Executive Selection in the European Union: Does the Commission President Investiture Procedure Reduce the Democratic Deficit? (*)
Simon Hix
European Integration online Papers (EIoP) Vol. 1 (1997) N° 21;
Date of Publication in EIoP: 14.11.1997
| Abstract | Back to homepage | PDF | PS |



1. Introduction: Executive Accountability in the European Union

The holders of political power in the European Union (EU) institutions must be made more accountable. This issue was forced onto the European agenda in the wake of the Maastricht Treaty, following the Danish rejection of the Treaty in a referendum, the narrow votes on the Treaty in the French referendum and the British parliament, and the German constitutional court ruling that the EU cannot be integrated further unless the institutions are made more democratic (cf. Weiler et.al., 1995). The so-called ‘permissive consensus’, where Europe’s leaders can build a European polity without requiring a priori consent from their voters, no-longer exists. EU Heads of Government consequently declared that a central goal of the Intergovernmental Conference preparing the reform of the Maastricht Treaty, that produced the Draft Amsterdam Treaty, was to make ‘the institutions more democratic’ (European Council 1996: 2).

The challenge for the EU institutions, however, is that a central element of the practice of democracy in all political systems is the ability of citizens to ‘throw out’ the holders of executive office, through the process of competitive elections (esp. Weber, 1918; Schumpeter, 1943; Schattschneider, 1960; King, 1981). In a parliamentary system, this operates through an indirect mechanism: where the executive must command a parliamentary majority, immediately following parliamentary elections. In a presidential system, in contrast, the mechanism is direct: where voters make a direct choice between rival candidates (Lijphart, 1992; Shugart & Carey, 1992; Sartori, 1994).(1)

But, the EU has a ‘dual executive’ (cf. Lenaerts, 1991). On the one hand, national governments (in the Council and European Council) possess long-term executive power: they set the overall political and legislative agenda (such as who qualifies for Economic and Monetary Union [EMU]); delegate short-term executive and regulatory power to the Commission; have the sole right of policy initiation in the intergovernmental provisions on the Common Foreign and Security Policy and Justice and Home Affairs; and maintain a monopoly over the forces of coercion (i.e. police and armed forces). On the other hand, in the everyday making of EU social and economic policies, the EU Commission exercises considerable executive power: it has an influential leadership role (as in the Delors Plan for EMU); has the sole right of legislative initiative; is the ‘guardian of the Treaties’; has powerful rule-making powers in the regulation of the EU Single Market (as in EU competition policy); and is responsible for the execution and administration of legislation (in cooperation with the member states) and the budget. In the exercise of these powers, moreover, whereas the Council is a true ‘collegial executive’, with a rotating President, the Commission is more like a national cabinet, where the Commission President is the primus inter pares: the ‘first among equals’.

Furthermore, prior to the Maastricht Treaty and the Draft Amsterdam Treaty, neither of these branches of the dual EU executive were accountable via the classic indirect or direct methods. When exercising executive power, the Council operates as a ‘consociational’ institution (cf. Taylor 1991; Chryssochoou 1994): decisions are made by unanimity, which allows each member of the executive a ‘mutual veto’ to protect a vital interest of its constituents (its ‘pillar’) (Lijphart, 1969) Qualified majority voting in the Council only takes place under certain legislative procedures, and especially where legislative power is shared between the Council and the European Parliament (EP), under the cooperation and co-decision procedures. Furthermore, the Draft Amsterdam Treaty mandates the Council to make a clearer distinction between when it is exercising legislative and executive power, as a prerequisite for making the EU legislative process more transparent.(2) As a result, when acting as the EU executive, the Council is not collectively accountable to Europe’s voters (either directly, or indirectly via the EP). Nevertheless, each individual member of the Council is accountable via the national channels of executive selection: i.e. through national general elections. Therefore, as long as Council decisions are by unanimity when exercising executive power, every European citizen has an indirect say in the process, through the election of the national government. In other words, the legitimacy of the Council in this regard is identical to that in other collegial-executives in consociational systems: where each ‘societal pillar’ is an equal participant in executive power - as in the fixed system of representation of the linguistic groups and familles spirituelles in the Swiss Federal Government.


However, the accountability of the Commission is even more problematic. Prior to the Maastricht Treaty, the Commission President was chosen unanimously by the national governments. This was more akin to choosing the head of an international organisation than selecting a chief executive. As a result, the EP argued in the 1984 ‘Draft European Union Treaty’, and in the subsequent reports to the Intergovernmental Conferences that produced the Single European Act and the Treaty on European Union (the Maastricht Treaty), that it should have a role in the selection process. The EP used the argument that Europe’s voters are not interested in EP elections because they have no impact on the make-up an executive. Hence, by giving the EP a say in choosing the Commission President, European elections would really start to matter (e.g. Lodge, 1995). Backed by a large majority of member states and the European Commission, this proposal was introduced in the Maastricht Treaty, and will be reinforced if the Draft Amsterdam Treaty is ratified. The result of these reforms is a quasi-parliamentary system (i.e. the classic indirect mechanism) for selecting this second branch of the EU executive: whereby the Commission President is nominated by the member states and ratified by the EP immediately following the EP elections.

