Introduction
The expenditure policy of the EU is perceived as a product of the past characterised by a strong status quo bias whereby changes in the budget are carried out through incremental additions (Laffan and Lindner, Reference Laffan and Lindner2010: 227; Rant and Mrak, Reference Rant and Mrak2010: 351). Status quo bias is similar to Scharpf’s theoretical expectation about a joint decision trap, in which decisions are taken by unanimity and accordingly, have a tendency to follow the logic of the lowest common denominator (Scharpf, Reference Scharpf1988: 257; Schild, Reference Schild2008: 533; Scharpf, Reference Scharpf2011: 223; Citi, Reference Citi2013: 1157). At the European Council’s meeting on 7–8 February 2013, the European heads of governments reached an agreement on the multiannual financial framework (MFF) for the period 2014–20. The deal deviates from the predictions of status quo bias, as it schedules cuts in the overall expenditure level for the first time since the MFF was introduced in 1988 (European Council, 2013: 46–47; Schild, Reference Schild2008: 536). Furthermore, the deal shows remarkable changes in the respective shares of the total expenditure under the different headings. Though significant reprioritisations appeared in the Delors I (1988–92) and Delors II (1993–99) periods, these were qualitatively different by the allocation granted on top, as opposed to actual cuts. In addition, these reprioritisations were linked to the major integration projects of the Single Market and the EMU (Laffan and Lindner, Reference Laffan and Lindner2010: 219). The Agenda 2000 package for 2000–06 reflected a concern of the member states for the stabilisation of expenditures, and the negotiations were linked to the accession of new member states (European Commission, 2008). Nevertheless, the agreement entailed incremental changes in the allocations under the different headings. The MFF 2007–13 marked a substantial status-quo bias in the light of the adoption of the Lisbon goals and significant calls for reprioritisations of the expenditure (Schild, Reference Schild2008).
The MFF 2014–20 contains significant reprioritisations, which challenge the prediction of a status quo bias, including a joint decision trap in systems with many veto players. The purpose of this paper is to explain this conundrum by identifying the key factors for making significant reprioritisations possible in a budget procedure that is characterised by a strong status quo bias. This puzzle is answered by an analysis based on actor-centered institutionalism developed by Scharpf (Reference Scharpf1988, Reference Scharpf1997, Reference Scharpf2006, Reference Scharpf2011). The issue of status quo bias in the MFF has been argued to be particularly important since the Treaty of Lisbon came into force, as this stipulates that the budget is continued in the case of a situation where no agreement on a new budget has been reached (OJ, 2007: 122). This is assumed to amplify the institutional basis for status quo bias in MFF (Laffan and Lindner, Reference Laffan and Lindner2010: 222). This study adds to the sparse literature, which has focused on the MFF negotiations prior to the Lisbon Treaty and before the probably worst economic crises in the history of the community took effect (Schild, Reference Schild2008; Dür and Mateo, Reference Dür and Mateo2010; Laffan and Lindner, Reference Laffan and Lindner2010; Rant and Mrak, Reference Rant and Mrak2010).
The analysis is delimited in two ways. First, the focus is on the intergovernmental bargaining, in which the governments of the member states were key actors until the agreement on the 8 February 2013 was reached. Second, focus is on explaining the expenditures, especially under the headings ‘Market Related Expenditures and Direct Payments’ (agriculture) and ‘Economic, Social and Territorial Cohesion’ (cohesion policy). Taken together, these two headings amount for two-thirds of the EU’s total expenditure. The revenue part of the budget will be included in the analysis, as the issue of rebates is inextricably linked to the negotiations on expenditure. The rest of the paper is divided into six parts. The institutional design of the MFF, as well as the deal that was reached, is described in the second section. In the third section, the theoretical framework is specified. The fourth section outlines method and data. In the fifth section, the agreement is analysed from the perspective of the theoretical framework. The sixth section summarises the results including their implications.
Negotiating the MFF
The first part of this section contains a general account of the MFF, after which the main elements of the agreement from the 8 February 2013 are outlined.
The MFF as an institution
The MFF lays down the EU’s financial course for periods of at least 5 years by specifying overall expenditure as well as under the different headings (OJ, 2007: 122). Revenue and expenditure must be balanced in the budget. The implementation of the appropriations is specified in the annual budgets and through separately adopted legislative acts. The allocation method and criteria for receiving funds from the cohesion policy are set up in the MFF, whereas the allocation of direct payments under the common agricultural policy (CAP) is decided primarily through legislative acts. Payment of funds from the budget generally requires an appropriation and a legislative act (OJ, 2007: 121), with the notable exceptions of direct payments and pensions. The legislative acts cannot be passed before the MFF is adopted. The MFF must be adopted by unanimity in the European Council, and hereafter approved by a majority of the members of the European-Parliament (EP), and by unanimity in the Council. The EP must ratify the revenue (OJ, 2007: 121).
The budget has traditionally amounted to ∼1% of the EU’s Gross Domestic Product (GDP), corresponding to 2.5% of the total public expenditure within the EU (see Table 1). Redistribution through the EU’s budget is transparent, because the member states can easily calculate their own net balance (Bache, Reference Bache2007: 395). The variation in member states’ net balances is shown in Figure 1.
