1
Challenging the Executive in EU affairs?
An explanation of cross-party variance in parliamentary scrutiny behavior
Menno Hoppen 10 February, 2020 MSc Public Administration
COMPASS
Nijmegen School of Management Word count: 11.497
2 Abstract
This paper seeks to explain cross-party variance in levels of parliamentary scrutiny in EU affairs. The paper builds on the relatively novel approach of studying how intensively parties scrutinize the government – rather than only if they do – by appreciating levels of scrutiny interventions. The paper takes a most-likely case approach, by analyzing plenary debates in the Netherlands around EU Summits during the Refugee Crisis. Parties’ scrutiny behavior depends on their willingness and their capacity to scrutinize. Especially opposition parties are willing to intensively scrutinize the government in EU affairs. National salience correlates with more scrutiny and can be related to higher levels of scrutiny, whereas there are indications that Eurosceptic parties and parties with vast substantial differences in preferences with the government use lower levels of scrutiny. Larger, institutionalized parties tend to have more capacity to scrutinize the government. Parties in government have information-advantages, but use these in backrooms, rather than in parliament. All in all, the analysis shows that – at least in debates concerning a salient topic like Refugee Crisis – the main political parties are both willing and able to scrutinize the government intensively.
3 Introduction
Critics of the European Union (EU) often point at its alleged ‘democratic deficit’. Rather than a deficit, however, this is better described as a ‘democratic disconnect’ (Lindseth, 2012, p. 3, Kröger and Bellamy, 2016, p. 133). Whereas policy is increasingly affected by decisions made in Brussels, national capitals are still at the center of politics and receive the most attention from media and citizens. National politics increasingly have the appearance of ‘politics without policy’, while European policymaking appears as ‘policy without politics’ (Schmidt, 2013, p. 33).
Consequently, national parliaments may have the potential to overcome this disconnect (Auel, Eisele, and Kinski, 2018, p. 628; Kröger and Bellamy, 2016, p. 133). However, these are precisely the institutions that scholars have labeled the ‘losers’ of the European integration process: whereas national executives gained powers, parliaments mainly remained sidelined (Maurer and Wessels, 2001; Follesdal and Hix, 2006, p. 535). Scholars have summarized this diminished role of national parliaments in comparison to the executives in a diagnosis of ‘de-parliamentarization’ (Maurer and Wessels, 2001; Schmidt, 2013, p. 63; Mastenbroek, Versluis and Spenzharova, 2014, p. 750; Winzen, 2013).
However, the tides may have turned for national parliaments since the Lisbon Treaty entered into force in 2009 (Auel and Christiansen, 2015, pp. 266-7). For the first time, national parliaments have an explicitly stated role in an EU treaty. The treaty has enhanced the information position of national parliaments. These now have access to more documents and are not dependent on their national executives anymore to provide these (Auel and Christiansen, 2015, pp. 266-7). Theoretically, this could give parliaments the resources to tackle their problem of de-parliamentarization.
Besides the institutional conditions for Members of Parliament (MPs) to engage actively in the European policymaking process, motivational conditions have changed as well. Whereas the initial attitude of the public vis-à-vis European integration reflected a permissive consensus in which public opinion was submissive, the recent decades are characterized by constraining dissensus in which the EU has become politicized in referendums and national elections (Hooghe and Marks, 2009). As a result, political elites ‘must look over their shoulders when negotiating European issues’ (Hooghe and Marks, 2009, p. 5). Due to this transformation of public opinion about the EU, it has become increasingly interesting for MPs to assert themselves on the topic of European integration (Raunio, 2017; Auel and Christiansen, 2015).
Yet, despite this evolved environment in which European integration is becoming more important and politicized, Rauh and De Wilde (2018) argue that we face an ‘opposition deficit’ in EU accountability from MPs (p. 210). A necessary condition for multi-level accountability, they argue, would be increased emphasis on issues relating to the EU in the national plenary debate (p. 197). Although MPs raise EU issues more in the plenary as more powers have been transferred to the European level, increasing public Euroscepticism is negatively related to EU salience in the plenary (p. 210). This correlation is counterintuitive: a more Eurosceptic public would expectedly incentivize (particularly opposition) MPs to establish themselves as a more Eurosceptic alternative to the government – thereby
4
raising EU issues more. Despite this expectation, government parties structurally outperform opposition parties in raising EU issues (Rauh and De Wilde, 2018). Thus, especially opposition parties fail to ‘give account’ (Bovens, 2007) to their constituents on EU issues. As a result, Rauh and De Wilde (2018) signal an opposition deficit in EU accountability (p. 194).
This observation is puzzling: institutional and societal changes have created an environment that enables and incentivizes MPs to actively scrutinize their government on EU issues, yet they do not seem to fully capitalize on this. However, Rauh and De Wilde (2018) merely focus on how many times MPs raise EU issues in the plenary – not the actual content of these contributions. It may well be that MPs do not raise EU issues more often, but they do raise them more critically. In order to establish whether there is an opposition deficit, we would need to know not only if but also how EU issues are discussed in parliament.
Moreover, the focus of Rauh and De Wilde (2018) on general salience of EU topics in the plenary begs the question of whether it is even desirable for MPs to engage in all EU affairs actively. As Winzen (2013) states: ‘What would it really tell us if we were to find that parliaments examine each and every of the EU’s legislative proposals in detail, instructing the government what goals to pursue in the negotiations?’(p. 153). The question should be ‘under what conditions and through what mechanisms parliaments decide to scrutinize’ (Winzen, 2013, p. 153). Although MPs of opposition parties may not be as vocal about EU issues in the plenary as one might expect in general, they may pick their battles and conduct ‘targeted scrutiny’ (Raunio, 2017, pp. 333-4). MPs have limited resources, which they may reserve for the cases with the highest stakes.
In order to gain more insight into the role national parliaments play in the EU, we would have to know when and how intensively MPs decide to scrutinize their governments in EU affairs. Therefore, the goal of this paper is to contribute to the academic debate about parliamentary scrutiny in EU affairs by investigating the conditions under which political parties choose to scrutinize their governments and how intensively they do so. This paper sees parliamentary scrutiny as ‘the exercise of power by the legislative branch to control, influence or monitor government decision-making’ (Holzhacker, 2007, p. 182).
This study takes a crucial case approach in order to attain this goal (Gerring, 2007, p. 115). This study analyzes the scrutiny behavior of political parties in the Dutch Tweede Kamer concerning the EU’s handling of the Refugee Crisis. This is a most-likely case study design because this is the most likely case for active parliamentary scrutiny to occur – except for the fact that it focuses on EU affairs. The selected country is the Netherlands because it has a comparatively strong parliament, which is a necessary condition for parliamentary scrutiny to occur (Raunio, 2005, p. 335; Auel and Christiansen, 2017). Besides the institutional conditions, this case also meets the conditions concerning incentives: the Dutch electorate is volatile, the parliamentary electoral threshold is low, and the EU has become increasingly politicized, thereby making it electorally attractive for political parties to assert themselves
5
in this area (Mastenbroek et al., 2014, p. 754). In sum, the study expects Dutch political parties to be both able and willing to actively scrutinize their government in parliament.
