On time or with a delay? Transposition of EU directives in the Czech Republic in relation to subsidiarity check

Abstract The government plays first fiddle in European Union (EU) decision-making processes, but a role in EU governance is also performed by the national parliament, which has gained additional competence to submit reasoned opinions based on the subsidiarity principle and participate in the political dialogue with the European Commission. The authors trace the policy-shaping and policy-taking processes and explore the impact of parliamentary and government involvement in EU policy-making on belated and timely transposition of EU directives in the Czech Republic. This comparative analysis of six directives, of which three were transposed on time and the other three from the same policy areas not, shows that the connection between ex-ante and ex-post stages still seems weak, and thus, greater involvement by parliament in EU affairs does not alone affect the time of transposition. Instead, the capacity of the government, determined partly by the salience of the legislation and its characteristics, is the main explanation for the transposition delays.

(EWM) if they perceive it breaches the subsidiarity principle (cf. art. 12 Treaty on the European Union/TEU; Protocol no. 1 on the role of the national parliament in the European Union). 1 At this early stage, parliaments can also send a nonbinding resolution to the European Commission (EC) on legislative and nonlegislative initiatives within the political dialogue (PD) 2 to support or raise concerns about a particular issue at hand. Later in the policy-making process, at the national level, they can also participate in the transposition process.
While research on the implementation of EU law is widespread (Treib 2014), it has mainly focused on the impact of the government and administrative capacity or the legislative characteristics for compliance with European law, either on time or correctly. National parliaments have been rather neglected. It was supposed as a common prerequisite that transposition is accomplished mainly through a government decree or ministerial order (ECPRD 2019;König 2007;Steunenberg and Rhinard 2010) or that the parliament only rubber stamps what the government prepares. The reason is that the latter negotiates the final legislation in the Council of the EU (hereinafter the Council) and thus is better informed about the final rules and should be able to transpose them effectively without the involvement of other veto players.
However, the national parliament can still play some important role both in exante control by scrutinising the EU legislative draft or by mandating the government for the decisionmaking in the Council (Winzen 2022) and in ex-post implementation by amending the transposing bill (Dörrenbächer et al. 2015;Zbíral 2017;Zbíral and Grinc 2020). Nonetheless, across the EU, the form and effectiveness of such procedures differ considerably (cf. Auel et al. 2015;Winzen 2013;COSAC 2017, and Gattermann and Hefftler 2015 on parliamentary oversight institutions; Börzel 2021 and ECPRD 2019 on transposition of EU law). Moreover, as demonstrated by Sprungk (2011) and later by Finke (2019), such ex-ante participation in the policyshaping phase hardly influences the parliament's ex-post role in the policy-taking transposition phase if the policy-making process is long or if the parliament lacks sufficient personal and procedural sources.
Since the introduction of PD and EWM, the prerogative of information and insight into the negotiation process is shared with the parliaments to a greater extent. The channel of information was made more formal and structured, managed not only from the national level but also from the EU level. As a result, parliaments started to scrutinise EU legislation more systematically (cf. Fromage 2020) and strengthened other control mechanisms in EU affairs (cf. Gattermann et al. 2016;Miklin 2017;Sprungk 2016;Winzen 2022). It might thus be expected they are better informed and can perform transposition faster than previously assumed. 1 Defining the subsidiarity principle is not always straightforward. Even though it is defined by the ToL (art. 5(3) TEU) as any action within non-exclusive EU competences taken at the EU level that must have sufficient added value compared to the action taken at the national, regional or local level, the interpretation of this by national parliaments varies considerably and ranges from judging the legal basis of the proposed legal act, through the principle of proportionality and policy substance, to the level of delegation of quasilegislative and executive powers upon the European Commission (cf. Granat 2017;Kiiver 2012).
The connection between the EWM and PD, compared to other parliamentary oversight procedures, and the timely transposition of EU directives has not yet been empirically tested, though. 3 There are only normative assumptions that these EU oversight tools can positively affect transposition of EU law (Fromage 2020). This article, therefore, explores this question with a case study of negotiation and transposition in the Czech Republic of six EU directives in three policy areas adopted after the ToL came into force. Besides the role of the national parliament in both ex-ante and ex-post legislative stages, the role of the government will also be traced for comparison to discover whether the EWM/PD can resolve the hitherto executive-legislative imbalance. Thus, the research question is to what extent can the EWM/PD influence transposition, compared to other factors related to the roles of the national parliament and government in the whole EU policy-making process?
