14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License JCMS 2022 Volume 60. Number 2. pp. 427–444 DOI: 10.1111/jcms.13244 Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making ANA MAR FERNÁNDEZ PASARÍN and NURIA FONT Department of Political Science and Public Law, Universitat Autònoma de Barcelona, Barcelona Abstract Do Member States behave inconsistently along the EU policy life cycle? Existing research has extensively examined patterns of conduct in the Council and in comitology. However, the question of whether national governments’ behaviours differ along both stages of EU policy-making is under-explored. This article examines whether and under what conditions voting outcomes along the Council–comitology cycle differ. Based on a dataset combining voting records in the Council and in comitology, the study tests hypotheses derived from delegation theory and demonstrates that legislative contestation, changing legislative preferences, and stringent delegation procedures increase inconsistency. Moreover, it demonstrates that inconsistency increases when there is contestation in the Council and changes in national governments’ preferences between the two votes increase. Ultimately, inconsistency is associated with the reversal of Council contestation in comitology. Keywords: council; comitology; delegation EU decision-making Introduction How inconsistent are national governments’ behaviours along the European Union (EU) policy life cycle? To what extent are member states’ decisions stable along the legislative–executive stages of EU decision-making? Existing research on EU institutions has extensively examined modes of deliberation and patterns of conduct in the Council of the European Union (hereinafter, the Council). Recent work has revealed that while decision-making in the Council is highly consensual at the voting stage, higher levels of contestation are found in early stages of negotiations (Hobolt and Wratil, 2020) and that unanimous outcomes may reflect a blame-avoidance strategy (Novak, 2013). This debate has been extended to national representatives’ behaviours in comitology. Related research has addressed the issue of whether committees that imitate the territorial and sectorial composition of the Council at the executive level actually reproduce the institu- tional dynamics that drive the Council’s decisions during the legislative process (Joerges and Neyer, 1997a; Pollack, 2003; Blom-Hansen, 2013; Dehousse et al., 2014). Our aim in this article is to reinforce the bridge between these two research agendas by posing the question of whether and under what conditions voting results adopting legislation in the Council and those on draft measures presented by the Commission in comitology to implement this legislation are inconsistent. In line with previous studies that have underlined the usefulness of adopting a multi-level approach to the analysis of EU policy-making (Burns and Tobin, 2020), our contribution goes beyond the unpacking of a single institution at the legislative or executive phase and tests the conditions under which the likelihood of discontinuity in the legislative–executive sequence of © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 428 Ana Mar Fernández Pasarín and Nuria Font intergovernmental decision-making increases. In doing so, it adopts a comprehensive approach to the study of EU policy-making with the objective of offering valuable insights into the question of how inconsistently member states behave when elaborating EU policies. The present study stands at the crossroads between two bodies of literature. First, it engages with research on EU policy-making highlighting the pertinence of exploring why governments change their positions between the deliberation and the voting stages of policy-making (Mattila, 2004; Novak, 2013; Dehousse et al., 2014; Hobolt and Wratil, 2020). Our concept of inconsistency draws inspiration from Hobolt and Wratil’s (2020) study identifying patterns of government consensus and contestation along Council deliberations and decision-making. Inconsistency is conceptualized in our study as the combination of intergovernmental consensus and contestation along legislative-executive sequences. More specifically, our notion of inconsistency captures changes in recorded voting results between the Council and comitology. Although such changes do not necessarily reflect shifting policy content or coalition patterns, our approach is relevant as it does not only complement previous separate studies on behavioural patterns in the Council and in comitology but also helps provide insight into EU decision-making dynamics. In this respect, while an important strand of research has argued that consensus is the general pattern of conduct at both levels of decision-making (Mattila, 2004; Heisenberg, 2005; Hayes-Renshaw et al., 2006; Blom-Hansen, 2014; Hagemann et al., 2017), other studies identifying different degrees of disagreement both in the Council (Thomson and Torenvlied, 2011; Hobolt and Wratil, 2020) and in comitology (Dehousse et al., 2014; Fernández-Pasarín et al., 2021) would challenge this view. The present article shows that inconsistent voting results occur more often than not, and it raises the question of what drives inconsistency. Second, this study builds on and derives hypotheses from theories on delegation in the US (Epstein and O’Halloran, 1999; Huber and Shipan, 2002) and related research in the EU context (Majone, 2001; Franchino, 2007; Thomson and Torenvlied, 2011; Brandsma and Blom-Hansen, 2016; Yordanova and Zhelyazkova, 2019). We demonstrate that legislative contestation in the Council, changing government preferences over time, and stringent comitology proce- dures increase the likelihood of inconsistency. Moreover, we show that inconsistency increases when there is contestation in the Council and when changes in national govern- ments’ preferences between the two votes increase. Ultimately, inconsistent outcomes are associated with the reversal of Council contestation at the comitology stage. The current study is based on a novel dataset merging data on Council voting records on the adoption of legislation under the Ordinary Legislative Procedure (hereinafter referred to as codecision) with the associated voting results in comitology during the 2009–18 period. The dataset comprises 2,726 comitology votes that refer to 184 Council votes adopting legislation. As under codecision the Council adopts legislation by qualified majority voting (QMV), focusing on this legislative procedure allows both unanimous and contested vote outcomes in the Council to occur and, hence, inconsistency along the Council–comitology sequence to be observed. The article is structured as follows. The first section identifies a literature gap and grounds the study in previous theoretical debates. Section II presents the theoretical framework and main hypotheses. Section III describes the data and methods employed in the study. Section IV presents the results of the analysis. The article concludes with © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 429 several remarks about the findings, their theoretical and normative implications, and suggestions for a further research agenda. I. Consensus and Contestation The debate on how national governments act when participating in the EU political system has given rise to an extensive corpus of literature partly focusing on the prevailing modes of deliberation and decision-making in the Council. In this vein, whether