Consensus Democracy and Bureaucracy in the Low Countries

• 1 Introduction

  The purpose of this special issue is to explore the degree to which The Netherlands and Belgium can still be regarded as consensus democracies as originally defined by Arend Lijphart (1968). Indeed, in current political analysis the question is being raised as to whether consensus politics might be eroding (Rosanvallon, 2018; Vollaard, Beyers & Dumont, 2015). This article contributes to the debate by enquiring about the degree to which Lijphart’s conceptual framework of consensus democracy, fifty years after its original publication, is still relevant to the analysis of present-day politics and administration in the Low Countries. Taking Lijphart’s work as our point of departure, we signal that the usefulness of Lijphart’s approach for the present-day analysis of political administration in the Low Countries has by no means disappeared but has a major limitation. This limitation is Lijphart’s almost exclusive focus on the political hardware of the state structure and his insufficient attention to the administrative and bureaucratic characteristics of the governance system, which we argue are critical for both the policymaking process and service delivery.

  Therefore, we problematize Lijphart’s normative evaluations of consensus systems as superior to majoritarian systems in terms of democratic quality and policy effectiveness, since he does not take into account the key features of the bureaucratic system. In order to remedy this, we propose to expand the model with seven additional features of the bureaucratic system (see also Daalder 1964; 1974a; 1974b; 1995; Van den Berg 2018a). In the context of this special issue, which focuses on The Netherlands and Belgium, we provide a comparative study of two consensus systems. Although this does not allow us to assess the superiority of such consensus systems over majoritarian systems, we can address the potential differences between the Dutch and the Belgian systems of consensus governance and suggest plausible explanations for these. In this study, we argue that these differences can be explained on the basis of the differences in the development of the political-administrative systems in the two countries, both when taking a long-term historical perspective and when focusing on recent decades.

  First, we briefly summarize Lijphart’s thesis and argue why his model is unduly limited to the political features of the governance system. We propose the elaboration of seven administrative features that in our view are crucial for truly understanding consensus democracy in fragmented societies. Second, we provide a concise discussion of the historical development of the political-administrative systems of The Netherlands and Belgium. In Section 4 we discuss the implications of the expansion of Lijphart’s model by applying it to the current Dutch and Belgian governance systems. We conclude with a summary of the new insights from our conceptual exercise and suggest avenues for further research.

• 2 Conceptual issues about Lijphart’s model

  One of Lijphart’s central claims is that the quality of democracy and the effectiveness of public policies are better guaranteed in consensus democracies than in majoritarian systems (Lijphart, 1977; 1994; 2002; 2012 & 2013). As a government’s performance is determined largely by the functioning of its bureaucratic apparatus, the quality of the civil service is likely to be an important factor in reaching that higher level of performance (Van der Meer 2011).
  Even in the second reworked edition of Patterns of Democracy (2012), Lijphart refers mainly to democracy and performance dimensions but remains inarticulate when referring to good governance indicators. Such indicators have recently been developed in more detail and with more conceptual clarity, for example by OECD (Government at a glance 201), World Bank (2019), Bertelsmann Foundation (2019) and the Blavatnik School at Oxford University (2019). To briefly summarize Lijphart’s argument (see also Van den Berg, 2018a), democratic systems can be categorized as more majoritarian or more consensus-oriented systems. The consensus model, in its purest form, has, according to Lijphart, ten institutional features, five of which relate to the executive branch of power and the remaining five to the structure of the state.
  With regard to the executive, these are as follows:

  1. An electoral system based on proportional representation;

  2. A multiparty system;

  3. Broad coalition governments;

  4. A strong controlling power of parliament on the government; and

  5. An institutionalized role for civil society groups in the policy process.

  With regard to the state structure, these are as follows:

  1. A federal or decentralized state structure;

  2. A parliament consisting of two more or less equal chambers;

  3. A hard-to-change constitution;

  4. A constitutional court that is key to the reading and interpretation of the constitution; and

  5. A strong and independent central bank.

  Why and how the occurrence of these features would lead to consensus government and the occurrence of their contraries to majoritarian government has been explained in the introduction to this special issue. Therefore, it suffices to state here that consensus systems are designed to share rather than concentrate power and to reach political decisions through power sharing and compromise rather than through adversarial or simple majoritarian decision-making. As a result, consensus democracy is less speedy but supported by a larger segment of society. For a clear understanding of the concept of consensus governance, two further remarks are relevant. First, consensus governance does not mean that there is a priori broad agreement about policies or that there are no conflicts between the involved parties. On the contrary, it means that there are clearly tensions and set-offs, but also that among the groups involved there is always the will and the effort to share power and to search for compromises, whether the contradictions in interests are big or small (Vollaard, Beyers & Dumont, 2015). Second, and related to this, consensus governance should not be taken to mean that dominant parties can impose their interest or vision on others and that every alternative direction is placed outside of the order of discussion.

