploit the opportunities of the knowledge 'turn' in contemporary public policy in their own ways, often reaching their decisions to do so with insights of how others have confronted similar scenarios.
This introduction to the thematic issue sets the stage for this conversation. It first introduces higher education as a policy sector that should be garnering far more attention from political, policy and administrative scientists interested in political, economic and social transformation. To structure this exchange, the sections that follow will present an analytical framework emphasizing the multi-issue, multi-actor and multi-level features of contemporary policy-making and policy coordination before describing how this volume's contributions highlight and bring to life these diverse facets. This introduction concludes with a discussion of what the 'multi-s' framework offers to studies of knowledge policy coordination. In so doing, the objectives of this thematic issue are to highlight what the case of higher education policy coordination offers to studies of public policy, and to initiate a dialogue between all scientists, scholars and practitioners interested in the increased complexity of governing, producing and using knowledge today.

Why is the case of higher education so interesting for public policy studies?
Higher education is a domain that is rapidly becoming more politicized, politically salient and an embedded element of other public policies (Busemeyer, Franzmann, & Garritzmann, 2013;Gornitzka & Maassen, 2014;Jungblut, 2015). Various factors contributed to making higher education central to the development of modern states and societies. First, there has been continuous massification of higher education, implying that an increasing percentage of the population is participating or expects to participate in some form of higher education (Andres & Pechar, 2013). While this has been a development in the making for many decades, it has generally led to growing public and private investments in the sector (Altbach, Reisberg, & Rumbley, 2009). What this means for studies of public policy is that the expectant demands for returns from these investments have made both politicians and citizens more sensitive to developments in the higher education domain.
Second, in a number of respects higher education has gradually become more relevant for other policy areas. For example, as societies face and become aware of a growing amount of grand challenges such as climate change, ageing and food security, policy-makers anticipate and want their (functioning) higher education and research systems to provide (some form of) policy solutions. Indeed, higher education sectors around the world are increasingly under pressure to 'export' policy solutions to other policy domains (Gornitzka & Maassen, 2014). This is in strong contrast to other policy areas with a strong transnational dimension, such as environment, which are 'imported' by other policy domains as a policy problem that must be tackled. Indeed, this turn to 'knowledge' for policy solutions is increasing and can be seen as integral to modern policy-making (Boswell, 2009). At the same time, higher education, universities and research institutions are under pressure for not offering the 'right' solution or not providing solutions 'on time' . In this, higher education can be seen as a policy area that is becoming more important but through this is also treated less special and facing increased political pressure to perform (Gornitzka & Maassen, 2014). Indeed, there are also many democratic legitimacy control issues at stake when discussing the usage of knowledge (Gornitzka & Holst, 2015), and questions about the relative and actual policy effectiveness of the 'scientific approach' and policy design (Chou & Ravinet, 2016;del Rio, 2016).
Finally, the move towards knowledge economies and societies additionally strengthened the role of higher education, as well as research and innovation, in the development of national economies (Maassen & Stensaker, 2011). What this means is that higher education has been integrated into debates in other policy areas (Braun, 2008), including reforms ushered in by New Public Management (NPM) (Paradeise, Bleiklie, et al., 2009;, and discussions about efficient public sectors and active welfare states (Gingrich, 2011(Gingrich, , 2015. These developments ensured that higher education has become a more relevant issue in many policy arenas, with new actors becoming more active in policy-making for the higher education sector. Indeed, a deeper understanding of higher education policy dynamics can be considered a necessary condition for the analysis of other public policies given the deep penetration of these dynamics into both public and private sectors. While higher education study is a thriving field of research (cf. specialized journals such as Higher Education, Higher Education Policy, Research in Higher Education, Studies in Higher Education and Higher Education Research and Development), it has yet to engage systematically with debates in the policy sciences despite the many implications that this sector has for public policy-making. It is the aim of this thematic issue to contribute to closing this gap in the literature. Research on higher education policy over the last two decades has especially focused on two aspects of particular interest for policy and administrative scientists. First, extant research has identified and examined the policy actors, such as the ministerial bureaucracy and organized sectoral interests (see e.g. Enders, de Boer, & Weyer, 2013;Goedegebuure & Van Vught, 1996;Goedegebuure et al., 1993), affecting higher education policies. Second, higher education policy research has investigated the importance of existing politico-administrative structures (see e.g. Bleiklie & Michelsen, 2013;Braun & Merrien, 1999;Capano, 2015b), singling out one part of a complex set of interactions within dense organizational set-ups in shaping higher education policy. Beyond these contributions, political, administrative and policy scientists have largely neglected higher education (and education) policy as fields for research in the last two decades (e.g. Busemeyer & Trampusch, 2011;Gift & Wibbels, 2014); but this is now changing (see also e.g. Chou & Gornitzka, 2014;Jarvis, 2014 introducing vol. 33, issue 3 of Policy and Society; Musselin & Teixeira, 2014).
