In Search of a Methodical Approach to Seat Apportionment in the European Parliament

The problem

The European Parliament is once again trying to tackle the problem of how to apportion its seats between member states. In one of those rare Treaty instances, Parliament is obliged to initiate this procedure itself [Article 14(2) TEU]. It has so far failed in this obligation, and finding a decent solution still proves difficult. The closest to an informed debate within the Parliament came in 2011 when it commissioned a study that produced the classic ‘base plus proportional’ formula, named the Cambridge Compromise (CamCom), which amply demonstrated the severe deviance of Parliament’s then arbitrary seat allocation.¹⁾

In the absence of agreement on a method, the practice has been to continue to trade parliamentary seats, often in an unseemly fashion, as part and parcel of the EU’s larger institutional bargaining. This has led the European Council to exhort the Parliament more than once (and now again for 2026) to “propose an objective, fair, durable and transparent seat distribution method”.²⁾ The prolonged failure to regularise the system for the composition of the European Parliament is becoming more serious. The EU never fails to criticise lapses in electoral law or practice where third countries are concerned. It stands guilty of hypocrisy unless and until it can put its own House in order.

Article 14(2) says: “The European Council shall adopt by unanimity, on the initiative of the European Parliament and with its consent, the decision establishing the composition” of the House. Changes to the relative allocation of seats should happen “in good time” before the end of each five-year term of the Parliament in order to take account of the accession or secession of a member state as well as of the continuing demographic churn.³⁾

The specific formal criteria are well established in Article 14(2). Parliament must have no more than 751 seats. No state shall be allocated fewer than 6 seats or more than 96. Apportionment is to be “degressively proportional”, which means that an MEP from a more populous state represents more people than an MEP from a less populous state. Size is calculated according to the internationally recognised standard of total resident population, regardless of nationality and excluding refugees. In accordance with good federalist principles, the larger states agree to be under-represented to allow for smaller states to be better represented.

Some more informal conditions should also be taken into account. The Union’s legislature is bicameral, with the citizens’ chamber of the Parliament and the states’ chamber of the Council [Article 10(2) TEU]. As a general rule — although there are important exceptions — nothing can become EU law unless it has been passed by both Houses [Article 294 TFEU].⁴⁾ The to and fro between Parliament and Council, in the grip of co-decision, is the stuff of Brussels politics. It follows, therefore, that any change to the voting weights or system in one chamber may affect the same in the other. The principle of “mutual sincere cooperation” between the EU institutions would seem to be relevant here [Article 13(2) TEU].

The federal dimension

It helps to see the EU as a youthful, yet contested, federal polity. Democratic federal systems are an inevitable compromise between, on the one hand, the constitutional principle of the “equality of Member States before the Treaties” [Article 4(2) TEU], and, on the other hand, the democratic principle of one person one vote or “equality of its citizens” [Article 9 TEU]. Every national of a member state is an EU citizen with the right to vote and to stand as a candidate in elections to the European Parliament wherever they live in the Union [Article 20 TFEU; Article 39(1) CFR].

No electoral system is perfect. But any directly elected federal parliament must reflect that basic compromise between statehood and citizenship in a satisfactory manner. Europe’s federal system must be able to confound sceptical or even hostile challenges to the European Court of Justice by certain litigious national constitutional courts.⁵⁾ The EU’s emerging federal constitutional order relies on “sincere cooperation” between the Union and its member states [Article 4(3) TEU]. Finding an equitable balance of power, therefore, between member states of different types and sizes is key to the successful functioning of the democratic Union and the mutual reinforcement of its twin European and national identities.

All these general principles, taken together, should be given practical application in the design of a permanent system to apportion seats between the states.

The goal

The objective is to agree on a mathematical formula that, by meeting the various criteria, will serve to enhance the stability and legitimacy of the European Union as a whole, whose very functioning is “founded on representative democracy” [Article 10(1) TEU]. Ideally, the method chosen should be settled and therefore durable, resistant to constant revision. Its entrenchment in EU law by regulation, allowing for no deviance, should further legitimate the European Parliament and quieten its critics who question its reputation.

