Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Abstract

This study analyses how preference heterogeneity across EU member states affects the adoption of soft-law acts over time. On the one hand, high diversity in policy preferences is expected to increase the proportion of soft-law instruments because governments are less likely to agree to binding measures. Conversely, preference heterogeneity could also decrease soft law due to the perceived threat of compliance problems. We test these competing arguments using a dataset on all EU soft-law and hard-law instruments adopted between 1967 and 2019. The results show that preference heterogeneity increases the share of soft EU instruments. However, more past heterogeneity prompts EU legislators to decrease the proportion of softer measures in areas that experience high levels of past non-compliance.

Keywords

EU compliance Council preference heterogeneity soft law

Introduction

There is a well-established view that more inclusive regimes impede co-operation due to the resultant heterogeneity in preferences (Buchanan and Tullock, 1962). As regimes expand their membership, the chance that at least one participant will not support co-operative outcomes increases. In the case of the European Union (EU or Union), by contrast, increased diversity has not impeded the continuous transfer of policy-making powers from the national to the EU level. Since its establishment, the EU has considerably expanded its involvement and authority in almost all policy areas. At the same time, membership expansion has made member states’ policy preferences more diverse. Thus, empirical evidence on the co-evolution of the EU's membership base and competences generally dismisses the existence of a trade-off between ‘deepening’ and ‘widening’ of supra-national systems (Leuffen et al., 2013: 21).

Nevertheless, the increasing heterogeneity in the EU could still make the EU's legal system less stringent. In addition to generally binding legislation, the EU also produces soft-law instruments (e.g. recommendations, resolutions, guidelines, opinions and communications) that lack supra-national enforceability and hence depend on the voluntary application by the member states. A novel database on soft law shows that non-binding instruments are on the rise in the EU (Cappellina et al., 2022). Soft law presents an alternative to hard law in areas where member states are unable to agree on binding legislation. In addition, the EU issues numerous communications and notices that complement hard law as they aim to facilitate the implementation of EU law in domestic settings.

A vast body of literature discusses the normative implications of soft law. Proponents of these ‘flexible’ modes of governance emphasise the lower legislative and sovereignty costs in the adoption of non-binding instruments (Radaelli, 2003; Tholoniat, 2010; Trubek and Trubek, 2005), whereas critics point out the danger of ineffective implementation and ‘race-to-the-bottom’ scenarios in the absence of enforcement mechanisms (Schäfer, 2006). Moreover, the increased use of soft-law instruments instead of binding legislation threatens the legitimacy of the EU because they lack the certainty, transparency and the legal protection provided by hard law, as the latter is adopted by the classic Community method (Idema and Kelemen, 2006; Majone, 2009).

However, it remains unclear how growing differences in government policy preferences affect the development of soft law. This is an important gap, as the development of non-binding instruments could explain how a supra-national system reacts to policy conflicts and the nature of co-operation in a highly diverse Union. Moreover, research shows that variation in member states’ policy positions has empirical implications for EU decision-making processes (König and Luig, 2012). This study addresses gaps in the literature by theorising and assessing the effect of member states’ preference heterogeneity on the evolution of EU soft law over time. Past research showed that coalition formation in the Council of the EU (Council) is shaped by the composition of political parties in national governments (Hagemann and Hoyland, 2008). Furthermore, political contestation on substantive policy issues is largely captured by the left–right dimension (Crombez and Hix, 2015: 488). Drawing on past research, we focus on preference heterogeneity between member states and conceptualise it as a function of policy-specific positions of government parties.

On the one hand, preference heterogeneity is expected to hinder policymakers’ agreement on binding legislation. In this case, soft law may offer a political alternative to binding legislation or supplementary measures designed to ameliorate conflicts between legislators. On the other hand, diversity of preferences signals to member states and the EU institutions that governments will ignore soft rules, unless there is a strong enforcement mechanism. Hence, soft-law instruments may create commitment problems in the EU, casting doubt on the credibility of EU co-operation. Consequently, heterogeneity of preferences may lead to the adoption of more binding instruments (relative to soft law) to ensure that member states comply. This is especially the case for areas that experience high levels of non-compliance, and where the adoption of soft-law measures may aggravate implementation problems.

We test these competing arguments using a novel dataset on soft-law and hard-law instruments adopted between 1967 and 2019 in 24 policy areas. To our knowledge, this is one of the most comprehensive datasets on the volume of non-binding supra-national acts. We complement our dataset with sector-specific and time-variant measures of member states’ preference heterogeneity on the left–right dimension based on the ParlGov database and the Comparative Manifesto Project. We find that preference heterogeneity increases the share of soft-law instruments. However, in the long run, it signals commitment problems in areas already experiencing high levels of non-compliance. In this case, EU member states are reluctant to exacerbate policy divergence by further softening EU acts. In addition, although the share of soft law increased over time, most non-binding instruments supplement, not substitute, binding EU acts. These findings alleviate concerns for the weakening of the EU's legal system. Policy diversity does not always lead to loose regulation.

Conceptualising soft law in the EU

Research on soft law finds its roots in the legal literature of different EU modes of governance. Soft law is generally defined as ‘rules of conduct which, in principle, have no legally binding force, but which nevertheless may have practical effects’ (Cappellina et al., 2022; Senden, 2004; Snyder 1993: 2). The literature is inconclusive about the meaning of soft law. European scholars mostly associate the concept with new modes of governance promoting a decentralised policy-making process to address domestic problems (Héritier and Lehmkuhl, 2011). Accordingly, soft law encompasses non-binding instruments such as codes of conduct, resolutions, communications, declarations, guidance notes and action plans among others (Cappellina et al., 2022).

In a recent work, Terpan and Saurugger (2021) propose more concrete criteria: obligations and enforcement, to differentiate hard and soft law. In particular, they define hard obligations as clear EU acts with limited discretion in the implementation process. Enforcement refers to whether EU acts are subject to judicial scrutiny and control (Abbott and Snidal, 2000; Saurugger and Terpan, 2021). In this study, we analyse different types of non-binding EU instruments across time and policy areas. Therefore, we adopt the more general conceptualisations proposed by Snyder (1993) and classify soft law as non-binding EU instruments with the potential to exert influence on member states, but are not subject to enforcement by supra-national institutions (also see Cappellina et al., 2022 for similar conceptualisation).

