Explaining Policy Convergence and Polity Divergence in Federal Systems: German and Swiss Higher Education

This article explores higher education reforms in the federally organized education systems of Germany and Switzerland. We argue that the increasing internationalization of educationç spearheaded by the Bologna Processçhas brought about changes not only in higher education policy , but also led to significant reconfigurations of the higher education polity . These processes present an empirical puzzle: while higher education policies in both systems have broadly moved in the same direction, they have significantly diverged regarding the higher education polity. While Germany has decentralized its higher education system, Switzerland has embraced more centralized structures. Drawing on historical institutionalism, we explain why Germany and Switzerland reacted to essentially identical external challenges with diametrically opposite polity reforms.We contend that this ‘‘polity divergence’’çdespite‘‘policy convergence’’çwas channeled by the preexisting institutional configurations of educational federalism which steered partisan conflicts along specific paths.

Increasing globalization is a catalyst not only for economic transformation but also for social and political change. Higher education (HE) is no exception and national systems have increasingly adapted to resulting pressures. Most prominently, the Bologna Process, initiated in 1999, aimed to make Europe the world's most attractive HE area by better coordinating the starkly differing systems in this region. Subsequently, national HE policies of forty-eight countries in Europe are increasingly coordinated through and influenced by non-hierarchical multilateral agreements.
Since the Bologna Process primarily addresses substantive policy issues, previous research has focused on the implementation of such decisions-e.g., study structures, mutual recognition of degrees, quality assurance mechanisms (e.g., Reinalda and Kulesza-Mietkowski 2006;Witte 2006) as well as international and national-level governance structures (Ravinet 2008;Dobbins and Knill 2014).
However, countries were often not only compelled to better coordinate educational programs and their recognition, i.e., HE policy, rather often also to adjust domestic decision-making institutions, thus HE polity. Yet, such changes cannot be characterized as a top-down process, through which the international level prescribes how national policy-making structures are to be modified; instead reforms of national political structures were essentially side-effects of perceived challenges of internationalization.
National HE systems vary significantly regarding the territorial organization of policy-making. While some systems have been historically subject to regulation by decentralized or local authorities (e.g., United States), others have been a domain of central-state authority (e.g. France). Multilateral agreements are particularly challenging for federal education systems. While whole countries generally sign these agreements, it is up to the subnational units to actually implement internationally agreed reforms.
Along these lines, federal systems are of particular interest as their reactions can differ internally regarding the HE polity; countries may experience centralization or decentralization, as political responsibilities for HE may be shifted to the central state, regions or HE institutions themselves. We argue that Switzerland and Germany are particularly interesting cases due to their oppositional HE polity reforms since the late 1990s: Switzerland has visibly tended towards centralization, while Germany has "re-decentralized" HE. This observation is remarkable for three main reasons. First, one might expect similar reform trajectories due to their relative socio-political and economic similarity. Germany and Switzerland are coordinated market economies (Hall and Soskice 2001), conservative welfare states (Esping-Andersen 1999) and generally characterized as reform-averse due to their federalist organization. Since HE is a joint responsibility of the federal government and the L€ ander/cantons, the subnational units could also act as veto players (Tsebelis 1995) to prevent undesired nationwide reforms (Behnke and Benz 2009). 1 Second, one might argue that the direction of observed polity change is counter-intuitive. Specifically, Germany could be expected to further centralize HE, as the decentralized German state is juxtaposed with a relatively culturally and linguistically homogenous society (Broschek 2014;Erk and Koning 2010). Particularly regarding education, demands for a greater federal role were constantly present insofar as it was perceived to enhance equal opportunities across the L€ ander . Against this background, the recent decentralization of German HE is rather ironic: In order to circumvent national policy-making obstacles induced by bicameralism and federalism, the central government actively pushed for the Europeanization of HE to enable more centralized coordination (Martens and Wolf 2006). However, the ensuing German convergence with European policy templates ultimately brought about the decentralization of the domestic HE polity. Swiss society, by contrast, is more culturally and linguistically heterogeneous and has traditionally attached great value to regional differences (Griessen and Braun 2010;Erk and Koning 2010). Accordingly, education has always been one of most decentralized policy areas in the highly decentralized federal system (Heidenheimer 1997;Hega 2011). Paradoxically though, the non-EU-member drew on Europeanization processes to centralize and thus to some extent homogenize its HE polity.
Third, with respect to recent overarching constitutional reforms, HE stands out as an outlier. Specifically, in 2006 and Germany embarked on a large-scale federalism reform to rearrange the competences of the federal government (Bund) and L€ ander. The then governing grand coalition's (CDU/SPD) reform aimed to reduce the influence of the L€ ander in federal policy-making, while simultaneously granting them additional decentralized legislative competences (Broschek 2014). Compared with constitutional reforms in other areas, HE decentralization was more far-reaching (Kaiser and Vogel 2017), as the federation has almost completely withdrawn from the policy area (Burkhart 2009;Behnke and Benz 2009). In contrast, Switzerland has displayed a general trend towards centralization over the past decades (Braun 2009). Yet, recent data show that Swiss HE has become more centralized-in legislative and administrative terms-than other key areas of the welfare state (e.g., healthcare, social welfare, employment relations) (Dardanelli and Mueller 2017).
Thus, we face a triple empirical puzzle: Why did two similar countries characterized by general reform inertia react to essentially identical challenges with similar policy reforms, yet nearly diametrically opposite polity reforms? Why did they ultimately introduce reforms which arguably were predestined to move in the opposite direction, at least from a political-cultural perspective? And why was the de/centralization of German/Swiss HE more far-reaching than the overall constitutional de/centralization process?
Our aim is to provide a conceptual framework for explaining the different HE polity trajectories in federally organized countries. While numerous researchers have focused on the widespread shift in decision-making autonomy to individual HE providers (e.g., Estermann et al. 2011), few authors have comparatively addressed changes in territorial authority over decision-making. We overcome this research gap by exploring the transformed relationship between the central government and subordinate territorial entities in Germany and Switzerland. We argue that while the overall reform process and ensuing policy convergence were largely driven by external factors, above all Bologna, polity divergence (i.e., German decentralization, Swiss centralization) can be explained by the respective dynamics of Swiss and German federalism and partisan politics. Drawing on historical institutionalism we show that institutional configurations of educational federalism channeled partisan interests and political conflicts along specific paths.
