Introduction

The study of higher education (HE) reform policy since the 1990s has been characterized by two tendencies. The first is to start out from what we may call the New Public Management (NPM) assumption that goes like this: Since the 1980s HE reform policies have primarily turned on attempts to make the sector and individual institutions more efficient and market oriented in accordance with characteristics of NPM as a public reform ideology (Pollitt 1993). The second tendency is that most studies of HE reforms have taken existing policies, as expressed in policy documents, legislation and regulations, as their point of departure. With few exceptions (e.g. Bleiklie 2006; Bleiklie et al. 2000; Kogan and Hanney 2000), questions about policy-making, such as how and by whom HE policies are made and designed, have been left out. The research questions have tended to focus on the extent to which and how declared policies are implemented and have affected universities and HE systems.

Thereby an important set of explanatory variables, structural characteristics of political-administrative systems, are left out. This is striking since policy studies in many other areas do in fact include this kind of variables, which is demonstrated by the literature that has emerged since the publication of Lijphardt’s Patterns of Democracy (1999). A number of studies have also demonstrated that recent HE reform policies come in different shapes although they claim to do roughly the same and are justified in terms of near identical ideas. Thus there is a considerable cross-national variation both in terms of the policy instruments that are devised and how they are being implemented (Bleiklie et al. 2011; Kogan et al. 2006; Paradeise et al. 2009). Time and again structural explanations are invoked to explain such differences. The abrupt and forceful shift in English HE policy introduced by the Thatcher government in the 1980s is a unique and often cited example of an atypical case that tends to be explained in terms of the peculiar opportunities offered by the winner takes all nature of Whitehall parliamentarism (Kogan et al. 2006; Paradeise et al. 2009).

The fact that structural explanations at times seem to stand out as candidates that cannot be overlooked in studies of HE policies, but never seem to have been included systematically in any research design that have been applied until now, suggests that they deserve to be tested systematically. We also believe it makes sense to include a broader canvas of explanatory factors than policy decisions and policy implementation, in order to properly understand variations in HE policies.

Furthermore, it makes sense to keep in mind a broader notion of reform policies than the rather simplistic notion of a move from traditional towards NPM forms of state steering and control. Paradeise et al. (2009) argue that a number of reform policies may be better understood when interpreted within alternative frameworks such as Network Governance and Neo-Weberianism . Bleiklie et al. (2011) demonstrate how NPM reforms may strengthen horizontal structures based on network power rather than stronger hierarchical structures. Both kinds of observations bear out the argument that it makes sense to include alternative perspectives to NPM in studies of reform policies.

The purpose of this chapter is, accordingly, to develop a conceptual framework for a comparative analysis of HE reform policies that enables us to investigate the explanatory power of structural characteristics of political-administrative systems . However, focusing on structural explanations it is important also to keep in mind that we are studying policies within a specific policy sector with its particular structural features and actor constellations  (Cf. Chap. 10). At this level there are actually stronger traditions for structural explanations. This is evident when researchers apply typologies distinguishing between “university traditions” such as the “Humboldtian”, “Napoleonic” and “Anglo-Saxon” traditions. The implication is that if we want a full appreciation of the potential of structural explanations, we need to include sector-specific characteristics in the policy analysis. Our aim is that the framework can address limitations that follow from the two tendencies mentioned initially. The policies we will study aim at improving the efficiency and quality of institutional performance. The approach is clearly structural, and we will focus in particular on structural changes in decision-making, funding and evaluation, and on how and to what extent they have affected the organization of universities in Europe.

The main research question: What implications does the organization of political systems have for policy content?

The general question may be specified like this:

  1. 1.

    What are the structural factors (independent variables) that drive policy-making in the field of HE?

  2. 2.

    What changes (dependent variable) have taken place in HE policy in the selected areas the last decades? Apart from easily observable declared policy goals, we intend to explore the questions of policy content by focusing on policy instruments and policy implementation.

Although we follow a structural approach in this chapter, it is important to emphasize that the approach has its limitations. We do not believe that policy outcomes can be deduced from structural characteristics of political-administrative systems, and like Kingdon (1995) we do not believe that the potential role of actors and process dynamics to affect policy outcomes can be overlooked. Given the absence of structural analysis in this field our aim is not to replace existing approaches, but to add the structural approach to the existing inventory of conceptual approaches in the study of HE policy. We want to clarify the possible impact of political-administrative structures in a more systematic way than the occasional ad hoc attribution of policy variation, which occasionally appears in the literature, to differences between majoritarian and consensual political systems or between federal and unitary states (Paradeise et al. 2009).

The following parts elaborate on these questions. In part 2, we focus on policy trends in HE in the eight countries in our study. In part 3, we discuss how the literature on comparative political and administrative systems can help us formulate assumptions about public policy-making and policy change. In part 4, these ideas are applied on public reform policies in general and in the area of HE in particular, followed by a test of these assumptions on available data on reform outcomes in seven countries. 1 The data indicate that a comparative political-administrative perspective is potentially useful with regard to explaining cross-national variation in HE reform policies in Europe.

Policy Change

The first research question we need to clarify is what reform policies have been devised and implemented. Based on existing knowledge, our main assumption is that policies with similar objectives have been put in place in the eight countries in our study.

One consistent tendency has been to strengthen vertical forms of steering through a process of hierarchization. In the area of decision-making and leadership the tendency has been to develop NPM policies aiming at stronger leadership and managerial structures at the expense of elected bodies controlled by academic staff. Thus executive leadership has been strengthened at the expense of collegial power in deliberative, representative bodies, while the academic community has (to varying extent) been transformed into staff and submitted to human resource management. In the area of funding, budgetary constraints have been tightened through reduced funding or by the introduction of new budgetary instruments based on indicators and output rather than on inputs. Thus budgetary reforms often implied heavier emphasis on performance and explicit performance measurement, assessment and monitoring in research and teaching while direct detailed regulation of funding decisions has been eased. Furthermore, there is a concentration of funds in the best-performing HE institutions and a broader vertical differentiation among HE institutions. Finally, evaluation of academic institutions and disciplines has been formalized and developed, partly in order to make decisions on accreditation and partly as an instrument to improve performance in management, research and teaching.