This study offers three main contributions to the current academic literature. Besides the investigation of the impact of the EWM/PD on transposition (and the related roles of parliament and government within these policy-making processes), the article analyses the situation in one of the Central and Eastern European (CEE) countries, which are generally less researched both in terms of scrutiny mechanisms and transposition of EU law compared to the Western and Southern Europe. Moreover, this study compares not only directives that were transposed with a delay but also those transposed on time. Previous studies have mainly focused on cases of delayed compliance when examining when and why it happened (e.g. König and Luetgert 2009). Comparing examples of timely (non) compliance can better indicate which factors influenced the delay in transposition and which might have not.
Nevertheless, despite this approach, this case study showed that the early involvement of the parliament in the policy-making through the EWM or PD seems not to have contributed principally to the timely transposition of EU law. This result implies that both of these EU scrutiny tools encountered similar obstacles known previously from the national oversight mechanisms, such as lengthy policy-making processes at both the European and national levels, between which the composition of the government or parliament can change, or insufficient parliamentary procedures that can link both ex-ante and ex-post stages. Moreover, both the negotiation and transposition of EU directives still depend mostly on the administrative capacity of the government rather than parliament. On the contrary, political factors, such as the party composition of the parliament and government and their positions vis-a-vis the EU, did not appear to play significant role in this study.
The analysis and results will be presented in the following steps: first, the literature on the role of national parliaments in EU politics and on the implementation of EU law will be discussed briefly to outline the research field. The specifics of the research design and case selection will be explained in the next two parts. Finally, the results will be interpreted. The concluding section summarises the article and presents suggestions for further research.

From scrutinisers to policy-shapers?
National parliaments can perform several roles and exercise several functions either at the national or at the European level, be it representation, deliberation/communication and legislation/policy-making (Cooper 2012), or scrutiny/control of the national government and cooperation with other (supranational and international) actors (Kinski 2021;Sprungk 2016). The multilevel governance of the EU has, however, presented some challenges for the accomplishment of these roles since the 1980s, after the first direct elections to the European Parliament (EP), and until the 2000s, no significant formal role was assigned to parliaments. Governments, which often represent the country at the supranational/international level and prepare transposition measures later at home (if necessary), naturally dominate in EU affairs over national legislative bodies.
Yet, over time parliaments have gradually tried to get involved, first by strengthening their national oversight institutions (Winzen 2012) and since the ToL also through the EU mechanisms (Miklin 2017). Much attention has been devoted to the institutional capacity of parliaments and their motivations to scrutinise EU affairs by mandating the government or discussing EU issues either in plenary sessions, EU committees or sectoral committees (Auel et al. 2015;Finke and Dannwolf 2013;Gattermann and Hefftler 2015;Winzen 2013), similar to the consequences of the introduction of the EWM/PD and other post-Lisbon (international) parliamentary roles on the functioning of national legislative bodies (Auel and Neuhold 2017;Cooper 2012;Gattermann et al. 2016;Högenauer and Neuhold 2015). In last decade, there were also some attempts to connect national scrutiny mechanisms to another role that parliaments can performtransposition of European law into national law (Finke and Dannwolf 2015;Sprungk 2011)and to question how the executive-legislative (im)balance might have been recalibrated by the new powers given to parliaments (Fromage 2020).
Within the implementation of EU law, parliaments act as one of the veto players, beside the government and its administrative bodies (Steunenberg 2007), subnational actors, such as regions and their government bodies (Paasch and Stecker 2021), and interest groups (Falkner et al. 2005), all of which may be able to constrain and delay smooth transposition. For parliaments to play an effective role in transposition therefore requires well-developed and strong ex-ante scrutiny (Finke 2019;Sprungk 2011), by which members of parliament (MPs) can get sufficient information on legislative drafts and the positions the government intends to defend in Brussels and possibly bring them in line with MPs' views. This can help later in (timely/correct) transposition (Finke 2019), but only if (1) the actors remain identical, (2) little time elapses during the whole (ex-ante and ex-post) EU policy-making process, and (3) the implementing actors are adequately involved in policy-shaping (Sprungk 2011). Such active participation in EU policy-making also depends on intensive cooperation between domestic and European actors, such as the lower chamber with the upper chamber or sectoral committees with the EU affairs committee (Sprungk 2016). However, early involvement by the parliament does not always have a positive impact on transposition, as it can also trigger unintended conflict between opposing interests at the ex-ante stage, which can later hinder proper transposition (Finke and Dannwolf 2015). In addition, the role of the parliament depends greatly on other independent factors that can influence the results of implementation, in particular the outcome of negotiations at the EU level in terms of the government position and voting, formulation of legislative acts (Zhelyazkova 2013) in specific policy areas and the later transposition measures (Haverland et al. 2011), which can take the form of government regulation, ministerial ordinance or statutory law.