We proceed by briefly outlining our theoretical framework and the concepts of centralization and decentralization, before discussing pre-existing federalist HE arrangements. Our case studies are based on process-tracing (Beach and Pedersen 2013), which allows us to establish a coherent sequence of causal mechanisms ultimately leading to the explanation of HE de/centralization. We rely on public debates, position papers from stakeholders, secondary sources, interviews with involved stakeholders, 2 and legislative documents.

De/Centralization, National Institutions, and Partisan Politics
While centralization describes the transfer of decision-making authority to the central government and its agencies, decentralization refers to the shift in authority from the central state to a subordinate organizational level (Men endez-Weidman 2001). The subordinate level may comprise subnational decision-makers (e.g., municipalities) or, in our case, education providers. This means that important decision-making responsibilities-such as funding and resource allocation, personnel management, student admissions, study content, and quality evaluation-are transferred to subnational state units or individual HE institutions. While administrative decentralization primarily pertains to the increase of selfgovernance capacities of individual universities only, i.e., university autonomy, territorial decentralization entails shifts in territorial competences from the central government to regional, state or local state authorities (e.g., the German L€ ander, Swiss cantons). Thus, university autonomy may remain low amid territorial decentralization. Such was previously the case in Germany until the mid-1990s, where universities' personnel and financial autonomy was limited despite the political responsibility of the L€ ander (Dobbins and Knill 2014). Dardanelli and Mueller (2017) also highlight crucial differences between the legislative, administrative, and fiscal dimensions of de/centralization and/or autonomy. Legislative autonomy pertains to the control of primary legislative powers granted to territorial entities, while their concept of administrative autonomy describes the control of individual territorial units over policy implementation. Fiscal autonomy comprises responsibility for revenue generation and assignment (Blöchliger and King 2006).
Analyses have shown that internationalization processes in education policy led to demands for modifications of both HE policies and polities (e.g., Martens et al. 2010). Faced with complex domestic decision-making structures, reform-oriented national policy-makers were able to garner support and legitimacy for the desired national modifications by referring to international HE agreements (Martens and Wolf 2006). It is against this backdrop that Germany's leading role in launching the Bologna Process can be understood. While interlocking federalist structures hampered domestic reform endeavors, the federal government sought to increase exogenous pressures to gain additional leverage to overcome the "federalist self-blockade" (Toens 2009;Niemann 2010). This Europeanization strategy of executive multilateralism served to increase external reform and legitimacy pressures over the sixteen L€ ander education systems and individual institutions.
Beyond these initial strategic considerations, the Bologna Process quickly established itself as the collective European answer to an array of interrelated challenges: the global knowledge economy, stagnant economic growth, the lacking attractiveness of European universities, and demographic changes. The intensified transnational communication inherent in the process has pressured national policy-makers to legitimize pre-existing HE institutions and contemporary reforms in a competitive institutional environment. This is facilitated by the benchmarking of jointly defined objectives and indicators. Studies have documented how institutional isomorphism (DiMaggio and Powell 1991), i.e., the alignment with policy models perceived as successful, has significantly impacted national HE policies (Dobbins and Knill 2014;Bieber 2016).
While the reform impetus to alter national HE decision-making structures to better respond to international challenges was identical, the domestic adaptation and the actual outcome differed. Hence, it is imperative to assess how national idiosyncrasies triggered different reactions. Drawing on historical institutionalism, which stresses the decisive character of the state's institutional organization in influencing collective action (Hall and Taylor 1996;Pierson and Skocpol 2002),w e argue that the reform impetus and the resulting challenges may consequently be perceived, evaluated, and implemented differently based on national institutions. In short, domestic institutions moderate external influences.
This explanatory framework is useful precisely with respect to the non-intended consequences of the internationalization of HE-and differing outcomes in Germany and Switzerland. While the Bologna Process provided a strong impetus for reforming national HE, its implementation was not explicitly defined: When there is a certain degree of leeway, national institutions may mediate external influences by shaping policy alternatives as well as actors' preferences and decisionmaking behavior (Steinmo 2008).
Federalist institutions also shape how domestic stakeholders, particularly parties, cope with (or utilize) international reform pressures and advance their interests in legislative arenas. The interests of sub-national party organizations may thereby differ from region to region and from national-level partisan positions (Stecker 2015). When making decisions, German or Swiss national parties have to take the L€ ander/cantonal party level into account as it may feature different preferences despite a common ideological background (B€ ack et al. 2016). If different regional political interests exist, or if subnational partisan government coalitions do not reflect the federal government coalition, the probability for a Land/cantonal government to veto a federal bill may increase (Debus 2008, 510). Consequently, in federal education systems the reform process may be particularly complex not only 5 due to their multi-level structures per se, but also heterogeneous partisan actors and preferences operating within them.

Obstacles to Reforming Federally Organized Education Systems
Almost all federally organized education systems exhibit features of shared rule and institutionalized collaboration that shape how policy-making is structured between different territorial levels. Such was partially the case in Switzerland and even more so in Germany before Bologna. The supremacy of the L€ ander in education was (almost) always a central feature of German HE polity since the foundation of the German Empire in 1871 (Führ 1997). Furthermore, in the post-war period, the allied powers enforced the "re-decentralization" of (West) German HE. Fearing a centralized state as well as the politicization and bureaucratization of education, they did not create a national educational administration (Dobbins and Knill 2014). The decentralized organization of HE is consequently firmly anchored in the German constitution, the Basic Law (Article 70; Article 30): the L€ ander are formally entrusted with authority over financial, structural, curricular, and personnel matters. Nevertheless, the federal government increasingly gained authority over the formal HE system by the mid-1950s. These centralization trends were fostered by the perceived technological lag vis-a-vis the Communist Bloc (Krücken 2005), increasing middle-class educational aspirations, and a growing consensus that German education was too elitist and rigorously structured (Teichler 2005).
The expansion of tertiary education and establishment of new universities at that time, however, required the financial support of the Bund. Although it already became involved earlier, in 1969 the first SPD/CDU grand coalition agreed on amending the constitution and redefining the joint responsibilities of the federal government and L€ ander. Eventually, the "great fiscal reform" in 1969 enabled the federal government to engage in joint education planning (Kaiser and Vogel 2017;Döring and Schnellenbach 2011). Centralization was seen here as an appropriate means of facing national and international HE challenges: After years of wrangling between the L€ ander and federal government, the Higher Education Framework Act (Hochschulrahmengesetz, HRG) of 1976 empowered the federal government to pass overarching HE legislation (Giesecke 2012). The HRG became the cornerstone of a relatively uniform, nation-wide HE policy, as it specified the legal status and responsibilities of HE institutions and, most notably, introduced standardized student selection and allocation procedures (Künzel 1982).