According to this reading, therefore, the Maastricht and Amsterdam Treaties have increased the democratic accountability of at least one part of the EU executive. But, has this quasi-parliamentary model of EU executive selection really allowed Europe’s voters to chose the EU executive?(3) Can European citizens throw-out the Commission President? Or, is more fundamental institutional reform needed? These questions remain unanswered without further theoretical and empirical investigation. First, however, the operation of the new Commission President investiture procedure, and academic consensus regarding the impact of the procedure, is discussed in more detail.

2. Choosing the Commission President: Towards a Parliamentary Model?

2.1. The New Commission President Investiture Procedure

As a result of the Maastricht Treaty, in December 1991, Article 158 (para. 2) of the EC Treaty states that:

The governments of the Member States shall nominate by common accord, after consulting with the European Parliament, the person they intend to appoint as President of the Commission.

Furthermore, the rest of the Article (para’s 1 & 3) specifies that the term of office of the Commission is five years, with the first Commission subject to these rules beginning its term on 6 January 1995. In other words, the terms of the Commission and the Parliament are concurrent. This enables the issue of the Commission President to be debated at the first plenary session following each set of EP elections.

When the EP drafted its new Rules of Procedure, however, the Members of the European Parliament (MEPs) deliberately took Article 158 to imply that they would be entitled to take a vote by a simple majority on the governments’ nominee, and that if the vote were negative the nominee would be withdrawn. This was a specific reading, as the word ‘consultation’ could imply that the Parliament simply be required to issue a statement on the nominee, which the governments could subsequently disregard. Nevertheless, the EP’s interpretation of the new investiture procedure was confirmed in a written response by the then President-in-office of the EU Council, Klaus Kinkel (the German Foreign Minister), to a question by the EP President, Klaus Hänsch. In this letter, Kinkel accepted that if the EP votes against the governments’ nominee, the Council will find a new candidate.

Furthermore, the Draft Amsterdam Treaty proposes to institutionalise the EP ‘interpretation’. On the recommendation of the Bourlanges and Martin Reports, the proposed Article 214 (para. 2) (ex Article 158) states that:

The governments of the Member States shall nominate by common accord the person they intend to appoint as President of the Commission; the nomination shall be approved by the European Parliament.

In effect, if the Amsterdam Treaty is ratified, ‘after consulting with the European Parliament’ will be replaced by ‘the nomination shall be approved by the European Parliament’. The Parliament’s right to vote by a simple majority on who should be the Commission President would thus be confirmed.


2.2. The Academic Consensus: A Major Reduction of the Democratic-Deficit

The ‘consensus view’ amongst scholars of the EU is that these changes significantly reduce the EU democratic-deficit. As three of the EP’s own institutional experts argue:

These changes will not in themselves make the relationship between the Community’s executive and its elected parliamentary body akin to that which exists between the government and the lower chamber of a parliament in member states, but it is certainly a step in that direction. The fact that European elections will be followed by the formation of a new executive potentially adds to the significance of the elections (Corbett et.al., 1995: 249).

Or, as a former Secretary-General of the Commission remarked:

Very soon the preliminary approval of [the Commission President] by the Parliament will be the decisive factor; the final appointment by the governments will retain only a formal character (Noël, 1992).

And, following the Draft Amsterdam Treaty, two other EU scholars argue:

... this might in the medium run change the character of the game of how the President is selected. His or her programme will be under parliamentary scrutiny. Due to this shift, the terms of the debate might be reshaped on party-political lines. This might, in turn prompt the design and marketing of alternative ideological designs of the European agenda (Falkner & Nentwich, 1997: 4).

In other words, the academic consensus is that the new Commission President investiture procedure establishes a new source of indirect legitimacy for executive authority in the EU, via the European elections and channelled through the EP. If this were in fact the case, it would eliminate a major element of the democratic-deficit.

However, this reading of the new procedure rests on two rather spurious assumptions:

  1. that in European Parliament elections, voters are able to choose between rival candidates for the Commission President, or at least rival platforms for European level action; and
  2. that the investiture of the Commission President will be determined by a majority in the European Parliament - which reflects an electoral majority - and not by (or at least, in addition to) the unanimity of national governments in the European Council.

Before these assumptions can be accepted, however, they must be subjected to theoretical analysis and empirical testing.