This paper solely focuses on scrutiny in plenary debates, since this is the forum where MPs hold their executives to account and simultaneously give account to their constituents in public (Bovens, Schillemans, and ‘t Hart, 2008, p. 231). The study only examines plenary debates about EU Summits, because these are the venues where the most important decisions are made at times of the highest salience (De Wilde and Raunio, 2016, p. 319, 323). Overall, plenary debates about EU Summits are the platform where MPs have the best chance to connect with their voters and fulfill their communication function about EU affairs, which has received comparatively little attention in previous research (Auel and Raunio, 2014).
The EU’s handling of the Refugee Crisis in 2015 and 2016 is at the center of the analyzed debates. In many West-European countries, immigration and integration have increasingly defined the political debate (Van Heerden et al., 2014, p. 120). Given the importance that both political parties and the electorate attach to these topics, this study expects that political parties are especially active in scrutinizing their government when it comes to the handling of the Refugee Crisis by the EU, since salience is an important precondition for parliamentary scrutiny to occur (Finke and Dannwolf, 2015; Hutten, 2016; Mastenbroek et al., 2014; Auel and Christiansen, 2015). Therefore, we may expect that – even though EU issues may not always be a top priority for MPs – issues concerning immigration will receive parties’ attention.
Altogether, this constitutes a most-likely case: if we would not find active parliamentary scrutiny during these plenary debates, chances are that it is nowhere to be found (Gerring, 2007, p. 115). At the same time, this means that this is the case where it is the least likely to find an opposition deficit. All in all, this results in the following research question: how can we explain cross-party variance in the level of parliamentary scrutiny concerning the EU’s handling of the Refugee Crisis in the Tweede Kamer?
In order to answer this question, this paper takes a case study-mixed methods approach (Guetterman & Fetters, 2018, p. 906) – similar to the study of Smeets and De Ruiter (2018). The study makes use of the sequential explanatory design, which is a version of mixed methods that consists of two phases (Creswell and Plano Clark, 2011). First, the study collects and analyzes the quantitative data about the scrutiny behavior of political parties in plenary debates around EU Summits during the Refugee Crisis. Second, the qualitative data are collected and analyzed in order to explain, elaborate, or even rebut the initial findings. The sequential explanatory design is especially suited for this study because it wants to assess relationships with quantitative data, as well as to explain the mechanisms behind these relationships (Creswell and Plano Clark, 2011, p. 82).
Specifically, this study assesses the levels of scrutiny by using a ‘ladder of scrutiny’, based on the research of Smeets and De Ruiter (2018) and, originally, Mastenbroek et al. (2014). The scrutiny ladder allows this study to capture insights about parliamentary scrutiny in more detail than previous research. This paper analyzes the cross-party variance in levels of parliamentary scrutiny with a
6
quantitative analysis. In order to validate and explain these initial findings, several elite-interviews with (former) Dutch MPs, their staff-members, and (former) staff members of the Tweede Kamer have been conducted.
Relevance
The role of national parliaments in European integration has been studied widely in Europeanization literature (Goetz and Meyer-Sahling, 2008). In general, scholars have argued that national parliaments have lost powers due to European integration (Raunio, 2011, p. 304; Goetz and Meyer-Sahling, 2008, p. 6; Follesdal and Hix, 2006, p. 535). The de-parliamentarization thesis may have become even more urgent in the last decade: the EU’s handling of the Eurozone crisis has been labeled as ‘emergency politics’, where executives increasingly make decisions in ad hoc decision-making venues – leaving national parliaments merely with a post hoc role to play (White, 2015, p. 306).
On the other hand, researchers have investigated whether and how MPs try to reassert themselves (Auel, 2007; Mastenbroek et al., 2014; De Ruiter, 2013). De-parliamentarization is not a fait accompli: national parliaments may also reinvent themselves in reaction to European integration (Winzen, 2013, p. 17). Some scholars have even described this process of adaptation as ‘re-parliamentarization’, which entails how national parliaments have attained instruments and regained powers to exert influence in European affairs (Kreilinger, 2019, p. 19).
By analyzing the scrutiny behavior of parties in national parliaments, this paper seeks to contribute to the scientific debate about the role of national parliaments in European integration, cross-party differences in parliamentary scrutiny, and the power dynamics between the parliament and executive in a European context. Although there already is a large body of literature about the institutional possibilities for parliaments to scrutinize their governments (e.g. Karlas, 2012), we know considerably less about the actual use of these possibilities (Auel et al., 2015; Winzen, 2013, p. 153). Furthermore, we know little about the content of parliamentary interventions in EU affairs. We may count the number of debates (e.g. Hefftler, Kreilinger, Rozenberg and Wessels, 2013), interventions, motions, or the motions that are adopted – but if we really want to gain insight in the (dis)balance of power between parliaments and executives in the EU, we need a more fine-grained approach to define levels of scrutiny interventions (Mastenbroek et al., 2014; Smeets and De Ruiter, 2018).
Given the increased dominance of the executives EU affairs, it is vital for students of public administration to paint a clearer picture of the political environment in which these operate. As Johansson and Tallberg (2010) state: ‘European integration serves to redistribute power among domestic political actors, rather than undermine state authority’ (p. 232). However, the dynamics of this have remained overlooked (Johansson and Tallberg, 2010, p. 232). While executives seem dominant given the shift of power to the Council, they may increasingly receive instructions from their national parliaments. Executive power means little if MPs send their governments to Brussels with clear goals to achieve and red lines that should not be crossed – possibly under threat of a no-confidence vote if
7
they fail to deliver (Van Middelaar, 2017, pp. 269-70). In a dynamic in which national parliaments increasingly try to wield power on European issues, it may be hard for executives to compromise. In extreme cases, this could even cause gridlock in the European machine of policymaking. Therefore, students of public administration, too, need more insights as to how parties in parliament behave and what this means for inter-institutional power dynamics in EU affairs.
From a democratic perspective, the role of parliaments in the EU is worth studying as well. Lindseth (2012) states that the EU has come a far way in creating government for the people (output legitimacy) and government by the people (input legitimacy) (Scharpf, 1999). However, people do not yet have the ‘perception that the institutions of government are genuinely the people’s own’ (Lindseth, 2012, p. 2). In other words, there is no European demos: the EU may have been created, yet remains insufficiently ‘imagined’ as a political community (Anderson, 1983) by the majority of its citizens to actually qualify as government of the people. Precisely because of this, national parliaments can serve a vital role in overcoming the democratic disconnect (Kröger and Bellamy, 2016; Lindseth, 2012). After all, national capitals are still the place where politics visibly take place for the public (Schmidt, 2013). If European citizens are to have a better connection with the EU as a polity, national parliaments have an essential role to play in discussing EU issues in order to provide a sense of political ownership for their citizens (Lindseth, 2012). In order to do so, MPs need to hold their executive to account in public, so that the electorate can make an informed choice on the basis of EU issues. If national parliaments do not sufficiently fulfill these functions, ‘national parliaments will contribute little to the democratic legitimacy of EU policy-making’ (Auel and Cristiansen, 2015, p. 263). Therefore, in this regard, it is especially interesting how MPs behave in plenary debates.