empha- sizing the intergovernmental bargaining dynamics or the deliberative supranationalism at work, an important stream of research has shown that a consensus-oriented culture prevails and that there is generally an absence of explicit opposition to legislative pro- posals presented by the Commission (Heisenberg, 2005; Hayes-Renshaw et al., 2006; Novak, 2013). Consistent with this, research on voting records in the Council has revealed that governmental opposition is rather rare (Hagemann et al., 2017). Nevertheless, several contributions have shown that consensus at the voting stage may be chosen as a blame-avoidance strategy (Novak, 2013) and may be preceded by contes- tation during the negotiation stage (Hobolt and Wratil, 2020). Along a different line of reasoning, other approaches have identified changes in the prevailing modes of decision-making from an explicit ‘consensus-driven’ approach to one where governments are less reluctant to express opposition (Naurin and Wallace, 2008; Hagemann and Franchino, 2016). More specifically, recent studies have shown that the percentage of op- position votes in the Council has increased over time (Hagemann and Franchino, 2016). In recent years, these core theoretical debates on decision-making in the Council have been extended to comitology (Héritier et al., 2013; Dehousse et al., 2014; Brandsma and Blom-Hansen, 2016). Actors in comitology committees differ from those in the Council in that they have an administrative rather than a political profile, the issues they deal with are highly technical, and experts’ attitudes are presumably more driven by scientific con- cerns than by political considerations. Yet, similarly to actors in the Council, committee members are national representatives invested with the mission of ensuring intergovern- mental control in EU policy making. In other words, the profiles and day-to-day tasks of actors in the Council and in comitology do not coincide, but the institutional design assigns them an essentially similar duty regarding the representation of member states’ interests in the EU structure. This has raised the question of how national representatives behave when supervising the implementing powers delegated to the Commission. Different interpretations have been formulated regarding this question, mainly influenced by intergovernmental and deliberative supranationalism theories. Thus, whereas some scholars argue that comitology committees primarily function as guardians of member states’ interests and carefully control the Commission’s use of its executive powers (Pollack, 2003), others maintain that this formal control is weakened in practice by at least two sets of factors: first, the epistemic community dynamics at work in experts’ meetings, which tend to privilege the quality of scientific arguments over strategy and, thus, deliberative over bargaining modes of interaction (Joerges and Neyer, 1997b); and second, the Commission’s interest in protecting its autonomy, its strategic anticipation of conflict with the Council, and its role as an informal agenda-setter in the day-to-day management of the committees’ work (Ballmann et al., 2002; Craig, 2016). Various studies have expanded on the idea that similar to what happens in the master institution, © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 430 Ana Mar Fernández Pasarín and Nuria Font consensual dynamics are the prevailing mode of decision-making in executive commit- tees. Contestation is not absent but anticipated and scaled back during negotiations at the Commission, which are rarely made explicit in voting by national representatives and, thus, are masked by the final outcome of comitology’s deliberations, which tend to be highly consensual (Ballmann et al., 2002; Dehousse et al., 2014). In sum, both literatures would lead us to expect consistency regarding voting results in both the Council and the comitology committees. In line with intergovernmental studies, work in executive committees would be a continuation of the bargaining dynamics observed in the Council (Blom-Hansen, 2011); its main feature would be consensus, and consequently, voting outcomes between the two institutional momentums should co- incide and be generally characterized by the absence of explicit opposition. This article aims to gain insight into this question by identifying and providing accounts for discon- tinuity in voting outcomes along the Council-comitology sequence of intergovernmental decision-making. Empirically, we expect two different patterns of inconsistency to be observed: on the one hand, a situation in which a unanimous decision in the Council is followed by a contested decision in comitology; on the other hand, a scenario in which a contested vote in the Council is followed by a unanimous position in comitology. Our empirical expectation is that inconsistency is more frequently observed when there is con- testation in the Council, as comitology is assumed to seek consensus-building, either by reversing explicit dissensus in the Council or by preserving minister-level consensus. In what follows, we elaborate our theoretical expectations on drivers of inconsistency. II. Drivers of Inconsistency Literature on delegation at the EU level and, more specifically, on comitology has shown that although highly specialized and with a less prominent political aura, comitology com- mittees are like ‘Councils in miniature’ that imitate the functional and decisional dynam- ics of their master institution (Blom-Hansen, 2013). This image of comitology as an extension of the Council at the regulatory level seems sound at first glance if we consider that the former was created by the latter with the concrete mission of controlling in its name the powers delegated to the Commission in the executive domain. The idea of comitology as the long arm and supervisory eye of the Council at the implementation stage would be consistent with the general argument in delegation theory that legislative actors delegate implementing powers to executive agents closer to their preferences (Thomson and Torenvlied, 2011). Hence, similar patterns of intergovernmental behaviour at the two levels of decision-making would be expected. However, this expectation is far from straightforward. We hypothesize that inconsis- tency is more likely to occur when there is contestation in the Council. We derive our expectation from research on transactional costs delegation theory (Brandsma and Blom-Hansen, 2016; Yordanova and Zhelyazkova, 2019) suggesting that member states delegate powers to the Commission as a means of reducing decision-making costs (Blom-Hansen, 2008) and enhancing policy efficiency (Majone, 2001). As legislators have limited expert knowledge, they rely on the intervention of executive actors in areas where they face uncertain consequences from legislative policy decisions (Ballman et al., 2002; Franchino, 2004). Hence, national governments would be incentivized to reach minimal consensus agreements at the Council and delegate the management of potential © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 431 conflict to more technical arenas such as comitology committees. Taking this into account, consensus in the Council does not necessarily translate into conflict in comitology, as legislative consensus would incentivize executive bodies to make policy adjustments within the boundaries of policy consensus. This expectation could be held even if unanimity in the Council is considered to reflect a blame-avoidance strategy (Novak, 