2.3 The Third Wave: Theoretical and Methodological Differentiation

It is conceptual shortcomings of this type, in conjunction with a growing uneasiness with the relatively narrow theoretical and empirical focus of earlier research that gave rise to the third wave of EU implementation studies. It is marked by a plurality of theoretical and methodological approaches. What ties the different contributions together is a desire to broaden the theoretical and empirical perspective in order to get a fuller picture of the conditions that drive domestic implementation processes. A first new development was that qualitative researchers in particular began to discover the importance of domestic politics in determining the speed and correctness of legal adaptation to European directives. Along these lines, Treib (2003a

, 2004

) showed that party political preferences of governments may have a decisive impact on transposition outcomes. In particular, he takes issue with the behavioural assumptions underlying the mainstream misfit argument, notably that domestic governments will always try to defend their existing policy traditions. Instead, his empirical case studies show that governments may well accept wide-ranging deviations from the status quo if the direction of the required reforms is in line with their party political preferences. Conversely, government parties, who by definition hold veto power over transposition laws, may also drag their heels on the realisation of rather minor adaptations if these modifications go against the grain of their party political goals. Similarly, Mastenbroek and Kaeding (2005 ) argued that EU implementation research needs to go “beyond the goodness of fit” and instead focus more thoroughly on the preferences of crucial players in the domestic political arena. In an empirical case study on the transposition of two directives with significant misfit in the Netherlands, Mastenbroek and van Keulen (2006 : 38) thus showed that favourable government preferences “may work wonders in overcoming misfit”.

The second remarkable development in the third wave was the growing popularity of quantitative studies. Thus, more and more scholars have come to use the easily available data on the Commission’s infringement proceedings against member states to measure the amount of non-compliance with EU law (Mbaye 2001 ; Börzel 2001 , 2003b ; Tallberg 2002 ; Börzel et al. 2004 ; Sverdrup 2004 ; Beach 2005 ). A second type of quantitative studies based their analyses on the transposition measures that member states officially notify to the Commission. One strand of this literature used transposition rates (Lampinen and Uusikylä 1998 ), sometimes also in combination with infringement data (Giuliani 2003 ). These rates, which represent the share of transposed directives against all applicable directives at a certain period of time, are regularly reported in the Commission’s annual reports on monitoring the application of Community law.⁷ Another group of scholars used the transposition instruments reported in the Celex database (Borghetto et al. 2005 ), in domestic legal databases (Mastenbroek 2003 ), or in a combination of both (Berglund et al. 2005 ), to determine when a particular directive was incorporated into national law.

In theoretical terms, most of these quantitative contributions are informed by compliance approaches developed in the international relations literature.⁸ These approaches revolve around the dichotomy between voluntary and involuntary non-compliance. Scholars stressing problems of voluntary non-compliance argue that the willingness of states to comply with international commitments depends on the domestic costs and benefits of adaptation and on the costs of defiance. Where the costs outweigh the benefits, states will try to evade these burdens by non-compliance. Therefore, effective monitoring and sanctioning by international supervisory authorities is required to force unwilling states into compliance. Therefore, this approach is known as the enforcement approach (see e.g. Downs et al. 1996 ). The management approach, in contrast, considers lacking administrative and financial capabilities at the domestic level or ambiguous norms the main sources of non-compliance. International organisations thus need to assist their members, by organising training programmes and by providing such things as financial aid and the like (see e.g. Chayes and Handler Chayes 1993 ).

