Modes of government responsiveness in the European Union: Evidence from Council negotiation positions

Abstract

Are national governments responsive to citizens’ opinions when negotiating policies in the Council of the European Union? Conceiving of the Council’s policy-making space as encompassing left-right and pro-anti integration issues, I argue that governments apply different ‘modes of responsiveness’ on these issues. As left-right issues are more reliably and intensely salient in domestic elections than pro-anti integration issues, governments’ responsiveness to left-right public opinion should be more systematic than to pro-anti integration opinion. Statistical analyses of 3700 policy positions of governments in the Council demonstrate that governments highly structure their responsiveness on left-right issues according to electoral cycles and systems (‘systematic mode’). However, they only sporadically respond to public opinion on pro-anti integration issues, when parties and events trigger the public salience of integration (‘sporadic mode’).

Keywords

Council of the European Union democratic deficit electoral systems government responsiveness salience

Introduction

Several studies have found diverse patterns of policy responsiveness in the European Union (EU) by demonstrating how aggregate policy outputs in the EU react to EU-wide public opinion (e.g. Bølstad, 2015; de Vries and Arnold, 2011; Toshkov, 2011). However, while these studies assume that policy responsiveness in the EU is a result of the responsive behavior of various actors in the EU institutions, such as national ministers and the European Commission, they have not investigated these micro foundations. Only very few studies have examined whether and how individual legislators in the EU institutions respond to the opinions of their constituents (but see e.g. Hagemann et al., 2017; Lo, 2013). Are EU-level legislators listening and acting upon their citizens’ substantive policy preferences when negotiating legislation in Brussels? Here, I address this question for the behavior of national governments in the Council of the EU.

This work makes three key contributions to the literature on policy-making and public opinion in the EU. First, while existing studies of EU-level legislators have focused on their voting behavior at the end of legislative negotiations (Hagemann et al., 2017; Lo, 2013), I investigate the responsiveness of initial negotiation positions formulated at the very beginning of the policy-making process. Whereas final passage votes in the Council are mostly consensual and do not influence the legislative text anymore, negotiation positions of governments shape which substantive policy is agreed (Thomson, 2011). Hence, while policy positions do not represent policy output, their responsiveness is highly important for understanding the influence of public opinion on the EU policy-making process. Second, while existing studies almost exclusively focus on whether policy or votes imply ‘more or less integration’ (e.g. Bølstad, 2015; Hagemann et al., 2017; Lo, 2013; Toshkov, 2011), I compare responsiveness on pro-anti integration issues with responsiveness on ‘left-right’ issues. Thereby, I provide a much more comprehensive view of government responsiveness in the EU, as left-right issues account for about 60% of all legislative conflicts discussed in the Council. Third, I provide an original argument of how differences in electoral competition over left-right versus pro-anti integration issues at the domestic level shape distinct ‘modes of responsiveness’ on these dimensions in the Council.

Theoretically, I start from the idea that when negotiating legislation in Brussels, governments also consider their accountability to domestic voters (Kleine and Minaudier, 2017; Schneider, 2013). Whereas these voters reliably care about left-right issues in domestic elections (van der Brug et al., 2008; van der Eijk et al., 2005), their interest in pro-anti integration issues is much more variable over time (de Vries et al., 2011; Stevens, 2013). This has important consequences for responsiveness in the Council. As governments know that left-right has a certain, rather stable level of salience in domestic elections, they engage in systematic, long-term planning and organization of their responsiveness efforts, and adjust them to national institutions and distribute them over time to optimize electoral returns (systematic mode). In contrast, since the electoral salience of pro-anti integration fluctuates, governments may not systematize their responsiveness efforts on such issues, as EU integration may be irrelevant at the time of the next electoral contest. The best governments can do is to take current increases in salience as a signal of potentially high salience at the next election, and respond sporadically when such increases occur (sporadic mode).

Statistical analyses of over 3700 Council policy positions adopted by governments reveal clear differences in responsiveness between issue dimensions. On left-right issues governments generally respond to public opinion, and their response is stronger if they face majoritarian electoral systems at home and when national elections are approaching. This indicates that governments systematically tailor and plan their responsiveness efforts with a view to national elections. In contrast, on pro-anti integration issues governments do not consistently or routinely respond to opinion. Instead, they only temporarily consider opinion at times when domestic opposition parties increase their emphasis on EU integration as well as when EU-related events increase the salience of integration in the public sphere.

These findings further challenge key aspects of the notion that the EU policy-making process is largely unresponsive to public preferences (e.g. Føllesdal and Hix, 2006). In contrast to this notion, the results suggest that national governments’ responsiveness in the Council is no ‘happy coincidence’ (Føllesdal and Hix, 2006: 556) but rather a reliable result of their electoral accountability at home.

Responsiveness in the EU policy-making process

Most studies of responsiveness in the EU policy-making process have examined ‘systemic responsiveness’ by ascertaining whether the amount of EU legislative acts reacts to changes in public Euroskepticism, and whether in turn the public ‘thermostatically’ adjusts its Euroskepticism in response to EU-level policy activity (Bølstad, 2015; de Vries and Arnold, 2011; Franklin and Wlezien, 1997; Toshkov, 2011). While Toshkov's (2011) landmark study finds that the volume of EU directives negatively reacted to EU-wide Euroskepticism before the mid-1990s but not afterwards, recent work by Bølstad (2015) identifies some general impact of opinion on policy, when decomposing opinion into ‘core’ (EU founding members) and ‘periphery’ (Denmark, Ireland, UK) trends. Commonly, systemic responsiveness is theorized as a result of the behavior of individual legislators in the EU institutions who react to public opinion (Toshkov, 2011: 172–173). However, the extant studies have not investigated these micro foundations empirically. Moreover, a central limitation of the systemic approach is that it focuses on the sheer quantity of acts that tells us little about the substantive content of policies.

