Fédéralisme Régionalisme

1374-3864 2034-6298


since 05 February 2011 :
View(s): 7580 (5 ULiège)
Download(s): 0 (0 ULiège)
Kris Deschouwer

Introduction - The Belgian federalism and federation on the move

(Volume 20 : 2020 — Trente ans de dynamiques fédérales et régionales)
Open Access

1This Journal has a somewhat special origin. The first issue of the actual Fédéralisme Régionalisme appeared in 1999-2000. That was however not the real beginning. There is a history that goes further back. A journal aiming at discussing issues related to federalism in Belgium was launched already in 1988. It was then called Coudenberg - Journal for Federalism and Democracy. The publisher was the Coudenberg Group, a collection of 47 prominent members of different sectors of Belgian society and 28 ‘young members’ who came mainly but not exclusively from the academic world. The Group was led and financed by Jean-Pierre De Bandt, chair of the law firm De Bandt, van Hecke & Lagae. That Coudenberg Group wanted to analyze and discuss the Belgian political system and its ongoing reforms in the light of theories of federalism. The Rapport Coudenberg1 that was published in 1987 put forward ‘twelve commandments’ of a classic federalism and concluded that many of these were not met by the way in which the Belgian state was being reformed. This classic federalism was however for the Coudenberg Group the way to go.

2We will not discuss the details of this rather negative evaluation of the Belgian system of the late 1980s, but must note that it was being evaluated against an ideal type of what a federal state should be, an ideal type modelled mainly on cases like the US and Canada, and to a lesser extent on Australia, Germany or Switzerland. The leading scholars on federalism were at that time indeed either based in the US or Canada, or doing research on these countries. As an offspring of the Coudenberg Group (and again with the financial and logistic support of Jean-Pierre De Bandt’s law firm) a Belgian Interuniversity Centre for Federalism was created, aiming at bringing together the expertise on federalism and on further developing it. Its members were André Alen, Jean Beaufays, Gonzales d’Alcantara (chair), Frank Delmartino, Kris Deschouwer, Rusen Ergec and Henri Tulkens. The Centre commissioned short reports on several aspects of federal systems2, published a handbook of federalism3 and immediately took over the Coudenberg journal that was from 1990 called Federalism and Democracy. The Centre for Federalism did not live long though. After a few years and because no solid and recurrent financial basis could be found it slowly faded away. Its journal also failed to produce a continuous stream of articles, and disappeared for a couples of years, to resurface in 1999 – thanks to the support of the Université de Liège – under the current title of Fédéralisme Régionalisme.

3Two elements in this short (pre)history of the journal are quite relevant. The first is that the early attempts to analyze the changing Belgian political system through the lens of federalism studies had a strong normative flavour. The Coudenberg Group was a somewhat strange combination of believers in true federalism and of people who did not like the dismantling of the unitary state and who tried – by proving that a real federal Belgium would never be possible – to find arguments to defend the status quo4. The Centre of Federalism did not at all share that latter argument and mainly wanted to boost the study of federalism in general, also by building links with other centres for the study of federalism in other federal countries. And while the Centre for Federalism did not survive, the number of Belgian scholars working on federalism has rapidly increased. A question that does remain very central in the research and the many publications until today is the question whether the Belgian political system – now transformed into (some kind of) federation – can really work. There is an almost absolute consensus about the fact that this is not the case, and therefore also a constant drive for identifying the bits and pieces of the institutional setup that contribute to its failure and a constant search for ways to improve it. We will elaborate on that below.

4The second element of the early history that is quite relevant is the use of the classic models as frames of reference and the subsequent move away from them as soon as the academic discipline of comparative federalism in political science, law and economics started to grow. That did not happen in Belgian isolation, and the Belgian story actually mirrors very similar developments in the study of federalism in Europe. The transformation of the Belgian state was part of a more general search in different European countries for a state structure that could respond to increasing demands for decentralization of decision-making. The relevant cases to compare with ceased to be the US and Canada (except then for the politics of linguistic identity in the latter) and became the United Kingdom, Italy and Spain. And within Europe renewed attention was also paid to Switzerland, Germany and Austria, not as ideal models to be followed, but to increase the understanding of the wide variety of rules, habits and institutions that could be found in federal systems. At the same time the European Union moved in a direction that increasingly looked like something that could be labelled ‘federal-type’ or ‘multi-level’5. The insights on the functioning of the EU – some of which (like the joint-decision trap or the notion of second order elections) came straight from the analysis of national federations – became part of the body of literature on comparative federalism.

