Shaping EU policy in regional parliaments: a configurational analysis of the Posted Workers Directive in Germany

ABSTRACT

Since the Lisbon Treaty, research on parliaments in EU affairs turned to the regional level, but few studies ask how subnational legislators engage with the substance of EU policies. We examine this topic based on statements by the parliamentary groups in all German Landtage concerning the reform of the Posted Workers Directive, which became particularly salient when the European Court of Justice liberalized wage clauses in state procurement law. Under which conditions did the parliamentary groups support the reform? Our configurational analysis reveals that a left party identity is the only necessary attribute for support, and that it becomes sufficient in conjunction with the group being in opposition or with state policy being affected by European jurisprudence. We find little evidence that the local economic context mattered. The results partly confirm research on the Europeanization of state procurement law but highlight the importance of policy shaping from below.

KEYWORDS:

• European Union
• Germany
• labour policy
• parliaments
• QCA

Introduction

The question of how domestic parliaments adapt to European integration remains an evergreen in EU research, matched only by the role of the regions. Yet studies on regional parliaments are still exceptions (Abels and Battke 2019, 9). It was only in the wake of the Lisbon Treaty that scholarly interest caught up (Bursens and Högenauer 2017), as the new early warning system (EWS) gave regional parliaments with legislative powers a collective and suspensive veto against EU bills on grounds of subsidiarity. Most work on regional parliaments examines how they adapted their internal organization to Europe, but much less attention is paid to how these changes show up in parliamentary practice, especially outside the EWS. Very few studies, moreover, ask how state legislatures and their members engage with EU policy in terms of substance, namely which positions they take and why. Our contribution engages with these topics based on a configurational assessment using Qualitative Comparative Analysis (QCA) of the activities and positions of the parliamentary groups in all sixteen German state legislatures concerning the reform of the posted workers directive (PWD).¹ Our specific research question asks if and how party identity, government-opposition dynamics, and the economic environment combine to explain support for the reform.

Although the PWD reform triggered the third EWS yellow card, it was concluded in 2018. The reform was among the last decade’s highly contested dossiers, an instance of ‘mega-politics’ (Martinsen and Blauberger 2021) that opened international and ideological rifts. Touching on fundamental market freedoms, it was not only politicized but also highly judicialized, as it set out to override several rulings by the Court of Justice of the EU (CJEU) (Lubow and Schmidt 2021). Among those, the Rüffert judgement² sanctioned how German federal states regulated wages under public works contracts. Faced with compliance pressure, the states either liberalized their procurement regimes or tried to contain the judgement by defining statutory minimum wages (Blauberger 2012; Sack 2012; Sack and Sarter 2018; Seikel 2015). Accounts of the Rüffert saga emphasize that the German Länder (states) adapted their policies to top-down Europeanization, but given the issue’s salience, one might expect active pushback from the bottom up. Furthermore, existing policy studies treat the Länder as monoliths. Our configurational analysis and case study contribute missing knowledge to this literature by probing how the German Landtage (state parliaments) and their party groups engaged in bottom-up Europeanization during the PWD reform.

We show that many German Landtage attempted to shape the reform rather than content themselves with adapting regional policy to EU law. Yet within the parliaments, positions were far from uniform. Based on comparative politics and political economy literature, we examine to what extent party identity, government-opposition dynamics, and the economic environment conditioned the support expressed by state party groups. We find that left parties supported the re-regulatory thrust of the Commission proposal. By contrast, market-liberal and right-wing party groups rejected it. The revealed preferences thus strongly aligned with the socioeconomic left-right party cleavage. Although our QCA identified left party identity as a necessary attribute for support, at least one additional attribute was required to form a sufficient explanation. Specifically, left party groups engaged in reform-oriented activities in states impacted by judicial deregulation or when in opposition. Left parties in west German states, however, only show up as an explanation in an alternative empirical model.

The following section discusses how our analysis informs the debate about the role of parliaments in EU affairs. Next, we recall the main issues surrounding the PWD reform and its relevance for wage and procurement policy in Germany. This is followed by a discussion of theoretically plausible conditions and mechanisms that might – jointly or individually – account for whether state lawmakers express reform support. After a discussion of our data and methods, we describe the QCA and interpret its results. The conclusion discusses limitations and implications.

State parliaments in EU affairs

European integration poses a challenge to domestic parliaments because EU policy making relies strongly on joint decisions by national executives and supranational institutions still further removed from parliamentary control (Bursens and Högenauer 2017, 1; Abels 2013, 353). To the extent that their competencies are concerned, this problem is even more acute for subnational parliaments, as they oversee regional governments and lack direct influence on the central executive. How parliaments cope with integration is a long-standing question. A first wave of scholarship examined to what extent national parliaments adapted institutionally by developing procedures and capacities to keep up with EU bills and to scrutinize and control their governments’ behaviour in EU negotiations. Over time, national EU parliaments increased their control potential (see Winzen 2022 for an overview). Differences were linked to institutional strength (Karlas 2011; Winzen 2012), euroscepticism (Raunio 2009; Winzen 2013), and constitutional preferences more broadly (Senninger 2020). In a second wave, researchers turned to the question of behavioural adaptation, asking under which conditions legislators put available instruments to practical use. Cross-national and individual differences were mostly associated with government-opposition dynamics (Auel and Benz 2005; Finke and Dannwolf 2013; Wonka and Rittberger 2014), party preferences (Wonka and Göbel 2016; Wonka and Rittberger 2014), and administrative resources (Högenauer and Neuhold 2015). After Lisbon, similar issues were explored regarding national parliaments’ application of the EWS (Malang and Leifeld 2021).

Regional parliaments might improve the EU’s fragile input and output legitimacy by bringing the EU ‘closer to its citizens’ and by attuning its decisions to the actors in charge of implementation (Bursens and Högenauer 2017, 3–5; Donat and Placzek 2023). Although regions are an important topic in EU scholarship, it was not until the Lisbon reforms that research turned to regional parliaments (Abels and Battke 2019, 9). Most work so far deals with institutional adaptation (e.g. Abels and Eppler 2016; Borońska-Hryniewiecka 2017; Buzogány and Häsing 2018), while behavioural adaptation still receives less attention (Meyer 2023; Randour, Bursens, and Laloux 2022; Schneider, Rittberger, and Wonka 2014). The factors that explain differences in adaptation across national parliaments and legislators are often also relevant at the subnational level, although here the salience of and regional parliaments’ perceived influence in EU affairs are additional factors (Randour, Bursens, and Laloux 2022; Schneider, Rittberger, and Wonka 2014). Moreover, government-opposition dynamics may span multiple levels (Borońska-Hryniewiecka 2017; Buzogány and Häsing 2018). So far, most studies on behavioural adaptation take the level of activity as their explanandum, paying little attention to substantive policy positions (but see Randour and Wolfs 2017).