  What becomes obvious is the absence of attention to the possible importance of specific types of bureaucratic models and arrangements in a given system for consensus governance. While we know that the nature and functioning of the civil service is an important factor in government performance (among others, OECD, 2018; Pollitt & Bouckaert, 2017), this remains an important blind spot in Lijphart’s approach. In this sense, Lijphart’s view of government appears to be similar to traditional constitutional legal thinking, where a strict separation between policy (politics) and implementation (administration) is believed to exist. In this train of thought, it is assumed that when the political institutions have made their decisions, implementation of these decisions will follow automatically in a rather straightforward manner. However, it is well known in the administrative sciences that during implementation binding decisions have to be made concrete and/or developed given the framework character of regulation (Lipsky, 1980; Pressman & Wildavsky, 1973). That implies a crucial role for bureaucracy not only in achieving results – thus in performance terms – but also in taking care of procedural aspects – thus in terms of responsiveness and democracy (see also Denhardt & Denhardt [2002], who argue in their New Public Service approach that this duty is to be considered a ubiquitous requirement of civil service).
  Therefore, in analysing the level of performance in both majoritarian and consensus-styled democracies, assessing the role, attitude and expertise of the civil service is of crucial importance. Moreover, if there are two institutional settings in which this is particularly the case, then it will be the Belgian and Dutch systems, given their pillarized and consensual structure of government and society dating back to previous centuries. As we shall argue later, the pillarized societal ordering created a societal and political institutional mould, in which the bureaucracy had to operate not only in the service of the political office holders of the day, but also in the service of society.
  We therefore criticize Lijphart’s work for underemphasizing the importance of the civil service arrangements in consensus systems. In order to remedy this shortcoming we look to the works of another well-known Dutch political scientist and founding father of (international and historical) comparative political research, Hans Daalder (1964; 1974a; 1974b; 1995). Lijphart and Daalder were contemporaries, fellow countrymen and colleagues at Leiden University from 1963 until 1968 and entertained a lively dialogue on the nature and meaning of consensus governance (Daalder, 1997). Daalder was – more so than Lijphart – sensitive to the administrative infrastructure. Following Daalder, we argue that specific characteristics of the bureaucratic system generate proportionality, depoliticization and pragmatism into the policy process, so as to enable the political actors to sustain and reconfirm the consensual mode of decision-making. Building on Daalder, Van den Berg, and Tromp (1990) and Van den Berg (2018a), we therefore propose to extend and refine Lijphart’s model, adding a third cluster of seven characteristics (administrative dimensions) that relate to the administration, pertaining to the various stages of the policy cycle (Howlett, Ramesh & Perl, 2009):

  1. A professional civil service based and recruited primarily on merit criteria (policy formulation and implementation);

  2. A civil service that supports the incumbent executive loyally but not from a party-political or ideology-dominated attitude (policy formulation and implementation);

  3. A civil service that operates as public servants, not as subservient followers or completely autonomous professionals (policy formulation and implementation);

  4. A public service that functions as a liaison between politics and society and serves both (policy formulation and implementation);

  5. A policy process that at the start is open to expertise from independent advisory councils, planning agencies and academic sources, as well as interest and pressure groups and civil society (agenda setting and policy formulation);

  6. A system of bodies for the implementation of policies, such as civil society organizations (policy implementation);

  7. A professional depoliticized bureaucratic system embodied in institutions and organizations at the end of the policy process, such as inspections and audit entities, research councils and committees evaluating and assessing results (policy evaluation).

  These seven public administration-related elements serve as the intervertebral discs between the aforementioned political characteristics with regard to the executive and the state structure. It is as if they are made of cartilage: somewhat elastic and thereby contributing to the shock absorbing capacity of the body as a whole. Under normal circumstances, they are not clearly seen or felt. But if they wear off or come to clog a nervous pathway, as in the case of a hernia, the pain is so severe that they can disable a great deal of physical functioning.
  Lijphart argues that in fragmented and segregated societies, consensus democracy helps to mediate conflicting interests and prevent violent conflict. In this sense, Lijphart views consensus democracy as not only empirically but also normatively superior to majoritarian rule. Given that religious and cultural cleavages in The Netherlands also follow territorial patterns to a certain degree, peaceful coexistence of multiple groups could also have been reached by means of a federal state structure (Toonen, 2005). However, when the emancipation of religious groups and of workers was completed in the Grand Bargain of 1917 (which regulated both universal suffrage and state funding of religion-based education), the governance model that emerged was a highly developed pillarized consensus (consociational) model within a decentralized unitary state. In Belgium the recent federalization of the state along language lines can be seen as an alternative way to make deep divisions politically manageable (see likewise Bosnia, the United Kingdom, etc.). Interestingly enough, Belgian federalization completed rather than substituted the country’s consociational set-up. A constitutional court is usually created as a support for a federal institutional design (as a referee) or after major conflicts such as after the South European dictatorial regimes and a way imposed in post-war Germany. Similarly, the creation of the Dutch bicameral parliament was the result of Belgian suspicion of domination by the North when the Kingdom of the Netherlands was formed in 1815. What remains is the issue of causality. Do these institutional features add to consensus building, or is having a need for consensus at the root of creating these institutions? Or might they even be mutually reinforcing themselves?

• 3 Historical context: Long-term and mid-range dynamics in The Netherlands and Belgium

