hapter 3 Legislative Politics Theories of Legislative Coalitions and Organizalion Development of the Legislative System of the EU Legislative Politics in the European Parliament Legislative Politics in the Council Bicameral Politics in the EU Summary: A Bicameral System The EU has a two-chamber legislature in which the Council represents the states and the European Parliament represents the citizens. The Council is more powerful. However, the introduction, revision and extension of the co-decision procedure to cover most legislative areas have moved the European Parliament towards parity with the Council. Although the main actors in the Council are governments and those in the European Parliament are political parties, internal politics and organization of the two chambers share some similarities. To help us understand how the system works, we first look at some general theories of legislative behaviour and organization. Theories of Legislative Coalitions and Organization The field of legislature studies in political science has generated many different theories of legislative bargaining, coalition formation and organization. One of the earliest was Riker's (1962) expectation that politicians are likely to form ;minirnum-winning-coalitions' and not 'surplus coalitions'. Riker argued that coalitions are unlikely lo include any groups that are not necessary for reaching a majority, as in a smaller coalition, fewer interests need to be appeased. But, an actor that is decisive (pivotal) for the formation of a winning coalition can demand a high price in return for participating. The more likely an actor is to be pivotal, the more 'power' he or she has in coalition bargaining (Shapley and Shubik, 1954; 58 The Political System of the European Union Banzhaf, 1965). However, if actors also care about policy outcomes as well as power, then 'minimum-connected-winning-coalitions' between legislators with similar policy preferences are more likely than coalitions between actors with radically different policy preferences (Axelrod, 1970), The theoretical foundation behind these ideas is the spatial model of politics (e.g. Hinich and Munger, 1997). In this framework, actors and policies can be represented as points in a policy space. The actors prefer to minimize the distance between their position (their ideal point) and the adopted policy. Consider a simple model with five actors, A, B, C. D and E, where the ideal point of A is to the left of B, whose ideal point is to the left of C, and so on, as illustrated in Figure 3.1. If a decision is made by a simple majority and all actors are allowed to make proposals, the outcome will be at C. A and B will not be able to convince D or E to move the policy further to the left, nor will D and E be able to convince A or B of moving the policy further to the right. However, if the policy is to the right (left) of C, she will be able to convince A and B (D and E) to move the policy to C, as this move will make them better off. This is known as the 'median voter theorem' (Hotelling, 1929; Black, 1958). Nevertheless, the outcome might be different if an oversized majority is required. In our example, if a majority of 4 out of 5 is required, any policy between B and D is impossible to move. The median voter theorem is a useful approximation of bargaining in a single policy dimension in a relatively institution-free environment. However, many policies are multidimensional. Where multiple issues are up for discussion at the same time, it is more difficult to find a single policy package that cannot be defeated by an alternative package. This is because in a multidimensional policy space it will always be possible to find an alternative combination of policies that can defeat a policy that a majority has agreed (McKelvey, 1976; Schofield, 1978). Nevertheless, this sort of policy instability is rare. The main reason for this is that legislative institutions generate what is known as 'structured-induced-equilibria': policy outcomes which are stable because of institutional rules (Shepsle, 1979; Riker, 1980). One set of rules are agenda-setting and veto rights (Romer and Rosenthal, 1978; Tsebelis 2002). For example, in Figure 3.1, if actor E is the sole agenda-setter _A_B_C_T)_E_ Figure 3.1 Legislative bargaining in one dimension Note: The figure shows the spatial location of five actors, labelled A, B, C, D and E, in a single-dimensional space. Legislative Politics 59 (with the power to make a proposal), and the existing policy (the status quo) is at B, then E can use her agenda-setting power to move the policy all the way to D, since C will be indifferent between having the policy at B and D. However, if A has veto power, it will not be possible to move the policy any further away from A. For actors interested in bringing about policy change, it is better to be an agenda-setter than a veto player, as the agenda-setter may be able to move the policy towards her, while a veto player can only prevent a policy from moving further away from her ideal point. However, for a committed conservative actor, it may be better to be a veto player as she can simply block any proposals, and not invest time and resources in drafting policy proposals. A second set of rules relates to specialization in the legislative process, for example in parliamentary committees. If agenda-setting power on specific policy areas is granted to committees, issues may be prevented from being linked across policies, which can then lead to stable policy outcomes (Shepsle and Weingast, 1987). Another motive for specializing in committees is the need for technical expertise and information in the legislative process. Legislators are often uncertain about the precise relationship between the policy instrument at their disposal and the final policy outcome. There is, hence, an incentive for legislators to grant agenda-setting rights to subsets of legislators on specific policy-fields in return for more precise knowledge about this relationship (Krehbiel, 1991). In turn, this need for information provides opportunities for interest groups to influence committee members (Crombez, 2002). Third, political