Agenda-setting by the European Commission. Seeking public opinion?

ABSTRACT

Historically, the European Commission has followed an expert-based depoliticized route to gain attention for policy issues and the credibility to deal with them. Given growing politicization, we ask whether the Commission might increasingly seek citizens’ views and whether there is patterned variation. We provide the first mapping of special Eurobarometers, the massive instrument for issue-specific public opinion. We found a steep increase and a curvilinear pattern: public opinion is rarely invited in areas of exclusive European Union competencies and exclusive national competencies. Most special Eurobarometers focus on shared competencies. Citizens are almost never asked about expenditure programmes and never on immigration. There is large variation across the Directorates General, which is only weakly related to the amount of planned legislation and the number of expert committees. Business-oriented Directorates General are much less likely to seek public opinion. These results open up promising avenues for research on agenda-setting strategies at times of politicization.

KEYWORDS:

• Agenda-setting
• Eurobarometer
• European Commission
• legitimacy
• politicization
• public opinion

EU Agriculture Commissioner Dacian Cioloș welcomed the EUROBAROMETER poll's findings, stating today: ‘I have always underlined that the CAP is not just for farmers, but for all EU citizens. This poll confirms that some of the key concepts of our reform … have wide public support.’ (European Commission 2011: 1)

It is important to remember that the Eurobarometer is an instrument created and financed by a political institution. It is therefore inconceivable that it could somehow damage that institution with the publication of adverse results in this regard. (Signorelli 2012: 69)

Introduction

The European Commission has privileged access to agenda-setting. Historically, the Commission followed a depoliticized route to gain attention for policy issues and credibility to deal with them. The Commission typically relied on selected experts from member state governments, interest groups and scientific institutions. This practice has been arguably enabled and legitimized by the existence of broad, though diffuse, citizen support for European integration. Over the last decades, however, this ‘permissive consensus’ has been replaced by increased politicization and lower levels of trust in European Union (EU) institutions (e.g., Hooghe and Marks 2009). Against this background, we wonder whether actors in the Commission increasingly reach out to the public, not in the sense of actually mobilizing citizens, but in seeking public opinion through surveys on specific topics. The Commissions’ Directorates General (DGs) could use this public opinion, on issues they want to move forward, to signal popular legitimacy to other actors in the decision-making process.

The Commission ran surveys since the early 1970s to enquire about general problem perceptions and attitudes towards the EU (Standard Eurobarometers). Parts of these very comprehensive surveys contain batteries of questions on concrete policy issues and the appropriate political level to deal with them. The results are communicated in comprehensive and publicly available reports (European Commission n.d.). The potential of these so-called Special Eurobarometers (special EBs) as a resource in EU decision-making seems obvious: a special EB demonstrating that a large majority of ‘European’ citizens wants the EU to act on a specific policy topic or supports the Commission’s solution to a policy problem would be a powerful resource for the Commission. Despite their resource potential, special EBs have so far received no systematic attention in the literature.

In this article, we will document the development of special EBs over time. Since the Commission can choose the topics for special EBs, but cannot prevent publishing the results if these are deemed unfavourable, we will investigate which topics have been included in the special EBs and which have been neglected. In addition, we will analyse whether the Commission is more likely to seek public opinion in areas where it has comparatively more competencies or in areas where it has weak or no competencies.

Additionally, we will zoom in on the DG level. Are some DGs more likely to invite citizens’ opinion than others? Are DGs who invite more citizen opinion also more active in terms of agenda-setting activities, such as planned legislative proposals? Do DGs who rely on expert group input rely less or more on public opinion as an alternative input source? And finally, do non-governmental organization (NGO)-oriented DGs, who are more susceptible to public frames, invite more public opinion than business-oriented DGs?

To be sure, there have already been critical examinations of the Eurobarometer surveys. However, these studies typically concern methodological issues such as sampling problems and cultural biases (e.g., Saris and Kaase 1997). The study by Höpner and Jurczyk (2012) takes a more ‘political’ approach, by linking methodological problems with a potential pro-integration bias of the Commission. This study did focus, however, on the framing of questions and not on the selection of topics.