The next section of this paper identifies the variables that may influence parliamentary scrutiny. The subsequent section lays out the methods have been used to test these hypotheses. The fourth section provides a case description of the EU’s handling of the Refugee Crisis and the Dutch position in this regard, in order to provide some context to the findings. The fifth section contains the quantitative and qualitative analyses. Finally, the main findings and their implications are summarized in the conclusion.
8
Explaining cross-party variance in parliamentary scrutiny
Scholars have identified several explanations of cross-party variance in the level of parliamentary scrutiny. The variables that are theorized to explain differences in cross-party scrutiny can be divided into two categories: incentives and resources (Auel and Christiansen, 2017). Incentive-variables are factors that increase or decrease the willingness of political parties to actively scrutinize the government, whereas resource-variables influence the extent to which parties are actually able to follow through on this wish.
The first incentive-variable that may explain cross-party differences in parliamentary scrutiny is party Euroscepticism. The rationale behind the expectation that party Euroscepticism increases the likelihood of active scrutiny is that the more skeptical a party is about European integration, the more skeptical it should be about its government’s behavior in European affairs. From both a policy-seeking perspective and a vote-seeking perspective, in times of ‘constraining dissensus’ when voters increasingly care about European affairs, it makes sense for Eurosceptic parties to engage more actively in parliamentary scrutiny in EU affairs (Marks and Hooghe, 2009). In contrast, as Auel and Christiansen (2015) note, ‘a party favorable to European integration may view parliamentary scrutiny as a factor impeding smooth European policy-making and thus as inappropriate’ (p. 271; Sprunk, 2003). Thus, this study expects that the more Eurosceptic a party is, the more actively it scrutinizes the government’s behavior in EU affairs. However, most studies have not found that party Euroscepticism influences the level of parliamentary scrutiny (Hutten, 2016; Raunio, 2005; Smeets and De Ruiter, 2018, pp. 576-577).
Nonetheless, Smeets and De Ruiter (2018, p. 576) put forward that, while Eurosceptic parties may be willing to engage in scrutiny more actively, they are not always able to do so due to information asymmetries. Smeets and De Ruiter (2018), however, studied the Banking Union Debate, which was highly technical. Because the debate on the Refugee Crisis was a lot easier to understand – for insiders and outsiders – this study expects that the more Eurosceptic a party is, the more it is inclined to actively scrutinize its government’s behavior in EU affairs. This leads to the first hypothesis:
H1: The more Eurosceptic a party, the more likely it is that it will actively scrutinize the government on EU issues
Second, the distance between a policy-issue and a party’s policy preferences may be positively related to the level of scrutiny (Mastenbroek et al., 2014, p. 753; Smeets and De Ruiter, p. 578; Auel and Christiansen, p. 270). When a proposal is entirely in line with a party’s preferences, it makes little sense to scrutinize that proposal actively. Conversely, if there is a considerable distance between a proposal and a party’s policy preferences, parties should be incentivized to draw contrast between these by scrutinizing the proposal. This expectation is in line with the conception that EU politics increasingly have the character of ‘normal’ politics (Smeets and De Ruiter, 2018, p. 569; Hooghe and Marks, 2009, p. 9; Holzhacker, 2005, p. 44). Therefore, the second hypothesis states:
9
H2: The greater the distance between a party’s policy preferences and the substance of a policy proposal, the more likely it is that a party will actively scrutinize the government on this proposal
Third, salience at the national level is an important explanatory variable for parliamentary scrutiny (Hutten, 2016; Mastenbroek et al., 2014; Auel and Raunio, 2014; Smeets and De Ruiter, 2018; Dörrenbächer et al., 2015). Issue salience refers to the importance that actors attach to issues (Warntjen, 2012, p. 169). The relationship between issue salience at the national level makes sense from both a vote-seeking and a policy-seeking perspective. When an issue is nationally salient, by definition, the electorate attaches importance to this issue and, therefore, there may be votes to be won for politicians on this issue. In a well-functioning democracy, political parties should want to influence policy about issues that the public deems salient. Therefore, the third hypothesis expects:
H3: The more nationally salient an issue is at a given moment, the more likely it is that parties will actively scrutinize the government on this issue
Fourth, whether a party is in government or the opposition may influence the degree to which its MPs engage in parliamentary scrutiny in EU affairs. Generally, MPs from opposition parties are more active scrutinizers than MPs from governing parties (Hutten, 2016; Mastenbroek et al., 2014). Even though MPs from political parties in government have more incentives to engage in scrutiny in EU affairs than in domestic affairs (since EU affairs generally are not part of coalition agreements) they may not want to do this publicly (Auel, 2007, p. 492; Holzhacker, 2005, p. 39). Empirically, governing parties are to be less active ex-post scrutinizers than opposition parties (Hutten, 2016; Mastenbroek et al., 2014). Smeets and De Ruiter (2018) find mixed results when analyzing ex-ante parliamentary scrutiny: although opposition parties tend to be more active scrutinizers in plenary debates, there is no significant difference between opposition and coalition parties when looking at all interventions, such as written meetings, committee meetings and plenary debates (p. 577). This finding lends leverage to the idea that scrutinizing endeavors in plenary debates are more informed by vote-seeking incentives than committee meetings (Holzhacker, 2005, p. 41). This leads to the fourth hypothesis:
H4: Opposition parties are more likely to actively engage in parliamentary scrutiny of the government than government parties
Besides incentives, resources such as information, expertise, and time may play a role as well. As Smeets and De Ruiter (2018) notice, opposition parties may find it harder to keep up with the relevant information than MPs from parties in government did. Information asymmetries may play a role here:
10
although MPs from outsider-parties may have the incentives to actively engage in parliamentary scrutiny, they may find it harder to reach higher levels of the scrutiny ladder since they mainly use their interventions to keep track of the developments in Brussels. However, in plenary debates, it is expected that this information-element of being either a government or opposition party weighs less heavily than the vote-seeking incentive.
As a fifth variable, party size is suggested to influence the level of scrutiny (Hutten, 2016; Dörrenbächer et al., 2015, p. 1023). Political parties with fewer seats have less capacity than parties with more seats in parliament. Moreover, larger parties tend to have more resources to collect information and keep an overview of what is going on in Brussels (Smeets and De Ruiter, 2018, p. 576). This leads to the final hypothesis:
H5: The more seats a party has in parliament, the more likely it is that it actively engages in parliamentary scrutiny
11 Methodology
This study employs both qualitative and quantitative methods to test the hypotheses and answer the research question. Thus, the method of this paper has the form of a crucial case study-mixed methods design (Gerring, 2007, p. 115; Guetterman and Fetters, 2018, p. 906). Specifically, this study makes use of the sequential explanatory design, which is a version of mixed-methods that consists of two phases (Creswell & Plano Clark, 2011). The first phase entailed the collection and analysis of the quantitative data. In the second phase, qualitative data were collected and analyzed in order to explain, elaborate, or rebut the initial findings. This design is well-suited for this study because it wants to assess estimations of relationships with quantitative data, as well as to explain the mechanisms behind these relationships (Creswell & Plano Clark, 2011, p. 82). If the study relied solely on either quantitative or qualitative methods, it would tell only a part of the story.