The theoretical insights of these statistical studies have hitherto been rather inconclusive. Some find support for the argument that structural properties of domestic polities, such as the number of veto players, have a significant impact on legal compliance (Lampinen and Uusikylä 1998 ; Giuliani 2003 ; Linos 2006 ), others do not (Mbaye 2001 ; Börzel et al. 2004 ; Borghetto et al. 2005 ). Some conclude that support for European integration is an important factor that facilitates compliance (Mbaye 2001 with regard to public support; Linos 2006 with regard to support by government parties), others do not (Lampinen and Uusikylä 1998 ), while still others find a statistically significant negative correlation between these two variables (Börzel et al. 2004 ). Some find a significant effect of indicators meant to measure the degree of changes required by the policies to be transposed (Borghetto et al. 2005 ; Linos 2006 ), others argue that misfit is a variable that cannot be operationalised adequately for the use in quantitative studies (Börzel et al. 2004 ), and so on. The only factor that seems to find support in most quantitative analyses so far is various aspects of administrative capabilities (Mbaye 2001 ; Börzel et al. 2004 ; Linos 2006 ; Borghetto et al. 2005 ; Berglund et al. 2005 ).

This presents us with an interesting paradox: While qualitative studies in the third wave of research have increasingly come to embrace the political character of transposition, the results of quantitative research seem to point back to the arguments of the pioneers of EU implementation research, who had highlighted the importance of efficient and well-co-ordinated administrations.

As a first step towards solving this puzzle, Bernard Steunenberg has recently begun to develop an interesting formal model of the transposition process, which encompasses both a politicised and a more bureaucratic mode. He conceptualises the process of incorporating a directive as a strategic game among several domestic players with distinct policy preferences (Steunenberg 2006 ; see also Dimitrova and Steunenberg 2000 for an earlier version of this argument). Crucial factors in this game are the type of instrument that needs to be adopted to transpose a directive and the preferences of the actors whose agreement is required for the adoption of this instrument. The type of legal instrument (legislation, decree etc.) decides on whether the actor constellation comprises the broad set of ministries, political parties and interest groups usually involved in enacting a piece of legislation (“horizontal coordination”) or whether the process is determined by a smaller set of actors, or even by a single ministry, as is sometimes the case if a ministerial decree is enough to transpose a directive (“hierarchical coordination”).

This model goes a long way towards a realistic conceptualisation of the variegated political constellations to be found in individual cases of transposition. However, it does not allow for theoretical expectations as to what typical processes of transposition in a given country and in a given policy sector might look like. Are there country or sector-specific patterns of the typical transposition instruments used? Does this result in typical patterns of rather politicised or rather bureaucratic transposition processes? And does a multi-actor constellation necessarily have to imply more problems than a single-actor constellation, or is it not possible that serious delays occur even if only one single ministry is in charge of transposition?

One answer to these questions was offered by a group of scholars who analysed the implementation of six directives from the field of EU social policy in the fifteen “old” member states prior to Eastern enlargement. The results of this study demonstrate that simple causal arguments, such as the misfit or veto player hypotheses, or the first-wave focus on administrative and procedural factors, do not hold across their cases. Instead, they argue that it is a complex web of administrative, institutional and actor-based factors that determines transposition outcomes (Falkner et al. 2002 , 2004 ; Falkner et al. 2005 : 277-316). Up to this point, the argument is not very different from Steunenberg’s model or from the heterogeneous results of quantitative studies.

However, the empirical results of this comparative research suggest that there are huge inter-country disparities, but strong similarities among members of different groups of countries, in the way they typically fulfil their EU-related duties.⁹ This results in a typology of three “worlds of compliance”. The three country clusters are characterised by the varying importance of a culture of compliance in member states’ political and administrative systems. In the world of law observance, which consists of the Nordic countries, the presence of a culture of respect for the rule of law among political and administrative actors usually ensures fast and correct transposition (Falkner et al. 2005 : 317-341; see also Leiber 2005 ). In the world of neglect, the absence of such a culture in both the political and administrative systems typically leads to long phases of bureaucratic inertia and rather apolitical transposition processes. Greece, France or Portugal conform to this pattern (Falkner et al. 2005 : 317-341; see also Hartlapp 2005 ). In the world of domestic politics, finally, administrations usually work dutifully, but since a culture of compliance is absent in the political realm, transposing EU law typically depends on the fit with the political preferences of government parties and other powerful players in the domestic arena. This is the largest cluster, involving countries like Germany, the Netherlands, Ireland and the UK (Falkner et al. 2005 : 317-341; see also Treib 2003a ,b , 2004 ).

