Comparing oppositional behaviour in European Union affairs: empirical evidence from six member states

While there seems to be widespread agreement in political as well as academic circles that national parliaments should play an active role in European Union affairs, no consensus has been reached on the extent to which they actually do so. For a long time, research efforts directed at examining the role of national parliaments in EU affairs were mainly focused on the formal scrutiny powers enjoyed by national parliaments (e.g. Raunio Citation2005; Strom et al. Citation2003; Winzen Citation2012). Studies on parliamentary activity in EU politics have only recently started to emerge (Auel and Raunio Citation2014; Auel et al. Citation2015a; Hoerner Citation2017; Senninger Citation2017; Rauh and de Wilde Citation2018). However, an important part of the puzzle, when it comes to the role played by national parliaments in EU politics, is still missing as we have seen very few studies that examine parliamentary behaviour in EU affairs from the theoretical perspective of political opposition. While there are some notable exceptions (Albrecht et al. Citation2019; Christiansen and Damgaard Citation2008; De Giorgi and Ilonszki Citation2018; Persson et al. Citation2019), there is still a glaring shortage of research on oppositional behaviour in EU affairs, and especially studies that address this issue from a comparative perspective.

This article seeks to bridge the existing gap in research on parliamentary opposition in EU affairs by offering a focused comparison of oppositional behaviour in European Affairs Committees (EACs) in six member states: Austria, Denmark, Germany, Ireland, Sweden and the United Kingdom. With the purpose of shedding new light on political opposition in EU affairs we develop an analytical framework grounded in classical scholarly work by Kirchheimer (Citation1957) and Dahl (Citation1966). Equipped with this framework, we examine the deliberations in six EACs in search of an answer to the overarching research question: how much and what kind of opposition exists in EU affairs? More specifically: has oppositional behaviour been directed at the content of EU policies or rather against the EU political system, and has opposition been voiced in the form of delivering critique or presenting alternatives? Throughout the analysis we will compare patterns of oppositional behaviour between political parties as well as examining cross-national variation.

While opposition in EU affairs can manifest itself during plenary debates and meetings of specialised standing committees in Parliament, the EAC is the one arena where the government and MPs meet regularly to discuss proposals from the European Commission, ongoing negotiations in Brussels and the positions taken by the government on policy issues. The EAC is thus an important arena for interaction between the government and the political parties represented in parliament, and as such a key forum to investigate if we want to improve our understanding of opposition in EU politics. By looking at EAC deliberations rather than plenary debates we are focusing on what goes on ‘behind the scenes’ (c.f. Lord Citation2018: 39) in national parliaments. This may prove advantageous from a methodological point of view as it is often argued (Lord Citation2018; Roger Citation2016) that what goes on backstage in the committees provides a more accurate picture of actors’ true preferences, than what plays out at the frontstage during plenary debates. There are thus a number of good reasons why a closer inspection of EAC deliberations is warranted as we seek to uncover patterns of oppositional behaviour. However, looking at previous research it is striking that we find very little research that is focused on examining what actually goes on inside the EACs when representatives of the government and MPs discuss EU affairs. The examination that follows will therefore contribute to opening up the black box of EAC deliberations in search of opposition in EU affairs.

It is worth emphasising that we are not content with simply examining activity that may or may not be evidence of oppositional behaviour, e.g. the frequency of EAC meetings, the number of mandates and resolutions, or the time spent discussing EU issues. Instead, we take things one step further by examining the actual content of one such key activity, namely the EAC deliberations. In this way we will get a more precise measurement of oppositional behaviour than had we simply relied on indicators of parliamentary activity. The study thus analyses statements made during EAC meetings in six countries for the period 2005–2016. In total, this comprehensive data set consists of 8945 hand-coded statements.

If citizens are deprived of opportunities to organise opposition within a political system, there is an obvious risk they will end up opposing the democratic polity as such (Dahl Citation1966; Mair Citation2013). Recent political developments in the European Union (EU) seem to illustrate this dangerous dynamic. The alleged ‘opposition deficit’ (Rauh and De Wilde Citation2018) in EU affairs has thus come to work as a breeding ground for Euroscepticism, right wing populism and anti-democratic sentiments in a number of EU member states. Against this backdrop, it is crucial that we improve our understanding of how much and what kind of oppositional behaviour exists in EU politics.

This article starts by providing an overview of the research field. Next, we discuss the concept of political opposition and present our analytical framework. This is followed by a section on case selection, data and coding. We then turn to an analysis of our data and report on the study’s empirical results. The article ends by us offering our conclusions on oppositional behaviour in EU politics.

Political opposition – survey of the field

For two decades scholars have argued that EU politics is characterised by an absence of opposition to the policies promoted by the EU (Mair Citation2007, Citation2013; Neunreither Citation1998), but we have yet to see much concrete evidence to support this claim. Looking at previous research we find a number of relevant studies that have examined the development of parliamentary EU activity and scrutiny at the domestic level (e.g. Raunio Citation2005; Winzen Citation2012). However, this important strand of research has mainly focused on the presence of formal oversight mechanisms and opportunity structures, rather than actual behaviour. Only recently have research on domestic parliamentary activity (e.g. Auel and Raunio Citation2014; Auel et al. Citation2015a; Rauh and de Wilde Citation2018) – the way national parliaments engage in debates and resolutions on EU affairs – and inter-parliamentary activity (e.g. Miklin Citation2013) started to emerge. These studies have delivered important insights regarding the role played by national parliaments in EU affairs and how parliamentary activity differs between member states (Auel et al. Citation2015a). Still, this literature has not confronted oppositional behaviour head on by examining the content of scrutiny activity. The frequency of scrutiny activity can of course be used as a proxy for oppositional behaviour but it does not speak directly to how much opposition actually exists in EU politics.