2013) as governments that reach a unanimous position at the end of a deliberation process are assumed to have resolved the most divisive issues and enabled the conditions for consensus in comitology. Our expectation in the event of a contested vote in the Council is that inconsistency between the Council and comitology voting outcomes is more likely than consistency. While some studies on delegation in the EU would lead us to expect that policy conflict at the legislative stage may reappear at the executive stage (Crombez and Hix, 2015), our expectation is derived from a general argument in delegation theory indicating that the presence of policy conflict at the legislative stage allows executive agents to operate under lower constraints (Huber and Shipan, 2002). Arguably, if contested policy issues were delegated to the sphere of technical expertise, comitology committees would have incentives to use their expert knowledge and maximize their discretionary powers to enhance policy efficiency and facilitate consensus; that is, executive agents’ discretionary power may operate as a consensus facilitator. An additional argument behind our expec- tation highlights the role of the Commission. We assume that the Commission’s search for an alignment of preferences and general agreement is driven not only by its reputation as an efficient and honest broker, but also by its search for autonomy (Craig, 2016) and empowerment as an institutional actor (Ballmann et al., 2002; Moury and Héritier, 2012). H1 Inconsistency between voting results in the Council and in comitology is more likely to occur when there is contestation in the Council. Classical debates in delegation literature have often highlighted that one of the main reasons for delegation is the need for credible commitments (for example, Majone, 1996; Gilardi, 2002). The core argument is that legislators delegate regulatory powers to independent bodies to secure credible commitments and, accordingly, the consistency of current policy preferences in the long term. However, empirical research also reveals that certain institutional conditions are associated with varying degrees of agency discre- tion (Gilardi, 2002; Huber and Shipan, 2002). In this respect, Huber and Shipan (2002) found a negative relationship between cabinet turnover and the level of detailed legisla- tion. As argued, increasing cabinet turnover reduces ministerial opportunities to control policy in the future. Turning to the EU, we know from previous studies that governments change their positions over time (Finke, 2017), whether because of changes in the composition of cabinets and government turnover (Häge, 2007) or, more generally, because of shifting political and economic contexts at both the domestic and global levels (Hagemann and Franchino, 2016). Yet, the effects of member states’ changing governments on policy outcome at the comitology stage are less known. If national governments’ motivations change over time, changing preferences along the legislative-executive stages of policy-making are also expected to occur. We hypothesize that changes in national governments’ preferences that occur between the vote adopting legislation in the Council and the associated vote on implementing proposals drafted by © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 432 Ana Mar Fernández Pasarín and Nuria Font the Commission in comitology increase the probability of inconsistency. The argument is that voting behaviours by national governments in the Council are more easily reproduced in comitology if national preferences at this stage are more similar than dissimilar to the ones at the ministerial level. H2 Inconsistency between voting results in the Council and in comitology is more likely to occur as changes in national governments’ preferences increase. Furthermore, we hypothesize that the impact of high government turnover on inconsistency varies depending on whether or not the Council adopts legislation unanimously. We expect that given contestation in the Council, the probability of unanimity in comitology is expected to increase as changes in national policy positions increase. The argument is that more changes in national preferences would make it less difficult for the Commission to maximise its problem-solving capacity and for committees, in general, to reverse Council contestation. In other words, changing national preferences would facilitate committees to unlock disagreements and reverse minister-level contestation. Considering this, it is debatable whether inconsistency would be more likely when government preferences change and the Council reaches unanimity. While one could argue that changing preferences would make it more difficult to maintain consensus in comitology committees, it is also reasonable to expect that the conditions needed to reach consensus are more favourable when consen- sus is already built into the Council. H3 Inconsistency between voting results in the Council and in comitology is more likely to occur as changes of national governments’ preferences increase and there is contestation in the Council. Previous studies have found that institutional actors tend to privilege certain comitology procedures, providing them with the maximum of individual control over delegated competences (Blom-Hansen, 2014). Thus, whereas the Council conceives comitology as a way of ensuring the presence of member states at the implementation stage and, hence, prefers strict procedures in order to benefit from delegation without un- duly empowering the executive (Blom-Hansen, 2013), the Commission dislikes a system that undermines its autonomy and prefers less-constraining procedures (Brandsma and Blom-Hansen, 2010, 2016; Blom-Hansen, 2014; Craig, 2016). In accordance with the in- stitutional design of comitology, the legislator (the Council and the European Parliament under co-decision) is the authority in charge of selecting the type of procedure and, thus, the degree of national control exerted on the implementing powers delegated to the Commission. Since the 2011 comitology reform (Regulation 182/2011), comitology is categorized into advisory and examination procedures. The main differences between the two involve the majority required for the adoption of the draft implementing proposal presented by the Commission and its binding effect on the Commission. Whereas a simple majority is required under the advisory procedure, a qualified majority is needed in the examination procedure. Furthermore, while the advisory procedure is merely consultative, examination has a more constraining effect on the Commission’s room for manoeuvre in that it can involve the intervention of an Appeal Committee in the event © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 433 of rejection by national representatives in the first instance. Moreover, the regulatory with scrutiny procedure (RPS) established in the 1999 Comitology Decision – which also re- quires a qualified majority and gives veto rights to the Council and the EP even in the event of a favourable opinion delivered by the committee – is still applied for those legal acts referring to it that were adopted up until 2013. The choice of the comitology procedure is goal-oriented and may influence voting results (Dehousse et al., 2014; Fernández-Pasarín et al., 2021). Dehousse et al. (2014) and Fernández-Pasarín et al. (2021) have shown that although contestation can be found in all categories, dissent appears to be more frequent and stronger in committees when