Our analysis contributes to the study of parliaments in EU affairs by examining three understudied areas: the regional level (all sixteen German state parliaments), behavioural adaptation (why state legislators adopt a specific EU policy position), and substantive issues of policy (a salient debate on posted work and judicial liberalization). We focus on Germany, a federal country in which legislative powers are shared between the federal Bundestag and the Bundesrat. The latter corresponds to a second chamber concerning most federal legislation, but it is composed of the Länder executives, which remain formal gatekeepers relative to their parliaments. In principle, the states are responsible for all policy areas not explicitly reserved for or pre-empted by the federal level. In practice, however, this leaves few genuine competences. In 2006, the Federal Constitutional Court confirmed that the states had authority to regulate the remuneration for public works (Sack 2012, 242). Whenever state powers are at stake, the federal government must give ‘prime consideration’ to the Bundesrat’s positions in EU negotiations, according to Art. 23(5) Basic Law. Since Lisbon, the Bundesrat may also trigger the EWS against EU bills that it regards as infringing on the principle of subsidiarity (Grotz and Schroeder 2021, 320). Although a few state parliaments can adopt binding mandates, their individual effect on the Bundesrat is limited, as subsidiarity complaints and Bundesrat resolutions require an absolute majority.

Landtage and posted work

The PWD is a cornerstone of the European single market. It regulates the ‘conditions under which employees may perform work across borders under […] the freedom to provide services’ (Lubow and Schmidt 2021, 325). The directive mediates between the services freedom and domestic labour regulation by allowing the host state to require certain minimum working conditions for employees posted from a company domiciled in another member state. As a result of EU enlargement, wages and working conditions within the EU became more heterogeneous, posted work became more frequent, and thus conflicts between (mostly western) net hosting states and (mostly eastern) net sending states emerged about the proper balance between the services freedom and social protection (Seikel 2022). Furthermore, the Court of Justice of the EU (CJEU) in 2007 and 2008 adopted a series of controversial judgments, known as the Laval Quartett, based partly on the PWD.³ Trade unions, left-wing members of the European Parliament, and several western governments responded by calling for the directive to be overhauled (Martinsen and Blauberger 2021, 53). As the CJEU had minimized the room for restrictions by domestic labour law on the free movement of services under the PWD, these voices demanded greater national discretion. In March 2016, the European Commission eventually proposed a revision. Following tough negotiations between east and west European states, and despite an EWS yellow card triggered by eleven national parliaments, the new directive was adopted in June 2018 (Martinsen and Blauberger 2021; Seikel 2022).

Germany was among the member states that supported the PWD reform, and it was affected by the Laval Quartett, in particular by the Rüffert judgement. The position of the federal government also reflected subnational politics. In half of the German Länder, regional procurement law stipulated that enterprises participating in public tenders must comply with collective agreements and pay union wages. Such wage compliance clauses were seen as important to protect regional companies from wage competition and to reinforce the shrinking collective bargaining coverage (Sack 2012, 242). Yet in the Rüffert decision on Lower Saxony’s procurement law, the CJEU declared wage compliance illegal under the services freedom in conjunction with the original PWD. Following the ruling, all Länder disapplied their wage compliance rules, but states (co-)governed by the Social Democratic Party (SPD) later introduced statutory minimum wages for procurement contracts, seen as compatible with the new case law (Blauberger 2012, 118–121).

Much research asks how state legislation adapted to European jurisprudence. This line of work shows that the introduction of statutory wages followed a clear partisan logic (Blauberger 2012; Sack and Sarter 2018; Seikel 2015), while the introduction of wage compliance was also driven by socioeconomic factors (Sack 2012, 247). By contrast, we do not know if and under what conditions state legislators supported changes in European legislation to override unwanted case law. Members of state parliaments (MPs) had reason to engage because attempts to contain the ruling by adapting regional legislation were fraught with legal uncertainty (Blauberger 2012, 121), and statutory wages could neither prop up bargaining coverage (Sack and Sarter 2018, 674) nor offer the same benefits as collective agreements (Seikel 2015, 1178). Our empirical analysis shows that their weak formal powers in EU affairs did not stop Länder parliaments from engaging. Whether they supported the PWD reform can be explained by their party membership in conjunction with other context factors. We discuss them in the following section.

Explaining reform support

In this article, we seek to identify the conditions under which Landtag party groups expressed support for the posted workers reform in their parliamentary activity. Based on comparative politics and political economy, this section discusses four potential drivers that, on their own or together, might lead parliamentary groups to express support: partisan identity, government-opposition dynamics, the economic environment, and the legal status quo. Since there is strong prior evidence that left parties support labour market regulation at different levels of governance, we focus on the configurations of party identity and other contextual factors that might jointly constitute sufficient conditions for reform support.

Partisan identity matters because parties unite legislators with similar ideologies and policy preferences. In Germany, most state legislators are elected on party tickets. Since the electorate contains different constituencies, parties distinguish themselves ideologically to maximize their vote share, and they substantiate their credentials by supporting policies that align with those views (Castles 1982; Schmidt 1995). The main dividing line between (centre-)left and (liberal-)conservative parties mirrors the class cleavage, which is why redistributive agendas like labour regulation are likely to elicit diverging partisan responses (Klingemann et al. 2006). While the effect of party identity on Länder policy choices depends on the issue area (Jeffery et al. 2014; Sack and Töller 2018), roll-call analyses indicate strong party discipline in state parliaments (Stecker 2011, 435). State parliaments lack direct control over policymaking at higher levels of governance, but rather than moderation this encourages even stronger positional demarcation (Stecker 2015). Empirical research on the adaptation of German state policy to the deregulatory Rüffert ruling also revealed a stark correspondence between local re-regulation and the governing party at the state level (Sack and Sarter 2018). Based on these considerations, we expect members of left party groups (SPD, Die Grünen, Die Linke, Piraten and SSW) to support the PWD reform. By contrast, members of market-liberal and centre-right to far-right party groups (CDU, CSU, FDP, Freie Wähler, AfD, and NPD) should reject the PWD reform.