Only very few studies have focused on the micro level of policy-making in the institutions and investigated the legislative behavior of national governments and members of the European Parliament (EP) (e.g. Hagemann et al., 2017; Lo, 2013). This literature has also paid more attention to the content of policy-making. For instance, Hagemann et al. (2017) use quantitative text analysis to distinguish between legislative acts that extend the EU’s authority and those that merely update existing legislation, and show that national governments are only responsive to public Euroskepticism on ‘integrationist’ acts. Yet, all work on individual legislators investigates the stage of voting, when representatives cannot influence the legislative text anymore – one reason why Hagemann et al. (2017) speak of mere ‘signal’ responsiveness in their study.

To my knowledge, none of the existing studies addresses substantive responsiveness of EU-level legislators at the negotiation stage, the very stage when several policy alternatives as options for the final output are on the table. I address this gap by investigating how governments’ initial policy positions in the Council, at the beginning of negotiations, are related to their domestic public opinion. From an empirical perspective, responsiveness in the Council is particularly important, since the Council has been consistently found to be the most influential institution in shaping EU policy output (compared to the EP and the Commission) (e.g. Franchino and Mariotto, 2012; Thomson, 2011).

Domestic electoral competition and the Council’s policy-making space

I argue that domestic electoral competition is key for understanding responsiveness in the Council, as governments’ accountability to voters at home provides the central incentives for them to respond to public opinion. The existing findings of responsiveness at the EU level (Bølstad, 2015; Hagemann et al., 2017; Toshkov, 2011) and national electoral cycle effects in Council decision-making (Kleine and Minaudier, 2017; Schneider, 2013) strongly suggest that governments consider their accountability in national elections when acting at the EU level. Considerations of electoral accountability should incentivize governments to be responsive to public opinion to forestall and avoid future electoral sanctions at the next election. This is what has been famously called ‘rational anticipation’ (Stimson et al., 1995) or ‘anticipatory representation’ (Mansbridge, 2003) in the literature.

The essence of rational anticipation is that governments’ incentives to respond to public opinion stem from the very ‘chance’ of being held accountable. In our context, that is, whether government members or officials are concerned that advocating ‘unpopular’ positions in the Council could potentially cost some votes at the next national election, as media,¹ interest groups, or competitors may raise voters’ attention to the government’s unresponsive behavior (either specifically or in diffuse terms, e.g. ‘the government acts against people’s interests in Brussels’). Crucially, the chance of electoral sanctions for governments’ behavior in the Council is related to but distinct from actual sanctions at the ballot box or actual politicization of Council politics in the public sphere. Where governments are highly responsive and follow the public’s preferences closely, the ‘need for actual electoral defeats’ (Stimson et al., 1995: 545) and sanctions is eliminated as well as competitors’ opportunities to politicize EU policy-making.

Note that responsiveness in the Council may even occur without intention or orchestration by national ministers. Governmental policy in the Council is often shaped by ministerial staff in the capitals and officials in the Permanent Representations in Brussels, who, in preemptive obedience, may adjust policy positions marginally, for instance, to forestall a negative news story about their superior’s unpopular EU policies. Note also that governments may not only be responsive because voters become aware of their behavior in the Council, but they may also fear that once policies are visibly implemented at the national level, voters may erroneously attribute disproportionate responsibility for unpopular EU policies to their government. Representing public opinion during Council negotiations may then serve to change these policies upfront, or at least, to be credibly able to claim to have worked for change with all efforts.²

In contrast to most studies of national systems that assess responsiveness on a single liberal-conservative (or left-right) dimension, responsiveness in the Council should be assessed on two issue dimensions, as the EU’s policy-making space, at least from an a priori perspective, is arguably two-dimensional (e.g. de Vries and Marks, 2012; Hix, 1999).³ Many EU policy-making controversies will naturally connect to conflicts about left-right ideological issues, encompassing classical economic and newly emerging left-libertarian versus right-authoritarian elements. Examples range from the liberalization of former state company sectors, over visa facilitations, to the treatment of laying hens in farms. But some pieces of legislation also touch upon genuine pro-anti integration conflicts, as they concern jurisdictional architecture and have direct implications for the scope and level of EU authority or for the EU’s geographical inclusiveness (see also de Wilde and Zürn, 2012). Examples range from the Commission’s authority in enforcing budgetary discipline to the member states’ involvement in the implementation of the European neighborhood policy. Sometimes both kinds of conflict intersect in the same issue.

To understand governments’ responsiveness on left-right and pro-anti integration issues in the Council, we therefore consider the role of such issues in domestic electoral competition, that is, their potential ‘chance’ to be raised by competitors and be important to some voters. On the one hand, left-right issues in the Council relate to left-right party competition as the ‘super-dimension’ of political competition in the member states, with a firm anchoring in classical political cleavages (de Vries and Marks, 2012; Lipset and Rokkan, 1967). Left-right has strong social bases, since citizens’ preferences on such issues derive in large part from socio-structural and social identity factors such as class, religion, or trade union membership (e.g. Freire, 2006). Parties themselves and experts assessing parties use the left-right semantic to summarize party positions, and left-right has a common core meaning across Europe from West to East (e.g. Bakker et al., 2014). Due to its centrality for party identity, left-right also takes primacy in party-strategic considerations, for instance, when parties adopt positions on new issues (Marks et al., 2006).

In contrast, pro-anti integration issues in the Council have no obvious relation to the classical Lipset and Rokkan (1967) cleavages in European democracies. Some argue that such issues are connected to a new integration-demarcation cleavage between winners and losers of globalization (Kriesi et al., 2008). But often integration is still conceived of as a fragile, evolving ‘issue’ of electoral competition (Carmines and Stimson, 1986) rather than a cleavage. Citizens’ pro-anti integration preferences do have some limited social bases (especially occupation, education, and income) (e.g. Gabel, 1998), but they remain highly malleable and subject to elite cueing (Gabel and Scheve, 2007; Steenbergen et al., 2007). As pro-anti integration issues cross-cut lines of party cleavage, they have caused major internal dissent in Europe’s mainstream parties, whose leaderships have tried to de-emphasize them (e.g. van de Wardt, 2014). As a consequence, such issues are mainly mobilized by issue entrepreneurs, i.e. fringe parties which are losers on the left-right dimension (e.g. Hobolt and de Vries, 2015).