However, other ideas influenced HE reforms over the same period of time (Ferlie et al. 2008), and the vertical form of steering inspired by NPM has been complemented by forms of network governance . First, some policies encouraged the inclusion of stakeholders in academic affairs, on institutional boards and decision-making on research funding, thus widening the networks of actors involved in decision-making and opening up for the introduction of non-academic criteria, principles and preferences in such processes. Second, centralized ways of steering have been challenged by participation of inter- and supra-national actors in HE. As a result, most teaching or research projects mobilize a combination of resources from different sources and rely on multiple levels and actors. This has been conceptualized as multilevel governance. As shown by Paradeise et al. (2009), in order to understand recent HE and research reforms in one country, one has to look at the relative influence of NPM and network governance , their interplay and sometimes conflicting influence.

Finally, academic autonomy is often explicitly promoted in terms of institutional autonomy, while autonomy in turn is increasingly perceived as the competency given to institutional leaders to make strategic decisions on behalf of their institution (Chap. 3). Thus budgetary reforms imply less detailed regulation and more leeway for institutional leaders to allocate funding as they wish. However, increased autonomy also tends to be circumscribed by increasing standardization in terms of procedures and performance criteria that may severely limit the space for strategic decision-making.

We might claim, therefore, that three major trends characterize policy change: policy movements pushing for stronger institutional hierarchies, for stronger inter-institutional networks and for standardization and formalization. However, we also know that there is considerable variation regarding the degree, and form of the actual changes that have taken place. Our point of departure is that these trends, and the extent to which they are implemented, may be better understood in terms of the structural characteristics through which policy processes take place.

A Typology of Political-Administrative Regimes

For our analysis of the impact of political-administrative structures on reform policies we have developed a typology that includes the following five dimensions: (1) state structure , (2) the nature of central level executive government , (3) actor constellations , (4) administrative traditions, (5) diversity of policy advice, i.e. the degree of diversity in the main political channels of political influence that fuel reforms. The combination of political system and central administration characteristics is based on the following considerations.

Firstly, our approach is based on a literature that focuses on whether variations in political systems and practices relate systematically to the products of government action (Hofferbert and Cingranelli 1996). Political systems provide distinct and relatively stable environments for policy-making in any particular policy field. While some systems may offer rich opportunities for actors to affect policy change, such opportunities may be less available in other systems that offer better opportunities (veto points) for actors to prevent change from taking place, modify reform proposals (consultation requirements) or at least delay decision and implementation processes (decentralized structures) for extended periods of time. For instance, what might be possible to accomplish in a unitary state like England, might be impossible in a federal state like Germany.

Secondly, an important point of departure is the Lijphart model of majoritarian and consensual political systems (Lijphardt 1999). 2 The two types of polities are associated with two different policy styles. Consensus politics is permeated by bargaining with many opportunities for a variety of actors to influence policies. Majoritarian polities are prone to more sweeping changes.

Thirdly, the Lijphartian approach is strongly focused on the relation between the legislature and the executive, and their proportional representation (PR) forms are at the centre of attention. Administrative systems are not. However, there are few indications that politicians usually engage in and fight over policy alternatives during the policy design and decision-making processes (Paradeise et al. 2009), and political conflict being absent it is a reasonable assumption that administrative agencies are allowed a much more prominent role in developing the policy alternatives that they subsequently are obliged to implement. Thus HE policies tend to become more like administrative policy, and better performance rather than new policies are at the center of attention (Pollitt and Bouckaert 2004: 8). Therefore, classical regime typologies, like the one of majoritarian and consensual systems, have to be supplemented by approaches that include the administrative apparatus of HE and its relations with the executive and other relevant actors.

Pollitt and Bouckaert (2004) offer an approach that meets this requirement. It integrates elements from the Lijphart political system approach with dimensions from public administration in different types of political-administrative regimes. They identify five different dimensions of political-administrative systems, which are likely to affect the processes of management reform. The main argument is that reform capacity and reform trajectories are broadly determined by regime type. Thus two political system dimensions—state structures and executive government —act in combination with the way in which the central bureaucracy is involved in policy-making. Pollitt and Bouckaert (2004) have pointed out three such dimensions; minister-mandarin relations, cultures of governance and sources of political advice. 3 While the former two focus directly on relations between politics and administration, the latter one opens up for influence from external institutions. Furthermore, we follow Verhoest et al. (2010) who in their study of state agencies employ a slightly different and extended version of the minister-mandarin dimension through the broader notion of actor constellations . Finally, we believe the administrative culture dimension, which is developed further by Painter and Peters (2010a) into a more nuanced dimension called administrative traditions, will provide us with a better and more nuanced tool for comparative analysis.

State Structure

Two basic dimensions of state structures are considered: The first is the vertical dispersion of authority, and how authority is shared between different levels (centralized versus decentralized). The second is the horizontal coordination at central government level (coordinated versus fragmented).

Vertical dispersion tends to be, but is not necessarily, the greatest in federal systems, where powers are formally delegated to sub-national tiers of government and least in centralized, unitary states. On the vertical dimension reforms in decentralized states tend to be less broad in scope and less uniform than in centralized states. Centralized states will also have a narrower focus on service delivery outputs and results.

On the horizontal dimension the question turns on the degree of horizontal coordination within central government in the various states. This is a difficult variable to estimate, but Pollitt and Bouckaert still hold that significant differences between countries can be identified in terms of horizontal integration. Thus they argue that in France the grand corps serves as strong glue at the top of a fragmented system, whereas Germany is even more fragmented than France. In England the treasury constitutes an integrating force.