Hence, the goal of the article is to systematically map public opinion-seeking through special EBs along a number of dimensions. Although these dimensions are theoretically informed, we do not aim to causally explain the temporal or cross-sectional patterns found. Causal analysis often implies a more narrow focus, limited by the demands of a specific theoretical debate (Gerring 2012). Though we fully agree that causal explanation is the core task of the social sciences, we side with Gerring (2012) that discerning descriptive patterns is an important scientific task in its own right. Comprehensive description is particularly valid when it concerns phenomena that are intrinsically relevant and where knowledge is limited, such as public opinion-seeking through special EBs. It is in this spirit that we have cast our net wide in this article.

The next section will embed our endeavour in the agenda-setting literature and substantiate our selection of dimensions for mapping the special EBs. We will then provide background information about the Eurobarometer surveys and further elaborate on the coding of special EBs. This is followed by the empirical analysis. In the conclusion, we sketch research questions for future research.

Agenda-setting in the EU: from a depoliticized route to public opinion?

The literature on agenda-setting in the EU has emphasized that the European Commission has followed a technocratic depoliticized route to agenda-setting, resulting into a ‘creeping’ task expansion of the EU (Pollack 1994; see also Princen and Rhinard 2006; Wallace and Smith 1995). To get attention to the issues it wants to move on and to build credibility as the actor to do so, the Commission utilizes both Commission expert committees, wherein scientific experts, member state civil servants and specialists from interest group are supposed to provide non-partisan advice, and stakeholder consultations (Gornitzka and Sverdrup 2011; Haverland and Liefferink 2012; Kohler-Koch et al. 2013; Rimkute and Haverland 2015; Van Schendelen 2003; Wille 2013).

There is ample case study evidence that the Commission has acted as a skilful policy entrepreneur, using its garnered expertise, and its advantage of institutional persistence, its unique position as process manager, and its ample mediating skills to exploit emerging windows of opportunity (Heritier 1997; Pollack 1997: 125–6, Smyrl 1998; Wonka 2008: 1146). For instance, Cram has observed that the Commission has acted as a ‘purposeful opportunist’, that is, as ‘an organization which has a notion of overall objectives and aims, but is quite flexible as to the mean of achieving them’ (1993: 143). Even when the Commission has been in a constitutionally weak position, such as in the social policy area, its officials have at times proved strategists ‘able to marshal innocuous-looking instruments to achieve surprising results’ (Wendon 1998: 340).

These ‘low politics’ agenda dynamics, revealed in studies focusing on the 1970s and 1980s, reasoned well with the ‘permissive consensus’ of citizens on the benefits of EU integration that characterized the EU until the early 1990s (Hooghe and Marks 2009). Therefore, it comes to no surprise that:

[w]hereas … public opinion play[s] important roles in studies of domestic agenda setting processes, they are hardly mentioned let alone analysed systematically in studies of EU agenda-setting. (Princen 2011: 940)

In recent years, however, the EU can no longer build on the diffuse support of its citizens. The EU has become more salient to voters and present in the media, and subsequently has become an issue of domestic party competition, as demonstrated by a rise in Eurosceptic parties. The ‘permissive consensus’ has been replaced by a ‘constraining dissensus’ and complaints about the democratic deficit of the EU have increased (Hooghe and Marks 2009; see also Loveless and Rohrschneider 2011). This change has inspired studies on the EU priorities of the general public (e.g., Bevan et al. 2016), on the link between public opinion and EU policy-making (e.g., Bolstad 2015; Toshkov 2011) and on the agenda responsiveness of the European Council (Alexandrova et al. 2016). This study focuses on public opinion as a tool for the European Commission in agenda-setting activities. As the Commission consists of unelected officials, it may have a particularly strong incentive to consider citizens’ views. Hence, citizen support may become an important resource in agenda-setting, in addition to expert and stakeholder involvement (Haverland 2013). It could be used to substitute the direct electoral link with citizens, as a weak form of ‘input legitimacy’ (Scharpf 1999). Therefore, Directorates General in the Commission may invite public opinion to signal popular legitimacy to other DGs, as well as to the Council and to the EP further down the road of decision-making. The special EB might not be the innocuous instrument it appears at first glance.

A systematic plausibility probe is needed to assess the probability that public opinion could be a tool for ‘input legitimacy’ in the Commission’s agenda-setting behaviour. Such a probe would require both a mapping of the overall picture of variation in the Commission’s public opinion-seeking behaviour and a more in-depth mapping of the differences between DGs within the Commission.