These differences between left-right and pro-anti integration issues are reflected in the issues’ salience for voters in national elections. Where longitudinal and cross-national data have been investigated, citizens’ left-right preferences prove universally salient for vote choice with not much variation between elections (van der Brug et al., 2008; van der Eijk et al., 2005). In contrast, the EU issue voting literature has demonstrated that the electoral salience of pro-anti integration issues is highly variable and often low. For instance, media information on European issues and events related to integration (e.g. referendums) trigger EU issue voting (de Vries, 2009; de Vries et al., 2011). However, comparable longitudinal data reveals that such effects are temporary, as the salience of EU integration for voters follows strong ebbs and flows without any long-term trend over time (see e.g. an analysis of UK elections from 1960s to 2010, by Stevens, 2013). Importantly, the electoral consequences are particularly limited for mainstream parties (de Vries, 2010), who staff most governments.

In sum, left-right issues are more reliably salient in electoral competition than pro-anti integration issues, whose importance is sometimes substantial but low on average and fluctuates starkly over time.

Electoral salience and modes of responsiveness

These differences in the variability in electoral salience of left-right and pro-anti integration issues should impact on responsiveness in the Council. When governmental actors at the EU level engage in the already mentioned act of ‘rational anticipation’ (Mansbridge, 2003; Stimson et al., 1995), they first and foremost want to know whether left-right and pro-anti integration issues from EU policy-making could be salient at the time of the next election. As the future is inherently uncertain, actors can only form (rational) beliefs about this chance and these beliefs can be true or false (Mansbridge, 2003: 517; Stimson et al., 1995: 545). This is where the variability in salience matters. If governments know that left-right issues are reliably important in elections, they face little uncertainty that there is, at least, some chance of electoral sanctions for their behavior on such issues in the Council. In turn, if the electoral salience of pro-anti integration issues varies starkly over time, the uncertainty about the chance of sanctions is substantial. I argue that this will affect the way in which governments organize their responsiveness efforts and lead to different stylized modes of responsiveness.⁴

As governments know that voters reliably care about left-right issues, they will adapt their responsiveness efforts to the domestic institutional framework, creating a long-term and structured plan (systematic mode). In particular, with a view to the next election they can adapt their responsiveness to electoral institutions and cycles. This is rational, as governments can expect advantages and pay-offs from their responsiveness at the next election with high certainty. In turn, as the electoral salience of pro-anti integration issues is uncertain, systematic responsiveness and long-term planning may not be efficient, since such issues may be entirely irrelevant to voters at the time of elections. Instead, I argue that governments will handle responsiveness more immediately by only reacting to changes in salience that demonstrate that EU integration could be important at the next election, and disregarding opinion otherwise (sporadic mode). Responsiveness should then primarily be triggered by signals of salience such as exogenous events, the activities of the opposition, and media attention.

To test whether governments’ modes of responsiveness differ between left-right and pro-anti integration issues, I formulate two testable hypotheses for each issue dimension that are specific to the systematic and sporadic mode respectively. I start with left-right and systematic responsiveness. First, if governments systematically plan their responsiveness efforts on left-right with an eye towards elections, their efforts should vary over the course of the electoral cycle. Such structuring should not occur if governments only respond sporadically. Several studies have demonstrated that the responsiveness of representatives on major issue dimensions increases with electoral proximity (e.g. Canes-Wrone and Shotts, 2004; Lindstädt and Vander Wielen, 2011). Organizing responsiveness efforts according to the electoral cycle is effective, as voters’ gratitude for government action declines over time and good deeds before elections count more than those in the distant past (Bechtel and Hainmueller, 2011). Moreover, voters may use the period leading up to the election as a heuristic for government performance overall and ‘substitute the end for the whole’ (Healy and Lenz, 2014: 33).

H1: Governments will be more responsive to domestic public opinion on left-right issues the closer national elections are.

Secondly, in the systematic mode, electoral systems should be the major institution according to which governments can optimize their responsiveness efforts, which should be of less concern if acting in an ‘ad hoc’ or sporadic manner. If the two major parties in a system are neck and neck, marginal changes in vote shares will typically transform into larger changes in seat shares in majoritarian electoral systems than in proportional systems. This means that the vote-seat elasticity,

e_{v} = \frac{Δ Seats / Seats}{Δ Votes / Votes}

, is larger under majoritarian systems (Kayser and Lindstädt, 2015). This should generally create stronger incentives for governments to respond to opinion in majoritarian as opposed to proportional electoral systems, as comparatively smaller electoral losses could dismantle government majorities.

However, there are two caveats to this general conclusion. First, the idea that seat-vote elasticities are strictly greater in single-member district (SMD) than in proportional (PR) systems only holds if the two major parties are neck and neck. If one party dominates by a wide vote margin, the seat-vote elasticity can approach 0 under SMD, while it will always be close to 1 under PR (see also Hobolt and Klemmensen, 2008: 313–314). Second, SMD also creates incentives for governments to over-respond to the preferences and opinions of citizens in marginal districts that may often markedly differ from average opinion (e.g. Rickard, 2012). Yet, Kayser and Lindstädt's (2015: 247) ‘loss probability’ model accounts for these caveats and still estimates a higher average electoral risk for Western post-war SMD than PR systems.

H2: Governments will be more responsive to domestic public opinion on left-right issues in majoritarian as opposed to proportional electoral systems.

