Engineering representation in multilevel democracies: The Belgian case illustrating reflections on bicameral systems

Organizar la representación en democracias multinivel: el caso belga para ilustrar reflexiones sobre los sistemas bicamerales

Frateur, Jakob

University of Antwerp. Faculty of Law (Bélgica – Belgium)

ORCID: https://orcid.org/0000-0002-5184-9405

jakob.frateur@uantwerpen.be

NOTA BIOGRÁFICA

Jakob Frateur is a doctoral researcher at the Faculty of Law and at the Department of Political Science (Faculty of Social Sciences) at the University of Antwerp. His research focuses on citizens’ political trust in multilevel systems in times of crisis with emphasis on (modes of) multilevel governance and institutional structures.

Meier, Petra

University of Antwerp. Department of Political Science (Bélgica – Belgium)

ORCID: https://orcid.org/0000-0003-4108-5875

petra.meier@uantwerpen.be

NOTA BIOGRÁFICA

Petra Meier is professor in Political Science, Faculty of Social Sciences, University of Antwerp. Her research focuses on the representation of gender and other social markers, with an emphasis on the reproduction of inequality through political systems, politics and policies. For an overview of her publications please see: https://www.uantwerpen.be/en/staff/petra-meier/publications/

Bursens, Peter

University of Antwerp. Departement of Political Science (Bélgica – Belgium)

ORCID: https://orcid.org/0000-0003-2096-6454

peter.bursens@uantwerpen.be

NOTA BIOGRÁFICA

Peter Bursens is professor in Political Science at the Faculty of Social Science of the University of Antwerp, member of the Politics and Public Governance research group, the GOVTRUST Centre of Excellence and chair of the Jean Monnet Centre STRATEGO. His research focuses on democratic legitimacy of multilevel political systems, federalism and EU-politics (see https://www.uantwerpen.be/en/staff/peter-bursens/publications/).

ABSTRACT

Objectives: In federal systems both the people as a whole (demos) and the statespeople (demoi) need to be represented, be it in a unicameral or bicameral set-up. In Belgium, this is increasingly debated, to the extent that the abolition of the second chamber is on the political agenda. What are the theoretical and empirical alternatives to organize representation in federal systems, and to what extent can they be applied to the Belgian federation? While literature on federal systems and most federations search for a more optimal representation of the demoi, Belgium is confronted with the opposite challenge: how to guarantee the representation of the demos in a set-up that favours representation of the demoi. Methodology: This is a theoretical and reflective text based on federalism theory and insights from, among others, EU studies. Results: In this article we argue that the solution to balance this representation can be both unicameral and bicameral. The crux in either is a change in the election of the parliamentarians representing the demos. This could be achieved by electing a part of the Chamber of Representatives through a nationwide, federal circumscription. Conclusions: The main take-away is the need to pay more attention to the representation of the demos at the federal level, especially in multi-level governance systems characterized by devolving tendencies.

KEYWORDS

Federalism; unicameralism; bicameralism; political representation; Belgium.

RESUMEN

Objetivos: en los sistemas federales, tanto el pueblo en su conjunto (demos) como los pueblos de las unidades constituyentes (demoi) deben estar representados, ya sea en un sistema unicameral o bicameral. En Bélgica, esta cuestión se debate cada vez más, hasta el punto de que la abolición de la segunda cámara figura en la agenda política. ¿Cuáles son las alternativas teóricas y empíricas para organizar la representación en los sistemas federales, y en qué medida pueden aplicarse a la federación belga? Mientras que la literatura sobre sistemas federales y la mayoría de las federaciones buscan una representación más óptima de los demoi, Bélgica se enfrenta al reto opuesto: cómo garantizar la representación del demos en una configuración que favorece la representación de los demoi. Metodología: se trata de un texto teórico y de reflexión, basado en la teoría del federalismo y en las aportaciones de, entre otros, los estudios sobre la UE. Resultados: en este artículo argumentamos que la solución para equilibrar esta representación puede ser unicameral o bicameral. El quid en ambas es un cambio en la elección de los diputados que representan al demos. Esto podría lograrse eligiendo a una parte de la Cámara de Diputados mediante una circunscripción federal de ámbito nacional. Conclusiones: la principal conclusión es la necesidad de prestar más atención a la representación del demos a nivel federal, especialmente en los sistemas de gobernanza multinivel caracterizados por la descentralización.

PALABRAS CLAVE

Federalismo; unicameralismo; bicameralismo; representación política; Bélgíca.

SUMARIO

INTRODUCTION. 1. REPRESENTATION IN THE BELGIAN BICAMERAL SYSTEM. 2. THEORY ON POLITICAL REPRESENTATION IN FEDERAL STATES. 3. EMPIRICAL EXAMPLES OF REPRE­SENTATION IN FEDERAL SYSTEMS. 4. WAYS FORWARD FOR THE BELGIAN CASE. CONCLUSION. ACKNOWLEDGEMENTS. REFERENCES.

INTRODUCTION

As federal constitutions divide the people into peoples (Benz and Sonnicksen, 2017Benz, A. and Sonnicksen, J. (2017). Patterns of federal democracy: tensions, friction, or balance between two government dimensions. European Political Science Review, 9(1), 3-25. https://doi.org/10.1017/S1755773915000259), federal systems tend to have two channels of representation to ensure that both the nation as a whole (“people” or demos) and the constituent units (“statespeoples” or demoi) are represented at the federal level (Meier and Bursens, 2021Meier, P. and Bursens, P. (2021). Belgium: The Democratic State of the Federation. In A. Benz and J. Sonnicksen (eds.), Federal democracies at work. Varieties of complex government (pp. 180-196). University of Toronto Press.). To this end, most federal states organize representation in a bicameral way, often with one chamber representing the constituent units and another one representing the nation as a whole. This twofold representation is put forward as a basic and to a certain extent even normative feature of federal systems and can be found in countries like Australia, Argentina, Austria, Brazil, Canada, Germany, India, Russia, Spain, Switzerland, and the US (Swenden, 2010Swenden, W. (2010). Subnational participation in national decisions: The role of second chambers. In H. Enderlein, S. Wälti and M. Zürn (eds.), Handbook on Multi-level Governance (pp. 103-123). Edward Elgar.).

The Belgian federal parliament also has two chambers, but the representation of the demos and the demoi differs from other federal states and is heavily disputed among political parties and academics. It makes Belgium an instrumental case to discuss and apply the theoretical and empirical models of representation in federal systems. For several years now, politicians in Belgium propagate the abolition of the Senate, which serves as locus of representation of the constituent units, thereby referring to a panoply of arguments, such as the lack of power of that institution, especially in relation to its costs, and its particracy facilitating incentives. One of these calls came from the then Senate president herself, the Dutch-speaking liberal Open VLD politician Stephanie D’Hose. She, too, argued that she had experienced that the Senate’s role is too limited to be a separate institution (Vanrenterghem, 2022Vanrenterghem, A. (2022). Senaatsvoorzitter Stephanie D’Hose (Open VLD) wil Senaat afschaffen. VRTNWS. https://www.vrt.be/vrtnws/nl/2022/01/08/senaatsvoorzitter-wil-senaat-afschaffen/; Arnoudt, 2023Arnoudt, R. (2023). Senaatsvoorzitter D’Hose wil akkoord over afschaffing Senaat: “In de praktijk niet zo simpel”. VRTNWS. https://www.vrt.be/vrtnws/nl/2023/05/22/senaatsvoorzitter-d-hose-wil-akkoord-over-afschaffing-senaat-i/). Indeed, the powers of the Senate were considerably scaled back during the sixth state reform in 2011. D’Hose followed in the footsteps of her party colleagues Guy Verhofstadt and Marc Verwilghen, who called for the Senate to be abolished in 2001 and 2008 respectively. While these early proposals met with a lot of resistance from some other parties, D’Hose’s recent appeal was more widely supported. The 2025 agreement underpinning the federal government following the 2024 election now even explicitly envisages a constitutional reform that will effectively abolish the Senate, thereby transforming Belgium into a unicameral system1.

