The phrase “composition of the European Parliament” is the legal term of the Lisbon Treaty to signify the allocation of the seats of the European Parliament (EP) between the member states of the Union (Art. 14 (2) TEU). Decisions in this regard are taken unanimously by the European Council on a proposal from the EP. So far, they have always been taken on an ad hoc basis for each upcoming electoral period, and there is no commonly accepted procedure that would allow to directly calculate the number of seats of a member state from its population. The privilege to initiate a decision for adopting such a permanent allocation procedure lies with the EP. Up to now, Parliament has taken no decisive action, despite debating the topic for decades. At least an overall agreement has emerged that the procedure sought must be objective, fair, durable, and transparent.
In February 2024, yet another workshop on a permanent system for the allocation of seats in the EP took place during an ordinary meeting of the Committee on Constitutional Affairs (AFCO). The speakers, Friedrich Pukelsheim, Victoriano Ramírez González and Manuel Müller, proposed three allocation methods: Power Compromise, FPS Technique, and Proportional Completion Through Transnational Lists. In this note, I briefly review the proposals and compare the resulting compositions with the ad hoc composition for the upcoming 2024–2019 Parliament.
Degressive representation
Any allocation procedure establishing the composition of the EP must be respecting and abiding by the rules which govern the Union. Primary and secondary Union law stipulates the criteria to be satisfied. The politically most sensitive and procedurally most challenging criterion, in Art. 14 (2) TEU, demands that
Representation of citizens shall be degressively proportional.
An EP resolution of 2007 interprets degressive proportionality to be an expression of the principle of solidarity whereby
the more populous States agree to be under-represented in order to allow the less populous States to be represented better.
This interpretation renders the term “degressive representation” more meaningful than the paradoxical notion in the Lisbon Treaty, “degressive proportionality”. To simplify wording, I shall speak of degressivity.
Degressivity requires an MEP from a more populous member state to represent more citizens than each MEP from a less populous member state. The concept is turned into a workable definition when decreeing the composition of the 2024–2029 EP: The representation ratio of a more populous member state is to be larger than the representation ratio of a less populous member state, where the representation ratio of a state is the ratio between the state’s population figure and its number of seats before rounding.
The other relevant criteria are straightforward. The states’ contingents of seats range between six and ninety-six seats (Art. 14 (2) TEU). The minimum of six seats is fully utilized for the smallest state. The capping of ninety-six seats is invoked only when necessary, such as at present; should the Union grow further, no state might qualify for ninety-six seats anyway. Finally, increase of population figures and seat contingents has to be concordant, i.e., the larger the population of a member state, the greater its entitlement to a large number of seats.
Population figures
Essential input data for any allocation method are the member states’ population sizes. The question of whom to count, or not to count, is subtle and delicate. However, the European Union has answered the question for its business when setting up the Qualified Majority Voting rule for its Council, which also requires an official definition of member states’ population.
These QMV-population figures are published every December for the subsequent calendar year. Since Council and Parliament are organs with joint governance responsibility, these figures suggest themselves as the input base also for the EP composition. Our sample applications below rely on the QMV-populations for 2023 as they were available when the 2024–2029 composition was decreed.
In case Parliament prefers another data type for the determination of its composition, care must be taken to ensure proper documentation. Current decisions shun any population figures whatsoever, whence neither degressivity nor concordance are verifiable. A passing remark that the population data are taken from EuroStat is insufficient and too vague as this source is in constant flux.
Degressivity is implemented by adjusting raw population figures in a way affecting larger states a bit more and smaller states somewhat less. The degressivity adjustments of the three proposals are fundamentally distinct. The Power Compromise transforms raw population figures into adjusted population units. The FPS Technique converts raw population figures into adjusted seat quotas. Proportional Completion substitutes any population figure by its square root.
Power Compromise
The Power Compromise pays tribute to the representation of member states’ citizenries as a whole as well as to the representation of Union citizens as individuals. Collective citizenries are accounted for in that every member state is allocated the same number of seats, called base seats. For instance, with four base seats for each of 27 member states, 108 seats are committed. With an EP size of 720 seats, 612 seats are remaining for further allocation.
The remaining seats are meted out taking into account the size of the citizenries, thus honouring individual citizens. Due to degressivity, raw population figures are reduced to adjusted population units. The adjustment is performed by raising all population figures to a certain power, the power parameter being calculated from the input data. The remaining seats are apportioned proportionally to adjusted population units using the divisor method with upward rounding. The presence of a power operation suggests the name Power Compromise.
To summarize, the Power Compromise has three parameters that jointly govern the system. The number of base seats is set so as to ensure every state a minimum allocation of six seats. The power parameter is calculated from the data at hand so that the largest member state is allocated just 96 seats. The divisor regulates the apportionment of the remaining seats in a way that the seat contingents of all states together exhaust the given EP size.
As an example, when applied to the 2024–2029 EP, the Power Compromise would allocate the 720 seats by way of rather transparent instructions: Every member state is assigned four base seats, plus one seat per 28,321 adjusted population units or part thereof, where the adjusted population units are obtained by raising the 2023 QMV-population figures to the power 0.8095.