Turning to pro-anti integration issues and sporadic responsiveness, two fundamental factors indicate increases in salience and should prompt governments’ occasional response: opposition parties emphasizing the issues, and events related to integration. First, as party elites play a crucial role in mobilizing issues with fluctuating salience in the domestic media and public sphere, governments’ responsiveness efforts should be sensitive to opposition parties’ emphasis on integration (cf. Carmines and Stimson, 1986; Hooghe and Marks, 2009). In contrast, this emphasis should not matter much if governments respond routinely. Most importantly, several studies demonstrate that party elites must often act as ‘vehicles’ for pro-anti integration issues to attain media coverage. If party leaders speak about the issues, reframe their domestic relevance, and communicate their stances on them, their news value increases (e.g. Adam, 2007; van der Pas and Vliegenthart, 2016). Therefore, when opposition parties place increased emphasis on integration issues, governments should interpret this as a signal of potential public and electoral salience, and increase their responsiveness.

H3: Governments will be more responsive to domestic public opinion on pro-anti integration issues whenever opposition parties increase their emphasis on such issues.

Second, events related to EU integration should add to the public salience of integration and therefore prompt governments’ responsiveness. Like party emphasis on integration, events can act as a vehicle for pro-anti integration conflicts to enter the public sphere. Research has demonstrated that events like referendums on EU integration, elections to the EP, or the introduction of the euro temporarily increase the coverage of integration issues in national media, as does the period in which a nation holds the Council presidency (e.g. Boomgaarden et al., 2010). These events often trigger intense public debates about pro-anti integration, for instance, about legislation that challenges national sovereignty (e.g. centralized refugee relocation). In the specific case of referendums, research has shown that such debates clarify actors’ positions to the public, affect public evaluations of leaders (de Vreese, 2004), and increase the electoral salience of pro-anti integration in the next national election (de Vries, 2009).

H4: Governments will be more responsive to domestic public opinion on pro-anti integration around major events related to EU integration.

Data and measurement

To test these hypotheses, I investigate the responsiveness of national governments from 26 EU countries on both issue dimensions in the Council.⁵ Specifically, I measure responsiveness as the effect of domestic public opinion on national ministers’ initial policy positions regarding concrete pieces of legislation negotiated in the Council. I take governments’ negotiation positions as the dependent variable from the Decision-making in the European Union (DEU) dataset (Thomson et al., 2012). Based on expert interviews, DEU provides governments’ initial positions (at the beginning of negotiations) on spatial policy scales regarding 331 controversial issues from two legislative terms, 1994–1999 and 2004–2009. It is the most widely used dataset on EU decision-making and has been employed in a diverse range of applications (see the Online appendix for more details). Importantly, DEU is a selection of the most salient issues in EU policy-making as proposals had to be mentioned in European media and actors had to be divided on key aspects to be included in the data.

I use an original coding scheme to identify those DEU issues that substantively relate to the two issue dimensions of interest (see the Online appendix for details and inter-coder reliability). The coding categories have been constructed to correspond to the DEU issues and at the same time closely reflect our best knowledge about what citizens associate with ‘left-right’ and ‘pro-anti integration’ conflicts. This guarantees a high substantive relevance between the dependent and independent variable. The scheme consists of eight categories representing the left-right dimension; they cover conflicts ranging from economic regulation, through consumer and environmental protection, to human and civil rights. Four categories capture the integration dimension, covering aspects ranging from harmonization over EU-level authority to the delaying of EU legislative acts.

Each DEU issue is coded into each category that applies. The coding reveals that about 61% of the DEU issues fall into at least one left-right category and 31% into at least one pro-anti integration category. While left-right issues are therefore more prevalent in the Council, governments deemed both kinds of issues as equally salient according to the DEU actor salience scores (average left-right issues = 57; average pro-anti integration issues = 53). Twelve percent of all issues are relevant on both dimensions. Twenty two percent relate neither to any left-right nor to any pro-anti integration category (e.g. they relate to geographical cleavages). Wherever necessary the DEU scales with a predefined range of 0–100 have been linearly rescaled so as to ensure that the most ‘right’ and the most ‘integrationist’ options advocated by any government in the estimation sample are represented by 100 and the most left/least integrationist by 0.⁶

Figure 1 illustrates the dependent variable with governments’ positions on two issues from the data. Left-right issues are illustrated with a proposal concerning the inclusion of the aviation industry in the EU’s CO2 emission trading system (COD/2006/304). In this case, the extent of emission allowances to be allocated to the industry was contested with some governments opting for smaller and others for larger allowances (left versus right position). Pro-anti integration issues are exemplified by the implementation arrangements for the European Neighborhood and Partnership Instrument (COD/2004/219). Governments were divided over the involvement of the member states (vis-à-vis the Commission) in the implementation of the neighborhood policy’s financing instrument.

Figure 1.

Illustration of the dependent variable.

To measure public opinion as the primary independent variable I use survey data from the Standard Eurobarometer series by linearly interpolating opinion between the six-monthly surveys. Following the study of government responsiveness in the Council by Hagemann et al. (2017), I lag the opinion variables by six months in order to reflect causal ordering, in which governments first observe opinion and subsequently react to it. Public opinion on left-right conflicts is operationalized as the country’s average of the ideological self-placement item that asks respondents to identify their own ideological position on a scale ranging from 1 (left) to 10 (right). Scholars have demonstrated that these self-placements are systematically related to attitudes on economic, cultural, and materialist versus post-materialist conflicts, including specific issues such as immigration, civil liberties and human rights as well as a variety of environmental issues (see the Online appendix for relevant references). Therefore, these placements are a substantive measure of the public’s left-right preferences.

In line with the extant literature, public opinion on pro-anti integration is measured as support for EU membership (‘Generally speaking, do you think your country’s membership in the European Union is a good thing, a bad thing, or neither good nor bad?’). Support is operationalized as the difference between respondents answering that the EU is a good thing minus those thinking the EU is a bad thing divided by all responses per country. This survey item has been shown to correspond with more specific attitudes on integration as well as with actual EU policy activity (e.g. Bølstad, 2015; Boomgaarden et al., 2011; Toshkov, 2011).