Government-opposition dynamics affect the position-taking of regional parliamentary groups because opposition parties face different restrictions and make different strategic calculations from parties in government. In parliamentary systems, positions of the parliamentary majority are more congruent with the executive than positions of the opposition. Opposition legislators, therefore, have more reason to take a position on the PWD reform, while majority legislators may expect that the government will act in their interest. Moreover, majority legislators can rely on informal intra-party channels for controlling the executive, while the opposition is confined to formal parliamentary instruments like questions and resolutions (Holzhacker 2002; Wonka and Rittberger 2014). Although oppositional resolutions hardly ever get adopted, they are useful signals to the opposition’s constituency and serve as a means for politicization. By contrast, majority legislators may refrain from public criticism out of concern for the reputation of their governing party (Raunio and Wiberg 2010, 80; Wonka and Göbel 2016, 218) and to maintain slack in negotiations at higher levels of governance (Auel and Benz 2005). This logic presumably conditions the positioning of left legislators whose party identity should lead them to support the PWD reform, as explained earlier. Thus, we expect left legislators in opposition but not in government to support the PWD reform. Put differently, left party identity combined with opposition status should be a sufficient condition for a party group to express support for the PWD reform.

The economic environment in which state legislatures considered the PWD reform differs markedly between east and west Germany. Due to the challenging transformation to a capitalist economy during reunification, east German states still have smaller fiscal latitude and suffer from a lack of competitiveness relative to their western counterparts (Auel 2014, 435–436), resulting in diverging wage levels and collective bargaining coverage (Statistisches Bundesamt 2022a; 2022b). These regional contexts also gave rise to partly conflicting employer interests, especially in the construction industry, which had been the main target of wage compliance. Employers in west Germany were supportive of a regulatory wage floor to shield against east German wage competition. Conversely, most eastern employers rejected wage compliance as a form of protectionism (Seikel 2015, 1170; Sack and Sarter 2018, 676). Regardless of their ideological beliefs, legislators can hardly ignore the realities of regulatory competition; indeed, not a single east German state had implemented wage compliance before Rüffert, and east Germany was hesitant to introduce statutory minimum wages afterwards (Sack and Sarter 2018). We thus expect left legislators in west, but not east German parliaments to support the PWD reform. In other words, left party identity combined with a western state should be a sufficient condition for a party group to express support.

The legal status quo shapes state legislators’ position on the PWD reform because it is the default against which they evaluate their preferred policy choice. As explained above, the Rüffert ruling declared wage compliance clauses illegal. However, since not all regional procurement laws contained wage compliance clauses when the court issued its judgement, not all states were equally affected. Only in states with wage compliance did the judgement create pressure for adaptation. Those states could respond in two ways. First, they could succumb to ‘top-down Europeanization’ (Börzel 2002) and replace wage compliance with statutory minimum procurement wages, which is what most affected states eventually did. Second, they could pursue ‘bottom-up Europeanization’ and attempt to override the unwelcome ruling by supporting a reform of its major legal basis, the PWD. These options were not mutually exclusive, and it was initially uncertain whether the first would comply with case law (Blauberger 2012, 121). In other words, legislators’ perception of the need to support a reform of the PWD depended on whether their state procurement laws were affected by Rüffert. We therefore expect left legislators in states with (but not without) wage compliance before the Rüffert judgement to support the PWD reform. Thus, left party identity combined with a state with wage compliance should also be a sufficient condition for a party group to express support.

In sum, we expect party identity to generate support for the PWD reform in conjunction with one or more context conditions. In line with extant research, a left party identity should be considered an important background condition and a potentially necessary component of explaining reform support. If a party is ideologically predisposed toward deregulation, we expect it not to support a reform that seeks re-regulation, regardless of contextual factors. However, the literature is less clear regarding the impact of government status, socioeconomic context, and specific policy issues. While none of these is expected to compel market-liberal or conservative party groups to support the PWD reform, each might provide sufficient motivation for a left party group to do so. Without an obvious theoretical basis on which to hypothesize about the relative importance of each context condition or about how many are required to produce the outcome, we leave these questions to exploration.

Data and method

To examine these expectations empirically, we compiled a dataset of all parliamentary activities that referred to reforming the PWD. The dataset contains information about the relevant activities and the positions taken by all ten parliamentary groups⁴ that were present in at least one of the sixteen German state parliaments between 13 February 2008, the day of the Rüffert ruling, and 28 June 2018, when the new PWD was adopted. For each group, we distinguished between legislative periods when it constituted the majority or the opposition, yielding a total of 125 cases. Thus, while the unit of analysis is the party group, cases from the overall population are defined as a regional party group’s legislative period. By focusing on this period, we also account for changes over time, which is often neglected in QCA.

We inferred each group’s position during a legislative period from a host of parliamentary activities that might be relevant to scrutinize EU politics (see Winzen 2013), namely motions for resolutions, statements in plenary, parliamentary questions, the convening of relevant hearings, and votes on motions by other party groups. When these activities called for or supported the idea of reforming the PWD to extend national discretion over posted work, we coded the position as support. Conversely, when these activities rejected such a reform, we coded the position as against. We also coded the position of a group as against if its MPs voted consistently against bills or resolutions in favour of reform without tabling alternatives. In 72 out of 125 total cases, a party group did not engage in relevant parliamentary activities. Since no position could be meaningfully inferred, we excluded these cases from our analysis. Therefore, our empirical sample consists of each regional party group’s legislative period with parliamentary activity on the PDW reform in a total of 53 cases. Across all activities, each group’s positions were remarkably consistent. Whenever a group engaged in multiple relevant activities in the same legislative period, these were almost always either in support or against.⁵

The data were obtained from each parliaments’ online archive (Parlamentsdokumentation). We queried the archives using search terms related to the posting of workers, whenever possible using the full text search. In addition, we conducted ten semi-structured interviews with MPs or their staff to acquire contextual information and triangulate the written material, especially when the archives documented little activity (see supplementary material, Table A1). Regarding the potential explanatory conditions, we determined whether each state is in east or west Germany, the existence of wage compliance clauses in each state before the Rüffert ruling (from Sack 2012) and whether a group was in the opposition or in the majority (based on Bundeswahlleiter 2022; Schakel 2021; Schakel and Massetti 2018).