While some of the arguments mainly serve political –public opinion nudging and agenda setting– purposes, all of them tend to ignore the representative function of the Senate, and generally of a bicameral system in a federal state. If at all, the representation argument runs that the Belgian Chamber of Representatives could also fulfil the representative functions of the Senate (Vanrenterghem, 2022Vanrenterghem, A. (2022). Senaatsvoorzitter Stephanie D’Hose (Open VLD) wil Senaat afschaffen. VRTNWS. https://www.vrt.be/vrtnws/nl/2022/01/08/senaatsvoorzitter-wil-senaat-afschaffen/). And, indeed, the Belgian system is particular in this respect. Officially the Senate is the locus of representation of the three Communities (and Regions, see next section) –the Dutch–, French-, and German-speaking language communities– and the Chamber of Representatives is the place where the Belgian people is represented. However, over time, Belgian constitutional and political dynamics have made the organization of political representation in the Chamber of Representatives reflect the constituent units at the disadvantage of rendering the voice of the people (FrateurFrateur, J., Bursens, P. and Meier, P. (2023). Exploring the Relationship between Federal Reform and the Representation of the People: The Case of the Belgian House of Representatives. Publius: the Journal of Federalism, 53(4), 618-641. https://doi.org/10.1093/publius/pjad006 et al., 2023Frateur, J., Bursens, P. and Meier, P. (2023). Exploring the Relationship between Federal Reform and the Representation of the People: The Case of the Belgian House of Representatives. Publius: the Journal of Federalism, 53(4), 618-641. https://doi.org/10.1093/publius/pjad006; Popelier and Lemmens, 2015Popelier, P. and Lemmens, K. (2015). The Constitution of Belgium: A Contextual Analysis. Bloomsbury Publishing.; Rummens and Sottiaux, 2014Rummens, S. and Sottiaux, S. (2014). Democratic legitimacy in the Bund or the “federation of states”. The cases of Belgium and the EU. European Law Journal, 20(4), 568-587. https://doi.org/10.1111/eulj.12053). Already in 1999, Tuschhoff (1999)Tuschhoff, C. (1999). The compounding effect: The impact of federalism on the concept of representation. West European Politics, 22(2), 16-33. https://doi.org/10.1080/01402389908425300 argued that the degree of autonomy of the substates has an effect on the representation of the constituent units at the federal level. More recently, Lacey (2017)Lacey, J. (2017). Centripetal democracy. Democratic legitimacy and political identity in Belgium, Switzerland and the European Union. Oxford University Press. argued in a similar way that the dominant frame of political reference in Belgium is shifting from the sovereign Belgian people as a whole to the constituent units. The overall dynamic of the federalization process thus had an impact on democratic practices such as representation, tending towards “demos constraining” effects (Sonnicksen, 2018Sonnicksen, J. (2018). Federalism and democracy: a tense relationship. In J. Tudella, M. Kölling and F. Reviriego (coords.), Calidad democrática y organización territorial (pp. 31-52). Marcial Pons.). In addition, and irrespective of the federalization process, MuellerMueller, S., Vatter, A. and Dick, S. (2021). A new index of bicameralism: taking legitimacy seriously. The Journal of Legislative Studies, 29(2), 312-336. https://doi.org/10.1080/13572334.2021.1996753 et al. (2021)Mueller, S., Vatter, A. and Dick, S. (2021). A new index of bicameralism: taking legitimacy seriously. The Journal of Legislative Studies, 29(2), 312-336. https://doi.org/10.1080/13572334.2021.1996753 categorized the Belgian Senate and Belgian bicameralism more broadly as having a low level of legitimacy, compared to other (federal) countries.

At first sight, the way representation is organized in the Belgian Chamber of Representatives makes the Senate seem obsolete. So, why not abolish it after all? But is the Senate really obsolete in the light of the prerequisite that both the people and the constituent units should be represented in federal systems? What costs would there be when it comes to the representation of the constituent units and that of the people if the Belgian federal parliamentary system were to become unicameral? Indeed, not all federations have a bicameral system, think of Micronesia or Venezuela. So, what lessons are there to learn, especially since Belgium is not the only country where the abolition of the Senate is on the agenda. Debates echoing the Belgian ones can also be found in Spain, for instance. Interesting in these debates is that they add a new focus to the literature, which mostly focuses on how to ensure representation of the demoi (and not the demos) at the federal level.

We take this Belgian debate as a starting point to explore alternative ways to organize representative democracy in federal states. This brings us to our research question: what are the theoretical and empirical options to organize representation in federal states, and to what extent can they be applied to the Belgian federation? To answer this question, the article is divided into four sections. In the first we briefly present the evolution of the Belgian parliamentary system with a focus on its bicameralism. Subsequently, we ­discuss ­literature rethinking the concept of representative democracy and political representation to establish ­possible alternative scenarios for a bicameral system in federal states. In the third section we analyze specific ­existing mechanisms in federal states meant to complement or replace the bicameral system. In the final ­section we evaluate the merits of the theoretical ideas and empirical examples for the Belgian case. The conclusion wraps up the main findings, sketches the broader lessons to learn from the Belgian case when it comes to engineering representation in multilevel democracies, and possible venues for further research.

1. REPRESENTATION IN THE BELGIAN BICAMERAL SYSTEM

Already at the time of the independence of Belgium in 1830, a bicameral system was chosen in a then unitary state following the examples of the Netherlands and the UK. As in many other cases, the Senate had to provide a counterweight to the new, progressive political elite that would sit in the Chamber of Representatives. Thanks to stricter election conditions (candidates had to be at least 40 years old and pay important property taxes), the more conservative Senate served as a reflection body that could subject the draft bills of the Chamber of Representatives to a second reading, thereby ensuring the interests of the (aristocratic) gentry (Passaglia, 2018Passaglia, P. (2018). Unicameralism, bicameralism, multicameralism: evolution and trends in Europe. Perspectives on Federalism, 10(2), 1-29. http://dx.doi.org/10.2478/pof-2018-0014; Popelier, 2018Popelier, P. (2018). Bicameralism in Belgium: the dismantlement of the Senate for the sake of multinational confederalism. Perspectives on Federalism, 10(2), 215-237.).

Throughout the twentieth century, the Senate was reformed on several occasions. While the Senate’s composition continued to differ from that of the Chamber of Representatives (including the direct successors of the Belgian crown, and provincial and co-opted senators), both chambers also increasingly resembled each other: the specific conditions to stand for Senate elections were abolished, both institutions were endowed with the right to initiate legislation, to approve a new government and its coalition agreement, to vote the annual budget, and to control the government and its members throughout the legislature.

The fully bicameral system lasted until 1993, when the fourth constitutional reform of the Belgian state, officially enshrining the federal system which had evolved since the 1960s (Deschouwer, 2012Deschouwer, K. (2012). The Politics of Belgium. Governing a Divided Society. Palgrave Macmillan.), introduced a functional split between the Chamber of Representatives and the Senate. The center of gravity of the legislature was assigned to the former while the Senate lost some of its powers. Based on the argument that ­federalism and bicameralism go hand in hand (Popelier, 2018Popelier, P. (2018). Bicameralism in Belgium: the dismantlement of the Senate for the sake of multinational confederalism. Perspectives on Federalism, 10(2), 215-237.), the Senate became a reflection chamber representing the constituent units. To that end, the provincial Senators were replaced by Senators to be appointed by the language groups. More particularly, these Senators combined a seat in respectively the Flemish Parliament, the parliament of the French-speaking community, and that of the German-speaking community with a seat in the Senate. The first two parliaments were entitled to ten Senators each while the latter was entitled to one Senator, reflecting the de facto bipolar character of the country and ensuring that the shares of the language groups in the new Senate were proportional to their share of the Belgian population.