the procedures are more constraining for the Commission. As indicated by Brandsma and Blom-Hansen (2010), in some committees, the Commission can be very active and act as both a mediator and a policy advocate. Member States can, thus, carefully delegate implementing powers to the Commission (Blom-Hansen, 2013). In addition, the Commis- sion tends to adapt its behaviour depending on the formal rules at work (Brandsma and Blom-Hansen, 2010). The qualified majority required in both examination and RPS procedures results in the Commission working under narrower constraints. At first glance, consistent Council-comitology voting outcomes could be expected to be a dominant pattern under more constraining procedures. However, our argument goes in the opposite direction. While debatable, comitology procedures imposing higher constraints on the Commission (that is, examination and RPS) are expected to increase the probability of inconsistency compared to advisory procedures. The argument assumes the rationalist idea that more- constraining procedures are chosen when the Council aims at retaining maximal control and reducing uncertainty over the use of delegated powers (Pollack, 2003; Blom-Hansen, 2014). This situation typically occurs when sensitive issues are at stake in comitology cases and the preferences of the Commission differ from those of the Council (Blom-Hansen, 2014). Our expectation is that when national stakes are high and preference divergence is anticipated, in which case stricter procedures are chosen, difficulties over negotiations in comitology may be more easily exacerbated and, as a result, uncertainty over the outcomes would increase. We hypothesize that when stricter procedures are chosen, uncertainty over voting outcomes increases, and this may, in turn, affect inconsistency. H4 Inconsistency between voting results in the Council and in comitology is more likely to occur when the comitology procedure is stricter for the Commission. Moreover, the effect of the comitology procedure on inconsistency is expected to vary depending on whether there is disagreement in the Council. On one hand, given unanim- ity in the Council, the Commission may face more difficulties maintaining consensus when national stakes are high, and, accordingly, the comitology procedure (examination or RPS) would be highly constraining compared with the less-demanding advisory procedure. On the other hand, in the event of dissent in the Council, the Commission may have more opportunities to arbitrate between conflicting national positions to unlock critical issues (thus activating its promoting role) and to negotiate consensual implementing proposals when national stakes are lower, and the chosen comitology procedure is thereby less restrictive for the Commission. © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 434 Ana Mar Fernández Pasarín and Nuria Font H5 Inconsistency between voting results in the Council and comitology is more likely to occur when the comitology procedure is less strict for the Commission and there is contestation in the Council. III. Data and Methods Our analysis is based on an original dataset that merges data on Council votes adopting legislation under codecision with the corresponding comitology votes on draft implementing proposals presented by the Commission. We focus on acts adopted under codecision for methodological reasons as the QMV required in the Council allows the observation of both unanimity and contestation. In this way, we can analyse vote incon- sistency, which is our dependent variable. In the case of comitology, we focus on commit- tees’ opinions delivered after the enactment of the Lisbon Treaty, which entailed the split of the old system of comitology into the delegated acts (quasi-legislatives, Art. 290 of the Treaty on the Functioning of the EU) and the properly implementing acts (Art. 291 of the TFEU) that remain under the system of comitology. Given that not all legislative acts adopted under codecision give rise to comitology opinions and that a single legislative act may give rise to several draft implementing proposals presented by the Commission, we selected all comitology votes during the period 2009–18 that referred to Council votes adopting legislation from 2006 to 2016. Note that this time period was chosen because 2009 is the first year with annual data on comitology votes in the new register of comitology that was established in April 2008. We then matched comitology votes with their corresponding Council votes. The data set covers 2,726 comitology votes on draft-implementing measures referred to 184 Council votes adopting legislation. Data were retrieved from the Council Open Data website and the Comitology Register. Note that data for 88 Council votes were not available, which may raise the question of whether the sample is representative of the entire population of Council–comitology pairs. For this reason, our results need to be interpreted considering these data limitations (see A1 in the Appendix for a more detailed description of data collection). This time period was chosen because 2009 is the first year with annual data on comitology votes in the new register of comitology that was established in April 2008. Our dependent variable is inconsistency between voting outcomes in the Council and in comitology, and our unit of analysis is a Council–comitology pair. It is worth noting that votes in the two arenas differ regarding their origin and content. In the case of the Council, votes are on legislative proposals defining policies, whereas in comitology, votes are on proposals for implementing measures that derive from the specific delegation pro- vision in the basic legal act. While we acknowledge this difference, we consider that the question of inconsistency between the two instances is relevant because comitology has been designed as an extension of the Council’s arm at the executive level. Committees in- tervene following a mandate of the Council. They represent the member states’ interests and apply the same rules for the adoption of their decisions. For operationalization, we take the aggregate voting results both in the Council and in comitology as disaggregated data on votes by member states in comitology are not made public. These data constraints make it impossible to compare individual votes and thus need to be considered when interpreting the results. Our dependent variable inconsistency © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 435 takes the value 1 when the voting outcome is unanimity at one stage of the policy and contestation at the other. It takes the value 0 when both voting outcomes are coincident. Note that our operationalization does not focus on the structure of contestation but on whether or not unanimity expressing the support of all member states has been reached. We acknowledge that this operationalization provides a simplified view of how member states behave as inconsistency can occur in quite different situations of contestation. Yet, as a previous study on comitology demonstrates that in practice extreme diversity is rare and that in general when there is contestation, this opposition is weak (Fernández-Pasarín et al., 2021), we consider it to be useful in that it allows visualization of four unanimity–contestation scenarios. By unanimity in the Council, we mean instances of votes in which legislative pro- posals are supported by all national governments attending the Council meeting. Note that the absence of Member States does not