Our research design assumes that certain combinations of conditions explain reform support. Because these explanatory conditions can natively be operationalized (‘calibrated’) in a binary manner and our data set includes a medium number of cases, crisp set Qualitative Comparative Analysis (QCA)⁶ is an appropriate method. QCA is a set-theoretic method developed by Charles Ragin with foundations in formal logic and Boolean algebra, constructed from logical and set-theoretic principles. It allows moving back and forth between theory and evidence to assess different explanations through combinations (‘configurations’) of conditions (Ragin 2000, 45). QCA is particularly well suited to assess our theoretical expectations due to the relation between the number of cases and the conditions in our sample, and because it provides insights into situations where different equally valid explanations exist for a particular outcome (‘equifinality’).

Analysis

The analysis begins with a descriptive outline of the relevant parliamentary activity, followed by a description of our coding decisions and the QCA of necessary and sufficient conditions that gave rise to support.

PWD reform in the state parliaments

Rather than watching from the sidelines, many German Landtage actively engaged with the posted workers reform. Often, they did so based on controversial debates and already before the European Commission had submitted its proposal. We observed 59 parliamentary actions concerning the PWD, comprising twelve of the 16 Länder, with most actions taking place in Lower Saxony (see Table A2 in the supplementary material).⁷ The most frequent type of activity were plenary statements (34), often made in the context of a motion, which was the second-most frequent activity (22). Hearings (2) and questions (1) about the PWD reform were rare. Among the motions debated in Bavaria and Saxony were initiatives directed at the Bundesrat to object to the draft directive under the EWS. Interestingly, although such an objection is formally not an amendment right, the Saxonian resolution by Die Linke supported the reform and proposed to extend the directive beyond the proposal. By contrast, the Bavarian CSU initiative sought to stop the proposal in its tracks. Another five resolutions equally targeted the Bundesrat, calling on the respective state government to support the reform on the federal level but without invoking the EWS. These descriptive results attest to the strong salience and controversy of the PWD reform in German state parliaments.

While our observation period covers more than a decade, one third (20) of all parliamentary actions took place within 18 months following the Rüffert judgement. This temporal clustering suggests that state legislators were not only aware of the ruling but also pushed for its reversal through EU legislation, rather than contenting themselves with regional adaptation. Only some motions or statements explicitly framed the PWD reform as a means to override unwanted case law, but the main thrust of most parliamentary activities in support was to extend the scope of application for domestic labour law over posted work, contrary to Rüffert.

Calibration and coding

Since in our study each explanatory condition and the outcome of interest are binary, we employed a crisp set QCA. The condition ‘partisan identity’ (PARTY) was coded as ‘1’ if the party in question is politically left and as ‘0’ if the party belongs to the right. Following Bakker et al. (2021) and Lehmann et al. (2022), we classified the SPD, the Grüne, Die Linke, the Piraten, and the SSW as left. Conversely, we classified the CDU, the CSU, the FDP, the Freie Wähler, the AfD, and the NPD as right. We hypothesized that (left) opposition groups would express reform support. Thus, the condition ‘government-opposition dynamics’ (DYNAMICS) was coded as ‘1’ whenever the party was in opposition, and as ‘0’ when it was in government. The condition ‘economic environment’ (REGION) refers to whether the state is in east or west Germany. Because it seems plausible that (left) parties in the more prosperous west of Germany but not necessarily those in the less competitive east support the PWD reform, we coded the condition as ‘1’ if the state is in west Germany and as ‘0’ for east Germany. The condition ‘legal status quo’ (REGULATION) refers to whether wage compliance clauses existed in each state before the Rüffert judgement. Since we expected that (left) parties in states with liberalization pressure would support the reform, we coded this condition as ‘1’ if wage compliance clauses existed before the judgement, and as ‘0’ if not.

The outcome is conceptualized as ‘support for the PWD reform’ and was coded depending on the position of the party group toward reforming the PWD as stated in a parliamentary activity: The outcome is thus coded as ‘1’ if the party group in question expressed support for a PWD reform in its parliamentary activities, and as ‘0’ if its activities were directed against reform. As mentioned, we excluded cases in which the party group did not engage in any relevant parliamentary activity, because no position could be inferred. Thus, the negation of our outcome is defined as ‘expressing opposition’ and not as ‘lack of support’.⁸ Out of the remaining 53 cases, 28 showed a positive (activity in favour of reform) and 25 a negative (activity against reform) outcome. The calibration of conditions and outcome as well as related theoretical expectations are summarized in the supplementary material (Table A3).

We compiled a data table of all cases along with their coded values for each condition and outcome. From the data table, we constructed a truth table of all logically possible combinations of conditions, the distribution of empirical cases across these combinations, and related statements about the consistency of each row in producing the outcome. QCA conducts a logical minimization of the truth table (Table 1). The data table (Table A4) is shown in the supplementary material, and the raw data is published at the Open Science Framework (Deters and Corcaci 2024). Descriptive statistics on condition and outcome skewness do not reveal any issues, with mean values between 0.528 and 0.641 and standard deviation between 0.480 and 0.5 (see supplementary material, Table A5). Moreover, the coding of the conditions and the outcome was cross-checked independently by both authors and by another expert to ensure inter-coder reliability. The QCA was conducted using fs/QCA 4.1 (Ragin and Davey 2023) and cross-checked with the R package ‘QCA’ (Dușa 2019).

Necessary conditions

First, we investigated whether any (combinations of) conditions were logically necessary.⁹ First, the consistency score is assessed as a quality measure that determines whether a condition can be considered necessary based on the occurrence of contradictory cases. A consistency score of 0.9 is widely accepted across the literature as the threshold for a condition to be necessary (e.g. Oana, Schneider, and Thomann 2021, 81; Schneider and Wagemann 2012, 143). One individual condition passes this threshold, namely PARTY. Because we employ a crisp set QCA, the consistency score of 1.0 for PARTY means that there are no cases that contradict the conclusion that PARTY is necessary. Empirically, this means that no right-wing or market-liberal party group expressed reform support.