First, the topics on which the Commission invites public opinion need to be mapped. The agenda-setting literature argues that political actors seek to use the agenda-setting stage to gain attention to issues they are interested in and to gain the credibility to deal with them (see Princen 2011: 929–30). In terms of public opinion, one would expect that the Commission focuses its public opinion-seeking on areas of weak or no competencies, as there is a more pressing need to gain attention and credibility for these topics. To shed more light on topic selection, we relate the frequency of topics in the special EBs to the treaty competencies of the EU and analyse whether the Commission is more likely to seek public opinion in areas where it has comparatively more competencies or in areas where it has weak or no competencies.

Second, bureaucratic politics within the Commission influence the overall agenda-setting activities of the Commission (Hartlapp et al. 2013). In what is probably the most comprehensive study of more recent agenda-setting activities within the Commission, Hartlapp and her collaborators hypothesized public legitimacy as one of the factors affecting DG power in the Commission’s internal position formation processes (Hartlapp et al. 2014). They systematically compared 48 legislative proposals, which were adopted between 1999 and 2009 and which spanned three policy areas. Public legitimacy was among the factors which shaped nine of those proposals. For some of these cases, officials have confirmed that special EBs have been explicitly commissioned to support the position of the DG (ibid.: 235). As Special EB are requested by individual DGs, we can compare the frequency of public opinion-seeking across components of the Commission. Are some DGs more likely to invite citizens’ opinion than others? Is this related to DGs’ activism in terms of agenda-setting activities or by the reliance on other input sources (such as expert groups and NGOs–business actors).

To gain a more fine-grained picture of the role of bureaucratic politics within opinion-seeking behaviour, we focus on three dimensions that distinguish DGs from each other. To start, we compare the variation in special EB requests by DGs to the number of planned legislative proposals per DG. After all, those DGs that have a more active legislative agenda might also more frequently gather public opinion. In addition, we compare the established variation with the degree to which the DG relies on expert groups – the main alternative source of input. Finally, previous research and expert surveys have pointed out that some DGs are more business-oriented, while others are more NGO-oriented (Bernhagen et al. 2015). In a similar vein, some DGs are supposed to be receptive to an economic interest frame, while others are more receptive to a public interest frame (Klüver et al. 2015). Our hunch is that business-oriented DGs will undertake less effort to seek public opinion as compared to NGO-oriented DGs. One could argue that ‘business-friendly’ DGs derive their legitimacy primarily from business support and are therefore less dependent on public opinion. NGO-oriented DGs, on the other hand, are more receptive to a public interest frame and have more stake in public opinion.

The Eurobarometer: standard and special

The Eurobarometer has been conducted on a regular basis since 1973. It is probably the most comprehensive government-led enquiry into public opinion worldwide. Questions are posed to a representative sample of about 1,000 citizens per member state at least two times a year.¹ The EB is co-ordinated by the Commission, more specifically by the DG Communication, and currently carried out by Taylor Nelson Sofres (TNS) Opinion and Social. The standard Eurobarometer consists of a set of ‘trend’ questions that are posed repeatedly over time, such as attitudes towards European integration and socio-demographic characteristics (GESIS n.d.). The standard EB does entail some questions about the relative salience of policy problems broadly defined.

However, the large-scale surveys conducted under the label Eurobarometer do not only contain the above-mentioned categories of questions. In fact, citizens are also surveyed on a great variety of specific policy topics. Typically, batteries of questions regarding these specific topics are requested by various DGs of the European Commission, or sometimes other EU institutions, for in-depth thematic studies (about 100 pages long). These batteries are the special EBs (Signorelli 2012: 26). According to an official of the Commission’s Eurobarometer Unit, the final list of questions is decided on a high hierarchical level involving, for instance, the heads of the Commissioners’ cabinets and highly ranked communication advisors. Selection of the final list is based on the Commission priorities, which currently include employment and growth. The official felt unable to comment on the pre-Barroso period, though he expected the process to be more decentralized (European Commission 2014).