While there are already some studies on the interlink between ex-ante and ex-post stages of EU policy-making processes, they have mainly focused on the role of the government or on national parliamentary scrutiny mechanisms. Empirical and systematic testing has not yet been carried out on the effect of the EWM/PD, which strengthened and equalised the national parliaments (and their chambers), but their use can be determined by factors that similarly affect national control mechanisms (cf. Hefftler 2015 vs. Winzen 2013;Auel et al. 2015;Sprungk 2016). The reason for the dearth of research may be the widespread sceptical view that the EWM cannot influence the final decision taken by the EU institutions (cf. de Wilde 2012;de Wilde and Raunio 2018;Raunio 2010). Nevertheless, few articles have already come to the conclusion that the EWM can be taken into account during the EU decisionmaking even though it does not reach the necessary threshold of the dissenting parliaments (e.g. Cooper 2019). Similarly, several studies also showed that the positions expressed by the parliament in reasoned opinions or PD resolutions are often in line with the position of the government on the particular legal act (cf. Borońska-Hryniewiecka and Grinc 2022; Cornell and Goldoni 2017), which might be positively reflected later in the transposition process.

Research design
To investigate this question, we traced the policy-making process (i.e. the mechanisms that function between the negotiation of the directive at the EU level and its transposition in the Czech Republic) starting with the writing of EC legislative drafts, their consultation by both chambers of the bicameral parliamentthe Chamber of Deputies (ChofD)/Poslanecká sněmovna as a lower chamber, and Senate/Senát as an upper chamber 4 in the Czech Republicthrough the deliberations of the Council, 5 to the national process for adoption of transposition measure(s), 6 and focused primarily on the variables specified below. 4 According to the Constitution of the Czech Republic (art. 10b), both parliamentary chambers are equal in authority over EU affairs, despite some procedural nuances specified in their rules of procedures. Moreover, the EWM/PD itself provides for equal rights for all parliamentary chambers in EU member states (Granat 2017). Therefore, the study focused on both chambers, even though the Senate is weaker in national politics because its decisions on legislation can be overturned by the lower chamber and it does not have a direct link to the national or the regional government(s). 5 The EP was not included in the analysis even though it participated in adopting all the directives in our study. The reason is that the members of the EP are not formally connected to the national government or parliament on a regular basis in the Czech Republic, especially as compared to the government representatives in the Council. 6 The transposition process in the Czech Republic follows the "ordinary" national legislative procedure for adopting legal acts. That means that the bill is prepared by a ministry, and before it is adopted by the government, it has to go first through the inter-ministerial consultation procedure and then through a review by the Legislative Council of the government at the expert level. After that, it is first debated and voted on in the lower chamber in three readings, within which sectoral committees can prepare amendment(s), and then, the Senate can either adopt, not act at all, propose amendment(s), or reject what the ChofD previously adopted. However, for the implementation of EU legislation (be it regulation, directive Since the subject of the article is the impact of the actions of the parliament and government on the amount of time taken for transposition when both can be involved in the decision-shaping and decision-taking stages, we analysed, mainly qualitatively, factors related to both powers. Specifically, and during the policymaking/implementation process (cf. Falkner et al. 2005, 6;Treib 2014, 6), the analysis focused on six factors: (1) The ex-ante parliamentary opinion of both chambers (and their committees) on each EU legislative proposal sent in the form of a reasoned opinion or as part of the PD (if available in the IPEX/online database for inter-parliamentary exchange or in databases of the Czech Parliament, and crosschecked also in the EC's reports both on relations with national parliaments and on subsidiarity and proportionality). (2) The position of the government as stated in the framework document indicating the salience of the act for the government (obtained either by personal request or by a manual search on Google). 7 (3) The vote of the government in the Council on the final version of the directive (tracked by the VoteWatch portalcurrently available in Hix et al. 2022), 8 predicting the final national support for the particular act. Later, we considered also (4) the ex-post involvement of the parliamentary committees of both chambers, 9 (5) the plenary debates on formulating and adopting the transposition measures, and (6) the number and types of measures for transposition 10 adopted after the directive came into effect (specified either in the EUR-Lex database or in the national table of concordance), indicating how complex the transposition was (Borghetto et al. 2006;Haverland et al. 2011). In addition to tracking both the EU and national procedures via document analysis, semi-structured interviews were carried out (in 2019) with two officials from the analytical department on EU affairs of the Senate, and one questionnaire was sent to the equivalent department in the ChofD, to triangulate the information on the ex-ante and ex-post stages. 