I H S — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — 1

Draft Constitution: The Double Implies a Massive Transfer of Power to the Large Member States – Is this Intended?

Institute for Advanced Studies, Vienna, June 2003

Bernhard Felderer (Director), Iain Paterson (Senior researcher), Peter Silárszky (Researcher, Prague)

In Brief In the wide-ranging debate surrounding the May publication of the Draft Constitution there has been surprisingly little comment on the proposal to change the system in the European Council from the system adopted in Nice based on weighted votes to a new system of double majority. This is surprising, first, because the issue of Council decision-making rules occupied the key position of the intergovernmental politics in Nice. The importance of the qualified majority system derives from it being widely recognised as being the main source of member states’ power in the key EU institution. The main surprise however, may actually come to some interested parties more as a shock, for a) the actual distribution of power implied by the proposed system is not immediately apparent; and b) the proposed system implicitly represents a shift of power away from smaller countries and towards the larger member states far in excess of the agreement reached at Nice.

The constitutional text could be amended to enforce the conditonal requirement of a majority of states and 60% of the poulation as the ‘twin sources of legitimacy’ on any system of qualified majority, without restricting future options by enforcing the rigid outcome of the double majority as presently defined.

Text of Draft Constitution and its Interpretation Article I-24 Qualified majority states: “When the European Council or the Council of Ministers take decisions by qualified majority, such a majority shall consist of the majority of Member States, representing at least three fifths of the population of the Union.”

As we will see, the requirement for 60% of population is a significant departure from the previous proposal of a double simple majority of more than 50% of both member states and population (proposed by the Commission during the IGC preceding Nice). 2 — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — I H S

We interpret this text, in agreement with, for example, contribution (4) of the Centre for European Policy Studies (CEPS) to this forum, that the proposed double majority is exhaustive, i.e., a necessary and sufficient requirement for a qualified majority, not merely a necessary conditional requirement. This distinction is crucial: the qualified majority system agreed in Nice guarantees explicitly that these conditions are met – in fact, any qualified vote will, perforce, mean that it is carried by at least 62% of the population of the enlarged EU and a majority of its states.

The Distribution of Power after Nice There is a long tradition of deriving the a priori constitutional power distribution implied by a voting procedure by calculating the power indices. In the literature there are two methods predominating, that of Shapley and Shubik (SH), which we consider to be appropriate for the consensus-seeking EU Council, and that of Banzhaf and Coleman (BC). For qualified majority driven by weighted voting in Council, as the number of member states increases, both sets of indices are in close agreement. This is illustrated in Table 1, for the Nice weightings in the EU of 25.

[Tables 1 and 2 about here]

The tables show how the total power (100 percent) is distributed between the member states. In general, due to the weighting of votes, the amount of a member state’s power correlates with its population, but there is a systematic and intentional ‘overweighting’ in favour of the smaller member state. The achievement of Nice was to introduce weights that reduced the overweighting . This effect is measured by the overall power gradient of the voting system1. On a scale of 0 (equal power regardless of size of country) to 100% (power exactly proportional to share of total EU population) the power gradient of the Nice system is 57%. This represents an increase of 14 percentage points on the extension of the current voting system of the EU of 15 members for the EU of 25 members, which will be in force between May 1, 2004 and October 31, 2004 (c.f. Table 2). This is a moderately substantial redistribution of power2. There are two other salient features common to both the current and the Nice voting system for the Council:

• The power of a member state is well approximated by its share of the vote;

• We can be confident of the power as calculated because of the close agreement of competing theoretical approaches.

1 Paterson, 1998, Paterson and Silárszky, 1999 2 c.f. Paterson, 2000 I H S — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — 3

The last column of Table 1 shows that the distribution of power is only marginally different from a hypothetical situation, which uses the Nice weighted votes without the state majority and population conditions. This reflects the fact that the weights alone are sufficient, in nearly all cases, to ensure fulfilment of the other two requirements.