  Lijphart’s study on The Politics of Accommodation (Verzuiling, pacificatie en kentering in de Nederlandse politiek, 1968) has been the stepping stone to his later work. The timing of this study is not too remarkable since the pillarized system as embodied in the political, socio-economic and sociocultural system came under pressure from the 1960s onwards. In both The Netherlands and Belgium this has led to the end of the pillarized version of the countries’ consensual design. Consociationalism as a form of consensus democracy (Andeweg, 2000) really pertained to a societal living apart together. The limited attention that is directed by Lijphart to the bureaucratic dimensions of the performance and functioning of the state is mirrored in the general state of affairs in political science and, for that matter, also in other administrative sciences at the time such as constitutional and administrative law. In The Netherlands and Belgium, public administration was still at a nascent stage and rather instrumental in nature. We can name exceptions like Gerrit van Poelje, Joris In’t Veld and later Aris van Braam, Antonie Klein, but also Henk Brasz who directed attention to the local level. In Belgium, one may think, for instance, of Andre Molitor, Hugo van Hasselt, Frank Delmartino and Roger Depré. Yet, despite these exceptions, research in public administration at that time was less well developed than that in political science and constitutional law, but still in public administration more attention was paid to bureaucracy than has been the case in Lijphart’s work.
  Many studies on consociationalism trace the Belgian and Dutch administrative tradition of consensual decision-making and the high degree of self-regulation by means of local organizations back to earlier, pre-modern times, to the early Middle Ages. Churches, diaconates, water authorities, guilds, militia and trading companies were the predecessors of what later became known as the civil society. On this age-old foundation the pillarized society was later built, a political-societal ordering of demarcated groups based on religious and ideological characteristics. Both Daalder (1974a; 1974b) and Groenveld (1995) have pointed out in their Leiden inaugural lectures that even in the 17th century we could find this politics of accommodation. It is important to note that in this sense, consensus governance predated democracy in the Low Countries. It is traced back even further to the struggle against the water, the origins of the water boards in the Middle Ages (Toonen, Van der Meer & Dijkstra, 2006). This goes for The Netherlands as much as it does for Belgium, given that the idea of a separation between the northern and the southern Netherlands did not come about before the second half of the 16th century. The separation into the Northern Netherlands and the Southern Netherlands resulted from the so-called ‘80 years war’ (1579-1648), after which both polities developed their own individual brands of consensus governance, each accommodating the particular political and social cleavages prevalent in their societies.
  Given the absence of detailed administrative historical and biographical studies, we need to reconstruct the bureaucratic contribution to decision-making. First, the role of bureaucracy in this consensus system – also in the consociational (pillarized) version – has been a supportive one, advising on policies and supporting implementation of the consensus reached. The bureaucracy has traditionally been merit based and supportive of the (majority) coalition of the day at the national level and its mirror image coalitions at the local and provincial level. Thus, bureaucracy was largely party-politically neutral and, as such, reasonably in accordance with what later would become known as the Weberian ideal-type. Even though from the very outset of the modern Dutch (1815; 1848) and Belgian state (1830), compartmentalization within and between government departments was mentioned as a problem hampering government operations (Van IJsselmuiden, 1988; Van Leeuwen, 1912), at the end of the day governance by consensus was also the dominant feature of this departmental system. What has been said of central government also applies to the other domestic levels of government (Van der Meer, 2018; Van der Meer, Raadschelders & Kerkhoff, 2011). In addition, when speaking of a pillarized system of governance and public service delivery, roughly in force from the late 19th century to the 1970s, we could more appropriately call it a pillarized system of multilevel governance with a need for public and private actors in pillarized (societal) organizations on different levels of scale to work together.
  Even though Belgium and The Netherlands have experienced a process of depillarization, the basic bureaucratic function of supporting consensus politics remained the same. The key difference between the pillarized and de-pillarized system of public service delivery was the absence in the latter of a natural link between public service delivery organizations and the communities they serve. That natural connection was replaced by a form of (bureaucratic) managerialism, which proved to have little to offer as an alternative in terms of accountability vis-à-vis the community (Van den Berg, 2018b; Van der Meer & Dijkstra, 2013). From this perspective, bureaucracy has partially replaced the political function and has become more dominant not only in the policy process but also in relations with societal actors within the multilevel governance system. In addition, in Belgium the protracted struggle for more regional (economic) autonomy and language emancipation resulted in a federalization of politics and government with effects on bureaucracy.
  The Belgian decentralization was a historical process that started in 1970 and took over four decades and six fundamental state reforms. Its origins lie in the tensions between the originally predominantly French-speaking central government and the economically dominant Dutch-speaking population that makes up over half of all Belgians. Gradually, the state reforms installed three communities and three territorial regions. At present, each has proper competences, directly elected parliaments and their own administration (Witte & Meynen, 2006). This growing apart has some similarities to consociationalism of old, yet with dividing lines linked not to political religious but now to community and language identity.
  Next, for both the Belgian and the Dutch cases, we will discuss the civil service arrangements in consensus systems according to the seven characteristics regarding bureaucracy as presented earlier. For the Belgian case, for reasons of focus and clarity, we will focus primarily on the federal level.