3. Theoretical Analysis: Party Leaders, MEPs and the Commission Presidency

The key political actors in the Commission President investiture procedure are the (national) party leaders (the main protagonists in European elections and the participants in the European Council), and the MEPs. These actors’ have certain political goals: for example, re-election. However, the ability to pursue these goals is mitigated by two factors: the strategic behaviour of competing actors (e.g. parties in the case of national party leaders); and by the structure of institutional constraints - which can determine, for example, which actors are more capable of achieving their goals (i.e. Shepsle 1989). In other words, the issue for theoretical investigation is: how does the structure of European elections and the Commission President investiture procedure shape the strategic behaviour of parties and MEPs?


3.1. Party Leaders: In Pursuit of National Government Office

Most theories of party behaviour assume an interaction between two primary goals: ‘political office’ and ‘public policy’ (e.g. Downs 1956, Riker 1962, Strom 1990). The structure of the EU system ensures that these goals are rewarded more at the national level than at the European level. The EU has an increasingly central role in setting social and environmental standards in the EU single market. However, the major areas of public policy and public expenditure, such as health, education and housing, are still controlled by national governments. Moreover, the focus of media attention is on the holders of national executive office rather than European executive office (such as the Commission President). And, national executive office-holders have a central (executive and legislative) role at the European level, in the EU Council and the European Council. In other words, the key political office for party leaders in the EU is still national government, the pinnacle of which is the national Prime Minister (or President in the case of France). The office of Commission President is of secondary importance, and is only important to national party leaders if it can help or hinder their chances of (re-)capturing national governmental office.

So, what does this mean for the process of selecting the Commission President? First, in EP elections, an implication of this national party preference structure is that regardless of whether the EP has a say in the selection of the Commission President, national parties will fight EP elections as a re-run of a national general election. Because national parties primarily seek national government office, the opportunity to use European elections (like regional and local elections) as a chance to vote ‘for’ or ‘against’ the party(ies) that holds national executive office will be simply too great to waste. National party leaders, particularly in opposition, will use European elections as a chance to voice a protest against governing parties.

This theoretical analysis was first proposed at the time of the first direct elections to the EP, in 1979, in Reif and Schmitt's (1980) famous ‘second-order national election’ model of EP elections. The expectation of the model is that in EP elections there will be a lower turnout than in national elections (because they do not influence the make-up of national government), and there will be more votes for opposition, minor and protest parties than in the previous general election (ibid., Reif 1984, Eijk & Franklin 1996). In addition, although the question is addressed by Reif et.al., the assumption that national governmental office is more important the Commission Presidency, implies that the second-order election model will still hold even if the EP has the power to choose the Commission President.

Second, in the actual Commission President investiture procedure, an implication of this national party preference structure is that national parties will be more concerned about (re-)capturing national governmental office than the party affiliation of the Commission President. Because parties also seek public policy, and the policy agenda of the EU can constrain the choices of national governments, national parties would like to have a Commission President with a similar policy agenda, in other words from the same party family (e.g. Socialist, Christian Democrat, Liberal, Conservative). However, because the policy-impact of the Commission President is weaker than the policy-impact of national governmental office, national parties are not prepared to jeopardise the chances of winning national governmental office, simply to secure a Commission President with the same policy platform.

However, this trade-off between office- and policy- goals has particular implications given the structure of the new Commission President investiture procedure. As the procedure is currently designed, parties in national government (represented in the European Council) collectively choose a nominee for Commission President. These parties consequently have a ‘vested interest’ to ensure that this candidate is approved by the EP in the subsequent EP investiture vote. If the MEPs from these parties vote against the European Council candidate, it would be embarrassing for the parties in national government. Opposition parties and the national media would point out that the governing party is divided, that the European policy of the party is unclear, and that the leader of the governing party is incompetent. Hence, parties in the European Council will do everything they can to ensure that their MEPs do not jeopardise their chances of holding on to their precious office-goal. Meanwhile, the incentive structure for parties in opposition in the domestic arena is much weaker, as they have no a priori commitment to any particular candidate for the Commission President.


3.2. MEPs: Enforcing National Party Loyalty

The other key actors in the Commission President investiture procedure are the individual MEPs. As with other parliamentarians (and like national party leaders), MEPs make trade-offs between office- and policy-goals: between re-election, and their ‘sincere’ ideological position (i.e. Mayhew, 1974; Cox & McCubbins, 1993). In terms of the ability of MEPs to secure re-election MEPs are dependent upon their national parties. In most member states (Austria, Denmark, Finland, France, Greece, Luxembourg, Netherlands, Portugal, Spain, Sweden), candidate lists are drawn up and controlled by the national party leaderships. In the others, MEPs are elected on regional lists (Belgium, Germany and Italy), in multi-member constituencies (Ireland) or in single-member constituencies (United Kingdom). However, even in these systems, MEPs acknowledge the elections are party-centred rather than candidate-centred; and hence that their re-election is more dependent upon national party (leaders’) fortunes than their individual popularity or performance in the EP (Bowler & Farrell, 1993). In other words, MEPs’ ability to secure their office-goals depends on national party support for their behaviour in the EP.