Next, we considered whether the necessary condition PARTY is actually relevant. For example, for a party group to be represented in parliament at all is clearly a necessary condition for the ability to express support for the PWD, but it is likewise trivial (see Schneider and Wagemann 2012, 144–145). This kind of trivialness is captured by the coverage value for necessary conditions (Ragin 2008, 60–63), which indicates situations where the condition set is much larger than the outcome set (Schneider and Wagemann 2012, 233). With a coverage value of 0.9333, however, PARTY is clearly relevant in this sense, whereas we do not consider other conditions to be relevant. Another kind of trivialness, which for large outcome sets is not captured by the coverage value, arises when a condition is almost constant (Schneider and Wagemann 2012, 233–234). For example, when the outcome of interest includes all parliamentary activities, not just those concerning the PWD reform, the aforementioned condition that a party group is represented in parliament at all remains trivial, but it would now pass the coverage test. We therefore also calculated Schneider and Wagemann’s (2012, 236) alternative relevance measure. With a value of 0.92, it confirms that the PARTY condition is relevant in the second sense as well. Therefore, PARTY is identified as the only necessary condition in a strict sense.¹⁰ The analysis of necessary conditions aligns with our theoretical expectation that party ideology should be part of all solution paths as a so-called INUS condition (insufficient but necessary part of a condition that is unnecessary but sufficient for the outcome; see Mahoney, Kimball, and Koivu 2009, 126). Parliamentary support for the posting reform always went along with a left party identity, which makes party identity a necessary (but insufficient) part of the explanation for support.

Sufficient conditions

We further applied a Standard Analysis of the sufficient conditions and thus of the possible explanations behind party group support (see Table 2). The frequency cut-off for the Standard Analysis was set to 1, meaning that all logical combinations of conditions for which at least one empirical case existed were included in the computation. The consistency cut-off, which is commonly set to a value between 0.75 and 0.8 according to the literature (Oana, Schneider, and Thomann 2021, 92; Ragin 2008, 46), was set to 0.8 with one non-perfect truth table row (Schneider and Wagemann 2012, 127). This is due to the crisp set truth table, which includes only rows with consistencies of 1 and 0, except for one row with 0.8 and one with 0.67 (see Table 1). The row with a consistency score of 0.8 includes one contradictory case (Saarland 02)¹¹ and was coded as ‘1’, while the row with a consistency of 0.67 was coded as ‘0’,¹² as were all remaining rows with consistency scores of 0.

Table 1. Truth table.

PARTY	DYNAMICS	REGION	REGULATION	Number	SUPPORT	Raw consist.	PRI consist.	SYM consist.
1	1	1	1	12	1	1	1	1
1	1	0	0	5	1	1	1	1
1	1	1	0	2	1	1	1	1
1	0	0	1	2	1	1	1	1
1	1	0	1	1	1	1	1	1
1	0	1	1	5	1	0.8	0.8	0.8
1	0	0	0	3	0	0.667	0.667	0.667
0	0	1	1	7	0	0	0	0
0	1	1	1	5	0	0	0	0
0	1	0	0	4	0	0	0	0
0	0	0	0	2	0	0	0	0
0	0	1	0	2	0	0	0	0
0	1	0	1	2	0	0	0	0
0	1	1	0	1	0	0	0	0
1	0	1	0	0	1	–	–	–
0	0	0	1	0	0	–	–	–

A crucial part of QCA is the analysis of ‘logical remainders’, logically possible combinations of conditions in a truth table that do not contain empirical cases. Such truth table rows illustrate the phenomenon of ‘limited diversity’ in social research, which QCA is uniquely able to address through different solution types and the manual analysis of logical remainders (Niikawa and Corcaci 2024). Our truth table contains only two remainders, namely the bottom rows 1*0*1*0 (PARTY*dynamics*REGION*regulation) and 0*0*0*1 (party*dynamics*region*REGULATION). In other words, apart from these two configurations, we find at least one empirical instance for all other truth table rows. Because of this unusually low number of remainders, all three solution types of the Standard Analysis (complex, intermediate, and parsimonious solutions) are identical for our data set, and thus we only discuss one set of solutions. After this first model (see Table 2), we also assess a second model that includes the two logical remainders (see Table 3).

Table 2. Model 1 (standard analysis of sufficient conditions).

Solution paths	Raw coverage	Unique coverage	Consistency
PARTY*DYNAMICS	0.714286	0.25	1
PARTY*REGULATION	0.678571	0.214286	0.95
solution coverage: 0.928571; solution consistency: 0.962963

The solution table provides the following overall model for explaining party group support: $\mathrm{PART}{\mathrm{Y}}^{\ast }(\mathrm{DYNAMICS}+\mathrm{REGULATION})\to \mathrm{SUPPORT}$The first model includes two alternative sufficient explanations that both include a left party identity as a necessary component (INUS condition), confirming our expectation that left party identity is a crucial part of the explanation for reform support among subnational MPs. The importance of party politics is consistent with extant policy studies on the PWD reform and on the Europeanization of German procurement law cited earlier. On its own, however, party identity is insufficient. The solution also suggests that left state MPs eschewed reform calls when their own party was in government, when regional procurement laws did not require wage compliance, or both. We find no solution path that includes the condition REGION in our data set. This implies that there is no explanation that includes REGION and covers cases not accounted for by the other two explanations. The absence of differences between east and west German states in our results contradicts our socioeconomic hypothesis and contrasts with earlier literature, according to which the economic environment shaped the introduction of wage compliance clauses. However, as we show later, the data permits an alternative explanation involving regional differences.

The two main solution terms explain a similar percentage of cases (raw coverage). Namely, PARTY*REGULATION explains 19 of the 28 cases where the outcome was present (∼68%), and PARTY*DYNAMICS explains 20 cases (∼71%). The QCA shows equifinality, i.e. several separate but equally valid explanations that partly overlap in the cases they can explain. Taken together, the two solution terms have a coverage of 0.9286, corresponding to 26 of 28 cases (∼93%) that show the outcome ‘reform support’, which is very high. The overall solution consistency, the main quality measure of QCA that is related to how many cases of the solution terms produce the outcome, is at a very high 0.963. This is because only one logical contradiction exists in the truth table.

The number of cases that are only explained by one particular solution term (‘unique coverage’) indicates the respective solution’s empirical relevance in relation to other explanations. Based on unique coverage, which here refers to the percentage of cases that are explained only by the solution in question, PARTY*DYNAMICS can be considered slightly more important. Its unique coverage is 0.25 (corresponding to 7 cases), while PARTY*REGULATION has a lower unique coverage of 0.2143 (6 cases). Altogether, the two solutions have very similar raw and unique coverage, highlighting the nexuses between the Rüffert decision or opposition status for one, and subnational support for reforming the PWD among left state MPs for another.