This proportionality distinguished the Belgian Senate from many other second chambers in federal parliaments where constituent units comprise the same number of Senators notwithstanding demographic differences. Formally, the Senate did not fully represent the constituent units as the majority of Senators was still directly elected by the people rather than being appointed by the parliaments representing the language groups or than being directly elected by the statespeople. In practice, however, it very much functioned as a second chamber representing the constituent units. Firstly, the directly elected Senators (as much as the other ones) represented linguistically split parties which mainly or exclusively focus on voters from the same language group (Caluwaerts and Reuchamps, 2014Caluwaerts, D. and Reuchamps, M. (2014). Combining federalism with consociationalism. Is Belgian consociational federalism digging its own grave? Ethnopolitics, 14(3), 277-295. https://doi.org/10.1080/17449057.2014.986866; Dandoy, 2013Dandoy, R. (2013). Belgium: toward a regionalization of national elections? In R. Dandoy and A. H. Schakel (eds.), Regional and national elections in Western Europe. Territoriality of the vote in thirteen countries (pp. 47-67). Palgrave Macmillan.; Dandoy and Schakel, 2013Dandoy, R. and Schakel, A. H. (2013). Introduction: territoriality of the vote: a framework for analysis. In R. Dandoy and A. H. Schakel (eds.), Regional and national elections in Western Europe. Territoriality of the vote in thirteen countries (pp. 1-26). Palgrave Macmillan.; Swenden, 2013Swenden, W. (2013). Conclusion: The future of Belgian federalism – between reform and swansong? Regional and Federal Studies, 23(3), 369-382. https://doi.org/10.1080/13597566.2013.773897, Mueller, 2024aMueller, S. (2024a). Shared Rule in Federal Theory and Practice. Oxford University Press. https://doi.org/10.1093/9780191991288.001.0001), and secondly, the mechanisms protecting language minorities at the federal parliamentary level stimulated Senators to think and act in terms of language groups (Hooghe, 2004Hooghe, L. (2004). Belgium: hollowing the center. In U. Amoretti and N. Bermeo (eds.), Does federalism matter? Political institutions and the management of territorial cleavages (pp. 55-92). Johns Hopkins Press.).

The sixth state reform of 2011 substantially changed the bicameral character by turning the Belgian Senate even more into a Senate of the constituent units, but at the same time also strongly diminishing its legislating and scrutiny powers, making it a non-permanent chamber meeting but nine times a year (Deschouwer, 2012Deschouwer, K. (2012). The Politics of Belgium. Governing a Divided Society. Palgrave Macmillan.). Senators were no longer directly elected. Except for the ten co-opted Senators all other fifty are now appointed by the parliaments of the constituent units. based on the results of the regional elections. The overall composition of the Senate is proportional (29 Dutch-speaking, 20 French-speaking and one German-speaking Senator, as well as six Dutch-speaking and four French-speaking co-opted Senators (DandoyDandoy, R., Dodeigne, J., Reuchamps, M. and Vandeleene, A. (2015). The New Belgian Senate. A (Dis)Continued Evolution of Federalism in Belgium? Representation, 51(3), 327-339. https://doi.org/10.1080/00344893.2015.1108358 et al., 2015Dandoy, R., Dodeigne, J., Reuchamps, M. and Vandeleene, A. (2015). The New Belgian Senate. A (Dis)Continued Evolution of Federalism in Belgium? Representation, 51(3), 327-339. https://doi.org/10.1080/00344893.2015.1108358), reflecting its role as the locus of representation of the constituent units at the federal level2. Furthermore, the Chamber of Representatives and the Senate no longer share equal powers. While the former has full powers, the Senate has legislative power only in the case of constitutional amendments, and legislation related to the state structure, party funding and the German-speaking community. In addition, the Senate can act in the event of conflicts of interest among constituent units, can draw up information reports on cooperation between those units, checks the principle of subsidiarity –especially in the context of the EU– and is involved in the appointments of a number of high courts’ judges.

All this leads Brenton (2022)Brenton, S. (2022). Does federalism enhance representative democracy? Perpetual reform and shifting power in a divided Belgium. Journal of Contemporary European Studies, 31(2), 500-518. https://doi.org/10.1080/14782804.2021.2023482 to argue, that while, all in all, the process of federalization has strengthened the features of representative democracy in Belgium (see also van Haute and Deschouwer, 2018Van Haute, E. and Deschouwer, K. (2018). Federal reform and the quality of representation in Belgium. West European politics, 41(3), 683-702. https://doi.org/10.1080/01402382.2017.1399320), it has also created some downsides. A lack of direct election of the Senators undermines accountability of representatives to their voters, while the curtailing of the Senate’s powers decreases the capacity to control the executive and thus the executive’s accountability to the legislative.

The sketched evolution increasingly triggered calls to turn the Belgian system into unicameralism, effectively endowing the Chamber of Representatives with all legislative powers. It is argued that this would not endanger the representation of the constituent units as the Chamber of Representatives is elected in unilingual provincial circumscriptions and consists of a French- and Dutch-speaking language group, both of which consist of different parties across the linguistic divide. There are two exceptions to this. The first is that MPs elected in the bilingual electoral circumscription of the Brussels Capital Region have to make explicit to which language group they belong, which can actually be deduced from the party to which they belong. The second is that in the Chamber of Representatives German-speaking MPs belong to the French-speaking language group. Unlike in the Senate, they do not constitute a separate language group in the Chamber of Representatives.

In addition, for Dutch-speaking Flemish nationalist parties such as N-VA and the populist radical right VB, the Senate is also the symbol of a unified Belgium, so they would rather see it abolished. Finally, critics point to the presence of co-opted Senators as a textbook example of the omnipresent particracy still strongly present in Belgium (BaudewynsBaudewyns, P., Brans, M., Reuchamps, M., Rihoux, B. and Van Ingelgom, V. (eds.) (2022). The Winter of Democracy. Partitocracy in Belgium. Press Universitaires de Louvain. et al., 2022Baudewyns, P., Brans, M., Reuchamps, M., Rihoux, B. and Van Ingelgom, V. (eds.) (2022). The Winter of Democracy. Partitocracy in Belgium. Press Universitaires de Louvain.). The fact that party leaders appoint part of the Senators does indeed strengthen their position and at the same time weakens the position of the individual politician who wants to have a chance of winning a seat as a Senator.

Opponents of unicameralism point to the role of the Senate with regard to constitutional amendments and changes to the state structure. Because of the control that the Senate can exercise over such bills, a proposal for a constitutional amendment must be well prepared and cannot be rushed. In addition, the Senate still has some functions at the international level. Not only does it check compliance with the principle of subsidiarity in the EU, but the constituent units also have access to international parliamentary organizations and the parliamentary assemblies of international organizations through their Senators. In addition, the Senate has a control and arbitration function vis-à-vis the constituent units, and between the federal state and the constituent units. By bringing together representatives of the constituent units as representatives of their statespeople or demoi in one chamber, potential conflicts between the regions and the communities can be resolved more easily.

It is the argument about representation that we turn to in this article. Many of the functions of the current Senate could indeed be executed by the Chamber of Representatives –or another institution. Its abolition would not prevent the Belgian political system from functioning. Our question, however, is what an abolition implies for political representation in a then unicameral federal system. In the next section, we discuss literature that discusses the concept of representative democracy and political representation to establish some alternative scenarios for a bicameral system in federal states, and Belgium in particular.