Net balances (2011) in percent share of GNI. Source: European Commission (2011a: 102).
Development in appropriations, billion euro (2011 prices)
The numbers for 2014–20 are referring to the agreement on 8 February 2013 at the European Council’s meeting.
Source: European Council (2013).
The deal
Existing literature characterises the MFF as an agreement on three conflict dimensions: (1) the overall expenditure level, (2) the allocation of the expenditure under different headings, and (3) the allocation of funds within these headings (Begg, Reference Begg2005: 14; Dür and Mateo, Reference Dür and Mateo2010: 559). That there actually was a deal at the European Council’s meeting of 8 February 2013 is remarkable. Negotiations in November 2012 ended without an agreement and it was not a given that one could be reached in February 2013. On the first conflict dimension, the total expenditure level was reduced, measured as a share of the EU’s GNI as well as in absolute numbers (fixed prices), which has never been seen before (Schild, Reference Schild2008: 536; EC, 2013: 46). The growth in total expenditure has been decreasing since 1993 (Schild, Reference Schild2008: 536), and the adoption of a flexibility mechanism will most likely give the EP power to allocate unspent money. This is expected to reduce the net effect of the budget cut, because unspent money has traditionally been sent back to the member states (European Council, 2013: 42). The already existing rebates given to the United Kingdom, Germany, the Netherlands, Austria and Sweden will continue to apply, while Denmark has been granted a rebate worth €130 million. Regarding conflict dimension number 2, more funds were allocated for the ‘Competitiveness for Growth and Jobs’, ‘Administration’, ‘Security and Citizenship’, and ‘Global Europe’ (see Table 1).
A key issue in the negotiations on cohesion policy was how the funds should be allocated (conflict dimension 3) among less developed regions, transition regions, more developed regions, the outermost and sparsely populated regions, and member states that are eligible for receiving funds (European Council, 2013: 11). For member states that were eligible between 2007 and 2013 but whose GDP per capita now exceeds 90% of the EU average, there will be a transition scheme (European Council, 2013: 17). Furthermore, a so-called safety net was adopted, which benefits regions that fall out of the category ‘least developed regions’ (European Council, 2013: 18). For regions in the less prosperous countries, the maximum allocation is capped at 3.15% of GDP (European Commission, 2013: 14). The special allocation provisions are qualitatively different from the other appropriations by being aimed at specific regions, instead of being allocated according to objective criteria (European Council, 2013: 19–20).
The criteria underlying the allocation of funds in the first column of the CAP are less specified compared with the cohesion policy. The level for direct payments per hectare in the member states is an exemption where farmers receive different amounts depending on which member state they live in. The level of payments is lower for the ‘new’ member states. The gap will be partly equalised with the agreement for 2014–20, which inflicts the greatest change for member states such as the Netherlands, Estonia, Latvia, and Lithuania, whose aid per hectare deviates markedly from the EU average (European Commission, 2011b: 12; European Commission, 2013: 26). Conversely, the net effect is much lower for member states like France and Germany, which are closer to the average.
Theory
Scharpf’s actor-centered institutionalism serves as the principal theoretical framework, which will be supplemented with Tsebelis’ (Reference Tsebelis2002) concept of veto players, Rant and Mrak’s (Reference Rant and Mrak2010) study of member states’ preferences and Putnam’s (Reference Putnam1988) two-level game approach.
Actors
The first elements for explaining policy outcomes are actors, who are determined by the institutional design (Scharpf, Reference Scharpf1997: 43). In the case of the MFF, the member states and the Commission are the main actors in the negotiations leading to the intergovernmental agreement. Unanimity as decision rule entails that the member states are veto players (Tsebelis, Reference Tsebelis2002: 19). A veto player is influential insofar as its threat of vetoing is credible. A credible veto player must have a possibility as well as a motive for exercising its veto. The concept of best alternative to a negotiated agreement (BATNA) is crucial for the credibility of a veto threat (Scharpf, Reference Scharpf1997: 122). A favourable BATNA gives an actor a strong position for negotiation, as (s)he will not incur losses in case of a negotiation breakdown, and will therefore be able to threaten a veto with a high degree of credibility. Conversely, an actor with an unattractive BATNA will be more dependent on an agreement, resulting in a low propensity to veto and consequently a weaker negotiation position. An actor’s preferences in the negotiations are given by the position on the three conflict dimensions outlined previously: the overall expenditure level; allocation of funds under headings; and the allocation of funds within these headings. Positions on these conflict dimensions constitute a win-set, that is, the room of acceptable negotiation outcomes, which deviate from the status quo (Tsebelis, Reference Tsebelis2002: 21). The potential for reform of existing policy lies within the credible veto players’ respective win-set. Theoretical development of the joint-decision trap, and empirical studies, show that the Commission shall be regarded as an important actor, as it can influence the outcome through agenda-setting power, and facilitate exit mechanisms by acting as an honest broker (Schild, Reference Schild2008; Scharpf, Reference Scharpf2011: 229–231). Thus, the Commission can possibly act to overcome a joint decision trap through a role as policy entrepreneur.