In addition, we recognise that soft-law and hard-law acts are not mutually exclusive (Slominski and Trauner, 2021; Terpan, 2015). In many policy areas, hard law cannot be replaced by non-binding instruments. Most soft-law acts complement binding EU rules by providing interpretation of existing EU legislation (Saurugger and Terpan, 2021). For example, when EU policymakers failed to amend the EU Return Directive of third-country nationals, they clarified the interpretation of the directive with the legally non-binding Return Handbook (Slominski and Trauner, 2021: 95). Moreover, the member states and EU legislative bodies are not the only authors of soft-law acts. For example, the European Commission (Commission) unilaterally produces various communications, notices and guidelines (Senden, 2004). A novel database on soft law shows that the Commission produced 83% of all non-binding instruments in the area of state-aid in the period between 2004 and 2019 (Cappellina et al., 2022: 13). These instruments generally aim to complement existing EU hard law (Zhelyazkova et al., 2015).

Explaining soft law in the EU

Most studies develop functionalist rationales to explain the choice of soft law during international and European negotiations (Abbott and Snidal, 2000; Schäfer, 2006; Slominski and Trauner, 2021; Terpan and Saurugger, 2021; Trubek and Trubek, 2005). Accordingly, non-binding instruments serve to reduce legislative and sovereignty costs in supra-national decision making. As a ‘compulsory negotiation system’, the EU fulfils its objectives through supra-national agreements, which are more difficult to realise with an increasing number of veto players in EU policymaking (Scharpf, 2006; Tsebelis, 1995). Heterogeneous policy preferences exacerbate the adoption costs during EU negotiations. In situations of political conflict, soft law helps EU policymakers overcome “EU-decision traps” and is a quicker alternative to legally binding EU measures. This is especially useful in times of crises, where policy disagreements can hamper the ability of policymakers to react swiftly. Research shows that member states could react swiftly in crisis management situations in external migration by excluding the European Parliament from the negotiations (Slominski and Trauner, 2021). In extreme cases, high diversity in member states’ preferences could cause legislative gridlock, where EU policymakers are unable to agree on binding legislation. Blauberger (2009: 725) shows that soft law dominated the area of state aid largely because the EU member states refused to agree on binding legislation (Blauberger, 2009). When member states disagree about substantive policy, soft law may prevent legislative gridlock.

Similarly, soft law makes co-operation less costly by allowing governments to deviate from supra-national agreements. Allowing for policy diversity in implementation could alleviate disagreements among member states during EU policymaking (Schäfer, 2006). For example, Slominski and Trauner (2021) highlight that non-binding instruments in external migration policy help EU policymakers signal their willingness to engage with third countries without making hard commitments (Slominski and Trauner, 2021).

Nevertheless, recent studies denote that efficiency-based explanations for soft law are conditional on the commitment and the importance that policymakers place on specific issues. For example, analysing the relevance of different EU crises, Terpan and Saurugger (2021) found that despite the ‘hardening’ of EU financial instruments following the sovereign debt crisis, the so-called refugee crisis triggered the adoption of non-binding measures (Terpan and Saurugger, 2021). The authors argue that the more severe European sovereign debt crisis threatened the existence of the Eurozone, prompting member states to make binding commitments to save it. In a similar vein, Schoenefeld and Knodt (2021) observe a trend of increased hardening in the area of EU renewable energy policy (Schoenefeld and Knodt, 2021). They explain the finding with the evolving prominence of international climate change negotiations and the entrepreneurial actions of the Commission and some member states to fight climate change.

In sum, high diversity in policy preferences during EU negotiations could either lead to more or less soft law depending on the perceived (in)effectiveness of non-binding measures to sustain EU co-operation (Saurugger and Terpan, 2021; Schoenefeld and Knodt, 2021: 51). Therefore, we formulate contrasting hypotheses and discuss the conditions under which policy heterogeneity facilitates or impedes the adoption of soft law.

Theorising how preference heterogeneity affects soft law

Models of legislative-bureaucratic relations yield different expectations about the effect of preference heterogeneity on the stringency of EU legislation (Franchino, 2007; Thomson and Torenvlied, 2011). An important characteristic of legal stringency is whether policymakers delegate monitoring and enforcement powers to independent institutions. In the EU, the Commission represents the centralised monitoring agency, and the European Court of Justice adjudicates on cases of suspected non-compliance. While member state governments want control over policy making, co-operative solutions bear commitment problems. National authorities would deviate from supra-national agreements in the absence of strong enforcement mechanisms. Under these circumstances, soft-law instruments are adopted when member states’ can preserve their discretion in policy making or value harmonised implementation across European countries.

Heterogeneity in policy preferences affects both conditions by increasing the legislative and compliance costs of designing binding policy instruments. First, delegation of enforcement powers to independent agents requires the support of coalitions of politicians (Bendor and Meirowitz, 2004). Under increased preference heterogeneity, it is difficult to form coalitions that would delegate enforcement competences to the Commission (Franchino, 2007; Thomson and Torenvlied, 2011), impeding the adoption of binding legislation. EU policymakers could overcome policy gridlock by permitting diversity in legislative arrangements at the national level. Scholars show that granting more discretion or allowing member states to opt out are crucial for consensus on EU policies (Dimitrova and Steunenberg, 2000; Thomson and Torenvlied, 2011; Winzen and Schimmelfennig, 2016). By the same token, member states are likely to agree on soft-law measures when their heterogeneous preferences prevent consensus on binding legislation. Second, preference heterogeneity affects member states’ compliance costs by increasing the likelihood that governments deviate from EU acts after their adoption. Soft-law instruments grant discretion to member states by allowing governments to preserve existing policies or implement outcomes that diverge from the EU objectives, without being prosecuted by the Commission.

The argument also holds for soft law that aims to complement existing binding EU legislation. Preference heterogeneity among legislators often leads to ambiguous legislative outcomes (Huber and Shipan, 2002). As a result, supplementary explanatory measures are needed because national governments are uncertain about how the EU rules should be implemented. Therefore, preference heterogeneity is also expected to increase the share of supplementary soft measures that clarify binding legislation.

H1: Heterogeneity of member states’ preferences is associated with an increase in the adoption of non-binding or soft-law measures relative to hard-law instruments.