The data collection phase consisted of selecting the relevant plenary debates in the Tweede Kamer around EU Summits about the Refugee Crisis, which resulted in the selection of seventeen plenary debates in total (Appendix I). In the subsequent part of the data collection phase, these debates were analyzed. In this stage, the interventions in the debates by MPs on behalf of their political parties were scored on a ladder of scrutiny, which consists of four levels of scrutiny (Mastenbroek et al., 2014 Smeets and De Ruiter, 2018) (Table 1). In total, the parties made 609 interventions in seventeen debates.
Subsequently, a regression analysis was conducted. Given the ordinal nature of the dependent variable, an ordinal logistic regression analysis would generally be the preferred way to assess the relationships between the variables (Osborne, 2015). This analysis, however, assumes that the effects of the independent variables are consistent for each order of the dependent variable (Osborne, 2015, p. 21). In this study, this proportional odds assumption is violated for some variables. In order to overcome this, the study makes use of an ordinal regression analysis with partial proportional odds (PPO model) (Williams 2006; Williams, 2016). The PPO model relaxes the proportional odds assumption for the variables that violate it but maintains it for variables that meet the assumption. Therefore, this model fits the data better and is more parsimonious than other alternatives (Williams, 2016, p. 13).
Another assumption for logistic regression requires attention as well. The observations may not be entirely independent of each other because interventions are clustered in debates. Therefore, it cannot be assumed that the errors are independent of each other. Several other analyses (ordinal regression, OLS, and multi-level ordinal regression) have been conducted as robustness checks and supported the original findings. The multi-level model did not provide a better fit for the data. Still, some caution is advisable in the interpretation of the quantitative results of this study.
The following phase of the study consisted of the qualitative data collection (Creswell & Plano Clark, 2011, p. 84). The main sources of the qualitative data are semi-structured elite-interviews. Semi-structured interviews are a well-suited method for this paper. First, these are especially useful to gain insight into the motivations and preferences of political parties and to examine the role of agency in political processes (Rathbun, 2008). Second, semi-structured interviews allow the researcher to ask
12
follow-up questions in order to enhance the internal validity of the study. Third, these interviews allowed room to gain inductive insights. In total, the qualitative analysis is based on seven interviews (Appendix II)
Operationalization
The dependent variable of this paper is the level of parliamentary scrutiny by political parties. Each intervention that a party makes in a debate is scored on a ‘ladder of scrutiny’. Mastenbroek et al. (2014, p. 756) created this scrutiny ladder, and Smeets and De Ruiter further developed it (2018). The original scrutiny ladder of Mastenbroek et al. (2014) comprised five steps: (1) expressing support, (2) gathering information, (3) signaling a position, (4) expressing disagreement, and (5) exerting influence. As Smeets and De Ruiter (2018) note, this ladder was implicitly biased in favor of opposition parties since these are inclined to disagree with the government. However, such a scrutinizing attempt does not necessitate more effort and information on the part of the MP, and neither does it necessarily challenge the government more. Therefore, in this study, expressing support is rated as highly as expressing disagreement. As interventions get more demanding for MPs, they will be rated higher on the scrutiny ladder: (1) expressing support or disagreement; (2) asking questions; (3) taking up alternative positions; (4) providing (specific) instructions. By not just examining whether political parties participate in plenary debates but also how they do this, the scrutiny ladder can give more detailed insights into cross-party differences in parliamentary scrutiny.
Table 1: Scrutiny ladder
Level of scrutiny 1.Expressing support/disagreement 2. Asking questions 3. Providing an alternative position 4. Providing instructions
(Mastenbroek et al., 2014, p. 756; Smeets and De Ruiter, 2018, p. 568)
An intervention is defined as a part of an MPs’ speech in which the MP tries to make a point on a specific issue (Smeets and De Ruiter, 2018, Web Appendix). Each of these interventions receives a score, usually resulting in multiple scores per political party per debate. When a party repeats the same point in one debate, only the intervention with the highest score will be taken into account. For instance, if a party first states it disagrees with mandatory reallocation scheme of the European Commission (level 1), then asks the government official why they support the scheme (level 2), and ultimately submits a motion that instructs the government to advocate a slightly altered relocation scheme in which
13
governments can buy off the obligation (level 4), the intervention receives the score ‘4’. When multiple parties address the same concerns, however, all of these parties receive a score.
Independent variables
The first incentive-variable is party Euroscepticism, which entails the extent to which political parties are opposed to European integration. A common operationalization for this variable is the Chapel Hill Expert Survey (CHES) (Polk et al., 2018). The variable is measured on a 7-point scale, where 1 indicates that a party is ‘strongly opposed’ and 7 indicates that a party is ‘strongly in favor’ of European integration (Polk et al., 2018, p. 9). This study takes the average of the parties’ level of Euroscepticism of the 2014 and the 2017 CHES because the analyzed period took place between these two expert surveys, and, for some parties, there are considerable differences between the two CHES scores. By taking the average, this study aims to increase the internal validity of this variable’s measurement.
The second incentive-variable concerns the distance between a party’s policy preferences and the substance of a policy proposal. A party’s preferences about immigration policy are taken as an indicator for a party’s policy preferences. The data are derived from the 2014 and 2017 CHES as well (Polk et al., 2018, p. 11). The possible value of this variable ranges from 1 (an open attitude to immigration) to 10 (a very restrictive attitude). The average score of the governing parties (PvdA and VVD) is the indicator of the substance of a policy proposal. This score is 6.05. Therefore, the distance between a party’s policy preferences and the substance of a policy proposal is the distance between a party’s mean score on the 2014 and 2017 CHES on immigration policy on the one hand, and the mean level of the governing parties on the other hand (Appendix III)
The third incentive-variable is national salience and is measured at the level of each debate. A common indicator of national salience is the amount of media coverage (Warntjen, 2012). Salience at the national level is therefore operationalized as the number of articles about the Refugee Crisis in the week before a debate in the following papers: Telegraaf, Algemeen Dagblad, Volkskrant, NRC Handelsblad, Financieel Dagblad, and Trouw. The following keywords are used on LexisNexis in order to establish the number of articles: vluchtelingencrisis, migratiecrisis, vluchtelingentop, migratietop, Turkijedeal, migratiedeal. Appendix I presents the salience of each debate.
The fourth incentive-variable is whether a political party was part of the government or the opposition. This variable is measured dichotomously. During the period of analysis, the People’s Party of Freedom and Democracy (VVD) and the Labour Party (PvdA) formed the government; the other parties were in the opposition.
The capacity-variable is party size. This variable is operationalized as the number of seats of a political party in the Tweede Kamer. This variable ranges from 1 to 41.
Finally, ex-post debates may differ from ex-ante debates. Therefore, this is included in the analysis as a control variable (Appendix I).