In the second model, we include the two logical remainders in our analysis based on theoretical expectations.¹³ This model also checks whether our socio-economic hypothesis is logically consistent with the data, while not being parsimonious. We included the logical remainder 1*0*1*0 (PARTY*dynamics*REGION*regulation) in the logical minimization process in line with the hypothesis that left party groups in west Germany support the posted workers reform. In contrast, we excluded the remainder 0*0*0*1 (party*dynamics*region*REGULATION) because theoretically, the former existence of wage compliance clauses should not lead to support on its own. Table 3 shows the results.

Table 3. Model 2 (sufficient conditions including logical remainders).

Solution paths	Raw coverage	Unique coverage	Consistency
PARTY*DYNAMICS	0.714286	0.178571	1
PARTY*REGION	0.642857	0	0.947368
PARTY*REGULATION	0.678571	0.0714285	0.95
solution coverage: 0.928571; solution consistency: 0.962963

Including the logical remainders produces a third solution term, namely PARTY*REGION, that is, left party groups in west Germany (with a raw coverage of 0.6429 or 18 of 28 positive cases). This solution is logically consistent with the data and corresponds to our hypothesized third explanation. It also provides an equifinal explanation for 18 empirical cases. To assess the relative plausibility of this additional solution, we compare the ‘within-case evidence’ (Goertz and Mahoney 2012, Ch. 7) in the next section.

Finally, although our study is not concerned with explaining opposition towards reform, which corresponds to the negated outcome, a brief assessment is considered good practice because of QCA’s inherent asymmetry between explanations for the occurrence of an outcome and its non-occurrence (Oana, Schneider, and Thomann 2021, 20). The analysis of the negated outcome ∼SUPPORT reveals that ∼PARTY constitutes the only relevant necessary condition, assuming a threshold of 0.75–0.8 for relevance,¹⁴ with a consistency of 0.92. Using the parsimonious solution, which is suitable due to the existence of only two logical remainders in the truth table, ∼PARTY is also the only sufficient condition with a consistency of 1.0 and a coverage of 0.92 (23 out of 25 negative outcome cases). Empirically, this means that being a right-wing party group is sufficient to express opposition to the PWD reform.

Discussion

In this section, we interpret the QCA results in terms of our theoretical expectations and complement them by within-case information.

Party identity matters

Our analysis reveals that party groups in German state parliaments actively engaged with the reform of the PWD, expressing support in about half of the cases in which they adopted a position, and often long before the draft directive had been tabled. Regarding our main question, under which conditions party groups came out in support, all solutions identified in the QCA contain party identity as a necessary component. Conservative and market-liberal parties, by contrast, did not support the reform, regardless of the context conditions.

The fact that party identity is a necessary component of all solution paths is striking, as non-trivial necessary conditions rarely occur in the social realm. At the same time, this finding chimes with the existing policy studies discussed earlier and confirms our expectation of a strong role for party politics. Wage and labour policy are deeply rooted in redistributive struggles along the class cleavage. Although affected by European regulation and market freedoms, they are prerogatives of national and, not least, subnational politics. The proposal for the new PWD rebalanced the European market freedoms of employers and the domestic social protection of workers in favour of the latter, and it was thus supported by unions and left parties in several member states (Martinsen and Blauberger 2021; Seikel 2022). Moreover, the PWD reform partly reversed the liberalization of procurement-related wage regimes that were mainly adopted under left Länder governments and that they defended using statutory wages (Sack and Sarter 2018). The support by left party groups for the PWD reform was consistent with these related positions. The nature of the policy area and the context of judicial liberalization thus go a long way in explaining why left and right party groups evaluated the PWD reform in stark contradiction.

Opposition dynamics trigger left reform support

Although we found a left party identity to be necessary, by itself it was not sufficient to generate reform support. Left groups only came out in support when they were in opposition or when the CJEU had struck down pre-existing wage compliance rules in their state. Both explanations cover roughly the same number of cases. We interpret the first mechanism, by which left party support is triggered by opposition status and inhibited by government participation, as a logic of institutional (dis-)incentives. Parliamentary majority groups may refrain from openly criticising ‘their’ government, simply because they expect the government to implement the majority mandate or out of concern for reputational damage and electoral punishment. Opposition groups, conversely, not only stand to gain from politicization, in scrutinizing the government and raising neglected issues they also perform their constitutional role. This result confirms the importance of opposition dynamics in the use of scrutiny instruments by domestic parliaments (Finke and Dannwolf 2013; Wonka and Rittberger 2014). However, it contrasts with research on plenary debates in national-level parliaments, which found opposition parties ‘dropping the ball by not debating Europe to a similar extent as government parties’ (Rauh and Wilde 2018, 210–211).

The seven Länder in our sample that due to the absence of wage compliance clauses were not affected by the Rüffert decision exemplify a ‘pure’ left opposition pathway. In North-Rhine-Westphalia, the SPD group expressed support for the PWD reform in parliament, urging the government to ‘work towards the reform on all political levels’,¹⁵ but it did so only (and shortly) after it had been ousted from government in the May 2017 elections. While in government, the red-green government consistently promoted reforming EU law to regain ‘manoeuvrability for wage compliance’ (MBEM 2010, 12), but its parliamentary groups showed no corresponding activity. Conversely, the SPD group in Mecklenburg-Vorpommern rejected a pro-reform motion while governing with the CDU.¹⁶ The coalition government of the SPD and Die Linke in Brandenburg is the only example among the states unaffected by Rüffert in which left party groups in government supported the PWD reform. This was part of ongoing deliberations about the impact of the free movement of labour on the Polish-German border region.¹⁷ The contrast between Mecklenburg-Vorpommern and Brandenburg also highlights that coalition politics may prevent left party groups in government from expressing support, in addition to the reputational considerations and preference (mis-)alignment that our hypothesis focused on.

Liberalization triggers left reform support

Left party groups also expressed support for the PWD reform when the procurement regimes in their state included wage compliance clauses. We interpret this mechanism as left party groups responding to the deregulation of state procurement law by the CJEU. Consistent with their prior support for wage compliance and their efforts to contain the ‘Rüffert shock’ (Schulten 2012) through state legislation, left parties pushed to override the judgement by means of EU legislation, even at the level of state parliaments far from Brussels and Luxembourg. For example, the SPD and Die Linke groups in the Brandenburg Landtag asked their government to ‘make it clear that the directive only includes minimum requirements and is not a ‘maximal directive’’.¹⁸ This demand sought to correct an especially problematic aspect of the Rüffert ruling that turned the PWD’s social provisions from minimum standards that the member states could exceed by requiring union wages into a legal ceiling on market restrictions (Seikel 2022, 509). A more explicit link between judicial deregulation and the PWD was drawn by a Die Linke MP in Mecklenburg-Vorpommern, who argued that ‘it is time, especially since the court ruling, for the federal government to fight for […] an employee-friendly reform of the posted workers directive’.¹⁹ Neither Brandenburg nor Mecklenburg-Vorpommern are among the states with pre-existing wage compliance clauses, but even there, left lawmakers associated the PWD reform with Rüffert.