Despite increased federal authority, the constitutional responsibility for education remained with the L€ ander, which maintained autonomy over funding, long-term planning, and personnel appointments. Nevertheless, the federal government succeeded in becoming increasingly involved in the policy area in which it originally had no constitutional authority (Erk 2003, 308), resulting in complex intertwined powers between both levels. Ultimately, the interlocking competences led to what Scharpf (1988) described as the "joint-decision trap", in which the two decision-making levels blocked each other due to contending preferences and ideological differences.
This interlocking structure was also reflected in the intertwined legislative processes. First, with the reform of 1969 the federal government gained more competences in framework legislation (Rahmengesetzgebung Article 75 Basic Law, prior to the federalism reform of 2006). It enabled the federal government to set key general aspects, while the individual L€ ander were tasked with detailed legislation. Second, concurrent legislative processes (konkurrierende Gesetzgebung Article 72 and 74 Basic Law, prior to the federalism reform) were established and reinforced by rulings of the federal constitutional court in areas where the federal government gained legislative authority although, originally, the L€ ander were exclusively responsible (Burkhart et al. 2008, 526). Subsequently, through framework legislation and concurrent legislative processes, the approval of the Bundesrat (the representative body of the L€ ander in federal decision-making) was required for the majority of laws. This interlocked decision-making structure where vertical cooperation between the Bund and the L€ ander was mandatory made Germany a typical case of cooperative federalism (at least before the federalism reform) (Döring and Schnellenbach 2011;Griessen and Braun 2010). However, the vertical intertwinement of decision-making structures made it difficult for Germany to sufficiently react to socio-economic challenges and internationalization/Europeanization processes Stecker 2015).
By comparison, Swiss HE polity was far less vertically interwoven. Since the foundation of the Helvetic Confederation (1848) education policy was the responsibility of the cantons. Strong decentralization was historically justified by the subsidiarity principle defined in the Constitution (Article 3): public matters are to be administered by the smallest and most decentralized possible political entity. Bieber (2012) argues accordingly that Switzerland previously pursued a "maximalist version" of educational federalism. If the lower levels are unable to effectively tackle the concerned issue, cantonal or national authorities may intervene. Subsequently, the cantons largely assumed responsibility for HE. Although this posed challenges regarding inner-Swiss mobility and increased demands for centralization as early as the 19th century, historically entrenched principles of cantonal sovereignty and direct democracy enabled the cantonal or local authorities to exert their veto rights in referendums (Bieber 2012). 3 Nevertheless, the Constitution gave the federation the option to establish technical universities, resulting in the Eidgenössische Technische Hochschule Zürich and Ecole Polytechnique F ed erale Lausanne. Yet, Switzerland still never operated a federal education ministry, making HE historically one of the most decentralized polities in its "decentralized federalism" (Braun 2003, 58 et. seq.).
However, in 1968 the Federal Assembly (Bundesversammlung) passed the first Federal Law on HE Promotion, which authorized the federation to allocate financial contributions and coordinate the gradual creation of a Swiss HE area. Moreover, the law mandated the federal government to foster cooperation between HE institutions and direct funding accordingly (Hochschulförderungsgesetz Article 1; Bieber 2016). Yet, further centralization attempts were again voted down in referendums in 1973 and 1978 4 with the effect that federal involvement was limited to pragmatic financial support based on the 1968 law. Subsequently, horizontal intercantonal coordination was prioritized over vertical coordination, as the federation essentially had no other legislative autonomy (Griessen and Braun 2010).
Regarding funding though, Swiss HE policy exhibits a distinct tradition of vertical and horizontal burden-sharing. The cantons contribute approximately 50 percent to university budgets, while the federal government provides approximately 15 percent to their basic budget and additional research funding (Simoleit 2016). An additional share (approximately 10 percent) is allocated to the universities as part of the so-called inter-cantonal financial equalization scheme: since the 1980 s the cantons have paid fixed amounts for students previously residing within their boundaries to cantons in which they pursue their studies. For example, if a student from Lucerne attends the University of Bern, the Canton of Lucerne transfers a share of the study costs to the Canton of Bern. This scheme reinforces horizontal inter-cantonal cooperation, which is further supported by numerous university cooperation agreements and diverse networks of national, (inter-)cantonal or joint federal-canton committees (Schmidt 2008, 121).
Altogether, the cantons have largely exercised legislative, funding and administrative autonomy. Thus, Swiss HE historically avoided the joint-decision trap inherent in German federalism. This was further facilitated by the fact that only qualified majorities (generally two-thirds) among the cantons are required for HE-related decisions at the federal level, whereas unanimity is required in Germany. Subsequently, Swiss HE federalism was less interlocked and more horizontally coordinated (Criblez 2008) than the German system. Taking the characteristics of Swiss and German HE federalism together, we follow Braun and Griessen's (2010) assessment that Germany was previously unable to adequately respond to burgeoning challenges in HE because it was overly vertically intertwined and blocking institutions prevented decision-making, while Swiss HE lacked cooperative vertical structures for joint responses. Hence, neither country had sufficient means to address HE challenges requiring nationwide action.
In the 2000s, the polity-related reforms in Germany and Switzerland took a markedly different trajectory with Switzerland moving towards greater centralization, while Germany manifestly decentralized HE to the L€ ander. We show how preexisting historical institutions of educational federalism created differing incentives for partisan actors to respectively centralize or decentralize the HE polity.

Germany: rediscovering Decentralized Higher Education
After encompassing reforms in the 1960 s and 1970 s resulting in overlapping Bund and L€ ander governance structures and the expansion of traditional universities and universities of applied sciences (Fachhochschulen), the (West) German HE system initially experienced a period of consolidation (Teichler 2005) within pre-existing policy and polity paradigms. 5 Despite a wave of management-oriented reforms in other North-Western European countries (Schimank and Lange 2009), the German (and Swiss) HE system remained rooted in the Humboldtian tradition, which stresses the unity of research and teaching as well as academic self-administration, largely without market-like mechanisms such as entrepreneurial leadership, systematized quality assurance, and highly competitive funding (Clark 1983; for Switzerland see Horvath et al. 2000).