parties also facilitate legislative stability by simplifying legislative bargaining (e.g. Aldrich, 1995; Cox and McCubbins, 2007). Because an individual legislator is unlikely to obtain her policy objectives by acting alone, and issue-by-issue co-ordination is costly, formal relationships that bind individual legislators together reduce the transaction costs associated with coalition formation. With uncertainty about other legislators' preferences and the impact of legislative decisions, legislators with similar policy preferences benefit from institutional arrangements that facilitate information gathering and development of policy expertise. The result is a division of tasks: backbench MPs provide labour and capital, while party leaders distribute committee and party offices and determine the party line on complex legislative issues. But, if the parties who control the executive (in government) do not have the power to dissolve the legislature, and so hold new legislative elections, their ability to enforce discipline amongst their legislative troops is limited (Huber 1996; Hix and Noury 2016). Fourth, legislative stability is also facilitated by the existence of multiple legislative chambers. In bicameral systems, for example, coalitions in both chambers have to be in favour of a proposal before it can become law. This restricts the 60 The Political System of the European Union set of possible policy choices and simplifies legislative bargaining (Riker. 1992; Tsebelis and Money, 1997). The legislative procedures regulating the interaction between the two chambers determine agenda and veto rights, thus determining the relative strengths of the two chambers in the bicameral bargaining (McCarty, 2000). Together, these theories help us understand EU legislative politics: what coalitions are likely to form, why the European Parliament and Council are organized in the way they are, and who is more powerful under the EU's legislative procedures. Development of the Legislative System of the EU The rules of the EU legislative process have evolved considerably since the Treaty of Rome established that legislation would be adopted through interaction between the Commission, the Council and the European Parliament. The Treaty of Rome did not set out a single procedure to govern this interaction. Instead, each article of the treaty specified what voting rule would be used in the Council - either unanimity or qualified-majority voting (QMV), a system of weighted voting - and whether or not the European Parliament should be 'consulted' by the Council. The right of the Council to make decisions by QMV was challenged in the mid-1960s, when the French President Charles de Gaulle objected to majority voting being used in a number of important areas. De Gaulle insisted that every member state should be able to veto legislation, even when the treaty specified that QMV could be used. In 1965 he provoked the 'empty chair' crisis by refusing to allow his ministers to participate in Brussels' business until the other member states accepted his position. The crisis was resolved in 1966 by the so-called 'Luxembourg compromise', which established the principle that if a member state declared that a 'vital national interest' was at stake, the Council should make every effort to reach a unanimous agreement. The Luxembourg compromise was not legally binding, as it was not included in the treaties, but it nonetheless ushered in nearly two decades of de facto unanimous agreement in the Council. Any member state faced with being outvoted on a key issue could invoke the Luxembourg compromise and threaten a veto. Nevertheless, by the early 2000s the EU legislative system had developed into something much closer to a traditional bicameral model. The first major development was the 1980 'isoglucose' ruling by the CJEU. During the first direct elections of the European Parliament in 1979, the Council had adopted a piece of legislation without consulting the European Parliament. The European Legislative Politics 61 parliament challenged the Council before the CJEU, and the court annulled the legislation on the grounds that the treaty required the Council to 'consult' the European Parliament. In the CJEU's opinion this meant that the Council could not act until the European Parliament had formally issued an opinion on a piece of legislation. This did not mean that the European Parliament could force its opinions on the Council, but it did mean that the European Parliament now had a 'power of delay'. A power of delay is not as strong as an agenda-setting or a formal veto power. Nevertheless, Kardasheva (2009) found that this power of delay did enable the European Parliament to substantively influence EU legislative outcomes under the 'consultation procedure'. The legislative powers of the European Parliament were significantly increased by four subsequent treaty reforms. First, in 1987 the Single European Act (SEA) introduced a new legislative procedure: the co-operation procedure This was the first procedure to be set out in a separate treaty article, to which other treaty articles referred. The procedure allowed the European Parliament a second reading, after the Council had adopted a common position, and reduced the ability of the Council to overturn European Parliament amendments made in the second reading. The SEA applied this procedure to only ten treaty articles, but these included most areas of the single market programme, plus some aspects of research and regional spending social and environmental policy. Together these constituted approximately one-third of all legislation. The SEA also introduced the 'assent procedure', whereby the approval of the European Parliament was required before the Council approved an association agreement with a non-European Community state or the accession of a new member state. Second, in 1993 the Maastricht Treaty introduced a fourth legislative procedure, the cd-decision procedure, which was also set out in a separate treaty article. This procedure introduced the rule that if the European Parliament and Council disagreed on a piece of legislation a conciliation committee would be