There are, however, a couple of recent exceptions that focus on the content of parliamentary activity. Senninger (Citation2017) examines EU-related parliamentary questions in search for differences between mainstream parties and Eurosceptic parties. The results show that Eurosceptic parties tend to emphasise general polity-related aspects of EU affairs in their parliamentary questions, whereas mainstream parties are more prone to focus on specific policy-related issues. Similarly, Hoerner (Citation2017) studies party variation in terms of support for the government’s position on EU policies and find that Eurosceptic parties (‘issue entrepreneurs’) to a higher extent than mainstream parties tend to criticise the government. At the same time, the Eurosceptic parties ‘are mostly concerned with general, politicized and Eurosceptic statements’ in their motions but ‘fail to make an impact on the “bread and butter” issues of EU politics’ (Hoerner Citation2017: 320).

Both these contributions provide important insights into political opposition in EU affairs, but neither provides a comprehensive picture of oppositional behaviour. Senninger’s study does not address the extent to which parliamentary questions contain any expressed disagreement with the government. Hoerner’s study does say something about the content of parliamentary motions and resolutions and the extent to which these are ‘critical or supportive of the government’ (Citation2017: 312). However, opposition is not only about delivering critique it is also, crucially, about presenting alternatives and this key dimension of opposition is not addressed by Hoerner. In another recent article Karlsson and Persson (Citation2018) do address opposition in the form of critique as well as presenting alternatives. Their results imply there exists more opposition in EU politics than what has often been assumed, but since this is a case study of only one member state (Sweden) it fails to provide a general picture of opposition in EU politics.

In conclusion, when looking at previous studies we find a shortage of empirical work on opposition, especially when it comes to examining actual oppositional behaviour in a comparative light.

Political opposition – analytical framework and theoretical expectations

This study proceeds from the theoretical understanding of opposition as an expression of disagreement with the government (Dahl Citation1966). This is a useful starting point, but two distinctions need to be made before it can serve the purpose of helping us uncover oppositional behaviour. First, we need to differentiate between two main types of opposition: disagreement in the form of critique directed towards the government, and disagreement in the form of presenting alternatives to the positions taken by the government. Expressing critique serves two purposes. On the one hand, it works as a means for controlling the government (Garritzmann Citation2017: 2), on the other hand, voicing critique allows the opposition parties to distance themselves from the government and make themselves more visible to the voters (Norton Citation2008: 238). However, opposition is also a matter of presenting alternatives, and this type of opposition is important since it introduces choice in politics. It is only by being presented with real choices that citizens will see elections as meaningful.

Second, we need to distinguish between three modes of opposition (cf. Kirchheimer Citation1957). What has been labelled ‘classical’ opposition is directed at the policies proposed and implemented by the government. We refer to this as policy opposition. The second mode of opposition contains expressions of disagreement directed at the political system. This has previously been labelled ‘opposition of principle’, but we shall simply refer to it as polity opposition. Opposition may also be directed towards the conducting of politics, such as how the government acts rather than what policies it promotes. This mode of opposition will be referred to as procedural.

Our research endeavour will thus be guided by a parsimonious analytical framework that examines the type and mode of opposition. To get a comprehensive picture of political opposition in EU politics, we will not be content with simply uncovering national variation when it comes to how much and what kind of opposition exists in EU politics. By also looking at variation in oppositional behaviour between political parties we will offer the most comprehensive study to date of opposition in EU politics. Whether a political party is in government or in opposition will very likely be the number one determinant of oppositional behaviour. However, since it is more or less a truism that opposition parties will express more disagreement with the government than governing parties, we have decided instead to focus our examination on a comparison between mainstream parties and Eurosceptic parties.Footnote¹ Following established practice we define parties with a CHES (Chapel Hill expert survey) score lower than 4 as Eurosceptic (see Meijers Citation2017: 417; Ray Citation2007: 159). For the sake of completeness, we include data on the differences in oppositional behaviour between governing and opposition parties in our six countries (see Online Appendix D).

What theoretical expectations can be formulated in view of previous research? When it comes to the most basic question of how much opposition we may expect to find there are two competing scholarly views in the literature. On the one hand we find a number of studies that argue opposition to be a scarce commodity in EU politics (Mair Citation2007, Citation2013; Neunreither Citation1998); that there is an ‘opposition deficit’ (Rauh and De Wilde Citation2018). Peter Mair has pushed this line of reasoning further than anyone else as he argues that ‘by becoming depoliticized itself, the EU also helps to depoliticize decision-making at the national level, tending therefore to bring even the member states to the status of polities without politics’ (Mair Citation2007: 8). According to what may be labelled the ‘depoliticization hypothesis’ we would thus expect that there is little opposition to be found in EU politics.