However, German HE was plagued by high drop-out rates, long study duration, and insufficient state funding well into the 1990 s, whereby modernization efforts failed due to the above-mentioned joint-decision trap. On the one hand, the specific allocation of authority between the Bund and L€ ander impeded the Bund from initiating reforms without the consent of the L€ ander. The latter, on the other hand, frequently held contrasting reform ideas and interests and were unable to collectively aggregate and "upload" their preferences to the central level. Consequently, any systematic responses to the above-mentioned challenges faced insurmountable institutional hurdles. Unlike the Swiss cantons, all L€ ander host several HE institutions which makes coordinating HE policy a complex bargaining matter.

Bologna as a Stimulus for Policy and Polity Change
Amid increasing globalization and transnational interdependencies, fears of lagging behind in the global race for educational excellence loomed high over German HE (Wissenschaftsrat 2000;Toens 2009). By the mid-1990s, it had become clear for policy-makers that a far-reaching reform was essential to remain internationally competitive (BMBF 2008, HRK 2012. The Bologna Process was seen as a suitable remedy for overcoming the existing reform obstacles. Despite the relative misfit between the Humboldtian academic self-rule tradition and Bologna's policy Analyzing the Reorganization of German and Swiss Higher Education recommendations largely based on Anglo-American HE practice (Dobbins and Knill 2014), the European-wide initiative unleashed a veritable reform fervor in German HE. Most apparently, pre-existing study structures were adjusted for most study courses to the Bachelor and Master models, the European Credit Transfer System (ECTS) and the diploma supplement scheme (Witte 2006). Besides direct HE policy reforms to increase international compatibility, governance structures were ultimately also altered and Germany has broadly aligned itself with policy guidelines promoted through the Bologna Process. Following the "less government and more governance" logic (de Boer et al. 2008, 21), both the L€ ander and federal government increasingly promoted steering arrangements which facilitate targetoriented governance, the incorporation of external stakeholders, inter-university competition, and quality assurance.
Beyond the general thrust towards market-oriented governance, Bologna also prompted countries to reassess the efficiency of pre-existing regulatory HE arrangements. It is in this context that the transfer of nearly all authority over HE to the L€ ander occurred. Policy and polity changes between 1949 and 2005 in Germany have primarily been in the direction of centralization, whereby globalization and Europeanization have compounded this trend (Kaiser and Vogel 2017). Specifically, the Social Democratic-Green federal government elected in 1998 aimed to expand the education policy authority of the Bund. For example, it introduced top-down regulations notably using the HRG framework regulation with financial incentives for universities to establish "junior professorships". However, the political resistance of some Christian-Democratic (and CSU) governed L€ ander, in particular, as well as two decisions of the German Constitutional Court in 2004 and 2005 set boundaries to these endeavors (Pasternack 2011). Subsequently, stronger federal regulation was impossible without fundamental constitutional amendments and the creeping centralization of German HE came to a halt.
Instead of reaching consensus on centralization, an encompassing reform of German federalism returned full autonomy over education policy to the sixteen L€ ander while the federal government withdrew as a decisive actor in this policy field. Under the CDU/CSU-SPD grand coalition, the Federalism Reform of 2006 and 2009 was (quantitatively) the most extensive reform of German Basic Law since 1949 and aimed to overcome the country's preexisting interlocked decisionmaking structures. 6 The main objective was to disentangle overlaps in territorial authority and clearly define responsibilities of the federal government and L€ ander, in order to enhance their capacity for action (Scharpf 2006;Burkhart 2009). Specifically, the L€ ander parliaments were to be given more distinct legislative competences, while the legislative clout of the Bundesrat was to be reduced. Education policy became a focus of the Federalism Reform because the shared powers of the federal government and L€ ander had visibly posed a reform obstacle and because in other policy areas, like financial issues (so called high politics), no compromises were initially reached in negotiations.
The Federalism Reform comprised two substantial changes in legislation that substantially impacted the HE polity. First, "framework legislation" was abrogated and legislative competencies were either delegated to the federal level or L€ ander (Pasternack 2011). In HE, the L€ ander gained full legislative authority while the joint task of education planning was abolished and the Bund no longer has any legislative competencies. Additionally, the HRG, which was to assure the uniformity of the sector, is no longer applied. Second, "concurrent legislation" was modified and legal matters were "assigned exclusively to either the Land or the Federation" and in areas where the Bund is entitled to pass legislation, e.g. university admissions and degrees, the L€ ander could enact deviating laws (Burkhart et al. 2008, 526).
The constitutional modifications also resulted in the dismantling of institutions for joint educational planning (Niemann 2010). Most prominently, the Bund-L€ ander Commission for Educational Planning and Research Promotion (BLK), which was the main coordinative body for HE matters, was replaced in 2008 by the Joint Science Conference (GWK). While the BLK dealt with all aspects of HE policy, including financing, the GWK limits itself to research promotion and general strategies (Auel 2014). The BLK's organizational structure was prototypical of the problems of German federalism: Consisting of representatives of all L€ ander plus the Bund, partisan conflicts were constantly present in the BLK and prevented swift decision-making (Griessen and Braun 2010). Thus, the pendulum of German HE policy clearly has shifted towards the L€ ander to the extent that there are currently no institutionalized forms of cooperation on equal footing.
Decentralization also affected new quality assurance mechanisms. With increasing internationalization in mind, the Standing Conference of the Ministers of Education and Cultural Affairs (KMK) and the German Rectors' Conference (HRK), two of the most important institutions for horizontal HE coordination, drew on developments in other countries and identified ex ante study program accreditation as the "gold standard" for sustaining the competitiveness of German HE (Witte 2006, 171). Hence, in 1998 a national Accreditation Council (AC) was established to secure minimal standards and labor market relevance of study programs across all L€ ander. 7 Despite its central position in the new quality assurance networks, the AC can be viewed as the outcome of a power struggle between the Bund and L€ ander, in which the L€ ander prevailed (Serrano-Velarde 2008). The AC has no authority to accredit study programs, but rather is at the center of a dual structure, in which it merely accredits decentral accreditation agencies, which in turn conduct discipline-specific evaluations of study programs (Kehm 2007, 89). The L€ ander thus succeeded in stripping the federal government of any authority over quality assurance, resulting in a patchwork of decentrally operating agencies.