However, preference heterogeneity could also negatively affect the production of soft-law acts because member states strive to make credible commitments to abide by supra-national policies (Majone, 2001; Thomson and Torenvlied, 2011). Based on the commitment perspective, governments care about harmonisation of national policies across countries because unequal practices could lead to negative externalities (Majone, 1999) and defeat the purpose of common EU policy (Franchino, 2007). For example, regulatory differences could create burdens for cross-border commercial activity – generating dispute between countries that prioritise uniform implementation over domestic freedom. Furthermore, there is a risk that member states fail to honour agreed European objectives if they are given extensive discretion on their implementation (Franchino, 2007; Thomson and Torenvlied, 2011). Even when member states agree on common objectives, their policy preferences may change over time. Member states are therefore unlikely to adopt soft law to prevent commitment problems.

The argument also holds for complementary soft law. Under situations of political conflict and lack of enforcement instruments, member states will not comply with the agreed objectives and will instead implement their own preferred policies. Member states are also less likely to follow any complementary interpretation of the EU hard law if they disagree with the legislative measures in the first place. Based on the commitment perspective, member states and the Commission are likely to push for more binding EU legislation (relative to soft-law measures) to ensure compliance with the EU acts. The logic is similar to the ‘fiduciary principle’ discussed by Majone (2001) where member states strive to resolve credibility issues in the EU by delegating enforcement powers to the Commission.

H2: Heterogeneity in member states’ preferences is associated with a decrease in the adoption of soft-law measures relative to binding EU legislation.

The commitment perspective has found empirical support in studies of compliance with supra-national policies and EU institutional reforms. In particular, the Commission more actively monitors compliance in EU directives that were highly disputed during the Council negotiations (Zhelyazkova and Torenvlied, 2009). Furthermore, supra-national actors and member states anticipated the negative effects from accession-related heterogeneity in member states’ policy preferences and capacities. Consequently, the EU has often changed its institutional arrangements prior to accepting new member states to avoid legislative gridlock (Kelemen et al., 2014). For example, the EU boosted the rigidity of its outputs prior to the accession of the Central and Eastern Europe countries (Kelemen, 2011; van der Veen, 2014). Conversely, Thomson and Torenvlied (2011) do not find support for the commitment perspective in the delegation of powers to the Commission.

The mixed empirical evidence suggests that the threat of commitment problems differs across policy areas and issues (Thomson and Torenvlied, 2011). Policymakers presumably perceive higher commitment problems if they fear that member states will not comply with an EU norm. Therefore, the effect of preference heterogeneity on the adoption of soft measures is conditional on the (perceived) threat of non-compliance. For example, certain policies are more susceptible to non-compliance due to lack of administrative capacities or changes in the preference constellations within governing coalitions. Moreover, non-compliance also varies over time. Whereas issues of non-conformity prevail in the areas of environment and internal market, agriculture policies used to be most infringement-prone in the early 1990s.

In a similar vein, the choice of policy instruments depends on the perceived efficacy of the adopted EU norms (Terpan and Saurugger, 2021). In situations of political conflict, non-binding acts could facilitate co-operation if EU legislators expect that implementing authorities will comply in the absence of enforcement mechanisms. High levels of compliance signal that the member states are committed to supra-national agreements and have the necessary resources to implement them. If the Commission rarely resorts to infringement proceedings to remedy non-compliance, policymakers are likely to prefer soft law to alleviate ‘EU-decision traps’. Conversely, policy areas that are prone to non-compliance increase perceptions that member states will not meet the EU objectives, if implementation is not subject to monitoring and judicial scrutiny. While we still lack detailed accounts on member states’ compliance with soft law, we expect that it leads to weaker policy implementation than binding EU instruments. Under these circumstances, preference heterogeneity increases perceptions that enforcement mechanisms are necessary to compel national authorities to comply with the EU policies (Zhelyazkova and Torenvlied, 2009).

H3: Heterogeneity in member states’ preferences is associated with a decrease in the adoption of soft-law measures in cases of high levels of non-compliance with hard-law measures.

Research design

Measuring soft law in a longitudinal framework

In this study, we opted for an all-encompassing measure that covers various non-binding EU instruments, including recommendations, resolutions, guidelines, action plans, communications and declarations. Soft-law instruments fulfil different functions and vary in their relationship with binding legislation and how they influence member states’ policies (Senden, 2005). Some soft-law instruments provide interpretation of hard law. For example, the Commission frequently adopts notices and communications, indicating how Community law should be implemented by the member states. Soft law is also drafted in the form of recommendations or resolutions that steer national authorities to take certain actions.¹ More precisely, steering soft law covers instruments that aim to establish closer co-operation between the EU member states or even harmonisation of national policies. This category can be further divided into formal steering instruments (e.g. recommendations), and non-formal steering instruments, such as conclusions, declarations, guidelines, opinions and others (Senden, 2005). Information-based instruments, on the other hand, notify actors about specific aspects of EU law, domestic implementation or other policy actions foreseen by member states or EU institutions. For example, in 2010 the Commission adopted a communication document that ‘aims at providing a factual basis’ about the consequences associated with the possible introduction of Security Scanners as a measure for screening persons at EU airports. In the Online appendix, we outline the procedure for extracting different types of soft-law instruments.

We extracted all soft-law and hard-law instruments from the Eur-lex database that were adopted between 1967 and 2019. Our approach differs from the methodology of a recently published dataset on soft-law measures (Cappellina et al., 2022). First, our dataset covers a longer time period (1967–2019) with a larger number of policy areas (24). This allows us to assess both short-term and long-term changes in the share of soft law. Second, we constrain the analysis to soft-law acts produced by the Commission, the Council, the member states and the European Parliament, excluding EU agencies. Our theoretical mechanisms are relevant for the incentives of ‘traditional’ EU legislative and executive institutions to adopt soft-law and hard-law acts, and they do not necessarily hold for EU agencies. Furthermore, the rise of EU agencies has occurred in recent decades, which could bias our analysis for earlier periods. Due to space limitations, additional information about data collection is provided in the Online appendix. Our approach more closely resembles the dataset by Zhelyazkova et al. (2015), while complementing it with parliamentary instruments. Nevertheless, Cappellina et al. (2022) find that theirs and Zhelyazkova et al.'s (2015) datasets showed similar trends despite employing different methodologies (Cappellina et al., 2022: 11).

The dataset further distinguishes between EU policy areas. Unfortunately, there is no uniform approach to coding policy sectors, which has led to different operationalisations (Börzel, 2021; Cappellina et al., 2022; König and Luig, 2012). Our selection is based on the Commission portfolios, as the Commission participated in the adoption of most soft-law instruments (80%). Similar to previous research, the measurement is based on directory codes (König and Luig, 2012) and subject matter when the directory code was obscure or missing. The selection procedure resulted in 24 different policy areas (see the Online appendix). The policy sectors reflect a wide variety of issues discussed at the EU level, while being sufficiently general to link with the Eur-lex directory codes.