We focus on the special Eurobarometer as it is the Commission’s main and a massive instrument for seeking public opinion on specific policy topics.²

Coding special Eurobarometers

This article provides a systematic description of dimensions of all special EBs conducted from 1970 until December 2014, marking the end of the second Barroso term. The population of special EB reports, 399 in total, is available through the Commission’s website (European Commission n.d). The website mentions both the dates of the fieldwork for the survey and the publication of the report, which allows tracing of the frequency of these surveys over time.

First, we have coded the topics according to the EU codebook of the Comparative Agenda Setting Project (Alexandrova et al. 2013). The topics of the special EB were inferred from the title of the reports. If the title was ambiguous, the report was downloaded and read. The topics were coded manually by two researchers independently. In a second step, we have converted the data according to the definition of policy areas provided by the Lisbon Treaty, which was adopted in 2007 and enacted in 2009 (see Table 1). We grouped the policy areas in five competency levels, using the terminology of the Lisbon Treaty: exclusive competencies of the EU; shared competencies; specific arrangement; national competencies – EU supports; and exclusive competencies of the member states. This allows us to map the frequency of topics on the division of competencies between the EU and the member states and to check whether there is a relationship between the degree of EU integration in an area and the number of special EBs commissioned by the Commission. Furthermore, we checked when (i.e., year of Treaty enactment) each policy area roughly ‘reached’ the identified competency level, in order to identify whether Eurobarometers have been requested before or after the EU gained that competency level.

Table 1. Topics addressed in Eurobarometer mapped on division of competencies.

	No. of special EBs since Lisbon competence level achieved	No. of special EBs prior to Lisbon competence level	No. of special EBs per area in total	Level of competence since roughly
Exclusive competences of European Union
Competition	0	0	0	1958
Common Commercial Policy (External trade )	3	0	3	1958
Conservation of marine biological resources	0	0	0	2009
Customs union	0	0	0	1958
Monetary policy	6	0	6	1993 ECU: 1958
Shared competences
Agriculture and fisheries.	27	0	27	1958
Area of Freedom, Security and Justice	37	2	39	1993
Consumer protection	13	2	15	1993
Development co-operation and humanitarian aid	18	0	18	1958
Economic, social and territorial cohesion;	4	0	4	1958
Energy	2	12	14	2009 Nuclear: 1958
Environment	18	2	20	1987
Internal market	21	0	21	1958
Public health as defined in TFEU	23	6	29	1993
Research, technological development and space	28	2	30	1987
Social policy, as defined in TFEU (social regulation)	6	0	6	1958
Transport	8	0	8	1958
Specific arrangements
Common foreign and security policy	6	1	7	1993 Defence: 2003
Co-ordination of economic policies	6	1	7	1993
Co-ordination of employment policies	5	3	8	1999
Co-ordination of social policy	7	5	12	1999
National competencies – EU supports
Civil protection	1	1	2	2009
Culture	3	0	3	1993
Education	2	3	5	2009
Protection and improvement of human health	3	6	9	2009
Sport	1	3	4	2009
Tourism	0	2	2	2009
Vocational training	1	0	1	1993
Youth	0	2	2	2009
Exclusive competencies member states
Housing and urban development^a	0	0	0	n.a
Public lands and water management^a	0	0	0	n.a

Source: European Commission (1970–2014), own calculation, total: 303.

Note: We focused on the 367 special EBs that are commissioned by the European Commission (in general or by a specific DG). However, 40 special EBs are too general to be meaningfully coded into one of the categories of the agenda setting codebook, such as ‘social climate’. Also, we excluded those that focus on EU governance and government operations as such which reduces the number of observations from 327 to 320. Due to additional considerations provided in the Online Appendix, the final number analysed amounts to 303.

^a Added by authors, not mentioned in the TFEU.

In a third step, we coded the name of the DG which requested the special EB. All special EB reports, save one, contain the name of the DG which requested the special EB. In addition, 15 reports (4 per cent) are not available for download, thereby resulting in missing values for the DG requesting the EB and in some cases for the topic as well, as the title was too ambiguous to infer the topic code. For labelling the DGs, we could not always use their proper names at the time when the special EBs were requested. Otherwise, it would not be possible to study DGs over a long period.³ The portfolio of the Barroso I Commission informs our labelling. This implies that, for instance, we pool the data for DG Energy and DG Move into the category DG ‘Energy and Transport’, for DG Home and DG Justice into ‘DG Justice, Freedom and Security’, for DG Agriculture and Rural Development and DG Fisheries into ‘DG Agriculture’, and for DG Environment and DG Climate Change into ‘DG Environment’.

