Theme
This analysis provides an assessment of the optimal voting threshold for Spain in the likely event that the Nice treaty is abandoned for a system in which the member States voting power in the Council of the European Union is decided by real population size.
Summary
If Spain should agree to abandon the Nice agreement for a system in which voting power in the EU is dictated by real population size, it will be necessary to reconsider the voting threshold for a majority. When the issue was last discussed there was considerable contention over the value of the new threshold. The current proposal, more than 60 % for a majority and more than 40 % for a blocking minority, was considered by the Spanish authorities unfavourable, although negotiations stalled before the question was put to the test. The issue is now up for renewed discussion. This analysis shows, using the latest Eurostat population data, that Spain has good reasons to oppose the proposed threshold. The findings here indicate that if the goal is to maximize the country’s power potential and at the same time maintain equal conditions in terms of minimum coalition size to form a majority or a blocking minority across the six largest countries, Spain should consider majority thresholds of between 54 %-58 %.
Analysis
Following the change of government in Spain and the 24 March EU summit there is a renewed interest in reaching an agreement on the EU constitution. As a result, the question whether the voting procedures established by the Nice agreement should be replaced by a system based on a double majority is again at centre stage.
The double-majority system consists of a first round of votes in which each country has one vote. Once this first stage is over a second round of voting takes place, whereby a country’s voting power is weighted by the size of its population (see “Votar y vetar en la Unión” by Jose Ignacio Torreblanca in this series for more details on the mechanisms of the proposal). This analysis deals exclusively with the second voting procedure and its purpose is to quantify the power implications of choosing different voting thresholds for Spain and the five other ‘large countries’.
The Two Systems
The fundamental difference between the Nice agreement and a voting procedure based on real population size is that the former gives an unproportional weighting to population. That is, under the Nice agreement, Germany, which has a population twice the size of Spain’s, is almost equal to Spain in terms of voting power. In a system based on real population size the difference in the voting power of the two countries increases substantially. The increasing difference is clear if the voting power in the two systems is compared in the form of a table:
| Table 1. Relative voting power comparing the Nice agreement with a proposal assessing voting power as a function of real population size | ||||
| Country | Population Size1 Jan 2004 | Voting Power Proposal | Votes Nice | Voting Power Nice |
| Germany | 82,545,100 | 17.04% | 29 | 8.41% |
| France | 59,895,500 | 12.36% | 29 | 8.41% |
| UK | 59,517,900 | 12.29% | 29 | 8.41% |
| Italy | 57,481,600 | 11.87% | 29 | 8.41% |
| Spain | 40,977,600 | 8.46% | 27 | 7.83% |
| Poland | 38,193,700 | 7.88% | 27 | 7.83% |
| Rumania | 21,716,000 | 4.48% | 14 | 4.06% |
| Netherlands | 16,257,600 | 3.36% | 13 | 3.77% |
| Greece | 11,046,900 | 2.28% | 12 | 3.48% |
| Portugal | 10,479,800 | 2.16% | 12 | 3.48% |
| Belgium | 10,396,700 | 2.15% | 12 | 3.48% |
| Czech Rep. | 10,211,000 | 2.11% | 12 | 3.48% |
| Hungary | 10,115,200 | 2.09% | 12 | 3.48% |
| Sweden | 8,974,900 | 1.85% | 10 | 2.90% |
| Austria | 8,091,900 | 1.67% | 10 | 2.90% |
| Bulgaria | 7,799,300 | 1.61% | 10 | 2.90% |
| Denmark | 5,397,600 | 1.11% | 7 | 2.03% |
| Slovak Rep. | 5,380,700 | 1.11% | 7 | 2.03% |
| Finland | 5,220,200 | 1.08% | 7 | 2.03% |
| Ireland | 4,024,600 | 0.83% | 7 | 2.03% |
| Lithuania | 3,447,100 | 0.71% | 7 | 2.03% |
| Latvia | 2,318,500 | 0.48% | 4 | 1.16% |
| Slovenia | 1,996,700 | 0.41% | 4 | 1.16% |
| Estonia | 1,350,900 | 0.28% | 4 | 1.16% |
| Cyprus | 727,700 | 0.15% | 4 | 1.16% |
| Luxembourg | 450,800 | 0.09% | 4 | 1.16% |
| Malta | 399,600 | 0.08% | 3 | 0.87% |
| Total | 484,415,100 | 100.00% | 345 | 100.00% |
| Note: population data refer to EUROSTAT’s first demographic estimates for 2004 and correspond to the population on 1 January 2004. |
As shown in Table 1, the difference in relative voting power between, for instance, Spain and Germany is smaller under the Nice agreement than under the proposed system. Another feature visible when comparing the two systems is that, under the proposal, smaller countries (of a size smaller than Poland) have less voting power and large countries have more voting power.
