dynamic federalism

Dynamic Federalism

Dynamic Federalism


Traditional federal theory seems no longer apt to grasp recent evolutions in state structures. By delimiting federal states in terms of defining institutional features, federalism scholars put themselves at the margin rather than the centre of where the action is: fragmenting dynamics in multinational states, secession movements, as well as centralist and decentralist tendencies within the European Union. In a dynamic approach to federalism, all multi-tiered systems are assembled with a common denominator being how they manage tensions between autonomy claims of territorial entities on the one hand, and the need for cohesion or efficiency of the central government on the other. In this approach, qualifying criteria to categorise state structures become mere indicators to rank multi-tiered systems on a gliding scale from the most central to the loosest systems. The ranking is based on three sets of indicators, one measuring autonomy, another measuring cohesion and a third, linking both, measuring participation. The core question examined in this contribution is: which mechanisms in the constitutional system have a centralising or decentralising effect?



It is an exciting era for scholars in federalism studies, with a succession of events that keep the world captive, from secession referendums in Scotland and Catalonia, the tiny Walloon Region in Belgium holding some 780 million persons hostage by vetoing a trade agreement between the EU and Canada, to the most pressing concern in the European Union: Brexit. That is – if the UK, Spain, Belgium and the EU are defined as federalist systems, worthy of study by federalism scholars. Books have been written about this question. Traditional federalists would argue that they are not, measured by criteria that have to be fulfilled and that are based on model states of past centuries, such as the USA or Germany. According to the traditional model, states are either unitary, federal or confederal, and systems that do not meet the defining criteria are called ‘regional states’, which, at best, are considered ‘immature federations’, implying that they aspire to become federal. However, excluding these systems leaves federalism scholars at the margin rather than the centre of activity.

At the same time, scholars realise that the federal/unitary distinction ‘is too crude to capture the complexity of contemporary governance’ (Loughlin, 2008: 473). In a new standard book on comparative federalism (Palermo and Kössler 2017), the authors list various definitions of federalism, one leading to a list of 23 federal states world-wide, the other to no less than 180 federal states. ‘The central question’, according to these authors, ‘is whether the question itself is meaningful’. I tend to agree. Whether Spain is a federal or a regional state is, frankly, a purely academic question, such as discussing the sex of angels. In the end, what matters, is to capture the essence of federal systems. The essence is not whether a given state has a bicameral system and a court to solve allocation of power disputes, or whether sub-units have their own constitution. What is at stake, was already defined by Friedrich (1968): the tension between autonomy of territorial entities on the one hand, and cohesion or efficiency of the central government on the other.

This brings us to a dynamic approach to federalism. In such an approach, qualifying criteria to categorise state structures become mere indicators, and the core question is: which mechanisms in the constitutional system have a centralising or decentralising effect?

The Use of Indicators to Define Federal Systems

In a dynamic approach, political systems are situated on a gliding scale, with unitary systems on the left side of the spectrum, and the loosest cooperative associations on the right. Systems are placed on this scale as soon as there is some tension between the central authorities and territorial sub-groups. While we could label these systems ‘federal’ in a broad sense, a more distinctive name that avoids confusion with ‘traditional’ federal systems, is ‘multi-tiered systems’, or simply MTS.

Political systems are situated on this gliding scale on the basis of a general score that relies on three axes: one measuring the autonomy of sub-national units, a second measuring the cohesiveness of the entire system, and a linking third focusing on participation. Political scientists have a longer tradition of scaling political systems on the basis of indicators. However, they are mostly interested in autonomy and they tend to ignore the role of courts (see, e.g., Hooghe et al. 2016). Indicative for the autonomy of sub-units is, amongst others, the entrenchment of subnational entities and competences in rigid acts, subnational representative bodies, financial autonomy, sets of competences and allocations techniques, and whether the entities are (directly or indirectly) involved in decision-making at the EU or the international level. In the literature, much less attention has been given to a second set of indicators that measure cohesion or integration. Indicators are, amongst others, free movement and a monetary and economic union within the legal system, mechanisms to deal with transboundary problems, instruments to prevent or solve conflicts of competences and conflicts of interests, or to prevent subnational entities from undermining central (national of international) policy. The third set of indicators is focused on mechanisms that ensure both autonomy and cohesion, with the subnational entities participating at the central level to ensure central legislative, administrative as well as judicial decision-making while paying attention to subnational specificities.