Increase in the production of public opinion?

The first question we posed was whether the European Commission increasingly seeks public opinion through special EBs. Indeed, the population of special EBs has increased dramatically over time. Before the 1990s, hardly more than five special EBs a year had been conducted. In the 1990s, the average was almost 10 special EBs a year, and in the 2000s the average further increased to almost 20 (see Figure 1). From the 1980s onwards, each Commission has executed more special EBs than its predecessors, with the exception of the short-lived Delors III Commission. The results also show increased requests of the European Parliament, and occasionally other actors, for special EBs. For the remainder of the article, however, we want to focus on the European Commission, which is responsible for 367 of the 399 special EBs.

Figure 1. Number of special EBs a year per Presidency. Source: European Commission (1970–2014), own calculations, total: 399.

Notes: The start and end dates of each EC Presidency were found at http://ec.europa.eu/archives/commission_2004-2009/president/history/. Eight special EBs were commissioned by the European Parliament and retrieved from http://www.europarl.europa.eu/aboutparliament/en/00191b53ff/Eurobarometer.html.

Topics addressed and neglected

The second question we posed was whether the Commission invited public opinion in areas in which it has exclusive or shared competencies or in areas with weak or no competencies.

As explained in the coding section, we mapped the frequency of topics on the division of competencies between the EU and the member states in order to see whether a relationship exists between the degree of EU integration in an area and the number of special EBs commissioned. In addition, we distinguished Eurobarometers that were requested before the policy area ‘gained’ the competency level, as identified in the Lisbon Treaty, from Eurobarometers that were requested after the policy area gained its current competency level.

Overall, we find a roughly curvilinear pattern with relatively fewer special EBs in areas where the EU already has far-reaching competencies and in areas where either policy issues are clearly (sub-) national or the EU involvement is limited to ‘support’ the execution of the national competence. The lion’s share of special EBs is conducted in areas of shared competencies.

Starting with the EUs exclusive competencies, we identified not a single special EB for either common commercial policy, competition policy (merger control, etc.), and common fisheries policy (conservation of marine biological resources). The EU has strong competencies in these areas since 1958. Six special EBs have dealt with monetary policy, specifically with the European Currency Unit or the euro. None of them were requested after 2002, however, despite the evolving crisis in the eurozone since 2008.⁴ Overall, the European Commission does not prioritize public opinion in areas where the EU (already) has exclusive competencies.

The same holds for areas of exclusive member state competencies. No special EBs have been conducted on housing and urban development or public lands and water management. A few special EBs have been conducted in areas that, until Lisbon, have been exclusive national competencies and where the EU now ‘supports’ national action. Most of these EBs have been requested prior to that treaty change. Special EBs on the protection of human health are the most frequent in this category, with six before and three after it has assumed ‘EU support’ competency level in the Lisbon treaty.

Moving ‘up the ladder’ again towards more integrated areas, we arrive at the so-called specific arrangements. Specific arrangements use intergovernmental bargaining and co-ordination, rather the Community method, for decision-making. In terms of economic co-ordination, six special EBs have been conducted since the EU assumed specific arrangements in this area. Some of these EBs deal with rather prosaic issues, such as the European citizens’ knowledge of economic indicators. There has never been a special EB on the budget. This implies that, at least up to 2015, the Commission did not invite citizens to provide their opinion on the economic crisis in the eurozone and recent European economic governance policies through the Special Eurobarometer. Note, however, that two special EBs on the economic crisis have been conducted on behalf of the European Parliament.⁵ The Commission has been more active in the area of social policies, in the sense of welfare state policies, which are since the Amsterdam Treaty (1999) subject to the Open Method of Co-ordination (OMC). There have been 12 special EBs in total, with some of them dealing with poverty and social exclusion. Most of the eight special EBs in the area of employment policies are close to social issues as well. Almost half of the special EBs in these two areas precede the establishment of competencies in the Amsterdam Treaty. Only seven special EBs focus on security and foreign policy, six of them after Maastricht has created the second pillar. Note that the Commission never conducted a special EB on defence.