Of course, references to the ruling were also made in states directly affected by it. One example is Bavaria, where the conservative CSU government seized the opportunity of Rüffert to abolish wage compliance and refrained from introducing statutory procurement wages instead (Sack 2012, 253). The oppositional SPD not only voted against a CSU motion on triggering the EWS against the draft directive,²⁰ it also sponsored a motion in support of the PWD reform that sought to make collective agreements without universal application legally binding on posted work,²¹ which directly contradicted a core element of the Rüffert decision (Sack 2012, 251). In six legislative periods across three states with pre-existing wage compliance clauses, we found left party groups that supported the PWD reform despite being in government. In Berlin, the governing SPD and Die Linke groups sponsored a joint motion in the Rüffert year that opened with several references to CJEU jurisprudence and a call for revising the PWD. The closing paragraph explicitly referred to the constraint that the ruling imposed on wage compliance.²² In Hamburg, the governing SPD group rejected a pro-reform motion by the oppositional Die Linke,²³ but it sponsored a motion in support of a Bundesrat resolution by the states of Hamburg and Rhineland-Palatinate in support of the PWD reform and stronger European worker’s rights. With explicit reference to Rüffert and other cases of the Laval quartet, the resolution criticized the CJEU for ‘reinterpreting’ the original PWD ‘at the expense of the employees’.²⁴ The example of Schleswig-Holstein is less clear cut. Here, the three groups of the left governing coalition called for a reform of the PWD, but we found no reference to jurisprudence. Instead, their parliamentary activity focussed on better control and implementation of the PWD’s social provisions.²⁵

Partial evidence that socioeconomic differences trigger left reform support

Our second model reveals another mechanism, by which left party groups in western states support the reform to protect their high wage levels from inter-regional competition. This mechanism corresponds to our third hypothesis, providing an equifinal explanation for 18 cases. However, not only is this model less parsimonious, the within-case evidence also shows that the other two explanatory mechanisms more convincingly reflect the data.

Left party identity in west Germany (PARTY*REGION) thus constitutes a potential alternative explanation for reform support. In North Rhine-Westphalia, the left party groups supported reform once they had gone into opposition. Yet concerns about competition from east German companies were absent from the relevant debates and resolutions, and employer representatives who participated in parliamentary hearings were sceptical of the reform instead of welcoming it as protection against competition (Landtag Nordrhein-Westfalen 2018, 22–25). In Lower Saxony, the Rüffert judgement, which explicitly targeted the state’s wage compliance regime, also figured prominently in the Landtag,²⁶ but all references to ‘wage dumping’ concerned low-wage EU countries instead of east Germany.²⁷ The same applies to Bavaria,²⁸ but in contrast to Lower Saxony, parliamentary support by the Bavarian SPD was related to its opposition status. As one SPD legislator explained, ‘I didn’t expect to get a majority […], but we tried […] to create political publicity’.²⁹ Overall, within-case information in our data set offers limited support for the socio-economic hypothesis. Instead, the solution PARTY*REGION might at least in part be attributed to an east-west bias in the adoption of wage compliance regimes, making it hard to distinguish from PARTY*REGULATION, the influence of the status quo.

Limitations

Finally, we consider our single deviant case, namely the position adopted by Die Grünen in the Saarland Landtag during the 14th legislative period. At this time, the small west German state was governed by an uncommon ‘Jamaica’ coalition of the CDU, the FDP, and Die Grünen. The oppositional SPD sponsored a motion against ‘wage and social dumping’ in the context of the free movement of labour.³⁰ It explicitly supported the PWD reform but called for even more comprehensive changes. The motion acknowledged that the Saarland government had already ‘encouraged’ a reform of the PWD in the Bundesrat, but it asked the executive to also support the above-mentioned Bundesrat resolution³¹ by Hamburg and Rhineland-Palatinate. The government parties, including Die Grünen, rejected the motion. The position of the Die Grünen party group contrasts with our hypothesis, confirmed by the other cases, that left parties in states with wage compliance support the reform. The SPD motion also did not refer to the Rüffert ruling, which suggests that its impact on wage compliance was not a salient concern. In fact, at the time of the motion, Saarland had already introduced statutory procurement wages that partly buffered the ruling’s deregulatory impact (Sack and Sarter 2018, 679). The deviant position of the Grünen group could thus be explained by a combination of factors. First, the Saarland government already supported the PWD reform, which made additional pressure by the Grünen group partly redundant, in line with our opposition hypothesis. Second, Die Grünen were in a coalition with right-wing parties, which constrained the position of their parliamentary group more than a leftwing coalition might have. And third, the impact of Rüffert was not sufficiently salient to overcome these countervailing tendencies. In sum, the influence of government participation prevailed over the influence of the legal status quo.

Overall, the Saarland example highlights some limitations of our results. While the PWD reform and the adoption of statutory procurement wages at the regional level were not mutually exclusive, the latter could have reduced the incentive to support the PWD reform in states with pre-existing wage compliance clauses. In a counterfactual scenario where the PWD reform was the only way to mitigate the deregulatory impact of CJEU jurisprudence, we might have witnessed stronger support. Moreover, our opposition hypothesis presumes that the views of majority party groups are more congruent with their governments than those of opposition groups, but we acknowledge that the level of congruence is subject to empirical but unmeasured variation. In Saarland but also in North-Rhine Westphalia, governments supported the PWD reform despite their different composition.

Conclusion

This study examined how regional parliaments respond to Europeanization pressure. In contrast to most existing studies, we focused on behavioural adaptation and substantive policy positions, namely on how and when state legislators try to shape EU decisions in their everyday parliamentary activity. We chose the reform of the PWD as an example because it was salient enough to trigger parliamentary activity and allowed us to complement available research on the top-down Europeanization of state policy. We found that the German Landtage attempted to shape and even instigate EU-law reform against the backdrop of constraining European jurisprudence rather than just making state law litigation-proof. More importantly, we also found that the revealed reform preferences of the Landtag party groups closely aligned with their party identity. The PWD reform was supported by left but not by right and market-liberal groups to an extent that made it appropriate to consider left party identity a necessary condition for support. But only the combination of a left party in opposition or of a left party in a state with European liberalization pressure emerged as sufficient conditions. While right and market-liberal parties disfavoured the reform, left parties supported it in response to the institutional incentives and constraints of their opposition status and to curb judicial deregulation.