• 4 Belgium and The Netherlands on an amended bureaucratic Lijphart framework

  4.1 Administrative dimensions 1, 2, 3, 4

  Discussing the dimensions of administration, when looking at the role and position of bureaucracy within Belgium, it has often been observed that the civil service is party-politically politicized in terms of appointment and careers (Brans & Hondeghem, 1999; Hondeghem, 2000). In addition to this politicized permanent bureaucracy, ministerial cabinets are extensively filled by political appointees involved in (substantive) policymaking (van Hassel, 1988). Recent years have witnessed efforts and reform initiatives aimed at placing more emphasis in the appointment procedures on the professionalization of the higher echelons in the permanent civil service. The career-based system is slowly incorporating elements of a position-based system based on new public management considerations (Bourgault & Van Dorpe, 2013; Hondeghem & Van Dorpe, 2013). Yet, as Pelgrims (2007) has observed, once candidates pass the assessment (tests), their appointment is often based more on political than on merit considerations. As a result, Belgium scores significantly lower than other EU countries on the EUPACK 2018 professionalism index. This index provides an indication of the extent to which a public administration is deemed more professionally oriented rather than politicized (Thijs, Hammerschmid & Palaric, 2018). In fact, the advancement to senior management positions is still a political decision made by the government of the day after an (merit) assessment of potential candidates. According to OECD standards, the appointment procedure (and its results) is not transparent, owing to political interference (Van Dooren, 2018). Belgian civil service employees are in touch with the political realities in the country and take these into account when conducting their work. Nevertheless, they operate as public servants, not as subservient followers of the political leadership of the day or as completely autonomous professionals.
  In the Dutch case there is (party-politically) depoliticized professional civil service at the various (hierarchical) levels within the bureaucracy. In the rule, the Dutch civil service supports the incumbent political office holders loyally (Bekker, 2012; Van den Berg, 2018b). That neutrality and depoliticization relates not only to the dimensions of a party-political recruitment and a partisan role of civil servants but also to the general absence of recruitment and appointment based on (‘forced’) policy congruence between office holders and bureaucracy (Dijkstra & Van der Meer, 2011). This (party) political neutrality and formal appointment has been the predominant state of affairs from the inception of the Kingdom of the Netherlands to date (Raadschelders & Van der Meer 1999; Rosenthal, 1979; 1983; Van der Meer, 1999). We emphasize the word predominant, in view of the few historical exceptions when political changes in the senior civil service have been made, for instance in the early stage of the Thorbeckian reforms around the middle of the 19th century.
  Although formal party politicization has never been the rule for the senior civil service appointments, there have been suggestions of informal politicization since the 1970s – though proof has been circumstantial (Raadschelders & Van der Meer, 1998; 1999; 2014). The same applies to a somewhat higher degree of policy-oriented politicization (Dijkstra & Van der Meer, 2011; Raadschelders & Van der Meer, 2014). Again, there are some cases of top civil servants recruited who had ideas compatible with the leading policy ideology of the day, but these cases have been limited in number. On the other hand, during the last few decades of the 20th century, attention (in public administration) has been focused on the emergence of proactive public entrepreneurs with their own policy agenda (Daalder, 1995; De Vries & Van Dam 1998). But this too has been the exception rather than the rule. Empirical evidence for the high degree of civil service neutrality and expertise can be found in the data presented in the Government at a Glance (2017) and the EUPACK reports (Thijs et al., 2018; Van der Meer, 2018). In addition, the Dutch civil service operates as public servants, not as subservient subordinates to the political office holder in charge or as entirely autonomous professionals. Perhaps not surprisingly, the public satisfaction with actual public service, from education to police and courts, is quite high (OECD, 2017), and so is the trust in government, ranking second in the EU 28 (Thijs et al., 2018; Van der Meer, 2018). Thus, the Dutch public service is performing its role to society as well as politics as a liaison that is continuously in touch with both of these worlds (Van der Meer, Van den Berg & Dijkstra, 2012).

  4.2 Administrative dimension 5

  In terms of a policy process that is open to external expertise, the Belgian political-administrative system has a long-established tradition of consulting trade unions, health insurers, associations of pharmacies and doctors, farmer’s organizations, associations of educational networks, environmentalists and other interest groups. At the same time, pillarization is, unlike in The Netherlands, still embedded in Belgium’s socio-economic life. From healthcare to insurance, down to trade unions and education, there are Catholic, Socialist as well as Liberal networks for consultation. All policy sectors have extensive consultation structures that established (admitted) civil society organizations have access to (Van Dooren, 2018). In addition, Belgian regulators are – compared with those of other OECD countries – among the least independent (OECD, 2017). With the exception of the economic planning bureau and knowledge centres for healthcare, universities are mostly assigned and commissioned with policy (evaluation) research. However, university research has a higher impact on agenda setting than on the phase of policy formulation (OECD, 2017; Van Dooren, 2018). The real policy decisions are made in the ministerial cabinets, which are large in size and power. Their influence also explains the limited role of the permanent bureaucracy (De Visscher & Solomonsen, 2013).
  In The Netherlands, the inclusion and input of interest and pressure groups in policymaking is and has been extensive. That has also been a crucial feature of the pillarized system and has remained so after depillarization. The main difference is that over time interest group access in the pillarized system was predominantly reserved for the neo-corporatist groups, and that changed to a more pluralized set-up. As one of the implications, civil servants operating within the relevant policy networks had to change their way of operation in order to acquire support; hence the search for new forms of network governance and civil service professionalism (see Hart, 2014). That openness also involves a widespread demand for external academic (and scientific) knowledge as input to policymaking and reform programmes. The existence and availability of think tanks, knowledge centres and university units specialized in public administration has similar features to the US situation (Van der Meer, Dijkstra & Kerkhoff, 2016). This openness makes Dutch government quite dependent on external expertise, an issue that has recently caused some concern. There is also an uneasiness involving the perception concerning an erosion of the knowledge base within government and bureaucracy. The latter is understood to be a negative side effect of the hiving off of departmental units, through staff cutbacks and hiring of external expertise since the early 1980s (Van den Berg, Schmidt & van Eijk, 2015a).