A combination of deductive and inductive approaches is used to map the preferences of the actors. In the case of congruence between the two approaches, the mapping of the actors’ preferences will appear reliable. If this is not the case, the credibility of the different sources will be assessed, including a consideration of the reasons for incongruence. First, the preferences will be deducted from the self-interests of the respective member states on the three conflict dimensions. The background for a systematic estimation of their preferences is to assume that they seek to maximise their own net balance, that is, that they act according to the logic of juste retour (Rant and Mrak, Reference Rant and Mrak2010; see Table 2). Second, an inductive strategy is used where we substitute the actors’ preferences with their negotiation positions (Jensen and Nedergaard, Reference Jensen and Nedergaard2012). Actors’ negotiation positions will be derived from documents from the process leading up to the agreement; interviews; and existing literature on the member states’ negotiation positions (see Method and data section). The different actors are treated unitarily, except for Germany, the United Kingdom, France, and Poland. Putnam’s (Reference Putnam1988) two-level games perspective will be applied to these four key member states to analyse how domestic politics impact the win-set of the respective heads of government.
Coalitions
Constellation
The second main element for the policy outcome is the constellation, which is a function of the composition of actors, their strategic options, and returns hereof (Scharpf, Reference Scharpf1997: 69–96). As no actor expects to be able to determine a policy outcome on their own, they tend to build coalitions. A coalition is characterised by its negotiation position, aggregated capabilities, internal homogeneity, and by the intensity of its preferences on a given conflict dimension (Scharpf, Reference Scharpf1997). A coalition’s capabilities equal the sum of its actors’ capabilities, defined as resources that enable an actor to influence policy on certain issues and to a certain degree (Scharpf, Reference Scharpf1997: 43). As indicators for capabilities, this paper will use the gross payments to the budget, combined with assessments from interviewees with insight in the negotiations. Though aggregate wealth should not matter in theory in unanimous decisions, it can be seen as a proxy for how many resources a member state can devote to the negotiations of the MFF. Previous research has demonstrated that political and economic weight is likely to increase member states’ abilities to shape history-making decisions (Moravcsik, Reference Moravcsik1998; Wallace, Reference Wallace2005).
A coalition is more influential the more internally homogenous it is (Scharpf, Reference Scharpf1997: 83). Internal homogeneity in a coalition is a measure of the degree of commensurability of the actor’s positions on the conflict dimensions. This is calculated as the standard deviation of a coalition’s net budgetary balances in percentages of GNI. Intensity of preferences denotes the degree to which a coalition’s members are occupied by a given conflict dimension (Listhaug et al., Reference Listhaug, MacDonald and Rabinowitz1995: 456).
Mode of interaction
The third element for explaining the political outcome is the mode of interaction. Whereas the constellation describes the level of potential conflict, modes of interaction describe the institutional framework within which the conflict must be resolved (Scharpf, Reference Scharpf1997: 46). The joint decision trap denotes a situation in which a composition of actors systematically produces suboptimal policy with a significant status quo bias (Scharpf, Reference Scharpf1988: 254–259, 271). The outcome is a function of unanimity as decision rule and of the absence of problem-solving as decision style. The decision rule can also be affected by informal institutions, such as a norm of consensus. Unanimity as decision rule creates a number of veto players, who are all likely to block any agreement that leaves them worse off than the status quo; therefore existing privileges will often be continued even if inefficient. The more veto players and the higher the heterogeneity of their preferences, the smaller the win-set (Tsebelis, Reference Tsebelis2002). Thus, the possibility of reaching an agreement, which deviates from the status quo is weakened (Scharpf, Reference Scharpf1997: 123).
Scharpf distinguishes between three decision styles: problem-solving, distributive bargaining, and confrontation (Scharpf, Reference Scharpf1988: 259). The decision mode of the MFF negotiations resembles distributive bargaining where the actors may bypass the logic of the joint decision trap through issue linkage and/or side payments (Scharpf, Reference Scharpf1997). By side payments, it is possible to compensate actors, who lose on certain parts of a particular agreement. This increases the possibility of reaching an agreement that lies within the win-set of the veto players, meaning that the logic of lowest denominator is bypassed.
Method and data
This paper is based on a case study of the MFF 2014–20, which can be perceived as a deviant case, because the result is unexpected from the basic theoretical framework in Scharpf’s joint decision trap (Gerring and Seawright, Reference Gerring and Seawright2008: 301–302). Furthermore, the outcome deviates empirically, as the total expenditure level is cut for the first time since the MFF was introduced as a budget procedure in 1988. Thus, the MFF 2014–20 is a useful case to refine the theoretical understanding of exit mechanisms from the joint decision trap (Falkner, Reference Falkner2011 b: 237–258). The mapping of actors’ preferences and strategies in the negotiations is based on statements from governments and national parliaments, statistics and governments’ pronouncements to the press, and existing literature on the individual member states’ preferences in the negotiations. Eight interviews have been carried out with key interviewees working for the member states, the DG Budget, and the Parliament’s Committee on Budgets. Interviewees have been selected in a way that represents the diversity of coalitions and preferences. To ensure the validity of the source’s statements, they are compared with other interviews, policy documents, announcements in the media, as well as the expectations that can be deducted from the logic of juste retour (see above).
Analysis
The analysis is structured in three sections on actors, constellations, and modes of interaction.