Although undoubtedly complex for the non-mathematician, the chosen method should be sufficiently transparent to be at least explicable by the average MEP as well as comprehensible to the media. As we have implied, the method should be litigation proof, a de jure realisation of the de facto federal compromise on which the EU is based.

A deputy elected at the general election of the European Parliament must not be deprived of his or her mandate because of the mid-term accession of a new member state. The House can be permitted to inflate temporarily above 751 seats until the subsequent election. In normal circumstances, it should be presumed that all 751 seats will be filled. With a decent formula in place, it will be unnecessary as well as misleading to leave vacant an arbitrary number of seats to cater for some putative future accession state. Expanding the size of the House on a permanent basis to accommodate enlargement would require treaty change. But there is no appetite for this among sitting MEPs. Already, speaking time for backbench deputies in plenary is heavily constricted: there is not much worth saying in a speech of less than 90 seconds.

Despite the artificial cap of 96 seats, discrimination against German interests should be as far as possible minimised until such time as future enlargement of the Union reduces Germany’s allocation — as it surely will when Ukraine joins (and/or the UK comes back).

Any apportionment method should be easily adaptable to the future introduction of transnational party lists from which a portion of MEPs will be elected for a pan-European constituency. Transnational lists will Europeanise the parliamentary elections and force into being proper federal political parties without which the development of EU representative democracy is handicapped. This is the official policy of the European Parliament.⁶⁾

The method deployed to reapportion seats every five years should aim to minimise disruption to the House. It should not provoke disputes, especially of an intractable nature between small, medium and large states claiming unfairness. The stability of the institution of Parliament is highly desirable, enabling it always to sustain the code of interinstitutional balance with the Council.

The options

With these stipulations in mind, Parliament’s Constitutional Affairs Committee (AFCO) organised a workshop (14 February) to consider three alternative formulae, all of which respect the principle of degressive proportionality.⁷⁾ (None touch the question of bicameral balance with the Council.) The first option is a Power variant of CamCom by Friedrich Pukelsheim and Geoffrey Grimmett. This well-rehearsed proposal — involving 4 base seats, plus power parameter and divisor — meets all the criteria discussed above and is future proof. Only 7 seats would need to be transferred out of the 751 compared to the official 2024-29 composition.

The second option is the elegant Fixed-Proportional-Square Root system (FPS) of Victoriano Ramirez Gonzalez. Only 8 seats would need to be switched from the current composition in a House of 720 MEPs. However, because the proposal allows for flexible interpretation, its adoption would risk constant battles between the interests of small states wanting a greater number of fixed seats, large states putting greater salience on proportionality, and middle-sized states prioritising the square root factor. And FPS does not provide for a fixed House of 751 MEPs.

The third alternative by Manuel Müller suggests one possible way to introduce a transnational element in the House. He would add extra ‘party’ seats on top of a square root proportional allocation for the larger portion of seats. But his system would count the votes for national parties within member states and accord them indirectly to EU parties. He would not give citizens a second ballot to vote directly for truly federal political parties.

Parliament now realises that the installation of transnational lists is unlikely to happen without treaty change, suitably crafted in a constitutional Convention [Article 48(3) TEU].⁸⁾ Parliament is also seeking to adjust the voting system in the Council, changing the threshold of QMV from the Lisbon formula of 55% of states representing 65% of the population to 60% of the states representing 60% of the population.⁹⁾ Another, and possibly better, approach would be to use the square root formula for Council QMV votes. The magic of the square root reduces the differentials between states of different sizes, thereby introducing an element of degressive proportionality into the Union’s second chamber. Respecting interinstitutional balance, the introduction of the Power variant of CamCom will reduce the clout of middling-sized states in the Parliament, while the square root voting system will advance their interest in the Council. On such exquisite deals are federal unions well built.

In the recent Brussels workshop, MEPs — missing the point — badgered the scientists to negotiate between themselves one single proposal, thereby relieving Parliament from having to take a political decision between alternative mathematical methods. In practice, therefore, a constitutional Convention is likely to be needed to settle all these matters in an open, comprehensive and consensual way. If AFCO manages in the meanwhile to produce an accurate and purposeful report, the European Parliament will have made an important contribution to improving its own legitimacy — and in preparing the Union for democratic enlargement.¹⁰⁾