necessarily prevent unanimity as states are as- sumed to agree with the legislative proposal unless they express their dissent in writing (tacit agreement). As abstentions may be resorted to by governments to avoid taking ex- plicit positions, we assume that abstentions do not reflect straightforward support for the legislative proposal. Moreover, as the rule at the Council level is that abstentions under qualified majority voting rules count as votes against and that the rule of the Council is extended to comitology for the calculus of the QMV (which is applicable to the stricter regulatory procedures examination and RPS), abstentions are coded as opposition votes in comitology too. In the case of the merely consultative procedure –advisory – the voting rule requires only a simple majority. The question of whether abstentions should be coded as favourable when a simple majority applies is discarded as our dataset contains only one case of advisory procedure with abstentions and without votes against (in which case the coding of our dependent variable would change). Most comitology votes in our study resulted in favourable opinions (96.4 per cent), with only a few resulting in no opinion (3.5 per cent) or in an unfavourable opinion (0.1 per cent). Data on voting results were not available for three favourable opinions and for the two unfavourable ones. No-opinions and unfavourable opinions reflect that the required majority for the adoption of the Commission’s draft implementing proposal in comitology committees was not reached, a fact that obviously involves the presence of contestation. Table 1 summarizes the frequency of voting outcome combinations in the Council and in comitology committees. As shown, inconsistency between Council and comitology voting is observed in 55 per cent of cases, the majority of which reproduce a contestation – unanimity pattern (43 per cent, with the inverse pattern of inconsistency representing 12 per cent). Consistent voting outcomes are observed in 45 per cent of cases, with a Table 1: Unanimity and Contestation in Council–Comitology Vote Outcomes (% in Parentheses) Comitology Unanimity Contestation Council Unanimity 751 (28) 327 (12) Contestation 1,185 (43) 463 (17) © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 436 Ana Mar Fernández Pasarín and Nuria Font predominance of maintained unanimity in the two voting stages (28 per cent) and a low frequency of continued contestation (17 per cent). We operationalize independent variables as follows. Contestation in the Council is operationalized with a binary variable taking the value 1 when the Council does not adopt legislation unanimously (there are votes against and/or abstentions) and 0 in case of a unanimous result. It must be noted that this independent variable reflects a voting out- come in the Council, whereas our dependent variable indicates whether there is a line of continuity in the sequence of intergovernmental decision-making along the legislative-executive stages of EU policy-making. Changing government preferences is operationalized as the number of changes in government involving a change in the leading party in the Member States between the vote in the Council and the vote in comitology. We assume that national policy preferences in a given period are more likely to change when the leading party in government changes than when not. Note that changes of government in Croatia are computed only for those votes in the Council that followed accession. Data are retrieved from ParlGov (Döring and Manow, 2019). A description of the variable government changes is included in point A3 of the Appendix. Other than this, while our operationalization allows determining whether the overall voting results in the Council and in comitology are inconsistent, the absence of individual data on comitology votes prevents us from providing a more nuanced picture of how government turnover affects inconsistency. Finally, a categorical variable indicates whether the comitology procedure is advisory, examination or RPS. Advisory procedure is taken as the reference category. Several controls are included. As the study focuses on legislative acts adopted under codecision, this involving that the EP and the Council are colegislators, we control for whether legislative bargaining between both institutional actors went smoothly or was controversial. For operationalization, inter-institutional controversy regarding acts adopted at first reading is assumed to be lower compared to those adopted at second or third reading. A binary variable takes the value 1 for first reading and 0 otherwise. While debatable, inter-institutional controversy might in principle increase inconsistency. We control for the type of legislative act by including a binary variable taking the value 1 for directives and 0 otherwise. The expected effect of directives on inconsistency is uncer- tain. While directives leave the adoption of implementation measures to the member states, it is far from clear whether national governments adopt more flexible bargaining approaches when negotiating directives as some directives may specify the ends to be attained in considerable detail. Based on recent studies identifying an association between complexity and delegation (Brandsma and Blom-Hansen, 2016; Yordanova and Zhelyazkova, 2019), we control for policy complexity. We use the number of committees consulted in the EP as an exogenous measure of policy complexity, here referring to the degree an issue cuts across several policy areas. We assume that more committees consulted reflects higher degrees of complexity (Reh et al., 2013). The number of recitals included in the legislative proposal is employed as an alternative measure. Note that neither measure may capture technical complexity and therefore do not necessarily reflect the technical dimension inherent in the delegation rationale. The expectation is that complexity increases inconsistency. We also control for whether the legislative act intro- duces new legislation or amends or recasts existing legislation (new = 1, otherwise = 0). Since new legislation covers a wider set of policy issues, as a result of which national © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 437 stakes are assumed to be higher compared with amending legislation, one possible expectation is that new legislation increases inconsistency. Yet, an opposite effect could occur if amending legislation covers issues with high stakes for the member states. Whether the legislative act is adopted in the post-Lisbon period is controlled for (after Lisbon = 1, otherwise = 0). Given the generalized involvement of the EP as co-legislator with the Council for most policy areas following the Lisbon Treaty, being post-Lisbon might increase inconsistency compared with being pre-Lisbon. If new and more- sensitive policy areas fall under codecision after Lisbon, contestation in the legislative stage is expected to emerge or re-emerge at the comitology stage after Lisbon. Table A1 in the Appendix summarizes the descriptive statistics of the dependent, independent, and control variables. Note that problems of collinearity are discarded as VIF coefficients are under 2 in all models. IV. Findings Given the multilevel structure of our dataset, in which comitology votes are nested with legislative acts, we perform multilevel regressions with crossed random effects for legislative acts. As we employ a binary dependent variable, we perform crossed random Table 2: Multilevel Logistic Regression