Actors: the positions of the leading member states
The member states and the Commission were the central actors in the process leading to the intergovernmental agreement on 8 February 2013 (Dür and Mateo, Reference Dür and Mateo2010: 558; Laffan and Lindner, Reference Laffan and Lindner2010: 213). However, the EP influenced the agreement, as the member states anticipated its preferences with the purpose of getting it approved in summer 2013 (Interview 4: 57m41). In concrete terms, this applied to the flexibility mechanism, which mandates that the EP allocates unspent money from the budget and to the adoption of a review clause in 2016 (EC, 2013: 42).
This section analyzes the preferences of Germany, the United Kingdom, France, and Poland based on Putnam’s (Reference Putnam1988) two-level games approach. Germany, the United Kingdom, and France are regarded in the literature as the most powerful member states in the EU (Moravcsik, Reference Moravcsik1991), while the Polish government has been a leading negotiator and coordinator for the member states with an interest in a strong cohesion policy (Interview 3: 1h03m40; Interview 6: 36m20). The positions of the remaining member states are analysed in the next section concerning the constellation. Besides these four member states, the preferences of the Commission will be mapped, because it has been portrayed as a policy entrepreneur that facilitates reforms (Schild, Reference Schild2008: 534; Scharpf, Reference Scharpf2011: 229–231).
From the perspective of juste retour, it was in the German government’s interest to decrease the total expenditure level, as Germany is the biggest net contributor to the budget. This is compatible with the German government’s austerity approach to economic policy in times of crisis. The domestic circumstances have narrowed the win-set of the German government. The German population does not want to finance other EU states in times of economic trouble, more than has already been the case. Thus, the external circumstances of the eurocrisis established an issue linkage between Germany’s ‘considerable efforts to stabilise the single currency’ and its net budgetary balance (Deutscher Bundestag, 2011: 1). Regarding the allocation of funds, it is favourable for Germany to maximise the appropriations for innovation and research at the expense of agriculture and cohesion policy. This was also the German position (Becker, Reference Becker2012: 17). The Länder from the former Eastern Germany have argued for the adoption of a so-called safety net as a transition mechanism, because they risk losing significant funds from the cohesion policy (Deutscher Bundestag, 2011: 4).
France has advocated a lower overall expenditure level, which follows its status as one of the biggest net contributors. France has marked an intense preference for keeping the direct payments in the CAP at the present nominal level, stating from the beginning that it would block any agreement that did not guarantee this (The Danish European Affairs Committee, 2011: 1). This is in line with existing studies, which portray France as the main protagonist of the CAP (Dinan, Reference Dinan2005). The position on convergence in the level for direct payments was less intensive as a result of the French level already being relatively close to the EU average (European Commission, 2011b: 12). The French government argued for a concentration of the cohesion funds for the least developed regions. This could be implemented by discontinuing the appropriations for ∼20 regions that did not meet the criteria for receiving funds, and by basing the allocation method on regional prosperity instead of national prosperity.
The United Kingdom had a strong wish to reduce the overall level of payments significantly, as well as a desire to continue the British rebate. This follows its position as a net contributor to the budget, but also mirrors the special domestic circumstances, which are marked by a great and increasing EU-skepticism, as clearly expressed in David Cameron’s announcement of a British referendum on the United Kingdom’s future membership of the EU. Thus, the demand for budget cuts was as much a wish for ‘less Europe’ as for an improvement of the British net balance (Interview 7: 9m10). Leading up to the European Council’s meeting in November 2013, the British government lost a vote in the House of Commons by a group that counted some of Cameron’s own conservative party colleagues, among others (BBC, 2012: 1). The narrow majority wanted the government to take a more radical position than Cameron had done with his demand that the expenditure should not increase by more than the inflation level. With this remarkable result, it was not unrealistic that the House of Commons could end up blocking an MFF agreement if this did not contain cuts. Cameron’s win-set on the overall expenditure level was very narrow, creating a strong British position in the negotiations.
Poland differed from Germany, France, and the United Kingdom by not advocating a reduction in the total expenditure level. The Polish government wanted a strong cohesion policy. Calculations based on the Commission’s first proposal showed that cohesion funds would account for two-thirds of the total funds that Poland was to receive (Kalan et al., Reference Kalan, Tokarski and Toporowski2012: 2). The concentration of cohesion funds for the least developed regions was not a crucial issue for the Polish government. On the contrary, it opposed transition schemes for regions that are about to fall out of the system (Kalan et al., 2012: 9). Regarding agriculture, there was a consensus about supporting the current level for direct payments. At the same time, there was an interest in achieving convergence, so that Polish farmers would not receive less aid per hectare than farmers in Western Europe. In general, the Polish government enjoyed considerable autonomy as it was not tied by domestic veto players, who would have weakened its negotiation position.
The Commission advocated higher expenditure, which is evident from its proposals. This follows existing literature that characterises the Commission as a budget-maximising actor with a preference for supranational expenditure (Laffan and Lindner, Reference Laffan and Lindner2010: 214). The Commission’s first proposal included a 5.7% increase in total commitments, reflecting a strong preference for the Connect Europe Facility and Horizon 2020 under heading 1, as well as Global Europe (heading 4) and Administration (heading 5) (European Commission, 2011c; Barroso, Reference Barroso2012).