11 The article examines the transposition time rather than its correctness, which is harder to measure. Therefore, our main interests are when the transposition process was completed, whether it complied with the transposition deadline required by the directive, and whether the infringement procedure (according to Art. 258-260 of the Treaty on the functioning of the EU/TFEU) was launched. A real-time analysis of when the transposition was completed is quite a challenging task because of insufficient data from the EC and EUR-Lex, which has been recognised in previous or other acts), there are specific rules setting the requirements for the preparation of the transposition measures, the monitoring of their implementation, notification to the EC and, among others, the timing in relation to the transposition deadline specifying when the bill should be sent to the consultation procedure (11 months before the transposition deadline in case of the status as the main transposition measures, and four months before for the government regulation) and to the government (nine months and two months, respectively) in order to ensure correct and timely compliance (Methodological guidelines for fulfilling legislative obligations arising from the membership of the Czech Republic in the European Union 2018). 7 The proposal of the EU act can be given either national priority (high), important priority (intermediate) or be followed priority (the low). studies (König and Luetgert 2009;Zhelyazkova et al. 2017, 222). On the one hand, the EC does not have the capacity to follow all cases of noncompliance and therefore chooses strategically to follow only those for which it has a good chance of success in litigation (cf. König and Mäder 2014). On the other hand, in the Eur-lex database, it is not always clear which transposition measures are the first and final acts that started and completed the whole implementation, respectively. Therefore, we decided to combine the transposition instruments adopted between the approval of the final version of the directive and the transposition deadline as identified in EUR-Lex and/or the national transposition tables, 12 together with the infringement procedures launched by the EC against the Czech Republic according to the Reports of the Government representative for the Czech cases before the Court of Justice of the EU (CJEU). That means that if no infringement procedure was started against the transposition of a particular legal act after the transposition deadline, we consider it as transposition on time.
Besides these formal procedural milestones, the party composition of the government and parliament in terms of majority/minority versus opposition and their positions on the EU were traced during the ex-ante and ex-post stages because they might have had an impact on the EWM/PD (Auel et al. 2015) or on the transposition (see Dörrenbächer et al. 2015 for relations between executive and legislative bodies; Börzel 2021 for Euroscepticism). However, such effects do not always have the expected influence. For instance, higher (party/public) Euroscepticism does not necessarily lead to more opinions expressed within the EWM/PD framework (Borońska-Hryniewiecka and Grinc 2022) or to more noncompliance (Toshkov 2019). Rather, the opposite can happen. In this article, the level of party Euroscepticism was measured according to the results of the Chapel Hill expert survey (CHES) for 2010, 2014 and 2019, between which the ex-ante and ex-post stages of policy-making of the analysed EU directives took place. We ranked the positions of the Czech political parties on the EU according to the mean values of the "EU_position" variable, defined by the CHES codebook as "overall orientation of the party leadership toward European integration" on a 7-point scale.

Case selection
The Czech Republic has been chosen as a case study for several objective reasons, besides the familiarity of the case to the authors, and ease of access to national sources. Firstly, the two chambers of the Czech Parliament belong to one of the most active national parliaments sending its PDs or reasoned opinions to the EU (cf. Borońska-Hryniewiecka and Grinc 2022; European Commission 2010-2020a; European Parliament 2017, 21-24;Grinc 2015;Hrabálek and Strelkov 2015, 502-503), 13 despite its relatively moderate oversight institutions for EU affairs (cf. Auel et al. 2015;Karlas 2011;Winzen 2013Winzen , 2022. At the same time, the Czech Republic, 12 The transposition tables were downloaded via Google or sent by request from the ministry responsible for the transposition. 13 The involvement of both chambers of the Czech Parliament in the EU scrutiny processes in comparison to the other national parliaments of the current 27 EU member states between 2010 and 2020 is demonstrated in the online appendix (part 3). together with Poland, represents an exceptional case among the "new" member states in terms of (non)compliance because both countries struggle with a high number of infringement procedures compared to other countries from the eastern enlargement (cf. European Commission 2010-2020b; Börzel and Sedelmeier 2017;Börzel 2021). For a long time, the Czech Republic has found itself around (and in last six years, above) the EU average for infringement cases (including late transposition) (cf. European Commission 2010-2020b; Single Market Scoreboard 2020). It holds a middle to lower position also for other characteristics, such as the functioning of the (government) coordination mechanism for EU affairs (Jensen 2014;Panke 2010;Perarnaud 2022) and bureaucratic capacity (Börzel 2021), including parliamentary administration (Högenauer and Neuhold 2015).