The dependent variable in this study is measured as the share of soft-law instruments adopted by the EU in a given year (t) relative to soft-law acts and secondary EU legislation regarding a specific policy area (p).

Share of soft la w_{t, p} = \frac{Soft {law}_{t, p}}{Soft la w_{t, p} + Directive s_{t, p} + Regulation s_{t, p} + Decision s_{t, p}}

The measure combines information-based and steering instruments. First, despite their different rationale, we derive similar theoretical expectations about the effect of preference heterogeneity on steering and information-based soft law. Second, a clear separation between information-based and steering measures has methodological limitations because many EU acts combine both steering and information purposes. For example, in a communication document to the European Parliament and the Council (an information-based instrument), the Commission outlined a strategy on clean and energy efficient vehicles in Europe, which seems to have more steering functions. Finally, we specifically assess the evolution of soft-law instruments relative to secondary EU legislative acts (e.g. directives, regulations and decisions adopted either by the Council alone or by the European Parliament and the Council). We exclude binding instruments that were unilaterally adopted by the Commission, as the member states do not directly participate in their adoption and they follow different logic from secondary legislation (Brandsma and Blom-Hansen, 2017). Nevertheless, we control for Commission legislation in the analysis. We also validated the findings using alternative measures for the dependent variable (e.g. number of soft-law measures or the share of steering instruments relative to regulations and directives). The results remain robust and are presented in the Online appendix.

Independent variables: Preference heterogeneity and non-compliance

The measure for heterogeneity in member states’ preferences is based on information from the Comparative Manifesto Project (Volkens et al., 2017) on party positions and Döring and Manow's ParlGov database on parties in government (Döring and Manow, 2019). We extracted policy-specific party positions regarding each policy sector based on parties’ election programmes in each EU member state with the exception of Malta.² We then computed the policy-specific government positions, weighted by parties’ share of seats in government (Crombez and Hix, 2015) and their voting power in the Council based on the Banzhaf index.³ Heterogeneity in the Council is measured as the standard deviation of member states’ preferences in a given year in relation to a specific policy area. Moreover, the analysis tests for both lagged and contemporaneous relationships between heterogeneity and the evolution of soft law. On the one hand, preference heterogeneity at the early stages of the Council negotiations could signal commitment problems in the future. On the other hand, heterogeneity could also hamper the adoption of hard-law measures during the final stages of negotiations. We account for these competing dynamics by including the contemporaneous and lagged (1 year) indicators for heterogeneity in separate models.

We obtained data about non-compliance with EU hard law from the Berlin Infringement Database (Börzel, 2021). The dataset covers all official cases opened by the Commission against the member states for violating EU law between 1979 and 2019. Policy-specific non-compliance is operationalised as the share of infringement cases in a specific policy area relative to all cases opened in a given year by the Commission.⁴ As European legislators are expected to consider past levels of non-compliance when assessing the severity of commitment problems, the variable is lagged (1 year) in the analysis. Furthermore, we interacted non-compliance with heterogeneity in member states’ policy preferences to test whether implementation failure reinforces the hardening of EU law in situations of the high-preference heterogeneity among member states.

Control variables

We control for the number of Commission executive acts in a given year in a policy area. These are tertiary binding instruments unilaterally adopted by the Commission. The function of these acts is to update and specify existing EU legislation based on explicit delegation of powers to the Commission.⁵ In particular, preference heterogeneity could curb the adoption of highly detailed EU rules (Huber and Shipan, 2002), requiring the Commission to further specify them.⁶ As a result, more Commission binding measures could replace the need for adopting soft law (especially information-based soft measures that are mostly adopted by the Commission). We extracted the Commission acts from Eur-lex based on the authorship of binding instruments (directives, regulations and decisions). The measure also includes Commission implementing and delegated acts that were introduced with the Lisbon treaty.

The analysis also controls for events that could interfere with the adoption of soft measures. First, treaty changes often induce significant policy activity. For example, the Amsterdam treaty transferred Justice and Home Affairs to the Community pillar allowing for harder measures in this policy area. Moreover, treaty changes can also increase soft-law measures if a new policy sector becomes subject to intergovernmental coordination at the EU level. Therefore, the analysis includes dummy variables indicating the first 2 years in which a new major treaty change came into force. Second, we also control for enlargement, which likely increases heterogeneity and non-compliance. Heterogeneity of preferences could be especially obstructive in the first year after enlargement, when member states are not used to working together.

Method of analysis

Our dataset encompasses 24 policy areas (N = 24) for the time period 1967 to 2019 (T = 53) and 1978 to 2019 (T = 40) for the non-compliance variable. The time-series-cross-section structure of the data could violate different assumptions in ordinary least squares analysis (Plümper et al., 2005). Specification tests reveal serial correlation between the observations in our data (i.e. errors correlate over time). In addition, long time series could be non-stationary, due to unit roots across the dependent and independent variables. Whereas relevant tests for panel data did not show unit roots in the overall dataset (Persyn and Westerlund, 2008), a more detailed analysis revealed that 16 out of the 24 individual policy areas have unit root in the series of the dependent variable. Regarding the independent variables, only the measure for Commission legislative acts has unit root across all series. Moreover, the tests do not unanimously show that the series are cointegrated.⁷ We address the issue of non-stationarity by regressing the differences of the non-stationary independent variables (i.e. Commission acts) on the first difference of the dependent variable. The model thus tests the effects of the independent variables on changes in the share of soft-law instruments.

In addition, we included a lagged dependent variable (in the level) to address serial correlation issues (Beck, 2001; Beck and Katz, 2011). This decision also has a theoretical rationale as member states resist hardening predominantly soft-law areas and are equally reluctant to soften existing hard laws (Saurugger and Terpan, 2021). Thus, past levels of ‘softness’ in a given policy area could influence policymakers’ considerations in the adoption of (non-) binding instruments. Due to the evidence of unit root and serial correlation, we employ the following model, where (p) denotes the policy areas and (t) the year of adoption:

\begin{aligned} Δ Soft law shar e_{p, t} = & α_{0} + α_{1} Soft law shar e_{p, t - 1} \\ + β_{1} Heterogeneit y_{p, t - 1} + β_{2} Noncomplianc e_{p, t - 1} \\ + β_{3} Δ Commission ac t_{p, t} + β_{5} Treatie s_{t} + β_{6} Enlargement s_{t} \\ + ε_{p, t} \end{aligned}

Panel time-series datasets suffer from additional methodological issues, including group-wise heteroscedasticity (i.e. the errors may have different variances across policy areas) and contemporaneous correlation of errors across policy areas due to common exogenous shocks. To account for heteroscedasticity and correlation in the panels, we apply Prais–Winsten models with panel-corrected standard errors (Beck and Katz, 1995, 1996). This method controls for outstanding serial correlation by modelling the serially correlated residuals as a first-order auto-regression or AR (1) process. Another common problem is the likelihood of omitted variables in panel data with correlated residuals within panels. To alleviate the problem of unobserved effects, we employ fixed-effects models by controlling for policy area average effects.