2. THEORY ON POLITICAL REPRESENTATION IN FEDERAL STATES

Federations need to find a balance between self-rule and shared rule (Elazar, 1987Elazar, D. J. (1987). Exploring federalism. University Alabama Press.), and this also applies to their processes of political representation. Indeed, an adequate representation is considered an essential part of the democratic process which is deemed crucial for the legitimacy of a democratic political system and is as such regarded as a main institutional feature of modern democracy (Scherz, 2018Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267). Representation is one of the functions through which democracy in its understanding as rule by the people, materializes, and the most explicit and main step through which most citizens partake in the political management of their society. Through their vote they delegate both political decision-making and the control of the executive. Their representative(s) are held accountable for their acts at the occasion of the next elections. “Democratic representation can then be understood as a dynamic process between moments of authorisation and accountability” writes Scherz (2018, p. 497)Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267, referring to Young (2002)Young, I. M. (2002). Inclusion and Democracy. Oxford Academic..

Without being much explained –rather it is mainly normatively put forward– federal theory holds the need for a twofold representation, that of the people as a whole and that of the constituent units, to find a balance between self-rule and shared rule. Therefore, most federal states organize representation in a bicameral way, often with a second chamber representing the constituent units and a first representing the people as a whole, so as to make sure that both the people and the constituent units are represented at the federal level (Meier and Bursens, 2021Meier, P. and Bursens, P. (2021). Belgium: The Democratic State of the Federation. In A. Benz and J. Sonnicksen (eds.), Federal democracies at work. Varieties of complex government (pp. 180-196). University of Toronto Press.). At least one reason can be put forward for this logic. As both the people as a whole and the constituent units are separately recognized as entities sui generis, both need a voice in the political management of their society as a whole. While the constituent units dispose of their own institutions, they, as much as the people as a whole, must be represented at the federal level. While Palermo (2018)Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016 argues that this link can also, or even better, be established through executives instead of legislatives, we contend that the latter implies participation of a majority which is not the same as representation of all. In the same vein, the people cannot be reduced to a sum of the constituent units and their interests in a federal system, but has to be able to express itself as such. This is, ideal typically, what distinguishes a federation from a confederation or an intergovernmental system in which the demos has no direct link with the encompassing executive.

Scherz (2018)Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267, arguing for the recognition of the EU as a multilateral democracy or demoicracy, makes exactly such an argument, when she emphasizes that “multilateral democracy is sometimes mistaken as an intergovernmental account that gives priority to the peoples and neglects the importance of individual citizens”. (Scherz, 2018, p. 496Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267). In the case of the EU, not only the interests of the member states, the demoi, should be represented. The demos, i.e. the transnational interests of the European citizens and the interests of these citizens in the system as a whole, should be represented as well. “Multilateral democracy as defended in this paper is based on the autonomy of persons in the sense of personal and political autonomy that is mutually realized through individual rights and collective self-determination. From this perspective, the intergovernmental way of representation disregards (…) the transnational interests of individuals and their interests in the system as a whole, which are not represented sufficiently by their respective states, [and] should also be taken into account”. (Scherz, 2018, pp. 497-498Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267). This makes her plead for a bicameral representative body as the basic representative structure for the EU. Such a body with equal direct representation by the peoples and the EU citizens as such, is backed by the argument that citizens “can have other interests in the overall system than they do in the domestic system”. (Scherz, 2018, p. 500Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267).

Indeed, such a difference in interests –and needs, considering the broader literature on the concept of political representation (Urbinati, 2006Urbinati, N. (2006). Representative Democracy. Principles and Genealogy. University of Chicago Press.)– could simply be the consequence of a division of policy competencies across different levels in a multilateral democracy. Such is the case in the EU but also in federal systems. And citizens might not have the same position on all these policy issues. But even without such a division of policy competencies across different policy levels, citizens can have other interests in the overall system than they do at the constitutive level. They might position themselves differently with respect to the citizens across the EU –or the people as a whole in the case of a federation– than with respect to their fellow citizens –or their statespeople in the case of a federation. Taking the broader literature on the concept and process of political representation into consideration, and considering the fact that interests and needs are not simply given but constituted (Saward, 2010Saward, M. (2010). The representative claim. Oxford University Press.), would allow to develop these arguments for a twofold representation of citizens at the upper level of multilateral democracies even further. But even without doing so, Scherz’ point becomes apparent. Citizens in multilateral democracies require a double representation at the upper level, that of the people as a whole and that of the constituent units. Otherwise, citizens as the people would not dispose of a voice in the governance of their society as a whole. They would only have a direct say in the constituent entities, not in what the latter –in an intergovernmental or confederal logic– decide to undertake jointly.

While Scherz argues for an improved system of political representation at the EU level, her argument also fits federal systems: the citizens of a federation can have other interests in the overall system than in the system of the constituent unit, for instance because the levels govern different policy domains. Therefore, federations need “to manage the relationship between a demos of some kind and multiple strong demoi” (Lacey, 2017Lacey, J. (2017). Centripetal democracy. Democratic legitimacy and political identity in Belgium, Switzerland and the European Union. Oxford University Press.). This explains not only the need for an adequate system and process of political representation at the level of the constituent unit. It also explains the need for a system of twofold representation at the federal level, of the people and of the statespeoples or constituent units. And while such mechanisms of representative democracy may be complemented by procedures and tools of participatory and/or direct democracy (cf. Brenton, 2022Brenton, S. (2022). Does federalism enhance representative democracy? Perpetual reform and shifting power in a divided Belgium. Journal of Contemporary European Studies, 31(2), 500-518. https://doi.org/10.1080/14782804.2021.2023482; Caluwaerts, 2012Caluwaerts, D. (2012). Confrontation and communication: Deliberative democracy in divided Belgium. Peter Lang Publishers.; Caluwaerts and Deschouwer, 2014Caluwaerts, D. and Deschouwer, K. (2014). Building bridges across political divides: Experiments on deliberative democracy in deeply divided Belgium. European Political Science Review, 6(3), 427-450. https://doi.org/10.1017/S1755773913000179; CaluwaertsCaluwaerts, D., Jacquet, V., Minsart, E., Junius, N. and Acar, L. (Forthcoming). Democratic innovation in Belgium. In M. Reuchamps, M. Brans, P. Meier and E. van Haute (eds.), The Oxford Handbook on Belgian Politics. Oxford University Press. et al., forthcomingCaluwaerts, D., Jacquet, V., Minsart, E., Junius, N. and Acar, L. (Forthcoming). Democratic innovation in Belgium. In M. Reuchamps, M. Brans, P. Meier and E. van Haute (eds.), The Oxford Handbook on Belgian Politics. Oxford University Press.; Junius, 2023Junius, N. (2023). Legitimacy in the participants’ eyes: A call for participants’ authorship over mini-public design in Brussels. Critical Policy Studies, 17(4), 524-543. https://doi.org/10.1080/19460171.2022.2157297, 2024Junius, N. (2024). From representative to represented mini-publics: How mini-publics’ outputs are shaped by representation. Political Studies Review, 22(4), 914-929. https://doi.org/10.1177/14789299231217619; Niessen and Reuchamps, 2022Niessen, C. and Reuchamps, M. (2022). Institutionalising citizen deliberation in parliament: The Permanent Citizens’ Dialogue in the German-speaking Community of Belgium. Parliamentary Affairs, 75(1), 135-153. https://doi.org/10.1093/pa/gsaa056), or bilateral relations between levels (Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016), the former nonetheless constitute a cornerstone of contemporary democracies. Representation, in this way, serves as a “vital bridging device between federalism and democracy” (Fossum and Jachtenfuchs, 2017Fossum, J. E. and Jachtenfuchs, M. (2017). Federal challenges and challenges to federalism. Insights from the EU and federal states. Journal of European Public Policy, 24(4), 467-485. https://doi.org/10.1080/13501763.2016.1273965).