Thanks to its middling status, the Czech Republic might reflect average characteristics for compliance among the EU member states. However, due to differences between countries in the length of EU membership, the Czech Republic can instead represent the smaller group, in particular that of the CEE/ Visegrád countries (cf. also Borońska-Hryniewiecka and Grinc 2022), whose oversight institutions are relatively strong (Karlas 2011) and whose compliance performance is overall better than in Western/Southern Europe (Zhelyazkova et al. 2017).
But, in general, the Czech Republic has rarely been a subject of implementation research, except for studies of all 27 member states (such as Toshkov 2012 andZhelyazkova et al. 2017), similar to the rest of the CEE. There have also been some (separate) studies comparing the Czech Republic with other countries on the EWM/ PD (Borońska-Hryniewiecka and Grinc 2022; Granat 2017), other tools of parliamentary scrutiny (Finke and Dannwolf 2013), transposition of directives (Zbíral 2017;Zbíral and Grinc 2020), and compliance with EU law (Falkner 2010), and a few individual analyses on implementation of specific policies (such as Baun and Marek 2013). Again, however, nobody has linked both stages of policy-making and tried to empirically investigate an impact of the ex-ante stage with EWM/PD on the ex-post transposition of several EU legislative acts.
Three policy areaspublic health (health risks), transport and economic policy (financial intervention), from which six directives were thoroughly examinedwere selected for having the highest number of infringement procedures launched against the Czech Republic. This is tracked by annual Reports on activities of the Government representative for the representation of the Czech Republic before the CJEU.
The research focused on directives because by definition they set a common EUwide goal, but the forms and measures for their implementation are, to a certain extent, left in the hands of national policymakers. Thus, compared to regulations, there is some room for manoeuvre that needs to be filled and takes some time. Moreover, the selected directives were to have been adopted after the entry into force of the ToL to ensure that the EWM was already in place. They also were to be negotiated through the ordinary legislative procedure (OLP, according to art. 289 and 294 TFEU), which allows a dissenting member state to vote against the final act (because qualified majority voting is applied), which might later cause a delay in transposition (Thomson 2010). The OLP also gives national parliaments more power to potentially trigger not only a yellow but also an orange card within the EWM, which would require the resolution of the case not only by the EC but also by both the Council and the EP, and thus might be more attractive for parliaments than a special legislative procedure.
We compared three directives that missed their transposition deadlines with three directives that the Czech legislature managed to transpose on time, i.e. no infringement procedure was initiated. Each policy area is thus represented by two directives, one for delayed and one for timely transposition. The delayed directives were selected according to the gravity of the infringement procedure in terms of the years of duration and/or the highest stage the infringement procedure reached, according to the Government reports on the Czech cases before the CJEU. Our main interest was to focus on serious infringement cases and look on the effect of the EWM/PD on them. The timely transposed directives were selected in the same policy areas according to the dates of their adoption at the EU level, which should have been in proximity to dates of the adoption of the directives that were not transposed on time. Such conditions shall ensure that the composition of the government and/or parliament in either of the two stages is the same or similar, in order to mitigate the effect of parliamentary elections or reshuffling of the government, which can change their preferences in regard to EU legislation and cause discontinuity as well as accountability and legitimacy problems in EU policymaking process (König 2007).
However, the selection criteria were not helpful in our analysis for two reasons. Firstly, the EU policy-making process, from the EC draft through its adoption at the European level to its implementation at the national level, is quite lengthy and can thus last longer than one four-year parliamentary term (applicable in the Czech Republic). Secondly, the unstable domestic political landscape, which is characteristic of the post-communist states like the Czech Republic in particular, can influence the process as well. Therefore, we were able to maintain similar conditions for the government and parliamentary composition between the cases, but not for the individual cases. As a result, we included these changes in the analysis since the shifts in government and the composition of parliament can be two important factors determining the nature of both the earlier parliamentary involvement (Auel et al. 2015;Gattermann and Hefftler 2015) and the transposition outcome (Dörrenbächer et al. 2015;Falkner et al. 2005).
The list of selected directives and their main characteristics is presented in Table 1. The formulation of the directives is considered an initial premise to be later confronted with the results of the analysis. As shown by previous studies, the novelty/amendment of the directive, its complexity (Kaeding 2008;Zhelyazkova 2013) and/or length of time for transposition (Borghetto et al. 2006;Haverland et al. 2011) can impose potential limits for smooth/timely transposition and therefore it's good to keep these characteristics in mind. 14

Discussion of the results
To answer the questions as to whether and how the involvement of the parliament in the EWM and/or PD can lead to timely transposition, compared to the government, we analysed the procedural, political and legislative indicators, the first two of which are summarised in Table 2 (the third set of factors is in Table 1 above). The complete qualitative analysis can be found in the online supplementary materials (part 1). Only systematic analysis of administrative and informal procedures and practices was left out as it would require more in-depth data.