Descriptive analysis

Figure 1 compares the evolution of different types of soft-law acts (steering and information-based), secondary legislation, and EU executive measures. It shows that Commission executive acts dominate the EU policy landscape. This is in line with research on the bureaucratisation of EU politics, indicating the increasing amount of executive acts produced at the EU level (Brandsma and Blom-Hansen, 2017). Moreover, the vast majority of non-binding instruments aim to clarify and inform the member states about their legislative obligations (80% of all soft-law acts). Only 20% of the acts are classified as ‘steering’ instruments. Whereas information-based soft law has abruptly increased since 2000, the yearly output of steering instruments has remained relatively stable. The increase of soft instruments in 2000 is in line with the formal establishment of the Open Method of Coordination, which triggered the emergence of more flexible governance approaches in the EU.

Figure 1.

Volume of soft law and hard law over time.

Moreover, the EU has steadily increased the production of soft-law instruments relative to secondary legislation since 1994 (see Figure 2). This suggests that the increase in soft law was already triggered after the adoption of the Maastricht treaty with the creation of new EU policy areas requiring more intergovernmental coordination. However, the majority of these instruments have informative purpose and do not replace binding EU legislation. In addition, the share of soft-law acts seems to have stabilised after the adoption of the Lisbon treaty and even decreased in recent years. This drop in the share of soft law is due to a decrease in the number of communications in Competition policy and especially instruments referring to state-aid applications by member states and authorisation by the EU Commission.⁸

Figure 2.

Share of soft-law acts relative to secondary legislation over time.

Results

We now turn to the main question of this study focusing on the effect of member states’ preference heterogeneity on the relative adoption of soft-law instruments over time. Models 1 to 3 (Table 1) test H1 and H2 on the relationship between member states’ preference heterogeneity and the share of soft-law instruments. Model 4 (Table 1) tests whether the effect of heterogeneity depends on the degree of non-compliance.

Table 1.

Prais–Winsten regression with correlated panels and corrected standard errors.

	Model 1	Model 2	Model 3	Model 4
Share of soft law (lagged)	−0.274***	−0.243***	−0.240***	−0.243***
	(0.032)	(0.034)	(0.034)	(0.034)
Preference heterogeneity	0.120***	0.108**
	(0.041)	(0.044)
Preference heterogeneity (lagged)			0.076*	0.115**
			(0.044)	(0.048)
Non-compliance (lagged)		−0.127*	−0.128*	1.076**
		(0.067)	(0.067)	(0.490)
Heterogeneity*non-compliance (lagged)				−1.016**
				(0.406)
Commission legislation (diff.)		−0.021*	−0.021*	−0.023**
		(0.011)	(0.011)	(0.011)
SEA		−0.036	−0.036	−0.039
		(0.035)	(0.035)	(0.034)
Maastricht treaty		−0.006	−0.006	−0.008
		(0.032)	(0.032)	(0.032)
Amsterdam treaty		0.046	0.045	0.044
		(0.031)	(0.031)	(0.031)
Nice treaty		0.028	0.027	0.024
		(0.046)	(0.047)	(0.046)
Lisbon treaty		0.053*	0.053*	0.051*
		(0.030)	(0.030)	(0.030)
Southern enlargement		0.069	0.074	0.072
		(0.049)	(0.049)	(0.048)
EFTA enlargement		0.045	0.047	0.043
		(0.043)	(0.043)	(0.043)
CEE enlargement		−0.069	−0.066	−0.063
		(0.069)	(0.069)	(0.068)
Policy fixed effects	YES	YES	YES	YES
Constant	−0.081	−0.060	−0.022	−0.07
	(0.051)	(0.060)	(0.059)	(0.064)
Observations	1115	901	901	901
R²	0.153	0.160	0.156	0.160
AR(1) coefficient (rho)	−0.098	−0.130	−0.135	−0.133

Notes: Standard errors in parentheses; the auto-correlation parameter (bounded −1 to 1) is relatively small in all models due to controlling for the lagged dependent variable. ‘Northern enlargement’ was dropped due to collinearity; *p < 0.10, **p < 0.05, ***p < 0.01. AR: auto-regression; CEE: Central and Eastern Europe; EFTA: European free trade area; SEA: single European act.

In particular, Model 1 (Table 1) shows that a higher level of member states’ heterogeneity is associated with an increase in the share in soft-law instruments from year to year. When we include all control variables statistical significance diminishes, but the effect remains robust (Model 2 of Table 1). We thus find that a high level of member states’ heterogeneity in a given year increases the share of soft-law instruments relative to the previous year, in line with H1. The coefficient for heterogeneity is 0.108 in Model 2 (Table 1), indicating that a point increase in member states’ preference heterogeneity results in a roughly 11% increase in the adoption soft-law instruments relative to secondary EU legislation. Figure 3 illustrates the marginal effect of policy heterogeneity on the share of soft-law instruments in the EU. The effect is significant especially at high levels of member states’ policy disagreement.

Figure 3.

Linear change in the share of soft-law measures due to changes in member states’ preference heterogeneity.

Model 3 (Table 1) examines the lagged effect of preference heterogeneity on the relative change in the share of soft-law acts. The coefficient drops in significance (p < 0.10) but remains positive. In other words, higher levels of heterogeneity in a previous year do not diminish the production of soft law. The results in Models 1 to 3 (Table 1) do not show support for the commitment perspective in H2, that higher past heterogeneity in member states’ preferences signals commitment problems and triggers the adoption of binding and enforceable EU acts. Nevertheless, the significant negative interaction effect in Model 3 indicates that higher levels of past non-compliance moderate the relationship between heterogeneity and changes in the adoption of soft law. Figure 4 illustrates the interaction between non-compliance and preference heterogeneity. In cases of full compliance (non-compliance = 0), high levels of heterogeneity in the previous year are associated with a roughly 8% increase in the share soft-law measures from year to year. Conversely, when past levels of non-compliance are high (non-compliance = 0.6), the share of soft-law instruments decreases on average by 30% due to preference heterogeneity. In short, Model 4 lends support to H3, that the (lagged) effect of heterogeneity on the production of soft law is conditional on levels of past non-compliance.