On this large scale of MTS, we can identify core ‘federal systems’. Federal systems in this narrow sense, find a balance between autonomy/differentiation and cohesion/integration. These systems will score moderate to high on all three axes. On the left side of the scale are those political systems that solve the tension by accentuating centralism. Such systems will score low on the autonomy axis but high on the cohesion axis. On the right side of the scale are those political systems that solve the tension by accentuating autonomy: more effort is made to preserving the autonomy of the subnational units rather than integrative mechanisms.

Within the sub-set of autonomy indicators, a political system can score low on one indicator and high on the other. While under the traditional approach the system should meet certain institutional requirements to be qualified as a federal system, under the dynamic approach other features can compensate for this.

While we preserve labels such as unitary states, decentralised states, regional systems, federal systems and confederations, a neatly cut categorisation of states is not always possible, and not even necessary. Spain, for example, would probably end up somewhere between regional states and federal states, but the exact label is not really important. Moreover, a categorisation of states is just a snap-shot. States evolve. Nowadays, in the centre of our attention are disintegrative states that move from the left side of the scale to the right. However, traditional integrative federal systems have their own, centralising, dynamics, bringing them from the right side of the scale to the left side, and sometimes leaving some doubt as to whether, in the end, highly centralised systems such as Austria can still be called ‘federal’ if defined as a system that upholds an equal balance between autonomy and cohesion.


Methodological Advantages

The methodological advantages of a dynamic approach are manifold.

First of all, it facilitates comparative research since we can easily group MTS with similar scores on one or more of the three axes. If we are mainly interested in the dynamics of state structures, we have a larger population than the twenty-something pure, traditional federal systems. We can look for factors that explain the position of specific types of states on the gliding scale. In addition, we can test whether the level of integration or disintegration is an explanatory factor for other things – for example the behaviour of courts in federal disputes, or the stability of the political system.

Secondly, we can examine phenomena that are not easily captured under traditional federal theory. An example is asymmetry. While traditional federal theory promotes symmetry for the sake of equality and stability, asymmetry is a growing trend in contemporary MTS. In a dynamic approach, we can measure the different sub-national entities on the autonomy-axis, so that we can give an exact score to the differences in status and competences between the most and the least autonomies subnational entities. This might instruct as to how asymmetric a system can be before it risks becoming unstable.

Thirdly, we can examine the impact of the international level on the relations between central authority and subnational units. This is especially important for MTS that are part of the European Union, as the impact of the European integration process upon the constitutional structure of the member states is more intense than that of any other supra-national organisation.


The Core Question: Which Mechanisms Have a Centralising or Decentralising Impact?

As mentioned, the core question in a dynamic approach to federalism concerns the process of integration or disintegration. In this approach, we examine the mechanisms that have a centralising or decentralising effect on the political system. For example, there is a common agreement that courts generally have a centralising effect when deciding on federalism disputes; yet some courts – mostly in multinational systems – take a more balanced approach (Popelier 2017). Political parties can have such effects as well. For instance, in Belgium, the break-up of national parties into regional parties had a disintegrative effect. Techniques to allocate powers may also have some impact: we can hypothesise that the predominance of concurrent powers has a centralising effect, whereas the predominance of exclusive powers has a decentralising effect.

If we have more insight into the conditions under which these mechanisms have a centralising or decentralising effect, we might be able to answer the question whether the dynamics of a specific political system can be turned through constitutional engineering. This is a topical question in the light of secessionist movements in countries such as Belgium, Canada, Spain and the UK as well as developments in, for example, Sri Lanka, where a devolutionary trend institutionalised at the end of the 1980s a form of multinational conflict management (Oberst 1988) but constitutional guarantees of national sovereignty and indivisibility are relied upon to break secessionist tendencies.


An Example: The Belgian Case

The Belgian case demonstrates the need for a dynamic approach to federalism. In the last five decades, it evolved from a unitary state into a federal state with confederal traits. Up until now, this has taken the shape of six state reforms, resulting, in 1993 in a constitutional provision that labelled Belgium as a federal state.