After the thorough examination, however, the results show that the original assumption that the early involvement of the national parliament through the EWM/PD facilitates timely transposition was not confirmed. The participation of both chambers in the early scrutiny of the draft of the directive, through the submission of their resolutions mainly within the PD (a reasoned opinion for the subsidiarity check was successfully raised only for the tobacco directive), seems not to be sufficient for timely transposition. The directives that were transposed on time gained much less attention from both parliamentary chambers during their negotiation at the EU level than those that missed their deadlines.
One problem in discontinuity is that the ex-ante scrutiny of EU legislative drafts is primarily in the hands of specialised EU parliamentary committees (EAC) in both Czech chambers, which are not, however, part of the later "ordinary" national legislative procedure that applies to the transposition process in the Czech Republic, Novelty is easy to trace as it is often indicated by, or it is obvious from, the legal act itself. But for complexity, several methodologies existsee Franchino (2007) for measuring discretion and related constraints provided by the directive to the member states, and also Kaeding (2008, 597, 600) and Steunenberg and Rhinard (2010, 501-502) for broader discussion. We chose to count the number of recitals as a widely used indicator of policy complexity that might reflect how difficult the transposition will be  Finally, there is no variability, but the variable was not preselected and thus the results were detected later in the analysis. Therefore, we decided to preserve it. However, it was adopted within the eight-week ('subsidiarity') period.

18
In its resolution, the ChofD committee for EU affairs explicitly called for the reasoned opinion but its resolution was adopted after the required eight-week period. 19 The Senate's EU committee proposed a reasoned opinion but only the opinion of the full Senate's plenary is valid and, in the latter case, was not achieved in time.  Accordingly, a government led by the soft Eurosceptic ODS was put together with pro-European TOP09 and VV (overall mean position from the CHES around 5). The government of CSSD, ANO and KDU-CSL was pro-European (overall mean position from the CHES: 5.7). The government of ANO and CSSD was also pro-European (overall mean position: 5.4) but supported by hard Eurosceptic KSCM (all three together ranked rather neutral with an overall mean position 4.4). In the first two governments, the chairmen of the ChofD's EU committee were from the government party; however, at the time of the less pro-EU government, the chairman was from soft Eurosceptic party ODS, and at the time of the more pro-EU government, the chairman was from the small pro-EU government party KDU-CSL. The composition of the EAC always reflected the composition of the parliament (i.e. ODS (6), CSSD (5), TOP09 (3), KSCM (2) and VV (2); ANO (9), CSSD (6), KSCM (3), TOP09 (2), KDU-CSL (2), ODS (2) and Dawn (1)). In Senate, the period 2010-2012 was dominated by opposition parties (however, the EAC was led by the governing ODS with a pro-EU majority); the period 2013-2017 was dominated by coalition parties (the EAC was led by the governing KDU-CSL with a pro-EU majority), and the period 2018-2020 was dominated by opposition parties.
similarly to several other member states (cf. ECPRD 2019). The rules of procedure for both chambers only determine their early involvement in EU affairs, including the EU policy-making process and ex-ante scrutiny of the government (Rules of procedure of the Chamber of Deputies, 1995, par. 109; the Standing rules of the Senate, par. 119). In the Senate, the EAC prepares the draft resolution and can consult relevant sectoral committees, although in practice this happens less than fifty per cent of the time (cf. Grinc 2015, 251). The final resolution must be adopted by the entire Senate with a debate in the plenary session. In the ChofD, the beginning of the procedure is the same, but the debate in the plenary is not necessary and is held only when a legislative file is politically significant (cf. Hrabálek and Strelkov 2015), such as the draft of the directive on the recovery of crisis in our analysis. The sectoral committees, in general, are rarely involved in the EWM/PD (as compared to the normal legislative procedure), and when they were involved in our cases (the directives on tobacco, railways, and recovery of crisis), the transposition was delayed. Neither the limited personal continuity between both stages of EU policy-making, which occurred in the case of delayed transposition of tobacco and railway directives, helped to avoid delay in transposition. Therefore, there is almost no involvement of the implementing actors in policy-shaping, which contradicts another necessary condition for the effective participation of the parliament in EU policy-making (Sprungk 2011). The representatives of the analytical department for EU affairs in the Senate who we interviewed confirmed an effort to more regularly stress the ex-ante stage during the ex-post stage in the future.