According to the competing view the importance of EU issues in national politics has been growing in recent years and we have seen a politicisation of the integration process (Hutter et al. Citation2016). EU politics has become an increasingly salient issue in national elections (Hoeglinger Citation2016; Hooghe and Marks Citation2009; Hutter and Grande Citation2014), and there is evidence to suggest that European issues are becoming increasingly contested in national parliaments (Auel et al. Citation2016). From these previous studies we may derive a counter-hypothesis to the one presented above. According to what may be labelled the ‘politicization hypothesis’ we would expect to find a significant amount of opposition in EU politics.

As we move to the party level it is more straightforward what we may expect to find on the basis of previous research. A key development in European politics in recent years is the birth and growth of new political parties that challenge the mainstream parties by focusing attention on the EU. These Eurosceptic challenger parties have successfully used their resistance to European integration to gain electoral success (de Vries and Hobolt Citation2012; Hobolt and de Vries Citation2015). Parallel to this development, previous research have argued that the political parties that are serious contenders for forming the next government have a common interest in not having ‘their hands tied by the constraints of popular democracy’ (Mair Citation2007: 7–8). In line with the cartel party-theory (Katz and Mair Citation1995) it is argued that this shared interest drives mainstream parties to abandon policy opposition on EU issues (de Vries Citation2007). From this line of reasoning we expect to find support for the hypothesis that Eurosceptic parties are the main drivers of opposition in EU affairs.

What theoretical expectations can be derived from previous research as we shift our attention from how much to what kind of opposition exist in EU politics? When it comes to the mode of opposition it seems logical to assume that Eurosceptic parties, which are fundamentally critical of the European integration project, will tend to focus more on polity than policy opposition. There is also some support for this assumption in previous research (Senninger Citation2017). As mainstream parties are in general strong supporters of European integration it stands to reason that they are likely to focus on policy rather than polity opposition to the extent that they voice disagreement with the government on EU politics. When it comes to the mode of opposition we thus expect to find support for the hypothesis that Eurosceptic parties will, compared to mainstream parties, engage more in polity opposition and less in policy opposition.

What, then, may be expected when it comes to the type of opposition in EU politics? As Eurosceptic parties are fundamentally critical of the integration project it seems reasonable to assume that they will focus more on delivering critique than presenting alternatives when they express opposition. In a study of parliamentary resolutions in five EU member states Hoerner found that resolutions initiated by Eurosceptic challenger parties are highly critical but ‘rather short and contain little detail’ (Hoerner Citation2017: 320). Mainstream parties, on the other hand, are contenders for forming the next government and as such they have stronger incentives to compete for votes by presenting alternatives to the positions taken by the government. When it comes to type of opposition we thus expect that Eurosceptic parties will, compared to mainstream parties, focus more on delivering critique and less on presenting alternatives.

Having presented the analytical framework and the theoretical expectations it is now time to turn to issues concerning research design and methods.

Case selection, data and coding

The current study examines EAC deliberations in Austria, Denmark, Germany, Ireland, Sweden and the UK. The selection of countries was made on the basis of a mix of theoretical and methodological concerns. It should be made clear that our purpose is not to explain cross-national variation, but to describe how much and what kind of opposition exists in EU politics. In order to obtain results that are generalisable beyond the scope of this study we have chosen cases strategically in light of factors that may reasonably influence variation in oppositional behaviour. If we look at member states characteristics often used to design comparative EU studies, it is far from obvious why factors such as country size, length of membership or geographical placement should matter in the context of political opposition. In the end we decided to focus on three factors that we believe are likely to drive the demand for and supply of opposition in EU politics.

First, we know there is substantial variation in public support for the EU. This means there is a higher demand, so to speak, for opposition in countries where a large proportion of the public is sceptical of European integration. Second, we may expect the supply of opposition in EU politics to co-vary with the strength of Eurosceptic parties in the national political system. Finally, national parliaments differ when it comes to the formal powers they are afforded in dealing with EU matters, and we expect to find more opposition the stronger the formal oversight mechanisms of national parliaments are.

In light of these potential drivers of opposition in EU politics, we have included member states where public support for the EU historically has been strong (e.g. Germany and Ireland) as well as member states where the public has been more EU sceptical (e.g. Sweden and the UK). In our six countries we also find variation when it comes to the share of seats held by Eurosceptic parties in parliament. In Austria, Denmark, Sweden and, in particular, the UK, Eurosceptic parties have held relatively many seats in parliament, whereas they have been less frequently represented in Ireland and Germany during the periods surveyed (see note to Table A2 in Online Appendix A). Finally, based on previous rankings of national parliaments in terms of their institutional strength in EU politics (cf. Auel et al. Citation2015b; Garritzmann Citation2017; Raunio Citation2005), our sample includes three high ranked (Denmark, Germany and Sweden, two mid ranked (Austria and the UK), and one low ranked parliament (Ireland).