  4.3 Administrative dimension 6

  With respect to policy implementation, Belgium relies heavily on the support of civil society organizations. There is a long history of a gradual institutionalization of (private) civil society initiatives in a (post-)corporatist welfare state (Pauly, Verschuere & De Rynck, 2018). In regard to education, for instance, Belgian municipalities set out the rules and standards governing the various educational organizations and networks in exchange for financial sustenance (Huyse, 1983). The situation in healthcare is comparable. For instance, facilities for the elderly are coordinated not only by municipal social services but also largely by the Catholic care network, which is again bound by legal provisions to maintain the same quality standards. Although the influence of the three main pillars in civil society has been in decline, their presence is still evident in many areas of life (Witte & Meynen, 2006).
  In The Netherlands, such an implementation system using civil society organizations has earlier been part and parcel of the old pillarized society. As argued earlier, the old implementation system has been subjected to a process of depillarization that began in the 1960s and 1970s. At first, it had as a side effect a transfer from civil society towards government-run and directed implementation policies. Depillarizing service provision led to semi-autonomous management-run education, healthcare and other welfare organizations, a movement towards more government involvement, and a growth of bureaucracy in terms of staff and rules. Tackling that growth and revitalizing civil society has been a central part of government policy since. This translated into more emphasis on the development of the enabling role of government and bureaucracy towards society through a more decentralized system emphasizing and stimulating civil society initiatives with a liaison role for civil servants as intermediaries (Van der Meer 2012; Van der Meer, Raadschelders & Toonen, 2015). This civil society involvement has recently been revamped and redeveloped through the ‘enhancement’ (by active citizens groups and government) of a so-called participatory society of active citizens and intermediate not-for-profit and civil society organizations (Bussemaker, 2019; Van den Berg et al., 2015b). Thus, the Dutch government depends largely on the involvement of citizens, civil society, non-profit organizations but also on private for-profit organizations to implement policies and provide actual service delivery. That enabling role and the need for citizen involvement are not limited to implementation issues alone but also to the input of citizens and communities, in short, civil society, at the very start of the policy process. Though still in its early stages, ideas such as co-creation and a so-called opt-out – the ‘Right to Challenge’ – that involves a takeover of, in particular, local government tasks by (groups within) the community point in this direction.

  4.4 Administrative dimension 7

  With respect to depoliticizing administrative institutions positioned at the final stages of the policy process, a major actor in Belgium is the Supreme Audit Office, which operates as an external auditor for the federal government, the regional and municipal governments and the provinces. The Audit Office advises on the budget, checks the accounts and monitors the legality of public expenses. Furthermore, “it engages in performance audits which report on economy, efficiency, and effectiveness of the implementation of public programs” (Van Dooren, 2018). These audits weigh in on the (political) assessment of policies and service delivery, and their impact is considered to be substantial (Desmedt, Morin, Pattyn & Brans, 2017). The overall level of transparency and accountability is high compared with the case in other EU countries (Thijs et al., 2018). Furthermore, adding to these transparency concerns, autonomous bodies have been created that are responsible for assisting whistle-blowing and the maintenance of integrity. The internal audit and control of the executive has a partial autonomous organizational status.

  A wide range of inspections, regulators, watchdogs and evaluators serve as examples of Dutch (depoliticizing) administrative institutions located at the final stages of the policy process. An important role is played by the National Audit Chamber. The reports of the audit chamber have broadened in scope in recent years from reporting primarily on the legality of expenditure to the inclusion of such criteria as effectiveness and efficiency. The latter tasks pertain not only to departmental policy programmes but also to system-wide performance and reform programmes (Van der Meer, 2018). Both the number and operational range of these institutions have increased since the 1980s, and their autonomy (at least in principle) has been strengthened. Some (inspectorates) have been consolidated in larger, more sustainable units in order to increase their operational force. Depoliticization and a more apolitical management-oriented approach have strengthened over the years, with new public management-related accountability, performance and management control systems. Nevertheless, cutbacks in budgets and staff have reduced the effects of these institutional capacity-reinforcing objectives.

  4.5 Quality of bureaucracy

  Finally, the perception of a high-quality bureaucracy in Belgium is coming to light in the positive citizens’ assessment of the standards of service delivery. For instance, the citizens’ satisfaction with healthcare provision in Belgium is the second highest of all OECD countries, with a satisfaction rate of 90% compared with an OECD average of 70%. The approval rate (80%) of the performance of the educational system ranks fourth in all OECD countries. The confidence in the judicial system and the courts meets the OECD average of 50% (OECD, 2017). Other than the highly perceived quality of provided services, this favourable appraisal of public service delivery can also be explained by a good access to services compared with the case in other OECD countries. This seems to accord with the observation that the access to and the size of public service delivery are relatively high in consensus democracies in order to accommodate the different parties involved. Belgium belongs to the top five OECD countries in respect to the affordability of healthcare, education and public justice (OECD, 2017).

  In IMF, OECD and EUPACK reports and in transparency indices, the Dutch scores on good governance rankings are among the highest in EU28 as well as worldwide (Dijkstra & van der Meer, 2012; Thijs et al., 2018; Van der Meer, 2018). The Worldwide Governance Indicators, for instance, show a top-ranking position on issues of voice and accountability (4), government effectiveness (9), regulatory quality (4), rule of law (4), control of corruption (12) and political stability, and a relatively lower position on the indicator of absence of violence (47). The latter is perhaps not surprising given the recent rise in political turbulence and, to a certain extent, political violence after 2002, starting with the assassination of Pim Fortuyn and, later, of Theo van Gogh. The public satisfaction with public service in areas such as education, police and courts is high when compared with the levels in other countries (see OECD, 2017; Thijs et al., 2018). Similarly, the distance of public servants towards society is relatively small (see OECD 2017; Thijs et al., 2018). The Dutch public service has a service attitude that tries to facilitate both politics and society and serves as a liaison (Van der Meer, Van den Berg & Dijkstra, 2012; Van der Meer et al., 2015; Dijkstra et al 2015). This is notwithstanding criticism from groups within society on bureaucratism within the public sector. Yet that criticism is directed more at government as a whole and/or government as an abstract concept rather than at specific actors in real-world public service delivery. This equates with a predominantly egalitarian society in search of consensus and pragmatism in order to overcome political and societal cleavages.