The lion’s share of special EBs is conducted in the areas of shared competencies. Perhaps not surprisingly, many of them concern the single market, the core project of EU integration. Twenty-one focus on various aspects of the internal market (about half of those related to services), taking mostly a consumer perspective. In addition, 15 special EBs focus on consumer protection in a narrower sense, 13 of them since the Maastricht Treaty turned this area into a shared competence.

Many surveys were requested in the area of freedom, security and justice. Thirty-nine have been adopted and almost all of them in the last decade. Eighteen deal with civil rights issues, such as EU citizenship, discrimination and data privacy. Twenty-one deal with law and crime issues such as violence and white collar crimes like corruption.

At the same time, there has not been a single special EB that has focused on immigration issues, such as refugees and asylum, the integration of immigrants, acquisition of nationality or border control.

There are many surveys concerning agriculture. Seven of them deal with food safety and seven with biotechnology. However, none of the 27 surveys dealt specifically with what might be regarded the core of the CAP, agricultural subsidies and the common organization of agricultural markets. Some surveys include issues in these domains, such as ‘capping’, but they are not a prominent part of these more general inquiries. This implies that citizen opinion is invited on the regulatory aspects of agriculture as opposed to the redistributive aspects of this large expenditure programme.

That special EBs are rather silent on expenditure is further confirmed by the low number of special EBs on cohesion. In addition to two rather generic special EBs on Europe of the Regions, there has been only one more directly related to territorial cohesion policies and there has been only one survey on the European Social Funds. Both surveys are of older vintage: for the last two decades, citizens’ opinion was not invited on the EU cohesion funds, the largest expenditure programmes of the EU.

In terms of timing, almost all special Eurobarometers have been adopted after the member states shared competencies with the European Union. The exception is energy; European citizens have been regularly quizzed about energy security long before the Lisbon treaty gave the EU a say in this area.

To summarize, special Eurobarometer have largely been conducted in policy areas in which the EU and the member states share competencies, particularly after the areas gained their current competency level. No Eurobarometers have been conducted on immigration and almost none on the redistributive expenditure programmes, even though these areas have shared competencies.

Variation within the Commission in gauging public opinion

The third question we posed was whether variation between DGs exists in terms of public opinion-seeking behaviour.

The general picture

As stated above, special EBs are not requested by the European Commission as such but rather by a specific Directorate General. Although the Commission is organized according to the principle of functional specialization, topics are not always ‘owned’ by a specific DG. Many topics cut across several DGs, requiring co-ordination and potentially encountering conflict and bureaucratic politics (e.g., Hartlapp et al. 2014). It is therefore worthwhile to map which DG has actually requested the special EB.

We found that DGs vary starkly in their effort to gauge public opinion (see Figure 2). Three DGs are responsible for about half of the special EBs. Quite a few special EBs are carried out by DG Communication itself. Further analysis reveals that these EBs typically focus on general topics or topics that cut across many issues, such as the future of the EU or German re-unification. From the DGs with a policy portfolio, DG Employment and Social Affairs and DG Health and Consumer Affairs are the directorates who request considerably more special EBs than the other DGs.

Figure 2. Number of special EBs per DG. Source: European Commission (1970–2014), own calculations, total no. 352.

On the other end, some DGs never or almost never seek the opinion of European citizens through special EBs. The following group of nine DGs are responsible for in total less than 5 per cent of all special EBs (16/352)⁶: DG Enterprise; DG Regional Policy; DG Trade; DG Economic and Finance; DG External Relations; DG Enlargement; DG Taxation and Customs Union; DG Competition; and DG Budget. The latter two have never requested a special EB.

In other words, in particular those DGs tasked with economic and foreign affairs issues rarely invite the opinion of European citizens. This is consistent with the results of our previous section, where we have seen that there are indeed only very few special EBs in the area of foreign and security policies and core areas of economic policies areas such as competition policy, external trade policy, taxation and the budget.

The large number of special EBs for DG Health and Consumer Protection focuses on activities in both core areas of its responsibility: health and consumer protection. This DG is also responsible for half of the 21 Special EBs dealing more specifically with the functioning of the single market. DG Internal Market and DG Enterprise are only responsible for a total of seven of the single market EBs. The large number of special EBs requested by DG Employment and Social Affairs results from the DG’s request of almost all special EBs in the areas of social policy and of employment policy. In addition, this DG is also responsible for five of the 14 special EBs in the category of civil rights, namely those focusing on discrimination of, for instance, disabled people.