Constellation: the key coalitions
Having mapped the central actors’ preferences, this section contains an analysis of the constellation, which describes the most important coalitions in the negotiations, their strategic options, and return hereof. The focus will be on the coalitions’ respective strength measured according to capabilities, internal homogeneity, and preference intensity. The starting point for the constellation is the character of the policy that is negotiated.
The coalitions in the negotiations are more or less unchanged since the negotiations on the MFF 2007–13, with juste retour still being a crucial factor for the preferences of member states. The groupings into net contributors and net recipients are almost unchanged (Rant and Mrak, 2010). This can be seen from Table 2, where the actors have been mapped deductively based on percentages of GNI (net position – see Appendix 1) and inductively based on documents and interviews.
The most notable coalitions have been ‘Friends of Better Spending’, ‘Friends of Cohesion’, and ‘Friends of Agriculture’ (see Table 2; Kölling, Reference Kölling2012). These coalitions are not theoretical constructs but labels used by participants in the negotiations and observers hereof to classify groups of like-minded member states. Different actors will very rarely have identical preferences on all different conflict dimensions. Therefore, the final agreement should not be perceived as a simple compromise between coalitions that are internally homogenous, but rather as a result of many specific interests and ad hoc groupings (Interview 5; 25m50; Interview 7: 42m50).
‘Friends of Better Spending’ were united by a demand that the overall expenditure level should not rise. The coalition consists of member states in the north and west, who are net contributors to the budget (Table 2) and there is a close relationship between the inductive and deductive mapping of preferences. This group of member states voiced its opinions publicly by common non-papers and letters (Europolitics, 2012). In December 2010, the group (excluding Sweden, Denmark, and Austria) advocated that total payment appropriations should not increase by more than inflation, while the total commitment appropriation should increase by less than inflation (Cameron, Reference Cameron2010). The homogeneity of the coalition on the overall expenditure level must be considered as high because there is not much variation in the members’ relative contribution to the budget. Furthermore, the cluster analysis reveals that a number of receivers from the budget are closer to the net-contributers in Euclidian terms (see Appendix 2). While being a theoretical abstract, this suggests that the net receivers are a much more heterogeneous group of member states when it comes to the overall expenditure level. If we view the MFF negotiations as an ongoing game, this may explain why some of the net receivers did not put up more of a fight to maintain the status quo because they knew that in the next round they might be net contributors.
The intensity of preferences regarding the total expenditure level varied internally in the coalition. France and Italy were not anxious about cuts, whereas the United Kingdom, the Netherlands, Sweden (Interview 4: 21m05) and Denmark gave higher priority to bargaining on the revenues (Interview 4: 18m05; Interview 3: 43m20). The coalition had high aggregated capabilities, as it counted a number of strong actors, including the four largest gross contributors to the EU budget: Germany, France, Italy, and the United Kingdom (European Commission, 2011a: 33).
Having budget cuts as a starting point, the coalition ‘Friends of Better Spending’ needed to find cutbacks, but when moving from the first conflict dimension concerning the total expenditure to the second and third conflict dimensions, consensus was hard to find. However, the broad outline was a preference for more funds to ‘Competitiveness for growth and jobs’ (Interview 5: 39m23; Interview 1: 16m45; Interview 4: 33m30). France had a strong preference for continuing the current level of direct payments under the CAP, while the Netherlands argued against the convergence in aid per hectare.
The other major coalition in the negotiations, ‘Friends of Cohesion’ (Table 2), consists of the EU’s least prosperous member states, which receive significant shares of their GDPs from the EU’s cohesion policy. These member states obviously have a strong preference intensity for reallocation. As for the coalition assessed above, there is almost perfect correspondence between the inductively and deductively derived coalition, which counts many member states (European Commission, 2011a: 33). The coalition was united by a preference regarding the composition of expenditure, while the preference regarding total expenditure level was less intense, though with a leaning towards a higher expenditure level than was the case in ‘Friends of Better Spending’. Furthermore, the group was much less homogenous compared with the other groups due to considerable variation in how much the individual members benefit from the heading and especially the sub-headings. The group of member states had official summits and reached joint statements on the MFF in Bucharest, Bratislava, and Brussels (Polish Government, 2012; Romanian Government, 2012; Slovakian Republic, 2012). The Czech Republic is almost a net contributor to the EU budget and has generally been reluctant to increase the expenditures (Kalan et al., Reference Kalan, Tokarski and Toporowski2012: 7); the Bulgarian government also worked against the general preference in the coalition by announcing a wish for new conditionalities for receiving cohesion funds. The preferences for allocation of funds within the different headings were intense, though not homogenous on the issue of how the cohesion funds should be allocated between the different recipient types. Spain advocated in favour of transition regions, as a number of Spanish regions risked losing all of their appropriations as they are relatively more prosperous compared with the regions in Central and Eastern Europe (Kölling and Leal, Reference Kölling and Leal2012: 6). ‘Friends of Cohesion’ generally supported the establishment of transition regions. However, the Bulgarian government demanded that this should not happen at the expense of the less prosperous regions in the East, and the Czech government worked for concentrating the funds in the least developed regions (Kovacheva, Reference Kovacheva2011: 1; Interview 5: 35m15). ‘Equal treatment’ was a dominating principle for the coalition, signifying a demand for convergence in aid per hectare and a rejection of the proposal to adopt a maximum for the size of shares that regions and member states can receive from the cohesion policy.