on Inconsistency (Odds Ratio) Independent Variables M1 M2 M3 M4 Contestation Council 8.4*** (2.89) 7.71*** (2.66) 6.3*** (4.32) 5.99*** (4.13) Government turnover 1.02*** (0.01) 1.02*** (0.01) 1.01* (0.01) 1.01** (0.01) Contestation Council * 1.02** (0.01) 1.02** (0.01) Government turnover Comitology procedure (Ref. = Advisory) Examination 1.78** (0.49) 1.83** (0.5) 1.72* (0.54) 1.77* (0.55) RPS 2.07** (0.72) 2.49** (0.9) 2.71*** (1.05) 3.18*** (1.28) Contestation Council * Comitology procedure Examination 1.19 (0.77) 1.17 (0.72) RPS 0.17* (0.16) 0.18* (0.17) Controls First reading 0.77 (0.27) 0.78 (0.27) Directive 0.98 (0.41) 1.05 (0.44) EP committees 0.93 (0.05) 0.94 (0.05) New 1.02 (0.38) 1.03 (0.38) Post-Lisbon 1.88 (0.72) 1.75 (0.67) Constant 0.15*** (0.05) 0.17*** (0.11) 0.17*** (0.06) 0.18*** (0.12) Random effects Estimate (se) Estimate (se) Estimate (se) Estimate (se) Legislative file 2.61 (0.54) 2.45 (0.53) 2.53 (0.53) 2.41 (0.52) Log likelihood 1362.92 1360.5 1356.68 1354.68 AIC 2737.85 2743.01 2731.37 2737.37 BIC 2773.31 2808.02 2784.56 2820.12 N 2,726 2,726 2,726 2,726 *** p < 0.01, ** p < 0.05, * p < 0.1. Standard errors in parentheses. © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 438 Ana Mar Fernández Pasarín and Nuria Font multilevel logistic regressions. We estimate four models. Model 1 includes the indepen- dent variables, and Model 2 adds the controls. Model 3 includes the two interaction terms, while Model 4 incorporates the controls. Results are reported as odds ratio and are shown in Table 2. Consistent with H1, the results show that contestation in the Council significantly increases the probability of inconsistency between voting outcomes in the Council and in comitology. Based on Model 1, we calculate the predicted probability of inconsistency for Council contestation, keeping all other variables constant. The predicted probability decreases from 0.76 in case of contestation to 0.27 in case of consensus (see Figure 1 summarizing the predicted probability of all independent variables on inconsistency). As expected, when ministers solve contentious issues at the legislative level, comitology committees are incentivized to make policy adjustments within the boundaries of consen- sus. Moreover, when issues have been contested in the Council, reaching a consensual comitology outcome is more likely than is maintaining contestation. Our interpretation is that when contested policy issues are delegated to executive agents, comitology committees have incentives to optimise expert knowledge and make use of their discretionary powers to improve policy efficiency and facilitate consensus. Moreover, our interpretation is suggestive of the Commission’s incentives to make the most of its Figure 1: Predicted Probability of Inconsistency, 95% CIs. [Colour figure can be viewed at wileyonlinelibrary.com] © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 439 problem-solving capacities with the aim of either sustaining consensus or reversing con- testation in the Council. The results confirm H2 on the effect of changing government preferences on inconsis- tency. They indicate that Council–comitology inconsistency is significantly associated with a higher government turnover between the Council and the comitology votes. Based on Model 1, we calculate the predicted probability of inconsistency at different levels of government change, keeping all other variables constant. As shown in Figure 1, the pre- dicted probability of inconsistency increases from 0.45 in the absence of government turnover to 0.74 at the maximum level of government changes. We acknowledge that the absence of individual data on comitology votes prevents us from observing the micro-mechanisms through which individual government turnover increases inconsis- tency. That said, our overall interpretation of the findings is that voting behaviour in the Council is more easily replicated in comitology when national preferences at the execu- tive stage are more similar than dissimilar to the ones at the legislative stage. In other words, the higher the number of government changes and the assumed increase of national preference change, the greater the difficulties committees will face in addressing government preference discrepancies. Accordingly, committees will be less likely to reflect initial Council preferences. Figure 2: Predicted Probability of Inconsistency: Combined Effects, 95% CIs. [Colour figure can be viewed at wileyonlinelibrary.com] © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 440 Ana Mar Fernández Pasarín and Nuria Font Regarding H3, the interaction between Council contestation and government turnover attains statistical significance in Models 3 and 4. In order to better assess the effect of the interaction and based on Model 3, we graphically plot the predicted probability of incon- sistency for contested and unanimous Council results at different values of government turnover, keeping continuous variables at their means and categorical ones at their modes. Figure 2 shows that, in the absence of unanimity in the Council, the expected probability of inconsistency between Council and comitology votes is 0.62 at the minimum level of government turnover and increases to 0.95 at the maximum level. That is, when there is Council disapproval, the predicted probability of inconsistency increases by 33 percent- age points between the minimum and the maximum levels of government turnover. In other words, when votes in the Council are contested, the probability of unanimous results in comitology (that is, inconsistency) significantly increases as government turnover in the member states between the two votes increases. More changes of government between the two votes, involving more changes in national preferences, would make it less diffi- cult for committees to reverse contestation in the Council and reach consensus. That is, higher government turnover would make it less difficult for committees and for the Com- mission in particular to make use of technical and problem-solving capacities to unlock dissenting positions and reverse minister-level contestation. The effect of the type of comitology procedure on inconsistency is in line with our the- oretical expectations as stated in H4. The results show that examination and RPS proce- dures are significantly more likely to increase inconsistency, taking advisory committees as the reference category. Based on Model 1, we calculate the predicted probability of in- consistency for each of the three types of comitology procedure, keeping all other vari- ables constant. As Figure 2 illustrates, the predicted probability of inconsistency for advisory committees is 0.40 and increases to 0.54 and 0.58 for examination and RPS committees, respectively. The results suggest that comitology procedures imposing higher constraints on the Commission (examination and RPS) increase the probability of incon- sistency compared to advisory procedures. Note, however, that H4 cannot be straightfor- wardly confirmed since confidence intervals of advisory committees, as opposed to examination and RPS committees, slightly overlap. Regarding H5 on the interaction between Council contestation and comitology proce- dure, and based on Model 3, we graphically plot the predicted probability of inconsis- tency for the three procedures when the Council voting outcome is either unanimous or contested, keeping continuous variables at their means and categorical ones at their modes. Figure 2 shows that when the Council adopts