Furthermore, a longitudinal perspective suggests that DGs of relative recent vintages and tasked with subjects that have not been part of the original European Economic Community (EEC) Treaty are relatively eager to invite public opinion. For example, since Barroso II, the DG in the area ‘Justice, Freedom and Security’ has been split into DG Justice and DG Home Affairs. Both have started to request special EBs. In addition, relatively recent DGs dealing with the Information Society, Communication Networks and Education and Culture have conducted special EBs from the start.

Variation across DGs in context

To further explore the variation between DGs in their effort to invite public opinion, we have taken into account three dimensions. First, we have related the number of special EBs to legislative agenda-setting activities. Perhaps DGs do not invite public opinion because they do not plan new legislation and have no agenda-setting goals. Second, we have looked at the number of expert committees per DG. It might be that those DGs that rely more on input of expert undertake less effort to gather public opinion. And finally, we have looked at the type of DG, in term of proximity to business and the economic framing of issues as opposed to proximity to NGOs and the public framing of issues. For reasons of data availability, this analysis will focus on the two Barroso presidencies only (2005–2014). We exclude DG Communication in this analysis, because the DG is not involved in legislative agenda-setting activities.

To measure legislative agenda-setting activities, we have taken the number of planned proposals per DG announced in the Commission work programmes. We consider this a relatively valid measure of legislative agenda-setting behaviour, as it ensures better comparability across DG’s than, for instance, the number of white papers, green papers and communications. We use the data provided by Osnabrügge (2015). Visual inspection reveals that the number of Special EBs are only weakly associated to the number of announced proposals (see Figure 3). Additional bivariate statistical analysis confirms this intuition and demonstrates that the association is not statistically significant (r = 0.34, sig. 0.13). To be sure, some DGs were both active legislative agenda setters and frequently request special EBs. Examples are DG Health and Consumer Protection and DG Justice, Freedom and Security. Others were also relatively active legislative agenda setters but (almost) never invited public opinion. Examples are DG Agriculture and Fisheries, DG Economic and Finance, DG Enterprise, DG Internal Market and DG Environment.

Figure 3. Special EBs per DG in context. Source: Special EBs: European Commission (2005–2014), own calculations, n=176; WP Planned proposals (2005–2012), Osnabrügge (2015: 245), n = 232; Expert committees (2010), Metz (2012), n = 896. Type of interest groups, Bernhagen et al. (2015: 577–8), Klüver et al. (2015: 490).

Is there a clear pattern when it comes to the sources of input for agenda-setting? Examining the results suggests a negative answer (see Figure 3). In general, public opinion is not used as an alternative to expert input. If at all, there is a very weak positive relationship in the population between the number of special EBs and the number of expert committees per DG, but it is not statistically significant (r = 0.26, sig. 0.26). If we concentrate on those DGs who are rather active legislative agenda setters (> 10 planned proposals), two DGs are responsible for a relatively large share of special EBs in this period: DG Health and Consumer; and DG Justice, Freedom and Security. They do rely less on expert committees, but not distinctly less than most others. DG Agriculture, DG Enterprise, DG Environment and DG Internal Market rely relatively more on expert committees than on special EBs. DG Energy and Transport roughly balances the number of special EBs and the number of expert committees.

We have found a rather strong pattern however, when it comes to the type of DG. The seven business-oriented DGs and the five public/NGO-oriented DGs score very similarly on legislative agenda activity. Directorate Generals in both groups are on average responsible for 7 per cent of the work programme proposals. Yet the business-oriented DGs on average commissioned 2 per cent of all special EBs in the period under study, while the public/NGO-oriented DGs commissioned on average 10 per cent of all special EBs. A comparison of means test demonstrates that a statistical significant result, despite the small number of observations (Eta = 0.75, sig. 0.005).

Discussion and conclusion

Against the background of an increased politicization of the EU, this article presents the first systematic description of public opinion-seeking through special EBs. We found that the amount of special EBs has dramatically increased from almost none in the 1980s to almost 20 a year in the 2000s. The effort to seek public opinion through special EBs is, however, not equally distributed across policy areas. Generally, citizens are not often invited to voice their opinion in areas in which the EU either already has far reaching competencies, such as external trade and competition, or areas in which the EU has none, such as public lands and water management. Citizen input is foremost sought in areas of shared competencies. Citizens are largely kept silent when it comes to EU expenditure and there has never been a special EB on immigration.