However, MEPs’ ability to secure their policy-goals depends on the EP party groups. EP party groups control the key committee appointments, the appointment of the President and Vice-Presidents of the EP, the choice of rapporteurs, and the legislative agenda of the plenary sessions (esp. Hix & Lord, 1997: Ch.5). If individual MEPs deviate from EP party group positions, their chances of securing these goals are consequently reduced. Although the EP party groups have a limited ability to suspend the group-membership of an individual MEP, the prospect of promotion to an important committee position or rapporteurship can be withdrawn, and whole national delegations can be expelled by a simple majority in a party group meeting. As a result, whereas the second-order structure of EP elections facilitates EP party group fragmentation, the structure of internal EP rules and incentives facilitates coalescence of individual MEPs and national party delegations into fewer EP party groups (Bardi, 1996). In addition, as this incentive structure has evolved, the cohesion of the EP party groups in roll-call votes has increased dramatically between the 1984-1989 and 1989-1994 sessions (Attiná, 1990; Quanjel & Wolters, 1993; Brzinski, 1996; Raunio, 1996).

The fundamental issue, therefore, is what happens when the office- and policy-goals of individual MEPs conflict? First, if national parties have no vested interest in MEP behaviour, MEPs can keep their national party leaders and EP party group leaders happy at the same time. Second, however, when national party leaders have a vested interest in MEPs behaviour, a trade-off must be made. For example, if a national party wants an MEP to vote in favour of a proposal, and the majority position of her EP party group is against the proposal, an MEP is torn: either to vote against the national party/with the EP party group and risk de-selection in the next EP election; or to vote with the national party/against the EP party group and risk removal of important privileges and policy influence in the EP. However, because of the structure of MEP candidate selection, the rational strategy in this situation is to side with the national party. The (national party) threat of de-selection far out-ways the (EP party group) threat of expulsion from the EP party group: de-selection not only means removal of an MEPs main office-goal, but also the removal of the ability to achieve policy aims through a seat in the EP; whereas expulsion from an EP party group only reduces the chances of achieving an MEP's policy goals.

There may be few occasions when national parties have a vested interest in EP votes, and hence when MEPs must make this trade-off. However, as from the previous discussion about national party leaders, MEPs clearly face this trade-off in the Commission President investiture procedure. This is because parties in national government make an a priori commitment for a particular candidate in the first round of the procedure, in the European Council. Because of the threat of de-selection, MEPs from these parties are unlikely to break from this support for the European Council candidate, regardless of the position of their EP party group. The remaining MEPs, from parties not holding the prime-ministers office in the national arena, can be expected to vote ‘for’ or ‘against’ the Commission President candidate on the basis of the candidate policy position. In this case, MEPs from the same party family, or on the same side of the Left-Right division in the EP, as the European Council candidate are likely to vote ‘for’, whereas the rest are likely to vote ‘against’.

The result, however, is that the coalition in the EP of ‘parties in the European Council plus EP party groups with similar policy preferences to the European Council candidate’ is almost always going to beat the coalition of ‘parties not in the European Council who also do not share the same policy preferences as the European Council candidate’. In other words, because of the structure of the investiture procedure (where parties in national government have a vested interest in the EP vote), and the structure of MEP candidate selection (by national parties rather than EP party groups), the European Council candidate for Commission President is unlikely ever to be rejected by the EP under the new investiture procedure. This theoretical analysis consequently suggests that there is little practical difference between the original Commission President selection process, where the EP did not have a say, and the procedure established by the Maastricht and Amsterdam Treaties.


Overall, this theoretical investigation produces two hypotheses:

  1. regardless of whether the EP has a say in the Commission President investiture procedure, EP elections will still be fought as second-order national contests; and
  2. even under the new Commission President investiture procedure, the European Council will be able to impose its wishes in the EP vote.

These propositions are directly contradictory to the consensus view about the new Commission President investiture procedure. However, before the consensus view can be rejected convincingly, these theoretical arguments need to be tested empirically.

4. Empirical Test: The Consensus View Refuted

4.1. European Parliament Elections: ‘Second-Order National Contests’

The second-order-national-election model of EP elections was clearly true for the first two direct elections to the EP, in 1979 and 1984. In the 1989 and 1994 EP elections, it has been argued that certain ‘European’ characteristics did emerge: a transnational concern for the environment and a consequent breakthrough for a number of Green parties in 1989 (Curtice, 1989); and a widespread anti-European vote in 1994 (Smith, 1996). As Table 1 shows, nevertheless, the general principles of the second-order model still held in these contests.

Table 1

Turnout continued to fall, from 68% in 1989 and 65% in 1984 to 63.4% in 1989 and 58.6% in 1994. And, a large proportion of voters in most member states continued to vote for parties they would not normally choose. In Denmark in 1989 and 1994 and in France in 1994, the high proportion of voters choosing other parties indicated the emergence of a different party system for European elections. In these member states, therefore, European elections began to take on their own dynamic: with voters and parties beginning to be aligned on the pro- and anti-European dimension as well as on the traditional left-right (cf. Hix & Lord, 1996). Nevertheless, for all the other member states, European elections continued to be fought by national parties about national executive power.

