Abstract
For the second time in just a few years, Belgium recently faced a profound political and institutional crisis. Since constitutions are the ultimate means of building and sustaining a just and stable politico-institutional order, these crises raise the question of what role the constitution plays in channeling and/or constraining the political state of affairs. This is a most pressing topic, especially since the dominant theory about why and how countries such as Belgium are able to function as stable polities also claims to be indicative of the democratic quality of these countries. In this article, it is submitted that the Belgian case can indeed be instructive in telling us a bit more about the constitutional conditions that are, if nothing else, at least not antagonistic to societal stability. The thesis proposed in this article is that for creating a stable society, there should be a connection between what is called negative and positive constitutionalism (S. Holmes): not only should a constitution disable political decision-making by building procedural roadblocks or by enacting bills of rights (negative constitutionalism); it can also help create the very demos which governs itself through the constitutional regime by including incentives for politicians to cooperate (positive constitutionalism). The Belgian Constitution has failed in connecting these types of constitutionalism. And since partition and secession, which might be the result of all this, usually come with an array of negative consequences (the possibility of violence being one of them), this is a problematic state of affairs.
1. Introduction
The title of this article is a testimony to the challenge that every academic active in the domain of Belgian public law has to face from time to time. Sooner or later, he or she will be confronted with questions about Belgian politics and its accompanying constitutional state of affairs. One question stands out prominently, and it has a tendency of leaving only bewilderment in the eyes of those raising it. It seems to point to a rather unusual situation, namely: “How does it work?” Explaining Belgian political and constitutional particulars and vicissitudes is no easy task at the best of times, and a foreign readership in particular would likely not see the forest for the trees. We are dealing with a very complex society, divided in many respects: linguistically, ideologically, culturally, regionally, economically, etc. These divisions are politically highly relevant and come with a complex constitutional design. “No country but Belgium offers so many job opportunities to lawyers specialising in constitutional law,” the Belgian Prime Minister Elio di Ruppo recently asserted during a presentation on a new treatise of constitutional law.1
In going back to the aforementioned bewilderment for a moment, one could observe that recently it has significantly amplified. This is due to the fact that for the second time in just a few years, the country has faced profound political and institutional crises. The most recent crisis resulted in Belgium not having even a federal government for 541 (!) days2—a world record, apparently. Therefore, the issue is no longer merely “How does Belgian politics function?,” but also “Does it function at all?”
This last question is of particular constitutional significance. Since constitutions are the ultimate means of building and sustaining a just political order, the Belgian crises just mentioned are also instructive as to whether or not the constitution, in reality, possesses independent or normative value: to what extent is the Belgian Constitution capable of governing the dynamics of the political power process? What role, if any, does it still play in channeling and/or constraining the political state of affairs?
These seem to me to be most pressing questions, especially since some theories about why and how countries such as Belgium are able to function as stable polities, also claim to be indicative of the democratic quality of the state it is meant to fit.3 Therefore, notwithstanding the fact that its political and constitutional ups and downs sometimes have a surrealistic quality about them (René Magritte was Belgian!4), the Belgian case can be relevant outside Belgium. Although every political and constitutional order is in some sense unique, trying to explain this uniqueness might well serve broader theory-building from the beginning.5 In any case, the question of how divided societies can build and sustain a just political order is one of the great questions of our time,6 and here I will deal with how this is attempted in Belgium and what role, if any, the constitution has to play in this.7
I will divide this article into five main sections: Section 2 will describe the constitutional and accompanying institutional organization of Belgium; Section 3 will sketch the historical roots of its current constitutional and institutional design; Section 4 will outline the more specific political drives behind this design; and Section 5 will analyze and evaluate the Belgian politico-societal developments currently challenging the country’s institutional set-up and constitutional shape. Before submitting some final observations in Section 7, I will deal with the question of what all of the preceding means for the state of constitutionalism and democracy in Belgium (Section 6). By approaching the topic of this article in this way, I hope to bring some order into what at first sight may seem like sheer chaos.
The approach taken here builds on both law and political science. Lawyers are used to studying constitutions as legal phenomena, and political scientists, if they do so at all, study constitutions as social phenomena. In this article I aim to integrate these two disciplines.8 “[K]nowledge of the constitutional text alone equate[s], [not] even nearly, to an understanding of political reality,” some colleagues wrote a few years ago. “[B]ut”, they added, “it is a necessary condition of that understanding.”9 I think this indicates reciprocity, and what we should pay attention to is how political reality and constitutional law inform and influence each other. In Sections 5 and 6 some space will be devoted to this reciprocity.
2. Constitutional and institutional layout —“warehouse” federalism and a blocked constitution10
Belgium today has some 10.6 million inhabitants; it is a small and affluent country and is organized around two dominant languages and cultural communities (demoi): the Dutch-speaking part in the north (called Flanders, with a population of some 6.1 million people, referred to as Flemings), and the French-speaking part in the south (called Wallonia, making up 55 percent of the land in the country and with 3.4 million residents known as Walloons). A language border, officially marked by the legislator in 1962–1963 and entrenched in 1970 in Article 4 of the Constitution,11 separates the two communities, and has resulted in a rigid division of the country into linguistic areas. The fact that within such an area a language is recognized as the main official language means that it is also the language of administration, government, education, and justice. As a result, Dutch is the official language in the north, and French in the south. However, Brussels, with some 1.1 million inhabitants, has been constitutionally recognized as a bilingual area—Dutch- and French-speaking—but has a predominantly francophone population.12 In effect, some 60 percent of the Belgian population are Dutch speakers and some 40 percent are French speakers. Constitutional recognition has also been granted to a small German-speaking part of the country (bordering Germany and with some 75,000 inhabitants).
In addition, there are a limited number of municipalities —situated on the language border and also bordering the so-called Brussels-Capital Region (see below)— where a substantial number of Dutch- and French-speaking people reside and where language facilities are granted to those who speak the officially recognized “minority” language, i.e., the language that is not the official language.13 These minorities can use the language of their choice in matters relating to, e.g., administration, schooling of children (6–12 years old), and justice. These municipalities are called “facility municipalities” and are also obliged to communicate with the citizens in the language of the “minority” language speakers’ choice. So, too, do all forms and official communications have to be available in―depending on the language area in which a municipality is situated― either Dutch and French, or German and French. However, in communications between such a municipality, on the one hand, and the regional or federal authorities to which the municipality resorts, on the other hand, the official language is the norm, as this deals with communications within the municipal administration itself. In other words, language facilities are a service for the citizens and not for the administration. To avoid any misunderstandings: individual linguistic liberty is protected by Article 30 of the Constitution (and has been since 1831 when Belgium was established as an independent state), and the aforementioned arrangements are only applicable to the relation between the citizens and the state; the choice of language used in private and social life, including commercial activities, is in principle free.
Belgium, as Article 1 of the Constitution has stated since 1993, “is a federal state composed of Communities and Regions.” Belgian federalism is organized, as is clear from this provision, on the basis of a double subdivision into Communities and Regions,14 each having its own legislative and executive jurisdiction. By way of several state reforms,15 since 1970, there has been a progressive transfer of powers from the federal level to these Communities and Regions. Between these Communities and Regions, on the one hand, and the federal level, on the other, there is no legislative hierarchy; all can enact legislation, for which they are exclusively competent, and all legislation enacted by the Communities and Regions is on a par with federal legislation.
There are three constitutionally recognized Communities, of which the so-called Flemish and French Communities are the most prominent ones. There is also a small German-speaking Community in the eastern part of the country. The Communities are mainly competent in relation to specific subject matters of cultural concern to the respective Dutch, French, and German-speaking populations in Belgium. However, they also have exclusive jurisdiction over what are termed personal matters: certain aspects of healthcare, family policy, education, and so on. This jurisdiction is territorially defined: it stops at the language border.16 For example, the Flemish Community is not competent in the French Community. The situation in Brussels, however, is an exception to the territorial rule. There, the Flemish Community has competence in matters with regard to the Dutch-speaking residents, and the French Community with regard to the French-speaking ones. By way of example: it is the Flemish Community that has competence with respect to schools in Brussels that teach in Dutch, and the French Community with respect to schools that teach in French.