convened, consisting of an equal number of representatives of the European Parliament and the Council. After a conciliation committee had reached an agreement, the deal would then have to be approved by both the Council and the European Parliament. The co-decision procedure originally applied to most areas of the internal market legislation that had previously been covered by the co-operation procedure, plus public health, consumer protection, education and culture. Third, in 1999 the Amsterdam Treaty reformed and extended the co-decision procedure. Under the new version of the procedure, legislation could be adopted at first reading if the European Parliament and Council already agreed at this stage. Furthermore, the conciliation committee became the last stage of the legislative process. Also, if the conciliation committee failed to reach an 62 The Political System of the European Union agreement, there would be no legislation. These reforms increased the power of the European Parliament within the procedure. The treaty also extended the procedure to most areas previously covered by the co-operation procedure. As a result, the version of the co-decision procedure established by the Maastricht Treaty is often referred to as 'co-decision I' while the Amsterdam Treaty version is called 'co-decision II'. Fourth, the 2009 Lisbon Treaty established the Amsterdam version of the co-decision procedure as the 'ordinary legislative procedure' and significantly extended its use to almost all areas of EU law. The Lisbon Treaty also replaced the assent procedure with the 'consent procedure', which gave the European Parliament a formal right of veto over agreement with a third country, such as an association agreement or a free-trade agreement. Despite the complexity of the EU's decision-making rules and the expanding policy agenda of the EL), the EU legislative system had adapted well. Although the involvement of the European Parliament has slowed down the legislative process, this effect has been offset by the increased use of QMV in the Council (Golub, 1999; Schulz and König, 2000; König, 2007). The other reason for the efficiency and effectiveness of the EU legislative system is that both the Council and the European Parliament have developed sophisticated strategies to maximize their influence vis-ä-vis each other in the various stages of bicameral bargaining. Legislative Politics in the European Parliament The European Parliament holds most of its plenary sessions in Strasbourg, and part of its secretariat is in Luxembourg. However, the bulk of the work of the Parliament is in Brussels, where some plenary sessions are held, the political groups and committees meet, and the offices of the MEPs, the political groups and the committees are based. The European Parliament operates like any other legislature: organizing and mobilizing to influence EU legislation and the EU executive. The institutional design of the EU - the separation of executive and legislative powers - means that the European Parliament is more similar to the US Congress than to national parliaments in Europe. There is no EU government that relies on the permanent support of a majority in the European Parliament, and neither the Commission nor the Council can dissolve the European Parliament, as the five-year electoral cycle of the Parliament is fixed in the treaty. The European Parliament can 'censure' the Commission, but this procedure is closer to a presidential style impeachment procedure than a vote of non-confidence in a government in a parliamentary system (see Chapter 2). As a result, the European Parliament is a relatively independent legislature, free Legislative Politics 53 to amend legislation proposed by the Commission and agreed by the Council, and the Commission and the Council must build coalitions in the Parliament issue-by-issue. MEPs: Agents with Two Principals The European Parliament is the only directly elected body of the EU. However, the link to the voters is weak. Because European Parliament elections are fought on national rather than European issues (see Chapter 6), and because few voters know anything about the MEPs, there is little an MEP can do to improve her chances of re-election. Moreover, for most MEPs the chance of standing as a candidate and being re-elected is determined not by her political group in the European Parliament or her behaviour on behalf of her voters, but by her party leaders 'back home', who control who gets to stand as a candidate in European Parliament elections (Bowler and Farrell, 1993; Farrell and Scully, 2007). As a result of the way European Parliament elections work, MEPs' behaviour is hence driven less by external re-election incentives than by career incentives and policy objectives inside the European Parliament. Traditionally, a career in the European Parliament was considered either as a training ground for a job in national politics or as a'retirement home' at the end of a national career. However, as the power of the European Parliament has grown, more MEPs have started to consider a political career in Brussels in its own right (Scarrow, 1997; Meserve et a!., 2009; Daniel, 2015; Pemstein et at, 2015; Hoyland et al„ 2019). MEPs intent on developing their career inside the European Parliament have two types of goal: • office - such as promotion to party leadership, a committee chair or vice-chair, or a position in the leadership of the parliament, such as president or vice-president; and • policy - pursuing their own policy preferences or interests of their constituency through influencing EU legislation or scrutinizing the Commission and Council. Whereas re-election is usually not dependent on MEP performance inside the European Parliament, the ability to achieve these goals is dependent upon gaining promotion within the European Parliament's committees and the political groups, and on being able to form coalitions with other legislators to secure common policy aims. So, MEPs have two principals they must respond to: to secure re-selection as a candidate and re-election they must cater to national party leaderships; but 64 The Political System of the European