The Nature of Executive Government

The nature of the executive is an important dimension in the structural analysis. Here Pollitt and Bouckaert adopt the main features of the Lijphart typology distinguishing between two main types of executive regimes:

  1. (a)

    Majoritarian regimes in which the power privileges of electoral majorities are emphasized, and accordingly relatively much power is vested in the executive

  2. (b)

    Consensual regimes where the importance of accommodating electoral minorities through negotiation and compromises are emphasized.

They then distinguish between four concrete categories of executive governments:

  • Single party, minimally winning or bare majority: one party holds more than 50% of the seats in the legislature

  • Minimally winning coalitions: two or more parties holds more than 50% of the seats in the legislature

  • Minority cabinets: government is supported by party or coalitions that hold less than 50% of the seats in the legislature

  • Oversized executives: additional parties are included beyond that of a minimally winning coalition

On the basis of these types of executive government characteristics, conventions are formed, which are regarded as fairly stable. They gravitate towards consultative practices the more one moves downwards in the direction of a minority or an oversized executive. The more one moves upwards the more practices tend to become majoritarian.

The two dimensions—state structure and nature of executive government—combine to exercise a significant influence on the formation of public policy.

  1. 1.

    Deep and rapid structural reform tends to be more difficult in consensual systems than in majoritarian systems. Majoritarian systems focus on political will, and generation of winners and losers. The more consensual the regime, the more likely is the opposite result. Consensual systems are less inclined to and less capable of radical reform.

  2. 2.

    Centralized countries find it less difficult to carry out radical reform than decentralized countries.

Abrupt policy changes produce winners and losers. The more consensual the regime, the more likely that losers will be represented in the executive leading to the proposition that polices will become diluted in the process. At the same time majoritarian policies may also fall victim to abrupt policy shifts. Reformers in more decentralized political systems will find it more difficult to carry out sweeping synoptic reforms than in centralized ones. All these features are well known. Pollitt and Bouckaert allow intermediary solutions, that is, the possibilities of hybrid s combining very different or even contradictory elements.

These two dimensions seem well suited to explain the English HE reform experience of rapid and radical change of the 1980s and increase our understanding of why it was different. It also makes sense in terms of reform experiences that Switzerland is locked at the opposite extreme on both dimensions. The position of the remaining countries in our sample is consensual, but centralized regimes (Italy, the Netherlands) and arguably intermediate regimes are found on both dimensions (France, Portugal, Norway and Germany). Over time there has been some movement, but none of the countries has changed position significantly (Lijphardt 1999).

Actor Constellation

Relations between politicians and bureaucrats vary across countries and over time (Peters 2008; Painter and Peters 2010b; Olsen 1983; Jacobsen 1967). A crucial issue is whether civil service decision-making is dominated by “technical” or “political” criteria, and what impact this has on careers. At one extreme, political careers are sharply separated from administrative careers while at the other extreme they are not clearly distinguished or “intermingled”. There is no accepted scale or classification available like the majoritarian versus consensual distinction. Politician-bureaucrat relations are complex and often difficult to specify. The role of top civil servants is a case in point. These positions are inherently political, and some contributions underscore the increasing weight of politicization, even in countries with a strong merit system and strong norms of civil servant neutrality (Peters and Pierre 2002). Furthermore, actor constellations at state level that have a bearing on policy reform include not just the minister-senior civil servant relations, but also the division of roles and responsibilities; cooperative or adversarial relations; ministerial capacity for policymaking and reform; ministry capacity for control; number of central agencies, degree of institutional differentiation; and position of the different institutions and their problem structures and reform ideas (Verhoest et al. 2010: 70–71).

The question is whether the formulation of reform proposals can rely on shared understanding and perceptions of policy problems, appropriate normative orientations and particular problem definitions and solutions. Shared perceptions are likely to facilitate a good match between reform designs and political preferences. One of the implications for administrative reform is that we may assume that countries with intertwined civil servant/politician relations will experience fewer problems advancing radical reform proposals. However, integrated elites at the top of the bureaucracy might also create problems in the following implementation process, as political decisions made at the top might be impeded by administrative opposition at a lower level generating implementation problems and implementation gaps as reform proposals trickle down the administrative ladder and reach the implementation stage (Lynn and Jay 1981). 4

Administrative Tradition

The question of administrative cultures turns on whether different cultures, each with their own specific values and assumptions, may be identified. Pollitt and Bouckaert (2004) distinguish between two main types—that of the “Rechtsstaat” and of the “Public interest”—based on how the role of the state is perceived. In the Rechtsstaat model the state is considered an integrating force, focused on the preparation and enforcement of law. Consequently, the bureaucracy emphasizes rule-following, correctness and legal control, its public servants will be trained in law, and a separate body of administrative law will be created. This type of administrative culture is often identified with the Weberian bureaucracy. The Public interest model on the other hand, envisages a less dominant role for the state. Government is a necessary evil, and ministers and officials have to be held accountable to the public through a variety of means. The law issue is not as dominant as in the Rechtsstaat model, and many civil servants will not have legal training. The process of governing and administration takes place in the context of competing interest groups, and the role of governing consists in being a fair and independent arbiter or referee, not that of a technical or legal expert.

The Rechtsstaat-public interest dichotomy is somewhat crude, since it leaves important dimensions out, as admitted by Pollitt and Bouckaert. Furthermore, it lumps a large number of countries into the Rechtsstaat category—in the TRUE sample we end up with one country (England) in the public interest category and the rest more or less clearly in the Rechsstaat category. A more fine-grained picture of administrative cultures can be explored through the literature of administrative traditions (Peters 2008; Painter and Peters 2010a: 20). The notion of “administrative tradition” 5 is close to that of “administrative culture”, but contains elements of institutionalism as well. The premise here is that administrative structures are engaged in political processes, but still distinguishable from state traditions. The nature of public administration may be influenced by political system characteristics, but may also develop independently. The assumption here is that specific traditions might influence contemporary reform and privilege certain policies rather than others (Peters 2008).