The insufficient role of the parliament throughout the policy-making process may be also connected with, and indicated by, the characteristics of the legislation. In our cases, the directives for which the parliament was not involved in the scrutiny (namely on electromagnetic fields, registration of vehicles, and debt instruments) were mostly revisions of the previous directives, were technical and not complex (in terms of the number of recitals). Such characteristics can, by themselves, facilitate timely transposition (as Borghetto et al. 2006 andKaeding 2008 found previously). As for the transposition deadline, it was sometimes helpful, sometimes not. In two of the timely transposed cases (the directives on electromagnetic fields and registration of vehicles), the transposition deadline was sufficiently long (around three years), whereas in the case of the railway directive, 31 months was not enough to avoid the delay in transposition.
Regarding the number and type of transposition instruments, the timely transposed directives required fewer transposition measures than the delayed transposed directives, which were instead converted into national law by new statutory acts, amending acts, and ministry ordinances (as predicted and confirmed by König and Luetgert 2009, for instance). Only the directive on the recovery of crisis was exceptional because it was transposed after the transposition deadline but with only a few transposition measures (new ones, though). Again, the number and type of transposition measures can be related to the characteristics of the directive, and thus, all these factors can affect the time of transposition.
Still, the parliament seemed not to be the main actor for the time taken for transposition. Instead, the time of transposition was determined by the capacity of the government to propose the transposition measures well in advance of the transposition deadlines, according to the national transposition rules, which Sprungk (2013) also concluded is an important factor for timely transposition.
Positive votes for the final versions of the directives by the government in the Council, which occurred in all cases, did not indicate how well the later transposition would proceed, which is again in line with previous findings (e.g. Toshkov 2011, 11). The degree of importance assigned to the EU legislative file by the government in negotiations was better associated with late transposition. In our analysis, the proposals for directives designated of high (national) priority in the early stage (specifically, the directives on railways and on the recovery of crisis) were more complicated and experienced delayed transposition. However, the tobacco directive, which was transposed after the deadline, was an "important issue" for the government, similarly to all three cases transposed on time.
Quite surprisingly and counter-intuitively, the composition of both parliamentary chambers and government between and within both stages of policy-making, affected by early/regular elections or by reshuffles in the government, and their positions toward the EU, were not found to be significant factors. Neither the majority pro-European government, which remained stable throughout the entire transposition period (as in the cases of delayed transposition of the tobacco directive and directive on the recovery of crisis), nor changes in the government due to elections or reshuffling during the transposition period (as in the cases of timely transposition of the directives on electromagnetic fields and debt instruments) seem to have affected the amount of time necessary for transposition. The effects of changes in government and parliament between ex-ante scrutiny and ex-post transposition are difficult to assess because for the analysed cases of both timely and late transposition neither the parliament nor the government remained entirely unchanged. We might assume that, in line with Sprungk's (2011) suggestion, when the occupants of policy-making positions change and a long time passes between the phases of EU policy-making, the connection between ex-ante scrutiny and ex-post transposition is weak. Nevertheless, these conditions are not always necessary because, as our case study showed, the positions of the parliamentary chambers toward the EU legislative draft often coincided with the position of the government, even though one parliamentary chamber (the Senate) was led by opposition parties. Such a consensual politics can later help with timely transposition irrespective of the changes in parliamentary and government composition. This did not, however, materialise in our cases.

Conclusions
The imbalance of power between governments and parliaments in EU affairs is a long-standing question that has attracted the interest of several researchers and practical reforms seeking to strengthen the parliamentary role. The ToL, which took one step in this direction by encouraging greater involvement of national parliaments in the decision-making stage, could affect another role parliaments perform in EU affairs, namely the transposition of EU law into national law, which can be another obstacle to the smooth running of EU affairs due to frequent delays. To date, however, both stages of EU policy-making were researched separately, without considering the connection between them.
Based on the case study of the Czech Republic in the negotiation and transposition of six EU directives from three policy areas, our study explored this problem and found that the empowerment of the parliament in policy-making through the EWM or PD seems to be negligible for the timely transposition of EU secondary law. In the cases where at least one parliamentary chamber was involved in the scrutiny of the EU legislative draft, delayed transposition later occurred, contrary to the initial expectation. The main reason for this appeared to be the distribution of tasks between the parliamentary committees on the EU, which almost exclusively participated in the ex-ante stage, and the sectoral committees, which rarely dealt with EU legislative drafts and instead tended to focus more on the ex-post stage. Even in those cases where the sectoral parliamentary committees were involved in both phases, the transposition was not completed on time. Hence, despite the high ambitions for the role of parliaments after the ToL, the legislature's position remains relatively marginal in the Czech Republic compared to the executive.