Our examination of oppositional behaviour requires high quality data that allow us to analyse the content of EAC deliberations in detail. Whereas some EU member states only provide short summary protocols of EAC meetings, all of the six countries in our sample provide access to extensive protocols that allow for a detailed examination of the EAC deliberations. One notable difference between the countries is that the protocols in Ireland, Sweden and the UK are stenographic while the Austrian, Danish and German protocols have the form of summary records of EAC deliberations. The latter, however, are very comprehensive and allow for the same detailed coding as the former. Also, after having been in contact with the parliamentary archives in these countries we have no reason to suspect that anything of importance for the coding and analysis is being left out of the summary protocols.

In order to minimise bias and obtain estimates of oppositional behaviour that are representative of the population of EAC statements we have drawn random samples of 30 protocols from the EAC meetings in Denmark, Germany, Sweden and the UK for two legislative periods. We have thus analysed statements made during 60 EAC meetings in these four member states. The Austrian and the Irish EACs hold fewer meetings than its counterparts; we have therefore examined all EAC meetings in Austria (the first period) and Ireland (both periods) (for an overview of examined legislative periods and cabinets, see Online Appendix A).

During the periods surveyed, several extraordinary political events occurred that could potentially affect domestic patterns of oppositional behaviour in EU affairs. Dramatic events such as the constitutional crisis that started with the rejection of the Treaty on a Constitution for Europe in 2005, the eurozone crisis which erupted at the end of 2009 and intensified in the following years, and the refugee crisis in 2015 when unprecedented numbers of migrants arrived in the EU, triggered domestic debates and increased the politicization of EU issues. Studies have shown, for instance, that the eurozone crisis did affect parliamentary scrutiny patterns and raised the level of partisan conflict (e.g. De Giorgi and Moury Citation2015). Although our study is not designed to isolate the effect of these events, it does make it possible to at least reveal differences in patterns of opposition across policy areas and countries which are affected to varying degrees by the crises (see Table AC10 in Online Appendix C).

The units of analysis in our study are statements made by MPs during the EAC meetings. These statements are derived from the individual contributions from the members of the EAC. In most cases a contribution contains only one statement. However, in some cases a member of the EAC may touch on two (or more) separate issues in a contribution and in those instances the contribution is divided into separate statements. Our dependent variable is labelled ‘type of statement’ and is constructed to tap the type as well as the mode of opposition. However, in order to get the full picture of the extent to which EAC deliberations are characterised by oppositional behaviour, we need to also examine the presence of support for the government’s EU position. By combining the distinction between alternative, critique and support, on the one hand, and policy, polity and procedural opposition, on the other, we have constructed the variable ‘type of statement’ which is coded: (1) policy alternative; (2) policy critique; (3) policy support; (4) polity alternative; (5) polity critique; (6) polity support; (7) procedural alternative; (8) procedural critique; (9) procedural support. A final category (10) other, contains all neutral statements made by MPs, e.g. questions directed at the government representative (for more information on the coding process, see Online Appendix B).

Before turning to an examination of our data we should acknowledge that the choice to study oppositional behaviour behind the scenes, in the EAC, rather than looking at the more visible plenary debates may impact on our results. As made clear in our conceptual discussion, to engage in oppositional behaviour is partly a means for political parties to communicate with voters. Compared to plenary debates EAC deliberations are arguably less public, which means that MPs by and large have weaker incentives to engage in oppositional behaviour in this institutional setting.

Comparing patterns of oppositional behaviour – what does the data show?

It is now time to analyse the data in order to find out how much and what kind of opposition exists in EU affairs in our six countries. We will start by analysing aggregate data and then continue by breaking down the data by party.

How much opposition exists in EU affairs?

Let us first look at whether the data support the ‘depoliticization hypothesis’ or better align with the counter-hypothesis that there is a significant amount of opposition to be found in EU politics. Which proposition is best supported by the data?

As Figure 1 shows, our data does not support the ‘depoliticization hypothesis’. To the contrary, we find a considerable amount of opposition expressed during EAC deliberations as nearly 40 per cent (39.9 per cent) of all statements contain opposition whereas supporting statements only add up to 8.2 per cent in the six EACs. Neutral statements in the form of questions or general remarks (‘other’), in turn make up 51.9 per cent of all statements (see Online Appendix C, Table AC1). The observed differences between the categories are statistically significant (p = 0.00). That our results in fact run counter to the ‘depoliticization hypothesis’ is underscored by the fact MPs, as discussed above, have weaker incentives to express opposition in the EAC compared to the plenary.

Figure 1. Share of EAC statements in six countries (percent, 2005–2016).

Note: Total n is 8945 statements: (Austria 818, Denmark 2350, Germany 1195, Ireland 1157, Sweden 1912 and the UK 1513). See Online Appendix E for information on which differences in proportions between countries that are significant (p-values less than 0.05).

Display full size

How then does opposition vary between countries? As seen in Figure 1, the highest share of opposition statements is found in Sweden (44.7 per cent), followed by Denmark (43.6 per cent) and Austria (40.5 per cent). In contrast, the lowest share of opposition statements is found in the German EAC (32.1 per cent), i.e. a difference of more than twelve percentage points compared to Sweden. In the EACs in Ireland and the UK we find 35.0 per cent and 37.7 per cent opposition statements, respectively. When looking at supporting statements, we again find that the EACs in Austria (11.2 per cent), Denmark (10.0 per cent) and Sweden (9.6 per cent) make up the top three, while we find a lower share of supporting statements in the UK, Ireland and Germany (4.2 per cent, 6.4 per cent and 7.0 per cent, respectively). The data in Figure 1 thus reveal a clear pattern across countries: In Sweden, Denmark and Austria we find that a majority of the statements express either opposition to or support for the government. In Germany, Ireland and the UK, on the other hand, we find a majority of neutral statements in the form of questions and general remarks.