A third coalition, ‘Friends of Agriculture’, focused on the level for the direct payments in the CAP (La Moncloa, 2012: 1). Here there is also a good correspondence between inductively and deductively mapped preferences, though France has played a much more leading role than one should expect from its net position, which is close to 0 (−0.006 or 0.033 of GNI if only market-related expenditure and direct aids is counted – see Appendix 1). However, the discrepancy must be attributed to the fact that the French have historically been the lead protagonist of the CAP and that French farmers are de facto veto players (Dinan, Reference Dinan2005).
France and Spain announced a common position on this issue, in particular after a meeting between their respective ministers of agriculture (La Moncloa, 2012), and Ireland was also part of the core in this coalition. The coalition’s homogeneity can be classified as medium as there is some variation in economic benefits, but not extreme. In its broadest sense, the coalition also counted a number of Central and Eastern European member states, but these mainly focused on arguing for convergence in aid per hectare (Orosz, Reference Orosz2012: 202). The Baltic countries in particular have been advocates of this, in contrast to the Netherlands, which would be severely affected by such a convergence (Interview 5: 39m23).
To understanding the bargaining constellation, the default outcome of each actor’s BATNA position, that is, the best alternative to a non-agreement, must be identified. The Lisbon Treaty stipulates that in the event of a non-agreement, the ‘existing ceilings and other provisions corresponding to the last year of that framework shall be extended until such time as that act is adopted’. The legal uncertainty was considerable in this regard, as a non-agreement would activate aspects of primary law that had not been used before, and the different actors tended to stress aspects of a non-agreement scenario that would benefit their own relative bargaining power. However, a continuation of the MFF itself would not necessarily imply a continuation of programmes, as ‘the framework programme shall be implemented through specific programmes developed within each activity’ (TEUF, article 182, 3). The agreements on the specific programmes would not be reached until the European Council’s agreement on the MFF. Therefore, the member states advocating for cohesion policy could not credibly pursue a strategy of vetoing until the existing MFF was continued into 2014, as this would seriously put in danger the flow of funds under the cohesion policy as of 1 January 2014. EU funds constitute a significant share of GDP for many of these member states in Central and Eastern Europe (Figure 1). On the other hand, the direct payments under the CAP would continue regardless of an agreement.
Modes of interaction: a distributive bargaining game with unequal outside options
Having mapped the most important actors’ preferences and the existing constellation, the following section will assess the institutional incentives that either systematically influence the tendency to status quo or put some coalitions in a favourable position. The section highlights four characteristics (a) the impact of the institutional design, (b) the joint decision trap, (c) issue linkage, and (d) the actors’ outside options.
The paradox of unity in ‘Friends of Cohesion’
By defining the issues on the table, the institutional setup is influential on which conflicts may unfold internally in the coalitions. Combined with the member states’ preferences, this created different possibilities for unity in the different coalitions. During the negotiations, the core members of the ‘Friends of Agriculture’ (France, Spain, and Ireland) had favorable conditions for maintaining unity in negotiating for agriculture. These member states had low preference intensity on the allocation of agricultural funds, because the aid per hectare, allocated to them, was already close to the EU average (European Commission, 2011b).
On the other hand, the member states in the ‘Friends of Cohesion’ had both high preference intensity and diverse preferences on the allocation of cohesion funds. While all of these member states advocated for cohesion funds in general, some preferred funds for the least developed regions, while others preferred funds for transition regions or for safety nets. Another option was to bargain for ad hoc appropriations under the cohesion policy’s ‘special allocation provisions’. Thus, the combination of divergent preference and the institutional design placed the member states in ‘Friends of Cohesion’ in a game that resembles the prisoner’s dilemma (Kalan et al., Reference Kalan, Tokarski and Toporowski2012: 8). Table 3 shows this in a simplified manner with Poland and Spain as actors. For both member states, the dominant strategy was to choose defection, that is, bargaining for special allocation provisions or the allocation method that secures the individual member state the largest possible share of the cohesion funds. The best collective outcome would be achieved if both actors negotiated for cohesion funds in general. However, this was not a likely scenario. If the Polish government chose to advocate more funds for cohesion in general (cooperation) it would expose itself to the worst possible outcome if Spain maximised its own utility by negotiating for a large share of the cohesion funds being allocated to it (defection; outcome no. 3 in Table 3). Thus, homogeneity was more difficult to maintain in ‘Friends of Cohesion’ than in ‘Friends of Agriculture’ and ‘Friends of Better Spending’. Defection from the common goal in ‘Friends of Cohesion’ was especially evident in relation to the special allocation provisions, which were granted in the final process during Van Rompuy’s bilateral negotiation with the different heads of government (Interview 6: 30m15; Interview 3: 1h09m43). The defections at the final stage indicates that the ‘Friends of Cohesion’ ended up having a suboptimal outcome (outcome no. 4 in Table 3), and did not succeed in maximising the aggregated funds for cohesion.