• 5 Conclusion

  Revisiting Lijphart’s book on the performance of consensus democracies is of particular interest inasmuch as the Low Countries are often hailed as examples of consociational expressions of consensus democracies. It is notable that Lijphart, as a true classical Dutch political scientist, hardly refers to the role of bureaucracy in his theoretical framework. Bureaucracy does not seem to matter as a core actor. In his fundamental question of how and why consensus democracy is performing better than a majoritarian system, the role of bureaucracy remains a black box. That might be considered a pity, but it has left us with the opportunity to amend this deficit – or, more appropriately, blind spot – in Lijphart’s work. In analysing the performance of either majoritarian or consensus-styled democracies the role, attitude and expertise of the civil service stand central. From a theoretical perspective we have therefore added to Lijphart’s overview of structural aspects of the executive and the state a classification of seven features of (the position and internal and external relationships of) bureaucracy within consensus states.
  Central to our argument is the re-examination of what the role of bureaucracy is and was within two (changing) consensus states: Belgium and The Netherlands. Using the latter description, however, raises the question of whether both countries can still be termed a consensus state. Though we argue that for both countries this can still indeed be done, the relatively more complete and speedier erosion of consociationalism in The Netherlands compared with Belgium is perhaps best understood by the nature and persistence of societal cleavages. Whereas in both countries religion has become less important as a defining cleavage, language is nevertheless still a prevailing dividing line in Belgium. In The Netherlands societal organizations (social organizations, trade unions, employers’ organizations, etc., with the exception of education and public broadcasting) have been depillarized, unlike in Belgium. According to Lijphart’s classification, Belgium and The Netherlands can still be characterized as predominantly consensus democracies, though perhaps weakened by the political and societal crises since the early 2000s. The consensus features still remain dominant, though there are pressures in both countries given the rise of populism through the rise of new (so-called populist) right-wing and left-wing parties. This development puts altogether more emphasis and reliance on the role of the civil servants as mediating actors (Van der Meer et al., 2015). The role of officials within both Belgian and Dutch governments is therefore still conceived as self-confident and autonomous. Yet, at the same time, being professional ‘public servants’ they are bound to politics and society (as formulated by Van der Meer et al., 2012), and they are operating within the context of the democratic rule of law. A difference is that in the Dutch case we have, in essence, a party-political neutral civil service (Raadschelders & Van der Meer, 2014), whereas in Belgium, party-political politicization, in particular in the policymaking echelons, has of old been more dominant. In Belgium, politicization cannot be conceived as ‘the winner takes all’, but as a result of accommodating the different segments in society and thus as a precondition for a consensus democracy (see also Van der Meer et al., 2015). In addition, Hondeghem (2000; following Dierickx & Majershof, 1993) has argued that Belgian civil servants consider themselves primarily as servants of the state rather than as servants to a political programme. This can be explained by differences in the institutional design, accommodating different versions of a pillarized society. That institutional design has been a persistent inheritance of the past. Nevertheless, both in the Belgian and in the Dutch cases we can detect a seminal role for bureaucracy as, in general (and more specifically excluding politicization in the Belgian case), it fits the bureaucratic features we added to the Lijphart model. Thus, we conclude, those bureaucratic features have over time become even more relevant as an addition to the original Lijphart model.

References

• Andeweg, R.B. (2000). Consociational democracy. Annual Review of Political Science, 3, 509-536.

• Bekker, R. (2012),. Marathonlopers rond het binnenhof. Amsterdam: Boom Lemma.

• Bertelsmann Stiftung. (2019). Sustainable governance indicators. Retrieved 19 February, 2019, from http://www.sgi-network.org/2018/.

• Blavatnik School of Government. (2019). International civil service effectiveness index. Retrieved 19 February 2019, from https://www.bsg.ox.ac.uk/about/partnerships/international-civil-service-effectiveness-index.

• Bourgault, J. & Van Dorpe, K. (2013). Managerial reforms, public service bargains and top civil servant identity. International Review of Administrative Sciences, 79(1), 49-70.

• Brans, M. & A. Hondeghem (1999). The senior civil service in Belgium. In V. Wright & E.C. Page (Eds.), Bureaucratic elites in Western European states. Oxford: Oxford University Press.

• Bussemaker, M. (2019). De zorg als sociale kwestie. Inaugural lecture Leiden University, 15th February 2019. Retrieved 21 February 2019, from https://www.lumc.nl/over-het-lumc/hoo/oraties-redes/2019/Oratie-Bussemaker/ .

• Daalder, H. (1964). Leiding en Lijdelijkheid in de Nederlandse Politiek [Elites and Apathy in Dutch Politics]. Assen, The Netherlands: Van Gorcum.

• Daalder, H. (1974, July). The consociational democracy theme. World Politics, 26(4), 604-621.

• Daalder, H. (1974). Politisering en lijdelijkheid in de Nederlandse politiek. Assen, The Netherlands: Van Gorcum.

• Daalder, H. (1995). Van Oude en Nieuwe Regenten. Politiek in Nederland. Amsterdam: Bert Bakker.

• Daalder, H. (1997). Universitair Panopticum. Herinneringen van een gewoon hoogleraar. Amsterdam, The Netherlands: De Arbeiderspers.

• Daalder, H., van den Berg, J. & Tromp, B.A. (1990). Politiek en historie: Opstellen over Nederlandse politiek en vergelijkende politieke wetenschap: Opstellen over Nederlandse politiek en vergelijkende politieke wetenschap. Amsterdam, The Netherlands: Bert Bakker.

• De Visscher, C. & Salomonsen, H.H. (2013). Explaining differences in ministerial ménages à trois: Multiple bargains in Belgium and Denmark. International Review of Administrative Sciences, 79(1), 71-90.

• De Vries, J. & Van Dam, M. (1998). Politiek-bestuurlijk management, Een blik achter de gouden muur. Alphen a/d Rijn, The Netherlands: Samson.

• Denhardt, J. V. & Denhardt, R. B. (2002). The new public service: Serving, not steering. Armonk, NJ: M.J. Sharpe.