In recent years, there is a growing interest in the role of politics, policy-making and bureaucratic structures in affecting outcomes in the higher education sector. For example, scholars applying a political economy and party politics approach have shown that political parties have distinct preferences concerning higher education policy and that these preferences are translated into policies when these parties come into government (e.g. Ansell, 2010;Garritzmann, 2015;Jungblut, 2016aJungblut, , 2016b. Additionally, there seems to be a link between the institutional structure of welfare regimes and education systems, which highlights arguments of institutional complementarity but also interlinkages between welfare and education policies (Busemeyer, 2015;Iversen & Stephens, 2008;Willemse & de Beer, 2012). The time is thus ripe to demonstrate how to conduct a systematic analysis of the multi-issue, multi-level and multi-actor characteristics of higher education policy coordination processes, as well as provide a first analysis of these very dynamics in a single volume. In the next section, we present a framework to reveal better the value-added of higher education policy as a case for studies of complexity and contemporary policy-making.

Unpacking the 'multi-s': a framework for studying higher education policy coordination
The recognition over the years that public policy-making and policies are increasingly complex has prompted political and policy scientists to revise their respective analytical frameworks to account for the 'multi-s' features discussed above. One such set of efforts has resulted in the vast literature on MLG, which has become a taken-for-granted perspective to describe policy coordination across different governance levels. In the light of this, the following begins with a critical assessment of the MLG literature, in particular with regards to its potential to adequately capture the relevant facets of contemporary policy dynamics. An extension of the framework is then offered by highlighting explicitly two additional 'multi-s' -multi-actor and multi-issue, which should be considered as independent from each other and the (refined) multi-level dimension.
The emphasis on the multi-level dimension reflects the preoccupation in comparative politics, policy analysis and European studies to account for the changes in governance arrangements away from the centralized state. According to one of the most cited contributions on MLG -'Unravelling the Central State, but How' by Hooghe and Marks (2003) -these shifts can be categorized as two distinct types of MLG: one in which different levels of authority are neatly nested within each other and which is designed to comprise an entire fixed system of governance (Type I, e.g. typical federalist structure) and the other in which the focus is on task jurisdictions which may change should the need arise and where jurisdictions may overlap (Type II). While in Type I, there is a limited number of governance levels and the same numbers exist across all public sector domains (e.g. local, state, federal), in Type II the number of governance levels is not limited and depends on the domain (e.g. more levels related to education than health).
This simple categorization of governance arrangements as Types I or II already reveals the many potential variations of MLG currently subsumed under Type II. Indeed, it can be argued that this dichotomous distinction is less useful when many governance arrangements around the world, especially international or intergovernmental cooperation, are policy sector-based and may therefore be qualified as Type IIs. Furthermore, these categories are unable to describe what happens when multiple Type II governance arrangements are in place for achieving an overarching common objective. For example, efforts to construct a common area of knowledge in Europe (see 'Europe of Knowledge' in Chou & Gornitzka, 2014) encompass developments in the higher education policy sector (i.e. European Higher Education Area [EHEA]), in the research policy sector (European Research Area [ERA]) and in the innovation policy sector. What is notable about these developments in Europe is that each set of sectoral governance arrangements follows a distinct method of coordination and upholds their individual sectoral rationales, even though policy reforms have been introduced to promote coherent coordination across these sectors. The starting point of the framework presented in this section is thus to challenge the utility of MLG as a framework to account for the growing complexity observed in contemporary governance. By demonstrating how MLG conflates multiple aspects of governance arrangements under the broad label of 'multi-level' , the discussions call for and present a framework of 'multi-s' to replace MLG as the preferred analytical approach to studying today's complex public policy coordination.