Belgium is further divided into three Regions: the Flemish, the Walloon, and the Brussels-Capital Regions. These Regions are mainly and exclusively competent in certain economic matters in their respective areas: environment, agriculture, energy, employment policy, public transport, international cooperation related to these competences, area development and planning, etc. There is no German Region; it is the Walloon Region that has competence in the German-speaking part of Belgium.
Complementing this federalized structure, the federal government and bicameral federal parliament have decision-making powers extending across the whole country, especially concerning issues that are not expressly left to the Regions or Communities. So, although the Regions and Communities have significant competences, federal institutions retain a fair share of power. Key issues like social security, police, immigration, foreign policy, justice and fiscal policy are still, to a large extent, in the hands of the federal government and parliament.
Graphically, the situation is presented in Figure 1.
The federal government depends on a majority in parliament; Belgium is thus a parliamentary democracy (as well as a—mainly ceremonial—monarchy). In federal parliament, which has competence with respect of the federal constitution (the term is a tad deceptive, as it is in fact the only one in the federation17), there is a clear Dutch-speaking majority: 62 of the members of the Chamber of Representatives belong to the French-speaking language group, while 88 belong to the Dutch-speaking language group; in the Senate the numbers are 29 and 44 respectively. From a quantitative point of view, Dutch-speaking members of parliament could impose their will on the minority of francophone parliamentarians. In reality, of course, things are not quite that simple in a federation whose main organizational principle is language. Not only has in the Federal Council of Ministers since 1970 linguistic parity been mandatory (Article 99 Constitution), but also decisions that concern the organization of the state have to be made by a two-thirds majority in federal parliament, as well as by a majority in each of the two language groups in this parliament (the requirement is described in art. 4 of the Constitution and referred to throughout the Constitution). Interestingly enough, this can be done without constitutional amendment, but by so-called “special legislation”; it is a technique that is prescribed for most Belgian institutional reforms. So, while a two-thirds majority can change the Constitution after a complicated and lengthy procedure, institutional reform must meet even stricter demands, even if such reform follows a “regular” legislative procedure, i.e., a procedure that doesn’t require change of the Constitution as such. Consequently, institutional reforms are impossible without a special majority in federal parliament, and 31 members of the federal Chamber of Representatives or 15 members of the federal Senate can block them. In other words, an institutional reform can never be achieved against the will of a majority in one of the language groups, even if this majority is representative of a minority in Belgium (as is the case of the French-speaking language group in federal parliament).
Moreover, whenever a 75 percent majority of one of these language groups in one of the chambers of federal parliament declares that the relations between the different communities in Belgium risk deteriorating as a result of a federal legislative proposal (as long as they are not dealing with budgets and do not concern laws requiring a special majority), the legislative procedure will be postponed (art. 54 Constitution). This is known as the alarm bell procedure, and the idea behind it is to prevent one of the language groups in federal parliament from pushing through a legislative proposal against the will of the other language group in parliament. When the alarm bell procedure is raised, the Council of Federal Ministers, which is composed on the basis of linguistic parity, must then, within 30 days, advise parliament on the issue concerned. As can be readily imagined, such a procedure puts enormous pressure on the federal government; if it cannot come to an agreement, it will almost certainly fall, with representatives of the two language groups subsequently having to negotiate a new government on a subject about which there is fundamental disagreement. Thus, the alarm bell procedure has a preventive or dissuasive quality and the political factions in parliament will not easily raise the alarm.18
There are, moreover, three highest courts: the Cour de cassation (in civil and criminal matters), the Conseil d’État (in administrative matters), and the Cour constitutionnelle (in constitutional matters). The Cour constitutionnelle deserves special mention here because its establishment was the result of Belgium’s transition, in the 1970s and 1980s, from a unitary to a federal state. This ultimately resulted in the introduction of judicial constitutional review, initially only to resolve conflicts of competence between the different levels of government. As indicated earlier, there is no hierarchy of legislative instruments enacted by the various entities of the Belgian state, and this may cause conflicts of competence between the Communities/Regions and the federal state. To resolve such conflicts, it was felt that a constitutional court was needed, since leaving this task to the parliaments would make Belgium ungovernable. Instead of assigning this role to the existing courts, the initiators of a constitutional amendment of 1980 chose to set up a special court, at the time called the Cour d’Arbitrage (Court of Arbitration). This court would adjudicate disputes between the various legislatures of the federal Belgian state.19 The then Cour d’Arbitrage (which was, and still is, composed of twelve members, on the basis of linguistic parity) could thus check whether legislation introduced by the federal state, the Communities or the Regions was in line with the competences assigned to them.20
It might be worth mentioning here that the demise of the federal government in 2007 was the direct result of a decision by the Cour constitiuonnelle on a highly sensitive issue: the constituency of Brussel-Halle-Vilvoorde (BHV), which for the federal elections comprised the central bilingual district of Brussels and some neighboring parts of Dutch-speaking Flanders (which, in fact, is home to many French-speaking Belgians), together containing some 1.6 million inhabitants. It has been an exceptionally—and politically—difficult topic for decades. In 2003, the Cour constitutionnelle ruled that, following a 2002 electoral reform, which was the result of a hard-fought political compromise, this constituency contravened the equality and non-discrimination clauses of the Constitution. It was the only constituency in the country that did not coincide with a single monolingual province. As a result of this 2002 reform, voters in parts of Flanders could vote for French-speaking politicians. The Cour constitutionnelle declared unconstitutional parts of the law concerning this compromise,21 also stating that a solution for this situation had to be reached before the 2007 federal elections. A solution, however, was not reached, and the contention made the post-election coalition negotiations very difficult.
In November 2007, the Dutch-speaking members of parliament almost unanimously voted to split up the constituency. This was the first time ever that a single language group did so in view of such a sensitive issue, which put a lot of pressure on the relations between the different language groups. The dire prospect of the conflicts of interest22 and alarm bell23 procedures loomed largely, and the “conflict of interest” procedure was actually resorted to. This caused considerable delay in resolving the issue. In 2010, the government fell again, before agreement on this topic between the different political factions was reached.24
What may not be immediately obvious to outsiders is to what extent difficulties surrounding this electoral district have to do with territorial-linguistic claims. Due to the constituency’s hybrid nature, the Dutch-speaking Flemish political parties wanted it to be split, which would create two separate electoral districts: one for the Brussels Region and one district for the Province of Flemish Brabant. This would prevent the francophone political parties from attracting French-speaking voters outside of bilingual Brussels and put a brake on the “Francization” of parts of Flanders (especially in the municipalities surrounding Brussels), something so greatly feared by the Flemish.
This is a highly sensitive issue in Belgian politics and society. Brussels has, over the centuries, developed from a Dutch-speaking city into a French-language enclave in Flanders. The legislative fixation in 1962–1963 of the language border (which was based on the language census of 1947), was meant to stop the Francization of Brussels and its surrounding area. The border was not to be based on language census anymore. The granting of linguistic facilities and rights to the French-speaking population in some municipalities was to compensate for this. The Francization, however, didn’t stop, resulting today in French-speaking majorities in officially Dutch-speaking municipalities. The debate, as it unfolds today, is partly put in terms of whether or not the language facilities were introduced as a temporary measure to give the French-speaking population the opportunity to learn to speak Dutch (a measure which, many Dutch speakers say, has been to no avail) or as some sort of fundamental right which almost by definition cannot be temporal (as many of the French speakers claim).
In sum, Belgium can, on the one hand, be characterized by what might be called “warehouse federalism”: a polity in which the different constituent parts of the state are competent in different matters, sometimes even within the same geographical area (i.e., Brussels). On the other hand, Belgium’s layout is also characterized by a blocked constitution, i.e., a constitution on the basis of which a minority can freeze any decision concerning institutional and constitutional reform. As far as state reforms and related matters are concerned, it is almost as if the majority is structurally in a minority position; it is a constitutional paradox.25 Today, all this makes for eight parliaments and eight governments,26 on top of ten provinces, a constitutional court, and two Kings (and three Queens actually).