The Belgian federation, however, does not meet several of the criteria that defines federal states under the Hamilton approach. For example, in 1993, discussions on the use of a second chamber did not result in the abolition of the Senate. Instead the federalism argument was used to maintain an institution that did not fulfil a federalist function: The Senate was reformed, but in the new constellation a minority represented the federated entities and they were appointed on the basis of the federal, not the regional, elections. Since the sixth state reform in 2012-2013, the Senate has been reformed into an actual Chamber of the sub-states, but is left with only few competences. On the other hand, the subnational entities can directly interfere, with a suspensory vote, in the federal decision-making process. Moreover, while not represented as such, they have a dominant say through the linguistic groups that structure the federal parliament, the federal government, as well as the administration and the courts. The federal government consists of an equal number of French- and Dutch speaking ministers; these ministers are nominated by regional parties that represent the interests of their language groups. Besides, the two language groups in Parliament have a suspensory veto right.

The federalism argument was also used in the Flemish fight for subnational constitutional autonomy, although the (little) constitutional autonomy that was acquired has not been used in a way that is substantially different from what was regulated at the federal level. On the other hand, the subnational entities do enjoy substantial autonomy when it comes to concluding international agreements or involvement at the EU level.

This demonstrates that the traditional criteria to qualify federations are not always functional: The Senate did not fulfil the function of involving the federated entities in central decision-making, but they got their say through other means, to the point that the system even acquired confederal traits, as no federal decision can be made without the consent of both language groups. It also shows how a low score on one indicator (subnational constitutional autonomy) can be compensated for by a high score for another indicator (international affairs).

The devolving dynamics in Belgium are based upon the allocation of exclusive powers, equality of federal and federated entities, regionalised political parties, and a general ambiance of distrust and conflict typical of dyadic federalism. The dominant political party, N-VA, is a Flemish-nationalist party that supports Flemish independence. According to surveys, Flemings support autonomy though not secessionism, the N-VA’s strategy is to ‘naturally’ end up at independence through confederalism. Legal scholars and political scientists are eager to point out that confederalism, according to traditional theory, means the association of independent, sovereign states, implying that technically confederalism cannot precede secession. This shows, once more, how traditional theory is unable to capture political reality. In a dynamic approach, the N-VA’s strategy makes perfect sense, and the core question for those who support the continuation of the Belgian system, is whether through constitutional engineering – be it federal districts, shared competences, or otherwise – we can turn the tide.



Friedrich, K. 1968. Trends of Federalism in Theory and Practice. Frederick A. Praeger.

Hooghe, L. et al. 2016. Measuring Regional Authority. Oxford: Oxford University Press.

Loughlin, 2008. ‘Federalism, regionalism and local government: comparative perspectives on transforming the nation-state’, 7 European Political Science, 472-482.

Oberst, R.C. 1988. ‘Federalism and Ethnic Conflict in Sri Lanka. 18 Publius 175-194.

Palermo, F. and Kössler, K. 2017. Comparative Federalism. Constitutional Arrangements and Case Law. Oxford and Portland: Hart.

Popelier, P. 2017. ‘Federalism disputes and the behavior of courts: explaining variation in federal courts support for centralization. 47 Publius, 27-48.


Further Reading

Aubert, J.F. 1963. ‘Essai sur le fédéralisme.’ 80 Revue du droit public et de la science politique. 401-452.

Popelier, P. 2015. ‘Secessionist and autonomy movements in Flanders: the disintegration of Belgium as the chronicle of a death foretold?’ in Belser, E.M. et al. (eds), States Falling Apart? Secessionist and Autonomy Movements in Europe. Bern: Stämpfli Verlag, 215-246.

Popelier, P. 2014. ‘Subnational multilevel constitutionalism.’ 6 Perspectives on Federalism, 1-23.

Popelier, P. 2012. ‘The need for sub-national constitutions in federal theory and practice.’ 4 Perspectives on Federalism 36-58.

Popelier, P. and Lemmens, K. 2015. The Constitution of Belgium. A Contextual Analysis. Oxford: Hart.



Posted by Patricia Popelier, 0 comments