Scholars interested in 'beyond the centralized state' developments (Zürn, Enderlein, & Wälti, 2010, p. 1) have embraced MLG as an approach to unpack the complexity between the different functional governance layers. MLG has found particular resonance among European integration scholars, with the 'unique setting [of the EU serving] as key catalyst in the development of multi-level approaches' (Zürn et al., 2010, p. 1). Indeed, MLG has variably been used to describe the institutional and issue complexity of EU policy-making (Egeberg, 2006), to highlight interdependence of the decision-making actors and processes across national and supranational governance levels and to tackle normative issues of legitimacy, democracy and accountability (see Stephenson, 2013 for a review of use of the concept over two decades). There are, however, strong efforts to generate more generic insights into MLG, both using the case of European integration as well as looking beyond Europe.
As for interesting innovations of MLG in recent European research, Piattoni (2010), for example, more explicitly connected MLG with upwards, downwards and sideways shifts in governance arrangements (cf. Hooghe & Marks, 2003). Specifically, she identified three conceptually distinct dimensions of MLG: (1) domestic-international (upward shift), (2) centre-periphery (downward shift) and (3) state-society (sideways shift) (Piattoni, 2010). Policy scholars have applied MLG to capture both opportunities and challenges of better coordination between governance domains as well as across governance levels, even if not labelled explicitly as such (Braun, 2008;Capano, 2015a). MLG has been criticized, among others, for its lack of predictive power (see Bache, 2008 for a review), the term still holds significant currency, including in higher education studies (see Fumasoli, 2015 for a review). Moreover, there are challenging accusations of concept stretching (Piattoni, 2010) and overuse of MLG in the literature: 'To permit a sartorial analogy, MLG has been thrown around by scholars like a favourite coat -a staple item in the European political science wardrobe, but perhaps one worn so often that it has now become threadbare' (Stephenson, 2013, p. 817).
As for using MLG to look beyond Europe, this has remained comparatively rare. There have been notable and stimulating exceptions of 'de-Europeanising' the MLG approach. Some scholars have adjusted and used the MLG approach to studying other regional governance systems (Sbragia, 2010): Dabène (2009), for instance, has applied the notion of 'regional multilevel integration' to the case of Latin America, and Schreurs (2010) characterizes what is MLG 'in the ASEAN way' . Different authors have also embarked on the path of defining global governance as MLG (Prado, 2007;Zürn, 2012). Yet beyond these contributions, there is a discussion of whether applying MLG beyond the European case can be 'especially tricky' (Sbragia, 2010, p. 268). In this thematic issue, it is argued that the way forward in this discussion must be conceptual; indeed, the conceptual extension of MLG research also goes hand in hand with a broadening of empirical applications. Conceptually, governance of complex systems is not only about multiple levels -there are various other 'multi-s' that are important to examine, and, empirically, there is of course no reason why this type of questioning should be confined to EU cases. A greater diversity of empirical cases will fuel conceptual creativity and ultimately even lead to a reassessment of European governance through a renewed conceptual lens. In order to generate more insights and examine the utility of the proposed framework, it is useful to focus first on an individual sector -higher education, for instance -to demonstrate its overall applicability. In the paragraphs below, the framework is presented and the extensions to the MLG approach are described.