3. How it has all come about: a very short socio-political history
As we have seen, even a fairly succinct description of the institutional organization of Belgium opens up a very complex multi-layered—“byzantine”—constitutional and institutional layout. How has it come about?27
After the Napoleonic wars, the United Kingdom of the Netherlands was established in the Low Countries in 1815. This Kingdom encompassed both the northern provinces that nowadays constitute the Netherlands, and the southern provinces that nowadays constitute Belgium. Support for the policies of the first monarch of the United Kingdom, King William I, was far less strong in the south than it was in the north. In the south, Roman Catholics and liberals pursued greater liberty in shaping their own identities. The Roman Catholics sought freedom of religion from the Protestant north, while the liberals wanted better representation: southern underrepresentation in the States General, i.e., the parliament based in the northern provinces, was indeed acutely felt by many in the region. In addition, the francophone population found it hard to stomach William’s Dutch-language politics.
Matters eventually came to a head when in 1830 the pursuit of liberty sparked a revolution, and the southern part of the country seceded from the northern part. A provisional government was formed, and shortly afterwards a National Congress was elected. The Congress then drafted and, in 1831, promulgated, a very liberal, for its time, Constitution.28 The result was what might be called a bourgeois state, which excluded a large number of its citizens from passive and active voting rights. Due to the previous Napoleonic occupation, these developments were largely, though not exclusively, French in influence. The resulting state was francophone in spirit, and as it happened, the poorer part of the fledgling country’s population was Flemish and Dutch-speaking. Dutch was considered to be the language of “common” people, while French was the language of high culture and refinement; French language dominated cultural, administrative, judicial, and economic life, as well as higher education.
A Flemish nationalist movement took shape in the nineteenth century as a result of at least three factors: (i) the influence of the Romantic movement, which greatly valued language, identity, tradition, and popular culture; (ii) the support of the Roman Catholic Church to the preservation of the Dutch language (Flanders was almost 100 percent Catholic, while in Wallonia, with its tangible French revolutionary influence, there was a strong anticlerical movement); and (iii) the threat of French annexation of Belgium. This last factor strengthened the desire to resist francophone influence, and for this reason, even the French-speaking King Leopold I supported the development of the Dutch language.29 As a result, in the nineteenth century, Flemish sentiment transformed into Flemish activism, and legislation was enacted that recognized the use of Dutch as one of the official languages of communication in local administration as well as the judiciary. Such a measure presupposed bilingualism in civil servants, which effectively put the Walloon population at a disadvantage, since Walloons were (and are) largely monolingual.
Belgium, like its northern neighbor the Netherlands, has only ever had a single constitution since its founding. The amendments this constitution has undergone, through the course of the country’s history, have, however, been substantial. These changes chiefly reflect the kingdom’s social segmentation and linguistic diversity. What is important to note here is that the Belgian state could only exist as an unqualified unitary nation state within the framework created by the tax-based electoral system, introduced by the 1831 Constitution:30 as voting power was in the hands of a wealthy French-speaking elite, electoral clout and “capital” for the most part coincided. The “language question,” as it was called, however, soon cast its shadow on all manner of struggles, such as those involving identity, class, economic development, and religion. When in 1893 the masses of Dutch-only Flemings were democratically and politically empowered for the first time—they were given the vote (although not yet universally, and with qualifications)—they began to realize more than ever before that with language division came socio-economic inequality.31 And they also realized that it was language discrimination that perpetuated the social divide. In the years to follow, more language rights were granted to the Flemish people. As a result of the gradual introduction of universal suffrage and of other electoral reforms (in particular, proportional representation, which allowed the Catholic and Socialist parties to dominate the political scene in the Walloon and Flemish parts of the country, respect- ively), regional monolingualism was legislatively recognized. For example, Dutch became the language of academic education at Ghent University in 1930, as well as of primary and secondary education (from 1932); also linguistic areas were defined, albeit initially in a flexible manner and in the context of administrative affairs (1921 and 1932). The development of universal suffrage acted as a catalyst for constitutional change in this respect, and for undermining the unitary Belgian state.
In this way, the language question gradually also gained an institutionalized identity. And it also developed into an emotional battlefield, building on the inherent social tensions within the Belgian state. Its scope was broadened by the accompanying socio-economic cleavage between the different communities. The conflict over language was as much a socio-economic struggle as it was a struggle for the institutional recognition of territorially defined identity.32 After the Second World War, Flanders developed quite rapidly economically, while, at the same time, there was a steady decline of Walloon industry, and the locus of economic power was transferred to the northern Dutch-speaking part of the country—a development that has stimulated Flemish consciousness and political power, but that has simultaneously invigorated the Walloon movement (by now the underdog!) and has fueled discontent between the two regions.
4. Political theory and culture
Although written law, especially the Constitution and special legislation, describes the official structures and procedures within which the main Belgian political institutions and its actors are supposed to function, it does not of course fully cover the actual decision-making processes.33 It is therefore important to deal with Belgian political culture, and clarify its relationship to the constitutional design (see Section 2), that is representative of it.
The political scientist Arend Lijphart has famously stated that what is remarkable about Belgium is “not that it is a culturally divided society . . . but that its cultural communities coexist peacefully and democratically. What is more, Belgium can legitimately claim to be the most thorough example of a consociational democracy, the type of democracy that is most suitable for deeply divided societies.”34 A consociational democracy can usually be found in socially, and possibly ethnically, divided or segmented societies. Whereas it was previously thought that a stable democracy could not exist in such societies, Lijphart has tried to show that this is not necessarily true. Consociational theory teaches us that segmentation raises the stakes of politics higher than they would be in a homogenous society, since segmentation is a potentially destabilizing force. Conducting politics as a zero-sum game is risky business, and the representatives of the different segments in society therefore actively strive for cooperation, consensus, and stability: they seek to find each other and to cherish common ground as much as possible. Political differences between the ruling groups are, as a result, not politicized or exaggerated and a substantial number of the political leaders cooperate in governing the country. A pragmatic political elite, organized into political parties, seeks to solve societal and political problems in such a way that all factions concerned can more or less accept the outcome of the political game, thus neutralizing destabilizing tendencies. This also prevents major political groups from becoming estranged from the political system. As a result, although political decision-making in consociational democracies is strongly affected by the interplay of past and present political and other tensions, in practice, so the theory goes, it operates in a way that defuses these tensions and encourages compromise.35
To generate widespread acceptance, a consociational democracy should ideally have four characteristics: a grand governing coalition (or power-sharing executive structure among relevant political groups); proportionality (not just in elections, but also in cabinets, parliament, the civil service, the media, advisory organs, etc.); mutual vetoes (decisions can only be taken with mutual consent); and segmental autonomy (each social segment has its own sphere of authority, either territorial and/or functional, which allows sensitive issues to be taken off the national agenda).
In Belgium, even before 1970, the year in which the nation’s linguistic division was constitutionally entrenched and the process of federalization started, there were clear elements of consociationalism, predominantly on matters of what might be called confessional/ideological and socio-economic concern, but not so much on matters of language divide.36 Consociationalism mainly manifested itself as “pillarization.”37 A pillarized society is vertically divided into several segments or “pillars” that represent different allegiances. In Belgium, this manifests as Catholics against socialists and liberals in the case of confessional issues (and all that comes with it) and socialists against liberals in the case of socio-economic issues (the Catholics being split on this).38 The three main political segments (liberals, Christian democrats, and socialists), which date back to the mid-nineteenth century, were reflected not only in political parties, but also in social organization. These segments enjoyed a certain amount of self-rule, especially concerning the just mentioned ideological fault line, which, as a result, gave more salience to the linguistic tensions in the country.39 In any case, ideological and socio-economic segmentation has strongly permeated virtually all socio-economic and religious institutions in the country. Thus, outside the strictly political framework, but mostly in close contact with it, there is a whole range of social org- anizations that reflect the various ideological and socio-economic divisions in Belgium and that possess a measure of autonomy in organizing their sphere of activity: there were and still are Catholic, socialist, and liberal trade unions, and Catholic and (neutral) state schools, hospitals, universities, etc. And although the societal differences on which these organizations were built, especially the confessional/ideological one, have eroded, the grip they have on the government, parliament, and public administration is still considerable. Socio-economic policies and the organization of the welfare state continue, to a large extent, to devolve into the deeply institutionalized pillar organizations.40
Nevertheless, it was only after the “pacification” of the regional-linguistic divide in 1970—the date, as we saw, of the institutionalization of the language divide in the Constitution—that Belgium could develop into a near-split country,41 with the linguistic and accompanying socio-economic divisions as its main driving force. At first, the Regions and Communities were only granted cultural autonomy, as it was believed that such “light” competences could not endanger national unity. Over the years, however, in the wake of subsequent state reforms, regional autonomy has grown. The Belgian state has further federalized, and the Regions and Communities have gained competences in many matters; consociationalism in Belgium today is therefore qualified in that it has coincided with federalization. As a result there is, it is fair to say, a double protection of minorities: by way of institutionalizing these protections in the Constitution at the federal level, and by way of granting autonomy to linguistic groups (including minority groups) through federalization.42 For Flemish nationalists, this double protection of the French-speaking minority is a major source of frustration, and this again has boosted Flemish identity and calls for more Flemish autonomy.