As we move on to examine variation in type of opposition we exclude the category ‘other’ and focus exclusively on the remaining statements. The data show that opposition statements containing critique (55.0 per cent) are roughly twice as many as those presenting alternatives (27.9 per cent, see Online Appendix C, Table AC2). The observed differences are statistically significant (p = 0.00). The EAC deliberations are thus primarily focused on controlling the government by criticising it, rather than providing the electorate with choice by presenting alternatives. As Figure 2 shows there are, however, significant differences between countries. In the Anglo-Saxon countries we find a huge difference between the share of statements delivering critique (73.5 per cent in the UK and 72.2 per cent in Ireland) and the share of statements presenting alternatives (16.4 per cent in the UK and 12.3 per cent in Ireland). In Germany and Denmark we also find that critique is the dominant type of opposition, but the differences between categories are much smaller. In contrast, we find in Sweden that the two types of opposition statements are almost equally common (41.6 per cent critique and 40.8 per cent alternatives), whereas presenting alternatives is actually the more common type of opposition in Austria (41.6 per cent against 36.6 per cent).

Figure 2. Share of EAC statements in six countries, by type (percent, 2005–2016).

Note: The category ‘Other’ has been excluded. Total n is 4301 statements (Austria 423, Denmark 1259, Germany 467, Ireland 479, Sweden 1038 and the UK 635). See Online Appendix E for information on which differences in proportions between countries that are significant (p-values less than 0.05).

Display full size

When it comes to the mode of opposition, our data show that policy opposition dominates over polity and procedural opposition. If we focus exclusively on statements containing opposition, we find that 76.9 per cent of these are directed at policies, whereas only 9.0 per cent are aimed at the EU polity and 14.1 per cent target the conducting of politics (see Online Appendix C, Table AC3). The observed differences between the categories are statistically significant (p = 0.00). This is hardly surprising since EAC deliberations are mostly focused on the government’s position in connection to ongoing policy negotiations. However, as we break down our data we find considerable variation between countries. As shown in Figure 3, policy opposition is most common in Austria (88.7 per cent) and Ireland (81.8 per cent), while it is least common in Germany (67.9 per cent). Polity opposition, on the other hand, is most frequent in the UK (12.4 per cent) and Denmark (11.2 per cent), while being virtually non-existent in Ireland (1.3 per cent). Procedural opposition, finally, is most frequent in Germany (21.2 per cent), and least common in Austria (4.0 per cent).

Figure 3. Share of EAC statements in six countries, by mode (percent, 2005–2016).

Note: The category ‘Other’ has been excluded. Total n is 4301 statements (Austria 423, Denmark 1259, Germany 467, Ireland 479, Sweden 1038 and the UK 635). See Online Appendix E for information on which differences in proportions between countries that are significant (p-values less than 0.05).

Display full size

In summary, our data on the aggregate level reveal patterns of oppositional behaviour that support the ‘politicization hypothesis’. We have also uncovered distinctive patterns of cross-national variation. First, when it comes to the mode of opposition, we find two clusters of countries. In Austria, Denmark and Sweden we find that a clear majority of statements express opposition or support, whereas neutral, information-seeking statements dominate deliberations in Germany, Ireland and the UK.

The second key difference concerns the type of opposition. In Ireland and the UK we find that opposition statements predominantly are expressed as critique rather than alternatives. Deliberations here conform to a Westminster style of interaction where the role played by the parties outside government is to ‘oppose everything, and propose nothing’ (Norton Citation2008: 238). In stark contrast to this, we find in Sweden roughly as much opposition in the form of presenting alternatives as expressing critique, and in Austria the latter category is in fact the largest one. We thus find three pairs of countries when it comes to type of opposition: In Ireland and the UK expressing critique is the dominant type of opposition whereas presenting alternatives is equally or more common than expressing critique in Austria and Sweden. Denmark and Germany fall in between these extremes. The country differences found may also be understood in terms of the familiar distinction between debating and working parliaments (Polsby Citation1975). Debating parliaments are characterised by a focus on plenary debates where the opposition criticises the government, whereas working parliaments are more consensual with a heavier emphasis on committee work. The patterns of oppositional behaviour that we find are by and large consistent with the fact that the House of Commons is the archetype of a debating parliament, whereas the Nordic countries are seen as the best examples of working parliaments (Arter Citation1999).

How does oppositional behaviour vary between political parties?

From previous research we derive the theoretical expectations that Eurosceptic parties will be the main drivers of opposition, whereas mainstream parties are assumed to keep a low profile in EU affairs. Is this the case? At the aggregate level we find that Eurosceptic parties account for little more than half of all opposition statements in our six countries (see Online Appendix C, Table AC4), which mean they definitely punch above their weight when it comes to expressing opposition, considering their seat share in Parliament (the observed differences between mainstream and Eurosceptic parties are statistically significant (p = 0.00)). As shown in Figure 4, our data confirm that Eurosceptic parties are the main drivers of opposition in Denmark and Sweden, and also play a significant role in Austria and the United Kingdom. In Germany and Ireland, however, the lion’s share of all opposition statements in the EACs emanate from mainstream parties.