To begin, MLG's reference to 'levels' as an all-encompassing descriptor is hugely problematic. Indeed, as the examples given above already show, the 'levels' within MLG have been opened to different interpretations, including functional, territorial and analytical distinctions (Piattoni, 2010). While part of the literature effectively uses the term to focus on the distribution of state authority across different levels (see e.g. Jordan, 2001 on multi-level government), the term has also been used to capture involvement of non-state actors (cf. Piattoni's state-society dimension). In that respect, the reference to 'levels' effectively conflates at least two distinct developments -distribution of authority across governance levels as well as division of authority between state and non-state actors. This is problematic for both theory-building and practical application of the MLG approach. The paragraphs below will parse out and distinguish the additional features of contemporary governance subsumed under the MLG label: multi-actor and multi-issue. In drawing out these two features, the framework presented in this section demonstrates how to extend MLG insights by incorporating insights common in public policy approaches to better capture the nuances of policy coordination processes and outcomes, which extant studies of (higher education) policy coordination have already uncovered. Therefore, the extensions of the MLG approach presented in the following paragraphs (the 'multi-s' framework) are known features in policy analysis, fuelled by recent developments in the field. The novelty of the 'multi-s' framework lies in the combination of these insights, which originated from different analytical lenses. It is argued that this combination of lenses is the key for capturing complex dynamics, given that it minimizes blind spots of each individual approach.
The first extension concerns the presence of multiple actors, which can be highlighted through the simple examples of vertical and horizontal tensions. Vertical tensions arise when different authorities (e.g. at the sub-national, national, supranational levels and so on) compete to exercise or defend regulatory competence on an issue and in a policy sector. Here, already, different cleavages can be identified: the North-South and East-West divide, as well as the Left-Right divide on the political spectrum. These classic tensions are the bedrock of the MLG approach; they are taken for granted assumptions about what happens when authority is devolved upwards, downwards and outside. In comparison, horizontal tensions refer to the same phenomenon between different sectors. Such horizontal tensions, first of all, direct attention to the observation that clashes and frictions can also emerge within a state and that this state, regardless of governance level, should not be treated as a unitary actor. The EU is an especially illuminating case for vertical and horizontal tensions: within this regional organization are complex relationships between the different member states, between the central institutions -the European Commission, Council of Ministers, European Parliament and the European Court of Justice, as well as within each of them. Governance arrangements emphasize the importance of attending to this multi-actor characteristic as a determining factor of political and policy outcomes, changes and evolution. Braun (2008), and Chou and Gornitzka (2014), for instance, have highlighted horizontal and vertical tensions in higher education, research and innovation policy coordination in Europe. What their findings show are that the presence of multiple actors, organized according to different principles and ideologies, encourages competition, promotes considerations for veto blocks, as well as fostering new alliances and partnerships. The very presence and importance of vertical and horizontal tensions suggests that the multi-actor feature of policy coordination needs to be made explicit.
The clashes and friction between the multiple actors mask another key characteristic of policy coordination, namely its multi-issue feature. Here, two questions are embedded in this confrontation: (1) who should be in charge or authorized to exercise regulatory competence? And, (2) how should a specific policy issue or problem be addressed? While the first question is about jurisdiction and ownership (and suggests that the distinction between the three 'multi-s' may not be so easily made, especially in practice), the second question is more concerned about how actors' articulated policy preferences could be reconciled or advanced, i.e. arriving at the 'best' and 'most acceptable' solution for those involved. It is important to acknowledge that policy actors may not necessarily differentiate between these two questions, often seeing them as one question when it comes to policy-making. This conflation is also seen in the principle of subsidiarity, which seeks to ensure that the central authority only performs policy functions that cannot be better carried out at the 'local' level. This multi-issue characteristic is particularly pertinent for the higher education domain because of the 'knowledge' turn in contemporary public policy-making. As discussed in the above sections, knowledge has become central to many policy sectors and to the transition towards knowledge-based economies and societies. It is thus essential to also make explicit the multi-issue feature in addition to the two other 'multi-s' , multi-level and multi-actor, because this characteristic often masks the hidden strategies that policy actors apply to achieve their sectoral goals and objectives in another policy domain (Chou, 2012).