Figure 4.

Linear change in the share of soft-law measures due to changes in member states’ preference heterogeneity at different levels of non-compliance.

The analysis also shows the relevance of some control variables. An increasing change in the Commission legislation relative to the previous year decreases the share of soft-law acts. Given that the vast majority of soft-law instruments are information-based and unilaterally adopted by the Commission, it is possible that Commission soft law substitutes binding executive measures that provide information about the application of executive hard measures. However, the effect is significant only under the 10% significance level in most models. Another result emerging from the analysis is that a higher proportion of soft measures in a previous year decreases the share of soft law in a current year. European legislators are thus reluctant to further soften policy areas with already high volume of non-binding instruments (Saurugger and Terpan, 2021).

Robustness analysis

Given that the analysis combines different types of soft-law and hard-law instruments, it is important to test the validity of the findings with alternative measures for the dependent variable. First, soft-law instruments are not the only viable solution to disagreements in the Council. Preference heterogeneity may also lead to legislative gridlock curbing any EU instruments’ adoption. Moreover, EU legislators are not faced with the binary choice between adopting secondary legislation or soft law. Preference heterogeneity may on average increase the volume of soft-law instruments within policy areas over time while there is simultaneously more hard law adopted as most soft law serves to complement hard law and not to replace it. We, therefore, replicated the analysis on the number of soft-law instruments adopted in a given year. In Table 2, we present three fixed-effects negative binomial models (the dependent variable is based on count data that is highly dispersed across panels), while additionally controlling for the volume of Council and European Parliament secondary legislation. The robustness analysis bolsters the finding that preference heterogeneity has a positive effect on soft-law acts produced in a given year (see Models 1–2 of Table 2). Similar to the main analysis in Table 1, the lagged effect of heterogeneity is negative and is conditional on past levels of non-compliance (see Model 3 of Table 2).

Table 2.

Fixed-effects negative binomial regression on the number of soft-law instruments adopted at a given year.

	Model 1	Model 2	Model 3
N soft-law instruments (lagged)	0.006***	0.005***	0.005***
	(0.0002)	(0.0003)	(0.0003)
Secondary legislation	0.464***	−0.250*	−0.272*
	(0.103)	(0.147)	(0.148)
Preference heterogeneity	1.380***	0.813***
	(0.207)	(0.272)
Preference heterogeneity (lagged)			0.961***
			(0.289)
Non-compliance (lagged)		1.044***	8.420***
		(0.352)	(2.878)
Preference heterogeneity *non-compliance (lagged)			−6.130**
			(2.396)
Commission legislation (diff.)		−0.182***	−0.183***
		(0.063)	(0.062)
SEA		−0.580***	−0.601***
		(0.168)	(0.168)
Maastricht treaty		−0.384**	−0.403***
		(0.154)	(0.154)
Amsterdam treaty		0.257**	0.253**
		(0.123)	(0.123)
Nice treaty		0.354**	0.322**
		(0.155)	(0.155)
Lisbon treaty		0.263**	0.272**
		(0.133)	(0.130)
Southern enlargement		−0.242	−0.238
		(0.212)	(0.212)
EFTA enlargement		−0.020	−0.053
		(0.198)	(0.200)
CEE enlargement		−0.099	−0.066
		(0.216)	(0.215)
Constant	−2.272***	−0.971***	−1.129***
	(0.252)	(0.335)	(0.354)
Observations	1248	946	946
Log likelihood	−3314.205	−2703.283	−2701.154

Notes: Standard errors in parentheses; *p < 0.10, **p < 0.05, ***p < 0.01. CEE: Central and Eastern Europe; EFTA: European free trade area; SEA: single European act.

Second, we recognise that there is substantial variation in the EU's soft-law instruments that is not captured by the main analysis. Arguably, the meaning, function and scope of soft instruments varies across policy areas. For example, the Council annually issues country-specific recommendations in the context of the European Semester. The production of soft law is more institutionalised in some policy areas and aggregate levels of preference heterogeneity may have limited impact on the yearly growth of some soft measures.

As discussed earlier, we make a rough distinction between information-based and steering soft law at the instrument level. Figure 5 illustrates the varying effect of heterogeneity on the share of these two types of instruments across policy areas. While the smaller number of observations decreases the significance of the effects, there are still some differences in the effect of heterogeneity. For example, in the areas: Employment, Research and Innovation and Enlargement, heterogeneity has a positive effect on the relative growth of information-based and steering soft-law instruments, whereas the effect is negative in the area Environment. In the former case, EU legislators often use soft-law measures because they are unable to agree on common policy (Schäfer, 2006). Conversely, environment policies experience low levels of compliance (Börzel and Buzogány, 2019) and require hard law. This finding is in line with case studies showing that EU legislators agreed to more binding targets in climate change policies in response to non-compliance (Schoenefeld and Knodt, 2021: 55). In other policy areas, the effect of heterogeneity varies across different types of instruments. For example, in the area of transport, heterogeneity is associated with a more information-based instruments, but less steering instruments. Thus, heterogeneity induces EU institutions to produce notices and communications to ameliorate disagreements between member states. Nevertheless, the overlapping confidence intervals show that the differences are not significant. Furthermore, the findings from the main analysis remain stable when employing different measures for preference heterogeneity and soft-law acts (see the Online appendix).

Figure 5.

Effect of member states’ preference heterogeneity on the share of information-based (a) and steering (b) soft-law instruments across policy areas.

Conclusion

In this study, we analysed the stringency of the EU's legal system across time and different policy areas. In particular, we investigated how the adoption of EU soft-law instruments has developed as a result of member states’ heterogeneity in policy preferences. It is widely acknowledged that EU integration has continued despite the increasing diversity in member states (Schimmelfennig et al., 2015). However, we know little whether the increased diversity in policy preferences has affected the EU's legal system through the increased adoption of non-binding instruments relative to hard-law acts.