14 Validity and reliability
Internal validity concerns the quality of the data, results, and interpretations that follow from these (Creswell and Plano Clark, 2011, p. 210; Yin, 2011, p. 79). This addresses the question of whether the indicators are meaningful representations of their theoretical constructs – whether they measure what they aim to measure. The quantitative internal validity of this study is guaranteed by the fact that the quantitative indicators are firmly grounded in theory. Furthermore, several statistical analyses have been conducted as robustness checks in order to ensure the quantitative causal validity of this study.
Moreover, this study triangulates different data sources in order to strengthen its internal validity (Yin, 2011, p. 81). Apart from the debate transcripts from which the quantitative data have been derived, the study makes use of semi-structured interviews. Semi-structured interviews are known for their internal validity since they allow the researcher to ask follow-up questions for clarification. In order to reduce the risk that the respondents give socially desirable answers, the interviews have been processed anonymously. The different sources of evidence have been compared to each other in order to corroborate the findings.
External validity concerns the extent to which the results of a study can be generalized to other situations (Yin, 2011, p. 226). Whereas internal validity is generally a strength of a case study, its external validity is often limited. Since the results of a case study are inherently contingent on its context, the findings of this study cannot be transferred to other contexts easily. However, this study makes use of a most likely case design which gives the study a degree of external validity. If there is no active parliamentary scrutiny in this case, it is probably nowhere to be found in EU affairs. Likewise, if we would find an ‘opposition deficit’ (Rauh and De Wilde, 2018) in this case, we will likely find it in other places, as well.
Reliability refers to the replicability of a study (Golafshani, 2003, p. 601). Moreover, it refers to the extent to which it can be repeated and yield the same results. A large part of this study – that is, the quantitative data collection from the debates – relies on subjective interpretation. Therefore, the reliability of this part of the study is especially a concern. Normally, this could be addressed by using multiple individuals as coders and ensure intercoder reliability, but given the individual nature of a thesis project, this was unfeasible (Creswell, 2007, p. 210). However, the study did rely on communications with a researcher who conducted a similar project to address several coding difficulties (Smeets, 2019, informal communication). Furthermore, the coding process has been registered in Excel documents. The reliability of this study is further strengthened by the fact that it documents the interview questions (Appendix IV). The interviews have been recorded and converted into verbatim transcripts. Thus, while the analysis depends on individual interpretation, the reliability of this study is guaranteed sufficiently.
15 Case Description
Three main issues dominated the EU policy agenda during the Refugee Crisis (Niemann and Zaunn, 2018, p. 5-12). The first issue entailed the attempt to close the EU’s porous external southern border by improving border control, creating hotspots, and offering assistance to Italy and Greece in managing the registration, sheltering, and the relocation of migrants. The second issue entailed the proportional relocation of eligible asylum seekers to other Member States, and the extent to which unwilling Member States could be forced to adhere to these relocation mechanisms. Finally, the third issue involved the attempts to strike deals with third countries, in order to further relieve pressure on the EU’s external borders and to create legal migration routes for eligible asylum seekers. The main example of this is the EU-Turkey agreement.
The Dutch government sought to reconcile its internal divisions about the Refugee Crisis. Whereas the PvdA had a moderately welcoming attitude to refugees, the VVD was more restrictive (Respondent 1; NU, 2015). Both parties, however, acknowledged the need for a collaborative European approach. The core of the government’s preferred EU approach was aimed at collaboratively combatting the root causes of migration, improving regional sheltering of refugees, establishing a proportional system to distribute asylum seekers among Member States, increasing the effectiveness of return-procedures for failed asylum seekers, and combatting human trafficking (Ministerie van Veiligheid en Justitie, 2015). The EU-Turkey deal fit in this strategy, which is exemplified by the leading role that the Netherlands played in the preparation of this deal (Peeperkorn, 2016).
The main opposition parties in de Tweede Kamer were the populist radical right Freedom Party (PVV), Socialist Party (SP), Christian Democrats (CDA), the progressive liberal democrats (D66), Christian Union (CU), the Greens (GroenLinks), and the orthodox Calvinist party (SGP). Whereas parties such as GroenLinks and D66 – and to a lesser extent CU and SP – advocated a Dutch Willkommenskultur, the CDA and SGP took a more cautious approach and voiced concerns about the limits to the abilities of the Dutch society to absorb the massive influx of refugees (NU, 2015). The PVV took an extreme position in parliament: it denied the notion that the Refugee Crisis required a European cooperative solution and urged the government to close the Dutch borders and deny entrance to every asylum seeker (PVV, 2015).
Leaving the PVV aside, the main parties agreed with the principles of the government’s envisioned EU approach to the crisis (Respondent 1, 3, 4, and 5). This does not mean that there was nothing left to discuss: parties differed in their conceptions of what number of refugees Europe could bear, under what terms deal with third countries could be made, the extent to which Turkey could be considered a safe country for refugees, and whether the Netherlands should lead by example in complying to the relocation settlements or whether Dutch compliance should be connected to other countries’ compliance. It does mean, however, that the PVV’s preferences were outside of the scope of the debate.
16 Quantitative Results
The parties made a lot of scrutiny interventions, resulting in clearly observable cross-party variance in both the amount (ranging from 2 to 73 scrutiny interventions per party) and levels of scrutiny (Figure 1). Especially the difference between the parties with multiple seats and the parties with only one seat in parliament stands out. The quantitative analysis does not take these smallest parties into account, due to a lack of observations and a lack of data on the independent variables. Given the limited involvement of these parties, their absence will not substantially harm the validity of the results.
The PPO model (Table 2) essentially produces three binary logistic regression analyses (Williams, 2006, p. 63). For instance, the first model contrasts the estimated likelihood that an intervention makes use of the first level of scrutiny – (dis)agreeing with the government’s position – with the higher levels of scrutiny. A positive coefficient indicates that an increase in the independent variable increases the estimated likelihood that an intervention makes use of a higher scrutiny level than the reference category. By doing this, the model enables us to discern different effects of variables from one level of scrutiny to another (Williams, 2006; Williams, 2016). For two variables – whether a party is in the opposition or government, and whether a debate is ex-ante or ex-post – the proportional odds assumption holds. This means that the effects of these variables are similar for each model. The quantitative analysis can only offer insights as to how the incentive-variables relate to levels of scrutiny, because the government-opposition variable and party size variable were multi-collinear in the analysis. The data-collection phase, in which the plenary debates were coded, indicated that the government-opposition variable is more important in influencing scrutiny behavior than party size. This initial impression was corroborated by the interviews (Respondent 1, 2, and 5). Therefore, party size has been excluded from the analysis, which means that the fifth hypothesis cannot be tested quantitatively.
The first hypothesis expects a positive relationship between party Euroscepticism and scrutiny levels. The quantitative analysis does not support this hypothesis. In the only model that reports a significant relationship between party Euroscepticism and scrutiny levels, the relationship is inverted: each unit increase of EU-favorability increases the log-odds that a party makes use of a higher scrutiny level than 1 – (dis)agreeing with the government – with .338. Thus, the analysis indicates that, for as far as we can speak of a relationship between party Euroscepticism and levels of scrutiny, it is negative. However, it is possible that this finding is disproportionally influenced by the behavior of the PVV in the debates. There seems to be a striking difference between levels of scrutiny between the PVV and the other parties. 45% of the interventions of the PVV – by far the most Eurosceptic party – remained at the first level of scrutiny, whereas for the other parties this was, on average, only about 12% (Appendix V).