The moderating factors hypothesized are measured as follows. First, to test whether the electoral cycle influences government responsiveness I operationalize the distance to the next scheduled national legislative election in 100-day units. Second, the impact of electoral systems is ascertained with a measure of logged district magnitude to map systems on a continuum from majoritarian to proportional.⁷ Third, whether an increase in opposition parties’ emphasis on EU integration is related to stronger responsiveness is tested with the help of the Comparative Manifesto Project’s (CMP) database of election manifestos (Lehmann et al., 2015). This is operationalized as the simple average percentage of quasi-sentences opposition parties devote to EU integration (positive as well as negative statements) in their manifestos. To capture times in which opposition parties increase their emphasis, I measure the change in opposition parties’ emphasis over the last four years. The central logic of this measure was developed by Hagemann et al. (2017).⁸ Fourth, to determine the effects of major EU-related events I use a dummy variable that takes the value of ‘1’ six months before and after the following events: national referendums on integration, final ratification of EU treaties, accession to the EU, introduction of the euro currency, elections to the EP, holding the Council presidency.

I also include measures of government parties’ left-right and pro-anti integration positions at the last election from the CMP. For left-right I use the CMP’s summative RILE measure, and for pro-anti integration I take the difference between the percentage of positive and negative quasi-sentences on the EU. Aggregated government positions are obtained by seat-weighting. This effectively controls for effects of the ‘electoral turnover’ mechanism (for a discussion, see e.g. Stimson et al., 1995). Not only do governments respond directly to public opinion, but voters also align governments’ preferences with their views by influencing the composition of parliament and government through elections. Including measures of governments’ promised policy positions guarantees that the effects of public opinion are controlled for any effects of the ideological composition of the government and exclusively represent governments’ responsiveness efforts, i.e. rational anticipation (Stimson, 1999).

In addition, I control for several factors that are known to influence the positioning of governments in EU policy-making and might at the same time have non-trivial relationships with public opinion. Most importantly, I account for a potentially relevant redistribution cleavage with a measure of countries’ annual net receipts from the EU budget (% of GDP) (e.g. Bailer et al., 2015). Second, I include national unemployment and inflation rates to ensure that the relationship between parties’ and government’s policy positions is more than a reflection of macroeconomic fluctuations (Ferguson et al., 2013). Third, I broadly capture the idea that member states may try to ‘upload’ their domestic policies to the European level (see e.g. Börzel, 2002) with a measure of domestic economic freedom from the Fraser Institute’s Economic Freedom of the World dataset (also used by Thomson, 2011). All data sources are linked on the date the Commission submitted the relevant proposal to the Council to reflect that DEU measures ‘initial positions’, just after the introduction of the proposal.

Analysis and results

To model the relationship between governments’ policy positions and opinion, I use linear regressions with two-way fixed effects for the DEU policy issues as well as for the countries. The fixed effects estimator is best suited to obtaining consistent estimates in face of the heterogeneous set of policy issues and potentially unobserved country-level confounders. Moreover, I employ standard errors clustered at the country level. Including interaction terms between public opinion and the respective moderating variable allows us to test the hypotheses.

In a first step, I restrict the sample to all issues that relate to the left-right dimension. Table 1 shows the major results. Model 1 regresses governments’ policy positions (0 ‘most left’ to 100 ‘most right’ option) on left-right opinion and the control variables. The results show that opinion exerts a systematic and significant effect on governments’ positioning in the Council. For a unit change in opinion to the right, governments shift their policy positions by 15.2 points towards rightist policy positions on the DEU scale. This effect is significant at the 5% level. In addition, governments take more rightist policy positions in times when they receive higher payments from the EU budget. The coefficients on all other control variables are not statistically significant.

Table 1.

Responsiveness on left-right dimension.

	Model 1	Model 2
Left-right opinion	15.21	47.94
	(7.20)**	(14.55)***
Days to election		5.58
		(2.45)**
Left-right opinion × Days to election		−1.03
		(0.45)**
District magnitude (ln)		43.91
		(18.98)**
Left-right opinion × District magnitude (ln)		−9.12
		(3.65)**
Government left-right position	0.08	0.12
	(0.13)	(0.14)
Net receipts from EU budget	3.25	4.04
	(1.79)*	(1.87)**
Unemployment	−0.67	−0.89
	(0.72)	(0.68)
Inflation	−1.52	−1.66
	(1.55)	(1.49)
Economic freedom	5.22	7.94
	(3.63)	(4.60)*
Fixed effects	Policy issues, countries	Policy issues, countries
Number of policy issues	172	172
Number of countries	26	26
N	2965	2965
R²	0.28	0.28

Notes: All are fixed effects regressions; Country-clustered robust standard errors in parentheses.

p < 0.1, **p < 0.05, ***p < 0.01.

Model 2 investigates the evidence for the two hypotheses specific to the systematic mode. First, it tests whether governments increase their responsiveness on left-right issues as national elections approach (H1). Indeed, there is a significant negative interaction effect between public opinion and the number of days to the next election. As elections approach, governments increasingly consider public opinion when taking policy positions in the Council. Second, the model also assesses the impact of electoral systems on responsiveness by allowing the opinion effect to vary by logged district magnitude (H2). It shows that governments in majoritarian electoral systems are indeed more responsive to left-right preferences than their counterparts in proportional electoral systems, as the opinion effect diminishes with district magnitude. Both results suggest that governments systematize and adjust their responsiveness with a view to national elections. The model also provides some limited evidence that governments with increasing economic freedom at home take more rightist positions, which is in line with the ‘uploading’ hypothesis.

Figure 2 illustrates these results. Panels (a) and (b) show how the marginal effect of public opinion increases closer to elections and with lower district magnitudes. Panels (c) and (d) combine the effects by showing the influence of the electoral cycle in SMD systems and in average PR systems (i.e. at the sample mean of district magnitude). From (c) we can see that governments facing SMD systems at home are generally responsive when taking positions in the Council, even five years ahead of elections, and when elections are imminent they change their policy positions by up to 48 points for a unit change in opinion. Considering the DEU scale from 0 to 100, this is a stark effect. In turn, (d) reveals that in the average, PR electoral system public opinion does not have any significant influence when planned elections lie more than 1300 to 1400 days in the future. However, from around three and a half years before the elections governments clearly consider public opinion when taking positions in the Council.