One implication of this is that all the six largest countries stand to gain and all the 21 smaller countries stand to lose from switching to the proposed system. However, and herein lies the problem, the proportional system introduces much larger relative differences within the group of the six gainers. The standard deviation, or the average difference between the six large countries, rises from 0.3 to a stunning 3.3 with the new system (expressed in percentage points).
Power Implications
While it is clear that abandoning the Nice agreement brings about substantial changes at country level, it is less clear how their role as political actors will change. The reason for this is that no single country can have a majority. Unless of course the threshold for a blocking minority is set, for example, as low as 17 % of the votes, which would give Germany unrestricted power over EU decisions. Thus, a country’s political role and manoeuvrability in shaping EU policy is dependant on the voting threshold adopted. In sum, the name of the game in EU decision-making is forming coalitions that can accumulate sufficient votes to obtain the percentages required for passing or blocking proposals subjected to the vote. While this is intuitive, forming coalitions is very much about multiple choices.
Previously, before enlargement, when the EU consisted of only 15 members, coalition building was relatively straightforward. The number of possible unique coalitions was only 16,384 for each country. These coalitions varied in size from 2 to 15 countries. Some of these potential coalitions were of no use since they failed to provide the requisite votes to form a majority or a blocking minority. With enlargement, however, forming successful coalitions has become much more complicated. With 27 members the number of possible unique coalitions is 67,108,864, varying in size from 2 to 27 countries. Although many of these are useless coalitions for the reason described above, there is still a very substantial number of possible coalitions that could make a difference when voting (the number of possible coalitions that a member state can form in a political system such as the EU is defined by the formula 2(n-1), where n is the number of member states).
We will now take a closer look at how each of the six largest countries fare in terms of their capacity to form coalitions resulting in a majority or blocking minority at each percentage threshold between 1 and 100. The general assumption is that each member state’s power potential is a direct function of the number of coalitions resulting in a majority or blocking minority that it can form at any particular percentage threshold. That is, the more coalitions resulting in a total vote over a specific percentage threshold a country can form, the larger is a country’s power potential and the greater is its scope for political manoeuvring.
Figure 1 compares the number of possible coalitions at different voting thresholds applying the voting power dictated by the Nice agreement and by the proposal scrutinised in this paper. If, for instance, the threshold for a majority is set at 60 % (the x axis), the corresponding value on the y axis shows the number of unique coalitions each of the six countries can form resulting in a total vote of over 60 %. Similarly, a majority threshold of 60 % corresponds to a 40 % threshold for minority blocking. Thus, knowing the voting threshold we can simultaneously assess the number of unique coalitions with a vote large enough to block decisions available to each country. The vertical lines in the graph correspond to the threshold established in the Nice agreement (over 26 % for forming a blocking minority and over 74 % for a majority). A line at the 50 % level has been added as a benchmark for comparisons.
There are several conclusions to be drawn from Figure 1. First, while there are slight differences between the two systems, their respective distribution of possible unique coalitions rendering sufficient votes for a majority or a blocking minority appears to be similar across all possible thresholds. The primary reason for this is that the two systems rely on a strong population component to decide each country’s relative voting power.
The second and perhaps most important conclusion is that there is only one sure winner among the six largest countries if the Nice agreement is jettisoned – Germany. Applying a voting system weighted by actual population size would mean that Germany would increase or equal its capacity to form coalitions resulting in a majority or a blocking minority across all possible threshold combinations (1 %-100 %) with respect to the Nice agreement. Depending on which voting threshold we apply for a majority, the five other large countries could see either a deterioration or an improvement in their capacity to form successful coalitions. As shown in Figure 1, it is primarily their capacity to form coalitions with sufficient votes to block decisions that could deteriorate, whereas their capacity to form coalitions resulting in a majority could be significantly improved. Not surprisingly, there are differences between the five countries. Spain and Poland could experience a greater deterioration in their capacity to form blocking alliances compared with France, the UK and Italy.