Additionally, the territorial decentralization gave the individual L€ ander various opportunities to introduce administrative decentralization by redefining the governance relationship between the Land's administration and the universities. In this regard and also inspired by developments spurred by the Bologna Process (BMBF 2012), the L€ ander relinquished varying degrees of autonomy to their HE institutions. Some went further than others with their reform efforts: Whereas, for instance, North Rhine-Westphalia gave universities extensive self-regulation capacities, the Bavarian government has largely maintained tighter control over universities and concludes detailed multi-annual target agreements with them (Lanzendorf and Pasternack 2008). Differences between the L€ ander also emerged regarding tuition fees. A ruling of the Constitutional Court in 2005 generally permitted the imposition of study tuition. While several L€ ander led by conservative parties quickly introduced fees (max. e500 per semester) and justified them by also referring to international competition (Bavarian Science Ministry 2012), socialdemocratic governments were far more reluctant since they feared that socioeconomically disadvantaged groups would be excluded from tertiary university education (Science Ministry NRW 2012). This illustrates that different HE ideas of center-right and center-left parties on distributional issues endure, making cooperative decision-making for the whole German HE sector challenging.
Competition between increasingly autonomous universities was also promoted as a crucial reform component of territorial and administrative decentralization. The federal government thereby regained some degree of authority as a co-financer of (temporally limited) HE-related projects, albeit only in close coordination with all L€ ander. Prominent examples are the Excellence Initiative (Hartmann 2006) and Higher Education Pact 2020, through which the federal government makes very substantial financial resources available. The Excellence Initiative awards extra funding to selected universities. As a cooperative Bund-L€ ander-project, it is reassessed every four to seven years and new funding parameters are regularly renegotiated. Consequently, institutional differentiation was promoted and subsidized by the state.
Taken together, the Bologna reforms and Federalism Reform constituted a parallel trend in times of increasing internationalization (Toens 2009, 254). Precisely, such external developments made German decision-makers aware that adaptations of HE polity were necessary to better cope with global competition and that the existing interlocked decision-making structures posed a significant obstacle to the adaptability of the system (BMBF 2012, HRK 2012, KMK 2012. In particular, the federalism reform modified the German HE polity (territorial decentralization), while the L€ ander often passed on some regulatory autonomy to the universities themselves (administrative decentralization).
The Bologna Process conveniently played into the hands of the L€ ander who were seeking to regain control over the HE sector (Pasternack 2011). Just as the federal government had (co-)initiated the Bologna Process to overcome obstacleprone decision-making structures and circumvent the individual interests of involved actors (Martens and Wolf 2006), the L€ ander also viewed the Bologna Process as a vehicle for indirectly asserting their sovereignty over educational policy. In these newly established intergovernmental-national-subnational layers, the L€ ander were in a strong position because the capacity to reform the system depended decisively on their authority.
Explaining the "re-decentralization" of the German HE polity Why did Germany reverse its path towards centralization and "re-decentralize" HE? The previous system with its two interlocked decision-making levels and forced cooperation between the L€ ander made encompassing joint HE reform undertakings extremely challenging (KMK 2012). Simultaneously, growing reform pressures made comprehensive reforms almost mandatory. Consequently, the interlocked decision-making had to be disentangled in order to facilitate reforms. From the two options, centralization or decentralization, only the latter was viable under the German institutional setting. Against the background of reform imperatives posed by Internationalization/Europeanization and the Federalism Reform the L€ ander were in a stronger position for negotiating their preferred outcome and safeguarded their constitutionally guaranteed authority. Based on these considerations, at least three complementary historical-institutional factors explain why Germany ultimately opted for decentralization.
First, opposing party preferences at the federal and the sub-state level fueled by asymmetries between the L€ ander hampered the introduction of necessary reforms. In this regard, two relevant lines of conflict in German federalism have to be considered when analyzing reform projects and path dependencies: a horizontal conflict of interest between the individual L€ ander and a vertical conflict between the federal government and L€ ander in cases where diverging interests on both levels come into play (e.g., regarding financial equalization) (Br€ auninger and Ganghof 2005, 156). Although Germany with its culturally homogenous society was characterized as a classic example of symmetric federalism (Auel 2014, 424), recent developments pushed the country in a more asymmetric direction. Against the backdrop of reunification in 1990, German federalism became more asymmetrical since the already prevailing economic and fiscal heterogeneity of the L€ ander was reinforced and challenged the norm of unitary living conditions. Moreover, divided government became more common since the regional party systems increasingly fragmented and cleavages between financially and economically weaker and stronger L€ ander deepened (Kaiser and Vogel 2017).

13
The relationship between parties on both levels is characterized by both competition and cooperation. Overall, parties at the L€ ander level are considered "fully functional political parties with an intent to shape all areas of public policy" (Stecker 2015(Stecker , 1309 and, hence, are held accountable by their constituencies for HE policies. Long-standing political cultures in German regions influence party competition and ideological differences between parties at the L€ ander and national level since parties sometimes adopt "tactical positions to recommend themselves as potential coalition partners" (Debus 2008, 535). However, within the political structure of cooperative federalism, Germany's party system is also territorially integrated across the L€ ander and, hence, party politics prevent the L€ ander governments from introducing excessively competitive elements (Benz 2007, 423).
Ideological differences between the center-left and center-right political camps regarding HE prevailed but, generally, party ideologies did not have a substantial impact on preferences for centralization or decentralization (Kaiser and Vogel 2017, 21) and " [t]here is no clear preference for decentralized policy-making associated with any party orientation" (Döring and Schnellenbach 2011, 94). A more important reform obstacle was the decreasing willingness of the wealthier L€ ander to financially subsidize the educational expenditures of poorer L€ ander by means of the German fiscal equalization scheme (L€ anderfinanzausgleich) in the case of centralization (Kaiser and Vogel 2017). The fact that the liberal FDP (in strong contrast to its Swiss counterpart) was a main proponent of decentralization and the conservative sister parties CDU/CSU also pushed for more decentralization (Pasternack 2011) was not so much ideologically motivated. Instead fiscal rationales were essential. Since the CDU/CSU and FDP tended to govern the wealthier L€ ander in the South, they pushed for more HE decentralization. At the same time, smaller and poorer L€ ander, often ruled by social-democratic-led government coalitions, were not interested in co-financing HE-related projects of the federal government, but were willing to trade in some decision-making authority for the provision of additional federal fiscal resources (Kaiser and Vogel 2017).