The findings from this study generally contradict the idea of declining stringency of the EU's legal supra-national system because of diversity in policy preferences. Although we observe relatively more EU soft-law instruments over time, most soft measures have information and interpretive functions. Thus, they complement existing EU legislation by providing further guidance to the member states on EU policy implementation. This is especially the case in policy areas with high preference heterogeneity and low compliance. Such measures supplement hard law and strengthen the EU's legal system by helping national implementers achieve common objectives. Moreover, whereas heterogeneity increases the share and amount of soft-law acts in the short term, EU legislators are reluctant to relax implementation requirements in areas of high levels of non-compliance in the long run.

The results presented in this article have theoretical and empirical implications for research on member states’ flexibility and the nature of supra-national legal systems. The analysis adds support for the commitment perspective beyond the study of discretion of hard law, by testing its relevance for the production of soft-law acts. Research on member states’ discretion in the EU has found mixed evidence that member states limit their own discretion in the implementation of EU hard law under situations of conflict in the Council (Thomson and Torenvlied, 2011). However, our analysis suggests that the relevance of the commitment perspective could be conditional on the policy context and the perceived efficacy of norms to achieve commonly agreed objectives. Furthermore, preference heterogeneity has a positive effect on the share of soft-law acts relative to secondary EU legislation. In other words, EU legislators turn to non-binding instruments either to further clarify the objectives of existing measures or because member states cannot agree on hard law. This finding suggests that the adoption of soft-law acts helps the EU react swiftly to sudden controversial events (Slominski and Trauner, 2021). On the other hand, soft-law acts have no legal status and it is unclear whether they de-facto facilitate the interpretation and implementation of EU policies through normative standard setting.

The findings from this study open new avenues for research on flexible modes of EU governance. First, the intricate relationship between preference heterogeneity, non-compliance and soft law deserves further attention. While we expected that preference heterogeneity prompts EU legislators to prefer binding legislation in areas with high levels of non-compliance, this may differ between issue areas. For example, non-compliance and soft measures both increased in asylum policy in the aftermath of the so-called refugee crisis (Terpan and Saurugger, 2021; Trauner, 2016). Future research should also consider the sovereignty costs associated with certain policy areas as a condition explaining policymakers’ use of soft law despite the threat of non-compliance. Second, the analysis suggests that the Commission adopts most steering and complementary soft-law acts without involving the member states and the European Parliament. Further empirical work is needed to explain the Commission's involvement in the adoption of soft-law instruments. Third, previous research identified a continuum from softer to harder law depending on the content of EU policy instruments (Saurugger and Terpan, 2021). For example, a recent dataset on soft law acknowledges the multi-dimensional nature of non-binding acts by distinguishing between formalisation and enforcement (Cappellina et al., 2022: 8). Future research should test the validity of our findings applying different conceptualisations of soft law. Finally, scholars may derive different conclusions about the strength of the EU's legal system if the focus shifts to other forms of flexibility. For example, optional or ambiguous EU legislation grant extensive leeway to member states in the application of EU rules. In a similar vein, some countries may be exempt from binding instruments. Thus, future work should analyse the EU's legal strength in terms of other aspects of flexibility.

Supplemental Material

sj-pdf-1-eup-10.1177_14651165231162510 - Supplemental material for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Supplemental material, sj-pdf-1-eup-10.1177_14651165231162510 for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time by Asya Zhelyazkova, Markus Haverland and Rik Joosen in European Union Politics

Supplemental Material

sj-dta-2-eup-10.1177_14651165231162510 - Supplemental material for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Supplemental material, sj-dta-2-eup-10.1177_14651165231162510 for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time by Asya Zhelyazkova, Markus Haverland and Rik Joosen in European Union Politics

Supplemental Material

sj-dta-3-eup-10.1177_14651165231162510 - Supplemental material for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Supplemental material, sj-dta-3-eup-10.1177_14651165231162510 for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time by Asya Zhelyazkova, Markus Haverland and Rik Joosen in European Union Politics

Supplemental Material

sj-dta-4-eup-10.1177_14651165231162510 - Supplemental material for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Supplemental material, sj-dta-4-eup-10.1177_14651165231162510 for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time by Asya Zhelyazkova, Markus Haverland and Rik Joosen in European Union Politics

Supplemental Material

sj-do-5-eup-10.1177_14651165231162510 - Supplemental material for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time

Supplemental material, sj-do-5-eup-10.1177_14651165231162510 for Legislating softly: The effect of preference heterogeneity on the share of EU soft-law instruments over time by Asya Zhelyazkova, Markus Haverland and Rik Joosen in European Union Politics

Footnotes

Acknowledgements

We are grateful to Thijs Stegeman for providing research assistance in the preparation of the study. We also thank Anne Rasmussen, Jonathan Zeitlin, Michal Onderco, Pieter Tuytens, Dimiter Toshkov and Mads Dagnis Jensen, the four anonymous reviewers and the editor Gerald Schneider for their feedback on earlier versions of the project.

Author contributions

AZ contributed to the idea of the paper, theoretical framework, hypotheses, data collection, research design, statistical analysis and drafted the initial manuscript. MH contributed to revising the manuscript, the framing of literature and developments in specific issue areas. RJ contributed to revising the manuscript, data collection and measurement of key factors, and the comparison with other datasets on soft law.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Asya Zhelyazkova

Supplemental material

Supplemental material for this article is available online.

Notes

References

Abbott

Snidal

(2000). Hard and soft law in international governance. International Organization 54(3): 421–456.

Beck

(2001) Time-series-cross-section data: What have we learned in the past few years? Annual Review of Political Science 4(1): 271–293.

Beck

Katz

(1995) What to do (and not to do) with time-series cross-section data. American Political Science Review 89(3): 634–647.

Beck

Katz

(1996) Nuisance vs. Substance: Specifying and estimating time-series-cross-section models. Political Analysis 6: 1–36.

Beck

Katz

(2011) Modeling dynamics in time-series-cross-section political economy data. Annual Review of Political Science 14: 331–352.

Bendor

Meirowitz

(2004) Spatial models of delegation. American Political Science Review 98(2): 293–310.

Blauberger

(2009) Of “good” and “bad” subsidies: European state aid control through soft and hard law. West European Politics 32(4): 719–737.

Börzel

(2021) Why Noncompliance: The Politics of Law in the European Union. Ithaca and London: Cornell University Press.

Börzel

Buzogány

(2019) Compliance with EU environmental law. The iceberg is melting. Environmental Politics 28(2): 315–241.

10.

Brandsma

G-J

Blom-Hansen

(2017) Controlling the EU Executive? The Politics of Dele-gation in the European Union. Oxford: Oxford University Press.