While confirming some existing research regarding the Europeanization of German wage and procurement regimes, these findings highlight certain issues that have been overlooked in existing policy studies and the EU parliaments literature. First, rather than silently adapting regional policy to judicial liberalization, left state legislators actively pushed for EU-level reform. Second, at the intra-institutional level, the more relevant difference between domestic legislators is not so much the intensity but the substantive quality of their activities. Third, domestic parliamentary activity not only responds to pending legislation but also to judicial policy making at the EU level. Finally, although the partisan logic emphasized in the policy literature extends to parliamentary politics, it is conditioned by differential adaptation pressure in each state and by the government-opposition dynamics that structure the parliamentary arena.

Our research design allowed us to discover different equifinal mechanisms behind reform support, but it also limits the external validity of our findings. The sixteen states we examined are all part of a single country, and we focused on a single dossier. Future research, qualitative or quantitative, could extend this line of inquiry to other countries and policy issues. We presume that the intensity of mobilization and the strength of partisan differences depend on issue characteristics (salience, regulatory or redistributive character), and that opposition dynamics play out differently in more complex federations and party systems like, for example, the Belgian one. In normative terms, our analysis finally highlights the importance of domestic opposition parties for the scrutiny of EU politics, especially since the EU lacks an institutionalized central opposition (Mair 2007). It moreover illustrates an underappreciated shortcoming of the EWS, which as a collective veto cannot initiate policy reform against unwelcome CJEU rulings. Rather than safeguarding subsidiarity, it may thus reinforce existing asymmetries between political and judicial policy making.

Acknowledgements

The authors would like to express heartful thanks to François Randour, who participated in useful discussions about earlier versions of the manuscript, Sho Niikawa for helpful methodological comments, Moritz Klock for research assistance, and the members of the Regioparl project for feedback during the conception of this study. Andreas Corcaci's work was co-funded by the European Union under grant agreement ID 101105696. Views and opinions expressed are however those of the author only and do not necessarily reflect those of the European Union or the European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. This research was funded in part by the Austrian Science Fund (FWF), grant-DOI: 10.55776/P33773. For open access purposes, the author has applied a CC BY public copyright license to any author-accepted manuscript version arising from this submission.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by FWF (Austrian Science Fund) [grant-DOI: 10.55776/P33773]; Andreas Corcaci's work was co-funded by the European Union [grant agreement id 101105696].

Notes

1 Directive 96/71/EC, replaced by directive 2018/95/EC.

2 C-346/06.

3 Laval (C-341/05), Viking (C-438/05), Rüffert (C-346/06), European Commission v Luxembourg (C-319/06).

4 AfD, CDU, CSU, Die Grünen, FDP, SPD, Freie Wähler, Die Linke, Piraten, NPD, SSW. When members of a regional parliament were too few to be accorded Fraktionsstatus, we still counted them as a group.

5 In three cases of inconsistent activities, we coded the position of the more significant activity (see Deters and Corcaci 2024).

6 ‘Crisp set’ refers to the variant of QCA where the conditions and outcomes can only take on two values, either 0 (‘false’) or 1 (‘true’). An extension (and generalization) of this variant that accounts for differences-in-degree (Schneider and Wagemann 2012, 27) is called fuzzy set QCA (Ragin 2008).

7 Only Baden-Württemberg, Bremen, Hesse, and Rhineland-Palatinate showed no activity whatsoever.

8 Consequently, our hypotheses do not make any claims about the negated outcome.

9 So-called SUIN conditions (sufficient but unnecessary parts of a factor that is insufficient but necessary for the result; Schneider and Wagemann 2012, 79) are not expected because only party group affiliation is theoretically predicted to possibly form a necessary condition and no SUIN conditions emerge from the theoretical assumptions. This is confirmed by analysing necessary conditions with the ‘superSubset’ function of the R package ‘QCA’ (Dușa 2019), which reveals no relevant SUIN conditions (see supplementary material, Table A6).

10 By ‘strict’, we refer to the requirement for necessary conditions to cross a threshold of 0.75–0.8 for both coverage and ‘Relevance of Necessity’ (RoN; Schneider and Wagemann 2012, 256 – the authors consider a RoN value of 0.56 as ‘low’) to be considered relevant (Corcaci 2019, 237).

11 Saarland 02 is a so-called ‘deviant case consistency in kind’, see the discussion below.

12 This row includes two so-called ‘deviant cases coverage’, Brandenburg 07 and 08, see the discussion below.

13 This model can be related to Theory-Guided Enhanced Standard Analysis (TESA), an extension of Ragin’s Standard Analysis by Schneider and Wagemann (2012, 211–212), which refers to two strategies for dealing with logical remainders. One is to include theoretically sound truth table rows as ‘good’ counterfactuals, another is to formulate directional expectations for conjunctions of conditions. TESA rejects the premise of Standard Analysis that parsimony is the ‘guiding principle for choosing logical remainders for counterfactuals’ (Schneider and Wagemann (2012, 212) – a contested approach in the QCA literature. Instead, it can be used to include theoretically sound remainders even though they do not make the result more parsimonious, which is the approach followed here (see Corcaci 2019: ch. 5.4).

14 See footnote 11.

15 E.g. Drucksache 17/1122.

16 Drucksache 5/2624.

17 Drucksache 5/3847.

18 Drucksache 5/3847.

19 Helmut Holter (Die Linke), 40th plenary session of the Landtag of Mecklenburg-Vorpommern, 24 April 2008.

20 Drucksache 17/10874.

21 Drucksache 17/19581.

22 Drucksache 16/1840.

23 Drucksache 20/7222.

24 Drucksache 20/526, BR-Drucksache 196/11.

25 Drucksache 18/746.

26 E.g. Drucksache 16/808.

27 E.g. Detlef Tanke (SPD), 28th plenary session of the Landtag of Lower Saxony, 15 January 2009, 3313.

28 E.g. Drucksache 17/6665.

29 Interview with a Bavarian state legislator of the SPD group, 23 April 2021.

30 Drucksache 14/448-NEU.

31 BR-Drucksache 196/11.