Figure 2.

Marginal effects of opinion by electoral cycle and system.

In a second step, I restrict the sample to all issues relating to conflicts about pro-anti integration. The results are reported in Table 2. Model 3 estimates the adopted policy position (0 ‘least integrationist’ to 100 ‘most integrationist’ option) as a function of pro-anti integration opinion and the control variables. In contrast to the findings on the left-right dimension, there is no unconditional effect of public opinion on governments’ positioning on pro-anti integration issues. The coefficient on opinion is positive but statistically indistinguishable from zero. This underscores that governments do not in general respond on pro-anti integration issues, as pro-anti integration is not reliably salient in domestic elections.

Table 2.

Responsiveness on pro-anti integration dimension.

	Model 3	Model 4
Pro-anti integration opinion	19.92	5.85
	(19.09)	(21.63)
Opposition parties’ emphasis (EU)		−3.40
		(1.43)**
Pro-anti integration opinion × Opposition parties’ emphasis (EU)		10.93
		(3.46)***
EU-related event		−16.80
		(7.42)**
Pro-anti integration opinion × EU-related event		32.91
		(14.79)**
Government pro-anti integration position	−0.21	−0.26
	(0.61)	(0.55)
Net receipts from EU budget	7.28	3.22
	(2.42)***	(2.91)
Unemployment	−1.03	−1.62
	(0.55)*	(0.53)***
Inflation	−0.52	0.03
	(1.05)	(1.03)
Economic freedom	−1.34	0.16
	(3.58)	(3.53)
Fixed effects	Policy issues, countries	Policy issues, countries
Number of policy issues	82	82
Number of countries	26	26
N	1448	1448
R²	0.30	0.31

Notes: All are fixed effects regressions; Country-clustered robust standard errors in parentheses.

p < 0.1, **p < 0.05, ***p < 0.01.

Model 4 tests the two hypotheses specific to the sporadic mode, investigating whether governments respond on pro-anti integration when issue salience increases. First, it tests whether governments are responsive when opposition parties increase their emphasis on pro-anti integration in their manifestos (H3). The hypothesis is supported by a significant interaction term (at the 1% significance level). Second, the model tests whether major events related to EU integration play a pivotal role in making governments wary of public opinion on pro-anti integration (H4). The estimated coefficients entirely support this conjecture, as the opinion effect is almost 33 scale points higher before and after major events on EU integration. The results demonstrate that governments do not respond systematically on pro-anti integration but sporadically when triggered by opposition parties’ emphasis on integration or EU-related events, both of which signal potential electoral salience.

In Figure 3, panels (a) and (b) demonstrate how the marginal effect of public opinion is higher when opposition parties’ emphasis on integration has increased and in the context of EU-related events. Panel (c) shows that in times without EU-related event governments do not significantly respond to opinion, irrespective of opposition parties’ emphasis on integration. However, panel (d) reveals that around events related to integration, opinion exerts a statistically significant effect on governments’ positions whenever opposition parties have focused more on integration in their manifestos.

Figure 3.

Marginal effects of opinion by party emphasis and events context.

In terms of the control variables, the negative and significant coefficient on unemployment in Models 3 and 4 suggests that if governments face more unemployed workers at home they are more reluctant to extend integration.⁹ Also note that neither on left-right nor on pro-anti integration government parties’ positions in their election manifestos prove predictive of the positions they take in the Council.

The substantive magnitude of these opinion effects can be best interpreted by considering distances in policy positions between national delegations. For instance, the average distance on the left-right scale between the UK, which advocated more ‘rightist’ policy positions, and France, with more ‘leftist’ positions, is about 10 scale points (57 against 47 respectively). Given an opinion effect of about 28–48 points in SMD systems and long-run standard deviations of left-right opinion between 0.1 and 0.2, regular fluctuations in opinion are sufficient to make the UK government adopt positions like its French counterpart. As example for pro-anti integration, we can compare Ireland, which has largely adopted integrationist policy positions, and the Netherlands, which has been more reluctant. An average of seven scale points separates them. Given an opinion effect of 15–55 points in most situations around EU-related events and long-run standard deviations of support for membership around 0.1, changes of policy positions by 3–11 scale points are quite frequent. Hence, under Eurosceptic tides of public opinion Irish governments facing, for instance, an EU referendum can be expected to agree with Dutch delegations.

In the Online appendix, I perform various robustness checks of these results. In particular, I demonstrate that the modes of responsiveness are dimension-specific, that is, governments do not apply both modes on both dimensions.

Conclusion

This study has investigated government responsiveness of substantive negotiation positions in the EU’s two-dimensional policy-making space. I have argued that the different roles left-right and pro-anti integration issues play in domestic electoral competition are central to understanding government responsiveness in the Council. While left-right issues are of high and stable salience in domestic elections, the electoral importance of pro-anti integration issues varies greatly over time. This affects governments’ rational anticipation of the chance that voters will care about such issues in the next election and leads to different stylized modes of responsiveness. While a single form of public opinion may have little influence across all realms of Council politics (Bailer et al., 2015), I find substantial responsiveness when partitioning legislative acts or issues by substance (i.e. pro-anti integration, left-right) and using the suitable public opinion measures on sub-samples (see also Hagemann et al., 2017).

The findings contribute to our understanding of policy representation in EU politics. Existing studies mostly demonstrate some link between public Euroskepticism and legislative output. But they could not establish whether individual legislators behave as actual ‘agents of the public’ at the negotiation stage, defending domestic opinions when different substantive policy options are on the table. The results here speak in favor of a central micro foundation of EU policy representation, namely that, in many situations, national governments attempt to align the very substance of policies to domestic public opinion when negotiating in the Council. Moreover, the results suggest that governments’ motivation for this responsiveness is their electoral accountability in national elections, since opinion is stronger reflected in government positions in contexts of high electoral pressures (i.e. election approaching, SMD system, increased opposition emphasis or EU-related events).