The second hypothesis expects a positive relationship between the distance between the government’s position and a party’s preferences and levels of scrutiny. However, like the previous variable, to the extent that there is a relationship, the quantitative analysis points at an effect in the opposite direction. The distance in preferences is only significantly related to scrutiny levels in the first model. Each unit increase in the distance in preferences, decreases the estimated likelihood that a party
17
makes use of a higher scrutiny level than 1 – (dis)agreeing with the government – with -.432. This is surprising: expectedly, a greater distance between a party’s preferences and the government’s position would increase a party’s willingness to use higher levels of the scrutiny ladder.
For the third hypothesis – concerning the relationship between national salience and scrutiny – it is worth it to not only look at the level of scrutiny but at the amount of scrutiny as well. The dynamics of the number of scrutiny interventions and national salience are generally similar and correlate significantly with each other (r = .765; p < .001) (Figure 2). This corroborates the expectation that political parties’ behavior is partly informed by media coverage.
As Smeets and De Ruiter (2018, p. 575) also noticed, however, different dynamics are at play when we look at the level of scrutiny. National salience is only significantly related to higher scrutiny levels in the third model. This means that higher national salience is associated with a higher estimated likelihood that political parties actively scrutinize the government by providing specific instructions. The coefficient is statistically significant, but its substantial effect is not strong (β = .008, p = .014). However, it is worth keeping in mind that this variable has a wide range (Appendix V). One standard deviation increase in national salience results in an odds-ratio of instructing the government in comparison to other levels of scrutiny of 1.305 (Table 2). Thus, it seems that parties are more inclined to actively instruct the government in times of high salience.
The fourth hypothesis – which expects that opposition parties are more active scrutinizers than government parties – is corroborated by the quantitative analysis. The effect of whether a party is part of the opposition is substantially strong, statistically significant, and consistent among all levels of scrutiny. The odds that a party in the opposition makes use of higher levels of scrutiny are 3.124 as high as those of a government party (Table 2). This variable is the strongest indicator of its scrutiny behavior. Opposition parties tend to actively scrutinize the government by asking questions, offering alternatives, and submitting motions in order to instruct the government. Government parties, on the other hand, spend most of their time during the debate defending the government’s positions by addressing the concerns of the opposition parties, rather than actively scrutinizing the government’s positions. This means that MPs from parties in government both make less scrutiny interventions and – when they do – that these interventions often do not transcend the first two levels of scrutiny: (1) voicing (dis)agreement and (2) asking questions.
The difference between government and opposition parties is even more striking, considering that the version of this paper’s scrutiny ladder ranks voicing disagreement and agreement with the government’s position both as level 1, in order to reduce possible bias in favor of opposition parties (cf. Smeets and De Ruiter, 2018; Mastenbroek et al., 2014).
Finally, the analysis indicates that levels of scrutiny are structurally lower in ex-post debates, as compared to ex-ante debates. Although much caution is warranted in drawing conclusions from this – there were only three ex-post debates in the sample –, this makes sense. Instructing the government, for instance, after a European Summit makes less sense than doing it in preparation for a Summit.
18 Figure 1: All interventions, by all political parties (N= 608) 1
Figure 2: National salience and amount of scrutiny interventions. (r = .765; p < .001)
1 Exact numbers can be found in Appendix VI 0 10 20 30 40 50 60 70 80 1 2 3 4 0 20 40 60 80 100 120 27-12-2014 6-4-2015 15-7-2015 23-10-2015 31-1-2016 10-5-2016 18-8-2016 26-11-2016 6-3-2017
19
Table 2: Dependent variable: the level of scrutiny of an intervention
Estimates Standard errors p-value Odds ratios
1 (1 vs. 2, 3, and 4) Constant .063 .704 0.928 1.065 Position on EU integration .338 .070 .000** 1.402 Difference in preferences -.432 .118 .000** 0.649 National salience Opposition party Ex-post debate -.007 1.139 -.063 .004 .267 .263 .096 .000** .010** 0.993 3.124 0.533 2 (1 and 2 vs. 3 and 4) Constant Position on EU integration Difference in preferences National salience Opposition party Ex-post debate -2.516 .0284 .034 .001 1.139 -.063 .648 .052 .081 .003 .267 .263 .000** .586 .672 .711 .000** .010** .081 1.029 1.035 1.001 3.124 0.533 3 (1, 2, and 3 vs. 4) Constant Position on EU integration Difference in preferences National salience Opposition party Ex-post debate -3.696 .071 .043 .008 1.139 -.063 .666 .056 .086 .003 .267 .263 .000** .203 .615 .014* .000** .010** .025 1.074 1.044 1.008 3.124 0.533
Ordinal logistic regression with partial proportional odds, N = 531, *p < 0.05, **p < .01; McFadden R² = .0589
Nagelkerke R² = .149, Likelihood Ratio Chi² = 77.82 (p < .001)
20 Qualitative findings
In order to gain insight as to why the independent variables are (not) related to each other and provide understanding of mechanisms that lie underneath the relationships, we need the qualitative data. This part of the analysis draws on seven interviews and the transcripts of the analyzed debates.
Like the quantitative analysis indicated, party Euroscepticism does not seem to influence cross-party variance in levels of scrutiny. Although two moderately Eurosceptic interviewees indicated that their critical attitude vis-à-vis European integration stimulated them to actively scrutinize the government in European integration, others did not identify this as a general dynamic (Respondent 1, 2, 3, 4, and 5). These findings are not necessarily mutually exclusive. Whereas some MPs may choose to actively scrutinize the government on EU affairs out of Euroscepticism, others may do so out of enthusiasm for the EU. The data from the interviews refutes the indication from the quantitative analysis that some parties may be less active scrutinizers out of Euroscepticism.
The second hypothesis expected a positive relationship between the difference in preferences and scrutiny levels, but the statistical analysis did not corroborate this. In fact, one model resulted in the opposite relationship (Table 2). Interviewees indicate that parties make cost-benefit analyses, based both on policy-seeking and vote-seeking motivations (Respondent 1, 2, and 3). Respondents argue that party preferences should not only include substantial preferences, but also issue saliency on the party level. For instance, one MP explained why pressure from within her political party stimulated her to spend a lot of attention to scrutinizing the government: ‘There was nothing else I could do, because my supporters would not allow it. They wanted to speak about immigration with me every week’ (Respondent 1). This indicates that the more saliency a party attaches to a subject, the more intensively it will scrutinize the government on these issues.