It became apparent that L€ ander were mainly concerned with issues of HE competitiveness (KMK 2012). While prior to the Federalism Reform all L€ ander provided almost the same framework for their HE institutions (e.g., salary of university professors, workload) under the auspices of a German-wide HE policy coordination through the HRG, the HE-related amendments to the German constitution compelled the L€ ander to compete for the best personnel, students, and infrastructure by establishing different legal regulations. While some wealthier L€ ander supported the idea of competitive federalism (Ziblatt 2002), the actual elements for enabling competition between the L€ ander were only marginally implemented in HE during the Federalism Reform. For instance, the remuneration of civil servants, e.g., university professors, is a Land matter. However, no significantly differing policies were implemented. 8 Hence, Germany partially shifted from cooperative federalism to competitive federalism in HE (Toens 2009).
Second, the original institutions of the decentralized education system were never legally revoked: the pathway of decentralization was still intact and the distribution of governmental power was generally immune to fundamental changes (Lehmbruch 2002). Instead, the shift toward centralization was a process of incremental change that eventually added a layer to the existing decentralized governance system by strengthening the federal government's competences (see Streeck and Thelen 2005). Privileging one layer over the other is possible as long as both are operative. The complete shift towards institutionalizing centralization would have required a fundamental amendment of the German constitution and the L€ ander to give up their authority over education-which they have always rejected. The L€ ander insisted on their exclusive authority over education matters during the negotiations over the Federalism Reform (Scharpf 2009, 101-102). Rescinding the HE competences of the Bund, in contrast, was less problematic taking the prevailing historically anchored institutional configuration into account. In other words, the pathway towards centralization could easily be reversed since it was just an additional layer to the existing decentral legislative pattern that was not fully institutionalized.
Third, the federal government had no opportunity to expand its limited competences in HE and did not vehemently defend the status-quo because the joint task of education planning was considered largely ineffective (Scharpf 2009, 104) and other opportunities emerged for the Bund to gain influence in HE (BMBF 2012). With the Excellence Initiative on the horizon, launched in 2009, the federal government regained influence and the close cooperation with the L€ ander was maintained. Because of limited funding capacities of the L€ ander, the Bund was needed as the main financer providing 75 percent of the funding (while the university Land has to cover 25 percent). The introduction of this selective HE funding program established new "robust links between federal and state authorities in the financing and organization of German higher education" (Burkhart et al 2008, 537) and legislative decentralization was replaced by new forms of executive centralization under the new constitutional provisions. Although the Excellence Initiative is not a permanent institution and must be renewed every few years, it nevertheless established a new policy pathway which is hard to reverse since the extra funding for HE institutions is required to make the German university sector internationally competitive (BMBF 2012). In other words, the Excellence Initiative can be seen as "backdoor centralization" as the Bund traded in legislative competences for financial influence.
Although HE modernization was frequently discussed, Swiss HE policy was characterized by significant reform backlog. Like Germany, Swiss HE was plagued by long study durations, poor student-teacher ratios, the absence of quality assurance structures and rapidly increasing student numbers. The 1990s indeed heralded a wave of policy reforms. One important modification was the 1995 Law on Universities of Applied Sciences, which defined the applied sciences universities as a new type of HE institution. Amid HE expansion, this had the effect that teacher training and higher vocational training were no longer carried out in specialized HE institutions, i.e., polytechnics or schools of commerce or public administration, rather at pedagogical and applied sciences universities. This significantly enhanced the status of these institutional types to the extent that they hold an equivalent institutional status to universities (Bieber 2016). Presently there are ten "traditional" universities in Switzerland, seven of which are large, multidisciplinary institutions (Basel, Bern, Fribourg, Geneva, Lausanne, Neuchâtel, and Zurich) and three of which are smaller, more specialized institutions (St Gallen, Lucerne, and Lugano) (Perellon 2001). The Swiss cantons now also operate eight applied science institutions (Fachhochschulen) 9 and seventeen pedagogical HE institutions (Swissuniversities 2016). However, only the host cantons of the universities actually hold the status of "university cantons" 10 , which means that (de-)centralization directly affects less than half of cantons.

Bologna as a stimulus for policy change
Due to its reserved stance towards European integration, Switzerland was an unlikely candidate for policy change stimulated by a voluntary, transnational process. Yet, the perceived lack of inner-Swiss compatibility and mobility as well as increasing global university competition (Aeberli and Sporn 2004) prompted Switzerland to increasingly engage with the Bologna platform. In fact, various authors have labelled Switzerland a Bologna "poster child" (Criblez 2008, 281;Bieber 2016) due to its swift implementation of Bologna-promoted policies. Besides its European dimension, the multitude of associational, political, and academic actors involved in the process lent it additional legitimacy and enabled Swiss policy-makers to put reforms in a broader transnational context. Switzerland's voluntary involvement in numerous HE-related epistemic communities and policy networks (e.g., regarding quality assurance) facilitated the joint development of guidelines and benchmarking standards (Osterwalder and Weber 2004;SBF 2012). Particularly decisive in this respect was precisely the non-EUcharacter of Bologna: it was not perceived as an overzealous EU intervention into internal affairs. This increased its acceptance in the non-EU member and policymakers' willingness to engage in lesson-drawing (Bieber 2016).
Subsequently, a domestic reform coalition consisting of public administration bodies, most notably, the Swiss State Secretariat for Education and Research (SBF), as well as universities, and university organizations such as the Swiss Rectors' Conference (CRUS) increasingly promoted Bologna as an internationalization process from which Switzerland should not isolate itself (CRUS 2012;Osterwalder and Weber 2004, 25-26;Benninghoff and Leresche 2009, 200). Like in Germany, Bologna functioned as a lever for policy change and alignment with externally promoted models. In terms of legislation, the direct Bologna influence over policy was unmistakable: the Swiss University Conference (SUK) passed its so-called "Bologna Directives" (SUK 2003) in 2003, which saw for the Swiss-wide introduction of Bologna-style study structures for all HE institutions by 2010 as well as the ECTS and diploma supplement.
Moreover, Bologna added a further impetus to other policy-related developments already on the horizon. Like in Germany, Switzerland shifted somewhat away from its traditional Humboldtian governance model towards what Lepori and Fumasoli (2010) define as "controlled competition", i.e. a balance between competition and (interuniversity and state-university) cooperation. This was complemented by more competitive institutional funding, enhanced university autonomy and a focus on institutional strategies (ibid. 2010). Greater entrepreneurial leeway for universities was counterbalanced by a new quality assurance and accreditation system following Bologna guidelines. Regarding policy, Switzerlandand Germany-thus experienced moderate convergence towards market-oriented structures.