All three European party federations - the Party of European Socialists (PES), the European People’s Party-Christian Democrats (EPP), and the European Liberal, Democratic and Reform Party (ELDR) - mentioned the connection between the EP elections and the selection of the Commission President in their 1994 European Election Manifestos. These manifestos were signed by the national party leaders of the member parties of the federations, at special party leaders’ meetings. And, at a meeting in Brussels in December 1993, the Socialist party leaders even discussed whether they should propose a candidate for the Commission Presidency prior to the European elections (Hix & Lord, 1996). However, Europe’s voters and domestic parties took no notice. As in all the other European elections, the European Party manifestos were ignored and European issues were conspicuous by their absence.

In other words, the theoretical hypothesis, that EP elections would still be fought as second-order contests regardless of the role of the EP in the Commission President investiture process, is confirmed. EP elections are not about who holds executive power at the European level, or even about rival agendas for European action. This may change with a uniform electoral procedure, where some candidates are nominated by the European party federations, and where the transnational Election Manifestos are used more widely. However, national parties would still have a strong incentive to fight the campaigns as mid-term ‘beauty contests’ for national government. National executive office (the Prime Minister) is a more prestigious prize than the Commission Presidency. Consequently, genuine ‘European’ elections could only exist if national electoral competition can be prevented from interfering in the EP election process. But, it is difficult to imagine how this could ever happen in more than a few member states, regardless of the electoral rules. This is thus a real blow to the consensus view about the new investiture procedure.


4.2. The Commission President Investiture Vote: A Coalition of Governments Not Voters

Turning to the second theoretical hypothesis, the first (and only) time the Article 158 Commission President investiture procedure was used was in July 1994, following the June 1994 EP elections. On 24-25 June, at the Corfu Summit, the European Council was unable to agree on a nominee, due to John Major’s refusal to back Kohl and Mitterrand’s choice - the Belgian Christian Democrat Prime Minister, Jean-Luc Dehaene - because the British government perceived him to be too ‘federalist’. Klaus Kinkel, the President-in-office of the Council, hastily organised an emergency European Council on 15 July, and Kohl persuaded Jacques Santer, the Luxembourg Christian Democrat Prime Minister, to put his name forward. He was subsequently supported unanimously by the heads of government of the member states.

On 21 July, at the first plenary session of the new Parliament, the MEPs voted on whether they would accept or reject the European Council nominee. The day before the vote, Jacques Santer attended the meetings of the PES group and EPP group, to urge them to vote ‘yes’. Following these meetings, however, the PES group voted overwhelmingly to oppose him, because they objected to the secretive nature of the nomination process in the European Council, because they feared that he was the ‘lowest common denominator’ candidate of the British Prime Minister, and because they argued that Europe’s voters had supported a centre-left majority in the 1994 EP elections. The EPP, in contrast, voted to back him, as a fellow Christian Democrat and former President of the EPP party federation. The other groups all took one or other position, or simply opposed Santer as a vote against the EU in general. As a result, the day before the vote, the line-up was as shown in Table 2.

Table 2

Would this be translated into a the real vote on 21 July? If the consensus view of the new procedure were right - that the EP electoral choices could be translated into action by the EP party groups - then Jacques Santer should have lost the EP investiture vote.

However, on 21 July, the MEPs voted 262 in favour and 244 against, with 22 abstentions and 39 absentees. In one of the highest roll-call turnouts in the EP’s history, Santer was elected as Commission President by 18 votes. Of those who voted, the line-up of the EP party groups on the previous day explained 88.9% of MEP behaviour: in other words, 56 MEPs (11.1%) voted against their party group positions. Had these MEPs voted with their party groups, Santer would have lost. So, what explains this breakdown of the ‘natural’ EP electoral/party group majority against Santer?

As our theory suggests, most of these MEPs who broke from their party group line did so under the instructions of their national parties. These MEPs were members of parties in government at the national level, who had already supported Santer in the European Council. For example, after the majority vote in the PES group meeting (the day before the vote), the leader of the Spanish PSOE delegation informed the PES meeting that his party would not abide by the PES group position. Madrid had instructed the PSOE delegation to support Santer, regardless of the PES group line. We can reasonably assume that this was repeated for all parties who had been in the European Council (i.e. who held the Prime Ministers’ office). The winning coalition in the investiture vote was hence ‘parties in the European Council plus on the right in the parliament’, and not a simple Left-Right majority from the EP elections. And, as Table 3 shows, this coalition structure, explains 96% of MEPs behaviour in the investiture vote: and 7.1% more than a simple Left-Right explanation.(4)

Table 3

This consequently confirms the theoretical hypothesis that because the European Council went first in the investiture process, national parties in government had an incentive to impose their wishes on their MEPs in the EP vote. National parties were able to do this because they control the selection of candidates in EP elections. For the MEPs torn between a national party position (for Santer) and an EP party group position (against Santer), the national parties could threaten to prevent the MEPs from standing in the next EP election.