Administrative traditions can be defined in various ways. Painter and Peters (2010a) group administrative traditions into four families: Anglo-American, Napoleonic, Germanic and Scandinavian. They follow the Pollitt and Bouckaert scheme to a considerable extent, but divide the Rechtsstaat category into three different traditions; Napoleonic, Germanic and Scandinavian, each with a different combination of features (Table 9.1). One of the advantages is the specification of a separate Napoleonic tradition. In comparative administration the Napoleonic tradition is underrepresented (Ongaro 2010). But still the same question applies as to whether it is possible to identify one coherent tradition of cases classified as Napoleonic or whether they differ too widely along significant dimensions.

Table 9.1 Families of administrative traditions
Full size table

In the Anglo-American tradition boundaries between state and civil society are far from clear. There are close ties to the Common law tradition, an inductive and procedural approach through the accumulation of case law, in contrast to the Roman law tradition with its deductive approach. Furthermore, there is a tendency to elevate political rather than legal accountability mechanisms. The profession of public administration is about management and policy, not the law. Still the doctrine of separation of politics and administration is prevalent.

Napoleonic traditions share the Rechtsstaat focus on law as a state instrument for intervening in society rather than serving as a means of conflict resolution. A separate system of public law regulates relations between state and citizen. Administration is closely related to laws, and the complex relations between constitutional law, statutes, regulations, administrative notes and circulars, define the scope and content of administration. Within the Napoleonic tradition, Peters (2008) asserts that the role that societal actors and networks legitimately can play in countries characterized by this administrative tradition is rather small. In fact, interest group participation is considered almost illegitimate interventions into state autonomy. Therefore, interests are not usually incorporated into public administration and there is considerable selectivity about participation. State autonomy is crucial in these administrative traditions . Still, Peters acknowledges the need to study whether the development of network governance has enhanced the role of society in these countries.

The crucial question is if systematic propositions on the relation between administrative structures and the character of reforms could be formulated within Napoleonic traditions. One important feature often associated with the Napoleonic traditions is the problem of “implementation gaps”. Such gaps may be observed in two different ways: firstly, the distance between what is prescribed by reform legislation and the actual existence of management tools; and secondly the distance between the mere presence and the actual utilization of management tools. The assumption of the existence of such a gap in the implementation of reforms in Italy is widely shared in international political science and public administration literature. Still, the problem of explaining implementation gaps remains (Ongaro and Valotti 2008). Although such gaps exist all over the place, they appear to be larger in Napoleonic countries than those experienced in northern Europe and in the Anglo-Saxon countries thus opening up for cross-national comparison as one way of arriving at explanations, e.g. associated with differences in typical reform trajectories.

It has been argued that the changing Italian system may provide a more favorable environment for more radical, wide and rapid transformations (Ongaro and Valotti 2008; Ongaro 2010). The picture of the relations between executive politicians and top bureaucrats is complex, but the closer connection between the two components determined by the spoils system might reinforce the tendency to radical and sweeping reforms, as well as sudden interruption or reversals of them. By comparison data from France seem to suggest a different reform trajectory. Management reforms in France seem more fragmented as management tools have been domesticated (Rouban 2008). The main characteristic of administrative reform processes in France is the rather steady progress through incremental changes and not through major legislative overnight changes. On this basis it is also often contended that only a strong political commitment may alter this structure. This suggests that public management reforms also differ even within the category of Napoleonic traditions (Suleiman 2003; quoted in Rouban 2008).

German traditions differ from Napoleonic traditions in several ways. While Rechtsstaat traditions are embedded in a decentralized and federalist order, Napoleonic traditions focus on the unitary organization of the state, a “technocratic” orientation towards decision-making and a nation-building role of government. A unified administrative arrangement produces uniformity as well as an exclusive administrative class. The southern variants are characterized by high degree of legal formalism, management by decrees coupled with clientelism. Legal formalism breeds double talk, and outcomes are arranged through informal relations. Furthermore, the Germanic cooperation between state and non-state corporations, which often are given a special legal status, a feature strongly related to the organic view of society also stands out, in addition to the fact that German administrative structures are more fragmented than their French counterparts.

Scandinavian traditions combine the German orientation towards the law with a strong universal welfare orientation. It operates within unitary states, but shares with Germanic traditions similar corporate mechanisms of cooperation between state and non-state actors. In addition, it is also characterized by small ministries supplemented by a variety of autonomous agencies.

In the following table, the TRUE cases are distributed according to families of administrative traditions (Table 9.2).

Table 9.2 Families of administrative traditions —TRUE sample
Full size table

Norway is an obvious example of Scandinavian administrative traditions. Usually France, Portugal and Italy are regarded as countries belonging to the family of the Napoleonic administrative traditions (Ongaro 2010). Ongaro underscores the significance of the interpenetration and porousness between the civil service and political careers in Napoleonic countries compared to the others, as well as the significance of law as opposed to management (Ongaro 2010). Again the problem of hybrid s and compounds comes up (e.g. Netherlands, Portugal and Italy). Italy seems to have “imported” the French model in its entirety (Ongaro and Valotti 2008). This does not apply to Portugal, where the administrative tradition may be regarded as a specific combination of British, French and indigenous influences (Corte-Real 2008).

Diversity of Policy Advice

The last item in the template is focused on the question of policy advice. Here we ask about the extent to which high level civil servants are directly involved in HE policy-making or whether important alternative sources of policy advice have emerged.