Figure 4. Share of EAC statements in six countries, by party category (percent, 2005–2016).

Note: Total n is 8942 statements (Austria 818, Denmark 2350, Germany 1193, Ireland 1156, Sweden 1912 and the UK 1513). Black refers to 'opposition', light grey to 'support' and dark grey to 'other'. See Online Appendix E for information on which differences in proportions between mainstream and Eurosceptic parties that are significant (p-values less than 0.05).

Display full size

On closer inspection, we find that the three Eurosceptic parties in Sweden, the Green Party (MP), the Left Party (V) and the Sweden Democrats (SD) account for roughly 60 per cent of all opposition statements made during EAC deliberations. In Denmark, the Eurosceptic parties are likewise prominent drivers of opposition as they together account for more than 60 per cent of all opposition statements, and in Austria the Eurosceptic Freedom Party (FPÖ) alone account for slightly less than a third of all opposition statements made.

The data in Figure 4 lend partial support to the proposition that Eurosceptic parties are key drivers of opposition in EU politics as they account for a considerable amount of opposition statements in four of the six countries. The data, however, also show that mainstream parties are considerably more active when it comes to expressing opposition in EU affairs than what the cartel party-theory would have us believe. Even in Sweden, where Eurosceptic parties account for a higher share of opposition statements than in any other country, MPs from the Social Democrats (S) express more opposition than representatives from any other single party. It is also noteworthy that three pro-European mainstream parties – Fianna Fail (FF), Fine Gael (FG) and the Labour Party (Lab) – each express more opposition in the Irish EAC than the Eurosceptic Sinn Fein (SF) (see Online Appendix C, Table AC5). These observations underscore that the role played by Eurosceptic and mainstream parties when it comes to expressing opposition in EU politics is not as straightforward as our theoretical expectations suggested. On this point, it is important to note that the oppositional behaviour of mainstream parties of course depends heavily on whether they are governing or opposition parties (see Online Appendix D, Tables D1–D3).

The data also reveal the existence of distinct within-group variation among Eurosceptic parties. On closer inspection it becomes obvious that two Eurosceptic parties – the Sweden Democrats (SD) and the Austrian Freedom Party (FPÖ) – stand out from the crowd. Whereas the share of opposition statements made by most Eurosceptic parties is balanced by roughly an equal amount of supportive or neutral statements, this is not the case when it comes to SD and FPÖ (see Online Appendix C, Table AC5). Instead, for both these parties we find that an overwhelming majority of all statements fall in the category ‘opposition’, and that they make very few supporting statements. The data thus suggest a distinct difference in oppositional behaviour between Eurosceptic parties. In fact, it seems we are able to make sense of this difference by introducing the well-known distinction between ‘hard’ and ‘soft’ Euroscepticism. Whereas the former implies outright rejection of the entire integration project, and resistance to one’s country remaining in the EU, soft Euroscepticism is defined as involving ‘contingent or qualified opposition to European integration’ (Taggart and Szczerbiak Citation2004: 4). Now, for the examined time periods, according to the widely used Chapel Hill Expert Survey (CHES), both the Sweden Democrats (SD) and the Austrian Freedom Party (FPÖ) scored very close to the low end of the scale (note that the Sweden Democrats was represented in the Riksdag only during the second legislative term, see Online Appendix A, Table A2). With CHES scores of 1.8 (FPÖ) and 1.6 (SD) the two parties definitely qualify as hard Eurosceptics and stand out from the remaining Eurosceptics parties in our six countries which have an average CHES score of 2.6.

We now proceed to examine variation in type of opposition in order to uncover whether there is support for the hypothesis that Eurosceptic parties, compared to mainstream parties, focus more on delivering critique and less on presenting alternatives. On this dimension, two interesting patterns emerge from the data in Figure 5 below.

Figure 5. Share of EAC statements in six countries, by party category and type (percent, 2005–2016).

Note: The category ‘Other’ has been excluded. Total n is 4299 statements (Austria 423, Denmark 1259, Germany 465, Ireland 479, Sweden 1038, and the UK 635). Black refers to 'alternative', dark grey to 'critique' and light grey to 'support'. See Online Appendix E for information on which differences in proportions between mainstream and Eurosceptic parties that are significant (p-values less than 0.05).

Display full size

First, we do find that Eurosceptic parties account for a high share of the statements expressing critique and it seems that our hypothesis is supported by the data (the observed difference between mainstream and Eurosceptic parties in this regard is statistically significant (p = 0.00)). In the Swedish EAC we accordingly find that the three Eurosceptic parties – the Left Party (V), the Green Party (MP) and the Sweden Democrats (SD) – account for more than half of all statements containing critique, and in Denmark 61 per cent of all statements containing critique are made by Eurosceptic parties. Likewise we find that slightly less than a third of all statements containing critique in the Austrian EAC can be attributed to The Freedom Party (FPÖ), and a similar pattern is found in Germany where the Left Party (Die Linke) alone account for a third of the statements containing critique. In the UK, finally, we find that the Conservatives (CON) alone account for a vast majority of all statements containing critique in the EAC (see Online Appendix C, Table AC7).