• Desmedt, E., Morin, D., Pattyn, V. & Brans, M. (2017). Impact of performance audit on the administration: A Belgian study (2005-2010). Managerial Auditing Journal, 32(3), 251-275. doi:10.1108/MAJ-04-2016-1368

• Dierickx, G. & Majersdorf, P. (1993). Politieke cultuur van ambtenaren en politici in België. Brugge, Belgium: Vanden Broele.

• Dijkstra, G.S.A. & van der Meer, F.M. (2011). The civil service system of the Netherlands. In F.M. van der Meer (Ed.), Civil service systems in Western Europe (pp. 149-185). Cheltenham/Aldershot, MA: Edward Elgar.

• Dijkstra, G.S.A., van der Meer, F.M. & van den Berg, C.F. (2015). Is er nu wel of niet sprake van een terugtredende overheid? In J.J.M. Uijlenbroek (Ed.), Staat van de ambtelijke Dienst (pp. 23-37). Den Haag, the Netherlands: CAOP.

• Groenveld, S. (1995). Huisgenoten des geloofs. Was de samenleving in de Republiek der Verenigde Nederlanden Verzuild. Hilversum, The Netherlands: Verloren.

• Hart, P. ‘t. (2014). Ambtelijk Vakmanschap 3.0: Zoektocht naar het Handwerk van de Overheidsmanager, Essay in opdracht van de Vereniging voor OverheidsManagement, Vereniging van Gemeentesecretarissen en de stichting IKPOB. Den Haag, The Netherlands: NSOB.

• Hondeghem, A. (2000). The national civil service in Belgium. In A.G.J.M. Bekke & F.M. van der Meer (Eds.), Civil service systems in Western Europe. Cheltenham, MA: Edward Elgar.

• Hondeghem, A. & Van Dorpe, K. (2013). Performance management systems for senior civil servants: how strong is the managerial public service bargain? International Review of Administrative Sciences, 79(1), 9-27.

• Howlett, M., Ramesh, M. & Perl, A. (2009). Studying public policy: Policy cycles and policy subsystems (Vol. 3). Oxford: Oxford University Press.

• Huyse, L. (1983). Breuklijnen in de Belgische samenleving. In L. Huyse & J. Berting (eds), Als in een spiegel Een sociologische kaart van België en Nederland (pp. 9-25). Leuven: Kritak.

• Lijphart, A. (1968). The politics of accommodation. Pluralism and democracy in the Netherlands. Berkeley, CA: University of California Press.

• Lijphart, A. (1977). Democracy in plural societies: A comparative exploration. New Haven, CT: Yale University Press.

• Lijphart, A. (1994). Electoral systems and party systems: A study of twenty-seven democracies, 1945-1990. Oxford: Oxford University Press.

• Lijphart, A. (1999). Patterns of democracy: Government forms & performance in thirty-six countries. New Haven, CT: Yale University Press.

• Lijphart, A. (2012). Patterns of democracy: Government forms & performance in thirty-six countries (2nd ed.). New Haven, CT: Yale University Press.

• Lijphart, A. (2013). Steps in my research and thinking about power sharing and democratic institutions. Taiwan Journal of Democracy, (Special Issue), 1-7.

• Lipsky, M. (1980). Street-level bureaucracy: dilemmas of the individual in public services. New York: Russell Sage Foundation

• OECD. (2017). Government at a glance 2017. Retrieved from http://dx.doi.org/10.1787/22214399.

• Pauly, R., Verschuere, B. & De Rynck, F. (2018). Steering civil society towards the market? An analysis of organisational strategies in the governance arrangement of WISE in Flanders. In 14th Trans-Atlantic Dialogue, Bucharest (Romenia).

• Pelgrims, C. (2007). The impact of politicians on administrative reform processes? Analyzing the introduction and implementation of mandates for top civil servants. Paper prepared for the Leading the Future of the Public Sector: The Third Transatlantic Dialogue University of Delaware, Newark, Delaware, USA, May 31-June 2, 2007.

• Pollitt, C. & Bouckaert, G. (2017). Public Management Reform: A Comparative Analysis. Into the Age of Austerity. Oxford: Oxford University Press.

• Pressman, J.L. & Wildavsky, A. (1973). Implementation: How Great Expectations in Washington are Dashed in Oakland; or, Why it’s Amazing that Federal Programs Work at All. Berkeley, Los Angeles, London: University of California Press.

• Raadschelders, J.C.N. & van der Meer, F.M. (1999). The senior civil service in the Netherlands: A quest for unity. In V. Wright & E.C. Page (Eds.), Bureaucratic elites in Western European states. Oxford: Oxford University Press.

• Raadschelders, J.C.N. & van der Meer, F.M. (2014). The social profile of administrative elites in the Netherlands, 1980-2010. In International Review of Administrative Sciences .

• Rosanvallon P. (2018). Good government: Democracy beyond elections. Cambridge, MA: Harvard University Press.

• Rosenthal, U. (1979). De secretarissen-generaal: politisering of verambtelijking? Acta Politica, 343-377.

• Rosenthal, U. (1983). De mandarijnen van de rijksdienst.; modieuze stellingen en harde feiten over de Nederlandse topambtenarij, Bestuurswetenschappen, 302-315.

• Thijs, N., Hammerschmid, G. & Palaric, E. (2018). A comparative overview of public administration characteristics and performance in EU28. Brussels, Belgium: Directorate-General for Employment, Social Affairs and Inclusion under the Contract VC/2016/0492 “Support for developing better country knowledge on public administration and institutional capacity building”.

• Toonen, Th.A.J. (2005) Theorie van de Provincie. The Hague: IPO.