The contemporary constitutional and institutional structures of Belgium, as set out in Section 2 of this article, can indeed be understood from the point of view of consociationalism: federal government coalitions usually have a fairly broad support base with a minimum of three coalition partners (and often more), and it is clear that changes in the country’s institutional set-up cannot possibly be decided by a simple majority in parliament; there are many institutionalized mechanisms in place to reach (or force!) political consensus or pacification on these issues. For this reason, Pinder has defined consociational arrangements as “institutions and procedures that encourage consensus rather than allowing the will of those who represent a simple majority of the population to prevail.”43 Majority-type democracy, from this point of view, is “the antithesis of consociational democracy.”44
One specific aspect of Belgian politics should finally be stressed here, namely the important role political parties and elites play in the political system and in society. Compromise-mindedness is a necessity, but not, as we will see in the next section, an easy requisite for political decision-making in a divided society like Belgium. Due to the complicated nature of reaching political compromises, the political elite mostly makes choices in secretive negotiations. Political parties also tend to control the allocations of positions and policy choices in government, as well as the appointment of many officials. They have, moreover, been able to maintain this prominent position by allocating public resources to themselves by legislative means (party financing mechanisms).45
5. Trouble in consociational paradise: two fault lines
Does Belgium’s constitutional layout promote societal consensus and stability? The short answer to this question is that its track record is not optimal: constitutional design in Belgium, as much as it fosters cooperation, structurally hinders accommodation techniques. As it develops, consociationalism can harden or rigidify pre-existing group differences.46 At least two problems stand out.
The first can be found on the level of political parties and elite estrangement. In a consociational democracy, Stephen Holmes has aptly observed, the political elites must both represent and not represent their constituencies: “They must hold their followers’ loyalty, but not reproduce their uncompromising attitudes in national negotiations. Such cross-sectarian cooperation among elites requires ‘a strengthening of the political inertness of the non-elite public and their deferential attitudes to the segmental leaders.’”47 The trouble is that in Belgium this seems hardly possible since national political parties no longer exist. Since the 1960s, Belgium has had regional political parties, and today, quite a few of these parties have no ideological counterpart on the other side of the language border. And even when they do (as is the case with the Christian-democratic, liberal, and social-democratic political parties) they are split on many important issues. Today, the discursive focus of politics and of public opinion is almost exclusively geared towards the regional level and its sentiments. Federal politicians are first and foremost representatives of their respective regions in which they are elected. This makes for a not necessarily Belgian minded political elite, with hardly any Belgian identity.48 All this has removed incentives for moderate consensual and electoral politics, and has resulted in a big gap between regional electoral pledges and the reality of national politics. This gulf has been widened by a split media landscape,49 and the fact that since the 1990s, it has been possible to form different or asymmetrical regional and federal government coalitions.50 Consociationalism asks that the language communities be willing and have the ability to compromise and reach consensus, but before that is possible, in Belgium, “elected politicians wanting to govern at the federal level must first solve the problems [among the language communities] that they have created themselves.”51
The second consociational fault line, which lines up to the first one in that both stimulate calls for more, in this case, Flemish autonomy (and in both cases the solutions desired become part of the problem for consociationalist theory and politics), has to do with the specificity of Belgian federalism. Federalism is mostly intended to bring the separate units of a state together.52 The Belgian variety of federalism, however, has the exact opposite aim of separating its two main components;53 Belgian federalism is thoroughly bipolar. While federalism tends to work well when there are more than two basically equal partners, “[Belgian federalism] is the juxtaposition of two peoples moving in different directions.”54 As a result, social tensions among the groups can easily worsen, since “the problem” can always be traced to “the same other”:55 political parties belong to one language group, as do members of parliament (including the accompanying veto mechanisms), federal government is based on linguistic parity, etc. This is even more so since the rather far-reaching allocation of competences to the Regions and Communities highlight the differences rather than the commonalities between the two main regions and language groups. To put it in financial terms, the fact that there are significant financial transfers from one economically solid region—Flanders—to the economically less well off region—Wallonia—has a profound impact on Flemish public opinion. Many would reason that the financing region should also have a more decisive say in the other’s state of affairs and in their institutional and constitutional relations (i.e., how they relate to each other).56 However, as we have seen, decision-making on the future of the Belgian state, including its financial allocation mechanisms,57 always involves the other veto player who is perceived to be radically different. And since calls for more autonomy tend to come from the Dutch-language part of the country, this “other” is also a linguistic minority.
In Belgium, the dynamics between institutional change and regional identity building seems to have a self-enforcing effect. Van Goethem points out that the growth of regional nationalism and awareness in Flanders and in Wallonia is constantly subjected “to the renewing influence of its own internal dynamic, so that the emotional element becomes stronger and stronger, almost in a kind of permanent escalation.”58 He suggests that the electoral reforms of 189359 sparked a development, in Flanders at least, from romantic nationalism to a hard political struggle for a monolingual and more autonomous Flanders. At the same time, this development hardened Walloon resistance against Flemish nationalism. What is important to note here is that an incremental development towards more demands for regional autonomy is embedded in this dynamic.60 The same is true of the process of federalization, which started off institutionally in 1970 and was introduced precisely to curb demands for more autonomy (or even separatism), but which seems to possess a virtually unstoppable devolutional or centrifugal dynamic exactly towards (demands for) more autonomy. Today, the Regions and Communities have indeed developed into highly autonomous parts of the country. The autonomy of the Regions and Communities also explains why Belgium can go without a federal government for long periods of time: governmental policy is to a considerable extent set and implemented by the sub-state governments and, not to forget, the European Union (with its headquarters in the Belgian capital, Brussels). Although this is certainly not an ideal situation, it does allow Belgium to survive federally with a so-called caretaker government.61 Thus understood, there is in Belgium less than optimal pressure to rapidly form a federal government—a state of affairs that is moreover not improved by the fact that the consociational elites have delegated many functions of government to semi-private segmental organizations (pillarization).62 In more general terms, institutional and constitutional patterns have reinforced more structural patterns; patterns that are not necessarily conducive to a consociationalist political culture.