In general, the civil service is regarded as the prime source of policy advice and normally considered the most obvious source of professional policy advice and feasible options. However, the monopoly on policy advising once held by the civil service has gradually been broken. We have seen a considerable increase in the number and prominence within government of political advisers appointed outside the civil service framework. Additional policy advice is also being provided by a variety of policy and planning units as well as special advisers and various forms of experts, consultants and (politically affiliated) think tanks. The old civil service prerogative of policy advice has also come under pressure from various external sources, most notably external challenges such as internationalization, Europeanization, and multilevel policy-making processes. It has been increasingly argued that a shift from government to governance has occurred, where formal hierarchies are being supplemented as well as challenged by alternative organizational forms such as networks (Peters and Pierre 1998; Rhodes 1996). The capacity of the national bureaucracy to deliver policy expertise and policy advice is reduced, forcing the state to extract policy advice from institutions and networks outside the civil service, at the national level from semi-autonomous administrative agencies and networks, but increasingly also from international and supra-national arenas and networks.

Nevertheless, the extent and form of this kind of development vary considerably across nations. Based on McGann (2009: 13), the TRUE countries could be divided into three groups according to where think tanks are most prevalent. The UK, Germany and France are at the top, a middle group consists of Italy, Switzerland and the Netherlands, while Portugal and Norway are at the bottom of the list.

Thus structures and processes of policy advice seem to vary across nation states, between highly competitive, adversarial, and politically partisan (UK), and more consensual and non-partisan (Germany and Switzerland), as well as technocratic ones (France). These differences seem to have affinity to differences in political-administrative systems . The UK case represents a type of system, where there are contested and more competitive processes of winners and losers. Here there is a tendency towards radical advice, of the provision of policy advice that facilitates more radical policy reform. Decentralized and federalist systems like Germany and Switzerland provide numerous access points for external policy advice. Yet what is produced is generally consensual, incremental and non-partisan processes of political advice. There is not much space for partisan political advice, which would require a decoupling of think tank networks and dominant corporatist structures. Thus there is little evidence to support the standard proposition that the wider the range of sources of advice, the more likely it is that new ideas might be adopted by policy-makers (Peters and Pierre 1998). Napoleonic systems provide a different environment for policy advice. The centralist features of the political system offer few entry points for external actors and expertise. Policy advice takes the form of technocratic and statist expertise. We conclude this part by presenting the TRUE sample according to how the countries might be classified in terms of the five dimensions and the illustrations that we have discussed above (Table 9.3).

Table 9.3 Types of political-administrative regimes, TRUE sample
Full size table

Political-Administrative Systems and Reform Policy

After having presented the five dimensions of the political-administrative system based on the contributions of Lijphardt (1999), Painter and Peters (2010a), Pollitt and Bouckaert (2004), and Verhoest et al. (2010) and classified the TRUE countries in terms of these dimensions, we are now in a position to formulate more systematic assumptions about the likely relative HE reform activity in these countries in terms of the timing, pace, rate and direction of reform.

We suggest the following propositions concerning characteristics of political systems and reform policy, including HE reform policy.

State Structure

Variations in state structures will produce reforms that vary in pace and scope. The most straightforward basis for this hypothesis is the different opportunities, pointed out by Lijphart (1999) that exist for actors to slow down, modify or prevent decisions from being made in unitary, centralized states (few) as opposed to federal and decentralized ones (many):

  • In centralized state structures HE reforms are expected to be more comprehensive in scope than in decentralized states where numerous veto points are more likely to limit the scope of reforms.

  • In centralized state structures HE reforms are expected to be more oriented towards policies that support hierarchization.

  • In decentralized/federal states HE reforms are more likely to be oriented towards policies that favor network governance .

Executive Dimension

Variation in executive regime, in party composition and durability of cabinets has consequences for HE reform policy. These hypotheses are based on Lijphart’s (1999) analysis of the characteristics of majoritarian and consensual systems suggesting variations both in the capacity to produce radical change and to keep up steady reform processes. However, the argument that one-party majority governments produce more effective and efficient policy-makers is based on reasoning that confuses the impact of the “strong hand” with that of the “steady hand”. Lijphart argues that coalition governments are just as able or even better able to provide sensible and coherent policy-making, and that policies supported by a broad consensus have better chances of being implemented. Policies that stay on a steady course are thus superior to policies implemented by a determined government that may be succeeded by an equally determined opposing party resulting in a stop-go pattern of policy-making:

  • Majoritarian political systems will produce HE reforms that are more comprehensive, radical and sweeping than consensual political systems.

  • Majoritarian systems tend to produce a stop-go pattern of reform while consensual systems tend to produce a more continuous and incremental pattern of reform.

  • Cabinet structures seem to have a weak explanatory power. All governments in our sample tended to seek and receive cross-party consensus. Thus it has to be acknowledged that consensual as well as majoritarian cabinet structures may be able to carry out significant reforms.

Administrative Tradition /Administrative Culture

The basic assumption here is that variations in administrative traditions tend to produce different reform trajectories along at least two dimensions. The first is the extent to which legal instruments are important or dominant compared to other policy instruments. The second is how administrative traditions induce different ways of practicing reform policies. These assumptions underscore the significance of legal traditions for HE reform. However, the relationship between administrative tradition and reform policy is not straight forward. Tentatively we suggest these hypotheses.

  • A strong administrative law component (seen in Rechtsstaat and Napoleonic traditions) seems to strengthen focus on control and regulations.

  • Administrative cultures of the Rechtsstaat tradition tend to induce incremental administrative HE reforms, whereas public interest traditions gravitate towards comprehensive HE reform.

  • Administrative cultures of the Rechtsstaat tradition tend to induce reform through legal procedures, whereas public interest traditions gravitate towards a more varied set of devices.

Actor Constellation

The actor constellation in focus here is the relationship between politicians and senior civil servants. It is usually assumed that this relationship is important for how reforms are implemented through public bureaucracies: They may be tightly integrated, e.g. through a spoils system, where senior civil servants act like politicians, or they may be separated in a constellation where politicians operating according to political criteria are sharply separated from administrators operating in accordance with “technical” criteria. As argued above, the implications of integration or separation are not straight forward. The underlying variable that is emphasized here is whether these actor groups enjoy trusting mutual relationships and how this interacts with the degree and form of civil servant involvement in policy-making (Peters and Pierre 2002; Verhoest et al. 2010).