5Publications on Belgium, both by Belgian scholars and by scholars from abroad, have become an integral part of these new and very rich developments. The study of federalism has moved way beyond the analysis of the old ‘coming together’ federations and has added numerous insights on the dynamics of ‘holding together’ federations, on processes of decentralization and devolution. Today we know a lot more about many actors, events and policies than we did 25 years ago. We have analyzed the organization of political parties, their varying strategies across levels (electoral strategies, coalition formation, …) and their electoral programmes and manifestoes. We have analyzed electoral results at the different levels and their relations to each other. We have analyzed members of parliaments at the central and substate level and understand now the degree in which the level at which they play their representational role affects the way in they understand it. We have analyzed political careers across levels and identified the different patterns and models. We have mapped the similarities and differences between first and second chambers, the forms and roles of intergovernmental relations, of intergovernmental formal and informal meetings, at conflicts of interest and at intergovernmental agreements. We have analyzed fiscal politics and policies, varieties of Finanzausgleich and social security systems. We have looked at supreme courts and other judicial techniques that regulate the relations between actors and institutions. We have looked at identity politics, at varieties of nationalism and regionalism, and at referendums for autonomy or independence.

6Belgium is but one of the many cases that now appear in all these studies on federal-type or multi-level systems. Belgium has in that sense become a ‘typical’ case of federalism. That is quite a change compared to the way in which the Belgian case was looked at a few decades ago. What has not changed that much though is the normative concern about the way in which the Belgian federation functions. The criticism of the early days is still very much present.

7There are several and strongly interrelated reasons why Belgium is seen – mainly by the Belgian scholars of federalism – as a suboptimal construction. The major indicator for that is the weak capacity for decision-making. Overlapping competencies and / or the need for different actors at different levels to collaborate for policy-making results too often in gridlock and therefore in the absence of decisions and absence of policy choices. One can refer to mobility, climate plans, the national airport, labour market or health policy (with the Covid-19 crisis as prime example) but also the state structure and state reform itself. The theoretical lens through which this can best be understood is the notion of the ‘joint-decision trap’, identified by Fritz Scharpf6, and inspired by the German political system and by the European Union. The short version says that when the federal level requires the input and participation of the substate levels for making a decision, and when the rule for decision making is consensus rather than majority, the chances are high that the actors fall into the trap and that no decision is being made. For policy programmes that require change, the result of the joint-decision trap is the continuation of a suboptimal situation. And for new policies it means that they cannot be developed. As a consequence of that, the actors build up levels of frustration, lose their trust in the system and try out unilateral actions to produce their own policies.

8This sounds and looks familiar to scholars of the Belgian system. Several of its characteristics make it fit nicely into this. In the very first place the bipolar nature of the federation opens the trap as wide as can be. While there are in Belgium formally three language communities and three regions, the dynamics of the political debate (the media and the public opinion), of political representation (the party system) and of political decision-making often involve two crucial actors: the Dutch and the French language communities. The decision-making at the federal level requires (constitutionally) the presence of both language groups and when only two actors are involved in decision-making the only rule that can be followed is unanimity. Each of the two groups thus has a veto right. It means that the state has to be governed together, or not governed at all. And while the obligation to move together does build in strong incentives to find a solution and to be extremely creative if needed (a creativity that has actually very often been displayed), the obligation to take into account the demands of the other – always the same other – makes it difficult to organize a good democratic representation. The latter indeed requires that the decision-makers are able to maintain a more than decent level of responsiveness, are able to keep their promises, and are able to defend the policies on which they have agreed.