Claims made for the Double Majority are a Red Herring The CEPS contribution (4) to the Forum echoes many commentaries in welcoming the double majority proposed by the Draft Constitution, having previously recommended such a solution as having “the advantage of simplicity, transparency, and of taking full account of the twin sources of legitimacy in the EU, people and states.”3 Taking these issues one by one, it may be somewhat simpler to describe voting procedure of double majority than the Nice system. The latter is technically a triple majority, for, in addition to the voting weights, there are provisos written into the system for ensuring a simple majority of states and 62% of population. However, in most cases a qualified majority of votes alone will automatically ensure the fulfilment the other two criteria. The practical result of this is that member states’ representatives may first and foremost concentrate their efforts on achieving the weighted vote majority – a simpler task than juggling two .

Regarding transparency, the double majority of the draft Constitution most definitely does not deserve to be termed transparent. It is indeed opaque in its effects. For we ask: what exactly is the implication for the distribution of power between member states? No-one, not a Council minister, not an EU citizen, can give an immediate answer to this question. One needs to have a grounding in the calculation of power indices, and we will encounter an area of contested answers in so doing (below).

Both kinds of qualified majority systems, Nice and Draft Constitution, take the number of states and percentage of population, the ‘twin sources of legitimacy’ into account, as has been mentioned, and is thus not the exclusive advantage of either solution.

Such claims of the supposed advantage of the double majority, which have been heard from several quarters and repeated often in the media are, in fact, when it comes to the all-important question of power, spurious. We show now what the Draft Constitution proposal really means, and ask the questions “is this intended by the Convention and will it be an acceptable solution?”

3 CEPS contribution (3) to the Forum 4 — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — I H S

The Distribution of Power Implied by the Draft Constitution ‘Double Majority’ The results of our computer programs for calculating the power indices and the associated power gradient are shown for the EU of 25 member states in Table 3. Without resort to such methodology and technology there is no approximate indication of power analogous to the share of total weighted votes.

[Table 3 about here]

The major effect of the Draft Constitution is a redistribution of power to the 4 largest Member States, resulting in a very high power gradient of 82%. There are also some gains made by some of the smallest member states, although the assessment of this varies, depending on which method is used to calculate the power indices. It should be noted here that the large redistribution of power arises primarily because of the particular choice of quotas for the qualified majority, in particular the requirement for 60% of the EU population.4 The main qualitative difference of the proposed double majority is that the larger population of Germany is reflected in its voting power relative to the other big member states (up to 50 percent more power than other large member states).

There is some disagreement in the results depending on the power index used, although the conclusions to be drawn are more or less the same. This can be readily seen by comparing the power gradients in Table 4. Whereas the redistribution of power to larger countries, and hence the power gradient, is higher for the SH power indices than BC, in both cases the proposed double majority represents a very high degree of reform compared to the Status Quo (a gain of 39 and 25 percentage points respectively) and far in excess of the Nice agreement (an increase of 25 and 13 percentage points respectively).

[Table 4 about here]

In summary, the salient features of the Draft Constitution double majority voting system proposed for the Council are:

• There is no ready tangible measure of power of a member state analogous to the share of the vote;

• There is a degree of uncertainty about the actual distribution of power because of the disagreement of competing theoretical approaches;

4 In contrast to this, a double simple majority (majority of states and 50% of population) results in a lesser reform than the Nice arrangement (i.e. it has a lesser power gradient) I H S — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — 5

• There is a massive redistribution of power to the large member states, in particular to Germany.

Consequences for Designing the Qualified Majority System In view of the lack of transparency of how power is allocated by the Draft Constitution double majority voting system, and the surprising results when power is calculated, we feel obliged to pose the question to the members of the Convention and other interested parties: is this high degree of redistribution of power an intended consequence of this proposal?