To sum up, studying policy coordination in today's higher education sector requires unpacking three distinct characteristics of this very coordination -multi-level, multi-actor and multi-issue -and addressing them separately from one another as an independent perspective and recognizing their interaction as likely to be responsible for the outcomes observed, as elaborated below. Specifically, the proposed framework calls for analysing how the 'multi-s' features affect the stability, changes and evolution of individual and collective higher education policy coordination under observation. In academic practice (i.e. theory-building, research design and empirical fieldwork), it means that it is essential to pay attention in the following ways when examining individual 'multi-s' characteristic: (1) Multi-level characteristic -focus on the antecedents and consequences of distribution or concentration of authority across governance levels; (2) Multi-actor characteristic -acknowledge both the heterogeneity of the 'state' and its many composite institutions as well as the involvement of non-state actors (e.g. stakeholder organizations, businesses, consumers) in this policy domain; and (3) Multi-issue characteristic -identify how clashes as well as complementarities between policy sectors and spill-overs move into and away from the policy domain of interest.
Differentiating these three conceptually distinct features of policy coordination also encourages an examination of their interaction. It is often their very interaction, and the aggregate outcome, that is observed and conflated under the MLG label. If political and policy scientists are serious about moving beyond 'complex' (or 'complexity') as the default description of interlocking processes, these interactions must be clearly delineated, sequenced and their relationships explained. Together there are seven potential variations of 'multi' features that are of interest when examining policy coordination: (1) multi-level, (2) multi-actor, (3) multi-issue, (4) multi-actor and multi-issue, (5) multi-actor and multi-level, (6) multi-issue and multi-level and (7) multi-actor, multi-issue and multi-level.
The 'multi-s' framework offers a solid first conceptual step to encapsulate and unravel the complexity observed within contemporary higher education policy-making and coordination. For instance, this framework enables an account as to why non-state actors such as stakeholder meta-organizations (organizations of other organizations, see Ahrne & Brunsson, 2008), are important key actors in higher education. Take the European University Association (EUA) and the European Students' Union (ESU) as two examples; these organizations are multi-level organizations, operating at both the European level and, through their members, at the national and local levels. To influence policy decisions, EUA and ESU utilize their multi-level characteristic to their benefit, choosing strategically which level to address (see e.g. Beyers & Kerremans, 2012 on multi-level venue shopping;or Nokkala & Bacevic, 2014;Yagci, 2014 on EUA and ESU policy advocacy). By drawing out the multi-level features of these meta-organizations, it is easier to see how they may behave as state actors and have multiple access points into the policy-making process with or without formal inclusion.
Similarly, the proposed 'multi-s' framework encourages an investigation of the interaction between the different 'multi-s' . Take, for example, the interaction between multi-level and multi-issue features: distribution of regulatory authority between sectors is often organized differently between governance levels that then leads to conflict in achieving sectoral objectives. Knowledge policy coordination in Europe is again an excellent illustrative case of this conflict. At the national level, there may be one department dealing with both higher education and research, while at the supranational level these issues are addressed by two distinct departments. This leads to specific coordination challenges, both with regard to policy formation and policy implementation, with potentially significant implications for changes in how distributions are organized across levels (Egeberg & Trondal, 2009). In the same way, examining interaction between multi-actor and multi-issue characteristics not only allows an identification of policy framing and venue-shopping at one governance level, but also does justice to the heterogeneity of policy preferences between actors that sets the stage for strategic behaviour, negotiation and bargaining between actors.
The proposed framework of the 'multi-s' thus promotes the study of distinct features of policy coordination that have hitherto been conflated as merely one characteristic. Parsing out these features will enable a more nuanced understanding of how and why contemporary public policy-making is indeed so multifaceted and 'complex' . The following section elaborates how the contributions in this volume apply this framework of 'multi-s' to bring out the multi-issue, multi-level and multi-actor features of contemporary higher education policy coordination around the world.

The 'multi-s' in action: contributions from this volume
The contributions comprising this volume cover a wide geographical spread of higher education contexts -Asia, Europe and North America, and include comparisons both across and within these macro-regions. The contributions also focus on a diverse set of themes, including funding, work-based higher education programmes, policy formation at state, federal or macro-regional levels, policy implementation and the role of local authorities and stakeholder organizations. At the same time, each contribution highlights one or more 'multi-s' , thus exemplifying their potential to provide nuanced accounts of the complexities of governance arrangements in higher education.