However, these findings leave the question unexplained why both party Euroscepticism and a larger difference in preferences are associated with an increased estimated likelihood of using the lowest scrutiny level in the quantitative analysis. It seems that in the debates, these two variables played in interaction with each other (cf. Mastenbroek et al., 2014, p. 764). For the PVV, the party’s Euroscepticism and its preferences amplified each other: it had less interest in incrementally influencing the government’s positions in the upcoming EU Council because it rejected the entire premise that the Refugee Crisis was a complex problem that required a European solution (NU, 2015). When a party’s position is ‘Netherlands out of the EU and refugees out of the Netherlands’ – there is really not much left to debate about. Conversely, SP and SGP are usually critical of the EU. However, these parties acknowledged that the Refugee Crisis concerned a situation that required an EU-wide approach (Respondent 3 and 4). In this regard, the parties agreed with the bottom line of the government’s vision which created sufficient common ground to debate on. This common ground may be a prerequisite for higher levels of scrutiny.
The third hypothesis expected a positive relationship between national salience and levels of parliamentary scrutiny. The quantitative analysis indicated that the correlation between national salience
21
and amounts of scrutiny is strong and significant. However, the relation between national salience was only significantly related to a higher estimated likelihood of using the highest scrutiny level in comparison to the other levels. The qualitative data shed more light on this finding. All respondents confirmed that media attention is an important factor on which MPs base their decision to participate in debates. However, this is related to whether they participate in debates (i.e. the amount of scrutiny) rather than to scrutiny levels of interventions in these debates.
One respondent acknowledged that the quantitative finding that higher national salience is related to a higher likelihood of using the highest level of scrutiny (providing instructions to the government) makes sense intuitively – you have to strike the iron while it is hot (Respondent 3). Also, it may be that important policy decisions are often on the table when national salience is high, and therefore these times call for higher levels of scrutiny. Nevertheless, none of the interviewees confirmed to consciously alter their behavior in the plenary this based explicitly on media attention. Therefore, the relation between national salience and levels of scrutiny may be correlational rather than causal. In sum, the evidence from both analyses do not provide sufficient evidence to state that national salience is positively related to levels of scrutiny, but it plays an important role in determining the amount of scrutiny.
Like in the quantitative analysis, the government-opposition divide appears to be the most influential variable in the qualitative analysis. MPs from parties in government have several advantages that allow them to keep track of the state of play in the national government and, thereby, the European Council. However, these advantages make them less likely to make use of higher scrutiny levels during plenary debates. ‘When your party is in government, you use other means to influence the govern. If you can negotiate behind closed doors, there is not much need to do this in the plenary anymore.’ (Respondent 1). In other words: MPs from parties in government do not scrutinize the government in the plenary but in backrooms. During plenary debates, these MPs spend most of their time defending the government’s positions, rather than scrutinizing them. In this regard, European affairs do not seem to differ from regular domestic politics (Respondent 1, 4, and 5). All in all, the qualitative findings confirm the quantitative analysis that the government-opposition divide between parties in government plays the most influential role regarding variance in scrutiny behavior.
The fifth hypothesis expected a positive relation between party size and active scrutiny behavior, but could not be taken into account in the quantitative analysis. In the interviews, the respondents are clear: size matters. MPs from parties with fewer seats in parliament have fewer colleague MPs to share their work with and have fewer staff members. Therefore, the cost-benefit calculations whether a debate is worth participating in is extra salient for smaller parties. One respondent whose party suffered a substantial loss in seats in recent elections stated: ‘We rarely participate anymore. If there is no great political necessity or no salient political decision on the agenda, we often decide not to show up. You have to weigh your options’ (Respondent 1). As stated, such a calculation may include the potential for media attention that such a debate offers or the potential to influence policy. This effect may be even
22
bigger in committee meetings: ‘If you want to have direct influence about the larger issues, you have to be present at the plenary debates. These debates are about crucial moments, the crucial topics in Europe. That is when you have to be present’ (Respondent 3). Therefore, small parties may choose to skip committee meetings, rather than plenary debates (Respondent 3, 6, and 7).
These comments, however, relate to the amount of scrutiny – not necessarily the level of scrutiny at the moments when the MPs do participate. One respondent stated that the substance of interventions does not suffer from a lack of resources (Respondent 1). Another one said: ‘Still, I think that with the idea of ‘pick your fights’ we can have substantial influence’ (Respondent 3). In sum, it seems that party size plays in interaction with incentive-variables, such as the perceived possibility of policy-influence, media attention, and salience that a political party attaches to a topic (Respondent 1, 2, and 3).
Apart from seats, respondents hinted that capacity should be understood in broader terms than merely party size in parliament. Apart from seats in parliament, traditionally larger parties have institutionalized party think-tanks and access to party members who work in local government, academia, and civil society organizations. Such parties can reach out to those supporters easily and, thus, have more expertise at their disposal (Respondent 3 and 5). This implies that a party’s size in parliament is an incomplete indicator of its capacity. Higher institutional capacity can increase both a party’s capacity to actively scrutinize and its willingness because it creates a form of internal party pressure to assert itself on more issues (Respondent 1, 2, and 3). Parties could also enhance their capacity by communicating with their MEPs. However, the interviews suggest that this does not happen systematically and effectively among all parties (Respondent 1, 4, and 5). Even when former MPs are elected into the EP, consistent inter-institutional information exchanges seem to be difficult to implement: ‘You have to understand that, once they are in Brussels, they are in Brussels. And not in the Hague.’ (Respondent 5).
Apart from cross-party variance, the capacity of the Tweede Kamer as a whole to scrutinize the government is essential to discuss. The willingness and capacity of MPs to assert themselves in EU affairs have increased considerably in the last fifteen years (Respondent 4). Still, especially in European affairs, there can be staunching information asymmetries. Interviewees admit that this asymmetry is partly eliminated by the supporting staff of civil servants on the House’s EU Affairs Committee but it still should not be overlooked. A former policy advisor for a small party stated: ‘The distributions are completely skewed. The government has a huge staff at its disposal, whereas a small or even medium-sized party has only one or one and a half staff members on EU affairs’ (Respondent 2). An MP confirmed that this is an especially daunting task in EU affairs. He claimed that, often, MPs can hardly follow European politics: ‘There’s a lot going on in Brussels, you know. Ministers travel to Brussels all the time – there are all kinds of Council meetings – and the House of Representatives isn’t aware of half of what’s going on there.’ (Respondent 4). This may be even more visible in emergencies like the Refugee Crisis, as a CDA MP, Pieter Omtzigt, articulated during a debate: ‘And this is my main worry for this European Summit: you never quite know for sure what is being discussed and when it is decided
23
upon. And when a decision has been made, it is presented here as a fait-accompli. In this process, the democratic legitimation of these decisions tends to slip away sometimes’ (Tweede Kamer, 25 May 2015).
Conclusion and discussion
This article analyzed cross-party variance in parliamentary scrutiny behavior in EU affairs. As was expected on the basis of the most-likely case selection, the EU’s handling of the Refugee Crisis invited extensive scrutiny during the plenary debates around EU Summits in the Tweede Kamer. Explanations for variance in scrutiny in these debates concern two main factors: capacity and willingness. Parties’ capacity to actively scrutinize the government in EU affairs depends on party size in parliament. Larger parties tend to have more resources and time at their disposal and can, therefore, attend more debates than smaller parties. It does not, however, seem to influence a party’s behavior when they do decide to participate. Furthermore, the results indicate that parties’ capacity should also include its broader institutional resources. Capacity also depends on the access that a party has to the key decision-makers. In this regard, MPs from coalition parties have more access and, therefore, information about the state of play in the EU. Yet, coalition parties tend to scrutinize the government in private, rather than in public. This finding is in line with the literature and explains why the coalition-opposition variable has the strongest substantial effect on a party’s scrutiny behavior (Auel, 2007, p. 492; Mastenbroek et al., 2014, p. 764).