Fenna and Schnabel (2024)Fenna, A. and Schnabel, J. (2024). What is Federalism? Some Definitional Clarification. Publius: The Journal of Federalism, 54(2), 179-200. https://doi.org/10.1093/publius/pjad034, in an attempt to conceptually clarify federalism as a principle of state organization, have a different reading of a bicameral system. To understand it, one has to return to their premise, which is “the autonomy of two orders of government as the essential, defining, feature of ­federalism” (Fenna and Schnabel, 2024, p. 180Fenna, A. and Schnabel, J. (2024). What is Federalism? Some Definitional Clarification. Publius: The Journal of Federalism, 54(2), 179-200. https://doi.org/10.1093/publius/pjad034). According to them, “This encompasses three main requirements: a ­division of powers between, archetypally, a central government and a set of constituent units; a direct relationship of each government with the people based on meaningful powers; and constitutional ­entrenchment. (…) Bicameralism exists in some form or another in all democratic federations and may represent an ­enhancement of federal practice, but it is not what makes a country federal”. (Fenna and Schnabel, 2024, pp. 180-181Fenna, A. and Schnabel, J. (2024). What is Federalism? Some Definitional Clarification. Publius: The Journal of Federalism, 54(2), 179-200. https://doi.org/10.1093/publius/pjad034). In sum, bicameralism is not a necessary feature of any federal system, and, indeed, not all federations do have a bicameral representation (Mueller, 2024bMueller, S. (2024b). Bicameralism. In A. Vatter and R. Freiburghaus (eds.), Handbook of Comparative Political Institutions (pp. 263-275), Edward Elgar Publishing.). The main reason for that lies in the fact that shared rule, ­according to them, does not entail the involvement of constituting units in central government ­decision-making. Based on this understanding of a federal system, constituent units need not be represented at the central level and thus there is no need for a bicameral system.

Swenden (2010, p. 120)Swenden, W. (2010). Subnational participation in national decisions: The role of second chambers. In H. Enderlein, S. Wälti and M. Zürn (eds.), Handbook on Multi-level Governance (pp. 103-123). Edward Elgar., also argues that “second chambers should neither be seen as a necessary nor as a sufficient condition for the proper functioning of multi-level governance” but is a bit more nuanced when it comes to the representation of constituent units at the federal level. Like Fenna and Schnabel (2024)Fenna, A. and Schnabel, J. (2024). What is Federalism? Some Definitional Clarification. Publius: The Journal of Federalism, 54(2), 179-200. https://doi.org/10.1093/publius/pjad034, he writes that it is not necessary for purely federal matters. However, “one would expect the second chamber to have a right to introduce, amend or veto constitutional changes that alter the balance of powers between the federal and regional levels of government; to co-decide on the regional distribution of federal grants or to influence the rate of federal taxes if part of their revenue accrues to the regions” (Swenden, 2010, p. 112Swenden, W. (2010). Subnational participation in national decisions: The role of second chambers. In H. Enderlein, S. Wälti and M. Zürn (eds.), Handbook on Multi-level Governance (pp. 103-123). Edward Elgar.). He underlines, though, that subnational representation can also be organized in another way than through bicameralism, for example through inter-executive relations (see also Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016). Popelier (2021)Popelier, P. (2021). Dynamic federalism. A new theory for cohesion and regional autonomy. Routledge., too, argues that there is no necessary link between federalism and bicameralism. Federalism, according to her, can be defined as seeking a balance between integrity and diversity. This balance is indeed reflected in bicameralism, yet especially in multinational states diversity prevails. Hence, bicameralism may not be the optimal design.

Returning to Belgium, a textbook case of executive federalism (Niessen and Popelier, forthcoming), this implies that the Senate could be abolished. However, as mentioned before, the crucial point is that the Chamber of Representatives is organized in a way that MPs represent the two main language communities. The issue is thus not only whether a unicameral system would suffice, the question is how to deal with the demos in such a system.

3. EMPIRICAL EXAMPLES OF REPRESENTATION IN FEDERAL SYSTEMS

Most federations organize twofold representation at the federal level in a bicameral way by means of a directly elected chamber as representative of the people and a chamber composed of representatives of the constituent units. Second chambers differ around the world, both in their composition and their powers, and hence also in the way the constituent units’ representation is performed at the central level. In some countries, members of second chambers are directly elected by substate constituencies, while in other countries, the members are selected in the substates’ parliaments or even governments.

In the US, but also in Argentina, Brazil, Switzerland, and Nigeria, the members of the second chamber are directly elected in substate constituencies (Belser, 2018Belser, E. M. (2018). The Swiss Ständerat: a model of perfect bicameralism. Perspectives on federalism, 10(2), 152-181.; Popelier, 2021Popelier, P. (2021). Dynamic federalism. A new theory for cohesion and regional autonomy. Routledge.). Each substate gets to elect an equal number of representatives that should represent the substate at the central level. In the most famous example, the US, each state elects two Senators (art. 1.3 US Constitution) and each bill should pass in both the House and the Senate to ensure the substate interests are taken into account (Russell, 2001Russell, M. (2001). The territorial role of second chambers. The Journal of Legislative Studies, 7(1), 105-118. https://doi.org/10.1080/714003852). The House, which is the venue for representation of the demos, is also directly elected but each state gets seats proportional to the size of its population (art. 1.1 US Constitution). Contrary to the Senate, states do not have an equal number of votes. Popelier (2021)Popelier, P. (2021). Dynamic federalism. A new theory for cohesion and regional autonomy. Routledge. argues that, when the members of the second chamber are directly elected, the second chamber does not represent the constituent units but rather the subnational electorate as there is no institutional link to the substate.

This is slightly different in countries where the second chamber consists of members appointed or selected by the substate parliaments, as is the case in India, Austria, BiH and Malaysia (Doria, 2006Doria, G. (2006). The paradox of federal bicameralism. European Diversity and Autonomy Papers, (5), 5-43. https://hdl.handle.net/10863/7391), as well as for the majority of the Senators in Belgium (art. 67 Belgian Constitution). In Austria, for example, the demos is represented in the National Council, which is directly elected by the federal nation (art. 26 Austrian Constitution). The demoi are represented in the Federal Council, whose members are appointed by and from the parliaments of the substates proportional to the size of the population of the constituent unit. In Belgium, representation of the constituent units is organized in the same way as in Austria. The Senate, composed of appointed members of the constituent units’ parliaments and coopted members, is a meeting place for the constituent units at the central level, while, at least in theory, the directly elected Chamber of Representatives is the venue of representation of the demos.

Germany is an example of constituent unit representation in a second chamber being based on its members being selected in the substates’ governments. The Bundesrat functions as the institution to ­represent the constituent units (Länder) at the federal level, while the directly elected Bundestag represents the German people (arts. 38 and 50 German Constitution). The Bundesrat is composed of members of the government of the Länder, with at least three and a maximum of six representatives per Land, depending on the size of the population (art. 51 German Basic Law). These members are instructed by their governments and vote as a bloc per Land (Niedobitek, 2018Niedobitek, M. (2018). The German Bundesrat and executive federalism. Perspectives on Federalism, 10(2), 198-214.). Furthermore, the Bundesrat has the right of initiative and a suspensive veto that can be overruled by a majority or a two third majority, depending on the majority that supports the suspensive veto, in the Bundestag (Niedobitek, 2018Niedobitek, M. (2018). The German Bundesrat and executive federalism. Perspectives on Federalism, 10(2), 198-214.). In certain matters that affect the Länder, the Bundesrat has an absolute veto (Niedobitek, 2018Niedobitek, M. (2018). The German Bundesrat and executive federalism. Perspectives on Federalism, 10(2), 198-214.). For these reasons, regarding the rather unique composition and the strong powers of the second chamber, the German federal system is often seen as an example of a federal system with successful constituent unit representation in the second chamber (e.g., Doria, 2006Doria, G. (2006). The paradox of federal bicameralism. European Diversity and Autonomy Papers, (5), 5-43. https://hdl.handle.net/10863/7391; Niedobitek, 2018Niedobitek, M. (2018). The German Bundesrat and executive federalism. Perspectives on Federalism, 10(2), 198-214.; Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016). Indeed, the representatives in the Bundesrat have a direct institutional link to the Land governments while being able to represent their Land in decision-making through the bloc vote (Popelier, 2021Popelier, P. (2021). Dynamic federalism. A new theory for cohesion and regional autonomy. Routledge.).