The thematic issue opens with a contribution by Lacy, Tandberg, Fowles and Hu, who analyse a specific outcome of higher education policy processes -public funding for higher education -across time and across the states in the US. Their approach is grounded in an institutionalist perspective and builds on prior research that shows how interest group constellations, partisanship and state leadership (governors in the US case) matter for policy processes and outcomes. They expand this by focusing also on executive agencies responsible for oversight and management of the sector, which are under-researched organizational actors in higher education policy. Based on quantitative analysis of data from 49 US states over the course of 25 years , Lacy et al. go beyond explaining the volatility, as opposed to overall levels, of public funding by overall state economic conditions and show how centralization of bureaucracy in a state and agenda-setting power of the leadership of the executive agency matters. Their contribution therefore highlights the multi-actor dimension of governance in higher education, in particular with regards to unpacking the state as a unitary actor.
The issue of centralization of governance is also the focus of the contribution by Mok and Han. In their analysis on how branch campuses in China are governed, they identify two aspects of decentralization: (1) a horizontal one focusing on privatization and marketization of higher education, which effectively introduces new actors in the governing process (multi-actor dimension), and (2) a vertical one which concerns how power within the state is dispersed from the central level to local authorities (multi-level dimension). Based on in-depth interviews with a variety of policy actors, Mok and Han show that decentralization in higher education has been accompanied by re-centralization, given the overall hierarchical approach to governance in China in general. Moreover, given the interaction with state and market forces, transnational higher education in particular requires collaboration across multiple sectors, thus highlighting primarily the multi-actor dimension of higher education governance given the importance of various actors from both within China and abroad but also, by extension, the multi-issue (given the diverse jurisdictions of these actors) and multi-level (given the focus on decentralization) dimensions.
Jungblut and Rexe complement the focus on the multi-actor dimension by analysing governing structures involved in policy coordination in federal systems, thus adding also a focus on the multi-level dimension of higher education governance. Relying on the policy analysis literature concerning dimensions, barriers and strategies for coordination, they explore bodies that were specifically set-up in two federal countries -Canada and Germany -to organize coordination of higher education policies. Through interviews as well as analysis of policy documents and secondary sources, they identify both similarities and differences in how higher education is coordinated, with differences being primarily linked to diverse constitutional arrangements, indicative of path dependency and embeddedness of higher education governance into more general public sector governance arrangements (an example for higher education becoming more important, but also less special).
Similarity and differences of governance arrangements are also the topic of the contribution by Dobbins and Knill, who trace changes in France, Germany and Italy in relation to both transnational pressures and cultural, historical and institutional legacies. Identifying three governance types -state-centred, market-oriented and academic self-governance, and focusing on three aspects -general arrangements, financial governance, personnel autonomy -based on policy documents, interviews and secondary sources, Dobbins and Knill show a mixed pattern of convergence and divergence. Effectively, the analysis shows far less convergence in particular with regards to personnel issues, which are not solely an issue of higher education governance but also of civil service law, thus highlighting also the multi-issue dimension of governance. Conversely, where stronger transnational pressures for change exist, which also includes promotion of stakeholder involvement, convergence is stronger, demonstrating the interaction of multi-level and multi-actor dimensions of governance.
While Dobbins and Knill focus on governance of higher education in general, Graf zooms in on the expanding sector of work-based higher education, its positioning in relation to other forms of advanced skill formation and the involvement of various stakeholders (including employers, state agencies and higher education institutions). He compares Germany and the US, two countries typically identified as 'different systems' in terms of the institutional configuration of education and training and the respective educationeconomy nexus. However, based on document analysis and interviews, Graf identifies significant similarities in the emergent governance mode of work-based higher education across the two countries. In particular, the multi-actor dimension is highlighted in his analysis, showing how involvement of diverse public and private actors provides new possibilities for cooperation but also new possibilities for conflicts, which may require explicit government policies on contested issues.