The government still seems to hold the reins of EU policies because it takes the final decision in the Council, and after that, it prepares the proposals of the transposition instruments, which may or may not be adopted by the parliament. On the other hand, the personal discontinuity caused by elections or reshuffling of the government during the long negotiation and transposition processes did not play a significant role. Rather, we found that other factors, such as the complexity of EU legislation, the duration of the transposition and the number and type of transposition measures can negatively affect the transposition process and lead to delay, as confirmed by previous studies. Another predictor of the nature of the transposition process was the importance attached to the draft directive by the government before the negotiation at the EU levelhigher priority indicates salience and can result in a difficult transposition. Often this is not due to the voting results in the Council, since consensual EU politics pushes representatives from member states to finally take (almost) all of them on board in the final legislative acts.
As a result, taking the whole process into consideration, the administrative capacity of the government, determined by its internal rules and procedures, can be a significant factor for timely transposition (cf. Börzel et al. 2010) as well as for linking the ex-ante and ex-post stages. Thus, the functioning of the national bureaucratic apparatus across the EU policy-making process requires further study, which is beyond the scope of this article. Similar attention should be devoted to the parliamentary administrative capacities for better connecting the ex-ante and ex-post stages, since political factors (including the party positions toward the EU) can be marginal in debates over technical policies, as EU rules so often are. Despite predictions that transposition processes are politicised in the CEE countries (Falkner and Treib 2008), our results, like those of some other recent studies (cf. Borońska-Hryniewiecka and Grinc 2022; Zbíral 2017), instead show that turning the EWM/PD (with the benefit of equal access of parliaments to information on EU legislative file early on) into an advantage for transposition would demand more expertise, similar to what has been previously found for national scrutiny mechanisms.
Still, these conclusions might be only preliminary as they are based on very limited data. While qualitative analysis has the benefit of thorough study of subjects and mechanisms working within them, it cannot replace the added value of quantitative research, which in future could include more and various cases (in terms of the complexity and novelty of the EU directives adopted not only by ordinary legislative procedure, but also by special legislative procedure, and transposed both on time and with a delay), as well as explanatory factors (such as the impact of crisis management, or the number and type of delegated and implementing acts linked to directives that are often criticised by national parliaments in reasoned opinions/PD), and search for correlations or causality between them. This article might be a stepping stone for such more extensive research.
Moreover, this proposed research design can be applied to other policy areas or other countries, especially those (1) which are similarly active in the EWM or PD, (2) whose parliamentary sectoral committees are more involved in EU affairs (through various parliamentary scrutiny instruments), and (3) which are more compliant with EU law than the Czech Republic. Finding a state fulfilling these criteria is not easy because the parliaments most active in ex-ante scrutiny are from southern Europe (e.g. Portugal and Italy), which are among the worst compliers (Börzel 2021, 30). The best example would be Sweden and its Riksdag, which produced, until recently, a high number of opinions and resolutions within the EWM/PD. Its sectoral committees are more involved in the scrutiny of EU affairs, and compliance with EU law is high as well. Also, Romania seems to be an appropriate candidate with an increased number of PDs/subsidiarity opinions (European Commission 2010-2020a), the involvement of sectoral committees in EU affairs (COSAC 2017;Gattermann et al. 2016) and an average-to-higher number of noncompliance cases (cf. European Commission 2010-2020b). In addition, there are several countries that could fulfil at least two of three proposed conditions and can thus be potential candidates for future researche.g. Denmark (cf. Högenauer and Neuhold 2015, 347-348;Kiiver 2012, 26;Sprungk 2016, 189), Germany (cf. Gattermann et al. 2016), Lithuania (Gärtner et al. 2011) or Luxembourg (cf. European Commission 2010-2020aCOSAC 2017;Börzel 2021, 30;Gattermann et al. 2016;Grinc 2015).
Considering the changing landscape of EU law, in terms of the decreasing number of legislative proposals (European Parliament 2017, 25-26), parliamentary dialogue with the EC (European Commission 2010-2020a), as well as noncompliance cases (cf. Börzel 2021), it is also necessary to carefully choose a timeframe for study when parliamentary engagement with EU affairs was higher (i.e. from 2010 to 2018, with the exception of 2015). In this way, we can try to discover the potential link between the ex-ante and ex-post stages of the EU policy-making process, including the EWM and PD, and see whether there might (have) be(en) any causal mechanism that can turn the national parliament into an effective actor in EU affairs at the European as well as national level.
Supplementary material. To view the supplementary material for this article, please visit https://doi.org/ 10.1017/S0143814X23000132 Data availability statement. This study does not employ statistical methods, and no replication materials are available.
Journal of Public Policy 573