• Toonen, Th.A.J., Van der Meer, F.M., Dijkstra, G.S.A. (2006). Modernization and reform of Dutch Waterboards: resilience or change? In: Journal of Institutional Economics 2(2), 181-201.

• Van den Berg, C.F. (2018a). Van Zuilen naar Bubbels: de toekomst van consensusbestuur in een geglobaliseerde samenleving. Groningen: Rijksuniversiteit Groningen.

• Van den Berg, C.F. (2018b). The Netherlands: The emergence and encapsulation of ministerial advisers. In R. Shaw & C. Eichbaum (Eds.), Ministers, minders and mandarins: An international study of relationships at the executive summit of parliamentary democracies. Cheltenham, MA: Edward Elgar.

• Van den Berg, C.F., Schmidt, A.J. & van Eijk, C.J.A. (2015a). Externe advisering binnen de Nederlandse overheid: Naar een empirisch en theoretisch onderbouwde onderzoeksagenda. Bestuurskunde, 24(3), 17-31.

• Van den Berg, C.F., van der Meer, F.M., Mannekes, M.V., Osch, D.V., Porth, J. & Schmidt, A. (2015b). Koers houden in turbulentie: De rol van de rijksoverheid op het gebied van infrastructuur en milieu internationaal vergeleken. Publicatiereeks overheid & arbeid. Den Haag: CAOP.

• Van der Meer, F.M. (Ed.). (2011). Civil service systems in Western Europe. Cheltenham/Aldershot: Edward Elgar.

• Van der Meer, F.M. (2012). Voorwaarden, waarborgen en ambtenaren. De gevolgen van de opkomst van de voorwaardenscheppende staat voor de publieke dienst. Inaugural lecture. Leiden: Leiden University.

• Van der Meer, F.M. (2018). Public administration characteristics and performance in EU28: the Netherlands. In N. Thijs & G. Hammerschmid (Eds.), The Public Administration in the EU 28. Brussel, Belgium: Directorate-General for Employment, Social Affairs and Inclusion under the Contract VC/2016/0492 “Support for developing better country knowledge on public administration and institutional capacity building”.

• Van der Meer, F.M. & Dijkstra, G.S.A. (2013). Hervormingen publieke sector en de democratische rechtstaat: consequenties voor de publieke dienst. Tijdschrift voor Openbare Financiën, 41, 41-48.

• Van der Meer, F.M., Dijkstra, G.S.A. & Kerkhoff, A.D.N. (2016). What does the changing institutional design of the state do to the recruitment, merit requirements and deployment of (senior) civil servants? Recruitment and deployment in a changing decentralized civil service system in the Netherland 1814-2016. Administory Journal for Administrative History/ Zeitschrift für Verwaltungsgeschichte, October 2016, 1-24.

• Van der Meer, F.M. & Raadschelders, J.C.N. (1998). Politicisation or normal situation? Civil servants and politics in The Netherlands. Revue Française d’Administration Publique, 86.

• Van der Meer, F.M. & Raadschelders, J.C.N. (2007). From the active to the enabling state: The changing role of the senior civil service in Dutch national government. In E.C. Page & V. Wright (Eds.), The changing role of senior service in Europe. Houndmills: Palgrave.

• Van der Meer, F.M, Raadschelders, J.C.N. & Kerkhoff, T.D.N. (2011). Van nachtwakersstaat naar waarborgstaat. Proliferatie en vervlechting van het Nederlandse openbaarbestuur in de lange twintigste eeuw (1880-2005). In F.P. Wagenaar, M.R. Rutgers & A.D.N. Kerkhoff (Eds.), Duizend jaar openbaar bestuur in Nederland: Van patrimoniaal bestuur naar waarborgstaat (pp. 221-290). Bussum: Coutinho.

• Van der Meer, F.M., Van den Berg, C.F. & Dijkstra, G.S.A. (2012) Het eigene van de overheid en haar personeel in een systeem van multi-level governance: naar een nieuwe ambtelijke status, Leiden: LUP Academic Series, Leiden University Press.

• Van der Meer, F.M., Raadschelders, J.C.N. & Toonen, Th.A.J. (2015). Comparative civil service systems: Comparative perspectives. Houndsmill: Palgrave/McMillan.

• Van der Meer, F.M. & Roborgh, R. (1993). Ambtenaren in Nederland, Omvang, bureaucratisering en representativiteit. Alphen aan den Rijn: Samsom Tjeenk Willink.

• Van Dooren, W. (2018). Public administration characteristics and performance in EU28: Belgium. In N. Thijs & G. Hammerschmid (Eds.), The Public Administration in the EU 28, Brussel, Belgium: Directorate-General for Employment, Social Affairs and Inclusion under the Contract VC/2016/0492 “Support for developing better country knowledge on public administration and institutional capacity building”.

• Van Hassel, H. (1988), Het kabinetssyndroom in historisch perspectief. Gemeentekrediet van België, 42, 11-36.

• Van IJsselmuiden, P.G. (1988). Binnenlandse Zaken en het ontstaan van de moderne overheidsbureaucratie in Nederland 1813-1940. Kampen: Kok.

• Verslag van de Staatscommissie (van Leeuwen). (1912). Over de inwendige organisatie van de departementen van Algemeen Bestuur. Staatscourant 10-08-1910 no. 26 report.

• Vollaard, H., Beyers, J., Dumont, P. (2015). European integration and consensus. Politics in the Low Countries. London: Routledge.

• Witte, E. & Meynen, A. (2006). De geschiedenis van België na 1945. Antwerpen: Standaard Uitgeverij.

• World Bank. (2019). World bank governance indicators. Retrieved 19 February, 2019, from http://info.worldbank.org/governance/wgi/#doc-intro.