  • Political-administrative relations, where minister-top civil servant relations are adversarial with little civil servant involvement in policy-making, induce radical reform and favor NPM-oriented policies.

  • Political-administrative relations, where minister-top civil servant relations are trusting with considerable civil servant involvement in policy-making, induce incremental reform and favor Network governance oriented policies.

  • Cooperative and trusting relations between ministers and civil servants combined with a low capacity for control/(small ministries) and a weak position of central ministries might provide a fruitful context for reform policies in the form of autonomous agencies.

Diversity of policy advice: As already noted the nature of policy advice in the TRUE countries seem to be related to and affected by the political-administrative structure and the policy environment in which they are embedded, and at this point we are not in a position to formulate hypotheses about the effect of the diversity of policy advice on policies.

Applying the Perspective on HE Reform Policy

The different strands of literature discussed here differ along a number of dimensions. Both the Lijphardtian literature on comparative political systems and policy, and the literature on political-administrative traditions are associated with the macro-institutional context of the state (Yesilkagit 2010). But we also know that policies, institutional setups and structures vary within a particular country, albeit to a different extent, and they remain lodged in different setups, norms and traditions of governance (Peters 2010; Lodge 2010). We may also distinguish between different sector traditions (Yesilkagit 2010). Obviously this applies to the liberal Anglo-Saxon countries where coordination has been weak, but we also find a variety of sector governance traditions in more tightly coordinated countries like the Netherlands and Norway. Such features may reflect older structures and traditions embedded in the various sectors. Thus, the relations between national political-administrative systems and sector systems can be regarded as somewhat loose or indeterminate.

These reservations, notwithstanding, we shall use secondary data from the comparison of HE reforms presented by Paradeise et al. (2009), the most systematic comparative analysis available, in an attempt to test the assumptions that follow from the comparative analysis of political-administrative systems discussed above. The main conclusions in the Paradeise et al. (2009) study were as follows: The pace, methods and extent of reform and policy change varied across countries. If one considers the period from 1980 onwards, the study thus identified one NPM outlier (UK), one group of Continental European countries, and the Netherlands in between. The authors claim that the UK must be understood as a NPM outlier, apart from which the diffusion of most radical NPM ideas proved problematic. Outside the UK and possibly the Netherlands, policy-makers did not have the ambition of building an exhaustive system of operational instruments based on an elaborated ex ante theory of action. They also suggested that interest in new policy instruments in the other countries resulted mostly from increasing cost-awareness. In this interpretation, reforms and instruments have been largely contained within national HE traditions, and the new levers of action were digested by the environment they were supposed to impact. Implementation processes have tended to follow incremental rather than radical trajectories. In many countries, such as France, Italy, Norway, Germany, and Switzerland, general legislation would typically pile up, whereas implementation tended not to follow policy objectives. Reforms that developed during the 1990s were mostly disjointed and incremental. Yet, deliberately or not, technical measures sometimes opened unexpected channels for change as was the case in France. Furthermore, tendencies towards systematic reform have become more perceptible since the turn of the century and more powerful instruments of funding, evaluation and governance have been introduced in countries like France, Norway, Germany and Switzerland (Bleiklie and Lange 2010; Bleiklie et al. 2011; Paradeise et al. 2009). Thus the net result of reform activity over a span of 30 years remains to be seen, as is the comparative reform histories of the “strong hand” UK-policies and the “steady hand” policies of France and Norway. These reforms span over quite a wide range of aspects such as legal status, funding, evaluation, institutional leadership, decision-making and internal organization. Thus it is important also to look at the focus or emphasis of reforms both as to the specific dimensions that are targeted in a country and the extent to which reform efforts are comprehensive or focused on one or a few dimensions (Paradeise et al. 2009). If we distinguish between NPM and NG policies, we find that network governance has contributed to limiting and modifying the effect of NPM policies in settings as different as the Netherlands and Switzerland (Bleiklie et al. 2011).

We argue that there are significant differences between countries with different administrative traditions. The Napoleonic and Rechtsstaat traditions have several features in common, when measured against the British public service tradition. But they differ substantially along other dimensions, and interact with other state level dimensions such as state structure . The difference between federalist and unitary states is a case in point. There is a big difference between the German federalist structure and the more unitary French organization of the state along the centralization/decentralization dimension. This also applies to social democratic systems normally associated with countries in northern Europe and Scandinavia. As a type of political-administrative regime, the social-democratic tradition shares the strong unitary state dimension with France, but still appears decentralized. Unlike the Napoleonic tradition, where the role that societal actors and networks legitimately can play is rather small and interest group participation is considered almost illegitimate, interventions into state autonomy and participation in policy networks is extensive and legitimate. This might suggest a more participatory pattern of modernization, including policy formation and implementation , rather than a managerial pattern.

In order to produce a tool that can help us analyse the complex picture rendered by available studies of HE reform, we have developed a procedure to reanalyze data from Paradeise et al. (2009) inspired by the approach used by Pollit and Bouckaert (2004) in that we find their Neo-Weberian distinction between modernizers and marketizers as representatives of alternative modernization strategies useful for our purposes. Thus Continental European countries, usually perceived as NPM laggards, were here conceptualized as actors pursuing a positive modernization trajectory.

The Paradeise et al. (2009) study developed narratives as a central principle for the organization and interpretation of data. Narratives are stories about actual or fictional events. Their strength is based on their internal coherence that affords cognitive frames used as policy models and theories for action. However, what is gained in coherence may be lost in nuance and prevent us from identifying important factors that affect reform policies. A reconceptualization and recalibration of the various reform dimensions from a narrative to an indicator oriented framework may therefore help us identify more empirically delineated reform routes which entail a different type of evaluation of the reforms and their various components.