This challenges parts of Føllesdal and Hix' (2006: 556, 545) prominent notion that relationships between public opinion and EU policy-making are unlikely to be more than ‘happy coincidence’ from elite bargains between ‘benevolent but non-accountable rulers’. While my analyses here are confined to one institution, they provide strong evidence that, at least, legislative negotiations in the Council are systematically connected to governments’ electoral accountability at home. Governments in the Council are not ‘benevolent authoritarians’ that do something for the people out of kindliness, but they strategically follow and shirk opinion depending on electoral incentives. Future work should investigate whether this behavior of national governments indeed aggregates and transforms into policy outputs that do not only quantitatively but also substantively match citizens’ demands.

Moreover, the findings potentially also contribute to our understanding of how issue characteristics influence responsiveness in general. While we have known for a long time that the level of electoral salience influences the strength of responsiveness in national politics, the findings suggest that we should also consider the consequences of variability in salience more thoroughly. For instance, it may be no coincidence that most studies reporting significant electoral cycle effects assess responsiveness on issues with strong and very stable salience, particularly a liberal-conservative/left-right issue dimension.

Null findings might also be reconsidered. For instance, so far scholars were unable to detect any relationship between responsiveness and variations over time in citizens’ responses to the ‘most important problem’ (MIP) question as a measure of public salience. While this has been attributed to MIP measuring ‘problem load’ rather than ‘importance of an issue’, it could also partly be due to the fact that the issues investigated – predominantly defense spending in the US – are at the core of the liberal-conservative cleavage (see Wlezien, 2005). Given the reliably high salience of such issues, representatives may systematize their responsiveness on such issues and respond irrespective of minor salience fluctuations.¹⁰

While these considerations indicate that different modes of responsiveness might also be relevant in other contexts and for other issues, future work is needed to determine the generalizability of the findings. This can involve the re-examination of existing datasets but also new studies comparing different issues and issue dimensions. Then, we will be able to determine whether distinct modes of responsiveness are an idiosyncrasy of policy-making in the Council, or whether they apply more generally beyond the EU context.

Footnotes

Acknowledgements

The author would like to thank Sara B. Hobolt, Sara Hagemann, Mark A. Kayser, Mareike Kleine, Christine Reh, Jonathan Slapin, and two anonymous reviewers for their excellent comments on earlier drafts of this article. The author is also grateful to Ida Popovski for her superb research assistance as well as to Julian M. Hörner for his reliability testing of the DEU coding scheme. An earlier version of this work won first prize in the LSE ‘Europe in Question’ Discussion Paper Series Doctoral Paper Competition 2014–2015. All remaining errors are my own.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The author would like to acknowledge the generous funding by the UK Economic and Social Research Council (W88918G) and the Fritz Thyssen Foundation (20.16.0.045WW).

Supplemental material

Supplementary material is available for this article online.

Notes

References

Adam

(2007) Domestic adaptations of Europe: A comparative study of the debates on EU enlargement and a common constitution in the German and French Quality Press. International Journal of Public Opinion Research 19(4): 409–433.

Bailer

Mattila

Schneider

(2015) Money makes the EU go round: The objective foundations of conflict in the Council of Ministers. Journal of Common Market Studies 53(3): 437–456.

Bakker

Jolly

Polk

et al. (2014) The European common space: Extending the use of anchoring vignettes. The Journal of Politics 76(4): 1089–1101.

Bechtel

Hainmueller

(2011) How lasting is voter gratitude? An analysis of the short- and long-term electoral returns to beneficial policy. American Journal of Political Science 55(4): 852–868.

Bølstad

(2015) Dynamics of European integration: Public opinion in the core and periphery. European Union Politics 16(1): 23–44.

Boomgaarden

Schuck

ART

Elenbaas

et al. (2011) Mapping EU attitudes: Conceptual and empirical dimensions of Euroscepticism and EU support. European Union Politics 12(2): 241–266.

Boomgaarden

Vliegenthart

de Vreese

et al. (2010) News on the move: Exogenous events and news coverage of the European Union. Journal of European Public Policy 17(4): 506–526.

Börzel

(2002) Pace-setting, foot-dragging, and fence-sitting: Member state responses to Europeanization. Journal of Common Market Studies 40(2): 193–214.

Canes-Wrone

Shotts

(2004) The conditional nature of presidential responsiveness to public opinion. American Journal of Political Science 48(4): 690–706.

10.

Carmines

Stimson

(1986) On the structure and sequence of issue evolution. American Political Science Review 80(3): 901–920.

11.

de Vreese

(2004) The effects of strategic news on political cynicism, issue evaluations, and policy support: A two-wave experiment. Mass Communication and Society 7(2): 191–214.

12.

de Vries

(2009) The impact of EU referenda on national electoral politics: The Dutch case. West European Politics 32(1): 142–171.

13.

de Vries

(2010) EU issue voting: Asset or liability?: How European integration affects parties’ electoral fortunes. European Union Politics 11(1): 89–117.

14.

de Vries

Arnold

(2011) Making the polity: Exploring the link between public opinion and European elites preferences for European Union public policy. In: Fuchs

Klingemann

H-D

(eds) Cultural Diversity, European Identity and the Legitimacy of the EU, Cheltenham: Edward Elgar, pp. 201–217.

15.

de Vries

Marks

(2012) The struggle over dimensionality: A note on theory and empirics. European Union Politics 13: 185–193.

16.

de Vries

van der Brug

van Egmond

et al. (2011) Individual and contextual variation in EU issue voting: The role of political information. Electoral Studies 30(1): 16–28.

17.

de Wilde

Zürn

(2012) Can the politicization of European integration be reversed? Journal of Common Market Studies 50(1): 137–153.

18.

Ferguson

Kellstedt

Linn

(2013) How does the economy shape policy preferences? Electoral Studies 32(3): 544–550.

19.