This typology means that the controversial political interactions between political parties, powerful interest groups and other important political actors, which were described by scholars like Treib, or by Steunenberg’s mode of “horizontal coordination”, are only typical for a certain group of countries. In other member states, transposition is usually a rather apolitical, bureaucratic process, as in Steunenberg’s mode of “hierarchical coordination” or as suggested by some of the quantitative findings. In a third group of countries, the actor constellation may be similar to Steunenberg’s multi-actor co-ordination, but it does not give rise to deadlock and delays since all of these actors are culturally inclined to comply with the law no matter what the short-term disadvantages should be. This also suggests that many of the existing theoretical propositions are only “sometimes-true theories” (Falkner et al. 2007 ), which are relevant in certain countries, but not in others. Political variables such as party political preferences, interest group pressure and veto players should have a major impact in the countries belonging to the world of domestic politics. Administrative factors should be particularly important in the member states forming the world of neglect. In addition, collectively shared cultural dispositions towards respecting the law should be able to explain the raft of transposition processes in the countries belonging to the world of law observance.

It remains to be established empirically whether and to what extent these country patterns hold beyond the specific cases studied by Falkner, Treib, Hartlapp and Leiber, especially whether they may be identified in other policy areas as well. However, some results derived from quantitative studies seem to point in a similar direction. First, the identification of a Nordic “world of law observance” ties in with the findings of Sverdrup (2004 ), who looked at infringement proceedings and showed that the Nordic countries are the subject of much less infringement proceedings and give in much faster than other countries to the pressure from the Commission. Second, a recent paper by Börzel et al. (2004 ) identified a statistically significant correlation between the number of infringement proceedings against a given member state and the level of public support for the rule of law among the citizens of this country. Although Börzel, Hofmann and Sprungk point out that the available opinion-poll data on public support for the rule of law are problematic since they are quite old and only refer to twelve member states, this result nevertheless lends some support to the argument that countries such as Denmark, where citizens strongly support the rule of law, are less non-compliant than countries where citizens do not take the issue of observing the law very seriously, such as in Greece, France or Portugal. Using a different, more indirect indicator for the acceptance of the rule of law within European societies, Berglund et al. (2005 ) also demonstrated that there is a significant relationship between the rule of law and member states’ transposition record.

The considerable proliferation of studies dealing with transposition should not conceal the fact that third-wave research has also looked at the later stages of the implementation process. Compared to earlier research, however, studies covering not only transposition but also enforcement and application have become a very small minority in recent years. Among the few exceptions is a study by Versluis (2003 , 2004 ), whose explicit focus is on the enforcement of two directives from the field of chemical safety in four countries. She discovers major enforcement problems in some of her cases and argues that issue salience is crucial in determining whether domestic inspectors take a particular directive seriously or whether they ignore it (Versluis 2004 ).

The study by Falkner et al. (2005 ) also included not only transposition but also enforcement and application. Informed primarily by the insights of the top-down school in domestic implementation research, they present a set of institutional conditions that determine the effectiveness of domestic enforcement systems (“co-ordination and steering capacity”, “pressure capacity” and “availability of information”), and they distinguish between different types of enforcement for different types of norms (Falkner et al. 2005 : 33-40). Applied to the fifteen member states included in their study, they find that the shortcomings of the domestic systems of enforcing labour law in four countries (Greece, Ireland, Italy and Portugal) “are so significant in overall terms that we regard these countries as neglecting their duty to ensure not only legal transposition, but also a reasonable level of practical compliance” (Falkner et al. 2005 : 275). Based on a new study on implementation processes in four Central and Eastern Europe countries, Falkner and Treib (2006 ) have recently suggested that the combination of political contestation at the transposition stage and significant shortcomings at the stage of enforcement and application, which was already found to be characteristic of Ireland and Italy, might even call for a fourth world of compliance: the “world of dead letters”.