Some federations have a unicameral parliamentary system at the federal level and try in different ways –and with varying degrees of success– to make the representation of the demos and the demoi coincide in that chamber. They either dispose of some kind of reserved seats for representatives elected in a statewide circumscription or of rather symbolic solutions.

In Micronesia, an island state consisting of three substates, for example, one member per state is directly elected at large on the basis of state equality, while the other representatives are elected in small electoral districts in the substates (art. 8 Micronesian Constitution). The number of the latter representatives per substate is proportional to the size of the substate, but at least one representative per substate is elected (art. 8 Micronesian Constitution). Also, bills must be voted by a two third majority and later by a two third majority of the state delegations –which comprise all MPs of a state voting in bloc. In Venezuela, a similar arrangement exists as the National assembly consists of directly elected members proportional to the population of the substates and of three additional appointed members per substate (art. 186 Venezuelan Constitution). In Saint-Kitts and Nevis, some members of the National Assembly are appointed by the prime minister and the opposition leader, while others are directly elected in smaller constituencies within each substate (arts. 29 and 30 Constitution of Saint-Kitts and Nevis).

A particular case is the United Arab Emirates (UAE), where in theory the constituent units themselves choose how the representatives at the federal level are selected (art. 69 UAE constitution). In practice, representatives are often appointed by the rulers of the Emirate. Depending on this, the individual Emirates delegate a number of representatives, while the rest of the parliament is indirectly elected in electoral districts within the constituent units. Also here, there are no representatives elected by the people as a whole. This is solved by a sentence in the constitution that says that representatives need to represent the entire people of the UAE and not merely their Emirates (art. 77 UAE constitution). In practice, of course, this cannot be controlled and there are no guarantees that the demos and the demoi are both adequately represented. Hence, this can be seen as a symbolic solution to the issue of twofold representation.

As underlined in the previous section, ways to realize double representation can also be found outside the realm of parliament. Palermo (2018)Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016 cites, for example, the existence of strong territorially bound ­parties as a way of representing the demoi at the federal level. Another way to realize twofold representation beyond the realm of parliament consists in strengthening the possibility of concluding bilateral respectively multilateral agreements between constituent units (Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016). Such agreements would be concluded between the governments of the constituent units mutually or between the governments of the constituent units and the federal government.

4. WAYS FORWARD FOR THE BELGIAN CASE

While bicameralism is not a necessary feature of federalism, many unicameral solutions implemented in other federal states may not really be applicable in the Belgian context. The main reason is that these unicameral designs seek to ensure the representation of the demoi on top of the already secured representation of the demos. In Belgium, the challenge is the other way around: as the demoi are already overrepresented compared to the demos in both chambers, it is the latter that needs to be provided with sufficient representation. In addition, an abolition of the Senate and the transfer of its functions to the Chamber of Representatives results in not even all constituent units being represented at the federal level, as the German-speaking community would lose its guaranteed representation. From a democratic point of view, this possible loss of representation of one of the Communities is problematic because the bipolar character of the Belgian federation would be further strengthened to the detriment of the other communities.

Similarly, the alternative channels discussed in the previous section serve the representation of the demoi at the federal level, not that of the demos. Furthermore, they would further exacerbate some features of the Belgian federation which are already heavily criticised. Ensuring the representation via territorially bound parties acting at the federal level is already the case in Belgium, where territorial –and at the same time language-based– parties stand for election in the different regions, though not in bilingual Brussels where parties of both language groups are present. Such a solution through party politics, however, might increase the dominance of parties even more. While party leaders have always been very influential and party discipline has always been very strong, the federalization process has actually strengthened the power of party leadership even more (Pilet and Meier, 2018Pilet, J.-B. and Meier, P. (2018). Ze halen hun slag wel thuis: Over particratie en het aanpassingsvermogen van Belgische partijen. Res Publica, 60(4), 321-345. https://openjournals.ugent.be/rp/article/id/74682/). In this respect, stronger regionally based parties that appoint MPs to represent the region might not necessarily strengthen representation of the demoi, but it would in any case lead to even stronger parties, enhance party discipline and thus further strengthen the partitocracy in Belgium (Mueller, 2024aMueller, S. (2024a). Shared Rule in Federal Theory and Practice. Oxford University Press. https://doi.org/10.1093/9780191991288.001.0001).

Likewise, the introduction of bilateral agreements between the constituent units (and the federal level) does not match the Belgian context either. Meier and Bursens (2021)Meier, P. and Bursens, P. (2021). Belgium: The Democratic State of the Federation. In A. Benz and J. Sonnicksen (eds.), Federal democracies at work. Varieties of complex government (pp. 180-196). University of Toronto Press. have shown that Belgian executives already dominate the legislatives. An enhanced executive federalism would distort this power balance even further at the disadvantage of the legislative. Legislative control over the executive would be reduced while also the legislative initiative of the legislature would be further severely restricted. In addition, there is a risk that relations between the units will be skewed as the constituent units’ governments can conclude different agreements with the federal level leading to an even more complicated state structure (Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016). Furthermore, there is a risk that a wealthier and larger constituent unit will dominate the other(s), and even the federal level, by strengthening its position in negotiating agreements (Palermo, 2018Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016). In Belgium, this risk is particularly high due to the limited powers that the federal level still holds.

In short, unicameralism as suggested above does not fit with the defining features of current Belgian federalism. A unicameral set-up would only satisfy the requirements of a representation of the demos when the composition of this sole chamber is drastically changed. If a representation of the statespeople should be maintained, one part of the MPs can be elected as it currently happens, i.e. via provincial districts which results in a de facto representation of the demoi. To this should be added a representation of the German-speaking demoi so as to make sure that all communities are represented. In order to fully respect the constitution, it would also require a representation of all regions, which might be especially challenging for Brussels. Representatives of the statespeople could also be appointed by the parliaments or the governments of the constituent units. However, the real challenge would be the representation of the people in such a unicameral chamber. This could be achieved by MPs representing the demos to be elected via an electoral district comprising the whole Belgian territory. A proposal in that direction has been put forward in the past by the Pavia group (Deschouwer and Van Parijs, 2009Deschouwer, K. and Van Parijs, P. (eds.) (2009). Electoral Engineering for a Stalled Federation [Re-Bel e-book 4]. Re-bel initiative. https://rethinkingbelgium.eu/ebook/electoral-engineering-for-a-stalled-federation/), whereby the idea was to elect some of the MPs of the Chamber of Representatives in a federal district. Proposals varied between 15 out of the existing 150 members of the Chamber of Representatives or adding another 15 to those 150. The underlying idea was twofold. The first was to trigger parties to consider the whole population and not only that of their own language community when presenting their electoral program. The second consisted in allowing the population to hold all parties of the previous legislature accountable and not only those of their language community (see also Van Parijs, 2000Van Parijs, P. (2000). Power-Sharing versus Border-Crossing in ethnically divided societies. In I. Shapiro and S. Macedo (eds.), Designing Democratic Institutions (pp. 296-320). NYU Press.) suggestion to give each voter two votes, similar to the logic of the German Bundestag, one for a smaller circumscription, one for the people as a whole in a country wide circumscription). In a unicameral set-up, the number of MPs elected to represent the people would need to be higher as compared to the proposal of the Pavia Group, but the logic of the proposal could be applied.