Since the threshold for voting is still under discussion, the only way to compare the two systems directly is by applying the Nice agreement thresholds to both systems.
| Table 2. Number of possible coalitions to pass or block decisions, Nice agreement thresholds | ||||||
| Pass decision <74 % | Blocking <26 % | |||||
| Proposal | Nice | % diff. | Proposal | Nice | % diff. | |
| Spain | 6,482,103 | 2,298,331 | 282.04 | 65,369,480 | 66,822,416 | -2.17 |
| Germany | 8,110,326 | 2,349,818 | 345.15 | 66,953,684 | 66,870,164 | 0.12 |
| France | 7,440,007 | 2,349,818 | 316.62 | 66,310,536 | 66,870,164 | -0.84 |
| UK | 7,418,295 | 2,349,818 | 315.70 | 66,289,452 | 66,870,164 | -0.87 |
| Italy | 7,328,397 | 2,349,818 | 311.87 | 66,202,044 | 66,870,164 | -1.00 |
| Poland | 6,357,258 | 2,298,331 | 276.60 | 65,245,608 | 66,822,416 | -2.36 |
Table 2 confirms that the proposed system is more flexible for approving decisions. All six countries would see an increase in the total number of coalitions resulting in a majority of several hundred percent. In terms of their capacity to form successful coalitions to block decisions, the difference between the two systems is marginal. Spain and Poland would see a deterioration of around 2 %, while all the others would only experience marginal change.
It is clear that the Nice agreement is extremely rigid when applying the current thresholds for voting. It is rigid insofar as it strongly favours the possibility of forming successful coalitions to block decisions, to the detriment of coalitions resulting in majorities. All six large countries can form well over 66 million coalitions with enough votes to block a decision under the Nice agreement, while the number of possible coalitions to approve decisions is only around 2 million. Considering that the maximum number of unique coalitions is just over 67 million, Figure 1 shows that the Nice agreement gives the six largest countries an almost unrestricted capacity to block decisions when required.
Since the proposed system implies a greater flexibility to form majority coalitions, it is some improvement as far as rigidity is concerned. However, in general, a switch to the new system would not introduce any substantial changes as far as rigidity is concerned as long as the same voting threshold is applied as in the Nice treaty.
It could even be argued that abandoning the Nice agreement without reconsidering the established thresholds could complicate rather than improve EU decision making. The reason is that assigning voting power according to real population size opens up the possibility of forming coalitions consisting of only three countries with sufficient votes to block a proposal subject to voting. This would have been impossible with the voting power given by the Nice agreement, which require coalitions of at least four countries to block a proposal.
For example, applying the voting powers suggested in the proposal, Spain, Germany and France would add up to 37 % of the total votes, which is well above the 26 % threshold established in Nice. Under the Nice regulations, the same three countries would only account for 24.65 % of the votes. Even if we were to substitute Spain for, say, the UK, the total vote of the resulting triumvirate would still be below the established threshold. That is, while it always takes four countries to block a decision under the Nice agreement, it would only require three countries to block a decision if voting power is weighted by real population and if the thresholds established in the Nice agreement are maintained. Table 3 shows the six countries’ capacity to form unique coalitions consisting of only three or four countries with sufficient voting power to block a proposal under the two systems if the voting thresholds of the Nice agreement are applied.
| Table 3. No of possible unique three- or four-country coalitions for blocking decisions by country, thresholds 74/26 | ||||
| Proposal | Nice | |||
| Coalition size | 3 countries | 4 countries | 3 countries | 4 countries |
| Spain | 25 | 471 | None | 140 |
| Germany | 98 | 1246 | None | 173 |
| France | 49 | 795 | None | 173 |
| UK | 49 | 788 | None | 173 |
| Italy | 47 | 765 | None | 173 |
| Poland | 20 | 431 | None | 140 |
Optimising Spain’s Voting Power in a System Based on Real Population Size
| Table 4. Number of possible unique 3-, 4-, 5-, or 6-country coalitions to form a majority or a blocking minority applying the 60/40 threshold | ||||
| Coalition size | 3 countries | 4 countries | 5 countries | 6 countries |
| Pass decisions more than 60 % of the votes | ||||
| Spain | None | None | 1 | 33 |
| Germany | None | None | 2 | 60 |
| France | None | None | 2 | 58 |
| UK | None | None | 2 | 58 |
| Italy | None | None | 2 | 53 |
| Poland | None | None | 1 | 33 |
| Block decisions more than 40 % of the votes | ||||
| Spain | None | 19 | 603 | 6,144 |
| Germany | 3 | 88 | 1,494 | 13,024 |
| France | 2 | 59 | 948 | 8,423 |
| UK | 2 | 58 | 929 | 8,359 |
| Italy | 2 | 57 | 882 | 8,057 |
| Poland | None | 16 | 484 | 5,569 |
When discussing the conditions for abandoning the Nice agreement it is generally agreed that the voting threshold also has to change. So far, the discussion concerning the value of the “new” threshold is subject to a great deal of guesswork and political horse trading, whereby each of the six large countries have defended different views with the excuse of not weakening their institutional power relative to the other countries in the group. We shall now look at the possible range of thresholds that are most beneficial for Spain in terms of both the country’s total capacity to form coalitions with sufficient votes to block or approve a proposal, as well as its capacity to form successful coalitions with only three or four countries. The question I try to answer is what threshold or thresholds maximize Spain’s voting power?