The Power Compromise is objective and durable. It permits control and variation of population figures, the number of member states, and the EP seat total. Moreover, it features some built-in dynamics towards enhanced simplicity and fairness. It may happen, in a growing Union, that the largest state falls short of 96 seats and capping is evaded. The power parameter would then equal 1, whence raw population figures and adjusted population units would plainly coincide.
FPS Technique
The FPS Technique uses a different type of degressivity adjustment by generating what I term adjusted seat quotas. A state’s adjusted seat quota is a fractional quantity close to the state’s eventual seat contingent. The adjusted seat quotas are taken to be a mixture of three parts, fixed part (F), proportional part (P), and square-root part (S). Hence the acronym FPS.
The fixed part splits the seat total equally between all member states. The proportional part subdivides the seat total proportionally to population figures, while the square-root part partitions the seat total in proportion of the populations’ square roots. The parts’ weightings must be preordained in advance. Author Victoriano Ramírez, by way of a comparison with former compositions, recommends a mixture of ten per cent fixed part, fifty per cent proportional part, and the remaining percentage (forty per cent) square-root part. I denote this mixture by F10P50S.
Finally, the available seats are apportioned proportionally to the adjusted seat quotas using the divisor method with standard rounding, observing the restrictions that a state’s contingent is at least six seats and at most ninety-six seats.
As an example, when applied to the 2024–2029 EP, the F10P50S Technique would allocate the 720 seats by way of the following instructions: The seat contingent of a member state is the whole number nearest to the quotient of its adjusted seat quota and the divisor 0.9955, except when an incrementation to six seats or a decrementation to ninety-six seats is called for. The adjusted seat quota of a member state with 2023 QMV-population pop is given by
0.1 · |
1 |
· 720 + 0.5 · |
pop |
· 720 + 0.4 · |
√pop |
· 720 |
27 |
447,533,143 |
91,209 |
The divisor 0.9955 is determined so that the seat contingents of all states sum up to the given EP size of 720 seats. The denominator 447,533,143 corresponds to the population of the entire EU, while 91,209 is the sum of the square roots of the population figures of all member states.
The FPS Technique permits control and variation of population figures, the number of member states, and the EP seat total, as does the Power Compromise.
However, the FPS percentages ten, fifty, and forty are static and insensitive to the data. They invite dispute. The groups of smaller, larger, or middle-sized states would benefit from raising the percentages of the fixed, proportional, or square-root parts, respectively. Since the percentage total is always one hundred and constant, one group’s joy is another group’s sorrow.
Moreover, the final stage should employ the divisor method with upward rounding rather than the divisor method with standard rounding. Upward rounding favours smaller states at the expense of larger states, which is what degressivity is all about.
Proportional Completion through transnational lists
The Proportional Completion system tackles EP elections in their entirety, not just the EP composition. Aiming at electoral equality among all Union citizens, the system strives for a higher level of equality than is currently achieved by separate elections in 27 member states. To this end, 75 seats, of the maximum EP size of 751 seats, are set aside to be filled via transnational lists of European political parties. These 75 seats have an overall European obligation whence they are exempt from being attributed to any particular member state.
This approach leaves 676 seats to be distributed between the member states. Future reinforcement of the European dimension by transnational seats mitigates domestic ambitions in the seat allocation. The author Manuel Müller proposes to allocate the seats proportionally to the populations’ square roots, using the divisor method with standard rounding. The largest member state would receive 68 seats, thereby eluding the 96 capping. The minimum guarantee of six seats remains relevant and would apply to the smallest member state.
The complexity of the Proportional Completion system reaches beyond the remit of the workshop. Narrowed down to the issues concerning the EP composition, sole reference to populations’ square roots entails a massive transfer of seats from the six largest member states to the other smaller states. An intervention at this scale would need to be embedded into a more extensive redesign of the European Electoral Act.
Only few changes from the ad hoc composition 2024–2029
How do the proposed methods compare to the ad hoc composition for the 2024–2029 EP? The Power Compromise yields a composition differing by a transfer of no more than seven seats altogether. All transfers touch upon a single seat, except for a sole instance that involves two seats.
The F10P50S Technique also transfers just seven seats provided the final stage uses the divisor method with upward rounding. Six transfers are identical to those required by the Power Compromise; the two-seat instance is the same. In other words, these two methods differ in their procedural instructions, but practically coincide in their results (using the 2023 QMV-population figures).
When the final stage of the F10P50S Technique applies the divisor method with standard rounding, as suggested by the author Victoriano Ramírez, a total of eight seats need to be transferred. One instance involves the transfer of two seats, another, three seats. The states affected are partly distinct from the previous two methods.
Time to finish the project
Prior to the workshop, AFCO rapporteurs Ana Collado Jiménez and Niklas Nienaß had issued a Draft Report noncommittal as to the choice of the allocation method. Since the current legislative period is drawing to a close, it seems too late to agree on a definitive decision.
Yet the final AFCO report may set the scene for the next Parliament to finish the project. As Andrew Duff emphasized during the workshop and in a subsequent Verfassungsblog article, further progress, such as the installation of transnational lists, may be addressed in a future constitutional convention.
Friedrich Pukelsheim is a professor emeritus at the Institute for Mathematics of the University of Augsburg. |
If citizen votes don't count the same it is not fair elections, it is not equality before the law. The European Parliament voting scheme is a scam.
AntwortenLöschen