The alternative to a unicameral solution would be a bicameral design with a preservation of the Senate, yet in a reformed version that optimally represents the demoi, in combination with the Chamber of Representatives, equally reformed to more optimally represent the demos. As the main cleavage in Belgian politics is defined by the language communities, the Senate could be organized as a chamber that represents the language communities. It seems appropriate to abolish the system of co-opted Senators as the latter decrease the demoi-representing character of the Senate to the advantage of (the leadership of) political parties. The Senate could be upgraded in terms of competencies to give the demoi legislative power. And its role of a venue in which the different interests of the states peoples can be discussed and reconciled could be strengthened.

Even in a bicameral system, it is necessary to strengthen the representation of the demos. The most logical way forward to do so is by electing the representatives via a federal electoral district (although such a reform may not be without flaws itself, see e.g. Stojanovic, 2021Stojanovic, N. (2021). Multilingual democracy: Switzerland and beyond. Rowman and Littlefield.). One crucial argument against this is, is that this can only work with statewide parties, which have almost all disappeared since several decades (only the radical left is organized in a federal way). The reinstalment of such federal parties is unlikely given the political dynamics. However, statewide parties (or integrated politics, Thorlakson, 2020Thorlakson, L. (2020). Multi-level Democracy: Integration and Independence Among Party Systems, Parties, and Voters in Seven Federal Systems. Oxford Academic.) are not a necessity in a federal circumscription. If designed carefully (cf. Deschouwer and Van Parijs, 2009Deschouwer, K. and Van Parijs, P. (eds.) (2009). Electoral Engineering for a Stalled Federation [Re-Bel e-book 4]. Re-bel initiative. https://rethinkingbelgium.eu/ebook/electoral-engineering-for-a-stalled-federation/ for details), independent politics as well (Thorlakson, 2020Thorlakson, L. (2020). Multi-level Democracy: Integration and Independence Among Party Systems, Parties, and Voters in Seven Federal Systems. Oxford Academic.) would allow parties to address voters beyond their language group without the fear of losing votes as a language group. Indeed, such a measure would stimulate party competition among the parties of the same language group to address voters of the other language group which could result in a representation of the demos through non-statewide parties.

CONCLUSION

In federal systems both the people as a whole and the statespeople need to be represented, but this should not necessarily be organized through a bicameral set-up. In Belgium, this is an increasingly debated issue, exemplified by the fact that the abolition of the second chamber is firmly on the political agenda. While this debate is not unique to the Belgian case, cues from other federal systems are hard to apply in Belgium as these tend to focus on how to strengthen the representation of the demoi at the federal level. In Belgium, the opposite challenge is true: not the representation of the demoi but of the demos is at stake. The solution to balance this representation can be unicameral or bicameral. The crux in either of these is to guarantee the election of parliamentarians that represent the demos as whole, i.e. to establish a nationwide, federal circumscription. To maintain such a representation of the constituent units at the federal level –which, strictly speaking, is not necessary– there are two scenarios. In a unicameral set-up this would require a representation of the German-speaking community, the only official community not as such represented in the current Chamber of Representatives, and a better representation of the other constituent units, the regions. In a bicameral set-up this would require an upgrade of the Senate to a fully-fledged chamber of the constituent units on top of the discussed electoral reform of the Chamber of Representatives.

We worked under the assumption that the political and sociological nature of Belgium fits a federal set-up. Indeed, in a confederal design, the Chamber of Representatives and the Senate could merge as the direct representation of the people would be unnecessary. However, as long as the Belgian state architecture is federal and the federal level disposes of (exclusive) competencies that directly affect the whole population a representation of the demos is required.

The main take-away beyond the Belgian case is the need to pay more attention to the representation of the demos at the federal level, especially in multi-level governance systems characterized by devolving tendencies, and especially in times when electoral democracy is eroding.

ACKNOWLEDGEMENTS

We would like to thank the anonymous reviewers and the editorial team of the journal for their suggestions and comments on earlier versions of this article.

REFERENCES

Arnoudt, R. (2023). Senaatsvoorzitter D’Hose wil akkoord over afschaffing Senaat: “In de praktijk niet zo simpel”. VRTNWS. https://www.vrt.be/vrtnws/nl/2023/05/22/senaatsvoorzitter-d-hose-wil-akkoord-over-afschaffing-senaat-i/

Baudewyns, P., Brans, M., Reuchamps, M., Rihoux, B. and Van Ingelgom, V. (eds.) (2022). The Winter of Democracy. Partitocracy in Belgium. Press Universitaires de Louvain.

Belser, E. M. (2018). The Swiss Ständerat: a model of perfect bicameralism. Perspectives on federalism, 10(2), 152-181.

Benz, A. and Sonnicksen, J. (2017). Patterns of federal democracy: tensions, friction, or balance between two government dimensions. European Political Science Review, 9(1), 3-25. https://doi.org/10.1017/S1755773915000259

Brenton, S. (2022). Does federalism enhance representative democracy? Perpetual reform and shifting power in a divided Belgium. Journal of Contemporary European Studies, 31(2), 500-518. https://doi.org/10.1080/14782804.2021.2023482

Caluwaerts, D. (2012). Confrontation and communication: Deliberative democracy in divided Belgium. Peter Lang Publishers.

Caluwaerts, D. and Deschouwer, K. (2014). Building bridges across political divides: Experiments on deliberative democracy in deeply divided Belgium. European Political Science Review, 6(3), 427-450. https://doi.org/10.1017/S1755773913000179

Caluwaerts, D., Jacquet, V., Minsart, E., Junius, N. and Acar, L. (Forthcoming). Democratic innovation in Belgium. In M. Reuchamps, M. Brans, P. Meier and E. van Haute (eds.), The Oxford Handbook on Belgian Politics. Oxford University Press.

Caluwaerts, D. and Reuchamps, M. (2014). Combining federalism with consociationalism. Is Belgian consociational federalism digging its own grave? Ethnopolitics, 14(3), 277-295. https://doi.org/10.1080/17449057.2014.986866

Dandoy, R. (2013). Belgium: toward a regionalization of national elections? In R. Dandoy and A. H. Schakel (eds.), Regional and national elections in Western Europe. Territoriality of the vote in thirteen countries (pp. 47-67). Palgrave Macmillan.

Dandoy, R., Dodeigne, J., Reuchamps, M. and Vandeleene, A. (2015). The New Belgian Senate. A (Dis)Continued Evolution of Federalism in Belgium? Representation, 51(3), 327-339. https://doi.org/10.1080/00344893.2015.1108358

Dandoy, R. and Schakel, A. H. (2013). Introduction: territoriality of the vote: a framework for analysis. In R. Dandoy and A. H. Schakel (eds.), Regional and national elections in Western Europe. Territoriality of the vote in thirteen countries (pp. 1-26). Palgrave Macmillan.

Deschouwer, K. (2012). The Politics of Belgium. Governing a Divided Society. Palgrave Macmillan.

Deschouwer, K. and Van Parijs, P. (eds.) (2009). Electoral Engineering for a Stalled Federation [Re-Bel e-book 4]. Re-bel initiative. https://rethinkingbelgium.eu/ebook/electoral-engineering-for-a-stalled-federation/

Doria, G. (2006). The paradox of federal bicameralism. European Diversity and Autonomy Papers, (5), 5-43. https://hdl.handle.net/10863/7391

Elazar, D. J. (1987). Exploring federalism. University Alabama Press.