Union Legislative Politics to secure promotion within Ihe European Parliament and policy outputs via EU legislation MEPs must cater to the most powerful politicians inside the European Parliament: their political group leaderships. Most of the time these two principals are not in conflict. But, if these principals ever are in conflict, MEPs tend either to abstain in a roll-call vote (Muhlbock and Yordanova, 2017) or to side with their national parties against their European political groups In particular, MEPs tend to vote more against their European political groups if they have been elected under a 'closed' form of PR in European Parliament elections, which gives national parties more power to influence which MEPs will get elected (see Chapter 6) (Hix, 2004), if they intend to return to national political office after the European Parliament (Meserve et al., 2009; Hoyland et al., 2019), and in the build-up to national or European Parliament elections, when national parties pay more attention to the behaviour of 'their' MEPs in the European Parliament (Lindstadt et al. 2011; Koop et al., 2018) MEPs who plan to vote against the political groups often choose to speak in a plenary debate, to explain their national party position to the European political group (Slapin and Proksch, 2010). Agenda Organization: Leaderships, Parties, Committees and Rapporteurs The European Parliament determines its own organization and writes its own rules. These rules are formalized in the European Parliament rules of procedure. The rules of procedure establish three main organizational structures to facilitate agenda control: the parliamentary leadership, political groups and the committee system. Regarding parliamentary leadership, the most senior offices in the European Parliament are the president and the fourteen vice-presidents, whose main responsibility is to chair the plenary sessions. There are also three leadership bodies: the Bureau of the Parliament (consisting of the president and the vice-presidents); the Conference of Presidents (consisting of the president, the leaders of the political groups, and the chairman of the Conference of Committee Chairmen); and the Conference of Committee Chairmen. Together these bodies involve all the senior figures in the European Parliament. The Bureau deals with the internal organizational and administrative matters, but is also active on political issues and meets almost every week. The Conference of Presidents is where most political issues are tackled, particularly with regard to the relationship between the European Parliament and the Commission and Council. It normally meets twice a month. The Conference of Committee Chairmen co-ordinates the committee agendas and decides which committee is responsible for which legislative dossier. Second, the European Parliament political groups are the central mechanisms for structuring debate and coalition formation in the legislative process (Hix and Lord, 1997; Raunio, 1997; Kreppel, 2002a). The rules of procedure set out how many MEPs are needed to form a party group: at least twenty-three MEPs from at least one-quarter of all member states (currently seven member states). The political groups have certain privileges, such as secretarial and research staff as well as financial resources. Table 3.1 shows the size of the political groups and the national memberships of the groups in March 2021 (just after the Hungarian Fidesz party left the EPP). The European People's Party (EPP), on the centre-right, and the Socialists and Democrats (S&D). on the centre-left, are the two dominant groups. However, neither of these groups commands more than 25 per cent of the seats, and so they each need several other groups to secure a winning majority. The political groups are important because their leaderships determine vital political issues in the European Parliament: such as the choice of the European Parliament president, the allocation of committee positions and legislative reports {rapporteurships), the agenda of plenary sessions, and the policy positions of the political groups. If a national party is not a member of a political group, it is unlikely to secure any office or policy goals for its MEPs. When deciding which political group to join, a national party takes two things into account: power - how influential they will be in the parliament as a whole, so being in a larger group is usually better, and policy - how aligned they are with the other members of a group in policy terms (McElroy and Benoit 2010; Hix and Noury, 2018; Martin, 2021). Over time, the power incentives have led to parties leaving the smaller groups to join the EPP or S&D groups. But power and policy incentives have also led to splits in the groups. For example, in 2009 the British and Czech conservative parties left the EPP because they increasingly found themselves voting against the pro-European and more corporatist policy positions of the EPP (Hix and Noury, 2009; Heyland, 2010). In general, individual MEPs try to avoid upsetting their political group leaderships. The party line is enforced by 'party whips', who issue voting instructions to the group members and monitor whether MEPs follow these instructions (Bailer et al, 2009). The result is a high level of group cohesion, despite the growing national and political heterogeneity of the Parliament after successive EU enlargements (Faas, 2003; Hix et al., 2005; Hix and Noury, 2009). Nevertheless, within the groups, the national party delegations remain powerful, with the larger national delegations dominating the key leadership positions inside the groups. Table 3.1 Membership of the political groups in March 2021 Member state EPP S&D Renew ID G/EFA ECR Left NA Total Belgium 4 3 4 3 3 3 1 Bulgaria / b 3 2 17 Czechia 5 6 2 3 4 I 21 Denmark 1 3 6 1 2 1 14 Germany 29 16 7 25 1 :> 2 96 Estonia 1 2 3 1 7 Ireland 5 2 2 ' 13 Greece 8 2 1 6 4 21 Spain 13 21 9 3 4 6 3 59 France 8 6 23 23 13 6 79 Croatia 4 4 I 1 2 12 Italy 8 18 2 27 4 7 10 76 Cyprus 2 2 2 6 Latvia 2 2 1 1 2 8 Lithuania 4 l 1 2 1 1 i ixembourg 2 I 2 1 6 Hungary 1 5 2 13 21 Malta 2 4 6 Netherlands 6 6 7 1 3 4 1 1 29 Austria 7 5 1 3 3 19 Poland