The only time the Article 158 investiture procedure was used hence suggests that the second assumption of the consensus view is also weak. Despite previous high levels of party group cohesion, the EP party groups were not able to translate the 1994 electoral majority into a positive vote either for or against a Commission President. Instead, the national governments were able to impose their choice on the Parliament, through the creation of an unholy alliance in the EP: of parties whose leaders had backed Santer in the European Council plus ‘natural allies’ (from the same party family as Santer) who would have voted for him anyway. Consequently, the Commission President was still the choice of Europe’s governments and not of Europe’s voters.


5. Normative Implications: Limits of the Parliamentary Model at the EU Level

Contrary to most academic commentators, therefore, these theoretical and empirical results suggest that the Maastricht and Amsterdam reforms do not in fact allow Europe’s voters to chose who holds executive office in the EU - and hence ‘throw them out’ next time round. At face value, the new procedure for selecting the Commission President pushes the EU towards a parliamentary model of executive selection, by attempting to connect executive selection to a majority in European Parliamentary elections. However, at a theoretical level, there are no incentives for parties to compete for the office of the Commission President. And, there is every incentive for individual MEPs to side with the wishes of their national party leaders (to preserve their re-election) rather than the majority opinion of their EP party group.

These theoretical arguments are also conformed at an empirical level. First, the 1994 EP elections were still fought by national parties on the performance of national executives and not on rival candidates for the office of Commission President. Second, in the EP investiture vote, the winning majority was a ‘prime ministerial coalition’, of parties holding the national Prime Ministers’ office plus minority (centre-right) partisan allies of their nominee in the EP, and not a ‘European party coalition’, of the (centre-left) European parties/EP party groups that had ‘won’ the June 1994 EP elections.

Presented with this evidence, Nentwich and Falkner (1997: 8) contest that this view ‘might underestimate the dynamics of European level politics because of the gradual increase of powers of the EP’. Moreover, Nentwich and Falkner remain ‘convinced that the system set in place by the Amsterdam Treaty will give forceful impetus with a view to politicising and Europeanising the selection of the European executive’.

But, the theoretical and empirical results suggest that this developmental argument only holds if there is a substantive shift from national to European politics: where national party leaders see the office of the Commission President as a partisan appointment, and are willing to jeopardise their capture of national government office to secure the Commission Presidency for their own ‘party family’. In practice, this means that prime ministers would be willing to allow their MEPs to oppose a candidate that they had already supported in the European Council, and face the wrath of the national media and opposition parties as a consequence of this volte face. Without this incentive, and this practical consequence, the ‘prime ministerial coalition’ in the European Council will always prevail on the ‘European party coalition’ in the EP.

Until this transformation from national to European level party goals, which I believe is a long way off, the Commission President investiture procedure, as set out in the Maastricht and Amsterdam Treaties, will do little to increase the accountability of the holders of EU executive power. Nevertheless, if one accepts that this outcome was a function of the institutional rules - that the outcome is a ‘structure-induced-equilibrium’ (Shepsle, 1989) - then further institutional reform may rectify this situation.

In terms of the institutional design of executive selection, the EU has two basic options. The first would be to stick with a parliamentary model of executive selection, but to allow EP party groups to exercise a genuine choice about the candidate for Commission President. This would require breaking the bond between domestic parties and EP party groups, by preventing national parties from controlling the selection of MEPs. For example, the EP party groups, the national delegations of MEPs from each party, the transnational party federations, or local and regional party organisations would be responsible for selecting candidates in EP elections, and national party head-offices would be actively prevented from influencing the candidate-selection process. This could, perhaps, be imposed through a legally-enforceable clause in a uniform electoral procedure for EP elections.


Alternatively, the bond between national parties and EP party groups could be kept, but the vested interest of national parties in the EP vote could be removed. For example, the EP could be allowed to choose from a list of Commission President candidates that has been approved by the European Council, and the European Council could subsequently vote on the EP’s choice. Either of these reforms would restrict the ability of the parties in government, in the European Council, to impose their wishes on their MEPs in an EP investiture vote. Without this interference, the EP party groups could act cohesively to transform the majority in EP elections into a majority vote for or against a Commission President. In other words, the procedure would operate as a genuine parliamentary model of executive selection, as the academic consensus envisages.

A fundamental problem, however, is that even if these reforms were introduced, EP elections would still be fought as ‘national contests’. Regardless of whether the EP party groups are cohesive, voters would still make their choices in EP elections based on competition for national executive office, even with a uniform electoral procedure. Despite the development of some ‘first-order’ characteristics in EP elections in Denmark and France, the probability of a transformation of EP elections in all member states into a contest about rival candidates for the Commission President is close to zero.