Centralization as a Bologna Side-Effect
Yet, Switzerland also seized on Bologna to reconfigure its HE polity. Although the cantons still hold a substantial degree of authority, essentially every recent polity modification moved towards centralization. In the early 2000s, there was a widespread conviction among Swiss policy-makers that Bologna reforms require national regimentation among heterogeneous actors to ensure unitary and coordinated implementation (Trampusch and Busemeyer 2010, 605). The Federation thus quickly drew on international developments to legitimize the alteration of power relations towards the federation, which in turn enabled Swiss HE to overcome pre-existing veto structures and dissolve the reform backlog (Benninghoff and Leresche 2009, 200).
This resulted in an extensive structural reorganization. The Federal Law on Financial Aid to Universities (UFG, 1999(UFG, ), constitutional reform (2006, Law on Universities of Applied Sciences (1999) and Federal Act on Funding and Coordination (HFKG, 2009) led to a strong shift of decision-making powers to the federal level. For example, the UFG modified the mode of governance to the extent that new cooperative structures between the federal government and cantons were introduced in this area which was previously dominated by the cantons (Bundesversammlung 1999). Specifically, the law turned the preexisting Swiss Higher Education Conference into the new Swiss University Conference and introduced the Swiss Agency of Accreditation and Quality Assurance (AQQ). To promote vertical cooperation, the Swiss University Conference functions as a joint body of the federal government and the university cantons (Bundesrat und Regierungen der Universit€ atskantone 2000). To this end, the federal government and cantons delegated several formal powers to the Swiss University Conference (Benninghoff and Leresche 2009). Responsibilities for the implementation of HE reforms were centralized, whereby the directives of the Swiss University Conference became mandatory for all universities. The federal government and cantons granted the Swiss University Conference the necessary financial resources to oversee the process, which in turn assured the binding character and effective implementation of the Bologna guidelines (Griessen and Braun 2010).
This shift in administrative autonomy to the federation expanded the institutional foundations for cooperative federalism, i.e., institutionalized cooperation between the federal government and cantons (Benninghoff and Leresche 2009). Despite the preexisting basis for intercantonal cooperation, this constituted a novelty in Swiss HE, not only because the federal government acquired extensive authority to intervene in issues related to HE administration, but in particular because the cantons were willing to relinquish some of their power. This development continued into the next decade. In 2000, the responsibility for implementing the Bologna recommendations was transferred to centralized intermediary organizations, specifically the three conferences of university rectors (CRUS), applied sciences universities (KFH) and pedagogical colleges (COHEP) (Bundesrat 2009;Lepori and Fumasoli 2010, 811). This was a decisive reform step because the federal government previously did not have the authority to intervene into cantonal educational policies despite its limited participation in institutional funding (Benninghoff and Leresche 2009, 208).
To facilitate the introduction of the Bologna system, the 1995 Law on Universities of Applied Sciences was revised. The new law effective in 2005 put all applied sciences universities under the supervision of the federal government (Bundesversammlung 2003). To enhance the national and international comparability of degrees, the law created legal foundations for the introduction of Bachelor and Master study programs: applied sciences universities were not required to offer Bachelor programs and authorized to introduce Master programs subject to the quality assurance and accreditation procedures of the federal government.
Against this background, the complete revision of the constitutional articles pertaining to education in 2006 can be seen as a further turning point towards centralization. This polity change was explicitly motivated by concerns over the international competitiveness of Swiss HE and research (Bundesrat 2007). Authority over quality assurance and the permeability of study programs was transferred to the federal government and cantons, which were called on to jointly to create a uniform Swiss HE. This required the creation of common institutions to promote the joint coordination of activities and to ensure the equivalence of general and vocational education (Bundesversammlung 2006). The constitutional reform gave the federal government subsidiary regulatory authority in previous domains of cantonal authority, and thus facilitated the partial harmonization of the cantonal HE systems.
Specifically, the new constitutional article 63a requires the federal government and cantons to co-regulate the HE system, to guarantee its quality based on contractual agreements and-in strong contrast to Germany-to relinquish authority to joint governing bodies. Unlike compulsory education, where the cantons still maintain authority over primary schooling, the constitutional modifications created the necessary preconditions for the simplified, centralistic governance of the entire Swiss HE area, transparent result-oriented funding as well as for strategic planning and a more effective distribution of tasks between HE institutions.
As a further sign of centralized legislative autonomy, the HE Aid and Coordination Law (HFKG) was drafted in 2009 and came into force in 2014 (Bundesrat 2009). Its main intention was to develop a simplified HE governance processes as defined in constitutional Article 63a, which enables the uniform regulation of the entire HE sector. In 2012, the law replaced two different federal decrees of the HFKG and Applied Sciences University Law (aFHSG), thus becoming the sole legal foundation of the federal government to support traditional and applied sciences universities at the cantonal level and co-regulate the entire Swiss HE area with the cantons.
An additional significant modification regarding administrative and fiscal autonomy (see Dardanelli and Mueller 2017) was the merger of the former rector conferences CRUS, KFH and COHEP into swissuniversities. Unlike Germany, the merger-along with the HFKG-fostered the centralization of HE governance by defining conditions for the allocation of federal funds to universities and applied science universities. Furthermore, principles and procedures for the central planning of Swiss HE policy and quality assurance were defined. In contrast to Germany, the law also defines joint aims and joint governance bodies. While up to now only applied sciences universities had to be accredited, the law prescribes all HE institutions and universities an obligatory centralized accreditation procedure.

Explaining the Centralization of Swiss HE Polity
Although the cantons have by no means entirely relinquished HE policy to the federal government and are still heavily involved in joint coordination and planning, the reforms reflect a marked trend towards the centralization and harmonization of Swiss HE. Why were the cantons willing to abandon or share a significant degree of their legislative, fiscal, and administrative autonomy over HE policy? First, one could make the trivial argument that political centralization is easier in smaller countries than in geographically larger countries. This dovetailed with functional considerations: there was a pronounced desire in Swiss policymaking circles for the small, but heterogeneous country to operate with a uniform HE system in this increasingly internationalized policy area (Bieber 2016). Hence, policy-makers embraced a "kill two birds with one stone" approach: centralization not only served to externally present the Swiss HE system effectively and uniformly, but also to ensure greater inner-Swiss mobility, transparency and coordination (Bundesrat 2007). The shift towards centralization and shared governance was further propelled by increasing demands for inter-university cooperation, the introduction of new (multilingual) study courses at several geographically proximate universities (e.g., Bern, Neuchâtel, Fribourg) as well as increased cantonal efforts to jointly coordinate study programs for efficiency reasons.