i7 7 1 27 52 Portugal 7 9 1 4 21 Romania 14 10 8 1 33 Slovenia 4 2 2 8 Slovakia 5 3 2 2 2 14 Finland 3 2 3 2 3 1 14 Sweden 6 5 3 3 3 1 21 Total 175 1-15 98 74 73 63 39 38 705 % 24.8 20.6 13.9 10.5 10.4 8.9 5.5 5.4 Source: European Parliament. Key: ECR European Conservatives and Reformists (national conservatives) EPP European People's Party (centre-right) G/EFA Greens/European Free Alliance (greens/left regionalists) ID Identity and Democracy (Eurosceptics/populist right) Left The Left in the European Parliament (radical left) NA non-attached (independents) Renew Renew Europe (liberals) S&D Socialists and Democrats (social democrats) 68 The Political System of the European Union Legislative Politics 59 The third organizational structure of the European Parliament is the committees. It is in the committees that the real scrutiny of EU legislation takes place. The committees propose amendments to legislation in the form of reports and draft resolutions, which are then submitted to the full plenary session. Amendments to the proposed committee resolutions can be made in the full plenary, but without the backing of a committee and a political group amendments are unlikely to be adopted by the Parliament. In line with the predictions from the informational theory of legislative organization (Krehbiel, 1990,1991), MEP membership of committees and the election of MEPs to senior positions in committees (chairs and vice-chairs) are largely representative of the European Parliament as a whole in terms of political preferences and nationality (Whitaker, 2005). Nevertheless, previous committee experience and policy expertise also matter when committee memberships are assigned (McElroy, 2006; Yordanova, 2009), and in terms of promotion to senior positions, such as chairs and committee chairs (Whitaker, 2011). Although, chairing any committee in the Parliament seems to be a stronger determinant of who becomes a committee chair than either length of time in a particular committee or voting loyalty to a political group (Chiru, 2020). Within committees, legislative reports are drafted by rapporteurs The allocation of these positions within committees is in general proportional between political groups, although larger national party delegations tend to write proportionally more reports than smaller national party delegations. At the individual MEP level, a small minority of highly active MEPs write the majority of reports (Mamadouh and Raunio, 2003; Yoshinakaetal., 2010). It is not clearto what extent the political group leaderships are able to use the allocation of reports as a disciplining tool, given the role national parties play in influencing voting behaviour (Kreppel, 2002a). However, MEPs who oppose political group and national party leaderships are unlikely to write important reports (Hausemer, 2006; Yoshinaka et al, 2010). MEPs' background, experience and links with interest groups do influence who within the different groups end up writing reports (Kaeding, 2004. 2005, Chiou et al., 2020). MEPs from national parties in government (who sit in the EU Council) are also more likely to win rapporteurships (Hoyland. 2006). Rapporteurs and other influential committee members are the target of lobbyists who trade information for influence at the committee stage (Marshall, 2010, see also Chapter 7). When preparing a report, a rapporteur has to trade-off the interest of her party, advice from lobbyists and suggestions from other committee members, in order to get the report adopted by both the committee and the plenary (Benedetto, 2005). But. them is evidence that rapporteurs are able to shape legislative outcomes in their preferred direction, particularly when agreement is reached between the European Parliament and the Council in the first reading of the co-decision procedure (see below) (Costello and Thomson, 2010). Coalition Formation There is no permanent coalition in the European Parliament, and without a government to support, legislative coalitions are formed formally or informally issue-by-issue and even vote-by-vote. On many issues the European Parliament behaves as if it were a single actor seeking to promote its own powers and interests against the interests of the second legislative chamber in the EU (the Council) or against the holders of executive power (the Commission). But, Hagemann and Hayland (2010) find that the Parliament is less able to present a united front if the Council is also divided on a particular legislative issue, as a national split in the Council can undermine the cohesion of the main political groups in the Parliament. An informal 'grand coalition', between the centre-left S&D and centre-right EPR is also facilitated by the relative sizes of the groups in the chamber, as neither an S&D-led centre-left coalition nor an EPP-led centre-right coalition has been able to command a majority of seats in the last few Parliaments. The rules of the EU legislative process also encourage a grand coalition. In the adoption of opinions in the early stages of the legislative procedures, when voting on 'own initiative' reports (non-binding invitations to the Commission to propose legislation on a certain issue) and when adopting amendments to resolutions on legislation, the European Parliament decides by a simple majority of those present at the vote. In contrast, in the second reading under the ordinary legislative (co-decision) procedure an absolute majority of all MEPs. not just of those turning up to vote, is required to propose amendments. This encourages co-operation between the two largest political groups throughout the legislative process (Kreppel and Hix 2003). Voting in the European Parliament is by show of hands, electronic vote, or by roll-call vote. The latter form occurs on all final votes on legislative dossiers and on any other issue if requested by a political group or by at least thirty-eight MEPs (5 per cent). Political groups that do not support a legislative proposal from a committee are more likely to request a rollcall vote in plenary (Thierse, 2016). This means that roll-call votes on individual amendments tend to be on the more contentious and high-profile issues. That said, political groups can also strategically use roll-call votes to monitor their behaviour of the members or to illustrate that another political group is not cohesive on a key issue (Carrubba et al., 2006; Hug 2016). How individual MEPs vote in a roll-call vote is recorded in the minutes and is now available on the internet (at www.VoteWatch.eu). The records of these roll-call votes have led to a large body of empirical research on voting behaviour and coalition formation in the European Parliament (Kreppel, 1999; Hix et al, 2005, 2007; Lindstadt et al, 2012). This research reveals that in roll-call votes, MEPs vote more along transnational party lines than national lines, that different coalitions form on different issues and that the two main political groups do not always vote together. 