Føllesdal

Hix

(2006) Why there is a democratic deficit in the EU: A response to Majone and Moravcsik. Journal of Common Market Studies 44(3): 533–562.

20.

Franchino

Mariotto

(2012) Explaining negotiations in the conciliation committee. European Union Politics 14(3): 345–365.

21.

Franklin

Wlezien

(1997) The responsive public: Issue salience, policy change, and preferences for European unification. Journal of Theoretical Politics 9(3): 347–363.

22.

Freire

(2006) Bringing social identities back in: The social anchors of left-right orientation in Western Europe. International Political Science Review/Revue internationale de science politique 27(4): 359–378.

23.

Gabel

(1998) Interests and Integration, Ann Arbor: University of Michigan Press.

24.

Gabel

Scheve

(2007) Estimating the effect of elite communications on public opinion using instrumental variables. American Journal of Political Science 51(4): 1013–1028.

25.

Hagemann

Hobolt

Wratil

(2017) Government responsiveness in the European Union: Evidence from Council voting. Comparative Political Studies 50(6): 850–876.

26.

Healy

Lenz

(2014) Substituting the end for the whole: Why voters respond primarily to the election-year economy. American Journal of Political Science 58(1): 31–47.

27.

Hix

(1999) Dimensions and alignments in European Union politics: Cognitive constraints and partisan responses. European Journal of Political Research 35(1): 69–106.

28.

Hobolt

de Vries

(2015) Issue entrepreneurship and multiparty competition. Comparative Political Studies 48(9): 1159–1185.

29.

Hobolt

Klemmensen

(2008) Government responsiveness and political competition in comparative perspective. Comparative Political Studies 41(3): 309–337.

30.

Hooghe

Marks

(2009) A postfunctionalist theory of European integration: From permissive consensus to constraining dissensus. British Journal of Political Science 39(1): 1–23.

31.

Kayser

Lindstädt

(2015) A cross-national measure of electoral competitiveness. Political Analysis 23(2): 242–253.

32.

Kleine M and Minaudier C (2017) Negotiating under political uncertainty: Domestic elections and the dynamics of international politics. British Journal of Political Science. Epub ahead of print 06 February 2017. DOI: 10.1017/S000712341600051X.

33.

Kriesi

Grande

Lachat

et al. (2008) West European Politics in the Age of Globalization, Cambridge: Cambridge University Press.

34.

Lehmann

Matthieß

Merz

et al. (2015) Manifesto Corpus. Version: 2015-a, Berlin: WZB Berlin Social Science Center.

35.

Lindstädt

Vander Wielen

(2011) Timely shirking: Time-dependent monitoring and its effects on legislative behavior in the U.S. Senate. Public Choice 148(1–2): 119–148.

36.

Lipset

Rokkan

(1967) Cleavage structures, party systems, and voter alignments: An introduction. In: Lipset

Rokkan

(eds) Party Systems and Voter Alignments: Cross-National Perspectives, New York: The Free Press, pp. 1–64.

37.

(2013) An electoral connection in European Parliament voting. Legislative Studies Quarterly 38(4): 439–460.

38.

Mansbridge

(2003) Rethinking representation. American Political Science Review 97(4): 515–528.

39.

Marks

Nelson

Edwards

(2006) Party competition and European integration in the East and West: Different structure, same causality. Comparative Political Studies 39(2): 155–175.

40.

Reh

Heritier

Bressanelli

et al. (2013) The informal politics of legislation: Explaining secluded decision making in the European Union. Comparative Political Studies 46(9): 1112–1142.

41.

Rickard

(2012) Electoral systems, voters’ interests and geographic dispersion. British Journal of Political Science 42(4): 855–877.

42.

Schneider

(2013) Globalizing electoral politics: Political competence and distributional bargaining in the European Union. World Politics 65(3): 452–490.

43.

Soroka

Wlezien

(2010) Degrees of Democracy: Politics, Public Opinion, and Policy, Cambridge: Cambridge University Press.

44.

Steenbergen

Edwards

de Vries

(2007) Who’s cueing whom? Mass-elite linkages and the future of European integration. European Union Politics 8(1): 13–35.

45.

Stevens

(2013) Issue evolution in Britain: The debate on European Union integration, 1964–2010. European Journal of Political Research 52(4): 536–557.

46.

Stimson

(1999) Party government and responsiveness. In: Przeworski

Stokes

Manin

(eds) Democracy, Accountability, and Representation, Cambridge: Cambridge University Press, pp. 197–221.

47.

Stimson

MacKuen

Erikson

(1995) Dynamic representation. American Political Science Review 89(3): 543–565.

48.

Thomson

(2011) Resolving Controversy in the European Union. Legislative Decision-Making before and after Enlargement, Cambridge: Cambridge University Press.

49.

Thomson

Arregui

Leuffen

et al. (2012) A new dataset on decision-making in the European Union before and after the 2004 and 2007 enlargements (DEUII). Journal of European Public Policy 19(4): 604–622.

50.

Toshkov

(2011) Public opinion and policy output in the European Union: A lost relationship. European Union Politics 12(2): 169–191.

51.

van de Wardt

(2014) Putting the damper on: Do parties de-emphasize issues in response to internal divisions among their supporters? Party Politics 20(3): 330–340.

52.

van der Brug

Franklin

Tóka

(2008) One electorate or many? Differences in party preference formation between new and established European democracies. Electoral Studies 27(4): 589–600.

53.

van der Eijk

Schmitt

Binder

(2005) Left–right orientations and party choice. In: Thomassen

(ed.) The European Voter, Oxford: Oxford University Press, pp. 167–182.

54.

van der Pas

Vliegenthart

(2016) Do media respond to party conflict? Debates on European integration in British, Dutch and German party manifestos and newspapers, 1987–2006. Political Studies 64(4): 260–278.

55.

Wlezien

(2005) On the salience of political issues: The problem with ‘Most Important Problem’. Electoral Studies 24(4): 555–579.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.55 MB

0.00 MB

0.02 MB

0.77 MB