Preference cohesion and bargaining satisfaction among Southern EU member states: a comparative perspective

AbstractTo what extent does being aligned in EU negotiations predict favourable outcomes for the Southern Euro countries (Italy, Spain, Portugal, Greece) in Council of the EU negotiations? This article attempts to answer this question by leveraging data from the DEU (Decision-Making in the European Union) dataset, and comparing this group of countries to two other well-established alliances at the EU level: the Nordics and the Visegrad Four. Results indicate that the Southern alliance generally lacks cohesion compared to the other groups, but when it is able to find a coherent position, it becomes more effective in obtaining its members’ desired legislative outcomes. Preliminary evidence also suggests that this effect occurs because aligning gives groups of countries more voting power, and because their social capital makes them more effective in negotiations.Keywords: European Unionlegislative bargainingCouncil of the EUbargaining satisfaction Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 This study, in accordance with Arregui’s (Citation2016) arguments, uses the terms ‘bargaining satisfaction’ or ‘bargaining utility’ as opposed to ‘success’.2 Given that we are interested in legislative bargaining, it is natural for us to focus on the systemic level, but readers should be aware that Peterson (Citation1995) identifies two further levels of decision making in the EU, i.e. the ‘super-systemic’ (which includes history-making decisions shaping the procedures and institutions of the EU) and ‘sub-systemic’ (which includes the more technocratic decisions about implementation) levels. The analyses from this article should not be applied to these other levels, as dynamics are likely to be very different.3 To our knowledge, no contribution to date has demonstrated that the Southern countries obtain especially negative outcomes in the Council. However, there is some evidence of a negative reputation for some Mediterranean MSs, usually obtained through expert interviews (e.g. Bailer Citation2004). It could also be speculated that the Southern countries under-perform in terms of network capital (see Naurin Citation2007 or Naurin and Lindahl Citation2008), but the evidence is tenuous.4 Importantly, preference alignment and cooperation are not one and the same. Indeed, having similar preferences does not necessarily imply cooperating (or in other words actively working together) to reach a mutually beneficial outcome. On the other hand, actors may, due to power politics considerations, elect to cooperate with others that they are not necessarily fully aligned with. Our three groups of countries constitute ‘preference’ or ‘policy’ networks, as they are found to often have similar positions on issues (e.g. Thomson Citation2009), but it has to be said that they also tend to indicate each other to be important cooperation partners (e.g. Naurin and Lindahl Citation2008). They can therefore be considered true alliances.5 These are: Agrifish (agriculture and fisheries), Compet (internal market, space, industry and research), Ecofin (economics and finance), Envi (environment), Epsco (employment, social and consumer protection, health and equal opportunities), EYCS (education, youth, culture and sport), FAC (foreign affairs), GAC (general affairs, dealing with institutional rules, co-ordination and issues crossing various policy areas), JHA (justice, home affairs and civil protection) and TTE (transport, telecommunications and energy).6 Some studies (e.g. Aksoy Citation2010) also take the position of the status quo into account when measuring bargaining utility. This approach, however, has been shown to make predictive models significantly less accurate (Thomson et al. Citation2006; Thomson Citation2011), and we therefore follow the majority of the literature in operationalising bargaining utility as the absolute distance between preferences and outcomes.7 Due to space concerns, we do not include all details about the operationalisation of control variables in the main text, though readers should turn to Table A1 for more information and descriptive statistics. It is however important for readers to be aware that all variables are taken at the alliance-issue level, with the aggregation procedure the same as the one used to construct the dependent variable.8 We do not report results for specifications including less controls in the main text, though the simple model without controls can be found in the Online Appendix (see Table A3).Additional informationFundingThis work was supported by the Spanish Ministry Science and Research under Grant PID2020-119716GB-I00; it was also supported by the Erasmus + programme of the European Union under Grants 101085465 – BACES and 101047889 – EU-GOV (Jean Monnet Chair in European Governance).Notes on contributorsNicolas BicchiNicolas Bicchi is a PhD student in the Department of Political and Social Sciences of the Pompeu Fabra University in Barcelona. His main research areas are legislative bargaining and elite behaviour in the European Union and the effect of automation on the European political landscape. [nicolas.bicchi@upf.edu]Javier ArreguiFrancisco Javier Arregui Moreno is an Associate Professor of Political Science at Pompeu Fabra University. He has been awarded a Jean Monnet Chair in EU Governance (2016–2018, 2019–2021 and 2021–2024). He is also the Head Director of the Barcelona Centre for European Studies (BACES). His main research areas are the political process and the policy making that take place in the EU, as well as the analysis of European public policies designed and approved in Brussels and their implementation in Member States. [javier.arregui@upf.edu]