70 The Political System of the European Union Figure 3.2 shows the voting patterns in the 2009-14 and 2014-19 parliaments in a two-dimensional space, as estimated from voting behaviour of MEPs in all the roll-call votes in these periods. Note that most MEPs are tightly clustered together with other MEPs from the same political group, which reflects the high I ' s « G/££? A'' L \ ■*'A w G/EFA *.NGL ENF : ia Left-Right hj u , L >1w E E c ,c Eg c NA " co co K 5 .P o Q. oj s £ S m cm ro o cm oj cm cm -§ Q- 2i, o cd co co co t- t; c cm cq co co co ro en is -o .í? co c cr power For example, in the negotiations on the Lisbon Treaty, the Spanish and Polish prime ministers initially refused to support the move to double majority, as this would significantly reduce their bargaining power relative to the voting weights they had under the existing QMV rules. Nevertheless, in general, compared to the pre-Lisbon QMV rules, the shift to a double majority increased the bargaining 'power' of the largest member states, and Germany in particular, relative to the power of the medium-sized and smallest member states. As the table shows, Brexit also affects the relative voting power balance in the Council, by moderately increasing the power of the larger member states further relative to the smaller member states (Gabon 2020). Despite the formal QMV rules and the battles over voting powers, coalitions in the Council are not formed randomly, but between member states with similar economic interests in EU politics (such as net contributors versus net beneficiaries from the EU budget), or similar geographic and cultural connections (Carrubba, 1997; Mattila, 2004; Zimmer et at, 2005; Naurin and Wallace, 2008; Thomson, 2009, Venn 2011). The Franco-German coalition has been at the heart of Council decision-making since the 1950s. The Benelux countries are more economically and politically integrated than any other grouping in the EU, and there are other emerging alliances, for example between the 'Visegrad 4' (Poland, Hungary, the Czech Republic and Slovakia), the 'Nordics' (Sweden, Denmark and Finland) or the 'Southern bloc' (Italy. Spain, Portugal and Greece) (Naurin and Lindahl, 2008). Equally, some member state governments are better connected than others. For example, while in the mid-2000s, the UK was on the losing side more than any other member state in recorded votes, Naurin and Lindahl (2010) found that the UK was the member state mentioned most by officials from other countries when asked 'Which member states do you most often co-operate with in order to develop a common position?' Also, although the Council now uses QMV to adopt legislation in most policy areas, decisions are often unanimous (Hayes-Renshaw and Wallace, 2006). This search for unanimous decisions is labelled the culture of consensus' (Heisenberg, 2005). For example, during the first two years after the Eastern and Southern enlargement in 2004, 90 per cent of all Council decisions were unanimous (Mattila, 2009). Nevertheless, Council decisions are often made in the 'shadow of a vote', where opposing governments know that they will be outvoted, and so refrain from recording a negative vote, as their opposition would not change the outcome (Golub, 1999). Finke (2017) offers an alternative explanation: that governments use the threat of voting against the other member states earlier in negotiations to secure consensual agreements, but these threats are only credible if the government 78 The Political System of the European Union faces opposition from a coalition partner or in their national parliament. Similarly, Hagemann et al. (2017) find that governments are more likely to vote against the majority in the Council when their domestic electorates are more Eurosceptic or when the EU is salient in domestic party politics. Meanwhile, van Guisen et al (2019) find that a government is more likely to vote with the majority in the Council if it held the presidency, or was one of the Trio of presidencies during the time the legislation was being negotiated. And, consistent with the idea that bargaining in the Council is policy-specific, Hayland and Woien Hansen (2014) find that governments' specific policy positions - on environmental or social policies, for example - are better predictors of their voting behaviour in the Council than their general left-right or pro-/anti-EU positions. Despite the incentives to reach consensus, the analysis of voting in the Council may offer insights into coalition patterns in the Council (Hagemann and Hoyland, 2008). Figure 3.3 shows the estimated positions of member states in the Council on the basis of their voting behaviour between 2009 and 2019. In general, in this period, the voting divisions in the Council were driven more by member state interests than by the party-political make-up of the governments. UK governments were the big outliers, as the Cameron government abstained or voted against a large number of issues in Council in the build-up to the Brexit referendum in 2016, and then the May and Johnson administrations continued this behaviour until the UK left the EU in 2020. Excluding the UK may suggest a geographical tendency in voting during this particular time period: whereby North member states on average vote against East member states, with the South member states somewhere between these