One of the most frequently debated voting thresholds put forward in the discussion of whether to substitute the Nice agreement by a system based on real population size is 60 % (majority)/40 % (blocking minority).
When the European Constitution was being debated, Spain opposed these thresholds: as it turns out Spain had good reasons to oppose them. Applying the 60/40 thresholds would strengthen the institutional power of the four largest countries – Germany, France, the UK and Italy, while Spain and Poland to some extent would see no improvement in their capacity to approve and block decisions.
As shown in Table 4, applying the 60/40 thresholds both Spain and Poland can still form several coalitions consisting of only four countries with enough votes to block a proposal. However, compared with the Nice treaty the number of possible coalition combinations for achieving this result decreases substantially for both countries (see Table 3 for a comparison). For the other four countries the case is similar. But, on the upside, and in contrast with the Nice regulations, with the 60/40 thresholds they can block decisions by invoking a very limited number of three-country coalitions. That is, accepting the 60/40 threshold gives the four largest countries substantial power leverage over Spain and Poland compared with the Nice treaty.
What are Spain’s options if it wants to avoid this split in power potential between the six large countries? In principle there are three solutions to the problem: (a) maintain the Nice agreement as it stands today; (b) lower the threshold for minority blocking to allow itself the option to block proposals relying on coalitions consisting of only three countries; (c) increase the threshold for minority blocking so that the four largest countries can no longer block decisions through coalitions consisting of only three countries.
Option (a), while being simple, is less attractive than it appears. The reason is that all the six large countries would benefit from having their voting powers dictated by their real population size. Thus, choosing option (a) is a lost opportunity to gain power over the 21 smaller countries in the European Union. Hence, for Spain and the other five large countries, the rational choice is between options (b) and (c).
Looking at the choices available to meet the conditions in option (b) first, it turns out that the range of possible minority thresholds that would allow Spain to successfully rely on three-country coalitions for blocking proposals are all those thresholds larger than 30 % (29.40 %) and smaller than 38 % (37.46 %). (1)
Minority thresholds for values below 29.40 % would imply that Germany and France could form a two-country blocking minority, and values below 29.33 % imply that Germany and the UK could form a two-country blocking minority and so forth. Minority thresholds above 37.46 % imply that Spain is forced to rely on four-country coalitions or larger to form a successful blocking minority, while the larger countries can still block decisions relying on coalitions between three countries.
Figure 2 shows the number of possible three-country coalitions resulting in enough votes to successfully block decisions with minority thresholds larger than 30 % and smaller than 39 %. Note that Spain and Poland cannot block decisions by relying on coalitions consisting of three countries if the threshold for minority blocking exceeds 38 %.
The other option available to Spain is to argue for an increase in the blocking minority threshold and to make it high enough to prevent all the six countries from blocking decisions on the basis of coalitions between only three countries (option (c) above). At first sight it might seem that this could be achieved by deciding on a threshold for minority blocking that is larger than 42 % and smaller than 51 %.
However, this option is more complicated. Raising the blocking minority threshold automatically lowers the majority threshold. When the majority threshold reaches a certain level it opens up the possibility for some countries of approving decisions by relying on coalitions consisting of only four countries. Germany, France, the UK and Italy together have 53.59 % of the votes. This means that the four countries together, without the help of a fifth country, can have a majority vote as soon as the threshold for a majority reaches a value equal to or below this level. For values above 53.59 % it takes at least five countries to achieve a majority. The possibility of achieving a majority relying on only four countries becomes available to Spain when the threshold for a majority is set at equal or below 50.18 %, since Spain, together with Germany, France and the UK have 50.18 % of the total votes. The split in the capacity to reach a majority through a four-country coalition occurs at all minority thresholds larger than 47 and smaller than 50.
Considering this, if Spain opts to raise the blocking minority threshold, while at the same time preserving its capacity to form majorities on an equal basis with the larger countries, it should consider minority thresholds of above 42 % (41.72 %) and below 47 % (46.41 %). Furthermore, with the population estimates used in this analysis, it could also consider a 50/50 threshold for majority and minority blocking.