9There are ways to get out of the joint-decision trap. One of these is the making of large and complex package deals, combining decisions on different and not necessarily related policy domains and thus neutralizing several mutual vetoes by allowing each partner to have enough points on which it can declare that promised have been kept. The flipside of this however is long processes of negotiation, opaque wheeling and dealing and the writing down of detailed agreements that because of their subtle equilibrium cannot be changed. Again this looks and sounds familiar for scholars of the Belgian system (and of the European Union), and it makes is a very elite-centred polity with strong powers for party leaders and for governments, to the detriment of the members of parliament who’d better not disturb the subtle compromises.

10Another lens through which this can be analyzed is consociational democracy. The Belgian federation is strongly consociational, by combining a large degree of autonomy for the subgroups with power sharing at the centre7. Power sharing means that not majority rule but the guaranteed presence of the subgroups is the rule for decision-making. A consociational setup has been defended at length by Arend Lijphart as a good way for governing divided societies. Other institutional solutions – implying that some groups are permanently winning and other are permanently losing – would exacerbate the conflict. The jury is still out though on the question whether consociational devices are under all circumstances the best answer to societal tensions. The paradox of federalism approach8 states that a federal solution for ethnically divided societies might be a temporary arrangement, but not a stable one in the longer run. The federal arrangements themselves, and in particular the formal and institutional recognition of the different identities that acquire their own institutions, build in the continuation of the tension. Subnational identities will become even stronger, and demands for autonomy will further increase. An ethno-federation or a consociational federation is then only a temporary stop towards the final dissolution of the state or towards ongoing and (in some empirical cases) violent clashes. While the latter are not present in Belgium and are also quite unlikely, the mechanisms that build in further tensions as a consequence of the federal structure are very visible9.

11This is not the place to develop all these arguments in full detail. That has actually been done at length by several scholars. We only wanted to draw attention to the evolutions in the analysis of the Belgian federal system during the lifetime of this journal. There is on the one hand a striking continuity. The early (also pre-scientific) origins of the journal are rooted in doubts and concerns about the trajectory of the reforms of the Belgian state. With the classic federations as the point of reference the already very complex and hybrid construction of Belgium was being questioned. These same questions are still being asked today. They are often implicit and sometimes explicit in the scientific research on federalism in Belgium. They are also permanently present in the pubic debate, which explains to a certain extent the choice of the research questions formulated by the scholars. In the public debate, the critical analysis of the Belgian federation – which we have quickly summarized above – leads to very different conclusion. The same arguments about the suboptimal functioning of the Belgian state are the source of inspiration for those who want to turn back the reforms and return somewhat or completely to the simple and straightforward unitary state. For others they are the proof that all attempts to reform the Belgian state are useless and hopeless and that therefore full separation would be an easier, quicker and less painful solution. And the same basic analysis inspires a wide range of suggestions and proposals to improve the current system, to correct what is wrong and to improve what goes well.

12Against the background of this ongoing public debate and the ongoing scientific interest for the analysis of the functioning of the state there has been an important evolution in the scientific analysis of Belgian federalism. That evolution has taken place as part of a more general (and actually quite spectacular) transformation of the way in which the comparative analysis of state structures in general and federal-type structures in particular has changed. That analysis has moved away from the purely institutional and legal analysis to a broader analysis of political institutions and their interaction with societal actors. And it has moved away from simple classifications and normative yardsticks to a full recognition of the fact that modern states and modern unions like the EU are actually very hybrid constructions that are constantly on the move and constantly in search of ways to deal with both internal and external demands. That increased variety of state structures has been reflected in the literature in general and – with the Belgian scholars being fully part of it – in the literature on Belgian federalism.

13This anniversary issue of Fédéralisme Régionalisme is a perfect illustration of that. The article by Peter Bursens and Petra Meier dives deep into the discussions about the most appropriate state structures, with a special focus on their democratic quality. By looking at models that have inspired the debates on the nature and possible future of the European Union, they reflect on how Belgium might be reformed and improved. Patricia Popelier looks at the interaction between the public debate and the scientific analysis of federal states. She questions the classic classifications – especially the one that draws a very neat line between federations and confederations – and defends a more subtle approach that allows to capture more variation and therefore also allows to better understand the alternative models that are put on the table by political actors.