The second major takeaway from Figure 5 is that Eurosceptic parties not only account for a high share of statements expressing critique, they also account for a substantial share of statements presenting alternatives – something not in line with our theoretical expectations. Also in this regard, the observed difference between mainstream and Eurosceptic parties is statistically significant (p = 0.00). The extent to which they engage in the two types of oppositional behaviour does, however, differ between countries. In Austria and Sweden we find that the Eurosceptic parties engage as much in presenting alternatives as in expressing critique. In Denmark, Germany and the UK, on the other hand, we find that the Eurosceptic parties are much more prone to expressing critique than presenting alternatives. It is obvious, then, that we do not only see a divide between hard and soft Eurosceptic parties when it comes to type of opposition, but rather a divide between countries. As noted above, patterns of oppositional behaviour in Austria and Sweden are more consistent with the practices of a working parliament and it seems as if even the hard Eurosceptic parties in these countries to a large extent have embraced this mode of interaction rather than be content with expressing critique. If we instead look at the two countries with typical debating parliaments, i.e. Ireland and the UK, we find a modus operandi where expressing critique clearly dominate over presenting alternatives when opposition is voiced.

Finally, what about the hypothesis that Eurosceptic parties, compared to mainstream parties, will engage more in polity opposition and less in policy opposition. The party differences regarding the mode of opposition reported in Figure 6 below reveal some distinct patterns of oppositional behaviour.

Figure 6. Share of EAC statements in six countries, by party category and mode (percent, 2005–2016).

Note: The category ‘Other’ has been excluded. Total n is 4299 statements (Austria 423, Denmark 1259, Germany 465, Ireland 479, Sweden 1038, and the UK 635). Black refers to 'polity', dark grey to 'policy' and light grey to 'procedure'. See Online Appendix E for information on which differences in proportions between mainstream and Eurosceptic parties that are significant (p-values less than 0.05).

Display full size

First, and most obviously our data show that policy opposition dominates over expressions of disagreements directed at the EU polity or the conducting of politics. Without exception, all parties in our six countries, whether they be mainstream or Eurosceptics, focus most of their time and energy on policy issues, i.e. on what previously has been referred to as ‘classical opposition’ (Kirchheimer Citation1957). When it comes to Eurosceptic parties the results at first sight seem to fly in the face of our theoretical expectations. The results, however, need to be interpreted against the backdrop of EAC deliberations being dominated by discussions over ongoing policy negotiations, rather than issues of a polity nature. It is therefore rather natural that policy opposition dominate over polity opposition, even for Eurosceptic parties (the observed difference between mainstream and Eurosceptic parties in this regard is not statistically significant (p > 0.05)).

The second pattern to emerge from Figure 6 is that Eurosceptic parties, while clearly focusing most of their attention on policy opposition, deliver considerably more polity opposition than mainstream parties (this difference is statistically significant (p = 0.00)). In this sense our data do lend support to our theoretical expectations. In Austria, the Eurosceptic Freedom Party (FPÖ) accounts for roughly as much polity opposition as the other parties combined. In Sweden we find that the three Eurosceptic parties account for a vast majority of all polity opposition, and that the Sweden Democrats (SD) is especially active in expressing opposition to the EU political system. Whereas only 7.5 per cent of all opposition statements in Sweden were directed at polity, almost a third (31.1 per cent) of all opposition statements made by SD representatives was directed at the EU political system. In a similar vein, the Danish People’s Party (DF) and the Red-Green Alliance (EL) account for a majority of all polity opposition in the Danish EAC (see Online Appendix C, Table AC9). In Germany and the UK, however, we do not find a similar dominance from Eurosceptic parties when it comes to the polity opposition expressed in their respective EACs.

When it comes to the dominance of Eurosceptic parties in expressing polity opposition we see clear signs of the dividing line between soft and hard Eurosceptic parties reappearing. If we focus on the share of statements directed at polity compared to policy, three hard Eurosceptic parties – the Freedom Party (FPÖ), the Sweden Democrats (SD) and the Danish People’s Party (DF) – stand out as especially active when it comes to voicing opposition to the EU political system. In contrast, parties closer to the soft end of the Eurosceptic spectrum – like the Left Party (Die Linke) in Germany, the Green Party in Sweden or Sinn Fein (SF) in Ireland – express polity opposition much less frequently (see Online Appendix C, Table AC9).

A third takeaway regarding the mode of opposition is that mainstream parties are relatively speaking much more active when it comes to procedural opposition than what is the case in connection to policy and polity opposition (the difference is statistically significant (p = 0.00)). In Austria, for example, we find that the Social Democratic Party (SPÖ) and the Green Party (die Grüne) deliver approximately as much procedural opposition as the Eurosceptic Freedom Party (FPÖ). In the Danish EAC we see that the Liberal Party (Venstre) and the Social Democrats (S) deliver more procedural opposition than the Danish People’s Party (DF) and almost as much as the Red-Green Alliance (EL). In Sweden, the Social Democrats (S) in fact deliver more procedural opposition than the Eurosceptic parties combined, and in Germany the Social Democratic Party (SPD) and the Green Party (die Grüne) deliver procedural opposition on par with the Left Party (Die Linke) (see Online Appendix C, Table AC9). In other words: when it comes to procedural opposition the parties act very differently than what is the case in connection to policy and polity opposition. The pattern of oppositional behaviour associated with procedural opposition is in fact consistent with the cartel party-theory (Katz and Mair Citation1995), according to which mainstream parties are assumed to spend proportionately more time discussing the way politics should be conducted rather than focusing on the content. The basic assumption is that political parties that are serious contenders for forming the next government have a common interest in not having ‘their hands tied by the constraints of popular democracy’ (Mair Citation2007: 7–8).