All in all, a party’s capacity to actively scrutinize plays in interaction with its willingness: parties make cost-benefit analyses as to whether they decide to participate in debates. They weigh the costs of preparing and participating in the debates against the possible benefits that participation can offer, such as media-attention and policy influence on salient topics. These calculations are especially of importance for small parties that have fewer resources. Therefore, the effect of a party’s capacity will likely have a more decisive influence on its scrutiny behavior in cases that are less salient or are less visible to the electorate, such as committee meetings.
Surprisingly, the quantitative analysis did not find that party Euroscepticism and the difference in substantial preferences between a party and the government positively relate to higher scrutiny across all levels. However, the interaction between party Euroscepticism and difference in preferences may help explain some of the variance, such as the striking difference in the behavior of the PVV and the other parties. It may be interesting for future studies to find out to what extent populism plays into this mix. Otjes and Andeweg (2019), studying cross-party differences in the Tweede Kamer, for instance, found that the behavior of populist parties ‘differ significantly from other opposition parties, in particular the smaller green, social-liberal and Christian parties’ (p. 492). This finding matches the results of this study, even though this paper ascribes the difference to the interaction of the party’s substantial and EU preferences. Therefore, populism could be included in future similar studies.
24
National salience correlates significantly with the amount of scrutiny in debates and it is significantly related to a higher estimated likelihood that parties actively instruct the government during plenary debates. However, due to the small number of observations in this study, it remains to be seen whether this effect can be transferred to other situations. It would be interesting to investigate this relationship in future studies.
To a large extent, parliamentary scrutiny about EU issues seems to be driven by the same variables that drive political behavior in domestic politics, where general opposition-coalition dynamics influence political behavior (Mastenbroek et al., 2014, p. 765; Smeets and De Ruiter, 2018, p. 581). Nevertheless, EU affairs offer an even greater challenge for MPs to stay up-to-date about decision-making processes in Brussels. Even though interviewees signal an increased assertiveness and willingness on the part of MPs to take on a role in EU affairs, it is hard to stay on track about the latest developments – especially in ‘emergency politics’ (White, 2015).
Methodologically, this paper has shown that independent variables may have different effects on different levels of scrutiny. Therefore, future studies that make use of the scrutiny ladder may consider using the PPO model as well (Williams 2006; Williams, 2016).
There are some limitations to this study that need attention. First, out of considerations of capacity, this paper draws upon a limited number of observations, during only one legislative term. Therefore, we should refrain from directly transferring conclusions from this study to another. Except for the finding that this study did not find signs of an ‘opposition deficit’ (Rauh and De Wilde, 2018) in the least likely case, still, a lot more research is needed to paint a more comprehensive picture of the role of national parliaments in EU policymaking processes. The crucial case approach ensures that the study has some external validity but the deeper specifics of this study can hardly be translated into other situations. A crucial case approach is interesting but constitutes the bare minimum for a healthily functioning parliament which fulfills its role in holding the executive to account in EU policymaking processes.
Furthermore, this study examined the extent to which political parties attempt to control their governments – not whether they succeed. Therefore, it still remains unclear to what extent the national executive’s ‘hands are tied’ when it enters negotiations in the EU Council. This could be interesting for future studies.
Finally, it became clear from this study that parties in the Tweede Kamer have quite some trouble in keeping track of the latest developments in European affairs. In an EU in which media and citizens are still primarily focused on politics on the national stage, national parliaments will have a vital role to play in legitimizing the EU and scrutinizing the executive in EU affairs in the foreseeable future. Especially in the Dutch political environment of increased fragmentation of parties, capacity limitations may hamper the ability of parties to fulfill these functions. Therefore, it would be wise to increase funds to expand this capacity. At the same time, political parties should aim to make smarter use of possibilities that are already available, such as improving communications with MEPs and MPs from family parties in other European capitals.
25
Bibliography
Anderson, B. (1983). Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso.
Auel, K. (2007). Democratic accountability and national parliaments: Redefining the impact of parliamentary scrutiny in EU affairs. European Law Journal, 13(4), 487-504.
Auel, K., & Christiansen, T. (2015). After Lisbon: National Parliaments in the European Union. West European
Politics, 38(2), 261-281.
Auel, K., & Raunio, T. (2014). Introduction: Connecting with the electorate? Parliamentary communication in EU affairs. The Journal of Legislative Studies, 20(1), 1-12.
Auel, K., Eisele, O., & Kinski, L. (2018). What happens in parliament stays in parliament? Newspaper coverage of national parliaments in EU affairs. JCMS: Journal of Common Market Studies, 56(3), 628-645.
Auel, K., Rozenberg, O., & Tacea, A. (2015). To Scrutinise or Not to Scrutinise? Explaining Variation in EU-Related Activities in National Parliaments. West European Politics, 38(2), 282-304. doi:10.1080/01402382.2014.990695
Bovens, M. (2007). Analysing and assessing accountability: A conceptual framework. European law journal,
13(4), 447-468.
Bovens, M., Schillemans, T., & Hart, P. T. (2008). Does public accountability work? An assessment tool. Public
administration, 86(1), 225-242.
Creswell, J. W., & Plano Clark, V.L. (2011). Designing and conducting mixed research methods. Thousand Oaks, California: SAGE Publications.
De Ruiter, R. (2013). Under the radar? National parliaments and the ordinary legislative procedure in the European Union. Journal of European Public Policy, 20(8), 1196-1212.
De Wilde, P., & Raunio, T. (2016). Redirecting national parliaments: Setting priorities for involvement in EU affairs. Comparative European Politics, 16(2), 310-329.
Dörrenbächer, N., Mastenbroek, E., & Toshkov, D. (2015). National parliaments and transposition of EU law: A matter of coalition conflict? JCMS: Journal of Common Market Studies, 53(5), 1010-1026.
Finke, D., & Dannwolf, T. (2015). Who let the dogs out? The effect of parliamentary scrutiny on compliance with EU law. Journal of European Public Policy, 22(8), 1127-1147.
Follesdal, A., & Hix, S. (2006). Why there is a democratic deficit in the EU: A response to Majone and Moravcsik.
JCMS: Journal of Common Market Studies, 44(3), 533-562.
Gerring, J. (2017). Case study research : Principles and practices. Cambridge, United Kingdom: Cambridge University Press.
Goetz, K. H., & Meyer-Sahling, J.-H. (2008). The Europeanisation of national political systems: Parliaments and executives. Living Reviews in European Governance.
Golafshani, N. (2003). Understanding reliability and validity in qualitative research. The qualitative report, 8(4), 597-607.
Guetterman, T. C., & Fetters, M. D. (2018). Two methodological approaches to the integration of mixed methods and case study designs: a systematic review. American Behavioral Scientist, 62(7), 900-918.