Based on thematic charts presenting reforms and changes in seven European countries from 1980 until 2005 in Paradeise et al. (2009: 247–290), we selected nine reform dimensions that are of particular interest to us and at the same time provide a broad range of reform activities in the HE policy field. The charts were originally developed to map changes between 1980 and 2005 and were developed after lengthy discussions were members of the project team behind the book discussed and selected dimensions they agreed were of importance to university governance and were researchable in terms of available data in the participating countries. 6 The thematic charts were then developed with brief verbal descriptions of each country’s position on each dimension during the 1980s and during the 2000s. Based on the 20 different dimensions in this study we selected 9 dimensions (Table 9.4). Each country was given a score on three reform characteristics on each of the dimensions: (1) the amount of change involved, (2) year of reform and (3) degree of implementation . Scores were given on the basis of the verbal characterization provided by the thematic charts about amount of change and degree of implementation as shown below in addition to the year relevant reforms were introduced in the countries in question.

Table 9.4 HE reform output in seven European countries 1980–2005
Full size table
  1. 1.

    Change: (0) None (1) Minor (2) Major

  2. 2.

    Year: (0) 2000s (1) 1990s (2) 1980s

  3. 3.

    Implementation: (0) Weak (1) Medium (2) Strong

Scores on each criterion may vary between 0 and 2 and on each dimension between 0 and 6. The scores thus give a condensed expression of whether a country has been a swift and sweeping reformer as opposed to a slow and incremental one, or something in between. With nine different reform dimensions the scores also give an impression of reform emphasis as well as reform persistence. The reform dimensions are: procedural reforms, external governance, role of stakeholders, internal decision-making, internal organizational structure, leadership, external funding, internal funding, and evaluation. The results are presented in Table 9.4.

The countries have been grouped in order to keep different administrative traditions together, based on the idea that administrative culture and tradition play an important part in distinguishing between HE reform outcomes. Therefore Public interest (England), Social democratic (Norway), Napoleonic (France and Italy) 7 and Rechtsstaat (Netherlands, Germany, and Switzerland) traditions seemed as a good point of departure to formulate assumptions about reform experiences. We would expect the former traditions to achieve high scores and the latter to achieve lower scores. We are not going to comment on all the 12 propositions presented above in detail, but rather focus on the actual scores of the seven countries and how they may be interpreted against the backdrop of the political-administrative characteristics discussed previously.

Not unexpectedly, England achieved top score, combining a majoritarian system, a unitary state structure, long cabinet life, and a public interest administrative culture. Norway’s high score is interesting. It may suggest that although consensual systems may be slow in making comprehensive reform, sustained activity across a broad range of issues may yield results over longer time spans that match those of majoritarian regimes. This case may testify to the thesis suggested by Olsen (1996) about the “triumph of the tortoise” in the context of national reform policies. Still, there is also much to suggest that the Norwegian political system has moved in the direction of a majoritarian system the last 10–20 years (Baldersheim and Rose 2010).

However, the position of the Netherlands and Norway with the same reform score needs further discussion. The Dutch position is not well explained by the characteristics of these models. Timing may be the major difference compared to other consensual regimes. The fact that the Dutch started reforming rather early has given new policies longer time to take effect than in countries that started later. In addition, the fact that the Netherlands is a unitary state makes a major difference compared to the two other federal Rechtsstaat regimes in which the federal governments firstly, play a minor role compared to the Länder and Cantons and secondly constitute a structure with far more veto points that limit the possibility for comprehensive reform policies at the federal level.

The concept of the Napoleonic family of administrative traditions is not often explored in comparative public administration (Ongaro and Valotti 2008). The two Napoleonic regimes look quite different when it comes to HE reform. The Italian structure is characterized by a combination of unitary and decentralized state structure and a tradition of large majoritarian coalitions. It has also been argued that the closer connection between Italian politicians and top civil servants forged by the spoils system have reinforced both the tendency to radical and broad reforms, as well as the sudden interruption or reversals of them (Ongaro and Valotti 2008). But Italy has arguably produced little in terms of HE reforms compared to France. The Italian reform pattern seems to represent the classical “Napoleonic” pattern where reform tends to focus more on legislation and procedure at the expense of other policy instruments that are needed in order to achieve political reform. In this sense the broader focus of the French reforms may demonstrate that with deployment of a broader range of policy instruments , this pattern is less prevalent than in Italy. This is somewhat at odds with prior descriptions of French administrative policies as rather fragmented and weakly influenced by NPM ideas (Rouban 2008). There seems to be contrasting and somewhat contradictory evidence in Napoleonic countries on reform experiences. The conventional view is that administrative reform has been more limited in France compared to most other countries, like the northern European countries or the UK. The Napoleonic administrative tradition creates an unfavorable environment for the implementation of NPM models and mechanisms. Still it has been maintained that path-breaking budgetary and procedural changes have been carried out and a full set of performance instruments implemented (Bezes 2010). There has been a slow but steady ideological conversion among top civil servants, and NPM has become the new orthodoxy, at least for some reform issues. Through a combination of displacement and layering processes reforms have taken new and surprising shapes, where “low profile” procedural reforms have been reshaped as budgetary reforms, taking on performance measurement elements in the process. But such layering and conversion processes where policy-makers work around the opposition by adding new institutions without dismantling older ones (Palier 2004), and where policy instruments change significantly during the policy process also makes the task of inventorying policy change along different policy areas problematic.

The substantial differences between the Italian and the French reform profiles are also interesting in relation to notions of implementation gaps. Often it is assumed that Napoleonic systems provide severe impediments for implementation of NPM reforms (Ongaro 2010). The variations discovered do not provide empirical support to such a view.