The fault lines described and analyzed in this section provoke two questions about the status of consociational theory. What is its value, and is Belgium (still?) a consociational democracy? To begin with the second question, it is clear that the rules of the political game in Belgium are consociational: unilateral action and majoritarian decision-making between the language groups on issues of state organization are in any case not legally possible. The ability to compromise is still the conditio sine qua non of Belgian politics and possibly even of the continuation of the Belgian state: the default option is that there will be a profound crisis, and thus possibly no decision-making in any policy domain.63 The rules of the political game are constitutionally defined and figure prominently in political discourse, especially, of course, when the disagreement is on very fundamental issues of state organization. In Belgium, the constitutional and institutional context thus seems crucial in determining how the country develops; the veto player has very powerful means at his disposal, and to date, state failure has not seemed to be an option for those politically in charge. The 1970 reforms—resulting in the constitutional fixation of the language border, splitting up parliament in language groups and granting these groups veto rights, and other mechanisms by which a parliamentary majority can be blocked64—have been crucial in this respect.65 But there is also a creeping, built-in dynamic towards more autonomy that threatens the Belgian state from within. It seems, as I have tried to show, that Belgian consociationalism and federalism function as a self-fulfilling prophecy towards splitting up Belgium. Yet, it is also fair to say that Belgian consociational democracy continues to function. True, there have been a number of deep crises as well as highly imaginative compromises (which explain the rather accidental and hybrid nature of Belgium’s federalism and its Constitution), but it is precisely the fact that these compromises are reached that can be understood as proof of the political will to keep the system going.66
But how long will this situation last? “Prediction is difficult, especially concerning the future,” Niels Bohr reportedly once said. The complexity of the Belgian situation makes any prediction an even more hazardous challenge than it normally is. From the perspective of this article, the one million dollar question is whether Belgium’s constitutional and institutional set-up is conducive to societal stability and peace (and in line with this: whether Belgium will survive its seemingly permanent political crisis). So far it seems as if it has indeed been conducive to societal stability: Belgium is not marked by riots and violent uprisings between the different communities. And even the relative instability of Belgian governments can be seen as systemically functional because governmental break-up might also be understood as one of the means of reducing the possibilities for political deadlock and blockage.67 Even so, there are indications that in Belgium consociational strategies can no longer be fully relied upon. For example, the cries for more autonomy have in recent years led to the rise of a strong democratic nationalist political party in Flanders, the NV-A (Nieuw Vlaamse Alliantie, the New Flemish Alliance), which is currently part of the Flemish government coalition. The NV-A would appear to have replaced the more extremist and separatist right-wing Vlaams Belang (Flemish Interest) as the voters’ favorite nationalist political party; until recently, the latter did rather well in elections, but politically it has remained a marginal force because of its xenophobic views and the resulting exclusion of any governing coalition (cordon sanitaire).68 It is also at heart an opposition party. The current success of the NV-A might be a sign of a changing political constellation—separatism is becoming a real issue in political discourse; maintenance of the status quo does not appear to be the only option any longer, at least not for the Flemish part of the country, nor is it realistic, given the institutional and historical weaknesses of the Belgian state.69 And although demands for more autonomy seem to come mainly from the Dutch-speaking side of Belgium, these demands are highly relevant to the French-speaking part. Previously, when tensions ran high and one of the actors insisted on institutional reform, the other side perceived that as a quasi-obligation to negotiate a solution for the problem concerned.70 Today, however, the positions are hardening and some development towards an even more federalized system seems to be the unavoidable next step.
As far as the first question about the value of consociational theory is concerned, it seems to me that it is at its strongest when it is understood as a descriptive, rather than explanatory, theory, i.e., if we understand it as being descriptive of a “fuzzy”71 state of affairs. As an explanatory theory, consociationalism is unclear as to why some elites opt for consociational techniques, and other elites for other techniques, to create societal stability. Consociational theory seems, moreover, to be tautological:
Consociationalism is defined by a deeply divided society and by elite cooperation; in other words, both the problem and the solution are part of the definition. . . . To say that elite cooperation leads to political stability is almost the same as saying that there will be no fights in the playground when children stop quarreling (almost, because there still remains the possibility that the masses will revolt spontaneously against their cooperating leaders). In that sense, consociational democracy is not a theory of cause and effect but rather a descriptive category defined by a problem, the reaction to the problem by political elites, and the consequences of that reaction.72
Logically, a desire to coalesce implies a need for it.73 So in light of what has been said so far, we might well wonder why Belgium does not split up. The most convincing explanation for the current situation is Brussels. There are at least two elements that are relevant here. The first is the economically vital position of Brussels: it is a significant place of employment for many people from outside of Brussels. Breaking up comes at considerable cost, possibly too high to even seriously consider. The second element is the emotional element: although nowadays Brussels is essentially a French-speaking city, geographically surrounded by Flanders, historically it is of Dutch-speaking origin. The Francization of Brussels, which started in the mid-nineteenth century, is a “romantic” source of frustration for many.74 “If Flanders were not to claim or need Brussels, the structure of Belgium could be much less complicated. Both communities see Brussels as their (capital) city and that is what has created the double federation of Regions and Communities.”75 In the meantime, tensions continue to build, Flemish calls for more autonomy are growing stronger, and the willingness to federally cooperate in a consensual manner seems to be losing ground.
6. Back to the Constitution
In Passions and Constraint, Stephen Holmes writes that a constitution disables political decision-making in that it sets up procedural roadblocks by, for example, introducing supermajority decision-making rules or by enacting bills of rights; the idea is thereby to prevent tyranny and other abuses of power.76 Holmes terms this “negative constitutionalism.” In a divided society, as in any liberal society, constitutions must fulfill this negative or disabling role.77 A constitution at the same time can, according to Holmes, also assume an enabling role, which he calls “positive constitutionalism”: it can help create the very demos which governs itself through the constitutional regime. In this sense, a state can also use its powers to achieve cooperation and support, rather than just compliance. Even stronger, in a divided democracy, negative and positive constitutionalism should be connected to each other. Choudhry puts it like this:
[B]ecause of a history of conflict or a lack of shared existence, the constitution is often the principal vehicle for the forging of a common political identity, which is, in turn, necessary to make that constitutional regime work. To some extent, the constitution can foster the development of a common political identity by creating institutional spaces for shared decision making among members of different ethnocultural groups. Concrete experiences of shared decision-making within the framework of the rule of law, and without resource to force or fraud, can serve as the germ of a nascent sense of political community.78
I submit that, in Belgium, negative and positive constitutionalism are not connected and thus cannot be mutually reinforcing. Belgian constitutionalism does not succeed in taking up an enabling role. The Constitution itself is at best a rational construction, possessing an accidental nature and a complex technicality; there is no clear Belgian project to be discovered in it, and it represents a form of federalism that does not create any Belgian public space. More than anything else, the Belgian Constitution has institutionalized political distrust and instability. We might as well talk about disintegrating federalism. As a result, it has not been capable of instilling a sense of nationhood—with mutual respect and common interests between the (populations of the) different communities—or Verfassungspariotismus in its nationals.79 The Belgian Constitution is an encumbering one, and the type of decision-making it encourages can hardly serve as the “germ of a nascent sense of political community,” or promote the general interest beyond the regional interests.
All this raises more general questions of a constitution’s endurance and normative value:80 under what conditions are political actors usually prepared to comply with constitutional requirements? Under what conditions can negative and positive constitutionalism be connected? The answer to these questions will depend on the socio-political environment a particular constitution is meant to serve, and on the integration of that constitution into the society of which it intends to be a part.81 In this vein, Richard Kay has recently written that the force of any constitution ultimately depends on it continuing to be acceptable over time to the real human beings whose lives it affects. That acceptability, Kay adds, will have two aspects: “First, some minimum part of the relevant population must find the constitution’s substantive rules satisfactory, or at least tolerable. Second, that population must regard the constitutional rules as having issued from a legitimate source.”82 In other words, there should be a general public willingness to accept the authority of the constitution, regardless of whether or not each of its elements corresponds more or less with every person’s individual interest or point of view of how the state should be organized and work.
Kay’s anti-Weberian observation83 leaves open many pressing questions. How many people, for example, does it take to constitute the necessary minimum part of the relevant population in a liberal democracy? Whatever the concrete answer to these questions may be, I understand his observation as a mainly sociological one; claiming that the tension between the aspirations of constitutional rules, on the one hand, and their capacity to affect political reality, on the other, must be constrained if the constitution is to play a normative or real role. And then, it is important to note that the two conditioning features Kay identifies for a constitution to possess “constitutional force” are, at least in Dutch-speaking Belgium, in steady decline. As I have tried to show in this article, the Belgian Constitution does play a significant role in channeling and/or constraining the local politico-institutional situation. But this comes at great emotional cost, and at least in Flanders there appears to be limited support for the Belgian construct as it is established by its Constitution.84 The Belgian Constitution today reflects a status quo, has hardly any expressive or aspirational role85 to play, and is deficient in its positive or enabling function and in generating general public support.