Prisoner’s dilemma applied to Spain and Poland (Friends of Cohesion)
The joint decision trap
The negotiation on the MFF 2014–20 shares characteristics with Scharpf’s theoretical concept of the joint decision trap, which is a product of unanimity/consensus as decision rule and the absence of problem-solving as decision style. The decision rule in the MFF requires decisions to be adopted by unanimity, and the culture among the governments is characterised by a norm of consensus. This decision rule constituted each of the 27 fully accessed member states as veto players in the negotiations. The decision style in the MFF came closer to Scharpf’s ideal type of bargaining than that of problem solving, as the negotiations mainly appealed to the member states’ individual interests rather than to a common utility function. A number of actors threatened to veto proposals that did not follow their interest. This indicates a distributive bargaining decision style rather than a problem-solving approach (Kirkup and Waterfield, Reference Kirkup and Waterfield2012: 1). As everyone expected cuts in the overall expenditure level, the different governments experienced an increased risk of ending up with clearly identifiable losses in the negotiations. This is contrary to previous MFF negotiations, in which the demands of all actors could be met to some extent by increasing the total budget (Schild, Reference Schild2008: 534).
The joint decision trap implies that no veto player is expected to approve an agreement on the MFF 2014–20 if it places the actor in a worse situation than without an agreement. For example, 11 member states lost 20% or more of their appropriations received from cohesion policy when compared with the period 2007–13. This raises the question of, how all heads of states and government could approve an agreement that leaves some member states worse off in terms of net balance.
Side payments and policy entrepreneurs
Side payments have been crucial for achieving consent from all governments. The three conflict dimensions, on which the negotiations take place, were not conceived by the actors as separate in practice, as the different parts of the agreement are inextricably linked. A reduction in total expenditure level raises the question of where to cut and which actors – if any – should be compensated through other policies. The eventual intergovernmental agreement on the MFF was predisposed to comprehensive issue linkage, as the negotiations were run by the heads of government, resulting in strong decision-making power and the possibility to trade wins and losses across policies areas (Scharpf, Reference Scharpf1997: 130). Compensation by side payments was most clearly seen in the agreement’s section on special allocation provisions (EC, 2013: 19–20). Allocation through such an ad hoc approach happens in stark contrast to the generally applied method for allocation of cohesion funds, in which criteria are defined to decide which regions and states can receive appropriations. These special allocation provisions were assigned at the final stage, and were decisive for reaching an agreement (Interview 1: 24m40; Interview 4: 55m05; Interview 6: 22m58). First, the special allocation provisions constituted an option to compensate actors who were about to lose from an agreement. Hungary was granted €1.56 billion to compensate for the adoption of an appropriation maximum on 3.15% of GDP, and the Czech regions were granted €600 million as compensation for an agreement, which was generally unattractive for the Czech Republic (Interview 1: 26m00). Second, the special allocation provisions can be of major significance for the heads of government, who had to defend the agreement, when they returned to the domestic veto players for ratification. This also applied for provisions with smaller amounts, such as the €50 million granted to Ceuta and Melilla (Interview 3: 28m18). In this manner, the heads of government can appear domestically as winners, although they might leave the negotiation table with a worse net balance than when they arrived.
Whereas side payments enabled an agreement that deviated from the status quo, the Commission did not mark itself in the role of a strong policy entrepreneur able to bypass the logic of the joint decision trap. The achieved agreement does not come close to the Commission’s first proposal, especially with regard to the overall expenditure level. Various factors seem to account for this. Firstly, from a theoretical perspective, unanimity as formal decision rule weakens the agenda-setting power of the Commission, as all member states must accept a given agreement (Hix and Høyland, Reference Hix and Høyland2011: 243). This is contrary to the QMV in the adoption of the annual budget, which enables the Commission to pass through proposals even if these do not follow the interests of some member states. Second, the Commission did not de facto hold a monopoly on initiative, as both the President of the European Council and the rotating presidency put forward negotiation boxes. Thus, the Commission’s potential to act as policy entrepreneur through agenda-setting power was weakened. Third, the Commission was perceived more as a politicised actor pursuing its own interests than as an honest broker for the member states. The Commission’s proposal of a 5.7% increase in expenditures was not seen by the respective governments as a credible starting point for negotiations due to the euro crisis and the significant fiscal constraints in many member states. In particular, member states with preferences for a lower expenditure level complained about the Commission not fulfilling its de jure role as an honest broker (Interview 4: 48m45). The role as an honest broker seems, to some extent, to have been taken over by the President of the European Council, who was influential because he, to a larger extent than the Commission, was regarded as a neutral actor (Interview 2: 14m17; Interview 6: 23m20). This is illustrated by the fact that Van Rompuy’s negotiation boxes came markedly closer to the eventual result than the equivalent one made by the Commission.
Unequal outside options
A credible veto player must necessarily have a formal option as well as a motive to play the veto card. If the member states are not credible veto players, the logic of the joint decision trap is weakened, because deviations from the status quo will not necessarily be blocked.