legislation unanimously, the proba- bility of inconsistency is significantly different for advisory and RPS procedures, moving from 17% to 36%, respectively. Given Council contestation, the expected probability of inconsistency is significantly higher for advisory and examination procedures. No signif- icant difference is found for RPS. Overall, the results partially confirm our theoretical ex- pectations by suggesting that given unanimity in the Council, comitology committees have more difficulty maintaining consensus when the comitology procedure is highly constraining (RPS) compared to the advisory procedure, which is the least demanding. Through a similar rationale, given dissent in the Council, the committees and in particular the Commission have less difficulty optimising expert knowledge and problem-solving capacities to reverse contestation and produce consensus when the comitology procedure is scarcely constraining. Finally, the results are robust when controls are introduced. As © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 441 Models 2 and 4 show, no significant effect on inconsistency is found for any of the con- trols introduced. We perform several robustness checks. All models were run without cases of ‘No opin- ions’ in comitology (N = 2,630) to check whether these cases, where higher contestation by national representatives is assumed, resulted in different outcomes. The findings remain practically unchanged. The regressions were also run using an alternative measure of government change, adding 0.5 if the same party leads two consecutive governments. The results remain virtually unaltered. In regard to the comitology procedure, an alterna- tive binary operationalization takes examination and RPS procedures as those in which national representatives are expected to exert stricter control over the Commission draft implementing proposals. Examination and RPS procedures take the value 1, and advisory committees take the value 0. Finally, we perform all models after adding controls for the policy area by including EP committees responsible for the legislative report. None of the policy areas reach statistical significance. Overall, the results are consistent with our findings. Discussion and Conclusion This article explores the question of whether and under what conditions vote inconsis- tently along the Council-comitology stages of EU policy-making occurs. The article describes patterns of inconsistency and identifies conditions under which member states’ voting outcomes along the EU policy life cycle are inconsistent. Our first empirical finding reveals that voting outcomes along the Council-comitology sequence are incon- sistent more often than not. This finding calls into question the idea that comitology imitates the behavioural patterns in the Council. Second, inconsistency mostly reflects a contestation–consensus pattern, thus showing that comitology is a regulatory arena with the practical capacity to reverse minister-level dissent. Building on delegation theory, we demonstrate that legislative conflict, changes in legislators’ preferences, and strict delegation procedures increase the probability of incon- sistency along the legislative-executive sequence of intergovernmental decision-making. Moreover, one of the main findings of this study is that the effect of changes in national governments’ preferences on inconsistency varies depending on whether dissent was expressed at the Council decision-making level. Arguably, more changes of government between the two votes would make it less difficult for committees and in particular for the Commission to unlock minister-level contested issues and produce a consensual outcome. Another finding is that the type of comitology procedure seems to affect inconsistency. Procedures imposing higher constraints on the Commission (that is, examination and RPS) are associated with higher degrees of inconsistency. Yet, while inconsistency increases when either of the two procedures apply in the case of Council dissent, our general expectations cannot be straightforwardly confirmed. This article makes two main contributions. First, by extending the analysis along the EU policy life cycle, it complements recent research on legislative behaviour in the Council. The study takes a step forward by de-emphasizing the dominant questions in the related literature of whether, to what extent, and under what conditions disapproval permeates intergovernmental decision-making. Instead, it redirects attention to the following questions: (1) how national governments behave with respect to the © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 442 Ana Mar Fernández Pasarín and Nuria Font consensus–contestation tension along the Council–comitology sequence of EU policy- making; and (2) what the main drivers of intergovernmental inconsistency are. Second, by identifying conditions under which inconsistency increases – namely, legislative con- testation, changing legislative preferences, and stringent delegation procedures – the study contributes to a better understanding of how delegation dynamics works in the EU. Further analyses would benefit from complementing these findings with qualitative case studies and from focusing on specific policy areas and changing policy content. In particular, more attention is needed on tracing with higher precision governments’ changing preferences between votes on legislative proposals in the Council and votes on subsequent proposal for implementing measures in comitology. In addition, and in connection with recent work on legislative control over executive law-making (Yordanova and Zhelyazkova, 2019), a comparison of bargaining dynamics between implementing acts and delegated acts would also contribute to improving our knowledge of policy-making in the EU. Our findings have important normative implications in terms of democratic account- ability as they engage with the significant question in the literature of how decisions are made in the Council and in comitology. By addressing the question of member states’ inconsistency along the political and regulatory levels of EU decision-making and by identifying, in particular, the conditions under which comitology, which is a non-principal institutional arena, may reverse Council’s unanimous outcomes when adopting draft implementing proposals or, conversely, may arbitrate contestation, this article engages with more classical normative debates on the role of bureaucrats in EU decision-making and, more generally, on democratic accountability at the meso-level of the EU political system. Moreover, by showing that comitology more often than not operates as a regulatory arena seeking consensus-building, either by preserving minister- level consensus or by reversing explicit disagreement in the Council, the study encourages a normative reflection on the nature of the relationships between delegated powers and consensual decision-making. In particular, our study engages with a debate on the implications of non-majoritarian actors operating as agents of consensus for democratic accountability in a multi-level polity. Funding information This study was supported by the funding of the Spanish Ministry of Economy and Competitiveness (MINECO) under research projects ’El proceso legislativo en la Unión Europea: análisis de las mayorías parlamentarias en los procesos de codecisión’ (2016-2019; Reference number: CSO2015-66050-P) and ’Delegation in the Post-Lisbon Era: A Quantitative Analysis of Comitology’ (2018-2021; Reference number: CSO2017-87683-P). Correspondence: Nuria Font Department of Political Science and Public Law, Universitat Autònoma de Barcelona, Building B, Campus de Belletarra S/N, 08193 Bellaterra, Barcelona, Spain. email: nuria.font@uab.es © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Unveiling Inconsistency: Consensus and Contestation along the Council–Comitology Cycle of EU Policy-Making 443 References Ballmann, A., Epstein, D. and O’Halloran, S. (2002) ‘Delegation, Comitology and the Separation of Powers in the European Union’. International Organization, Vol. 56, No. 3, pp. 551–74. Blom-Hansen, J. (2008) ‘The Origins of the EU Comitology System’. Journal of European Public Policy, Vol. 15, No. 2, pp. 208–26. Blom-Hansen, J. (2011) ‘The EU Comitology System: Taking Stock before the New Lisbon Regime’. Journal of European Public Policy, Vol. 18, No. 4, pp. 607–17. Blom-Hansen, J. (2013) ‘Legislative Control of Powers Delegated to the Executive: The Case of the EU’. Governance, Vol. 26, No. 3, pp. 425–48. Blom-Hansen, J. (2014) ‘Comitology Choices in the EU Legislative Process: Contested or Consensual Decisions?’ Public Administration, Vol. 92, No. 1, pp. 55–70. Brandsma, G. and Blom-Hansen, J. (2010) ‘The EU Comitology System: What Role for the Commission?’ Public Administration, Vol. 88, No. 2, pp. 496–512. Brandsma, G. and Blom-Hansen, J. (2016) ‘Controlling Delegated Powers in the Post-Lisbon European Union’. Journal of Common Market Studies, Vol. 23, No. 4, pp. 531–49. Burns, C. and Tobin, P. (2020) ‘Crisis, Climate Change and Comitology: Policy Dismantling Via the Backdoor?’ Journal of Common Market Studies, Vol. 58, No. 3, pp. 527–44. Craig, P. (2016) ‘Comitology, Rulemaking, and the Lisbon Settlement. Tensions and Strains’, in C. F. Bergström and Ritleng, D. (eds), Rulemaking by the European Commission: The New System for Delegation of Powers (Oxford: Oxford University Press), chapter 9, https://doi. org/10.1093/acprof:oso/9780198703235.003.0009. Crombez, C. & Hix, S. (2015) Legislative activity and gridlock in the European Union. British Journal of Political Science, Vol. 45 No. 3, pp. 477–99. Dehousse, R., Fernández Pasarín, A. and Plaza, J. (2014) ‘How Consensual is Comitology’. Journal of European Public Policy, Vol. 21, No. 6, pp. 842–59. Döring, H. and P. Manow (2019) ‘Parliaments and Governments Database (ParlGov): Information on Parties, Elections and Cabinets in Modern Democracies’. Development version. Epstein, D. and O’Halloran, S. (1999) Delegating Powers: A Transaction Cost Politics Approach to Policy Making under Separate Powers (New York: Cambridge University Press). Fernández-Pasarín, A., Dehousse, R. and Plaza, J. (2021) ‘Comitology: The Strength of Dissent’. Journal of European Integration, Vol. 43, No. 3, pp. 311–30. Finke, D. (2017) Underneath the culture of consensus: Transparency, credible commitments and voting in the Council of Ministers. European Union Politics, Vol. 18 No. 3, pp. 339–61. Franchino, F. (2007) The Powers of the Union: Delegation in the EU. (Cambridge: Cambridge University Press). Franchino, F. (2004) Delegating powers in the European Community. British Journal of Political Science, Vol. 34 No. 2, pp. 269–93. Gilardi, F. (2002) ‘Policy Credibility and Delegation to Independent Regulatory Agencies: A Comparative Empirical Analysis’. Journal of European Public Policy, Vol. 9, No. 6, pp. 873–93. Häge, F. (2007) ‘Committee Decision-Making in the Council of the EU’. European Union Politics, Vol. 8, No. 3, pp. 299–328. Hagemann, S. and Franchino, F. (2016) ‘Transparency vs Efficiency? A Study of Negotiations in the Council of the European Union’. European Union Politics, Vol. 17, No. 3, pp. 408–28. Hagemann, S., Hobolt, S.B. and Wratil, C. (2017) ‘Government Responsiveness in the European Union: Evidence from Council Voting’. Comparative Political Studies, Vol. 50, No. 6, pp. 850–76. Hayes-Renshaw, F., Van Aken, W. and Wallace, H. (2006) ‘When and Why the EU Council of Ministers Votes Explicitly’. Journal of Common Market Studies, Vol. 44, No. 1, pp. 161–94. © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd. 14685965, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jcms.13244 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [16/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 444 Ana Mar Fernández Pasarín and Nuria Font Heisenberg, D. (2005) ‘The Institution of ‘Consensus’ in the European Union: Formal versus Informal Decision-Making in the Council’. European Journal of Political Research, Vol. 44, No. 1, pp. 65–90. Héritier, A., Moury, C., Bischoff, C.S. and Bergström, C.F. (2013) ‘Changing Rules of Delega- tion’. In A Contest for Power in Comitology (Oxford: Oxford University Press). Hobolt, S.B. and Wratil, C. (2020) ‘Contestation and Responsiveness in EU Council Delibera- tions’. Journal of European Public Policy, Vol. 27, No. 3, pp. 362–81. Huber, J.D. and Shipan, C.R. (2002) Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy (Cambridge: Cambridge University Press). Joerges, C. and Neyer, J. (1997a) ‘From Intergovernmental Bargaining to Deliberative Political Processes: The Constitutionalisation of Comitology’. European Law Journal, Vol. 3, No. 3, pp. 273–99. Joerges, C. and Neyer, J. (1997b) ‘Transforming Strategic Interaction into Deliberative Problem- Solving: European Comitology in the Foodstuffs Sector’. Journal of European Public Policy, Vol. 4, No. 4, pp. 609–25. Majone, G. (1996) Regulating Europe (London: Routledge). Majone, G. (2001) ‘Two Logics of Delegation: Agency and Fiduciary Relations in EU Governance’. European Union Politics, Vol. 2, No. 1, pp. 103–22. Mattila, M. (2004) ‘Contested Decisions: Empirical Analysis of Voting in the European Union Council of Ministers’. European Journal of Political Research, Vol. 43, No. 1, pp. 29–50. Moury, C. and Héritier, A. (2012) ‘Shifting Competences and Changing Preferences: The Case of Delegation to Comitology’. Journal of European Public Policy, Vol. 19, No. 9, pp. 1316–35. Naurin, D. and Wallace, H. (2008) Unveiling the Council of the European Union (Houndsmill: Palgrave Macmillan). Novak, S. (2013) ‘The Silence of Ministers: Consensus and Blame Avoidance in the Council of the EU’. Journal of Common Market Studies, Vol. 51, No. 6, pp. 1091–107. Pollack, M. (2003) ‘Control Mechanism or Deliberative Democracy? Two Images of Comitology’. Comparative Political Studies, Vol. 36, No. 1–2, pp. 125–55. Reh, C., Héritier, A., Bressanelli, E. and Koop, C. (2013) ‘The Informal Politics of Legislation: Explaining Secluded Decision Making in the European Union’. Comparative Political Studies, Vol. 46, No. 9, pp. 1112–42. Thomson, R. and Torenvlied, R. (2011) ‘Information, Commitment and Consensus: A Comparison of Three Perspectives on Delegation in the European Union’. British Journal of Political Science, Vol. 41, No. 1, pp. 139–59. Yordanova, N. and Zhelyazkova, A. (2019) ‘Legislative Control over Executive Law-making: Delegation of Quasi-legislative Powers to the European Commission’. Journal of Common Market Studies, Vol. 58, No. 2, pp. 345–64. Supporting Information Additional supporting information may be found online in the Supporting Information section at the end of the article. Table S1. Descriptive statistics of the dependent, independent and control variables. Figure S1. Distribution of government changes between Council–comitology votes. © 2021 The Authors. JCMS: Journal of Common Market Studies published by University Association for Contemporary European Studies and John Wiley & Sons Ltd.