Should the answer to this question be negative, the foregoing discussion shows clearly that there is an alternative way of preserving the requirements that intentionally lend legitimacy to decisions by ensuring that a qualified majority is backed by a majority of states and 60% of the population – namely by attaching these conditions to a system of weighted votes. A set of weighted votes may be suitably chosen, as in the Nice agreement, to effect a distribution of power that is close to the share of the total votes. Such a Constitutional text would treat the “double majority” as a necessary conditional requirement, not an exhaustive statement. A suitable reformulation of the text could read:

Article I-24 Qualified majority: “When the European Council or the Council of Ministers take decisions by qualified majority, using a schedule of votes as agreed intergovernmentally, such a majority shall also ensure a majority of Member States, representing at least three fifths of the population of the Union.” [changes in italics]

The question of whether the power of Germany should reflect its higher population or whether the ‘banding’ system of equality of large member states should still pertain, is one which can be settled entirely, explicitly and transparently within the chosen vote weighting mechanism.

The often reviled vote weighting system turns out to have many practical, not just theoretical, advantages. This does not mean, of course, that the settlement of Nice is the perfect answer ‘for all time’ (and thus deserving of constitutional status). But it would seem unwise at this stage, as implied in the text of the Draft Constitution, to rule out a review, and possible redesign, of the vote weighting system as an option for the period after 2009 or 2011. 6 — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — I H S

Table 1: The Nice Voting System

Weighting with state majority and (without population conditions * conditions) * SH Population weighted Vote Power (in '000s) votes Share index BC index SH index 82038 Germany 29 9.03 9.49 8.56 9.29 59247 UK 29 9.03 9.37 8.56 9.29 58966 France 29 9.03 9.37 8.56 9.29 57612 Italy 29 9.03 9.37 8.56 9.29 39394 Spain 27 8.41 8.67 8.12 8.61 38667 Poland 27 8.41 8.67 8.12 8.61 15760 Netherlands 13 4.05 3.95 4.23 3.98 10533 Greece 12 3.74 3.61 3.91 3.65 10290 Czech 12 3.74 3.61 3.91 3.65 10213 Belgium 12 3.74 3.61 3.91 3.65 10092 Hungary 12 3.74 3.61 3.91 3.65 9980 Portugal 12 3.74 3.61 3.91 3.65 8854 Sweden 10 3.12 2.99 3.27 3.02 8082 Austria 10 3.12 2.99 3.27 3.02 5393 Slovakia 7 2.18 2.07 2.31 2.10 5313 Denmark 7 2.18 2.07 2.31 2.10 5160 Finland 7 2.18 2.07 2.31 2.10 3744 Ireland 7 2.18 2.07 2.31 2.10 3701 Lithuania 7 2.18 2.07 2.31 2.10 2439 Latvia 4 1.25 1.17 1.33 1.19 1978 Slovenia 4 1.25 1.17 1.33 1.19 1446 Estonia 4 1.25 1.17 1.33 1.19 752 Cyprus 4 1.25 1.17 1.33 1.19 429 Luxembourg 4 1.25 1.17 1.33 1.19 379 Malta 3 0.93 0.88 0.99 0.89 Totals 321 100.0 100.0 100.0 100.0 Power Power Power quota (%) 72.27 Gradient Gradient Gradient i.e. threshold of 232 votes 57% 49% 56%

* i.e. a simple majority of states plus at least 62% of population source: Annex IV to Annex V, Council of the EU (see references); IHS I H S — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — 7

Table 2: The Status Quo extrapolated

Population Weighted Vote SH Power (in '000s) votes share index BC index 82038 Germany 10 8.06 8.30 7.60

59247 UK 10 8.06 8.30 7.60 58966 France 10 8.06 8.30 7.60 57612 Italy 10 8.06 8.30 7.60 39394 Spain 8 6.45 6.51 6.36 38667 Poland 8 6.45 6.51 6.36

15760 Netherlands 5 4.03 3.97 4.14 10533 Greece 5 4.03 3.97 4.14 10290 Czech 5 4.03 3.97 4.14 10213 Belgium 5 4.03 3.97 4.14 10092 Hungary 5 4.03 3.97 4.14

9980 Portugal 5 4.03 3.97 4.14 8854 Sweden 4 3.23 3.25 3.38 8082 Austria 4 3.23 3.25 3.38 5393 Slovakia 3 2.42 2.34 2.52 5313 Denmark 3 2.42 2.34 2.52