Figure 3 shows the number of possible unique four-country coalitions with sufficient votes to block a decision available to the six countries under the thresholds indicated above. Note that the special case where the threshold is set at 50/50 would imply that Poland is deprived of the possibility of blocking decisions on the basis of four-country coalitions, while Spain has one such option.
The question that remains is which option is the most attractive? Both options (b) and (c) are feasible in so far as they would prevent Spain from losing institutional power relative to the four largest countries. Or expressed differently, the six countries have the same restrictions in forming small coalitions to approve or block EU decisions. In option (b), forming a blocking minority require coalitions of only three countries or more. Forming a majority requires coalitions of at least six countries (except at the 70/30 level when seven countries are required).
In option (c), a blocking minority is only achieved by forming coalitions involving four or more countries, while a majority requires coalitions involving five or more countries. The only difference is that Spain and Poland have fewer coalition options than the four larger countries because of their weaker voting power.
Both options make it possible for the six largest countries to control most of the EU’s decisions. That is, approving and blocking decisions could be achieved within the group of the six largest countries and without the consent of the smaller countries. Recall, however, that the proposal consists of a double majority, and that a majority of countries, regardless of population size, first have to approve an initiative before it becomes subject to a vote based on population size. Regardless of the number of countries it takes to approve or block an initiative in this way, the smaller countries will not be redundant even if the Nice agreement is abandoned.
One way of assessing which option is most attractive is by returning to the concept of flexibility in approving and blocking decisions. Regardless of which threshold is adopted, the cut-off level dictates each country’s flexibility to form coalitions with enough votes to achieve either a blocking minority or a majority. The more options available for the two alternatives, the greater the room for political manoeuvrability.
Figure 3 shows the sum of the number of unique coalitions that would result in a majority or a blocking minority for Spain at different threshold values. The light grey areas in the Figure represent the range of thresholds available to Spain in options (c) and (b) discussed above. Option (c) is on the left and option (b) on the right. The dark grey area represents thresholds that would give the four largest countries –Germany, France, the UK and Italy– power leverage over Spain in terms of minimum coalition sizes to approve or block a proposal.
Figure 4 show that option (c) brings Spain closest to its maximum power potential, without having to consider yielding power to the larger countries in terms of minimum coalition size to block decisions.
The question is whether choosing option (c) would also mean that the larger countries maximize their power potential? In principle, the answer to this question is yes. The closer the 50/50 threshold, the closer every other country in the European Union would be to its maximum power potential. However, since the power potential is exponential, regardless of which threshold value Spain agree to, as long as the majority threshold is larger than 50 %, Spain, because of its lower voting power, will always be closer to its maximum power potential than countries with a higher voting power. Thus, Spain is likely to be more successful in maximizing its power potential compared with the larger countries, although the larger countries will always have more options to choose from in absolute terms.
Conclusions
While the present analysis is complex given the extremely large number of coalition possibilities implied by an enlarged European Union, the conclusions, if Spain accepts to abandon the Nice treaty for a system based on real population size, are simpler.
To avoid a deterioration of its power relative to the four largest EU countries when negotiating new voting thresholds, Spain should consider only majority thresholds of over 54 % and under 59 %, or alternatively majority thresholds of over 62 % and under 71 %. All other majority thresholds would imply a more or less significant loss of power to the four largest countries compared with the Nice treaty, since they would imply that Spain would be forced to rely on more countries to approve or block decisions than required by the four largest countries.
This analysis has also indicated that if the goal is to maximize the country’s power potential and maintain equal conditions in terms of minimum coalition size to achieve a majority or a blocking minority with the six largest countries, Spain should consider majority thresholds of between 54 %-58 % rather than between 62 %-70 %.
Finally, a note of caution in interpreting the results presented in this analysis. I have used the latest population estimates provided by Eurostat in assigning voting power. While this analysis has shown that it is possible to prevent a divide in the six countries’ power relations by manipulating the voting thresholds, if the population data should change this could imply that the thresholds discussed also change. Thus, before making a final assessment of which thresholds should be used in a system based on real population size it is first necessary to agree on the true population data.
Notes:
(1) For the sake of simplicity I only consider threshold values rounded upwards to the nearest integer when assessing the optimal cut-off for thresholds discussed in this part of the analysis. This means that I analyse thresholds as intervals between 1 and 100, rather than as a continuous values, where cut-offs could occur at any value between 1 and 100. There is no implication as long as the agreed threshold is also rounded in the same way. However, if for some reason the agreed threshold is set at a value between, for instance, 37 % and 38 % (say 37.589 %), this could change the number of possible coalitions a country can form compared with this analysis.
Rickard Sandell
Senior Analyst, Demography Population and International Migration, The Elcano Royal Institute