11.

Buchanan

Tullock

(1962) The Calculus of Consent. Ann Arbor: University of Michigan Press.

12.

Cappellina

, Ausfelder A, Eick A, et al. (2022) Ever more soft law? A dataset to compare binding and non-binding EU law across policy areas and over time (2004–2019). European Union Politics 23(4): 741–757.

13.

Crombez

Hix

(2015) Legislative activity and gridlock in the European Union. British Journal of Political Science 45(3): 477–499.

14.

Dimitrova

Steunenberg

(2000) The search for convergence of national policies in the European Union: An impossible quest? European Union Politics 1(2): 201–226.

15.

Döring

Manow

(2019) Parliaments and governments database (ParlGov): Information on parties, elections and cabinets in modern democracies. Development version.

16.

Franchino

(2007) The Powers of the Union. Cambridge: Cambridge University Press.

17.

Hagemann

Hoyland

(2008) Parties in the council? Journal of European Public Policy 15(8): 1205–1221.

18.

Héritier

Lehmkuhl

(2011) New modes of governance and democratic accountability. Government and Opposition 46(1): 126–144.

19.

Huber

Shipan

(2002) Deliberate discretion? The Institutional Foundations of Bureaucratic Autonomy. Cambridge: Cambridge University Press.

20.

Idema

Kelemen

(2006) New modes of governance, the open method of co-ordination and other fashionable red herring. Perspectives on European Politics and Society 7(1): 108–123.

21.

Kelemen

(2011) Eurolegalism: The transformation of law and regulation in the European Union. Cambridge: Harvard University Press.

22.

Kelemen

Menon

Slapin

(2014) Wider and deeper? Enlargement and integration in the European Union. Journal of European Public Policy 21(5): 647–663. http://doi.org/10.1080/13501763.2014.897745.

23.

König

Luig

(2012) Party ideology and legislative agendas: Estimating contextual policy positions for the study of EU decision-making. European Union Politics 13(4): 604–625.

24.

Leuffen

Rittberger

Schimmelfennig

(2013) Differentiated Integration. Explaining variety in the European Union. Basingstoke: Palgrave Macmillan.

25.

Majone

(1999) The regulatory state and its legitimacy problems. West European Politics 22(1): 1–24.

26.

Majone

(2001) Two logics of delegation: Agency and fiduciary relation in EU governance. European Union Politics 2(1): 103–122.

27.

Majone

(2009) Europe as the would-be world power: The EU at fifty. Cambridge: Cambridge University Press.

28.

Persyn

Westerlund

(2008) Error-correction – based cointegration tests for panel data. The Stata Journal 8(2): 232–241.

29.

Plümper

Troeger

Manow

(2005) Panel data analysis in comparative politics: Linking method to theory. European Journal of Political Research 44(2): 327–354.

30.

Radaelli

(2003) The Open Method of Coordination: A New Governance Architecture for the European Union. Stockholm: Swedish Institute for European Policy Studies.

31.

Saurugger

Terpan

(2021) Normative transformations in the European Union : On hardening and softening law. West European Politics 44(1): 1–20.

32.

Schäfer

(2006) Resolving deadlock: Why international organisations introduce soft law. European Law Journal 12(2): 194–208.

33.

Scharpf

(2006) The joint-decision trap revisited. Journal of Common Market Studies 44(4): 845–864.

34.

Schimmelfennig

Leuffen

Rittberger

(2015) The European Union as a system of differentiated integration: Interdependence, politicization and differentiation. Journal of European Public Policy 22(6): 764–782.

35.

Schoenefeld

Knodt

(2021) Softening the surface but hardening the core? Governing renewable energy in the EU. West European Politics 44(1): 49–71.

36.

Senden

LAJ

(2004) Soft law in European Community law. Oxford: Hart publishing.

37.

Senden

LAJ

(2005) Soft law and its implications for institutional balance in the EC. Utrecht Law Review 1(2): 79.

38.

Slominski

Trauner

(2021) Reforming me softly – how soft law has changed EU return policy since the migration crisis. West European Politics 44(1): 93–113.

39.

Snyder

(1993) The effectiveness of European Community Law : Institutions, processes, tools and techniques. The Modern Law Review 56(1): 19–54.

40.

Terpan

(2015) Soft law in the European Union – the changing nature of EU law. European Law Journal 21(1): 68–96.

41.

Terpan

Saurugger

(2021) Soft and hard law in times of crisis: Budget monitoring, migration and cyber security. West European Politics 44(1): 21–48.

42.

Tholoniat

(2010) The career of the open method of coordination: Lessons from a “soft” EU instrument. West European Politics 33(1): 93–117.

43.

Thomson

Torenvlied

(2011) Information, commitment and consensus: A comparison of three perspectives on delegation in the European Union. British Journal of Political Science 41(1): 139–159.

44.

Trauner

(2016) Asylum policy: The EU’s “crises” and the looming policy regime failure. Journal of European Integration 38(3): 311–325.

45.

Trubek

(2005) Hard and soft law in the construction of social Europe: The role of the open method of co-ordination. European Law Journal 11(3): 343–364.

46.

Tsebelis

(1995) Decision making in political systems veto players in presidentialism, parliamentarism, multicameralism and multipartyism. British Journal of Political Science 25(3): 289–325.

47.

van der Veen

(2014) Enlargement and the anticipatory deepening of European integration. Journal of European Public Policy 21(5): 761–775.

48.

Volkens

, Volkens A, Lehmann P, et al. (2017) The Manifesto Data Collection. Manifesto Project (MRG/CMP/MARPOR). Version 2017b. Berlin: Wissenschaftszentrum Berlin fuer Sozialforschung (WZB). https://doi.org/10.25522/manifesto.mpds.2017a.

49.

Winzen

Schimmelfennig

(2016) Explaining differentiation in European Union treaties. European Union Politics 17(4): 616–637.

50.

Zhelyazkova

Torenvlied

(2009) The time-dependent effect of conflict in the council on delays in the transposition of EU directives. European Union Politics 10(1): 35–62.

51.

Zhelyazkova

Borzel

Schimmelfennig

, et al. (2015) Beyond uniform integration? Researching the effects of enlargement on the EU’s legal system. MAXCAP Working Paper Series 8.

Supplementary Material

Please find the following supplemental material available below.

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

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

0.00 MB

0.02 MB

0.24 MB

0.36 MB

1.47 MB

0.42 MB