This is a problematic state of affairs, since partition and secession, which might be the result of all this, usually comes with an array of negative consequences, the possibility of violence being one of them.86 So, although there is no right or objective answer to the question why Belgium should continue to exist, there are nevertheless reasons to try to foster moderation and cooperation between the communities and its politicians. One way of doing this is by means of electoral systems. “If election depends, at the margin, on the ability to gain some votes from members of groups other than one’s own, then political leaders will behave in an ethnically conciliatory fashion for that purpose. One thing we know is that politicians like being elected and reelected. If consociational theory provides no motive for compromise behavior, incentive theory, by definition, does not share that defect,” Horowitz writes.87 One could, for example, (re)introduce a federal electoral district for the federal parliament, which should create incentives for political parties to cross internal linguistic borders and be able to achieve a stronger kind of political integration within Belgium and a better developed federal public sphere.88
7. Final observations
I do not think that there exists an objective set of rules for matching a people and their situation with a set of institutions, nor is there an inherently stable or objectively superior constitutional system.89 The Belgian case can nevertheless be instructive in telling us a bit more about the constitutional conditions that are at least, if nothing else, not antagonistic to societal stability (or possibly even conducive to it). It seems as if the Belgian Constitution is developing into a nominal constitution: i.e., a constitution that lacks existential reality.90 Loewenstein compares this to a badly tailored suit, but in the Belgian case, it might be better to talk about a worn out suit. In any case, a constitution like the Belgian one is in danger of no longer being able to govern the dynamics of the power process and becomes, instead, governed by it:91 a warning of the demise of a state’s capacity to effectively act as the agent of its citizens’ well being.
In such conditions, the Constitution runs the risk of losing its legitimacy and falling prey to political manipulation. Consociationalism à la Belgique does not seem to have been able to prevent this development, and has possibly even stimulated it. At the end of the day, this is a problematic state of affairs because democratic politics without the normative leverage of a constitution is powerless. No democracy can survive such a state of affairs.92 It is therefore not enough to have an idea about what role the Belgian Constitution still plays in channeling and/or constraining the local politico-institutional situation (i.e., the research question this article started with). It is just as important to know under what conditions a constitution can function as a normative document. The thesis proposed in this article is that there should at least be a connection between what is being called negative and positive constitutionalism, and that Belgium might benefit from installing political incentives to encourage moderation between the communities and its politicians.
Quoted in newspaper De Standaard, Feb. 27 2012.
From Apr. 26, 2010. The new federal government was installed on Dec. 6, 2011. The government formation process of 2007 took 194 days.
I am hinting here at Arend Lijphart’s theory. See Section 4 of this article.
Sophia Powers, René Margritte, The Art Story, www.theartstory.org/artist-magritte-rene.htm.
Cf. Mark Tushnet, The Continuing Significance of ‘Country Studies’ in Comparative Constitutional Law, Paper presented at the World Congress of the International Association of Constitutional Law, Mexico City, Dec. 6–10, 2010, available atwww.juridicas.unam.mx/wccl/ponencias/17/313.pdf.
Cf. Kris Deschouwer, The Politics of Belgium. Governing a Divided Society 242 (2009); Richard Pildes, The Legal Structure of Democracy, inThe Oxford Companion to Law and Politics 332 (Keith Whittington et al. eds., 2008). The absence of Belgium as a case study in many a volume on constitutional design and change seems to me a missed opportunity from the perspective of theory building. Recently this was for example the case in Constitutional Design for Divided Societies. Integration or Accommodation? (Sujit Choudhry ed., 2008); and How Constitutions Change. A Comparative Study (Dawn Oliver & Carlo Fusaro eds., 2011). Belgium indeed is a Fundgrube for developing ideas about how to sustain a just political order and about regulating societal conflict.
As may be inferred from the above, this article will specifically address those constitutional norms that concern the institutions and organization of the Belgian state and the rules of the political game. Of course, a democratic constitution also significantly contains civil rights. That is why I want to make it clear at this early point that the topic of this article does not preclude that I will pay attention to this other dimension of a constitution, although mainly as far as language rights are concerned.
Cf. Andrew Harding & Peter Leyland, Comparative Law in Constitutional Contexts, in Comparative Law. A Handbook 313, 322–323 (Esin Örücü & David Nelken eds., 2006).
Samuel E. Finer, Vernon Bogdanor & Bernard Rudden, Comparing Constitutions 5 (1995).
An English version of the text of the Belgian Constitution can be found at http://legislationline.org/documents/section/constitutions/.
Since in the context of this article this is a telling constitutional provision, I quote the text in full:
Belgium comprises four linguistic regions: the Dutch-speaking region, the French-speaking region, the bilingual region of Brussels-Capital and the German-speaking region. Each municipality of the Kingdom forms part of one of these linguistic regions. The boundaries of the four linguistic regions can only be changed or corrected by a law passed by a majority of the votes cast in each linguistic group in each House, on condition that a majority of the members of each group is present and provided that the total number of votes in favor that are cast in the two linguistic groups is equal to at least two thirds of the votes cast.
Historically, Brussels was a Dutch-speaking city. Nowadays, during office hours, with many Flemish commuters coming to work there, it is a bilingual city. For more on Brussels, see Section 5.
Which in practice might well be the majority language (linguistic borders have been entrenched on the basis of a decades old language count!). Where, for example, French is not the official language, this does not rule out the possibility that sometimes the majority of the population in these municipalities are first and foremost French-speaking. This can especially be the case in the municipalities surrounding Brussels.
In the remainder of this article I use capitals whenever I refer to these officially recognized Communities and Regions and their competences. When these capitals are not used I refer to the non-official, i.e., colloquial, meaning of the words.
Five in all: 1970, 1980, 1988–1989, 1993, and 2001. At the time of this article’s composition there was agreement on a sixth state reform, but it had yet partly to be laid down in legislation and implemented. Suggestions for a seventh state reform pop up in the media already.
Although there still is political disagreement on the issue of territorial competence: some French-speaking politicians claim that the competences of the Communities are personally rather than territorially defined. This would imply that the French Community would be competent with respect to the French-speaking population of Flanders. There is, however, an authoritative line of case law by the Belgian Cour constitutionnelle as well as the Conseil d’État that confirms the principle of territoriality in this context, although there are exceptions to this.
But see infra note 84.
This alarm bell mechanism exists next to the so-called conflict of interests procedure, which is applicable between the different entities of the Belgian federal state. More specifically, it can be called upon when the profession of competences by one of the parliaments of the Communities, Regions or Federation is believed to be harmful to the interest of one or more of the other entities (thus possibly contravening federal loyalty, as laid down in art. 143 of the Constitution). When, after consultations between the different entities no solution for the conflict can be reached, the case is sent to a “consultation committee,” composed of 12 members of the federal, Regional and Community governments, which has to find an agreement. If it does not, the issue is tabled again in the assembly where it was first dealt with (after which another assembly might initiate the conflict of interests procedure again). This procedure can delay a legislative procedure for quite some time. A similar procedure exists for conflicts of interest between the different governments.
Of course, since this procedure concerns the different autonomous entities of the Belgian federal state, the political consequences are not necessarily as drastic when compared to the alarm bell procedure (which, as we saw, deals within one and the same level of the Belgian federal state): the continued existence of any of the governments is not dependent on a political majority in a parliament to which it does not have to pay responsibility to.
For more on this, see Patrick Peeters, Expanding Constitutional Review by the Belgian “Court of Arbitration”, 11 Eur. Pub. L. 475, 475–479 (2005).
Since 1989, the Court has also been competent to determine whether the right to equal treatment, the right not to be discriminated against, and the right to freedom of education have been violated (encapsulated in arts. 10, 11, and 24 of the Belgian Constitution, respectively). In 2003, the Belgian constitutional legislature decided to extend the review powers of the Court of Arbitration to include, especially, all the rights and freedoms listed in Title II of the Belgian Constitution (On Belgians and Their Rights) and to a number of other constitutional provisions. Finally, in 2007, the Court’s name was changed to Cour constitutionnelle (Grondwettelijk hof in Dutch, Verfassungsgerichtshof in German). On these reforms, see Maurice Adams & Gerhard van der Schyff, Political Theory Put to the Test: Comparative Law and the Origins of Judicial Constitutional Review, 10(2) Global Jurist 8 (2010).