Fenna, A. and Schnabel, J. (2024). What is Federalism? Some Definitional Clarification. Publius: The Journal of Federalism, 54(2), 179-200. https://doi.org/10.1093/publius/pjad034

Fossum, J. E. and Jachtenfuchs, M. (2017). Federal challenges and challenges to federalism. Insights from the EU and federal states. Journal of European Public Policy, 24(4), 467-485. https://doi.org/10.1080/13501763.2016.1273965

Frateur, J., Bursens, P. and Meier, P. (2023). Exploring the Relationship between Federal Reform and the Representation of the People: The Case of the Belgian House of Representatives. Publius: the Journal of Federalism, 53(4), 618-641. https://doi.org/10.1093/publius/pjad006

Hooghe, L. (2004). Belgium: hollowing the center. In U. Amoretti and N. Bermeo (eds.), Does federalism matter? Political institutions and the management of territorial cleavages (pp. 55-92). Johns Hopkins Press.

Junius, N. (2023). Legitimacy in the participants’ eyes: A call for participants’ authorship over mini-public design in Brussels. Critical Policy Studies, 17(4), 524-543. https://doi.org/10.1080/19460171.2022.2157297

Junius, N. (2024). From representative to represented mini-publics: How mini-publics’ outputs are shaped by representation. Political Studies Review, 22(4), 914-929. https://doi.org/10.1177/14789299231217619

Lacey, J. (2017). Centripetal democracy. Democratic legitimacy and political identity in Belgium, Switzerland and the European Union. Oxford University Press.

Meier, P. and Bursens, P. (2021). Belgium: The Democratic State of the Federation. In A. Benz and J. Sonnicksen (eds.), Federal democracies at work. Varieties of complex government (pp. 180-196). University of Toronto Press.

Mueller, S. (2024a). Shared Rule in Federal Theory and Practice. Oxford University Press. https://doi.org/10.1093/9780191991288.001.0001

Mueller, S. (2024b). Bicameralism. In A. Vatter and R. Freiburghaus (eds.), Handbook of Comparative Political Institutions (pp. 263-275), Edward Elgar Publishing.

Mueller, S., Vatter, A. and Dick, S. (2021). A new index of bicameralism: taking legitimacy seriously. The Journal of Legislative Studies, 29(2), 312-336. https://doi.org/10.1080/13572334.2021.1996753

Niedobitek, M. (2018). The German Bundesrat and executive federalism. Perspectives on Federalism, 10(2), 198-214.

Niessen, C. and Popelier, P. (Forthcoming). Belgian Federalism. In M. Reuchamps, M. Brans, P. Meier and E. van Haute (eds.), The Oxford Handbook on Belgian Politics. Oxford University Press.

Niessen, C. and Reuchamps, M. (2022). Institutionalising citizen deliberation in parliament: The Permanent Citizens’ Dialogue in the German-speaking Community of Belgium. Parliamentary Affairs, 75(1), 135-153. https://doi.org/10.1093/pa/gsaa056

Palermo, F. (2018). Beyond second chambers: alternative representation of territorial interests and their reasons. Perspectives on Federalism, 10(2), 49-70. https://doi.org/10.2478/pof-2018-0016

Passaglia, P. (2018). Unicameralism, bicameralism, multicameralism: evolution and trends in Europe. Perspectives on Federalism, 10(2), 1-29. http://dx.doi.org/10.2478/pof-2018-0014

Pilet, J.-B. and Meier, P. (2018). Ze halen hun slag wel thuis: Over particratie en het aanpassingsvermogen van Belgische partijen. Res Publica, 60(4), 321-345. https://openjournals.ugent.be/rp/article/id/74682/

Popelier, P. (2018). Bicameralism in Belgium: the dismantlement of the Senate for the sake of multinational confederalism. Perspectives on Federalism, 10(2), 215-237.

Popelier, P. (2021). Dynamic federalism. A new theory for cohesion and regional autonomy. Routledge.

Popelier, P. and Lemmens, K. (2015). The Constitution of Belgium: A Contextual Analysis. Bloomsbury Publishing.

Rummens, S. and Sottiaux, S. (2014). Democratic legitimacy in the Bund or the “federation of states”. The cases of Belgium and the EU. European Law Journal, 20(4), 568-587. https://doi.org/10.1111/eulj.12053

Russell, M. (2001). The territorial role of second chambers. The Journal of Legislative Studies, 7(1), 105-118. https://doi.org/10.1080/714003852

Saward, M. (2010). The representative claim. Oxford University Press.

Scherz, A. (2018). Representation in multilateral democracy: How to represent individuals in the EU while guaranteeing the mutual recognition of peoples. European Law Journal, 23(6), 495-508. https://doi.org/10.1111/eulj.12267

Sonnicksen, J. (2018). Federalism and democracy: a tense relationship. In J. Tudella, M. Kölling and F. Reviriego (coords.), Calidad democrática y organización territorial (pp. 31-52). Marcial Pons.

Stojanovic, N. (2021). Multilingual democracy: Switzerland and beyond. Rowman and Littlefield.

Swenden, W. (2010). Subnational participation in national decisions: The role of second chambers. In H. Enderlein, S. Wälti and M. Zürn (eds.), Handbook on Multi-level Governance (pp. 103-123). Edward Elgar.

Swenden, W. (2013). Conclusion: The future of Belgian federalism – between reform and swansong? Regional and Federal Studies, 23(3), 369-382. https://doi.org/10.1080/13597566.2013.773897

Thorlakson, L. (2020). Multi-level Democracy: Integration and Independence Among Party Systems, Parties, and Voters in Seven Federal Systems. Oxford Academic.

Tuschhoff, C. (1999). The compounding effect: The impact of federalism on the concept of representation. West European Politics, 22(2), 16-33. https://doi.org/10.1080/01402389908425300

Urbinati, N. (2006). Representative Democracy. Principles and Genealogy. University of Chicago Press.

Van Haute, E. and Deschouwer, K. (2018). Federal reform and the quality of representation in Belgium. West European politics, 41(3), 683-702. https://doi.org/10.1080/01402382.2017.1399320

Van Parijs, P. (2000). Power-Sharing versus Border-Crossing in ethnically divided societies. In I. Shapiro and S. Macedo (eds.), Designing Democratic Institutions (pp. 296-320). NYU Press.

Vanrenterghem, A. (2022). Senaatsvoorzitter Stephanie D’Hose (Open VLD) wil Senaat afschaffen. VRTNWS. https://www.vrt.be/vrtnws/nl/2022/01/08/senaatsvoorzitter-wil-senaat-afschaffen/

Young, I. M. (2002). Inclusion and Democracy. Oxford Academic.

---

1. 1 https://www.belgium.be/en/about_belgium/government/federal_authorities/federal_government/policy/government_agreement

2. 2 To further complicate matters, the Belgian Senate is also meant to represent the three Belgian Regions (Brussels, Wallonia, and Flanders). Indeed, the Belgian federation contains a double level of constituent units, and both are supposed to be represented in the Senate. While the Senators are appointed by and among the MPs of the parliaments of the Communities and the Regions, they sit in the Senate as representatives of their language group (the Communities), not of a territory (the Regions). This is less of an issue for Flanders, as the institutions of the community and the region are merged, and the Dutch-speaking Senators are to a large extent ­appointed from that single parliament. Similarly, a part of the French-speaking Senators is appointed from the Walloon Parliament and might ­represent the ­interests of their region. However, the situation is different when it comes to the Senators appointed from Brussels. Officially a ­bilingual region, they are chosen along the lines of language groups, but the presence of MPs from both language groups of the Brussels ­parliament, is no guarantee of a representation of the Brussels region (see also Brenton, 2022Brenton, S. (2022). Does federalism enhance representative democracy? Perpetual reform and shifting power in a divided Belgium. Journal of Contemporary European Studies, 31(2), 500-518. https://doi.org/10.1080/14782804.2021.2023482). This second dimension of the representative function of the Senate is blurred in the actual functioning of the Senate and overall tends to be forgotten.