14Géraldine Rosoux discusses the role of judges in multi-layered political systems. The judiciary is always part of the political, but that is even more the case in complex and multi-layered systems. Their very nature requires the constant need for arbitrage, for solving conflicts, and for providing a homogenous interpretation of the rules that are valid across all the territories of federal-type polities.

15Sébastien Santander and Antonios Vlassis focus on the European Union and on the way in which it tries to develop its commercial policies in a context where major powers like the US and China are also defending their interests. In doing so they illustrate the need for concerted and coordinated action in federal-type union. Forming a union does add to the common strength, but also opens up the risks and pitfalls of non-decision and unilateral actions. The European Union is also prominently present in the contribution by Samuel Defacqz and Claire Dupuy. They look at the way in which pressure groups are adapting their organization and strategies in an ever more complex environment in which national states at the same time decentralize powers internally and pool them externally at the European level.

16Nadia Fadil and Marco Martiniello describe the developments in racism and anti-racism in Belgium. While they are dealing with more general topics like decolonization and islamophobia, their analysis of the Belgian case cannot avoid pointing at the differences in the public debates and political actors in the Dutch-speaking and French-speaking parts of the country. And they acknowledge the fact that the identity struggles and debates inside Belgium are not free from gross stereotypes and expressions that can be labelled as racism.

17Maxime Counet, Geoffroy Matagne and Pierre Verjans also look at the identity debates in Belgium. They compare two books by the chairs of the autonomist parties in Belgium – one from the chair of the Volksunie in 1992 and one from the chair of the N-VA in 2019, and are able to show that these parties and their leaders are constantly searching for a story that not only defends a larger autonomy for Flanders, but also searches for coherent plans to govern that political entity in a complex Belgium that also belongs to a complex European Union.


1 Rapport Coudenberg. Naar een nieuw België? Lannoo, Tielt, 1987 ; Quelle Belgique pour demain. Le rapport Coudenberg, Duculot, Gembloux, 1987.

2 All the reports can still be consulted on the website of Fédéralisme Régionalisme.

3 Interuniversitair Studiecentrum voor Federalisme: Federalisme. Staatkundig, politiek en economisch, Maklu, Antwerpen 1994 ; Centre d’Etude du Fédéralisme: Le fédéralisme. Approches politique, économique et juridique, De Boek, Bruxelles, 1994.

4 There was also a tendency to defend a ‘true’ federalism that was not based on the language identities but on the provinces, the latter believed to be more real and historical entities and identities than French-speakers and Dutch-speakers. This defense of a multi-polar federation was of course also a way to say that the ongoing reforms were not to be supported or continued.

5 We are using these different terms her interchangeably, only to say that many terms are now being used. A more elaborate discussion of the terminology is offered by Popelier in this issue.

6 Scharpf (F.), ‘The Joint-decision trap. Lessons from German federalism and European integration’, Public Administration, vol. 66, n°3, 1988, pp. 239-278.

7 Lijphart (A.), Conflict and coexistence in Belgium. The dynamics of a culturally divided society, University of California, Berkeley, 1981.

8 Erk (J.) and Anderson (K.), ‘The paradox of federalism: does self-rule accommodate or exacerbate ethnic divisions’, Regional and Federal Studies, vol. 19, n°2, 2009, pp. 191-202.

9 Caluwaerts (D.) and Reuchamps (M.), ‘Combining federalism with consociationalism: Is Belgian consociational federalism digging its own grave?’, Ethnopolitics, vol. 14, n°3, 2015, pp. 277-295.

To cite this article

Kris Deschouwer, «Introduction - The Belgian federalism and federation on the move», Fédéralisme Régionalisme [En ligne], Volume 20 : 2020, Trente ans de dynamiques fédérales et régionales, URL : https://popups.uliege.be/1374-3864/index.php?id=2044.

About: Kris Deschouwer

Emeritus Professor, Vrije Universiteit Brussel.