5160 Finland 3 2.42 2.34 2.52 3744 Ireland 3 2.42 2.34 2.52 3701 Lithuania 3 2.42 2.34 2.52 2439 Latvia 3 2.42 2.34 2.52 1978 Slovenia 3 2.42 2.34 2.52

1446 Estonia 3 2.42 2.34 2.52 752 Cyprus 2 1.61 1.57 1.69 429 Luxembourg 2 1.61 1.57 1.69 379 Malta 2 1.61 1.57 1.69 Totals 124 100.0 100.0 99.9

Power Power quota (%) 70.9 Gradient Gradient i.e. threshold of 88 votes 43% 37% source: Annex III to Annex V, Council of the EU (see references); IHS

8 — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — I H S

Table 3: The Draft Constitution Double Majority

Population SH Power (in '000s) weight#1 weight#2 index (a) BC index (b) 82038 Germany 1 82038 16.51 + 13.36 +

59247 UK 1 59247 11.27 + 9.52 + 58966 France 1 58966 11.22 + 9.49 + 57612 Italy 1 57612 10.93 + 9.30 + 39394 Spain 1 39394 7.73 - 6.85 - 38667 Poland 1 38667 7.60 - 6.79 -

15760 Netherlands 1 15760 3.36 - 3.59 - 10533 Greece 1 10533 2.52 - 2.93 - 10290 Czech 1 10290 2.48 - 2.90 - 10213 Belgium 1 10213 2.47 - 2.89 - 10092 Hungary 1 10092 2.45 - 2.87 -

9980 Portugal 1 9980 2.43 - 2.86 - 8854 Sweden 1 8854 2.26 - 2.72 - 8082 Austria 1 8082 2.14 - 2.62 - 5393 Slovakia 1 5393 1.73 - 2.27 - 5313 Denmark 1 5313 1.71 - 2.26 -

5160 Finland 1 5160 1.69 - 2.24 - 3744 Ireland 1 3744 1.47 - 2.06 - 3701 Lithuania 1 3701 1.47 - 2.06 - 2439 Latvia 1 2439 1.27 + 1.89 + 1978 Slovenia 1 1978 1.20 + 1.83 +

1446 Estonia 1 1446 1.12 - 1.76 + 752 Cyprus 1 752 1.02 - 1.67 + 429 Luxembourg 1 429 0.97 - 1.63 + 379 Malta 1 379 0.96 + 1.63 + Totals 25 450462 100.0 100.0

Power Power quota (%) 50.01 60.0 Gradient Gradient 82% 62% (a) gain over Nice +, loss compared to Nice - (b) gain over Nice +, loss compared to Nice - source: text of Draft Constitution; IHS I H S — Felderer, Paterson, Silárszky / Short Policy Paper for EU Convention Forum — 9

Table 4: Power Gradients (degrees of reform)

Power Gradients SH power BC power Convention Draft Constitution Double Majority 82% 62%

StatusQuo extrapolation May - Oct 2004 43% 37%

Nice weighting system in full 57% 49% source: IHS

References

Council of the , 11 Dec. 2002, Report on Enlargement, Annex V, 15524/1/02 REV 1 LIMITE ELARG 418, Brussels

Giovanni Grevi and Kirsty Hughes, Contribution (4) to Forum, What Prospects for compromise on Institutional Questions in the Convention endgame?, Centre for European Policy Studies (CEPS), Brussels

Kirsty Hughes, February 2003, Contribution (3) to Forum, The Battle for Power in Europe - Will the Convention Get it Right?, CEPS/EPIN Working Paper

Iain Paterson, 1998, Vote Weighting in the European Union - Confronting the dilemma of Dilution, Institute for Advanced Studies, East European Series No. 54, Vienna

Iain Paterson and Peter Silárszky, 1999, Redesigning the Institution of the Council of Ministers in Advance of EU Enlargement – Issues and Options, Journal of Institutional Innovation, Development, and Transition (IB Review), Vol. 3

Iain Paterson, Dec. 2000, Nice produces substantial Reform, but with some ragged edges, IHS Press Information.