A second basic option, therefore, would be to abandon the parliamentary model and transform the EU into a (partial) presidential system, by introducing some form of direct election of the Commission President - such as a two-ballot contest (along the lines of the French model), or via an electoral college (along the lines of the American model).(5) In such an election, it would be much harder for political parties to fight a campaign on the performance of the national governments, because the electoral choices on offer would be very different from the national and EP contests. Each national party would be forced by the national media to ‘come out’ in favour of one or other candidate. Moreover, this non-second order dynamic could be enhanced through specially designed rules for nominating candidates for the Commission Presidency. For example, each candidate could be required to be supported by one party (or group of MPs and MEPs) in every member state. This would guarantee that one section of the elite in every member state would be responsible to their electorates for the actions of the winning candidate (cf. Laver et.al., 1995). In practice, this would probably mean that each European party family would put forward a separate candidate. For example, if this were the case in 1994, the Socialists, with the approval of the PES party leaders’ summit, would probably have backed Felipé González, whereas the Christian Democrats, with the approval of the EPP party leaders’ summit, would have backed Jean-Luc Dehaene or Ruud Lubbers. Of course, the result of such an election would be a significant politicisation of the post of Commission President, which the national governments probably do not want. Nevertheless, a politicisation of executive power, whether via a parliamentary or presidential model of executive selection, is a necessary prerequisite for establishing political accountability.

Overall, the exercise of political power at the European level is increasingly questioned by Europe’s citizens. Part of the problem is that the current institutional framework, including the reforms of the Maastricht and Amsterdam Treaties, does not allow voters to decide who holds (or does not hold) executive office in the EU. This can only be rectified by further institutional changes, either to give the parliamentary model a real chance of working or to introduce an element of presidentialism into the system. Either way, partisan competition, between EP party groups or the transnational party federations, rather than national competition, between the governments in the European Council, is the only way Europe’s citizens will be able to choose ‘who governs’ at the European level.


Attiná, F. (1990) ‘The Voting Behaviour of the European Parliament Members and the Problem of Europarties’, European Journal of Political Research, 18/3, 557-79.

Bardi, L. (1996) ‘Transnational Trends in European Parties and the 1994 Elections of the European Parliament’, Party Politics, 2/1, 99-114.

Bowler, S. & Farrell, D.M. (1993) ‘Legislator Shirking and Voter Monitoring: Impacts of European Parliament Electoral Systems upon Legislator-Voter Relationships’, Journal of Common Market Studies, 31/1, 45-69.

Brzinski, J. Bay (1996) ‘Political Group Cohesion in the European Parliament, 1989-1994', in C. Rhodes & S. Mazey (eds) The State of the European Union, Vol. 3, London: Longman.

Chryssochoou, D.N. (1994) ‘Democracy and Symbiosis in the European Union: Towards a Confederal Consociation’, West European Politics, 17/1, 1-14.

Corbett, R., Jacobs, F. & Shackleton, M. (1995) The European Parliament, 3rd edn, London: Catermill.

Council of the European Union (1997) Consolidated Version of the Treaty Establishing the European Community, CONF 4005/97 ADD 2.

Cox, G.W. & McCubbins, M.D. (1993) Legislative Leviathan: Party Government in the House, Berkeley: University of California Press.

Curtice, J. (1989) ‘The 1989 Elections: Protest of Green Tide?’, Electoral Studies, 8/3, 217-30.

Downs, A. (1956) An Economic Theory of Democracy, New York: Harper & Row.

Eijk, C. van der & Franklin, M. (eds) (1996) Choosing Europe? Ann Arbor: University of Michigan Press.

European Council (1996) Turin European Council, 29 March 1996, Presidency Conclusions, DOC/96/2.

Falkner, G. & Nentwich, M. (1997) ‘The European Parliament - Winner or Loser of the Amsterdam Treaty? A Preliminary Assessment’, paper presented at the International Political Science Association seminar on ‘Amsterdam and Beyond’, 9-12 July 1997, Brussels.

Gabel, M. & Hix, S. (1997) ‘The Ties that Bind: The European Parliament and the Commission President Investiture Procedure’, unpublished mimeo.

Hix, S. & Lord, C. (1996) ‘The Making of a President: The European Parliament and the Confirmation of Jacques Santer as the President of the Commission’, Government & Opposition, 31/1, 62-76.

Hix, S. & Lord, C. (1997) Political Parties in the European Union, London: Macmillan.

King, A. (1981) ‘What do Elections Decide?’, in D. Butler, H.R. Penniman & A. Ranney (eds) Democracy at the Polls, Washington, D.C.: American Enterprise Institute.

Laver, M.J., Gallagher, M., Marsh, M., Singh, R. & Tonra, B. (1995) Electing the President of the European Commission, Trinity Blue Papers in Public Policy: 1, Dublin: Trinity College.