Judgment 73/2003 of May 26, 2003, available in Dutch, French, and German atwww.const-court.be/public. See Dave Sinardet, From Consociational Consciousness to Majoritarian Myth: Consociational Democracy, Multi-level Politics and the Case of Brussel-Halle-Vilvoorde, 45 Acta Politica 346 (2010); and Servaas Lindemans, Het probleem Brussel-Halle-Vilvoorde. Analyse van een staatsrechtelijke doos van Pandora [The problem Brussel-Halle-Vilvoorde. Analysis of a constitutional Pandora’s box], Jura Falconis 473 (2005–2006).
See Peeters, supra note 19, and supra text accompanying note 18.
See supra text accompanying note 18.
All this raises the question whether the 2007 and 2010 elections in the aforementioned constituency were conducted in a legally correct manner. Opinions differ.
Jan Velaers, De crisis van de staat en de achillespees van het staatsrecht [The crisis of the state and the achilles heel of constitutional law], Rechtskundig Weekblad (2011–2012), at 24.
Each government comes with a parliament. The parliaments are: Federal Parliament (which is the only one with two chambers); Flemish Parliament (for both the Flemish Region and Community); Parliament of the Walloon Region; Parliament of the French Community; Parliament of the German-Speaking Community; Parliament of the Brussels-Capital Region; Combined Assembly of the Common Community Commission; Assembly of the French Community Commission; and Assembly of the Flemish Community Commission. The last four parliaments/assemblies are composed of the same people, each time exerting other competences. Depending on the definition of parliament, the count may differ from the one given here. Vuye, for example, refers to nine parliaments, counting Federal Parliament, because of its two Chambers, double. See Hendrik Vuye, België: een Staat in hervorming of in ontbinding? [Belgium: a State in reform or in corruption?], Ars Aequi 719 (2008). To avoid making matters even more complex, I will disregard the ten provinces (each of which also has an assembly) and the 589 Belgian municipalities.
It is impossible here to give anything more than a very rough sketch of Belgian socio-political history. For details see, e.g., Herman Van Goethem, Belgium—Challenging the Concept of a National Social Security. A Short History of National Partition, in Social Federalism: The Creation of a Layered Welfare State: The Belgian Case 21 (Bea Cantillon et al. eds., 2011); Herman Van Goethem, Belgium and the Monarchy. From National Independence to National Disintegration (2010) (in this paragraph I strongly build on these two publications); Deschouwer, supra note 6, at 16–40. See alsoEls Witte et al., Politieke geschiedenis van België. Van 1830 tot heden [Political History of Belgium. From 1830 until Today] (1997).
A careful analysis of the various influences on the Belgian Constitution can be found in Jean Gilissen, Die belgische Verfassung von 1831. Ihr Ursprung und Einfluss [The Belgian Constitutution of 1831. Its Origin and Influence], inBeiträge zur deutschen und belgischen Verfassungsgeschichte im 19. Jahrhundert 38 (Werner Conze ed., 1967).
Van Goethem, supra note 27, at 25.
Van Goethem, supra note 27, at 251 et seq.
Id.
Van Goethem, supra note 27, at 25. I am not going into the question of whether or not either Flemish, Walloon or Belgian identity is historically accurate or merely a construct. What is important is that it is real to many. Van Goethem in this context talks about an “imagined community”; a fiction which can develop into a reality, especially when it is given shape within a framework of concrete institutions, rituals and actions. SeeVan Goethem, supra note 27, at 17. This seems certainly true for the Flemish identity.
Xavier Mabille, Political Decision-making, inModern Belgium 201, 201 (René Bryssinck & Michel Boudart eds., 1990).
Arend Lijphart, The Belgian Example of Cultural Coexistence in Comparative Perspective, inConflict and Coexistence in Belgium. The Dynamics of a Culturally Divided Society 1, 1 (Arend Lijphart ed., 1981) (emphasis added). See alsoArend Lijphart, Democracy in Plural Societies. A Comparative Exploration 223–238 (1977); and Arend Lijphart, Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries 243–310 (1999).
Mabille, supra note 33, at 202–203 and 215.
Deschouwer, supra note 6, at 4–9.
The term, from the Dutch “verzuiling,” was as far as I can trace introduced in this context by Lucien Huyse, Pillarization Reconsidered, 19 Acta Politica 145 (1984).
See Wilfried Swenden & Maarten Theo Jans, “Will It Stay of Will It Go?” Federalism and the Sustainability of Belgium, 29 West Eur. Pol. 877, 878 (2006).
Cf. Lucien Huyse, Political Conflict in Bicultural Belgium, inConflict and Coexistence in Belgium,supra note 34, at 107–126.
Deschouwer, supra note 6, at 8.
Liesbet Hooghe, Belgium: Hollowing the Center, inFederalism, Unitarism and Territorial Cleavages 55, 71 (Ugo Amoretti & Nancy Bermeo eds., 2004).
This double protection is also strongly criticized in Flemish legal and political academic doctrine. See, e.g., recently Stefan Sottiaux, De Verenigde Staten van Belgie. Reflecties over de toekomst van het grondwettelijk recht in de gelaagde rechtsorde [The future of Belgium. Reflections on the future of constitutional law in a multilayered legal order] (2011) (pleading for either more federalism or a more firm democratization at the federal level, i.e., less extensive protection of the French-speaking minority). Compare the observation of Guy Peters, who, on the occasion of the publication of a special issue on Belgium of the journal West European Politics, said that it was remarkable to note a number of academics discuss “questions bordering on survival of the system and the maintenance of the system, rather than merely fine tuning and coping with more immediate policy problems.” See B. Guy Peters, Consociationalism, Corruption and Chocolate: Belgian Exceptionalism, 29W. Eur. Pol. 1079, 1080–1081 (2006).
John Pinder, Multinational Federations, inMultinational Federations 1, 9 (Michael Burges & John Pinder eds., 2007).
Lijphart, The Belgian Example, supra note 34, at 1.
Guy Peters, supra note 42, at 1081.
Pildes, supra note 6, at 333.
Stephen Holmes, Passions and Constraint. On the Theory of Liberal Democracy 212 (1995) (referring to Lijphart). Peters describes this as follows: “The federal nature of Belgian politics means that the party leaders must be thinking about their role in the context of a multi-level governance arrangement, and must conceptualise their role as the cement that binds the various elements of the system together.” Guy Peters, supra note 42, at 1081.
Deschouwer therefore talks about distrust at the top: the elite does not trust the system, because the political system is perceived to benefit the other more. See Deschouwer, supra note 6, at 233.
Which, if it even reports on this issue, gives a very one-sided and even partial account of events and political life on the other side of the language border. See David Sinardet & Martina Temmerman, Political Journalism Across the Language Border: Communicative Behaviour in Political Interviews by Dutch- and French-speaking Journalists with Dutch- and French-speaking Politicians in Federal Belgium, inMultilingualism and Applied Comparative Linguistics. Vol. 2: Cross-Cultural Communication, Translation Studies and Multilingual Terminology 110 (Jeroen Darquennes ed., 2008).
Deschouwer, supra note 6, at 233–236; Van Goethem, supra note 27, at 269–271. Guy Peters rather neutrally observes that “there is not necessarily a greater identification of the public with Belgium as an entity than with the constituent parts.” He adds that the regions have become the arena in which the conflict between both parts of Belgium are played out. See Guy Peters, supra note 42, at 1083–1084.