two groups. Nevertheless, within the North group, the main divide was between the more 'corporatist' member states (namely Germany and Austria) and the liberal' member states (such as the Netherlands. Sweden and Denmark). Nevertheless, as the uncertainty associated with these positions, illustrated by the lines, shows substantive overlap, we should be careful not to read too much into these maps. Perhaps geography is a poor predictor of voting behaviour in the Council. Also, as noted by Plechanovova (2011), because negative votes are rare, these models tend to do a poor job at predicting negative votes correctly. In short, as the EU has enlarged, the Council when passing laws has become increasingly like a 'normal' legislature, with coalition bargaining, strategic agenda-setting, position-taking and voting splits that result. Nevertheless, politics inside neither the Council nor the European Parliament is conducted in isolation from politics in the other institution. Under the ordinary legislative procedure, both institutions need to agree in order for new legislation to be adopted. We consequently now turn to bicameral politics in the EU. Legislative Politics 7B I Netherlands Sweden Denmark Belgium France Fil;l:nn.l Ireland Luxembourg Austria Germany - Greece Italy Cyprus l\u uigal Malta Lit! mania Croatia Sli>\v:lia Estonia Latvia Slovakia Romania Czech Rep. Bulgaria h, ,■,.■.;„, PoMr.J UK Netherlands Sweden Ivcnmo rk l-ielginni I.iLiu-'.hki l''inlřll,l Greece i'-elar.J Oc.:t ■ Italy Slovenia Cyprus Luxembourg Portugal Ksroni,: Latvia Malta S o'.v.kia Romania Spain (>ech Ken, Bulgaria ! iungarv Austi- .:' Poland Germany North without UK Figure 3.3 Voting behaviour in the Council Afofe: The figure shows the results of an item-response model of voting to votes in the Council between 2009 and 2019. Abstentions were counted as negative votes. The dots are the ideal points (mean estimates), the bars are 50 per cent (25%-75%) and 95 per cent (2.5%-97.5%) confidence intervals The model is estimated in R using pscl (Jackman, 2020). The top panel orders the member states by thoir revealed one-dimensional voting position, while the bottom panel orders the members states by geographic region, with the UK excluded. 80 The Political System of the European Union Bicameral Politics in the European Union Under the Lisbon Treaty, there are three main legislative procedures that regulate the relationship between the European Parliament and the Council: consultation, consent, and the ordinary legislative procedure (otherwise known as the co-decision procedure). Under consultation, the Commission proposes, the European Parliament then issues an opinion by a simple majority, and the Council then adopts, amends or rejects the proposal. Under consent, the Council (acting by OMV or unanimity, and on the basis of a Commission proposal, depending on the policy area) adopts a common position. The Parliament then gives its assent by a simple majority (although an absolute majority of all MEPs is required for admission of new member slates and amendment of the rules governing European Parliament elections). Under the ordinary legislative procedure, the Commission submits a proposal to the European Parliament and Council, and the European Parliament then issues an opinion by a simple majority. The Commission then decides whether to incorporate the amendments from the Parliament and sends the revised proposal to the Council. The Council then either accepts the proposal or adopts a common position by QMV or unanimity (depending on the policy area). If the Council incorporates all the proposed amendments from the European Parliament, the proposal becomes law. If the Council adopts a common position that differs from the European Parliament proposal, the bill goes to a second reading. In the second reading, the European Parliament can amend or reject the common position by an absolute majority of the MEPs. If there is no absolute majority in favour of amending or rejecting, the common position becomes law. If the European Parliament amends the common position, the Commission issues its opinions on the amendments and sends the proposal back to the Council. The Council then adopts amendments supported by the Commission by QMV, but needs unanimity to adopt European Parliament amendments that are not supported by the Commission. Again, if the Council adopts all of the European Parliament amendments at this stage, the proposal becomes law. It the Council and the European Parliament texts still differ, a conciliation committee is established. The conciliation committee is composed of one representative from each member state (usually the Permanent Representatives), an equal number of MEPs (which include a vice-president, the chairs of the relevant European Parliament committees, and the rapporteur) and a representative from the Commission. The Commission has no voting rights in the Conciliation Committee. If there is no agreement at this stage, the legislation falls. But, if an agreement is reached, it then has to be adopted by the Council and the European Parliament (by simple majority) in the third reading; otherwise it falls. Legislative Politics 81 These three procedures are the result of several decades of procedural politics in the EU. The European Parliament has managed to increase its power through a series of new legislative procedures introduced in the 1980s and 1990s. This generated a large body of formal models of EU legislative politics (e.g. Steunenberg, 1994, 1997a; Tsebelis, 1994; 1997; Crombez, 1996, 1997, 2000, 2001; Moser, 1996, 1997; Tsebelis and Garrett, 1997, 2000, 2001; Tsebelis and Kreppel, 1998; Tsebelis and Yataganas, 2002). The main controversy in this literature relates to the powers of the European Parliament under the co-operation procedure from the Single European Act and the Maastricht version of the co-decision procedure (co-decision I). While Tsebelis and Garrett (Tsebelis, 1994,1997; Tsebelis and Garrett, 1997, 2000) argue that the European Parliament was more powerful under the co-operation procedure than under co-decision I, Crombez (1996, 1997), Steunenberg (1994, 1997a) and Moser (1996, 1997) find that co-operation did not empower the European Parliament whereas co-decision I did. There is, nevertheless, consensus that the European Parliament and the Council are now equal legislators under the Amsterdam version of the co-decision procedure (co-decision II) - which after the Lisbon Treaty is now the ordinary legislative procedure. However, some scholars point out the second reading absolute majority requirement for amending or rejecting still puts the European Parliament at a disadvantage vis-á-vis the Council, as it is easier for the European Parliament to accept the common position of the Council than to reject or amend it (Hagemann and Hoyland, 2010). The best-known model of legislative politics in the EU is developed by Tsebelis and Garrett (2000). Figure 3.4 is a simplified version of their model. The model is based on several assumptions about EU legislative politics: • for QMV, the Council is deemed to have seven members, and a qualified majority is five out of seven; • there is a single dimension of legislative bargaining, between 'more' or 'loss' European integration; • the actors have ideal policy preferences on this dimension and want outcomes to be as close as possible to these positions, • the member states are aligned at different points along this single dimension; • the Commission and the European Parliament are more pro-integrationist than most member states; and • the sfafus quo (SQ), the policy outcome if no new policy is adopted, is less integrationist than any member state. 86 The Political System of the European Union reading agreements, but still 89 per cent of bills passed at first reading in the 2014-19 session. Finally, to understand EU legislative politics, it is vital to know both the position of actors in both the Council and the European Parliament, as well their interaction. Thomson et al. (2006) and Thomson (2011) collected positional data on all governments as well as the European Parliament and the Commission and use the data to test competing theories of EU decision-making. Their findings warn against focusing too much on the formal rules of decision-making at the expense of compromises and bargaining. In their data, EU legislative outcomes tend to be located closer to the policy position of the average actor in the negotiations, weighted by the amount of 'salience' each actor places on a particular issue, rather than to the policy position of a pivotal government in the Council, the European Parliament or the Commission (cf. Achen, 2006). Nevertheless, there continues to be a lively debate about whether 'procedural models' (which focus on the location of the status quo, the legislative rules, and the powers of the Commission, Parliament and Council) or 'bargaining models' (which focus on the positions of the EU governments and the EU institutions) better predict the location of EU legislative outcomes (cf. Leinaweaver and Thomson, 2014; Slapin, 2014; Franchino and Mariotto, 2021a). In the bicameral bargaining between the European Parliament and the Council, the cohesiveness of the chambers also matters. Hagemann and Hoyland (2010) find the European Parliament is less likely to meet the absolute majority requirement necessary for amending the common position of the Council when the Council is divided, than when the Council adopts the common position unanimously - which suggests a close connection between the political parties represented across the institutions. Nonetheless, the European Parliament, if it has the support of the Commission, tends to be more successful than the Council in the conciliation committee (König et al., 2007; Franchino and Mariotto, 2013). Legislative Politics 8? also an increasingly formalized system of bicameral interaction, which has been capable of coping with an increasing membership of the EU in terms of national party delegations (in the European Parliament) and governments (in the Council). Consensus in recorded voting in the Council and oversized majorities in coalition formation in the European Parliament are less a response to diverse social interests than a consequence of the institutional rules and policy preferences of the actors. The informal grand coalition between S&D and EPP in the European Parliament is fostered by the similar policy preferences of these parties on many EU issues as well as by the need to form a united front to secure amendments in the Council. Meanwhile, most governments in the Council are led by social democrats or Christian democrats/conservatives. As a result of the dominance of social democrats, Christian democrats and liberals in the EU's legislative institutions, policy outcomes from the legislative process tend to reflect the liberal' and pro-European preferences of these parties: a relatively deregulated single market, fiscal conservatism in economic and monetary union, external free trade agreements, high environmental and social standards, and liberal policies on gender equality, the internal free movement of people and the rights of refugees and migrants. Moreover, the checks-and-balances of the EU's bicameral system make it difficult to change these policies once they have been adopted, as the mainstream centre-left or the mainstream centre-right can usually find a way to bloc any major policy changes they do not like. This has implications for the democratic accountability of the EU, as well as for public support for the EU, particularly amongst social groups who either disagree with or do not benefit from the particular policy mix the EU legislative system produces. Conclusion: Bicameral Politics in the EU The European Union has a sophisticated and effective legislative system. Contemporary theories of legislative behaviour see the internal organization of the European Parliament and the Council and the processes of bargaining and coalition formation as products of the rational self-interests of the EU legislators: the MEPs and political groups in the European Parliament, and the governments in the Council. Both institutions have developed strong systems of specialization to facilitate information gathering, bargaining and coalition building. There is