The low reform scores of Germany and Switzerland is clearly what one might expect in federal political regimes that produce many veto points compared to the rest. In addition, in the German case the combination of minimally winning coalitions and the powerful and reluctant position of the professoriate should be mentioned, and in the Swiss case, the dominant perception that their system works rather well in terms of quality as well as efficiency. Still, from a Swiss point of view it could be argued that policies have worked not in spite of, but because they are enmeshed in a system of consultations and organized interests. One major reform, therefore, introduced after prolonged and extensive debate and consultation, could achieve at least as much as ten smaller initiatives, and a broad consensus improves the chances of having policies implemented. Even though Swiss reforms might seem modest from a comparative cross-national perspective, commentators have characterized changes brought about in the 1990s as a paradigmatic shift that deeply modified the HE system (Perellon 2001). Federal authorities have managed to gain influence over the definition of the general conditions within which HE policy has to be developed in Switzerland. The road that had to be followed in order to get there has been characterized as tortuous (Perellon 2001).

Specific explanations aside, the structure of the findings does not conform well to conventional comparative divisions like the consensual-majoritarian divide or the Rechtstaat-public service divide. There are, however, indications that the Netherlands as well as Norway has been moving from the Rechtstaat-consensual regime towards a majoritarian-public service oriented regime. The high reform scores of the two northern European states also conform to the Pollitt and Bouckaert findings on administrative policies, which identified major implementation differences between continental European countries on the one hand and the UK, as well as between Northern and Southern Europe.

We argue that a more differentiated grid of political-administrative regimes, where administrative traditions and cultures are the most enduring features, might be more fruitful for our purposes. This allows a more fine-grained interpretation based on four different traditions; the public service, Rechtsstaat, social-democratic and Napoleonic traditions normally associated with distinct geographical denominations. One might suggest that the Netherlands perhaps should be treated as a North European social democratic country comparable to Norway, as these states arguably share a specific participatory modernization trajectory which will have to be explored further.

It might be objected that our use of quantitative methods in comparative research in order to assess policy outputs and performance is problematic. The greater number of policy initiatives are taken as unquestioned achievements. In a generic sense this type of quantification postulates equal effectiveness of German and British HE policies. Policy counting obviously has its shortcomings. So does equating outputs with performance. The actual impact of these reforms is a question that needs further research. There is obviously a danger that policy outputs in consensual/PR and decentralized countries, with their structural characteristics, might be underestimated in this type of analysis. Still we would argue that this kind of policy counting, which includes both the scope of reform in a number of well-selected areas as well as their implementation , nonetheless provide a useful measure of policy-making capacity as well as policy output.

Conclusion

We have argued in this chapter for the usefulness of testing structural approaches in comparative studies of HE reform policies. Based on contributions by Lijphardt (1999), Politt and Bouckaert (2004), Painter and Peters (2010a), and Verhoest et al. (2010) we outlined a typology of political-administrative regimes, and tested the approach empirically based on data from Paradeise et al. (2009). The results demonstrate that there is no straightforward unequivocal relationship between political-administrative structures and reform activity. Nevertheless, we have demonstrated its usefulness because it helps us clarify how structural conditions may have deep implications for policy outcomes, but in a much more flexible and ambiguous way than one might have expected if the goal had been to identify one to one relationships between structure and outcome. Thus the next step is to link agency to reform policy.

Accordingly, the structural approach can help us make sense of some of the variation we have observed. The three high score reformers belong to three different administrative traditions , and all four traditions are represented if we include France which is not far behind, among the high score reformers. This suggests that the traditions per se cannot explain the variation we have observed in a deterministic way. The only common characteristic is the unitary state structure we find in all the high performing countries. However, the lowest score reformer also has a unitary, although (increasingly) decentralized, state. Among the low score reformers Germany and Switzerland share characteristics on a number of dimensions, but Italy hardly share any with the two former countries. We are left with the impression that structural characteristics offer different conditions for reform processes that may limit or be exploited by actors who may want to promote, to slow down or downright prevent reforms from being introduced. Thus there are different paths to high reform scores, one based on the ability of actors to implement swift and sweeping reform (England, Netherlands) and another on the ability to keep up a relatively steady incremental process over a broad range of issues (France, Norway). Similarly, there seems to be two main roads to low reform scores, one characterized by federal structures and many veto points (Germany, Switzerland), the second by decentralized structures with a reform focus on legal and procedural issues and a relatively strong separation between formal procedure and informal practice. This leads to the unsurprising conclusion that the relationship between structural conditions like the ones studied here and policy processes is not deterministic but stochastic. It also demonstrates the necessity of studying empirically how structures and agency interact in affecting the outcome of policy processes.

Notes

  1. 1.

    Data on one of the eight countries in the TRUE project, Portugal, are not complete (cf. note 7).

  2. 2.

    In addition to the literature on comparative political systems, there are various strands of literature that might be relevant for our purposes: literature on welfare state regimes (Esping Andersen 1990) and partisan politics and redistribution policy (Iversen and Soskcise 2006; Iversen and Stephens 2008); on public administration and public policy (Pollitt and Bouckaert 2004); and on networks and their role in policy formation (Bleiklie 2006; Koppenjan et al. 2009; Rhodes and Marsh 1992; Van Waarden 1992).

  3. 3.

    This approach provided the template for the comparative analysis of public management reforms in 12 different countries as well as the EU: Australia, Belgium, Canada, Finland, France, Germany, Italy, Netherlands, New Zealand, Sweden, England and USA.

  4. 4.

    Also role perceptions of top civil servants will be important, e.g., whether they consider themselves as judges, arbiters or negotiator/mediator or as party politicians or accountants (Olsen 1983).

  5. 5.

    By administrative tradition we mean “a historically based set of values, structures and relationships with other institutions that define the nature of appropriate public administration within society” (Peters 2008: 118).

  6. 6.

    The project, Steering of Universities (SUN) was funded by the EU PRIME program, led by Catherine Paradeise and the team had members from seven different European countries.

  7. 7.

    Data from Portugal were not collected in the SUN project.