Lenaerts, K. (1991) ‘Some Reflections on the Separation of Powers in the European Community’, Common Market Law Review, 28/1, 11-35.

Lijphart, A. (1969) ‘Consociational Democracy’, World Politics, 21, 2, 207-25.

Lijphart, A. (1992) Parliamentary Versus Presidential Government, Oxford: Oxford University Press.

Lodge, J. (1995) The 1994 Elections to the European Parliament, London: Pinter.

Mayhew, D. (1974) Congress: The Electoral Connection, New Haven: Yale University Press.

Nentwich, M. & Falkner, G. (1997) ‘The Treaty of Amsterdam: Towards a New Institutional Balance’, European Integration online Papers, 1, 15, http://eiop.or.at/eiop/texte/1997-015a.htm.

Noël, E. (1992) ‘Reflections on the Maastricht Treaty’, Government & Opposition, 27, 2.

Quanjel, M. & Wolters, M. (1993) ‘Growing Cohesion in the European Parliament’, Paper presented at the Annual Joint Sessions of the European Consortium for Political Research, April 1993, Leiden.

Raunio, T. (1996) Party Group Behaviour in the European Parliament, Tampere: University of Tampere.

Reif, K.-H. & Schmitt, H. (1980) ‘Nine Second-Order National Elections: A Conceptual Framework for the Analysis of European Election Results’, European Journal of Political Research, 8/1, 3-45.

Reif, K.-H. (1984) ‘National Election Cycles and European Elections, 1979 and 1984’, Electoral Studies, 3/3, 244-55.

Riker, W.H. (1962) The Theory of Political Coalitions, New Haven: Yale University Press.

Sartori, G. (1992) Comparative Constitutional Engineering, London: Macmillan.

Schattschneider, E.E. (1960) Semisovereign People: A Realist’s View of Democracy in America, Hinsdale: Dryden Press.

Schumpeter, J. (1943) Capitalism, Socialism and Democracy, London: Allen & Unwin.

Shepsle, K.A. (1989) ‘Studying Institutions: Some Lessons from Rational Choice’, Journal of Theoretical Politics, 1/2, 131-47.

Shugart, M.A. & Carey, J.M. (1992) Presidents and Assemblies, Cambridge: Cambridge University Press.

Smith, J. (1996) ‘How European are European Elections?’, in J. Gaffney (ed.) Political Parties and the European Union, London: Routledge.

Strom, K. (1990) ‘A Behavioural Theory of Competitive Political Parties’, American Journal of Political Science, 34/2, 565-98.

Taylor, P. (1991) ‘The European Community and the State: Assumptions, Theories and Propositions’, Review of International Studies, 17/1, 109-25.

Weber, M. (1942 [1918]) ‘Politics as a Vocation’, in H.H. Gerth & C. Wright Mills (eds) From Max Weber: Essays in Sociology, Oxford: Oxford University Press.

Weiler, J.H.H., Haltern, U.R. & Mayer, F.C. (1995) ‘European Democracy and Its Critique’, J Hayward (ed.) The Crisis of Representation in Europe, London: Frank Cass.


(*) A previous version of this paper was presented to the IPSA workshop on 'Amsterdam and Beyond' in the European Parliament Brussels, 10-12 July 1997.

(1) There are also several hybrid systems: for example, where the head of the executive is directly-elected, but the cabinet requires parliamentary approval (e.g. the semi-presidential or directly-elected-premier models); or were the executive is approved by parliament but is not required to command a permanent majority (e.g. the Swiss collegial-executive model).

(2) Article 207 (ex Article 151) of the Draft Amsterdam Treaty states that: ‘the Council shall define the cases in which it is to be regarded as acting in its legislative capacity, with a view to allowing greater access to documents in those cases’ (Council of the European Union 1997)

(3) Strictly speaking, the reforms established a particular form of parliamentary model, whereby the executive requires the support of a parliamentary majority only upon investiture but not through its whole term in office. This is hence more similar to the Swiss collegial-executive system than the classic parliamentary models that exist in most national settings in the European Union.

(4) A large proportion of the other 4% is explained by the nine Portuguese Socialists who voted for Santer despite the fact that their party was in opposition in Lisbon. An interview with one of these members suggested that they were concerned that a negative vote would postpone EU enlargement to Austria, Sweden and Finland - scheduled for 1 January 1995 - and that such a postponement would jeopardise the agreement by the potential new members to make a significant contribution to the EU Cohesion Fund, of which Portugal is a large net beneficiary.

(5) I use the word ‘partial’ to reiterate that executive power would still be shared with the national governments (that the EU would have a ‘dual executive’), and hence that the consociational (non-majoritarian) part of the EU executive would remain in tact, to counterbalance the natural majoritarian tendency of a direct election of the Commission President.

©1997 by Hix
formated and tagged by MN, 12.11.1997