Conclusion

By drawing on a unique data set charting EAC deliberations in six countries, this study has thrown new light on opposition in EU politics. The examination has focused on two analytical dimensions: the type and mode of oppositional behaviour. In the course of our analysis, we have tested a number of theoretical expectations regarding how much and what kind of opposition exists in EU politics. We have uncovered variation in oppositional behaviour between Eurosceptic and mainstream parties as well as between countries. What conclusions, then, can be drawn from our examination?

First, the results shed new empirical light on the often repeated but rarely investigated claim that there exists an opposition deficit in EU politics (Mair Citation2007, Citation2013). In fact, nearly 40 per cent (39.9 per cent) of all statements made during EAC deliberations express opposition. The low visibility of EAC deliberations suggests we cannot dismiss the observed opposition behaviour as theatrics or grandstanding and the results rather lend support to the ‘politicization hypothesis’. We should keep in mind that our results build on data from only six countries during a turbulent time in European integration history, which warrants caution when interpreting the findings. But the fact that these cases were strategically chosen should go some way to increase the generalizability of our results. The mere existence of opposition in the EACs does not, however, mean that we can lay to rest the concerns over an opposition deficit. For if the opposition expressed in the EAC, stay in the EAC it will do little in terms of providing citizens with political alternatives and electoral choice.

A second conclusion to be drawn from our results is that there exists considerable cross-national variation when it comes to opposition in EU politics. When it comes to the mode of opposition we find in Sweden, Denmark and Austria that a majority of statements in the EACs express either opposition to or support for the government. In Germany, Ireland and the UK, on the other hand, we find that a majority of statements are delivered in the form of open questions or neutral remarks. The data also reveal considerable differences between countries when it comes to the type of opposition. In Ireland and the UK we find that opposition statements expressing critique outnumber those presenting alternatives by a wide margin. In sharp contrast we find that the two types of opposition statements are roughly equally common in Austria and Sweden. In the EACs in Denmark and Germany we find patterns of opposition that fall in between these extremes.

Third, we find at the party level support for the proposition that Eurosceptic parties are indeed key drivers of opposition in EU politics. The data, however, simultaneously show that mainstream parties express more opposition than what is to be expected according to the cartel party-theory. The picture is furthermore complicated by the fact that existing patterns of oppositional behaviour reveal a distinct within-group variation among Eurosceptic parties. Whereas most Eurosceptic parties tend to balance their share of opposition statements by roughly an equal amount of supportive or neutral statements, this is not the case for two of the hard Eurosceptic parties: the Sweden Democrats (SD) and the Austrian Freedom Party (FPÖ). We find instead that an overwhelming majority of all statements made in the EACs by these parties contain opposition in some form or the other.

Fourth, when it comes to type and mode of oppositional behaviour at the party level, our data partly support our theoretical expectations. As for the mode of opposition we find that even Eurosceptic parties focus more on policy than polity issues, but at the same time they are more prone than mainstream parties to express opposition directed at the EU political system. It is noteworthy that the difference in oppositional behaviour between soft and hard Eurosceptic parties is manifest also when it comes to the mode of opposition. We thus find that parties closer to the hard end of the Eurosceptic spectrum express much more polity opposition than soft Eurosceptic parties. When it comes to type of opposition our data by and large support the hypothesis that Eurosceptic parties, compared to mainstream parties, engage more in expressing critique than presenting alternatives. However, we also find evidence to support that there is national variation when it comes to the behaviour of Eurosceptic parties. In Austria and Sweden we find that all Eurosceptic parties, including hard Eurosceptics, engage as much in presenting alternatives as in expressing critique. It would thus seem that even the Eurosceptic parties in these member states have embraced the prevailing mode of interaction in these countries’ EACs.

Peter Mair once argued that the third milestone in the development of democratic institutions was missing in the EU in the absence of an arena in which to organise opposition (Mair Citation2007: 7). This study has shown that such an arena does in fact exist as we find much more opposition being voiced in the EACs than what the proponents of the ‘depoliticization hypothesis’ would have us believe. This fact by itself, however, does not justify the conclusion that opposition today plays a pivotal role in EU politics and that the concerns raised by Mair are no longer valid. Opposition is such a vital component of democracy because it is an instrument for providing voters with choice. However, opposition can only play this crucial role if alternative proposals and agendas reach the public sphere and are clearly communicated to voters. The challenge that lies ahead, then, is making sure that the opposition that do exist in the EACs break out of the parliamentary arena. Only then can we confidently claim that there is no longer an opposition deficit in EU politics.