Yet, the configurations of Swiss federalism and partisan politics within them were crucial in explaining the centralization trend. As hinted above, the decentralization of HE and the abolishment of central coordination bodies would not only have endangered the flow of funds from non-university cantons to university cantons, but also the relatively high federal funds for HE. In other words, there were few reasons for wealthy university cantons such as Zurich, Bern and Geneva to reclaim sole authority over HE, because they rely on financial subsidies from both the non-university cantons and the federal government. The (less populous) non-university cantons also feared an excessive re-distribution of funding to their disadvantage. However, due their size and generally weaker financial clout, they would be powerless in negotiations with the larger university cantons, which are well organized through the so-called Conference of Donor Cantons (Konferenz der NFA-Geberkantone). Along these lines, the central government has historically played the role of ensuring a more or less level financial playing field between the cantons. Cantons who perceive their financial burden as excessive direct their complaints towards the central government-often with great success (Basler Zeitung 2013). Heavier investments in education by one canton are often compensated for by funding grants for other areas (e.g. infrastructure, transportation) (Keller 2014). Therefore, in view of HE expansion, all cantons became increasingly reliant on institutionalized centralized structures to ensure an equitable redistribution of funding and had little incentive to withdraw from the existing scheme. In other words, the interlocked funding system enabled further centralization, as fiscal decentralization would have led to a "free-for-all" over funding redistribution. Hence, centralization functioned to some degree as a buffer against opportunism (see Braun 2009). 11 Unlike in Germany, partisan constellations within the confederation also did not pose a major obstacle to centralization. Switzerland exhibits a very unique federal executive branch. The federal executive consists of a seven-member council (Bundesrat), in which the three strongest parties in the National Council (Nationalrat) and Council of States (St€ anderat) hold two seats and the fourth strongest party one seat. Up to 2008, this was the Social Democratic Party (SP), Liberal Party (FDP) and Christian Democrats (two seats each), with the Swiss People's Party (SVP) holding one seat. After a strong electoral turnout, the SVP was offered two seats to the detriment of the CVP in 2008. This consociational arrangement forces the government to seek consensus on all policy matters. This consensus orientation is further facilitated by the fact that diverse, yet relatively stable center-right or center-left coalitions have held electoral majorities in many cantons for decades. These heterogeneous multiparty constellations serve to institutionalize partisan cooperation at the federal level and between cantons, as Swiss parties-unlike their German counterparts-are less fearful of the dominance of their partisan rivals at the federal level.
Contrary to the German case, where partisan actors feared that their rivals would "upload" their preferred HE policies to the federal level, the divide over HE policy and polity was much less pronounced in Switzerland. This is reflected, in particular, in the pro-centralization stance of the Swiss FDP, which holds political majorities in several (primarily French and Italian-speaking cantons). Generally affiliated with economic liberalism, the FDP-in strong contrast to its German counterpart-essentially pushed for HE centralization for "social-democratic" reasons. Specifically, it argued that federal harmonization is imperative to prevent unhealthy competition between the HE institutions and cantons due to different tuition fees and study conditions (Interpellation, Bundesversammlung, Randegger 2004). Like their German counterpart, the Swiss Social Democrats also advocated greater federal involvement in HE funding and coordination. Besides their shared position with the FDP that centralization would promote equality, the SP advanced the argument that if the federation negotiates transnational agreements demanding structural modifications from the cantonal universities, it should bear a significant share of the financial burden (Interpellation Bundesversammlung, Plattner 2003).
While the CVP generally supported centralization with the argument that it would facilitate transnational HE coordination, the SVP's stance was more complicated. A vehement opponent of the centralization of school policy, regarding HE policy it generally has been more concerned with upholding vocational education and the special status of applied science institutions, thus arguing against "academization" (SVP 2010). Skeptical of European integration, it also has expressed its opposition to the high share of foreign academics at Swiss universities. The SVP garners most of its support in Zurich and Bern, but is consistently weak in urban French-speaking university cantons such as Geneva and Vaud. As large donor and university cantons, Zurich and Geneva both had strong incentives to uphold the equalization scheme. Thus, the SVP did not put up any significant resistance to HE centralization despite its general aversion to Swiss-wide harmonization and trans-European policy initiatives. This resulted in a situation in which the largest and most influential (university) cantons (Zurich, Geneva, Bern, Vaud) were dependent on centralized HE structures due to their reliance on the intercantonal flow of funds and federal subsidies, while the richest German L€ ander preferred decentralization.

Conclusions
In this article we showed that the increasing internationalization of education policy-in Europe primarily driven by the Bologna Process-functioned as an important catalyst for national reform processes in two federally organized states. In HE, these reforms spanned not only to structural issues such as study degrees and their international comparability or issues of quality assurance, thus policy, but also the distribution of decision-making authority, thus polity: While such processes are reflected in increased decentralization of the polity in Germany through the increased authority of the L€ ander, the reverse development has taken place in Switzerland where the federal government and joint coordination bodies at the federal level with strong cantonal participation have accumulated more powers. We argued that the pre-existing organization of federalism is crucial in shaping the financial and political incentives and preferences of partisan actors as well as subnational units, i.e. L€ ander and cantons. Established patterns of federalism were responsible for pushing the HE system towards greater centralization (Switzerland) or decentralization (Germany).
We also demonstrated that the Bologna Process brought about much more change than the mere harmonization of study structures across Europe. Hence, international initiatives may produce unintended consequences, which were hardly foreseen by the initiators of the process-in our case "polity impacts". Further research should therefore focus on intended and unintended consequences of intended actions at different levels for educational policy. Our contribution examinedx the interplay between the international, the national and the subnational level and showed how intergovernmental agreements signed by sovereign states are actually spelled out in federal systems. In both cases, this threelayer interplay was strategically exploited by involved actors to overcome political bottlenecks-irrespective of whether this led to more decentralization or more centralization. In this sense, and extending Robert Putnam's "two-level game" (1988), our contribution can be seen as an advancement to federalism scholarship by highlighting the connectedness of intergovernmental relations and institutional path dependencies with strategic choices of actors at the "third level", namely the subnational, federal level.

Notes
(generally 1,000-2,000 CHF per annum) paid by foreign students, as they are not covered equalization scheme (Brinck 2010). This increased the university cantons' reliance on federal support and subsequently their acceptance of further centralization.