For several reasons, none of the relevant actors had an actual interest in delaying the agreement on the MFF (Interview 5: 55m40). First, a delay would endanger the disbursement of funds to EU-financed projects from 2014, with the notable exception of direct payments and pensions. Second, the EU as a political system would show a lack of decision power during times of crises, which might be punished by the market and the citizenry. In the ‘shadow of future cooperation’, some member states have diverged from their short-term national interests to support the credibility of the EU as a decision-making system and the long-term benefits it provides in terms of lowering transaction costs of negotiations and managing economic interdependence (Moravscik, 1998; Falkner, Reference Falkner2011a: 248). Third, the heads of government and their officials would have to use more resources on the negotiations. However, there are differences in how great the costs such a scenario with delay would cause to the different coalitions. The European Parliament’s approval of the agreement takes time, as does the agreement of the legislative acts that are required for the implementation of the MFF. Even with the agreement in February, it was a serious challenge to disburse cohesion funds from the beginning of 2014 (Interview 7: 58m15). The big recipients of cohesion funds would accordingly be relatively inclined to give up on their priorities in the negotiations, compared with the member states in ‘Friends of Better Spending’ (Interview 7: 59m05). The member states in ‘Friends of Better Spending’ could, consequently, wait for ‘Friends of Cohesion’ to relax their priorities because of the perspective of losing several percent of their GDP due to defaulting disbursements from 2014.
In the form of unequal outside options, the mode of interaction created an uneven power relationship between the two coalitions, with ‘Friends of Better Spending’ having a more favourable negotiation position than ‘Friends of Cohesion’. This seems to imply that an agreement was adopted even if it placed some actors as worse off than the status quo. However, the tendency of unequal outside options is not clear cut: the German government was interested in finishing before the upcoming German Federal election in September 2013 and Spain would, to some extent, benefit from a delay, as the data underlying the allocation of cohesion funds would then be updated to the exacerbated economic situation in the country (EC, 2013: 12; Interview 3: 34m50 and 49m45; Interview 5; 54m55; Interview 7: 57m10).
Conclusions
The European Council’s agreement from 8 February on the MFF 2014–20 lays down the EU’s financial course for the coming 7-year period. The overall expenditure was reduced for the first time since the introduction of the MFF. Furthermore, a smaller share of the budget was allocated for cohesion policy and agriculture. This paper has aimed at explaining the conundrum of how significant changes could be adopted in a setup, which is characterised by a strong status quo bias through the theoretical lens of actor-centered institutionalism.
The member states in ‘Friends of Better Spending’ were crucial for the reduction in total expenditure level. The coalition had a high preference intensity on this conflict dimension; it included leading actors; and given the domestic constraints, the governments of Germany and the United Kingdom entered the negotiation table with win-sets that did not allow a larger budget. The economic crisis and tight budgets in the individual member states had the effect of changing preferences towards lower spending. ‘Friends of Cohesion’ and ‘Friends of Agriculture’ had less intense and less homogenous preferences regarding the level of overall spending, as these coalitions were more focused on – and united by – the issue of allocation of funds between headings.
With the reduction in overall expenditure level, and the ‘Friends of Better Spending’s’ general preference for appropriations to research and innovation, the appropriations for agriculture and cohesion were put under pressure. The first column of the CAP was cut by 13.7% compared with MFF 2007–13. ‘Friends of Agriculture’, led by France, avoided a further reduction in agricultural subsidies. The coalition had a relatively high degree of preference intensity and was able to accept increasing convergence in the allocation of aid per hectare, as France, Spain, and Ireland are all close to the EU average on this parameter.
The member states in ‘Friends of Cohesion’ did not succeed in avoiding reductions in the expenditure for cohesion. The heading was cut by 8.4% compared with MFF 2007–13. This happened even though the coalition counted a great number of veto players and had a high intensity of preference in this particular issue. However, the institutional set up combined with the preferences of these member states made it difficult for the coalition to negotiate as a unit, because the individual member states had incentives to free ride on each other’s bargaining for more cohesion funds, while themselves bargaining for a greater share of the appropriations. This scenario can be expected to recur in future MFF-negotiations. Furthermore, the member states in ‘Friends of Cohesion’ had weak incentives to block an agreement, because this would have delayed the implementation of cohesion funds in the coming period. This lowered the credibility of their formal veto powers, consequently causing weaker negotiation positions. The deviations from the status quo were thus not only the result of changed political and economic factors, but also indicated that the group of net contributors had, to a large extent, forced through its preferences as was also the case for MFF 2007–13 (Schild, Reference Schild2008: 545).
This study has a number of implications. First, the unequal dependency on funds from the EU’s budget can be expected to cause difficult bargain conditions for member states that incur serious losses from delay. The classic option to compensate possible losers through side payments enabled an agreement that deviated from the status quo. Second, the Commission did not appear to have played a role as policy entrepreneur to the extent that it has done in former MFF negotiations. In contrast, future studies should take into account the President of the European Council’s role as broker, and thus the person’s agenda-setting power at the expense of the Commission. Third, while this paper has focused on the intergovernmental negotiations in the European Council in February 2013, studies could examine the subsequent negotiations in the EP leading to the agreement that was passed by the plenary in early November 2013. In that way an additional piece will be added to the puzzle of how an agreement was reached despite a strong status quo bias.
Acknowledgements
The authors would like to thank Anders Woller Nielsen and Benjamin Carl Egerod for valuable research assistance and the three reviewers for comments that have significantly improved the article. The standard disclaimer applies.
Supplementary Material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S175577391500020X