Deschouwer, supra note 6, at 237. I believe that this is the core of one of the main criticisms of consociationalism by Donald Horowitz: the granting of autonomy to different communities not necessarily promotes attitudes and behavior that promotes stability. Donald Horowitz, Constitutional Design: an Oxymoron?, inDesigning Democratic Institutions 253 (Ian Shapiro & Stephen Macedo eds., 2000); and Donald Horowitz, Constitutional Design: Proposals versus Processes, in The Architecture of Democracy. Constitutional Design, Conflict Management, and Democracy 15 (Andrew Reynolds ed., 2002). For an excellent overview of the debate between Lijphart and Horowitz, see Sujit Choudhry, Bridging Comparative Politics and Comparative Constitutional Law: Constitutional Design in Divided Societies, inConstitutional Design for Divided Societies,supra note 6, 15. See also Rudy Andeweg, Consociational Democracy, 3 Ann. Rev. Pol. Sci. 509 (2000).
Interestingly, Lijphart himself seems to have recognized this:
A multiple balance of power among the segments of a plural society is more conducive to consociational democracy than a dual balance of power or hegemony by one of the segments, because if one segment has a clear majority its leaders may attempt to dominate rather than cooperate with the rival minority.
SeeLijphart, Democracy in Plural Societies, supra note 34, 55.
Belgian federalism might have this in common with Canada or Spain, or possibly the United Kingdom (devolution).
As former Prime Minister Tindemans said in 1971: Parliamentary Proceedings, Senate (July 7, 1971) at 2368.
Deschouwer, supra note 6, at 235.
Van Goethem, supra note 27, at 271.
Which are not fully federalized: the division of competences between the two Regions and Communities does not always come with the full fiscal autonomy necessary for employing these competences; the financing mechanisms remain in federal hands. On this, see Swenden & Jans, supra note 38, at 885.
Van Goethem, supra note 27, at 255.
See supra Section 2.
Hooghe, supra note 41, at 36.
For example, during the caretaker administration of 2010–2011, an important agreement concerning a wage norm for employees was negotiated between the relevant actors (employers and employee organizations). On the caretaker government, see Carl Devos & Dave Sinardet, Governing without a Government: The Belgian Experiment, 25 Governance: Int’l J. Pol’y, Admin., & Institutions 167 (2012); and Geert Bouckaert & Marleen Brans, Governing without Government: Lessons from Belgium’s Caretaker Government, 25 Governance: Int’l J. Pol’y, Admin. & Institutions 173 (2012).
Hooghe, supra note 41, at 37. See also supra text accompanying note 37.
Velaers, supra note 25, at 25. Also Kris Deschouwer, And the Peace Goes On? Consociational Democracy and Belgian Politics in the Twenty-First Century?, 29 W. Eur. Pol. 895, 904 (2006); and Maarten Theo Jans, Leveled Domestic Politics. Comparing Institutional Reform and Ethnonational Conflicts in Canada and Belgium (1960–89) 4 Res Publica 37 (2001). Although the most recent caretaker government was capable of functioning for quite a while. See supra text accompanying 61. My suggestion is that this is partly due to the economic crisis, which created an extraordinary situation.
See supra Section 2.
Deschouwer, supra note 63, at 903.
Id. at 903.
Guy Peters, supra note 42, at 1088.
In 2004, some of its satellite organizations were convicted for racist crimes. See Maurice Adams, Geen woorden maar daden? De vrije meningsuiting van het Vlaams Blok [Not Words, But Actions. Freedom of Opinion of the Flemish Interest], Rechtsfilosofie en Rechtstheorie 189 (2004).
Van Goethem,supra note 27, 276.
Deschouwer, supra note 63, at 906.
Id. at 895–896 and 899 (Deschouwer also notes that the term consociationalism has been constantly taking on new meanings).
Andeweg, supra note 51, at 520 (referring to Barry).
As Lijphart himself admits in Arend Lijphart, The Wave of Power-Sharing Democracy, inThe Architecture of Democracysupra note 51, at 43.
See also Section 1.
Deschouwer, supra note 6, at 244.
Holmes, supra note 47, at xi–xiii, 5–7, and 161–164. Also Choudhry, supra note 51, at 5–6.
Choudhry, supra note 51, at 6.
Id. Choudhry adds that a constitution can “constitute a demos by encoding and projecting a certain vision of political community with the view of altering the very self-understanding of citizens” (id.).
Koen Lemmens, De Belgische Grondwet, een model voor Europa? Over natie en identiteit [The Belgian Constitution; A model for Europe? About nation and indentity], Rechtsfilosofie en Rechtstheorie 179, 185 (2005). Elazar puts it like this: “Federalism can exist only where there is a considerable tolerance of diversity and willingness to take political action through the political arts of negotiation even when the power to act unilaterally is available.” Daniel Elazar, Exploring Federalism 181 (1987).
On this normative role, seeKarl Loewenstein, Political Power and the Governmental Process 147–148 (1957): normative constitutionalism “focuses on the concordance of the reality of the power process with the norms of the constitution. It proceeds from the recognition that a written constitution does not operate automatically once it has been adopted by a nation.” See also Dieter Grimm, Types of Constitutions, in The Oxford Handbook of Comparative Constitutional Law 98, 107 (Michel Rosenfeld and András Sajό eds., 2012).
Loewenstein, supra note 80, 148.
Richard Kay, Constituent Authority, 59 Am. J. Comp. L. 715, 756 (2011). Cf. Choudhry, supra note 51, at 6–7.
In the sense that it runs counter to Max Weber’s view that the formal-rational character of law could be the dominant source of political legitimacy.
Hence also the many pleas and proposals for a Flemish “Constitution,” and this for many years already. As can be imagined, these proposals are fiercely debated, even among those sympathetic to such a project. Most recently, in May 2012, a resolution for a Charter for Flanders was proposed in Flemish Parliament by the three political parties that also participate in the Flemish government coalition. See Handvest Voor Vlaanderen, http://docs.vlaamsparlement.be/docs/biblio/opendigibib/monografie/2012/280_handvest_voor_vlaanderen_20120524.pdf (only in Dutch). This Charter is a collection of existing provisions, stemming from, e.g., the Belgian Constitution and the European Charter, and has no legal value. It has yet to be put to the vote, but the mere fact of it being proposed (by the parties forming Flemish government moreover) of course has political significance.
Cass Sunstein, On the Expressive Dimension of Law, 144 U. Pa. L. Rev. 2012 (1995–1996) and The Importance of Ideals: Debating Their Relevance in Law, Morality, and Politics esp. 11–38 (Wibren van der Burg & Sanne Taekema eds., 2004).
See Donald Horowitz, The Cracked Foundations of the Right to Secede, 14 J. Democracy 5 (2003–2002).
Id. at 15.
Dave Sinardet, Direct Democracy as a Tool to Shape a United Public Opinion in a Multilingual Society? Some Reflections Based on the Belgian Case, inPublic Opinion in a Multilingual Society Institutional Design and Federal Loyalty 34, 39–41 (Dave Sinardet & Marc Hooghe eds., 2009). About this, see Donald Horowitz, A Federal Constituency for Belgium: Right Idea, Inadequate Method, in Electoral Engineering for a Stalled Federation 25–28 (Dave Sinardet & Marc Hooghe eds., 2009).
Donald Lutz, Principles of Constitutional Design, at ix (2006). Loewenstein also discards the idea of trying to devise a “perfect theoretical constitution,” instead claiming that “an ideal constitution has never existed, and will never exist.” Loewenstein, supra note 80. This brings to mind Aristotle’s well-known observation that “[t]he attainment of the best constitution is likely to be impossible for the general run of states; and the good law-giver and the true statesman must therefore have their eyes open not only to what is the absolute best, but also to what is the best in relation to actual conditions.” See Aristotle, The Politics 134 (Bk IV, 1288b21) (Ernest Barker & Richard Stalley trans., 1998).
Loewenstein, supra note 80, at 148–149. Nominal constitutions, so Loewenstein asserts, find their limits in the given power structures, political as well as economical.
Id. at 149. The Belgian political sociologist Huyse noted more or less the same in the wake of the significant institutional crisis in Belgium in the middle of the 1990s, which was sparked by the high-profile criminal case against Marc Dutroux (the perverse murderer of a number of Belgian children): Luc Huyse, De lange weg naar Neufchâteau [The Long Road to Neufchâteau] 7 and 192 (1996).
Cf. Herman van Gunsteren, Het staatsrecht in de politiek [Constitutional Law in Politics], Nederlands Juristenblad 1111–1114 (2010).