Stakeholders are also the focus of Vukasovic's contribution, though on the European level. Building on the literature on interest groups, policy analysis and institutional isomorphism, she maps changes in policy positions of stakeholder organizations in European higher education. Distinguishing issues, preferences and ideology in these policy positions, she in particular analyses the extent to which the expectation of policy convergence -given the structuration of the European higher education policy arena -can be indeed identified. While quantitative and qualitative analysis of policy documents do not provide a neat picture of policy convergence, there is similarity in terms of issues on which stakeholder organizations focus. While this indicates some influence of the environment these organizations operate in, a lack of clear cut convergence of preferences and ideology is potentially suggesting stronger importance of policy dynamics within these organizations. This serves as a further reminder that the involvement of European stakeholder organizations is not only linked to multi-actor dimension of higher education governance, but also to the multi-issue (given the variety of issues these organizations focus on) as well as to the multi-level dimension (given that members of these organizations are national-level actors in their own right).
European policy-making is also the focus of Elken's contribution which analyses specific approaches to European integration in higher education in relation to the five policy modes traditionally used in European studies -Community Method, regulatory mode, distributional mode, intensive transgovernmentalism and policy coordination. While earlier scholarship identified the latter as the mode corresponding to higher education, Elken goes beyond such considerations of subsidiarity and limited role of European institutions (and the European Commission in particular) and stresses that soft governance comprising standardization may actually constitute a distinct policy mode. Building on the example of the European Qualifications Framework, Elken shows how developing standards creates new governance arenas, opens the doors for new actors (primarily experts), and also changes roles and positions of existing actors (e.g. Commission going from a marginal player to a standardizer), thus highlighting both the multi-level and the multi-actor aspects of European higher education governance.
In the final contribution to this thematic issue, Chou and Ravinet compare regional cooperation in Europe and South-East Asia. Emphasizing the importance of ideas on policy-making, they go beyond the often used argument (in the literature on policy diffusion and comparative regionalism) that highlights certain regions as idea-exporters (in this case Europe) and others as idea-importers (in this case South-East Asia). While European integration in the area of higher education has indeed garnered attention from policy-makers in other parts of the world, in the view of Chou and Ravinet this should not be equated as simple ideational diffusion from Europe to elsewhere. Rather, a closer look based on document analysis and extensive interviews with policy actors shows that such a Eurocentric approach effectively neglects endogenous regional developments. In other words, while it is important to do justice to the complexity of inter-regional and intra-regional governance dynamics (cf. multi-level dimension), diffusion of policy ideas is neither uni-directional nor all-powerful.

Moving forward and linking research agendas: what the 'multi-s' framework offers to studies of other public policy sectors
In a world that is now being increasingly characterized as volatile, uncertain, complex and ambiguous, the need, empirically as well as analytically, to delineate dimensions of interest and their interactions is great. Indeed, this was the starting point of this thematic issue, which introduced the higher education policy sector as a case of particular interest for political and policy scientists researching contemporary transformation around the world. With the knowledge 'turn' in policy-making, economy and society, the reforms and 'different ways of doing things' sweeping across the international higher education landscape are not part of a neutral process, un-orchestrated by an invisible hand. These are purposeful changes meant to induce particular shifts towards specific goals and objectives, and, in so doing, may be creating new winners while not succeeding in bringing forward those who have been left behind. To truly comprehend these many changes, unfolding at rapid speed, this thematic issue called for embracing the multi-level, multi-actor and multi-issue characteristics of contemporary public policy-making as the entry point for analysis.
The 'multi-s' framework is a portable device that scholars and practitioners can apply to unpacking developments across all policy sectors. It is especially applicable for examining those policy and issue areas with a strong transnational dimension such as sustainability, trafficking (of people, drugs, arms and so on) and terrorism, as well as sectors (e.g. health) marked with high political salience, significant professional autonomy and importance of the organizational level for dynamics and outcomes of change. Moreover, such portability implies that the 'multi-s' framework is also promising for cross-sectoral comparisons, which should become as central to policy analysis as international comparisons. While this thematic issue did indeed focus on a specific policy domain, it did so through an analytical framework that is non sector specific. It is hoped that the potential of the 'multi-s' approach would attract attention beyond researchers specialized in higher education policy, and pave the way for more systematic comparisons across policy sectors. A promising comparison is between higher education and health. An approach such as the 'multi-s' framework could be instrumental in engaging with this agenda of cross-sectoral comparison and, in so doing, provide greater coherence to policy analysis now currently missing.