We also see a large variation between DGs. DG Communication, DG Employment and Social Affairs and DG Health and Consumer Affairs are responsible for almost half of all special EBs. Most DGs who focus on economic issues and issues of external affairs almost never request special EBs. The variation is only weakly associated with legislative agenda-setting activity, in terms of planned legislative proposals. No clear pattern is visible in DGs’ relative reliance on public opinion and on expert advice in terms of number expert committees. However, a strong pattern emerges when it comes to the DGs proximity to business interests versus NGOs. The ‘business-friendly’ DGs are on average five times less likely to invite public opinion than NGO-oriented DGs, although they work on a similar number of legislative proposals.

This systematic description of the important but neglected phenomenon of the Commission’s public opinion seeking through special EBs raises puzzles which merit further research. Can the curvilinear relationship between the degree of EU competencies and the frequency of special EBs be married with an agenda-setting perspective? To be sure, it fits our expectation that the Commission has shown little appetite to inquire citizens’ opinion in areas of exclusive EU competencies. However, why has the Commission not been more active in areas that belong to the national domain? That special EBs are most frequent in areas where competencies are shared suggests a more subtle relationship than initially expected. The Commission invites public opinion as a resource in areas where its agenda-setting authority is neither obvious nor unachievable and where it is in clear competition with the member states for gaining attention and credibility. There is additional variation that also requires further research.

Can variation be explained by characteristics of the policy area and the likelihood of public opinion results that are to the liking of the Commission? Will this explain why the Commission eschews surveys on redistributive issues, which relatively visibly produce winners and losers? Will this explain the neglect of immigration issues, which may touch too much on identity and therefore be classified as too sensitive? These puzzles open up promising avenues for research on agenda-setting strategies at times of politicization.

Acknowledgements

We thank René Karens, Kyra Luchtenberg and Reinout van der Veer for able research assistance, and Petya Alexandrova, Miriam Hartlapp, Steven Van Hecke, Beate Kohler-Koch, Sebastiaan Princen, Arco Timmermans and Anchrit Wille for their insightful comments on previous versions of the article. We would also like to thank two anonymous reviewers for their helpful comments. This is a much revised, extended and updated version of LEQS working paper 2015, No. 88.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Markus Haverland is professor of political science at the Department of Public Administration, Erasmus University Rotterdam.

Minou de Ruiter is a PhD candidate at the School of Governance, Utrecht University.

Steven Van de Walle is professor of public management at the Public Governance Institute, KU Leuven.

Additional information

Funding

We acknowledge funding from the European Union’s Seventh Framework Programme under Grant Agreement No. 266887 (Project COCOPS), Socioeconomic Sciences and Humanities.

Notes

1 The exceptions are Luxembourg, Malta and Cyprus, with about 500 interviews per country.

2 In addition, since 2000, ‘Flash’ EBs have also been conducted. They consist of telephone interviews, rather than face-to-face contact, and often focus on specific target groups, such as entrepreneurs or ‘the youths’, or on one or a few countries. Therefore, they pose difficulties in terms of comparability of topics over time. There were 405 flash EBs until November 2014. Given their small-scale focus, they are unlikely to carry similar weight as the special EBs. We have initially included them in the analysis, but we found no patterns that would contravene our conclusions from the special EBs. There have also been 32 qualitative studies under the umbrella Eurobarometer. Given their small number, small scope and the diversity of topics addresses and of initiating DGs, they do not impact our conclusions either.

3 Since the start of the Eurobarometer in the early 1970s, new DGs have been created, DGs have been split up, DGs have assumed new tasks and discarded old ones, and accordingly the number, their tasks and their names have constantly changed.

4 Note, however, that since 2000 flash EBs annual monitor public perceptions concerning euro-related issues for eurozone citizens (Commission n.d.).

5 Note also that the annual flash EBs concerning euro-related issues also contain questions about European economic governance (Commission n.d.).

6 We were unable to identify the responsible DG for 15 special EBs, because these EBs were not available for download.